Theme Section

Doing institutional analysisimplications for the study of

innovations

J Rogers HollingsworthUniversity of Wisconsin

ABSTRACTThe study of institutions and innovativeness is presently high on theagenda of the social sciences There is increasing concern with how asocietyrsquos innovativeness is associated with its international competitive-ness And as scholars study why the innovative styles of societies varythere has been increasing concern with how the institutional makeup of a society inuences its particular style of innovativeness Howeverbefore there can be signicant advance in the study of this problem it isimportant that we have a better understanding of what constitutes insti-tutional analysis Every social science discipline ndash with the exception ofpsychology ndash has at least one distinctive strategy for doing institutionalanalysis And it is because of the lack of consensus as to the appropriateboundaries and content of institutional analysis that we have limited ability to make theoretical advances in understanding how the institutionalmakeup of a society impacts on its innovativeness Recognizing that thisis a serious problem for the social sciences this article attempts to struc-ture the eld of institutional analysis and takes the rst steps in relatingit to the study of a societyrsquos style of innovativeness

KEYWORDS

Incremental and radical innovations institutional arrangements institu-tions organizations path-dependency social system of production

Review of International Political Economy 74 Winter 2000 595ndash644

Review of International Political EconomyISSN 0969-2290 printISSN 1466-4526 online copy 2000 Taylor amp Francis Ltd

httpwwwtandfcouk

INTRODUCTION

This article addresses two issues which are currently high on the agenda of the social sciences (l) why do societies vary in their style ofinnovativeness and (2) how should we go about doing institutionalanalysis While these are often treated as separate issues this articlemakes some effort to relate the two subject areas to one anotherFor some years the economics literature has argued that a countryrsquos

innovative capacity is linked to its international competitiveness (Landes1969 1998 Nelson 1993) Yet we are greatly lacking a theoretical under-standing of why countries vary in their innovative styles Why forexample do some advanced industrial countries time and time againmake radical breakthroughs in basic and applied science and developradically new products ndash and even new industries And why do othercountries rarely make any radical breakthroughs in basic and appliedscience or in product development but continuously make incrementaladvances in knowledge improve upon newly developed products andeven become the dominant producers in these market segmentsWhy the innovative styles of countries vary is a complex problem But

much of the variation in innovative styles across societies is due to theirinstitutional conguration Institutions may either constrain or facilitateinnovativeness (Hage and Hollingsworth 2000 Edquist 1997 Langloisand Robertson 1995) but at present we do not have a good under-standing of how the institutional makeup of a society is associated withits style of innovativeness This is due to the fact that the comprehen-sion of the institutional structures of societies is in a state of confusionHence a major argument of this article is that before we can understandhow the institutional conguration of a society inuences its style ofinnovativeness we must rst identify the various components of theinstitutional makeup of a society and understand how these componentsare related to each other The latter part of the article explains how theinstitutional makeup of a society is related to its style of innovativenessAt present it is difcult to relate institutional analysis to innovative-

ness because our theories of both areas are poorly developed Even sowe are now at a strategic moment because we can build on a rich bodyof literature on institutionalism and innovativeness in order to advanceour theory of each and to integrate the two eldsFor some years much of the literature on technological innovation has

emerged from a focus on the rm (Dosi 1988 Fransman 1994 Langloisand Robertson 1995 Whitley 2000) For example Alfred Chandlerrsquosgreat corpus of work (1962 1977 1990) has tended to emphasize howthe success of a rmrsquos technological innovativeness ndash both across coun-tries and over time ndash has been inuenced primarily by whether it hasthe right strategy and structure For Chandler rms which have had theright strategy and structure have ended up having the organizational

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

596

capabilities which permitted them to have the economies of scale andscope to develop cost advantages over their competitors (Teece 1993)Chandlerrsquos work has had a profound inuence on the management liter-ature of the past couple of decades Hence in the Chandlerian traditionmuch of that literature has suggested that the key to understanding thecompetitive advantage of rms is to identify their strategies and orga-nizational structuresOver time another literature has emerged which emphasizes the

importance of the institutional environment of rms for understandingwhy rms in some countries excel in some industries but not in othersand why the rms in a specic country may excel in a particular industryat one time but may eventually lose that advantage (Landes 1969 1998North 1981 1990) More recently Richard Nelson and his co-workershave been advancing this literature by working at the frontiers of inte-grating the literature on institutions rm strategy and technologicalinnovation (Nelson 1994 1995a 1995b 1996 Mowery and Nelson 1999Murmann 1998 Arora et al 1998)The Nelson school correctly assumes that since we do not presently

have an adequate theory on how institutions rms and technologies co-evolve we are not at a stage to test a set of formal hypotheses whichow from some well-dened model Hence the Nelson school have collec-tively been developing descriptive studies of how institutions rmcapabilities and technologies co-evolve so that particular societies andrms at specic moments in time excel in particular kinds of innova-tions The goal of this kind of work has been to develop by workinginductively a better understanding of the processes of how technolo-gies rms and institutions co-evolve across a number of industries andcountriesA variety of endowments in the institutional environment provide

economic actorsrms with initial advantages or disadvantages forparticular types of technological activity (Murmann 1988 ch 7) Butover time everything is dynamic and the larger global and institutionalenvironments the capabilities of rms and the performance of rms allco-evolve and feed back onto one another However institutional envi-ronments differ widely from one society to another and the successfulrms and organizations are those which can best adapt their activitiesto the institutional environment within which they are embedded Oncea number of rms in a particular industry are successful however theymay be able to engage in collective action to modify their institutionalenvironment in order to enhance their innovativeness and their techno-logical competitiveness The studies of the Nelson group have beenparticularly important in advancing the argument that the interactionamong actors and their institutional environment is a multi-facetedprocess and that successful actors over time must not simply respond

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

597

to the institutional environment in which they are embedded but mustmodify their environment in order to maintain competitive advantagesAs these comments suggest the work of Nelson and his colleagues is

very promising for understanding why societies vary in their capacityto excel in different industries However to extend and complementtheir work we need a greater understanding of and consensus on themeaning of institutions and the institutional environment of rms Hencethis article is written in the spirit of complementing the agenda of Nelsonand his colleagues One of its goals is to develop a map of what can becalled the terrain of institutional analysis It assumes that until we havea map of this terrain scholars working in the broad eld of institutionalanalysis will not comprehend how their work interrelatesInstitutional analysis is now rather high on the research agenda of

the social sciences However we need to be aware of the obstaclesconfronting us as we attempt to advance an agenda of institutionalanalysis There is no consensus as to what is meant by institutions orby institutional analysis These terms are very widely used but they areused with different conceptualizations and the scholars who use themshare little common ground Until scholars have some consensus aboutthe meaning of the concepts they use their potential to bring about effec-tive advancement of knowledge is somewhat limited Thus thewidespread interest in several academic disciplines in the concepts lsquoinsti-tutionsrsquo and lsquoinstitutional analysisrsquo may well promise more than it candeliver given the organizational and disciplinary fragmentation ofcontemporary universitiesThere are many different approaches to the study of institutions

(Nelson and Sampat 1998) there are the new and the old institution-alisms (Stinchcombe 1997 Hodgson 1998 Langlois 1986 1989) thereis historical institutionalism (Steinmo et al 1992 Zysman 1994

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

598

ORGANIZATIONALCAPABILITIES INNOVATIVENESS

INSTITUTIONS INSTITUTIONAL ARRANGEMENTSAND INSTITUTIONAL SECTORS

Figure 1 Institutional environments organizations and innovativeness

Immergut 1998 Katznelson 1998) and several of the social scienceshave their own distinctive approaches to the study of institutions (Halland Taylor 1996 Hodgson 1988 Eggertsson 1990 Finnemore 1996Scott 1994 Calvert 1995 Hechter and Kanazawa 1997)The following comments reect some of the confusion in utilizing the

concept lsquoinstitutionrsquo Nobel laureate Douglass North in his bookInstitutions Institutional Change and Economic Performance (1990 3) denesinstitutions as lsquorules of the game in a societyrsquo To North institutions areconstraints which shape human interaction and the way that societiesevolve through time On the other hand Andrew Schotter argues thatinstitutions lsquoare not rules of the gamersquo Rather institutions are thebehavior that follows from rules Briey he is concerned with whatactors do with rules but not with what the rules are (Schotter 1981155) Many other examples might be given to illustrate the heterogeneityof approach to institutions and institutional analysis Even if scholarswere to agree with North that rules and norms are institutions theywould not necessarily agree on what a rule is Shimanoff for examplehas identied more than 100 synonyms for the concept lsquorulersquo (Shimanoff1980 57 Ostrom 1986 5)Another critical issue in the institutional literature is the relationship

between institutions and organizations North (1990) following from hisdenition of institutions argues that institutions and organizations aredistinct entities Which organizations come into being and how theyevolve through time is inuenced by a societyrsquos rules and norms thatis by its lsquoinstitutionsrsquo On the other hand a number of recent organiza-tional sociologists see very little difference between institutions andorganizations For them rules and norms are institutions which unfoldin tandem with organizational structures and processes and changes inorganizational forms internalize and reect changes in the societyrsquos rulesand norms Using this perspective a whole subdiscipline within soci-ology called the lsquonew institutionalismrsquo in organizational analysis hasemerged The lsquoinstitutionalistrsquo perspective on organizations assumes thatthe kinds of organizations which actors create are dictated by the culturalnorms and rules in which they are embedded (Powell and DiMaggio1991 Zucker 1988 1991)The importance of this disagreement about institutions is obvious If

institutions are so critical for understanding our societies it is impor-tant that we come to some systematic consensus as to what institutionsare and how they inuence social actors and the organizations that theycreate If we cannot do so we risk talking past one another and losingthe opportunities for cumulative knowledge based on articulation ofwidely shared theoretical understandingWe need not only conceptual clarication as to what institutions are

but also greater consensus as to how to study institutions No scientic

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

599

eld can advance very far if the practitioners do not share a commonunderstanding of the key concepts used in their analysis (Ostrom 19864) But with our universities and academic associations so fragmentedinto different disciplines and into various subspecialties within disci-plines it is difcult to advance the theoretical agenda of institutionalanalysis within the academy Indeed the disciplinary fragmentation ofthe modern university is a major barrier to the theoretical advancementof the study of institutions and innovations as well as most other hybrid elds of research (Hollingsworth 1984 Hollingsworth andHollingsworth 2000) And it will be only as a result of effective commu-nication across diverse elds of knowledge that our study of institutionsand innovativeness will be effectively advanced

MULTIPLE LEVELS OF INSTITUTIONAL ANALYSISAND INSTITUTIONAL CHANGE

There are innumerable signs that we are living in a time of great insti-tutional change the demise of the Soviet empire the processes ofEuropean political and economic integration rapid transformation inparts of the global economy the disintegration of the family structurethe weakening of voluntary associations and the decline in politicalparticipation in a number of advanced capitalist societies the weakeningof welfare states European law is superseding national law and is evenchanging complete national legal systems The list could go on and onEven though scholars discuss institutional change at length their

ability to measure the rate of institutional change is very limited Andmore crucial than the limited ability of scholars to measure institutionalchange is their very limited understanding of how to build new insti-tutions One of the reasons for these shortcomings is that the socialsciences are decient in a theory of institutions The building of newinstitutions and redressing the decline of some of the most importantinstitutions of our societies are among the most important problems ofour time If we are to advance in the development of a theory of insti-tutions we need to work collaboratively across the social sciences andwe need to dene the parameters of institutional analysis

First level of analysis

This article attempts to make some modest contribution to outlining theparameters of institutional analysis At the outset we need to recognizethat when we engage in institutional analysis we must be sensitive to multiple levels of reality As suggested above most scholars whoengage in institutional analysis do not participate in any coordinatedactivity with each other and their activity is fragmented into a variety

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

600

of disciplines and subdisciplines To establish some coherence to the eldof institutional analysis we need a map of the eld so that those workingin one area can see where their research ts in relation to other areasand other practitioners on the mapTable 1 presents such a map with multiple levels at which institu-

tional analysis occurs Theoretically each of these areas on the map isinterrelated with each other level However the various areascompo-nents on the map are arranged in descending order of stability orpermanence Those components at higher levels of reality are morepermanent and durable while those at lower levels change more rapidlyWere scholars doing institutional analysis able to reach some consensus

about where their own work ts in relation to all other practitioners inthe eld there would be increased potential for all practitioners tocommunicate with each other By analogy once geneticists crystallo-graphers biochemists etc had a good understanding of how theirinvestigations were related to each area of molecular biology the eldquickly was able to make theoretical advances (Judson 1979)At the rst level there are the basic norms rules conventions habits and

values of a society These are the most fundamental properties of insti-tutions and are the most enduring and resistant to change Rules normsconventions etc are institutions but are only one component of whatconstitutes institutional analysis As Burns and Dietz (2001 forthcoming)point out most human activity is organized and regulated by normsand rules and systems of rules These concepts are extremely importantfor institutional analysis as they exert the greatest inuence on the natureof the components of institutional analysis at the next four levels which

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

601

Table 1 Components of institutional analysis

1 Institutions = norms rules conventions habits and values (see North1990 Burns and Flam 1987)

2 Institutional arrangements = markets states corporate hierarchies networksassociations communities (Hollingsworth and Lindberg 1985 Campbell et al 1991 Hollingsworth et al 1994 Hollingsworth and Boyer 1997)

3 Institutional sectors = nancial system system of education businesssystem system of research (Hollingsworth 1997)

4 Organizations (Powell and DiMaggio 1991)5 Outputs and performance = statutes administrative decisions the nature

quantity and the quality of industrial products (Hollingsworth 1991 1997)sectoral and societal performance (Hollingsworth and Streeck 1994Hollingsworth et al 1990 Hollingsworth and Hanneman 1982)

Note The ve components in this table are arranged in descending order of permanenceand stability That is norms conventions etc are more enduring and persistent than eachof the other components of institutional analysis Each component is interrelated withevery other component and changes in one are highly likely to have some effect in bringingabout change in each of the other components

are depicted in Table 1 In most forms of social analysis it is extremelyimportant that we understand the social and cognitive conditions thatlead to compliance or non-compliance with rules and the conditionswhich lead to changes in rulesThe approach to the study of institutions employed here argues that

norms rules habits conventions and values both reect and shape thepreferences of actors Norms rules habits conventions and values inu-ence who and what are included in different types of decision makinghow information is processed and structured what action is taken(Shepsle 1986 1989) It is through norms and rules that behavior isjudged to be democratic fair or egalitarianBurns and Flam (1987) point out that in any society there are multiple

rule systems Within a family there are rules for decision making oftenquite different from rules and norms for decision making for a professorin a classroom or for the customer in a bank Despite the heterogeneityof rule systems there are meta rules and norms which encompass lesserrule systems Otherwise there would be such contradictory rule systemsthat a society would be paralyzed The existence of meta rule systemspermits different rule systems to intersect with each other so that ambi-guities can be resolved The greater the pluralism and complexity of asociety the more ambiguity there is about meta rules and norms in asociety and of course all ambiguities never completely disappearOverall the degree to which separate rule systems are interlinked is anempirical problem There are different sectors groups and interestspursuing their own action logic but through higher-order meta princi-ples and rules there can be order consensus and coherence in a societyIt is through a set of meta rules that class and ethnic conict in societiesare contained (For elaboration see Burns and Flam 1987)In many respects our understanding about norms rules habits con-

ventions and values inuences our perspective on how societies are con-structed and how they change lsquoNew institutionalistsrsquo (Posner 1992Schotter 1981 Williamson 1975 1985) tend to assume that at one timethere was a state of nature and that there was a movement from individ-uals to institutions ndash an approach often called methodological individual-ism (Popper 1961 Hodgson 1998 1999) And of course there areinnumerable instances which methodological individualists cite to demon-strate that individuals create new rules of behavior For example it is pos-sible for actors to change the rules of driving so that instead of drivingon the left side of the road drivers adopt a new rule and drive on the rightThis article however tends to equate social habits and institutions

As Hodgson and others remind us (Hodgson 1988 1989 1997 1999Grafstein 1992 Camic 1986 Johnson 1992 Nelson and Winter 1982Veblen 1899) social habits are the results of earlier choices and are ameans of avoiding endless deliberation Because cognitive frameworks

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

602

are learned through habit individuals rely on the acquisition of suchcognitive habits before reason communication choice or action arepossibleWhereas Schotter (1981) and other game theorists take the individual

as an agent unencumbered by previous habits Field (1984) and othershave stressed that there can be no games without prior norms and rulesand thus a set of norms and rules must be presumed at the start Thosewho attempt to explain institutions from individual behavior alone areusing a bad strategy (Hodgson 1998)The position here ndash heavily inuenced by Hodgson (1998 1997 1988)

ndash is that individuals are embedded in a complex institutional environ-ment and that institutions not only constrain but also shape individuals(also see Hollingsworth and Boyer 1997) It would be a mistake howeverto get involved in an innite regress in order to determine which camerst ndash individuals or institutions Of course institutions are formed andchanged by individuals just as individuals are shaped and constrainedby institutions But at a macro level it is institutions that provide acognitive framework whereby individuals can cope with their reality Inthis sense the micro and macro worlds are intertwined At the macrolevel there is considerable stability but at the micro level individualshave a signicant level of autonomy and there can be widespread diver-sity As Hodgson (1988) reminds us most institutions in a temporal senseexist prior to the individuals in any given societyIt would be a serious mistake to downplay the importance of indi-

viduals and micro level analysis as we study institutions In the nalanalysis it is at the level of individuals that norms rules habits conven-tions and values exist An individual is born into and socialized intogroups and a society and this is how one early in life acquires a senseof appropriate forms of behavior Because of the way that individualsare socialized into a world of rules norms habits conventions andvalues it is unnecessary for individuals to restructure the world anewevery day (Douglas 1987 Elster 1989) Every action does not have tobe seriously reected upon For this reason institutions provide cogni-tive frameworks for individuals make their environments predictableprovide the information for coping with complex problems and envi-ronments In the words of Johnson (1992 26) lsquoInstitutions reduceuncertainty coordinate the use of knowledge mediate conicts andprovide incentive systems By serving these functions institutions providethe stability necessary for the reproduction of societyrsquo However eachsociety has different forms of habits rules and norms and hence differentincentives and disincentive systems for learning and forgetting forprocessing information But because individuals have varying degreesof autonomy individuals and groups can deviate from the prescribedforms of behavior in a society And of course these changes at the level

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

603

of individuals become important in understanding processes of societalchangeThese views are not meant to imply that the type of institutional

analysis proposed herein approximates a general theory of society Thisis clearly not the case However it is intended to represent the rst stepsin a mapping exercise of the boundaries of institutional analysis and tosuggest a few methodological insights for studying institutionsOne should try to see norms and rules as continuous and not as

dichotomous entities to recognize that they come in varying strengthsLegro (1997) has suggested that we assess the robustness of norms andrules with three criteria their simplicity their durability and their concor-dance Simplicity refers to how well actors understand norms and ruleshow well they can be applied within a specic situation Some normsand rules can be so complex that actors can have considerable difcultyin applying them in specic cases Durability addresses the issue of howlong norms and rules have been in effect ndash in short in order to assesstheir level of legitimacy While the position of this article is that normsrules and values are quite durable they do vary in this respectConcordance refers to how widely applied a norm or rule is Thisaddresses the degree to which a rule is a meta rule the degree to whichit incorporates the heterogeneity of other norms and rules In sum theclearer the norms and rules of a society the longer they have been inexistence and the more widely applicable they are the greater theirimpact on a society Hence the more robust the norms and rules thegreater their impact on a society and the less their robustness the greatertheir exibility and the less their effect on shaping a societyrsquos outcomesand performancesBecause norms rules and values are quite durable they play an impor-

tant role in shaping the history of societies thus contributing to a greatdeal of path-dependency Actors attempt to adjust to their contempo-rary environment but since they are products of the past the historicallegacy of norms rules and values inuences the decisions they makeAlthough actors have some capacity to alter the course of their historythey are constrained by their past and the degree to which they canmove beyond their past is limited As Lanzara (1998) Johnson (1992)and others have argued societal inertia is a basic feature of institutionsThey provide the basic stability necessary for change Degrees of historyare continuously reproduced by the way in which the inhabitants ofsocieties are socialized History matters but at critical points in historythere is punctuated equilibrium (Somit and Peterson 1992) During mostperiods of history there is considerable stability in the norms rules andvalues of a society but at critical moments norms rules and values canquickly and dramatically be redened At all times norms rules habitsconventions and values are inuencing each of the other components in

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

604

the institutional framework discussed below but these other institutionalcomponents feed back and can modify rules norms conventions etc(Murmann 1998)

Second level of analysis

The norms rules habits conventions and values of a society lead to thenext level of analysis ndash the institutional arrangements which are involvedin the coordination of various economic actors producers and suppliersof raw materials knowledge etc processors of raw materials infor-mation workers customers of raw materials nished productsinformation etc nanciers governmental and other types of regulatorsThese actors regularly engage in contests to resolve various economicproblems in virtually all sectors of society how are prices to be setWhat quantity of various products is to be produced How are stan-dards of various products processes etc to be set How is the qualityof products and processes to be determined How are various societalprocesses to be nanced In order to confront these problems and toaddress the conicting positions of economic actors as they address these problems societies develop various institutional (governance)arrangements for coordinating different actors These consist of marketsvarious types of hierarchies and networks associations the state commu-nities and clans (see Hollingsworth and Boyer 1997 Campbell et al1991 ch 1)Each of these particular kinds of coordinating mechanism has many

different types For example there are many types of states (eg theregulatory state developmental state authoritarian state welfare state)on which there is an extensive literature (Kim 1997) Similarly there aredifferent types of markets networks different kinds of associations etc(Boyer 1997 Hage and Alter 1997 Schneiberg and Hollingsworth 1990)When we do institutional analysis we must engage in congurativeanalysis recognizing that actors are not coordinated or governed by asingle type of institutional arrangement Some of the literature discussesindustrial sectors as though they were coordinated or governed by asingle institutional arrangement whereas in fact each sector of aneconomy is coordinated by a conguration of institutional arrangementsSome congurations coordinate actors in certain problem areas whileother congurations of institutional arrangements coordinate actors inaddressing other problems The types of congurations which are domi-nant in a society are somewhat stable and tend to persist over timewithin a society (Hodgson 1999)When one mode of coordination is dominant in a society it will

inuence the role which other coordinating modes will play Hence thestrong role of the state in the Soviet Union inuenced the role of markets

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

605

associations etc in the governance of the Soviet economy Similarly theprominence of particular modes of coordination in a society inuencesits style of innovationMuch of the literature on institutional arrangements (eg forms of eco-

nomic coordination) remains fragmented and unintegrated It is helpfulto array these various forms of coordination in a two-dimensional tax-onomy as in Figure 2 On the vertical dimension the economistrsquos viewof a self-interested agent is contrasted with a more sociological perspec-tive according to which social rules obligation and compliance shapehuman actions On the horizontal dimension there is another continuum(ie the distribution of power) At one extreme of the dimension onends many and relatively equal agents interacting (eg as in a well-organized spot market) At the other extreme inequality in power resultsin a hierarchical form of coordination which structures the interactionbetween principals and agents or between leaders and followers Wherea society falls along the distribution of power is inuenced by the rulesnorms values etc which are dominant in a society at a particular momentin timeInstitutional arrangements can be visualized on two dimensions the

nature of action motive on the one hand and the distribution of poweron the other Markets (cell 1) combine self-interest with horizontalcoordination transactors and they reect sensitivity to concerns aboutsupply and demand thus providing ex post an unintended equilibriumParadoxically the more pure and perfect the market competition thegreater the need for codied rules of the games for coordinating economictransactions Thus collective associations (cell 6) andor various formsof state intervention (cell 4) are required to develop and enforce rulesfor transacting partners (Schneiberg and Hollingsworth 1990 Streeckand Schmitter 1985a) This is an example of how the norms and rulesand the institutional arrangements of a society are intertwined asreected in a conguration of institutional or governance arrangementsAlong the horizontal axis actors can be joined in an organization or

a rm a hierarchy is the generic terms for this institutional arrangement(cell 2) Along the horizontal line one recognizes the difference betweentransactions in a market and transactions within a rm The well-knownworks of Coase (1960 1981) and Williamson (1975 1985) utilize theconcept of transaction costs in explaining the emergence of corporatehierarchiesThere are also various types of networks (cell 5) which exhibit mixes

of self-interest and social obligation with some actors being formallyindependent and equal Yet in some networks (the large rms and theirsubcontractors) there is unequal power and initiative Networks maycomprise all kinds of actors some consist only of rms but others includeassociations and the state (Hage and Alter 1997)

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

606

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

607

DISTRIBUTION OF POWER

VERTICALHORIZONTAL

1 MARKETS 2 HIERARCHY

6ASSOCIATIONS

5NETWORKS

4STATE

3COMMUNITIES

SELF INTEREST

SOCIAL OBLIGATION

ACTION MOTIVE

Figure 2 A general taxonomy of institutional arrangements

The vertical axis deals with action motives Toward the upper part ofFigure 2 actors are engaged in individualistically oriented behaviorwhereas toward the lower part actors are more engaged in collectivebehavior and strive to cope with problems of common interest Cell 3 ndashcommunities and clans ndash consists of institutional arrangements based ontrust reciprocity or obligation and thus not derived from the pure selshcomputation of pleasures and pains This is an unconventional form ofcoordination for most neoclassical economists (however see Arrow1974) but not for many anthropologists political scientists and sociolo-gists (Streeck and Schmitter 1985a Polanyi 1944 Gambetta 1988Fukuyama 1995 Sabel 1992 Putnam 1993)In the neoclassical paradigm theorists argue that actors engage in

forms of exchanges that best promote their individual interests If somestructural conditions are fullled (absence of increasing returns to scalethe reversibility of transactions absence of uncertainty and completecontingent markets with no collusion between actors) then the invis-ible hand theorem applies and market-type activity functions quite welland also provides the optimum for society thereby combining efciencyharmony and order However an excess of market activity may welllead to ruinous competition and excessive conict There is variation inthe extent to which ruinous competition occurs depending on the socialcontext in which transactions take place Thus it is important that webe sensitive to the institutional context in which transactions areembedded and that we understand the degree to which social bondsexist at both the micro and macro levels of analysis Micro bonds facil-itate exchanges in a society but at the societal level social bonds existat the level of the collective ndash in the community or region and amongmembers of racial religious and ethnic groups All other things beingequal the more powerful the social bonds among transacting partnersthe more economic competition is likely to be restrained Thus mosttransactions occur not simply in an impersonal calculative system ofautonomous actors unrestrained by social ties ndash as implied by theneoclassical paradigm ndash but in the context of social ties variation in the strength of which leads to variation in levels of trust and transac-tion costs (Etzioni 1988 211 Granovetter 1985 Streeck and Schmitter1985a Hollingsworth and Boyer 1997 Hodgson 1988 1999 Schneibergand Hollingsworth 1990)Another form of multilateral institutional arrangement is various

types of associations (cell 6) Unlike networks clans and communitiesassociations are more formal organizations Whereas markets corporatehierarchies and networks tend to coordinate economic activity amongdifferent types of actors (eg producers with suppliers capital withlabor) associations typically coordinate actors engaged in the same orsimilar kinds of activities Business associations and labor unions are

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

608

some of the most common forms of associations for coordinatingeconomic activity in capitalist economies (Schneiberg and Hollingsworth1990 Streeck and Schmitter 1985a)Finally there is the state (cell 4) which is an institutional arrangement

quite unlike any of the others It is the state that sanctions and regu-lates the various non-state coordinating mechanisms that is the ultimateenforcer of rules of the various mechanisms that denes and enforcesproperty rights and that manipulates scal and monetary policy Overallit is the state that inuences the total incentive system of a society(Lindberg and Campbell 1991 Johnson 1992 40) At the same time thestate may also be an economic actor by engaging directly in productionand exchange relationsThe choices of institutional arrangements in Figure 2 are constrained

by the social context in which they are embedded Depending on thenature of that embeddedness there is variation in the collective formsof governance some of which are specied in the lower part of thegure Some modes of coordination in the bottom and upper levels ofthe typology are often mixed together though one coordinating modeis likely to be more prominent than another But actors are often engagedin complex congurations involving several kinds of institutionalarrangementsEach of these various institutional arrangements has its own logic ndash

its own rules its own procedures for enforcing compliance its ownnorms and ideologies which help to reduce the costs of enforcementThese are summarized in Table 2 which provides further elaborationabout the various coordinating mechanisms one nds in almost everycapitalist society Table 2 lists the organizational arrangements and theirstructures rules of exchange and means of enforcing compliance asso-ciated with each type of coordinating mechanism While each type ofinstitutional arrangement has various positive features each institutionalarrangement also has failures and these are featured in Table 3 It is thecontest between those who support and those who oppose these variousinstitutional arrangements that tends to lead to transformations in insti-tutional arrangements over time (Campbell et al 1991 Campbell 1997)Again it is difcult to conceive of pure institutional arrangements ndash

either exclusively markets or exclusively hierarchies ndash since it has beenwell known since Adam Smithrsquos Wealth of Nations that the division oflabor within the rm cannot be disentangled from the existence andextent of the market Institutional arrangements have their own distinc-tive form of efciency (good static efciency for the market dynamicefciency for rms) and inefciency often leading to considerableinequality Neither networks nor communities are panaceas for economiccoordination without being congured with other types of institutionalarrangements Networks and communities may solve certain issues but

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

609

Table 2

Log

ics of ins

titution

al arran

gemen

ts

Coordination

Organizational

Rules of

Individual means

Collective means

mechanisms

structure

exchange

of compliance

of compliance

Marke

tsEasy en

try an

d exit

Voluntary sp

ot excha

nge

Leg

al enforcemen

t of

Norm of pr

ivate

Bila

teral ex

chan

ge or

control

prope

rty

marke

t site (W

all Street)

Reg

ulations

to

Leg

itim

acy of m

arke

t en

force contracts

men

tality

Com

mun

ities

Inform

al m

embe

rship

Voluntary exch

ange

based

So

cial norms an

d m

oral

Highly institutiona

lized

ev

olve

s ov

er lon

g period

on social solid

arity an

d

princ

iples im

pose

norm

s an

d rules requ

ire

of tim

ehigh

deg

ree of tru

stob

ligations

mem

bers to accept

Kno

wledg

e of others an

d

lsquocorpo

ratersquoo

bligations

recipr

ocity ov

er tim

e

Networks

Semi-form

al m

embe

rship

Voluntary exch

ange

ove

r a

Con

tractual bon

dsPersona

l relation

sBila

teral or m

ultilateral

time pe

riod

Resou

rce de

pend

ence

Tru

st built outside

the

ex

chan

geecon

omic arena

Associatio

nsFo

rmal m

embe

rship

Restricted to mem

bers

Self-interest

Some deg

ree of

Multilateral excha

nge

Emph

asis on insider

Rep

utation

effects

compu

lsion

outsider or we

they

Priva

te interest type

of

men

tality

gove

rnan

ce

Private

Com

plex organ

izations

Restricted to mem

bers

Rew

ards to ind

ividuals

Highly institutiona

lized

hierarch

ies

which

ten

d to becom

e ex

chan

ge based

on

Asymmetric po

wer threat

rules

bureau

cratic hierarchies

asym

metric po

wer

of san

ctions

Mem

bers soc

ializ

ed into

bureau

cratic rules

corp

orate cu

lture us

e of

sanc

tions

State

Pub

lic hierarchy

Unilateral action

Exit vo

ice (vote lob

bying)

Coe

rcion

De jure

and impo

sed

Indirect political an

d

loya

ltyNorms an

d pu

blic rules

mem

bership

econ

omic excha

nge

Table 3

Failu

res of coo

rdinating mecha

nism

s

Coordinating mechanisms

Type of failure

Associations

Private hierarchies

State

Market

Communities

Networks

Enforcem

ent

Usu

ally relies on

Might enh

ance

Needs co

ntrol

Needs an

internal

Needs

tru

st and

Need an external

the state as an

oppo

rtun

istic

extern

al to state

enforcem

ent

loya

lty often

en

forcem

ent

enforcer

beha

vior

bureau

cracy

authority

coming from

au

thority

Resem

bles

The

ideal o

f (ju

dge

s

Facilitates

outside (fam

ily May

facilitate

enforcem

ent

internal m

arke

ts

parliamen

t collu

sion

and

relig

ion

cartelization an

d mecha

nism

might hur

t marke

t) to

impe

rfect

ethn

icity)

mon

opoly

of cartels

incu

mbe

nt

correct state

compe

titio

nCom

patible with

worke

rsab

uses

variou

s types of

Lob

bies can

co

mpe

tition

captur

e pu

blic

interest goa

ls

Public goo

d

Useful for

Gov

erna

nce co

sts

Can

provide

Can

not provide

Can

internalize

Useful for

and

establishing

might exceed

public goo

ds

colle

ctive go

ods

some co

llective

enha

ncing qu

ality

externality

stan

dards an

d

the be

nets of

but ha

s or deal w

ith

good

s (qua

lity

and training bu

t qu

ality for

internal division

difc

ultie

s in

externalities

training

) bu

t no

t no

t ve

ry goo

d in

setting ru

les of

of la

bor

prov

iding them

Inad

equate

othe

rs (welfare

prov

iding for

compe

titio

n in

Slow

to react to

in precise

mon

itoring of

gene

ral p

ublic

societal gen

eral

the indu

stry

chan

ges in the

am

ounts

tech

nical ch

ange

go

ods)

welfare

Useful for

environm

ent

Might fail in

and inno

vation

Mem

bers tightly

Weak in the

prov

iding

induc

ing

integrated

into

prov

ision of

man

y go

ods

tech

nical

community

colle

ctive go

als

colle

ctively

chan

geha

ve limite

d

that in

dividua

lcapa

city for

mem

bers can

not

inno

vatio

nsprov

ide for

them

selves

Table 3

continue

d

Coordinating mechanisms

Type of failure

Associations

Private hierarchies

State

Market

Communities

Networks

Efcien

cyFa

cilitates

De

cien

t in

Can

be high

ly

Some ba

sic social So

me go

ods

Slow

to en

hanc

e co

operation an

d

coop

eration an

d

bureau

cratic

relation

s cann

ot

cann

ot be

efc

ienc

y an

d

X-efcien

cy but

X-efcien

cyan

d can

not

be provided

by

delivered

at

speed of

not allocative

easily deliver

pure m

arke

t su

fcien

tly

adap

tive

ness

efcien

cy

good

s at

mecha

nism

slow costs

except in

low cost

indus

tries whe

re

tech

nology

is

complex an

d rapidly cha

nging

Equity

Narrow

Excessive

Might enh

ance

Facilitates

Might lead to

Whe

n widely

encompa

ssing

multiplication

ineq

uality

ineq

uality in

retard

ed

dev

elop

ed in

to

associationa

l of con

trollers

(pow

er

income an

d dev

elop

men

tindus

trial

stru

ctures lead

(fru

stration

and

an

d privilege

)wealth

districts

to in

come

ineq

uality)

netw

orks m

ay

equa

lity

facilitate greater

equality an

d

inco

me

distribution

whe

n wea

kly

dev

elop

ed

netw

orks ten

d to

increase soc

ial

ineq

uality

Note For allo

cativ

e efc

ienc

y an

d X

-efcien

cy see Leibe

nstein (19

66 19

76)

they raise other no less severe problems It is important to recognizethe imperfection of any single institutional arrangement in order tocomprehend the origin and transformation of any other institutionalarrangement (Hodgson 1999 ch 3)As suggested above each institutional arrangement is congured with

other institutional arrangements Usually one particular institutionalarrangement is more dominant in a particular conguration but becauseeach type of institutional arrangement has its own strengths andweaknesses there is no simple structural logic in the governance or coor-dination of a society Each institutional arrangement constrains the otherbut the inherent tension among the various institutional arrangementswithin a conguration contributes to changes in the governance orcoordination of a society The routines and logics of each of these cong-urations of institutional arrangements provide constraints and incentivesfor actors The way that actors perceive the incentives and constraintsof these governance congurations leads to a particular market logicand it is the specic market logic of a society which inuences its speciccapacities and weaknesses The inherent strain among the different logicsin a conguration of coordination helps to provide the exibility for asociety to adapt to new circumstancesIf there were a society with pure markets or a society coordinated

only by the state there would be too much rigidity and too little diver-sity to cope with the vast uncertainty in the global environment A societywith very little diversity in its coordination mechanisms would havelittle capacity to adapt to new circumstances Congurations with consid-erable diversity of institutional arrangements provide for a certainamount of incoherence in governance but they also provide for thecapacity to adapt to new circumstances While the Soviet Union wasdominated by the state as a coordinating mechanism there were alwaysfunctioning markets in the system Moreover the feudal society of thetwelfth and thirteenth centuries consisted not only of hierarchical rela-tionships between serfs and masters but also of urban guilds clericalhierarchies and markets (Johnson 1992 38) In sum the robustness ofinstitutions often depends on multiple and diverse principles and logicsof action on the inconsistency of principles and procedures on patternedforms of disorder (Lanzara 1998 Orren and Skowronek 1991 320 329)

Third level of analysis

The next level consists of the institutional sectors of a society The rulesnorms and values of a society inuence the array of institutional arrange-ments and levels one and two inuence the nature of and the relation-ships among various institutional sectors and all three levels areintricately linked together to form a social system of production Together

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

613

all of these components inuence the performance of economic sectorswithin a society as well as the performance of the total society A socialsystem of production is the way that a societyrsquos institutions (see rstlevel of analysis above) its institutional arrangements (see second levelof analysis above) and its institutional sectors are integrated into a socialcongurationAn institutional sector includes all organizations in a society which

supply a given service or product along with their associated focal orga-nizations (eg major suppliers funding sources regulators and so forth(Scott et al 1994 108 117 Campbell et al 1991 Hollingsworth et al1994) Institutional sectors include but are not limited to the societyrsquossystem of education system of research business system nancialmarkets legal system and the state The structure of the nancial marketssystem of training and education industrial relations system system ofresearch and state and legal systems are distinct and idiosyncratic ineach society In short they are system specicAll of these institutional sectors tend to cohere with each other

although they vary in the degree to which they are tightly coupled intoa full-edged system While each of the institutional sectors has someautonomy and may have some goals that are contradictory to the goalsof other institutional sectors with which it is linked an institutional logicin each society leads institutions to coalesce into a complex social con-guration (Hollingsworth 1991) This occurs because the institutionalsectors are embedded in a culture in which their logics are symbolicallygrounded organizationally structured technically and materially con-strained and politically defended The conguration of institutionalsectors usually exhibits some degree of adaptability to new challengesbut continues to evolve within an existing style But under new circum-stances or unprecedented disturbances these institutional congurationsare exposed to sharp historical limits as to what they may or may notdo (Schumpeter 1983 David 1988 Arthur 1988a 1988b Haringkanssen andLundgren 1997)Why do all of these different institutions coalesce into a complex

social conguration which is labeled here as a social system of pro-duction The literature suggests two contrasting interpretations Part ofthe answer ndash indeed a controversial one ndash is that these institutionalsectors are functionally determined by the requirements of the practiceof capitalism in each time and place (Habermas 1975) Another expla-nation emphasizes the genesis of the actual conguration via a trial and error process according to which the survival of rms regions orcountries is the outcome of complex evolutionary mechanisms (Maynard-Smith 1982 Nelson and Winter 1982) However the problem is evenmore complex Markets and other mechanisms for coordinating rela-tionships among economic actors place constraints on the means and

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

614

ends of economic activity to be achieved in any society The othercoordinating mechanisms include different kinds of hierarchies varioustypes of networks and associations ndash trade unions employers and busi-ness associations (Hollingsworth and Lindberg 1985 Campbell et al1991) The logic of these various institutional arrangements providesactors with vocabularies and logics for pursuing their goals for deningwhat is valued and for shaping the norms and rules by which theyabide (see Tables 2 and 3) In short in contrast to the logic of theneoclassical paradigm the argument here is that the dominant type ofinstitutional arrangements places severe constraints on the denition of needs preferences and choices of economic actors Whereas the neo-classical paradigm assumes that individuals and rms are sovereign thisarticle is based on the assumption that rms are inuenced by the hold that the institutional congurations making up a social system ofproduction have on individual decision making (Campbell et al 1991Etzioni 1988 Streeck and Schmitter 1985a Hollingsworth et al 1994Hollingsworth and Boyer 1997 Magnusson and Ottosson 1997 North1990 Hodgson 1999)Standard neoclassical economic theory has tended to downplay the

role of production and consequently of rms Even transaction cost theo-rists who are concerned with analyzing the rm as a coordinatingmechanism have been relatively unconcerned with the various compo-nents of a social system of production Indeed as long as there waswidespread optimism about the efcacy of Keynesian economics therewas relatively little concern among neoclassical economists with thesupply side of the economy Even in the opinion of most Keynesians agroup of experts should ideally be able to shape the size of aggregatedemand while the supply side of the economy would be left to the twominimalist institutions of neoclassical economics ndash markets and manage-rial hierarchies For several decades however it has become increasinglyobvious that some of the most competitive and successful patterns ofindustrial output and industrial production in capitalist economies donot derive from the neoclassical prescription of unregulated markets andcorporate hierarchies complemented by a neoliberal democratic stateIndeed it is now well understood that certain highly successful produc-tion patterns require for their emergence and survival institutionalarrangements and environments the very opposite of the prescriptionsfound in the neoclassical paradigm (see especially Streeck 1992 but alsoHollingsworth and Streeck 1994) Thus social scientists have come tounderstand that if they are to comprehend the behavior and performanceof contemporary economies concerns about social systems of produc-tion must be brought into the pictureBecause production involves more than technology a number of social

scientists have an increasing concern with social systems of production

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

615

The same equipment is frequently operated quite differently in the samesectors in different countries even when rms are competing in the same market (Maurice et al 1980 Sorge and Streeck 1988 Sorge 1989Hollingsworth et al 1994) Variations in production and process tech-nologies are inuenced partly by variations in the social environmentsin which they are embedded In other words rms are embedded intocomplex environments which among other things place constraints ontheir behavior Thus a social system of production is of major impor-tance in understanding the behavior and performance of an economyHow the state and other institutional arrangements (eg marketsnetworks private hierarchies associations) coalesce and are related toparticular social systems of production are important determinants ofeconomic performanceThe congurations of institutional arrangements that coordinate or

govern the behavior of actors in one society and its structure of specicinstitutional sectors cannot easily be transferred to another society forthey are embedded into a social system of production that is societallydistinct (Hollingsworth 1997) Societies borrow selected principles offoreign management styles and work practices but the effectiveness ofsuch borrowing is generally limited Economic behavior and performanceare shaped by the entire social system of production in which actors areembedded and not simply by specic principles of particular manage-ment styles and work practices Moreover a societyrsquos social system ofproduction tends to limit the kind of goods that it can produce and withwhich it can compete successfully in international marketsSome scholars (Kenney and Florida 1993 Oliver and Wilkinson 1988)

have assumed that the diffusion of particular forms of management stylesand work practices across societies could lead to system convergenceReferring to the Japanese production system Kenney and Florida haveargued that it consists of organizational practices whose fundamentallsquogenetic codersquo can successfully be inserted into another society and beginsuccessfully to reproduce its behavior in the new environment (1993 8)Their position is in the intellectual tradition of Antonio Gramsci (1971)who contended decades ago that the American system of mass produc-tion would most certainly diffuse to Europe over timeHowever the argument of this article is that even though British

French and American rms may adopt certain aspects of Japanesemanagement styles (eg just-in-time production self-managing teamsquality circles the use of statistical process controls) or some variant ofthe German vocational system their social systems of production willnot converge Moreover the overall conguration of a societyrsquos socialsystem of production inuences its sectoral and overall national economicperformance This helps to explain why societies have different systemsof innovation and why they vary in the clusters of industries in which

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

616

they are highly competitive in the international markets (Nelson 19931999 Hage and Hollingsworth 2000 Berger and Dore 1996)In order to understand how and why a societyrsquos economy performs

as it does it is necessary to understand its entire social system of produc-tion If a society is to modify substantially the performance of itsinstitutional sectors and its entire economy it cannot adapt only someof the management and work practices of its foreign competitors Ratherit must alter its entire social system of production Because a societyrsquosmodes of economic governance and coordination and its institutionalsectors develop according to particular logics and are system specicthere are serious limitations to the extent to which a society may mimicthe institutions (eg the rules and norms) institutional arrangements andinstitutional sectors of other societiesIn the history of modern societies there are logics by which institu-

tions coalesce into a social system of production (Hollingsworth 1991)Though institutions are constantly changing there are sharp limits tothe type and the direction of change that any particular institution canundergo because of its linkages with institutional arrangements and insti-tutional sectors Thus a societyrsquos business rms educational systemnancial markets industrial relations system etc can engage in seriousrestructuring only if most of the other institutional sectors also changeIn social systems of production there are pressures toward consistencyin the norms rules and values across institutional sectors though in anycomplex society social systems are obviously imperfectly integratedIndeed the degree to which the institutional norms and rules makingup a social system of production are loosely or tightly coupled is a vari-able of considerable importance In general the institutional sectorsmaking up a social system of production are interdependent and changesin one generally result in changes in the others Since each institutionallevel is dependent on the others for various types of resources there isinterdependence among the differing institutional spheres Moreovereach society has its norms moral principles rules and laws and recipesfor action as well as its own idiosyncratic customs traditions and prin-ciples of justice (Burns and Flam 1987)There are also other inherent obstacles to convergence among social

systems of production for where a system is at any one point in timeis inuenced by its initial state Systems with quite different initial statesare unlikely to converge with one anotherrsquos institutional practicesExisting institutional arrangements block certain institutional innova-tions and facilitate others (Roland 1990) Thus the institutions makingup a social system of production provide continuity even though insti-tutional arrangements are always changing but with a particular logicWhile Williamson (1975 1985) suggests that actors tend to choose theinstitutional arrangements which are most efcient North (1990) is

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

617

much closer to the mark in his argument that most societal institutionalarrangements exist as a result of custom and habit and are thereforeinefcient At any moment in time the world often appears to its actorsas very complex and uncertain For this reason actors often engage incontradictory forms of behavior pursuing different strategies as hedgesagainst a very uncertain world (Lanzara 1998) And their hedging andcontradictory forms of behavior may lead in somewhat different soci-etal directions all constrained by the institutional fabric within whichthe actors are embedded This kind of contradictory behavior occurs inpart because of the contradictions inherent in the dominant congura-tions of institutional arrangements with their contradictory logics (seelsquoSecond level of analysisrsquo above)Despite the emphasis on the logic of institutional continuity this is

not an argument that systems change along some predetermined pathThere are critical turning points in the history of highly industrializedsocieties but with the choices limited by the existing institutional terrainBeing path-dependent a social system of production continues along aparticular logic until or unless a fundamental societal crisis intervenes(Milgrom et al 1991 Krugman 1991 Durlauf 1991 Hollingsworth 1991Pred 1966 David 1988 Hodgson 1998)At this point it is important to confront the question of how social

systems of production evolve And this question gets to the heart of theproblem of building complex societal structures Certainly social systemsof production did not emerge from some process of social engineeringMoreover the various component parts of each social system of produc-tion have often not been designed to be part of a social system ofproduction The component parts of each social system of productionhave emerged more often than not as unintended by-products of goalswhich various actors had in mind at earlier moments in time It is usuallythe case that actors at time lsquotrsquo fail to comprehend the long-term conse-quences of their actions Wolfgang Streeck the author of several brilliantpapers on social systems of production argues that institutional sectordesigns created for one purpose generally address that goal but overtime those designs have quite unintended consequences (Streeck 1997a1997b Wright 1998) Describing the conguration of the German socialsystem of production of the 1970s and 1980s Streeck points out that itwas the unintended by-product of multiple points in history Someelements were pre-Wilhelmian others were introduced by the Alliedpowers after 1945 and others emerged during the years of the GermanFederal Republic All component parts of the German social system ofproduction

were and continue to be changing for their own reasons as wellas in reaction to each other and certainly there can be no presump-tion of a pre-established t between them even though one might

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

618

want to allow for some reinforcement effects of the modelrsquos histor-ically contingent social and economic success

(Streeck 1997b 54)

In sum the emergence of social systems of production is a long-termevolutionary process with each part interacting with its environment andresulting in a congurative whole but with no previous design by eithera single actor or a collectivity of actors As suggested above actors do not recreate their world anew At any moment in history the totalinstitutional complex of a society necessarily contains the resources and legacies of its past Preexisting institutional complexes are nevercompletely wiped out Societies consist of multiple layers of history withtheir diverse logics of action However the cumulative effect of smallperipheral changes in altering particular institutional sectors and insti-tutional arrangements can be substantial (Murmann 1998 Lanzara 1998)lsquoInstitution building is affected more by the ways in which people havecodied the past than by how they envision the future There are manymore potential resources in pre-existing institutional arrangements thanit is usually assumed or suspectedrsquo (Lanzara 1998 30)

Fourth level of analysis

The next level of institutional analysis organizational structures is some-what more controversial As suggested above Douglass North (1990)draws a sharp distinction between institutions and organizations Morerecently many organizational theorists (Powell 1991 Powell andDiMaggio 1991 DiMaggio and Powell 1983 Baum and Oliver 1992Kondra and Hinings 1998 Townley 1997) argue that institutional rulesnorms and conventions unfold in tandem with organizational structuresand this is my positionThe literature which focuses on how institutions inuence organiza-

tions is quite different from two other theoretical literatures which alsoare concerned with how organizational environments shape the behaviorof organizations There are resource dependency theorists (Pfeffer andSalancik 1978) who emphasize the role of environmental resources inshaping organizations while Hannan and Freeman (1977 1984) and otherpopulation ecologists emphasize the survival of organizations givencertain kinds of organizational conditions (Orru et al 1991)The institutionalist perspective as implied above emphasizes the

normative environment in which organizations are embedded It is a perspective which focuses on the way in which organizations in their behavior tend to conform to the institutional rules and normswhich are dominant in the organizational environment However allthree perspectives ndash the institutionalist the population ecology and

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

619

the resource dependency ndash emphasize how the environment inuencesorganizations and how organizations which are subject to the sameenvironment tend to converge in their behavior to have what DiMaggioand Powell have labeled organizational isomorphism (DiMaggio andPowell 1983)There has emerged a vast literature which demonstrates that within

each society there are scal political judicial and other regulatory normswhich limit and shape the culture and structure of organizationalbehavior The normative institutional environment of organizations limits the options of what organizations do in a particular society and inuences the patterns of ownership relations with suppliers andcustomers In short it is the normative environment of organizationswhich denes within a particular society what is socially acceptablebehavior for organizations (Hamilton and Biggart 1988 Hollingsworthand Hollingsworth 2000 Meyer and Rowan 1991 Zucker 1987 1988Orru et al 1991 Townley 1997)Thus far most of the analyses on isomorphic pressures operating on

organizations have been cross-sectional in nature However the entireconcept of isomorphism implies that there are strong environmental pres-sures exerted on organizations and in order to observe this phenomenonit is important that we have longitudinal studies which assess howchanges in an organizationrsquos institutional environment inuence changesin the structure and culture of organizations Moreover the expectationhere is that when one analyzes the patterns of organizational changeover time the isomorphism will be holistic ndash that is within a particularsociety there will be a tendency for the internal structure of particulartypes of organizations to converge in large part because of the pres-sures to conform to the changes in the external norms rules and valuesA close reading of Aoki (1990) and Slack and Hinings (1994) nds thatchanges in the national institutional environment of organizations haveinuenced changes in the internal structure of organizations within aspecic industry of a particular societyHowever it is inappropriate to view organizations as changing only

in relation to exogenous environmental change There is a kind of co-evolution which occurs between organizations and their institutionalenvironments Co-evolution implies nonlinear feedback between orga-nizations and their institutional environments This is of course quitedifferent from the more common way of studying organizations bymodeling relations between independent and dependent variables Ofcourse the more common type of analysis is quite appropriate whenthere are simple relationships which do not involve complex feedbackprocesses But in the strategy for institutional analysis which is proposedherein it is less useful to separate independent from dependent vari-ables and more useful to understand the interacting and co-evolutionary

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

620

processes among all of the levels of institutional analysis discussed inthe various sections above (Baum and Singh 1994a 379ndash80 Mowery andNelson 1999 Murmann 1998)The logic of the perspective herein suggests that within every

society there is variation in the structure and culture of business rmsuniversities and other complex organizations (Kondra and Hinings1998) However this variation takes place within parameters which aresystem specic For example every German rm and every Germanuniversity is different from every other German university and rm but there are a set of cultural and structural characteristics which distin-guish German rms and universities from those in America Similarlyeach Japanese rm and university is unique yet they have a set ofcultural and structural characteristics which are system specic andwhich differentiate them from business rms and universities in other countriesIn all societies each organization has its own distinctive organizational

rules norms and conventions which are subordinate to the meta norms and rules of the larger society within which they are embeddedHowever the strength of the institutional environment within whichorganizations are embedded varies from society to society Some societies have multiple institutional environments and there is hetero-geneity in terms of what constitutes appropriate behavior oforganizations In such societies there can be conict over whatinstitutional logics should regulate specic organizational functions(Friedland and Alford 1991 Townley 1997 262ndash3) In those societies inwhich the institutional norms habits and rules are most developed and in which the institutional pressures to conform are greatest thereis less variation in the structure and culture of business rms and variouskinds of research organizations In such societies the connectednessbetween research organizations and their institutional environments is sufciently strong for organizations to have low autonomy to pursue independent strategies and goals in these societies there is a great deal of organizational isomorphism Conversely the weaker the institutional environment in which research organizations areembedded the greater the variation in the structure and culture of business rms and research organizations Moreover where the institu-tional environments are more weakly developed organizations havegreater autonomy and exibility to respond to the development of new knowledge and to be highly innovative Hence it is in those societies where the institutional environments are most developed and most rigid and when there is less organizational autonomy and exibility that fewer radical innovations in basic and applied science as well as in totally new products and industrial sectors havebeen made

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

621

Fifth level of analysis

We come next to the outputs and performance of the various institutionalcomponents of a society It is at this level that institutional componentsare more pragmatic and exible For example in the legal sector thereare specic statutes and court rulings in the state sector there are specicpolicies in the business sector there are new products new technolo-gies and market strategies It is at this nal level of analysis thatinstitutional spheres are most open and susceptible to change and cross-national mimicry is easiest and most commonThrough the outputs of the society we can obtain some assessment as

to how well a society is performing Moreover we can assess how inno-vative it is how egalitarian it is in the distribution of its resources howegalitarian it is in terms of levels of health education etc Just as it iscomplicated to measure how egalitarian a society is similarly it is dif-cult to assess how innovative it is For example societies vary in makingincremental and radical breakthroughs in basic and applied science in developing totally new products and new kinds of organizations inincremental and process innovations in existing products and organiza-tions in developing new and different forms of marketing ndash bothdomestically and globally (Hage and Hollingsworth 2000) Of coursesocieties vary in their rates of economic growth in their rates of economicproductivity in their quality of life (eg rates of crime life expectancyetc) All of these performance criteria feed back and inuence each ofthe levels of institutional analysis discussed above ndash rules norms valuesetc institutional arrangements institutional sectors the structure andform of organizations Moreover performance measures may inuenceeach other For example if a society has very low rates of growth itmay not be very innovative in some of the types of innovative activitiesmentioned above Good or poor performances of certain kinds inuenceother performance indicators (Hollingsworth et al 1990)Different institutional arrangements and different social systems of

production result in different types of economic performance Hence solong as societies have different social systems of production there areserious constraints on the degree to which they can converge in theirinnovative styles Different social systems of production tend to maxi-mize in a more or less explicit manner different performance criteriausually mixing considerations about static and dynamic efciency protsecurity social peace and economic andor political power In short incontrast to the implications of neoclassical economic theory in real-worldeconomies there are no universal standards all economically rationalactors attempt to maximize Economic history provides numerous exam-ples of how a great variety of principles of rationality are implementedin different societies

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

622

A critical question in institutional analysis is whether an existing socialsystem of production which supports a set of routines for a particularkind of technology and industry can shift from old practices to newones There are numerous historical accounts of how the social systemof production in a particular society which worked so well for a numberof years could not adapt to new technologies (Lazonick 1990 Veblen1915 Nelson 1994) Schumpeter (1983) and Freeman (1991) have devel-oped the idea of a techno-economic paradigm which suggests thatdifferent technologies require specic forms of organizational arrange-ments ndash or what I called social systems of production Because oflsquoinstitutional inertiarsquo the social system of production of a society maynot be able to adapt to a new techno-economic paradigm (Nelson 199458 Lanzara 1998)Whether a social system of production can sustain its performance

standards depends not only on its intrinsic economic rationality but alsoon where it ts into a larger system If a particular social system ofproduction is immune from the innovativeness and competition of analternative system survival can be long lasting But if different socialsystems of production with diverging criteria of good economic perfor-mance meet in the world arena the arbitrariness of nationally imposedconstructed performance standards may be superseded by alternativeperformance criteria as a result of international competitivenessThe world economy is also socially constructed just as are national

economies Even if different social systems of production are competingin the international arena it is not always possible to determine whichis more competitively effective at any moment in time Hegemonicnation-states in the short run shape the rules of trade to favor their indus-trial sectors and rms But the history of hegemonic powers suggeststhat in the longer run such social systems of production sustainedlargely by military and political power eventually give way to moreinnovative and competitive social systems of production In our ownday as nation-states are increasingly integrated into a world economyeven hegemonic powers lose their innovative and competitive advan-tage and their share of world output decreases Such a country mayattempt to restructure its institutional arrangements and to readjust itsperformance preferences But to restructure its system of innovationgenerally calls for a major redistribution of power within a societyLargely for this reason societies have historically had limited capacityto reconstruct a system of innovation in the image of their major competi-tors Each countryrsquos social system of production is a conguration of ahost of norms rules and values as well as of institutional arrangementsEach system is constantly changing and is open to inuence from othersystems And indeed many technologies and practices diffuse from onesociety to another but a societyrsquos capacity to be innovative is constrained

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

623

by the existing social system of production Thus the same technologymay exist in numerous countries but how it is employed and how itinuences societal outcomes and performances varies from one institu-tional conguration or society to another (Hollingsworth 1998)

PERSPECTIVES ON INSTITUTIONAL CHANGE

A fundamental problem which remains unresolved in institutionalanalysis involves the nature of institutional change There is much confu-sion and miscommunication in the social sciences about institutionalchange in large part because we do not even know how to measure therate of institutional change Of course we know that there are bothexternal and internal focuses for change and we have had a good bitof scholarship on this subject What is not clearly understood is thatinstitutions and institutional arrangements within societies are histori-cally rooted that there is a great deal of path-dependency to the waythat institutions evolve Moreover the more intricately linked each sectoris with each other sector and with a societyrsquos rules and norms the lesschoice actors have to devise new institutions and institutional arrange-ments On the other hand if a sector or actor is too isolated it may betoo weak within a society to bring about any kind of effective changeThese considerations are very critical to the problem of how muchfreedom actors have to create new institutions afresh to what extentinstitutions diffuse from one society to another (Murmann 1998 Lanzara1998) This is a fundamental question for students of institutionalanalysis as we attempt to understand how the different institutionallevels are linked togetherThere are a variety of reasons why there is confusion in studying insti-

tutional change There is disagreement over the extent to which actorshave the freedom to build new institutional arrangements and if theydo to what extent the new arrangements and organizational patternsmay depart from past practices Among some institutional economists(Williamson 1985) there is a basic assumption that actors build efcientinstitutions by pursuing their self-interest and by promoting theirstrategic goals Moreover there is the assumption that a mixing of fullrationality with market competition tends to produce all of the optimalinstitutions that are needed in order to coordinate a complex capitalisteconomy (Hollingsworth and Boyer 1997 ch 14 Hodgson 1999) Incontrast the argument developed here is much more complexSocial institutions are historically rooted and there is a great deal of

path-dependency in the way that various institutional componentsevolve The shape of institutional congurations at any moment limitsthe type of options for change Because there is a great deal of institu-tional inertia radical change in the institutional components of a society

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

624

is uncommon (Lanzara 1998) Recognizing the constraints existinginstitutions exert on actors we note that existing congurations of insti-tutional arrangements institutional sectors and organizational propertiescan limit but also inuence the degree to which norms and rules canchange Many scholars (North 1990 Ostrom 1990) have emphasized therole of norms rules ideologies and values in limiting the ability of actorsto develop new institutional arrangements which radically depart fromexisting ones However there is a two-way process of interaction aprocess of co-evolution (Baum and Singh 1994a Nelson 1994) Thus asFriedland and Alford (1991 244) argue actors develop new institutionalarrangements by recognizing organizational failures and low perfor-mance but these new arrangements do not necessarily change a societyrsquosnorms rules habits and its underlying value systemCampbell (1997) advances our understanding of how existing institu-

tions enable actors to construct radically new institutions He recognizesthat most institutions and institutional arrangements embody a degreeof inertia from which it is difcult for actors to depart Moreover thedifferential power relations among actors make it unlikely that lesspowerful actors can change the existing power structure of a societyMost change in the institutional components of a society evolves througha process of constrained selection which reects to a considerable degreethe existing arrangements and power relations This becomes the inter-pretative frame for social actors with institutionalized scripts and ritualswhich tend to be taken for granted and appear to be quite ordinary andnatural This institutionalized scripting of the social world is an extremelyimportant source of social stability and inertiaCampbell has developed a very fruitful explanation of change in terms

of the various institutional components outlined here He distinguishesbetween incremental and radical institutional change and argues thatradical institutional change occurs when social actors with widelydiffering norms cultural scripts and rituals for action engage in intenseinteraction with each other (see also Knight 1992) The more diverse theinteractions the greater the potential for institutional change He arguesthat as shifts in the composition of interaction among social actors occurchanges in the interpretation of problems and interests will follow Ifthere are only minor changes in the extent to which diverse actorsinteract there will be only modest institutional change But the morefundamental the changes in the interaction of diverse social actors thegreater the change in the way that actors interpret their world And thegreater the change in the way that actors interpret their world the morelikely that radical changes in various institutional components willfollow In other words diverse decision making tends to create a widerrange of interpretative frameworks of the social world than is likely tobe the case if social actors operate in isolation

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

625

This perspective of diverse interactions among social actors facilitatesour ability to move beyond the traditional view that existing institutionsconstrain the range of institutional alternatives which actors face Theolder view has tended to suggest that actors engage in institutionalchange by extending existing institutional principles habits and conven-tions Specifying the conditions which facilitate fundamental institutionalchange permits a much richer perspective for institutional analysisCampbellrsquos perspective about the conditions under which fundamentalor radical change in institutional components occurs is consistent withmy own research about the conditions under which major discoveriesor fundamental new knowledge occur the more scholars with diversebackgrounds interact with intensity and frequency the greater the like-lihood that they will develop new and alternative ways of thinking(Hollingsworth and Hollingsworth 2000)These views of institutional change are also consistent with the recent

work of Nelson and his colleagues (Mowery and Nelson 1999 Murmann1998) They report the research on a variety of industrial sectors in anumber of different countries revealing that networks of actors in partic-ular industries business and professional associations universityresearch departments and institutes are frequently able to use their collec-tive power to modify a societyrsquos property rights research system andgovernmental policies It is the intense and frequent interaction of thesediverse complex sets of actors ndash sometimes constituting a technologysystem ndash that often leads to incremental changes in a societyrsquos total socialsystem of production It is this kind of collective action that brings aboutorganizational change change in institutional sectors institutionalarrangements and even in the rules and norms of a society And if weare to understand how the style of innovativeness of a society changeswe must rst understand the changes in its institutional makeup It ishighly unlikely that a society will develop a new style of innovativenesswithout changing its institutional structure

FIRMS THEIR INSTITUTIONAL ENVIRONMENT ANDNATIONAL SYSTEMS OF INNOVATION

Variation in the innovative style of societies is shaped by their insti-tutional makeup ndash the various levels which are discussed aboveinstitutions institutional arrangements the structure and cohesion of institutional sectors which constitute a societyrsquos social system of pro-duction and the structure and culture of its organizations ndash especiallyits business rms and other research organizations By undertaking aninstitutional analysis of a society one can begin to understand in whatkinds of organizations the production of specic kinds of knowledgetakes place and how this is linked to particular kinds of innovativeness

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

626

With such a perspective one can gain a rich understanding of why somesocieties excel in the production of radical breakthroughs in basic andapplied science and in developing radically new products and why othersocieties excel in more incremental innovationsReecting on these issues one cannot help being impressed with the

contrasts between the innovative styles of German and American orga-nizations For example Germany has continued to be very successful in making incremental innovations in many industries in which it wasvery competitive well before World War II paper printing materialsmachinery electrotechnical products motor vehicles chemicals textileyarns and fabrics But the Germans have been much less successful indeveloping totally new products or related innovations in biotechnologyelectronics telecommunications aircraft as well as other newer indus-trial areas In contrast the Americans have been very innovative in these and other industries with products having very short lives andwith technologies which change very rapidly and are very complex(Hollingsworth 1991 Soskice 1997 1999)The importance of institutions and innovativeness for international

competitiveness has led to the recognition that there are national systemsof innovation (Lundvall 1992 Nelson 1993 Edquist 1997 Hage andHollingsworth 2000) But even before the recent studies of nationalsystems of innovation Landes (1969) was a proponent of these viewswhen he made the argument that Germany had developed a nationalsystem of innovation by building on its education and scientic researchsystems which were lacking in Britain and elsewhere However the concept of a national system of innovation immediately poses theproblem of what kind of innovations should be consideredDavid Soskice an economist at the Wissenschaftszentrum Berlin has

addressed a set of problems somewhat complementary to the perspec-tive raised here His work (Soskice 1990 1997 1999) suggests that weare sensitive to the institutional environment in which the followingactors are embedded the employees of rms the owners and nanciersof rms the competitors of rms and nally their collaborators ndash aperspective complementary to Hollingsworthrsquos work on social systemsof production (Hollingsworth 1997 1998) When one reects on the insti-tutional contextsocial system of production within which these actorsare embedded one becomes aware that those countries which have some-what inexible labor markets external to rms but a business system in which rms have long-term commitments to their employees arestrongly associated with incremental but hardly at all with radical inno-vations In the rms of such countries (eg Germany and Japan) thereis much more consensus decision making than in societies where a lsquohireand rersquo set of practices is quite pervasive And this kind of consensusdecision making ndash while highly conducive to incremental innovativeness

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

627

ndash limits the capacity for radical innovations to occur This kind of lsquolongtermismrsquo set of norms and rules occurs not only in the employmentpractices of incrementally innovative rms but also in their nancingand ownership patterns (Hollingsworth 1997) During the past thirtyyears these were characteristics of many rms in Germany and JapanSignicantly in both of these countries venture capital markets werenot highly developed nancing tended to be based more on long-termbank loans than on the equity markets and the relationships of rmstended to be relatively stable And these stable institutional arrange-ments also contributed to a high degree of innovativeness ndash butincremental rather than radical innovationsMeantime we should also pursue Soskicersquos ideas (1999) about how

rmsrsquo relationships with their competitors and collaborators shape theirstyles of innovativeness In those advanced industrial societies in whichbusiness rms are highly mobilized into business associations theretends to be a high degree of standardized quality controls over technicalnorms and a high degree of reliance on legal sanctions as monitoringinstruments vis-agrave-vis association members As a result of this high degreeof integration of rms into associations member rms are more likelyto be associated with incremental innovations than is likely to be the case with rms in those societies where associations are less welldeveloped In the latter type of society radical innovations are morelikely to occur In those advanced industrial societies where rms havelong-term stable relationships with collaborators in product design andproduction the style of innovation is likely to be incremental And it is in those societies in which there is a weaker tradition of strongcollaboration among rms that radical innovations are more likely tooccurObviously even in a country where there may be a great deal of

radical innovativeness most innovations will be of an incremental typefor there are simply fewer radical innovations But Germany and therest of Europe not only have had relatively few radical innovations incomparison with the United States but they also have lagged behindthe USA in developing relatively totally new industries Thus there hasbeen much less innovative activity in Europe in high-technology andnewly emerging industries than in the USAIn complementary research Audretsch and Feldman (1995) have

demonstrated that the USA has been more successful in spawning rmsengaged in the early stages of the industry life cycle but relative toGermany had an environment unfavorable to innovative activity byestablished rms especially in more mature industries German rmshave tended to be associated with a routinized technological regime ndashwhere established enterprises have had strong incentives to engage inincremental innovations but new rms have tended to have fundamental

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

628

innovative disadvantages relative to the Americans (see Audretsch 1995Winter 1982 Casper 2000)On both sides of the Atlantic universities engage in basic and applied

research But signicantly the institutional environment in which univer-sities are embedded inuences the degree to which the research is likelyto represent major breakthroughs or is likely to have more modestimplications Recent research in biomedical science demonstrates that inthose societies in which university professors andor graduate studentshave the following characteristics there are likely to be relatively majorresearch breakthroughs graduate students have a long training periodhigh dependency on a major professor low autonomy in dening thenature of a research project restricted job options and a university envi-ronment where there is a very modest or nonexistent entrepreneurialculture For example in Germany ndash a country with relatively few radicalbreakthroughs ndash the scholar who nishes a habilitation will have spentmany years on a research project before qualifying for a permanent acad-emic position will have been in a high-dependency position with a seniorprofessor will have had limited autonomy to dene a research projectand will be constrained from pursuing a high-risk research project Theyoung German will have an extraordinarily high level of training butwill have faced a number of disincentives to undertake high-risk researchAnd as soon as heshe has completed the habilitation and becomes aprofessor administrative responsibilities constitute another disincentivefrom engaging in high-risk or long-term research projects Moreover theGerman professor is a civil servant with a more stable and secure incomethan hisher American counterpart In a very different institutionalenvironment American professors in large research universities tend tobecome much more entrepreneurial Today many American professorsbehave as though they were operating their own rms within univer-sities and thus devote considerable time to generating funds for researchprojects assistants secretaries supplies etc And the variation in thisentrepreneurial spirit across countries is associated with variation in thetaking of high-risk research strategies and the making of major researchbreakthroughsIn sum societies vary in terms of the institutional environment in

which their organizations are embedded and this variation inuencesthe degree to which they have industrial activities in the early or latestages of the industry life cycle successful new or mature industriesand radical vs incremental innovations And it is this variation whichis encouraging rms in countries with relatively few new industries and with few or no radical innovations to invest in those countries with many new industries and numerous radical innovations ndash as astrategy to be able to transfer new technologies from the latter institu-tional environment to the former

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

629

Table 4 summarizes the way that the norms rules and values of a society are associated with Americarsquos and Germanyrsquos institutionalarrangements their social system of production and their styles of inno-vativeness

CONCLUSION

Several themes deserve to be reemphasized Most importantly institu-tional analysis is emerging as a dominant issue in several social sciencedisciplines The study of various institutional components discussedabove is a complex subject requiring multiple levels of analysis and aknowledge of multiple sectors within societies It requires a theoreticalperspective that evolves from an empirical analysis in different societiesIf we are to advance the study of institutional analysis we need cross-disciplinary collaboration and team research which are not readilyachieved given the structure of most of our research universitiesAt present scholars in each academic discipline have their own distinc-

tive strategies for studying institutional analysis Thus economists tendto use several approaches (North 1990 Williamson 1985 Hodgson 19881998 1999) anthropologists another (Geertz 1995) and sociologistspolitical scientists and historians employ other strategies (Campbell etal 1991 Powell and DiMaggio 1991 Katznelson 1998 Hall and Taylor1996) With practitioners in various disciplines pursuing their distinctiveapproaches to institutional analysis there is low potential for collabo-rative institutional analysis across the social sciences The lessons fromthe biological sciences over the last 40 years are very instructive oncepractitioners from various biological disciplines began to cooperate instudying biological phenomena by using the same concepts at the mole-cular level theoretical advance was very rapid Comparable phenomenatook place in the development of other hybrid elds eg biochemistrybiophysics (Judson 1979 Olby 1979 Kohler 1982 1994) Similarly weresocial scientists in various disciplines to work collaboratively in the areaof institutional analysis there would very likely be an acceleration intheir theoretical knowledgeThis is not a plea that everyone engaged in institutional analysis should

do the same kind of research Indeed just as biologists working at themolecular level study many different kinds of problems similarly socialscientists engaged in institutional analysis would also work on manyseparate problems but in the spirit of a collective enterprise Some wouldconcentrate their attention on the study of rules norms habits conven-tions and values while others would study how these are associatedwith congurations of various institutional arrangements (types ofmarkets hierarchies networks associations communities clans statesetc) Others would work on specic institutional sectors (eg education

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

630

Table 4

Rules norms hab

its an

d values associated w

ith sp

ecic co

ngu

ration

s of coo

rdinating mecha

nism

s soc

ial sy

stem

s of

prod

uction

an

d nationa

l styles of inno

vation

USA

Germany

Norms rules values which provide incentives for

Norms rules values which provide incentives for

short-term horizons

long-term horizons

Con

gu

rative

forms of

Marke

ts c

orpo

rate hierarchies reg

ulatory state

Assoc

iation

s (busine

ss assoc

iation

s la

bor union

s)

coordinating

corp

oratist-type

networks

marke

ts c

orporatist-

mecha

nism

s (listed in

type

state

order of im

portanc

e in

the particu

lar social

system

)

Social sys

tem of

prod

uction

Bus

iness system

Firm

struc

tures

Con

glom

erates b

ut with mov

emen

t toward

Firm

more co

ngru

ent with pa

rticular prod

ucts

elim

ination of produc

tion

in som

e sectors

emph

asis on long

-term strateg

ies rm

s are well

emph

asis on sh

ort-term

strateg

ies high

integrated

differentiation

between divisions

of rms

Firm

s mob

ilized in

Low

High

and in

tegrated

into busine

ss

associations

Indu

strial relations

Low

deg

ree of jo

b secu

rity rigid in

ternal lab

or

High deg

ree of jo

b secu

rity

exible interna

l lab

or

system

marke

t ex

ible externa

l labo

r marke

tsmarke

t rigid externa

l lab

or m

arke

tPatterns of

Dispersed owne

rship frequ

ent turnov

erMore stab

le owne

rship patterns

owne

rship

Table 4

continue

d

USA

Germany

Norms rules values which provide incentives for

Norms rules values which provide incentives for

short-term horizons

long-term horizons

Social system of

Prod

uction

Bus

iness system

Training for labo

r an

dHigh deg

ree of voluntary decision mak

ing by

Collective an

d m

ore co

mpulso

ry d

ecisions

abo

ut

man

agem

ent

individuals as to ho

w m

uch

tim

e an

d ene

rgy

how m

uch ed

ucation

is ne

eded

he

avy em

pha

sis

to inv

est in educ

ation heavy

emph

asis on

on techn

ical training (eg en

gine

ering)

busine

ss sch

ools training

in marke

ting

and

sales accou

nting

Fina

ncial marke

tsEqu

ity an

d ven

ture cap

ital m

arke

ts highly

Equ

ity an

d ven

ture cap

ital m

arke

ts w

eakly

dev

elop

ed na

ncing from

cen

tral gov

ernm

ent

deve

lope

d he

avy em

pha

sis on

corporate loan

sto priva

te rm

s for dev

elop

men

t of new

tech

nologies related

to na

tion

al security

Unive

rsity research

High individual auton

omy high poten

tial for

Low

ind

ividua

l au

tono

my until almost 40

yea

rs of

system

individual creativity u

nive

rsity pr

ofesso

rs

age poten

tial for ind

ividual creativity high

ly

high

ly entreprene

urial

cons

traine

d unive

rsity pr

ofessors not so

entrep

rene

urial

Inno

vative

styles

Highly su

ccessful in m

ajor break

through

s in

High pe

rforman

ce in proc

ess an

d inc

remen

tal

basic an

d app

lied scien

ce go

od in dev

elop

ing

inno

vation

s in existing produ

cts not so su

ccessful

totally

new

lines of pr

oducts an

d in m

ajor break

through

s in basic scien

ce or in

tech

nologies high

deg

ree of entrepr

eneu

rship

deve

loping

totally new

lines of prod

ucts of

amon

g research

ers

tech

nologies low d

egree of entrepr

eneu

rship

amon

g research

ers

Table 4

continue

d

USA

Germany

Norms rules values which provide incentives for

Norms rules values which provide incentives for

short-term horizons

long-term horizons

Prod

ucts in w

hich

Pr

oducts with sh

ort liv

es prod

ucts in w

hich

Older produ

cts with long

history machine

ry pap

er

society excels

tech

nology

is bo

th highly co

mplex

and

pr

inting

textile

s m

otor veh

icles m

achine

too

lsch

ange

s rapidly en

tertainm

ent indus

try

pha

rmaceu

ticals biotechn

olog

y softw

are

aerosp

ace

business systems nancial markets and systems of research) Distinctivewould be the recognition of how these separate studies are linkedtogether how each institutional sector is linked to a societyrsquos normsrules values etc and its conguration of institutional arrangementsSimilarly other institutional analysts might focus on the study of orga-nizations with a major concern for how the institutional environmentof organizations (eg norms rules values congurations of institutionalarrangements institutional sectors) inuenced the structure culture andoutcomes of organizations And nally others might study how all ofthe aforementioned aspects of institutional analysis inuenced a societyrsquosoverall performance Nelson and Sampat (1998) have appropriatelyreminded us that we should engage in institutional analysis in order toaddress specic problems but the view developed in this article is thatwe rst need some mapping of the terrain of institutional analysis inorder for our diverse enterprises to become a meaningful collective enter-prise especially if we are to understand how institutions inuence theinnovative processThe argument developed here is not that all social scientists should engage

in institutional analysis Obviously there are numerous other importantareas of research which lie beyond the eld of institutional analysis Nordoes this article argue that institutional analysis is the most importantsubject for social scientistsSince the structure and culture of our research universities reect an

enormous amount of disciplinary fragmentation what are the prospectsfor promoting the kind of collective and interdisciplinary research agendaproposed above In one respect the prospects are not encouragingDuring the past half century the efforts of American universities to have effective communication across the biological sciences have variedgreatly in their success At the University of California-Berkeley it was not until the 1980s that there was a serious effort to integrate thebiological sciences and at that time the campus abolished seventeendepartments to promote an integrated biological program But at theUniversity of Wisconsin the biological sciences are still very fragmentedinto dozens of different departments in botany plant pathology zoologyand physiology with two departments of biochemistry and two depart-ments of geneticsThere have been other approaches which have worked very well histor-

ically for promoting scientic integration amidst scientic diversity Forexample several summer schools and special institutes were organizedin the United States during the 1940s in an effort to overcome the frag-mentation in the biological sciences These programs made it possiblefor scientists in diverse academic disciplines and from different univer-sities and countries to come together to engage in intense and frequentinteraction Perhaps the most effective program was that held in the

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

634

summers at Cold Spring Harbor New York where a research agendafor the study of molecular biology was developed (Fischer and Lipson1988 Stent and Watson 1966) Scientists went there summer aftersummer and when they returned to their universities many began therestructuring of the biological sciences in their universitiesIn our own day we need summer institutes and other similar programs

so that scholars from diverse backgrounds can come together anddevelop a collective agenda for doing institutional analysis and forstudying the impact of institutions on the performance of societies ndashinnovativeness being one such subject If we can develop such instituteswe will increase the prospects for transforming our universities Thereis reason to be optimistic that once an eclectic group of scholars workwith a common set of concepts we can have genuine theoretical advanceif researchers can bring their diverse expertise to bear on a common setof problems There is considerable evidence that institutional analysisoffers high potential for advancing our knowledge of innovations andtechnological change (Hollingsworth and Hollingsworth 2000) If we aregoing to make substantial theoretical advance it is not likely to occurfrom the vantage point of a multiplicity of disciplines Rather it is likelyto occur as we develop a new intellectual framework and our own acad-emic journals This is a tall agenda which requires us to reassess theway that knowledge is presently produced in our societies

NOTE

This article is part of a much larger agenda involving the study of institutionsorganizations and innovations in which I am involved My debts are many andvaried The rst version was written while I was in residence in the ResearchUnit for Institutional Change and European Integration of the Austrian Academyof Sciences in Vienna in the summer of 1998 I am especially grateful to ProfessorsEgon Matzner and Sonja Puntscher-Riekmann for providing the stimulating envi-ronment in which to think through the ideas developed here Egon Matzner hasengaged in very stimulating discussions with me about the importance of insti-tutions in the general eld of socio-economics The nal version of the articlewas completed during my residence as a fellow at the Neurosciences Institutein La Jolla California I also wish to thank my colleagues in the InnovationTheme Project at the Netherlands Institute for Advanced Study (NIAS) during1998ndash9 and Edgar Grande and Karl Muumlller for stimulating conversations aboutthe interaction of institutions and organizations in facilitating particular typesof innovation I am especially grateful to Jerald Hage for helping me to developa number of issues in the article and for his useful comments on the rst draftFrans van Waarden early on helped me to work through the idea of multiplelevels of institutional analysis Others who have helped to develop the ideas inthe article have been my colleagues in the research project lsquoComparative socialsystems of productionrsquo especially Tom Burns Christel Lane Yoshitaka OkadaWolfgang Streeck and Richard Whitley My colleagues Robert Boyer SteveCasper and David Soskice have taught me much about how incentives providedby the institutional environments of rms inuence individual decision making

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

635

But my most important debt is to Ellen Jane Hollingsworth who has repeatedlyquestioned every single assumption of this article Without her rich and contin-uous tutoring this article would be even less developed than it is at presentFinally without the insightful assistance of David Gear in many phases of thisarticle it would not have been completed

REFERENCES

Aoki Masahiko (1990) lsquoTowards an economic model of the Japanese rmrsquo Journalof Economic Literature March 1ndash27

Arora A Landau Ralph and Rosenberg Nathan (eds) (1998) Chemicals and Long-Term Economic Growth Insights from the Chemical Industry New York JohnWiley amp Sons

Arrow Kenneth (1974) The Limits of Organization New York NortonArthur W B (1988a) lsquoSelf-reinforcing mechanisms in economicsrsquo in P W

Anderson K J Arrow and D Pines (eds) The Economy as an Evolving ComplexSystem Redwood City Calif Addison-Wesley

mdashmdash(1988b) lsquoCompeting technologies an overviewrsquo in G Dosi C Freeman RNelson G Silverberg and L Soete (eds) Technical Change and Economic TheoryLondon Pinter Publishers pp 590ndash607

Audretsch David B (1995) Innovation and Industry Evolution Cambridge MassMIT Press

Audretsch David B and Feldman Maryann P (1995) lsquoInnovative clusters andthe industry life cyclersquo Discussion Paper Wissenschaftszentrum BerlinMarket Processes and Corporate Development Research Unit FS-IV-95-7

Baum Joel A C and Oliver Christine (1992) lsquoInstitutional embeddedness andthe dynamics of organizational populationsrsquo American Sociological Review 57540ndash59

Baum Joel A C and Singh Jitendra V (1994a) lsquoOrganizationndashenvironment coevo-lutionrsquo in Joel A C Baum and Jitendra V Singh (eds) The EvolutionaryDynamics of Organizations New York Oxford University Press pp 379ndash402

mdashmdash(eds) (1994b) The Evolutionary Dynamics of Organizations New York OxfordUniversity Press

Berger Suzanne and Dore Ronald (eds) (1996) National Diversity and GlobalCapitalism Ithaca NY Cornell University Press

Boyer Robert (1997) lsquoThe variety and unequal performance of really existingmarkets farewell to Doctor Panglossrsquo in J Rogers Hollingsworth and RobertBoyer (eds) Contemporary Capitalism the Embeddedness of InstitutionsCambridge and New York Cambridge University Press pp 55ndash93

Burns Tom R and Dietz Thomas (2001 forthcoming) lsquoSocial rule system theorysocial action institutional arrangements and evolutionary processesrsquo in JRogers Hollingsworth and Karl Muumlller (eds) Advances in SocioeconomicsLanham Md Rowman amp Littleeld

Burns Tom R and Flam Helena (1987) The Shaping of Social Organization BeverlyHills Calif Sage

Calvert Randall L (1995) lsquoThe rational choice theory of social institutionsrsquo inJeffrey S Banks and Eric A Hanuskek (eds) Modern Political Economy NewYork Cambridge University Press pp 216ndash66

Camic Charles (1986) lsquoThe matter of habitrsquo American Journal of Sociology 911039ndash87

Campbell John L (1997) lsquoMechanisms of evolutionary change in economicsgovernance interaction interpretation and bricolagersquo in Lars Magnusson

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

636

capabilities which permitted them to have the economies of scale andscope to develop cost advantages over their competitors (Teece 1993)Chandlerrsquos work has had a profound inuence on the management liter-ature of the past couple of decades Hence in the Chandlerian traditionmuch of that literature has suggested that the key to understanding thecompetitive advantage of rms is to identify their strategies and orga-nizational structuresOver time another literature has emerged which emphasizes the

importance of the institutional environment of rms for understandingwhy rms in some countries excel in some industries but not in othersand why the rms in a specic country may excel in a particular industryat one time but may eventually lose that advantage (Landes 1969 1998North 1981 1990) More recently Richard Nelson and his co-workershave been advancing this literature by working at the frontiers of inte-grating the literature on institutions rm strategy and technologicalinnovation (Nelson 1994 1995a 1995b 1996 Mowery and Nelson 1999Murmann 1998 Arora et al 1998)The Nelson school correctly assumes that since we do not presently

have an adequate theory on how institutions rms and technologies co-evolve we are not at a stage to test a set of formal hypotheses whichow from some well-dened model Hence the Nelson school have collec-tively been developing descriptive studies of how institutions rmcapabilities and technologies co-evolve so that particular societies andrms at specic moments in time excel in particular kinds of innova-tions The goal of this kind of work has been to develop by workinginductively a better understanding of the processes of how technolo-gies rms and institutions co-evolve across a number of industries andcountriesA variety of endowments in the institutional environment provide

economic actorsrms with initial advantages or disadvantages forparticular types of technological activity (Murmann 1988 ch 7) Butover time everything is dynamic and the larger global and institutionalenvironments the capabilities of rms and the performance of rms allco-evolve and feed back onto one another However institutional envi-ronments differ widely from one society to another and the successfulrms and organizations are those which can best adapt their activitiesto the institutional environment within which they are embedded Oncea number of rms in a particular industry are successful however theymay be able to engage in collective action to modify their institutionalenvironment in order to enhance their innovativeness and their techno-logical competitiveness The studies of the Nelson group have beenparticularly important in advancing the argument that the interactionamong actors and their institutional environment is a multi-facetedprocess and that successful actors over time must not simply respond

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

597

to the institutional environment in which they are embedded but mustmodify their environment in order to maintain competitive advantagesAs these comments suggest the work of Nelson and his colleagues is

very promising for understanding why societies vary in their capacityto excel in different industries However to extend and complementtheir work we need a greater understanding of and consensus on themeaning of institutions and the institutional environment of rms Hencethis article is written in the spirit of complementing the agenda of Nelsonand his colleagues One of its goals is to develop a map of what can becalled the terrain of institutional analysis It assumes that until we havea map of this terrain scholars working in the broad eld of institutionalanalysis will not comprehend how their work interrelatesInstitutional analysis is now rather high on the research agenda of

the social sciences However we need to be aware of the obstaclesconfronting us as we attempt to advance an agenda of institutionalanalysis There is no consensus as to what is meant by institutions orby institutional analysis These terms are very widely used but they areused with different conceptualizations and the scholars who use themshare little common ground Until scholars have some consensus aboutthe meaning of the concepts they use their potential to bring about effec-tive advancement of knowledge is somewhat limited Thus thewidespread interest in several academic disciplines in the concepts lsquoinsti-tutionsrsquo and lsquoinstitutional analysisrsquo may well promise more than it candeliver given the organizational and disciplinary fragmentation ofcontemporary universitiesThere are many different approaches to the study of institutions

(Nelson and Sampat 1998) there are the new and the old institution-alisms (Stinchcombe 1997 Hodgson 1998 Langlois 1986 1989) thereis historical institutionalism (Steinmo et al 1992 Zysman 1994

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

598

ORGANIZATIONALCAPABILITIES INNOVATIVENESS

INSTITUTIONS INSTITUTIONAL ARRANGEMENTSAND INSTITUTIONAL SECTORS

Figure 1 Institutional environments organizations and innovativeness

Immergut 1998 Katznelson 1998) and several of the social scienceshave their own distinctive approaches to the study of institutions (Halland Taylor 1996 Hodgson 1988 Eggertsson 1990 Finnemore 1996Scott 1994 Calvert 1995 Hechter and Kanazawa 1997)The following comments reect some of the confusion in utilizing the

concept lsquoinstitutionrsquo Nobel laureate Douglass North in his bookInstitutions Institutional Change and Economic Performance (1990 3) denesinstitutions as lsquorules of the game in a societyrsquo To North institutions areconstraints which shape human interaction and the way that societiesevolve through time On the other hand Andrew Schotter argues thatinstitutions lsquoare not rules of the gamersquo Rather institutions are thebehavior that follows from rules Briey he is concerned with whatactors do with rules but not with what the rules are (Schotter 1981155) Many other examples might be given to illustrate the heterogeneityof approach to institutions and institutional analysis Even if scholarswere to agree with North that rules and norms are institutions theywould not necessarily agree on what a rule is Shimanoff for examplehas identied more than 100 synonyms for the concept lsquorulersquo (Shimanoff1980 57 Ostrom 1986 5)Another critical issue in the institutional literature is the relationship

between institutions and organizations North (1990) following from hisdenition of institutions argues that institutions and organizations aredistinct entities Which organizations come into being and how theyevolve through time is inuenced by a societyrsquos rules and norms thatis by its lsquoinstitutionsrsquo On the other hand a number of recent organiza-tional sociologists see very little difference between institutions andorganizations For them rules and norms are institutions which unfoldin tandem with organizational structures and processes and changes inorganizational forms internalize and reect changes in the societyrsquos rulesand norms Using this perspective a whole subdiscipline within soci-ology called the lsquonew institutionalismrsquo in organizational analysis hasemerged The lsquoinstitutionalistrsquo perspective on organizations assumes thatthe kinds of organizations which actors create are dictated by the culturalnorms and rules in which they are embedded (Powell and DiMaggio1991 Zucker 1988 1991)The importance of this disagreement about institutions is obvious If

institutions are so critical for understanding our societies it is impor-tant that we come to some systematic consensus as to what institutionsare and how they inuence social actors and the organizations that theycreate If we cannot do so we risk talking past one another and losingthe opportunities for cumulative knowledge based on articulation ofwidely shared theoretical understandingWe need not only conceptual clarication as to what institutions are

but also greater consensus as to how to study institutions No scientic

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

599

eld can advance very far if the practitioners do not share a commonunderstanding of the key concepts used in their analysis (Ostrom 19864) But with our universities and academic associations so fragmentedinto different disciplines and into various subspecialties within disci-plines it is difcult to advance the theoretical agenda of institutionalanalysis within the academy Indeed the disciplinary fragmentation ofthe modern university is a major barrier to the theoretical advancementof the study of institutions and innovations as well as most other hybrid elds of research (Hollingsworth 1984 Hollingsworth andHollingsworth 2000) And it will be only as a result of effective commu-nication across diverse elds of knowledge that our study of institutionsand innovativeness will be effectively advanced

MULTIPLE LEVELS OF INSTITUTIONAL ANALYSISAND INSTITUTIONAL CHANGE

There are innumerable signs that we are living in a time of great insti-tutional change the demise of the Soviet empire the processes ofEuropean political and economic integration rapid transformation inparts of the global economy the disintegration of the family structurethe weakening of voluntary associations and the decline in politicalparticipation in a number of advanced capitalist societies the weakeningof welfare states European law is superseding national law and is evenchanging complete national legal systems The list could go on and onEven though scholars discuss institutional change at length their

ability to measure the rate of institutional change is very limited Andmore crucial than the limited ability of scholars to measure institutionalchange is their very limited understanding of how to build new insti-tutions One of the reasons for these shortcomings is that the socialsciences are decient in a theory of institutions The building of newinstitutions and redressing the decline of some of the most importantinstitutions of our societies are among the most important problems ofour time If we are to advance in the development of a theory of insti-tutions we need to work collaboratively across the social sciences andwe need to dene the parameters of institutional analysis

First level of analysis

This article attempts to make some modest contribution to outlining theparameters of institutional analysis At the outset we need to recognizethat when we engage in institutional analysis we must be sensitive to multiple levels of reality As suggested above most scholars whoengage in institutional analysis do not participate in any coordinatedactivity with each other and their activity is fragmented into a variety

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

600

of disciplines and subdisciplines To establish some coherence to the eldof institutional analysis we need a map of the eld so that those workingin one area can see where their research ts in relation to other areasand other practitioners on the mapTable 1 presents such a map with multiple levels at which institu-

tional analysis occurs Theoretically each of these areas on the map isinterrelated with each other level However the various areascompo-nents on the map are arranged in descending order of stability orpermanence Those components at higher levels of reality are morepermanent and durable while those at lower levels change more rapidlyWere scholars doing institutional analysis able to reach some consensus

about where their own work ts in relation to all other practitioners inthe eld there would be increased potential for all practitioners tocommunicate with each other By analogy once geneticists crystallo-graphers biochemists etc had a good understanding of how theirinvestigations were related to each area of molecular biology the eldquickly was able to make theoretical advances (Judson 1979)At the rst level there are the basic norms rules conventions habits and

values of a society These are the most fundamental properties of insti-tutions and are the most enduring and resistant to change Rules normsconventions etc are institutions but are only one component of whatconstitutes institutional analysis As Burns and Dietz (2001 forthcoming)point out most human activity is organized and regulated by normsand rules and systems of rules These concepts are extremely importantfor institutional analysis as they exert the greatest inuence on the natureof the components of institutional analysis at the next four levels which

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

601

Table 1 Components of institutional analysis

1 Institutions = norms rules conventions habits and values (see North1990 Burns and Flam 1987)

2 Institutional arrangements = markets states corporate hierarchies networksassociations communities (Hollingsworth and Lindberg 1985 Campbell et al 1991 Hollingsworth et al 1994 Hollingsworth and Boyer 1997)

3 Institutional sectors = nancial system system of education businesssystem system of research (Hollingsworth 1997)

4 Organizations (Powell and DiMaggio 1991)5 Outputs and performance = statutes administrative decisions the nature

quantity and the quality of industrial products (Hollingsworth 1991 1997)sectoral and societal performance (Hollingsworth and Streeck 1994Hollingsworth et al 1990 Hollingsworth and Hanneman 1982)

Note The ve components in this table are arranged in descending order of permanenceand stability That is norms conventions etc are more enduring and persistent than eachof the other components of institutional analysis Each component is interrelated withevery other component and changes in one are highly likely to have some effect in bringingabout change in each of the other components

are depicted in Table 1 In most forms of social analysis it is extremelyimportant that we understand the social and cognitive conditions thatlead to compliance or non-compliance with rules and the conditionswhich lead to changes in rulesThe approach to the study of institutions employed here argues that

norms rules habits conventions and values both reect and shape thepreferences of actors Norms rules habits conventions and values inu-ence who and what are included in different types of decision makinghow information is processed and structured what action is taken(Shepsle 1986 1989) It is through norms and rules that behavior isjudged to be democratic fair or egalitarianBurns and Flam (1987) point out that in any society there are multiple

rule systems Within a family there are rules for decision making oftenquite different from rules and norms for decision making for a professorin a classroom or for the customer in a bank Despite the heterogeneityof rule systems there are meta rules and norms which encompass lesserrule systems Otherwise there would be such contradictory rule systemsthat a society would be paralyzed The existence of meta rule systemspermits different rule systems to intersect with each other so that ambi-guities can be resolved The greater the pluralism and complexity of asociety the more ambiguity there is about meta rules and norms in asociety and of course all ambiguities never completely disappearOverall the degree to which separate rule systems are interlinked is anempirical problem There are different sectors groups and interestspursuing their own action logic but through higher-order meta princi-ples and rules there can be order consensus and coherence in a societyIt is through a set of meta rules that class and ethnic conict in societiesare contained (For elaboration see Burns and Flam 1987)In many respects our understanding about norms rules habits con-

ventions and values inuences our perspective on how societies are con-structed and how they change lsquoNew institutionalistsrsquo (Posner 1992Schotter 1981 Williamson 1975 1985) tend to assume that at one timethere was a state of nature and that there was a movement from individ-uals to institutions ndash an approach often called methodological individual-ism (Popper 1961 Hodgson 1998 1999) And of course there areinnumerable instances which methodological individualists cite to demon-strate that individuals create new rules of behavior For example it is pos-sible for actors to change the rules of driving so that instead of drivingon the left side of the road drivers adopt a new rule and drive on the rightThis article however tends to equate social habits and institutions

As Hodgson and others remind us (Hodgson 1988 1989 1997 1999Grafstein 1992 Camic 1986 Johnson 1992 Nelson and Winter 1982Veblen 1899) social habits are the results of earlier choices and are ameans of avoiding endless deliberation Because cognitive frameworks

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

602

are learned through habit individuals rely on the acquisition of suchcognitive habits before reason communication choice or action arepossibleWhereas Schotter (1981) and other game theorists take the individual

as an agent unencumbered by previous habits Field (1984) and othershave stressed that there can be no games without prior norms and rulesand thus a set of norms and rules must be presumed at the start Thosewho attempt to explain institutions from individual behavior alone areusing a bad strategy (Hodgson 1998)The position here ndash heavily inuenced by Hodgson (1998 1997 1988)

ndash is that individuals are embedded in a complex institutional environ-ment and that institutions not only constrain but also shape individuals(also see Hollingsworth and Boyer 1997) It would be a mistake howeverto get involved in an innite regress in order to determine which camerst ndash individuals or institutions Of course institutions are formed andchanged by individuals just as individuals are shaped and constrainedby institutions But at a macro level it is institutions that provide acognitive framework whereby individuals can cope with their reality Inthis sense the micro and macro worlds are intertwined At the macrolevel there is considerable stability but at the micro level individualshave a signicant level of autonomy and there can be widespread diver-sity As Hodgson (1988) reminds us most institutions in a temporal senseexist prior to the individuals in any given societyIt would be a serious mistake to downplay the importance of indi-

viduals and micro level analysis as we study institutions In the nalanalysis it is at the level of individuals that norms rules habits conven-tions and values exist An individual is born into and socialized intogroups and a society and this is how one early in life acquires a senseof appropriate forms of behavior Because of the way that individualsare socialized into a world of rules norms habits conventions andvalues it is unnecessary for individuals to restructure the world anewevery day (Douglas 1987 Elster 1989) Every action does not have tobe seriously reected upon For this reason institutions provide cogni-tive frameworks for individuals make their environments predictableprovide the information for coping with complex problems and envi-ronments In the words of Johnson (1992 26) lsquoInstitutions reduceuncertainty coordinate the use of knowledge mediate conicts andprovide incentive systems By serving these functions institutions providethe stability necessary for the reproduction of societyrsquo However eachsociety has different forms of habits rules and norms and hence differentincentives and disincentive systems for learning and forgetting forprocessing information But because individuals have varying degreesof autonomy individuals and groups can deviate from the prescribedforms of behavior in a society And of course these changes at the level

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

603

of individuals become important in understanding processes of societalchangeThese views are not meant to imply that the type of institutional

analysis proposed herein approximates a general theory of society Thisis clearly not the case However it is intended to represent the rst stepsin a mapping exercise of the boundaries of institutional analysis and tosuggest a few methodological insights for studying institutionsOne should try to see norms and rules as continuous and not as

dichotomous entities to recognize that they come in varying strengthsLegro (1997) has suggested that we assess the robustness of norms andrules with three criteria their simplicity their durability and their concor-dance Simplicity refers to how well actors understand norms and ruleshow well they can be applied within a specic situation Some normsand rules can be so complex that actors can have considerable difcultyin applying them in specic cases Durability addresses the issue of howlong norms and rules have been in effect ndash in short in order to assesstheir level of legitimacy While the position of this article is that normsrules and values are quite durable they do vary in this respectConcordance refers to how widely applied a norm or rule is Thisaddresses the degree to which a rule is a meta rule the degree to whichit incorporates the heterogeneity of other norms and rules In sum theclearer the norms and rules of a society the longer they have been inexistence and the more widely applicable they are the greater theirimpact on a society Hence the more robust the norms and rules thegreater their impact on a society and the less their robustness the greatertheir exibility and the less their effect on shaping a societyrsquos outcomesand performancesBecause norms rules and values are quite durable they play an impor-

tant role in shaping the history of societies thus contributing to a greatdeal of path-dependency Actors attempt to adjust to their contempo-rary environment but since they are products of the past the historicallegacy of norms rules and values inuences the decisions they makeAlthough actors have some capacity to alter the course of their historythey are constrained by their past and the degree to which they canmove beyond their past is limited As Lanzara (1998) Johnson (1992)and others have argued societal inertia is a basic feature of institutionsThey provide the basic stability necessary for change Degrees of historyare continuously reproduced by the way in which the inhabitants ofsocieties are socialized History matters but at critical points in historythere is punctuated equilibrium (Somit and Peterson 1992) During mostperiods of history there is considerable stability in the norms rules andvalues of a society but at critical moments norms rules and values canquickly and dramatically be redened At all times norms rules habitsconventions and values are inuencing each of the other components in

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

604

the institutional framework discussed below but these other institutionalcomponents feed back and can modify rules norms conventions etc(Murmann 1998)

Second level of analysis

The norms rules habits conventions and values of a society lead to thenext level of analysis ndash the institutional arrangements which are involvedin the coordination of various economic actors producers and suppliersof raw materials knowledge etc processors of raw materials infor-mation workers customers of raw materials nished productsinformation etc nanciers governmental and other types of regulatorsThese actors regularly engage in contests to resolve various economicproblems in virtually all sectors of society how are prices to be setWhat quantity of various products is to be produced How are stan-dards of various products processes etc to be set How is the qualityof products and processes to be determined How are various societalprocesses to be nanced In order to confront these problems and toaddress the conicting positions of economic actors as they address these problems societies develop various institutional (governance)arrangements for coordinating different actors These consist of marketsvarious types of hierarchies and networks associations the state commu-nities and clans (see Hollingsworth and Boyer 1997 Campbell et al1991 ch 1)Each of these particular kinds of coordinating mechanism has many

different types For example there are many types of states (eg theregulatory state developmental state authoritarian state welfare state)on which there is an extensive literature (Kim 1997) Similarly there aredifferent types of markets networks different kinds of associations etc(Boyer 1997 Hage and Alter 1997 Schneiberg and Hollingsworth 1990)When we do institutional analysis we must engage in congurativeanalysis recognizing that actors are not coordinated or governed by asingle type of institutional arrangement Some of the literature discussesindustrial sectors as though they were coordinated or governed by asingle institutional arrangement whereas in fact each sector of aneconomy is coordinated by a conguration of institutional arrangementsSome congurations coordinate actors in certain problem areas whileother congurations of institutional arrangements coordinate actors inaddressing other problems The types of congurations which are domi-nant in a society are somewhat stable and tend to persist over timewithin a society (Hodgson 1999)When one mode of coordination is dominant in a society it will

inuence the role which other coordinating modes will play Hence thestrong role of the state in the Soviet Union inuenced the role of markets

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

605

associations etc in the governance of the Soviet economy Similarly theprominence of particular modes of coordination in a society inuencesits style of innovationMuch of the literature on institutional arrangements (eg forms of eco-

nomic coordination) remains fragmented and unintegrated It is helpfulto array these various forms of coordination in a two-dimensional tax-onomy as in Figure 2 On the vertical dimension the economistrsquos viewof a self-interested agent is contrasted with a more sociological perspec-tive according to which social rules obligation and compliance shapehuman actions On the horizontal dimension there is another continuum(ie the distribution of power) At one extreme of the dimension onends many and relatively equal agents interacting (eg as in a well-organized spot market) At the other extreme inequality in power resultsin a hierarchical form of coordination which structures the interactionbetween principals and agents or between leaders and followers Wherea society falls along the distribution of power is inuenced by the rulesnorms values etc which are dominant in a society at a particular momentin timeInstitutional arrangements can be visualized on two dimensions the

nature of action motive on the one hand and the distribution of poweron the other Markets (cell 1) combine self-interest with horizontalcoordination transactors and they reect sensitivity to concerns aboutsupply and demand thus providing ex post an unintended equilibriumParadoxically the more pure and perfect the market competition thegreater the need for codied rules of the games for coordinating economictransactions Thus collective associations (cell 6) andor various formsof state intervention (cell 4) are required to develop and enforce rulesfor transacting partners (Schneiberg and Hollingsworth 1990 Streeckand Schmitter 1985a) This is an example of how the norms and rulesand the institutional arrangements of a society are intertwined asreected in a conguration of institutional or governance arrangementsAlong the horizontal axis actors can be joined in an organization or

a rm a hierarchy is the generic terms for this institutional arrangement(cell 2) Along the horizontal line one recognizes the difference betweentransactions in a market and transactions within a rm The well-knownworks of Coase (1960 1981) and Williamson (1975 1985) utilize theconcept of transaction costs in explaining the emergence of corporatehierarchiesThere are also various types of networks (cell 5) which exhibit mixes

of self-interest and social obligation with some actors being formallyindependent and equal Yet in some networks (the large rms and theirsubcontractors) there is unequal power and initiative Networks maycomprise all kinds of actors some consist only of rms but others includeassociations and the state (Hage and Alter 1997)

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

606

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

607

DISTRIBUTION OF POWER

VERTICALHORIZONTAL

1 MARKETS 2 HIERARCHY

6ASSOCIATIONS

5NETWORKS

4STATE

3COMMUNITIES

SELF INTEREST

SOCIAL OBLIGATION

ACTION MOTIVE

Figure 2 A general taxonomy of institutional arrangements

The vertical axis deals with action motives Toward the upper part ofFigure 2 actors are engaged in individualistically oriented behaviorwhereas toward the lower part actors are more engaged in collectivebehavior and strive to cope with problems of common interest Cell 3 ndashcommunities and clans ndash consists of institutional arrangements based ontrust reciprocity or obligation and thus not derived from the pure selshcomputation of pleasures and pains This is an unconventional form ofcoordination for most neoclassical economists (however see Arrow1974) but not for many anthropologists political scientists and sociolo-gists (Streeck and Schmitter 1985a Polanyi 1944 Gambetta 1988Fukuyama 1995 Sabel 1992 Putnam 1993)In the neoclassical paradigm theorists argue that actors engage in

forms of exchanges that best promote their individual interests If somestructural conditions are fullled (absence of increasing returns to scalethe reversibility of transactions absence of uncertainty and completecontingent markets with no collusion between actors) then the invis-ible hand theorem applies and market-type activity functions quite welland also provides the optimum for society thereby combining efciencyharmony and order However an excess of market activity may welllead to ruinous competition and excessive conict There is variation inthe extent to which ruinous competition occurs depending on the socialcontext in which transactions take place Thus it is important that webe sensitive to the institutional context in which transactions areembedded and that we understand the degree to which social bondsexist at both the micro and macro levels of analysis Micro bonds facil-itate exchanges in a society but at the societal level social bonds existat the level of the collective ndash in the community or region and amongmembers of racial religious and ethnic groups All other things beingequal the more powerful the social bonds among transacting partnersthe more economic competition is likely to be restrained Thus mosttransactions occur not simply in an impersonal calculative system ofautonomous actors unrestrained by social ties ndash as implied by theneoclassical paradigm ndash but in the context of social ties variation in the strength of which leads to variation in levels of trust and transac-tion costs (Etzioni 1988 211 Granovetter 1985 Streeck and Schmitter1985a Hollingsworth and Boyer 1997 Hodgson 1988 1999 Schneibergand Hollingsworth 1990)Another form of multilateral institutional arrangement is various

types of associations (cell 6) Unlike networks clans and communitiesassociations are more formal organizations Whereas markets corporatehierarchies and networks tend to coordinate economic activity amongdifferent types of actors (eg producers with suppliers capital withlabor) associations typically coordinate actors engaged in the same orsimilar kinds of activities Business associations and labor unions are

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

608

some of the most common forms of associations for coordinatingeconomic activity in capitalist economies (Schneiberg and Hollingsworth1990 Streeck and Schmitter 1985a)Finally there is the state (cell 4) which is an institutional arrangement

quite unlike any of the others It is the state that sanctions and regu-lates the various non-state coordinating mechanisms that is the ultimateenforcer of rules of the various mechanisms that denes and enforcesproperty rights and that manipulates scal and monetary policy Overallit is the state that inuences the total incentive system of a society(Lindberg and Campbell 1991 Johnson 1992 40) At the same time thestate may also be an economic actor by engaging directly in productionand exchange relationsThe choices of institutional arrangements in Figure 2 are constrained

by the social context in which they are embedded Depending on thenature of that embeddedness there is variation in the collective formsof governance some of which are specied in the lower part of thegure Some modes of coordination in the bottom and upper levels ofthe typology are often mixed together though one coordinating modeis likely to be more prominent than another But actors are often engagedin complex congurations involving several kinds of institutionalarrangementsEach of these various institutional arrangements has its own logic ndash

its own rules its own procedures for enforcing compliance its ownnorms and ideologies which help to reduce the costs of enforcementThese are summarized in Table 2 which provides further elaborationabout the various coordinating mechanisms one nds in almost everycapitalist society Table 2 lists the organizational arrangements and theirstructures rules of exchange and means of enforcing compliance asso-ciated with each type of coordinating mechanism While each type ofinstitutional arrangement has various positive features each institutionalarrangement also has failures and these are featured in Table 3 It is thecontest between those who support and those who oppose these variousinstitutional arrangements that tends to lead to transformations in insti-tutional arrangements over time (Campbell et al 1991 Campbell 1997)Again it is difcult to conceive of pure institutional arrangements ndash

either exclusively markets or exclusively hierarchies ndash since it has beenwell known since Adam Smithrsquos Wealth of Nations that the division oflabor within the rm cannot be disentangled from the existence andextent of the market Institutional arrangements have their own distinc-tive form of efciency (good static efciency for the market dynamicefciency for rms) and inefciency often leading to considerableinequality Neither networks nor communities are panaceas for economiccoordination without being congured with other types of institutionalarrangements Networks and communities may solve certain issues but

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

609

Table 2

Log

ics of ins

titution

al arran

gemen

ts

Coordination

Organizational

Rules of

Individual means

Collective means

mechanisms

structure

exchange

of compliance

of compliance

Marke

tsEasy en

try an

d exit

Voluntary sp

ot excha

nge

Leg

al enforcemen

t of

Norm of pr

ivate

Bila

teral ex

chan

ge or

control

prope

rty

marke

t site (W

all Street)

Reg

ulations

to

Leg

itim

acy of m

arke

t en

force contracts

men

tality

Com

mun

ities

Inform

al m

embe

rship

Voluntary exch

ange

based

So

cial norms an

d m

oral

Highly institutiona

lized

ev

olve

s ov

er lon

g period

on social solid

arity an

d

princ

iples im

pose

norm

s an

d rules requ

ire

of tim

ehigh

deg

ree of tru

stob

ligations

mem

bers to accept

Kno

wledg

e of others an

d

lsquocorpo

ratersquoo

bligations

recipr

ocity ov

er tim

e

Networks

Semi-form

al m

embe

rship

Voluntary exch

ange

ove

r a

Con

tractual bon

dsPersona

l relation

sBila

teral or m

ultilateral

time pe

riod

Resou

rce de

pend

ence

Tru

st built outside

the

ex

chan

geecon

omic arena

Associatio

nsFo

rmal m

embe

rship

Restricted to mem

bers

Self-interest

Some deg

ree of

Multilateral excha

nge

Emph

asis on insider

Rep

utation

effects

compu

lsion

outsider or we

they

Priva

te interest type

of

men

tality

gove

rnan

ce

Private

Com

plex organ

izations

Restricted to mem

bers

Rew

ards to ind

ividuals

Highly institutiona

lized

hierarch

ies

which

ten

d to becom

e ex

chan

ge based

on

Asymmetric po

wer threat

rules

bureau

cratic hierarchies

asym

metric po

wer

of san

ctions

Mem

bers soc

ializ

ed into

bureau

cratic rules

corp

orate cu

lture us

e of

sanc

tions

State

Pub

lic hierarchy

Unilateral action

Exit vo

ice (vote lob

bying)

Coe

rcion

De jure

and impo

sed

Indirect political an

d

loya

ltyNorms an

d pu

blic rules

mem

bership

econ

omic excha

nge

Table 3

Failu

res of coo

rdinating mecha

nism

s

Coordinating mechanisms

Type of failure

Associations

Private hierarchies

State

Market

Communities

Networks

Enforcem

ent

Usu

ally relies on

Might enh

ance

Needs co

ntrol

Needs an

internal

Needs

tru

st and

Need an external

the state as an

oppo

rtun

istic

extern

al to state

enforcem

ent

loya

lty often

en

forcem

ent

enforcer

beha

vior

bureau

cracy

authority

coming from

au

thority

Resem

bles

The

ideal o

f (ju

dge

s

Facilitates

outside (fam

ily May

facilitate

enforcem

ent

internal m

arke

ts

parliamen

t collu

sion

and

relig

ion

cartelization an

d mecha

nism

might hur

t marke

t) to

impe

rfect

ethn

icity)

mon

opoly

of cartels

incu

mbe

nt

correct state

compe

titio

nCom

patible with

worke

rsab

uses

variou

s types of

Lob

bies can

co

mpe

tition

captur

e pu

blic

interest goa

ls

Public goo

d

Useful for

Gov

erna

nce co

sts

Can

provide

Can

not provide

Can

internalize

Useful for

and

establishing

might exceed

public goo

ds

colle

ctive go

ods

some co

llective

enha

ncing qu

ality

externality

stan

dards an

d

the be

nets of

but ha

s or deal w

ith

good

s (qua

lity

and training bu

t qu

ality for

internal division

difc

ultie

s in

externalities

training

) bu

t no

t no

t ve

ry goo

d in

setting ru

les of

of la

bor

prov

iding them

Inad

equate

othe

rs (welfare

prov

iding for

compe

titio

n in

Slow

to react to

in precise

mon

itoring of

gene

ral p

ublic

societal gen

eral

the indu

stry

chan

ges in the

am

ounts

tech

nical ch

ange

go

ods)

welfare

Useful for

environm

ent

Might fail in

and inno

vation

Mem

bers tightly

Weak in the

prov

iding

induc

ing

integrated

into

prov

ision of

man

y go

ods

tech

nical

community

colle

ctive go

als

colle

ctively

chan

geha

ve limite

d

that in

dividua

lcapa

city for

mem

bers can

not

inno

vatio

nsprov

ide for

them

selves

Table 3

continue

d

Coordinating mechanisms

Type of failure

Associations

Private hierarchies

State

Market

Communities

Networks

Efcien

cyFa

cilitates

De

cien

t in

Can

be high

ly

Some ba

sic social So

me go

ods

Slow

to en

hanc

e co

operation an

d

coop

eration an

d

bureau

cratic

relation

s cann

ot

cann

ot be

efc

ienc

y an

d

X-efcien

cy but

X-efcien

cyan

d can

not

be provided

by

delivered

at

speed of

not allocative

easily deliver

pure m

arke

t su

fcien

tly

adap

tive

ness

efcien

cy

good

s at

mecha

nism

slow costs

except in

low cost

indus

tries whe

re

tech

nology

is

complex an

d rapidly cha

nging

Equity

Narrow

Excessive

Might enh

ance

Facilitates

Might lead to

Whe

n widely

encompa

ssing

multiplication

ineq

uality

ineq

uality in

retard

ed

dev

elop

ed in

to

associationa

l of con

trollers

(pow

er

income an

d dev

elop

men

tindus

trial

stru

ctures lead

(fru

stration

and

an

d privilege

)wealth

districts

to in

come

ineq

uality)

netw

orks m

ay

equa

lity

facilitate greater

equality an

d

inco

me

distribution

whe

n wea

kly

dev

elop

ed

netw

orks ten

d to

increase soc

ial

ineq

uality

Note For allo

cativ

e efc

ienc

y an

d X

-efcien

cy see Leibe

nstein (19

66 19

76)

they raise other no less severe problems It is important to recognizethe imperfection of any single institutional arrangement in order tocomprehend the origin and transformation of any other institutionalarrangement (Hodgson 1999 ch 3)As suggested above each institutional arrangement is congured with

other institutional arrangements Usually one particular institutionalarrangement is more dominant in a particular conguration but becauseeach type of institutional arrangement has its own strengths andweaknesses there is no simple structural logic in the governance or coor-dination of a society Each institutional arrangement constrains the otherbut the inherent tension among the various institutional arrangementswithin a conguration contributes to changes in the governance orcoordination of a society The routines and logics of each of these cong-urations of institutional arrangements provide constraints and incentivesfor actors The way that actors perceive the incentives and constraintsof these governance congurations leads to a particular market logicand it is the specic market logic of a society which inuences its speciccapacities and weaknesses The inherent strain among the different logicsin a conguration of coordination helps to provide the exibility for asociety to adapt to new circumstancesIf there were a society with pure markets or a society coordinated

only by the state there would be too much rigidity and too little diver-sity to cope with the vast uncertainty in the global environment A societywith very little diversity in its coordination mechanisms would havelittle capacity to adapt to new circumstances Congurations with consid-erable diversity of institutional arrangements provide for a certainamount of incoherence in governance but they also provide for thecapacity to adapt to new circumstances While the Soviet Union wasdominated by the state as a coordinating mechanism there were alwaysfunctioning markets in the system Moreover the feudal society of thetwelfth and thirteenth centuries consisted not only of hierarchical rela-tionships between serfs and masters but also of urban guilds clericalhierarchies and markets (Johnson 1992 38) In sum the robustness ofinstitutions often depends on multiple and diverse principles and logicsof action on the inconsistency of principles and procedures on patternedforms of disorder (Lanzara 1998 Orren and Skowronek 1991 320 329)

Third level of analysis

The next level consists of the institutional sectors of a society The rulesnorms and values of a society inuence the array of institutional arrange-ments and levels one and two inuence the nature of and the relation-ships among various institutional sectors and all three levels areintricately linked together to form a social system of production Together

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

613

all of these components inuence the performance of economic sectorswithin a society as well as the performance of the total society A socialsystem of production is the way that a societyrsquos institutions (see rstlevel of analysis above) its institutional arrangements (see second levelof analysis above) and its institutional sectors are integrated into a socialcongurationAn institutional sector includes all organizations in a society which

supply a given service or product along with their associated focal orga-nizations (eg major suppliers funding sources regulators and so forth(Scott et al 1994 108 117 Campbell et al 1991 Hollingsworth et al1994) Institutional sectors include but are not limited to the societyrsquossystem of education system of research business system nancialmarkets legal system and the state The structure of the nancial marketssystem of training and education industrial relations system system ofresearch and state and legal systems are distinct and idiosyncratic ineach society In short they are system specicAll of these institutional sectors tend to cohere with each other

although they vary in the degree to which they are tightly coupled intoa full-edged system While each of the institutional sectors has someautonomy and may have some goals that are contradictory to the goalsof other institutional sectors with which it is linked an institutional logicin each society leads institutions to coalesce into a complex social con-guration (Hollingsworth 1991) This occurs because the institutionalsectors are embedded in a culture in which their logics are symbolicallygrounded organizationally structured technically and materially con-strained and politically defended The conguration of institutionalsectors usually exhibits some degree of adaptability to new challengesbut continues to evolve within an existing style But under new circum-stances or unprecedented disturbances these institutional congurationsare exposed to sharp historical limits as to what they may or may notdo (Schumpeter 1983 David 1988 Arthur 1988a 1988b Haringkanssen andLundgren 1997)Why do all of these different institutions coalesce into a complex

social conguration which is labeled here as a social system of pro-duction The literature suggests two contrasting interpretations Part ofthe answer ndash indeed a controversial one ndash is that these institutionalsectors are functionally determined by the requirements of the practiceof capitalism in each time and place (Habermas 1975) Another expla-nation emphasizes the genesis of the actual conguration via a trial and error process according to which the survival of rms regions orcountries is the outcome of complex evolutionary mechanisms (Maynard-Smith 1982 Nelson and Winter 1982) However the problem is evenmore complex Markets and other mechanisms for coordinating rela-tionships among economic actors place constraints on the means and

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

614

ends of economic activity to be achieved in any society The othercoordinating mechanisms include different kinds of hierarchies varioustypes of networks and associations ndash trade unions employers and busi-ness associations (Hollingsworth and Lindberg 1985 Campbell et al1991) The logic of these various institutional arrangements providesactors with vocabularies and logics for pursuing their goals for deningwhat is valued and for shaping the norms and rules by which theyabide (see Tables 2 and 3) In short in contrast to the logic of theneoclassical paradigm the argument here is that the dominant type ofinstitutional arrangements places severe constraints on the denition of needs preferences and choices of economic actors Whereas the neo-classical paradigm assumes that individuals and rms are sovereign thisarticle is based on the assumption that rms are inuenced by the hold that the institutional congurations making up a social system ofproduction have on individual decision making (Campbell et al 1991Etzioni 1988 Streeck and Schmitter 1985a Hollingsworth et al 1994Hollingsworth and Boyer 1997 Magnusson and Ottosson 1997 North1990 Hodgson 1999)Standard neoclassical economic theory has tended to downplay the

role of production and consequently of rms Even transaction cost theo-rists who are concerned with analyzing the rm as a coordinatingmechanism have been relatively unconcerned with the various compo-nents of a social system of production Indeed as long as there waswidespread optimism about the efcacy of Keynesian economics therewas relatively little concern among neoclassical economists with thesupply side of the economy Even in the opinion of most Keynesians agroup of experts should ideally be able to shape the size of aggregatedemand while the supply side of the economy would be left to the twominimalist institutions of neoclassical economics ndash markets and manage-rial hierarchies For several decades however it has become increasinglyobvious that some of the most competitive and successful patterns ofindustrial output and industrial production in capitalist economies donot derive from the neoclassical prescription of unregulated markets andcorporate hierarchies complemented by a neoliberal democratic stateIndeed it is now well understood that certain highly successful produc-tion patterns require for their emergence and survival institutionalarrangements and environments the very opposite of the prescriptionsfound in the neoclassical paradigm (see especially Streeck 1992 but alsoHollingsworth and Streeck 1994) Thus social scientists have come tounderstand that if they are to comprehend the behavior and performanceof contemporary economies concerns about social systems of produc-tion must be brought into the pictureBecause production involves more than technology a number of social

scientists have an increasing concern with social systems of production

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

615

The same equipment is frequently operated quite differently in the samesectors in different countries even when rms are competing in the same market (Maurice et al 1980 Sorge and Streeck 1988 Sorge 1989Hollingsworth et al 1994) Variations in production and process tech-nologies are inuenced partly by variations in the social environmentsin which they are embedded In other words rms are embedded intocomplex environments which among other things place constraints ontheir behavior Thus a social system of production is of major impor-tance in understanding the behavior and performance of an economyHow the state and other institutional arrangements (eg marketsnetworks private hierarchies associations) coalesce and are related toparticular social systems of production are important determinants ofeconomic performanceThe congurations of institutional arrangements that coordinate or

govern the behavior of actors in one society and its structure of specicinstitutional sectors cannot easily be transferred to another society forthey are embedded into a social system of production that is societallydistinct (Hollingsworth 1997) Societies borrow selected principles offoreign management styles and work practices but the effectiveness ofsuch borrowing is generally limited Economic behavior and performanceare shaped by the entire social system of production in which actors areembedded and not simply by specic principles of particular manage-ment styles and work practices Moreover a societyrsquos social system ofproduction tends to limit the kind of goods that it can produce and withwhich it can compete successfully in international marketsSome scholars (Kenney and Florida 1993 Oliver and Wilkinson 1988)

have assumed that the diffusion of particular forms of management stylesand work practices across societies could lead to system convergenceReferring to the Japanese production system Kenney and Florida haveargued that it consists of organizational practices whose fundamentallsquogenetic codersquo can successfully be inserted into another society and beginsuccessfully to reproduce its behavior in the new environment (1993 8)Their position is in the intellectual tradition of Antonio Gramsci (1971)who contended decades ago that the American system of mass produc-tion would most certainly diffuse to Europe over timeHowever the argument of this article is that even though British

French and American rms may adopt certain aspects of Japanesemanagement styles (eg just-in-time production self-managing teamsquality circles the use of statistical process controls) or some variant ofthe German vocational system their social systems of production willnot converge Moreover the overall conguration of a societyrsquos socialsystem of production inuences its sectoral and overall national economicperformance This helps to explain why societies have different systemsof innovation and why they vary in the clusters of industries in which

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

616

they are highly competitive in the international markets (Nelson 19931999 Hage and Hollingsworth 2000 Berger and Dore 1996)In order to understand how and why a societyrsquos economy performs

as it does it is necessary to understand its entire social system of produc-tion If a society is to modify substantially the performance of itsinstitutional sectors and its entire economy it cannot adapt only someof the management and work practices of its foreign competitors Ratherit must alter its entire social system of production Because a societyrsquosmodes of economic governance and coordination and its institutionalsectors develop according to particular logics and are system specicthere are serious limitations to the extent to which a society may mimicthe institutions (eg the rules and norms) institutional arrangements andinstitutional sectors of other societiesIn the history of modern societies there are logics by which institu-

tions coalesce into a social system of production (Hollingsworth 1991)Though institutions are constantly changing there are sharp limits tothe type and the direction of change that any particular institution canundergo because of its linkages with institutional arrangements and insti-tutional sectors Thus a societyrsquos business rms educational systemnancial markets industrial relations system etc can engage in seriousrestructuring only if most of the other institutional sectors also changeIn social systems of production there are pressures toward consistencyin the norms rules and values across institutional sectors though in anycomplex society social systems are obviously imperfectly integratedIndeed the degree to which the institutional norms and rules makingup a social system of production are loosely or tightly coupled is a vari-able of considerable importance In general the institutional sectorsmaking up a social system of production are interdependent and changesin one generally result in changes in the others Since each institutionallevel is dependent on the others for various types of resources there isinterdependence among the differing institutional spheres Moreovereach society has its norms moral principles rules and laws and recipesfor action as well as its own idiosyncratic customs traditions and prin-ciples of justice (Burns and Flam 1987)There are also other inherent obstacles to convergence among social

systems of production for where a system is at any one point in timeis inuenced by its initial state Systems with quite different initial statesare unlikely to converge with one anotherrsquos institutional practicesExisting institutional arrangements block certain institutional innova-tions and facilitate others (Roland 1990) Thus the institutions makingup a social system of production provide continuity even though insti-tutional arrangements are always changing but with a particular logicWhile Williamson (1975 1985) suggests that actors tend to choose theinstitutional arrangements which are most efcient North (1990) is

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

617

much closer to the mark in his argument that most societal institutionalarrangements exist as a result of custom and habit and are thereforeinefcient At any moment in time the world often appears to its actorsas very complex and uncertain For this reason actors often engage incontradictory forms of behavior pursuing different strategies as hedgesagainst a very uncertain world (Lanzara 1998) And their hedging andcontradictory forms of behavior may lead in somewhat different soci-etal directions all constrained by the institutional fabric within whichthe actors are embedded This kind of contradictory behavior occurs inpart because of the contradictions inherent in the dominant congura-tions of institutional arrangements with their contradictory logics (seelsquoSecond level of analysisrsquo above)Despite the emphasis on the logic of institutional continuity this is

not an argument that systems change along some predetermined pathThere are critical turning points in the history of highly industrializedsocieties but with the choices limited by the existing institutional terrainBeing path-dependent a social system of production continues along aparticular logic until or unless a fundamental societal crisis intervenes(Milgrom et al 1991 Krugman 1991 Durlauf 1991 Hollingsworth 1991Pred 1966 David 1988 Hodgson 1998)At this point it is important to confront the question of how social

systems of production evolve And this question gets to the heart of theproblem of building complex societal structures Certainly social systemsof production did not emerge from some process of social engineeringMoreover the various component parts of each social system of produc-tion have often not been designed to be part of a social system ofproduction The component parts of each social system of productionhave emerged more often than not as unintended by-products of goalswhich various actors had in mind at earlier moments in time It is usuallythe case that actors at time lsquotrsquo fail to comprehend the long-term conse-quences of their actions Wolfgang Streeck the author of several brilliantpapers on social systems of production argues that institutional sectordesigns created for one purpose generally address that goal but overtime those designs have quite unintended consequences (Streeck 1997a1997b Wright 1998) Describing the conguration of the German socialsystem of production of the 1970s and 1980s Streeck points out that itwas the unintended by-product of multiple points in history Someelements were pre-Wilhelmian others were introduced by the Alliedpowers after 1945 and others emerged during the years of the GermanFederal Republic All component parts of the German social system ofproduction

were and continue to be changing for their own reasons as wellas in reaction to each other and certainly there can be no presump-tion of a pre-established t between them even though one might

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

618

want to allow for some reinforcement effects of the modelrsquos histor-ically contingent social and economic success

(Streeck 1997b 54)

In sum the emergence of social systems of production is a long-termevolutionary process with each part interacting with its environment andresulting in a congurative whole but with no previous design by eithera single actor or a collectivity of actors As suggested above actors do not recreate their world anew At any moment in history the totalinstitutional complex of a society necessarily contains the resources and legacies of its past Preexisting institutional complexes are nevercompletely wiped out Societies consist of multiple layers of history withtheir diverse logics of action However the cumulative effect of smallperipheral changes in altering particular institutional sectors and insti-tutional arrangements can be substantial (Murmann 1998 Lanzara 1998)lsquoInstitution building is affected more by the ways in which people havecodied the past than by how they envision the future There are manymore potential resources in pre-existing institutional arrangements thanit is usually assumed or suspectedrsquo (Lanzara 1998 30)

Fourth level of analysis

The next level of institutional analysis organizational structures is some-what more controversial As suggested above Douglass North (1990)draws a sharp distinction between institutions and organizations Morerecently many organizational theorists (Powell 1991 Powell andDiMaggio 1991 DiMaggio and Powell 1983 Baum and Oliver 1992Kondra and Hinings 1998 Townley 1997) argue that institutional rulesnorms and conventions unfold in tandem with organizational structuresand this is my positionThe literature which focuses on how institutions inuence organiza-

tions is quite different from two other theoretical literatures which alsoare concerned with how organizational environments shape the behaviorof organizations There are resource dependency theorists (Pfeffer andSalancik 1978) who emphasize the role of environmental resources inshaping organizations while Hannan and Freeman (1977 1984) and otherpopulation ecologists emphasize the survival of organizations givencertain kinds of organizational conditions (Orru et al 1991)The institutionalist perspective as implied above emphasizes the

normative environment in which organizations are embedded It is a perspective which focuses on the way in which organizations in their behavior tend to conform to the institutional rules and normswhich are dominant in the organizational environment However allthree perspectives ndash the institutionalist the population ecology and

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

619

the resource dependency ndash emphasize how the environment inuencesorganizations and how organizations which are subject to the sameenvironment tend to converge in their behavior to have what DiMaggioand Powell have labeled organizational isomorphism (DiMaggio andPowell 1983)There has emerged a vast literature which demonstrates that within

each society there are scal political judicial and other regulatory normswhich limit and shape the culture and structure of organizationalbehavior The normative institutional environment of organizations limits the options of what organizations do in a particular society and inuences the patterns of ownership relations with suppliers andcustomers In short it is the normative environment of organizationswhich denes within a particular society what is socially acceptablebehavior for organizations (Hamilton and Biggart 1988 Hollingsworthand Hollingsworth 2000 Meyer and Rowan 1991 Zucker 1987 1988Orru et al 1991 Townley 1997)Thus far most of the analyses on isomorphic pressures operating on

organizations have been cross-sectional in nature However the entireconcept of isomorphism implies that there are strong environmental pres-sures exerted on organizations and in order to observe this phenomenonit is important that we have longitudinal studies which assess howchanges in an organizationrsquos institutional environment inuence changesin the structure and culture of organizations Moreover the expectationhere is that when one analyzes the patterns of organizational changeover time the isomorphism will be holistic ndash that is within a particularsociety there will be a tendency for the internal structure of particulartypes of organizations to converge in large part because of the pres-sures to conform to the changes in the external norms rules and valuesA close reading of Aoki (1990) and Slack and Hinings (1994) nds thatchanges in the national institutional environment of organizations haveinuenced changes in the internal structure of organizations within aspecic industry of a particular societyHowever it is inappropriate to view organizations as changing only

in relation to exogenous environmental change There is a kind of co-evolution which occurs between organizations and their institutionalenvironments Co-evolution implies nonlinear feedback between orga-nizations and their institutional environments This is of course quitedifferent from the more common way of studying organizations bymodeling relations between independent and dependent variables Ofcourse the more common type of analysis is quite appropriate whenthere are simple relationships which do not involve complex feedbackprocesses But in the strategy for institutional analysis which is proposedherein it is less useful to separate independent from dependent vari-ables and more useful to understand the interacting and co-evolutionary

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

620

processes among all of the levels of institutional analysis discussed inthe various sections above (Baum and Singh 1994a 379ndash80 Mowery andNelson 1999 Murmann 1998)The logic of the perspective herein suggests that within every

society there is variation in the structure and culture of business rmsuniversities and other complex organizations (Kondra and Hinings1998) However this variation takes place within parameters which aresystem specic For example every German rm and every Germanuniversity is different from every other German university and rm but there are a set of cultural and structural characteristics which distin-guish German rms and universities from those in America Similarlyeach Japanese rm and university is unique yet they have a set ofcultural and structural characteristics which are system specic andwhich differentiate them from business rms and universities in other countriesIn all societies each organization has its own distinctive organizational

rules norms and conventions which are subordinate to the meta norms and rules of the larger society within which they are embeddedHowever the strength of the institutional environment within whichorganizations are embedded varies from society to society Some societies have multiple institutional environments and there is hetero-geneity in terms of what constitutes appropriate behavior oforganizations In such societies there can be conict over whatinstitutional logics should regulate specic organizational functions(Friedland and Alford 1991 Townley 1997 262ndash3) In those societies inwhich the institutional norms habits and rules are most developed and in which the institutional pressures to conform are greatest thereis less variation in the structure and culture of business rms and variouskinds of research organizations In such societies the connectednessbetween research organizations and their institutional environments is sufciently strong for organizations to have low autonomy to pursue independent strategies and goals in these societies there is a great deal of organizational isomorphism Conversely the weaker the institutional environment in which research organizations areembedded the greater the variation in the structure and culture of business rms and research organizations Moreover where the institu-tional environments are more weakly developed organizations havegreater autonomy and exibility to respond to the development of new knowledge and to be highly innovative Hence it is in those societies where the institutional environments are most developed and most rigid and when there is less organizational autonomy and exibility that fewer radical innovations in basic and applied science as well as in totally new products and industrial sectors havebeen made

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

621

Fifth level of analysis

We come next to the outputs and performance of the various institutionalcomponents of a society It is at this level that institutional componentsare more pragmatic and exible For example in the legal sector thereare specic statutes and court rulings in the state sector there are specicpolicies in the business sector there are new products new technolo-gies and market strategies It is at this nal level of analysis thatinstitutional spheres are most open and susceptible to change and cross-national mimicry is easiest and most commonThrough the outputs of the society we can obtain some assessment as

to how well a society is performing Moreover we can assess how inno-vative it is how egalitarian it is in the distribution of its resources howegalitarian it is in terms of levels of health education etc Just as it iscomplicated to measure how egalitarian a society is similarly it is dif-cult to assess how innovative it is For example societies vary in makingincremental and radical breakthroughs in basic and applied science in developing totally new products and new kinds of organizations inincremental and process innovations in existing products and organiza-tions in developing new and different forms of marketing ndash bothdomestically and globally (Hage and Hollingsworth 2000) Of coursesocieties vary in their rates of economic growth in their rates of economicproductivity in their quality of life (eg rates of crime life expectancyetc) All of these performance criteria feed back and inuence each ofthe levels of institutional analysis discussed above ndash rules norms valuesetc institutional arrangements institutional sectors the structure andform of organizations Moreover performance measures may inuenceeach other For example if a society has very low rates of growth itmay not be very innovative in some of the types of innovative activitiesmentioned above Good or poor performances of certain kinds inuenceother performance indicators (Hollingsworth et al 1990)Different institutional arrangements and different social systems of

production result in different types of economic performance Hence solong as societies have different social systems of production there areserious constraints on the degree to which they can converge in theirinnovative styles Different social systems of production tend to maxi-mize in a more or less explicit manner different performance criteriausually mixing considerations about static and dynamic efciency protsecurity social peace and economic andor political power In short incontrast to the implications of neoclassical economic theory in real-worldeconomies there are no universal standards all economically rationalactors attempt to maximize Economic history provides numerous exam-ples of how a great variety of principles of rationality are implementedin different societies

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

622

A critical question in institutional analysis is whether an existing socialsystem of production which supports a set of routines for a particularkind of technology and industry can shift from old practices to newones There are numerous historical accounts of how the social systemof production in a particular society which worked so well for a numberof years could not adapt to new technologies (Lazonick 1990 Veblen1915 Nelson 1994) Schumpeter (1983) and Freeman (1991) have devel-oped the idea of a techno-economic paradigm which suggests thatdifferent technologies require specic forms of organizational arrange-ments ndash or what I called social systems of production Because oflsquoinstitutional inertiarsquo the social system of production of a society maynot be able to adapt to a new techno-economic paradigm (Nelson 199458 Lanzara 1998)Whether a social system of production can sustain its performance

standards depends not only on its intrinsic economic rationality but alsoon where it ts into a larger system If a particular social system ofproduction is immune from the innovativeness and competition of analternative system survival can be long lasting But if different socialsystems of production with diverging criteria of good economic perfor-mance meet in the world arena the arbitrariness of nationally imposedconstructed performance standards may be superseded by alternativeperformance criteria as a result of international competitivenessThe world economy is also socially constructed just as are national

economies Even if different social systems of production are competingin the international arena it is not always possible to determine whichis more competitively effective at any moment in time Hegemonicnation-states in the short run shape the rules of trade to favor their indus-trial sectors and rms But the history of hegemonic powers suggeststhat in the longer run such social systems of production sustainedlargely by military and political power eventually give way to moreinnovative and competitive social systems of production In our ownday as nation-states are increasingly integrated into a world economyeven hegemonic powers lose their innovative and competitive advan-tage and their share of world output decreases Such a country mayattempt to restructure its institutional arrangements and to readjust itsperformance preferences But to restructure its system of innovationgenerally calls for a major redistribution of power within a societyLargely for this reason societies have historically had limited capacityto reconstruct a system of innovation in the image of their major competi-tors Each countryrsquos social system of production is a conguration of ahost of norms rules and values as well as of institutional arrangementsEach system is constantly changing and is open to inuence from othersystems And indeed many technologies and practices diffuse from onesociety to another but a societyrsquos capacity to be innovative is constrained

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

623

by the existing social system of production Thus the same technologymay exist in numerous countries but how it is employed and how itinuences societal outcomes and performances varies from one institu-tional conguration or society to another (Hollingsworth 1998)

PERSPECTIVES ON INSTITUTIONAL CHANGE

A fundamental problem which remains unresolved in institutionalanalysis involves the nature of institutional change There is much confu-sion and miscommunication in the social sciences about institutionalchange in large part because we do not even know how to measure therate of institutional change Of course we know that there are bothexternal and internal focuses for change and we have had a good bitof scholarship on this subject What is not clearly understood is thatinstitutions and institutional arrangements within societies are histori-cally rooted that there is a great deal of path-dependency to the waythat institutions evolve Moreover the more intricately linked each sectoris with each other sector and with a societyrsquos rules and norms the lesschoice actors have to devise new institutions and institutional arrange-ments On the other hand if a sector or actor is too isolated it may betoo weak within a society to bring about any kind of effective changeThese considerations are very critical to the problem of how muchfreedom actors have to create new institutions afresh to what extentinstitutions diffuse from one society to another (Murmann 1998 Lanzara1998) This is a fundamental question for students of institutionalanalysis as we attempt to understand how the different institutionallevels are linked togetherThere are a variety of reasons why there is confusion in studying insti-

tutional change There is disagreement over the extent to which actorshave the freedom to build new institutional arrangements and if theydo to what extent the new arrangements and organizational patternsmay depart from past practices Among some institutional economists(Williamson 1985) there is a basic assumption that actors build efcientinstitutions by pursuing their self-interest and by promoting theirstrategic goals Moreover there is the assumption that a mixing of fullrationality with market competition tends to produce all of the optimalinstitutions that are needed in order to coordinate a complex capitalisteconomy (Hollingsworth and Boyer 1997 ch 14 Hodgson 1999) Incontrast the argument developed here is much more complexSocial institutions are historically rooted and there is a great deal of

path-dependency in the way that various institutional componentsevolve The shape of institutional congurations at any moment limitsthe type of options for change Because there is a great deal of institu-tional inertia radical change in the institutional components of a society

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

624

is uncommon (Lanzara 1998) Recognizing the constraints existinginstitutions exert on actors we note that existing congurations of insti-tutional arrangements institutional sectors and organizational propertiescan limit but also inuence the degree to which norms and rules canchange Many scholars (North 1990 Ostrom 1990) have emphasized therole of norms rules ideologies and values in limiting the ability of actorsto develop new institutional arrangements which radically depart fromexisting ones However there is a two-way process of interaction aprocess of co-evolution (Baum and Singh 1994a Nelson 1994) Thus asFriedland and Alford (1991 244) argue actors develop new institutionalarrangements by recognizing organizational failures and low perfor-mance but these new arrangements do not necessarily change a societyrsquosnorms rules habits and its underlying value systemCampbell (1997) advances our understanding of how existing institu-

tions enable actors to construct radically new institutions He recognizesthat most institutions and institutional arrangements embody a degreeof inertia from which it is difcult for actors to depart Moreover thedifferential power relations among actors make it unlikely that lesspowerful actors can change the existing power structure of a societyMost change in the institutional components of a society evolves througha process of constrained selection which reects to a considerable degreethe existing arrangements and power relations This becomes the inter-pretative frame for social actors with institutionalized scripts and ritualswhich tend to be taken for granted and appear to be quite ordinary andnatural This institutionalized scripting of the social world is an extremelyimportant source of social stability and inertiaCampbell has developed a very fruitful explanation of change in terms

of the various institutional components outlined here He distinguishesbetween incremental and radical institutional change and argues thatradical institutional change occurs when social actors with widelydiffering norms cultural scripts and rituals for action engage in intenseinteraction with each other (see also Knight 1992) The more diverse theinteractions the greater the potential for institutional change He arguesthat as shifts in the composition of interaction among social actors occurchanges in the interpretation of problems and interests will follow Ifthere are only minor changes in the extent to which diverse actorsinteract there will be only modest institutional change But the morefundamental the changes in the interaction of diverse social actors thegreater the change in the way that actors interpret their world And thegreater the change in the way that actors interpret their world the morelikely that radical changes in various institutional components willfollow In other words diverse decision making tends to create a widerrange of interpretative frameworks of the social world than is likely tobe the case if social actors operate in isolation

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

625

This perspective of diverse interactions among social actors facilitatesour ability to move beyond the traditional view that existing institutionsconstrain the range of institutional alternatives which actors face Theolder view has tended to suggest that actors engage in institutionalchange by extending existing institutional principles habits and conven-tions Specifying the conditions which facilitate fundamental institutionalchange permits a much richer perspective for institutional analysisCampbellrsquos perspective about the conditions under which fundamentalor radical change in institutional components occurs is consistent withmy own research about the conditions under which major discoveriesor fundamental new knowledge occur the more scholars with diversebackgrounds interact with intensity and frequency the greater the like-lihood that they will develop new and alternative ways of thinking(Hollingsworth and Hollingsworth 2000)These views of institutional change are also consistent with the recent

work of Nelson and his colleagues (Mowery and Nelson 1999 Murmann1998) They report the research on a variety of industrial sectors in anumber of different countries revealing that networks of actors in partic-ular industries business and professional associations universityresearch departments and institutes are frequently able to use their collec-tive power to modify a societyrsquos property rights research system andgovernmental policies It is the intense and frequent interaction of thesediverse complex sets of actors ndash sometimes constituting a technologysystem ndash that often leads to incremental changes in a societyrsquos total socialsystem of production It is this kind of collective action that brings aboutorganizational change change in institutional sectors institutionalarrangements and even in the rules and norms of a society And if weare to understand how the style of innovativeness of a society changeswe must rst understand the changes in its institutional makeup It ishighly unlikely that a society will develop a new style of innovativenesswithout changing its institutional structure

FIRMS THEIR INSTITUTIONAL ENVIRONMENT ANDNATIONAL SYSTEMS OF INNOVATION

Variation in the innovative style of societies is shaped by their insti-tutional makeup ndash the various levels which are discussed aboveinstitutions institutional arrangements the structure and cohesion of institutional sectors which constitute a societyrsquos social system of pro-duction and the structure and culture of its organizations ndash especiallyits business rms and other research organizations By undertaking aninstitutional analysis of a society one can begin to understand in whatkinds of organizations the production of specic kinds of knowledgetakes place and how this is linked to particular kinds of innovativeness

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

626

With such a perspective one can gain a rich understanding of why somesocieties excel in the production of radical breakthroughs in basic andapplied science and in developing radically new products and why othersocieties excel in more incremental innovationsReecting on these issues one cannot help being impressed with the

contrasts between the innovative styles of German and American orga-nizations For example Germany has continued to be very successful in making incremental innovations in many industries in which it wasvery competitive well before World War II paper printing materialsmachinery electrotechnical products motor vehicles chemicals textileyarns and fabrics But the Germans have been much less successful indeveloping totally new products or related innovations in biotechnologyelectronics telecommunications aircraft as well as other newer indus-trial areas In contrast the Americans have been very innovative in these and other industries with products having very short lives andwith technologies which change very rapidly and are very complex(Hollingsworth 1991 Soskice 1997 1999)The importance of institutions and innovativeness for international

competitiveness has led to the recognition that there are national systemsof innovation (Lundvall 1992 Nelson 1993 Edquist 1997 Hage andHollingsworth 2000) But even before the recent studies of nationalsystems of innovation Landes (1969) was a proponent of these viewswhen he made the argument that Germany had developed a nationalsystem of innovation by building on its education and scientic researchsystems which were lacking in Britain and elsewhere However the concept of a national system of innovation immediately poses theproblem of what kind of innovations should be consideredDavid Soskice an economist at the Wissenschaftszentrum Berlin has

addressed a set of problems somewhat complementary to the perspec-tive raised here His work (Soskice 1990 1997 1999) suggests that weare sensitive to the institutional environment in which the followingactors are embedded the employees of rms the owners and nanciersof rms the competitors of rms and nally their collaborators ndash aperspective complementary to Hollingsworthrsquos work on social systemsof production (Hollingsworth 1997 1998) When one reects on the insti-tutional contextsocial system of production within which these actorsare embedded one becomes aware that those countries which have some-what inexible labor markets external to rms but a business system in which rms have long-term commitments to their employees arestrongly associated with incremental but hardly at all with radical inno-vations In the rms of such countries (eg Germany and Japan) thereis much more consensus decision making than in societies where a lsquohireand rersquo set of practices is quite pervasive And this kind of consensusdecision making ndash while highly conducive to incremental innovativeness

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

627

ndash limits the capacity for radical innovations to occur This kind of lsquolongtermismrsquo set of norms and rules occurs not only in the employmentpractices of incrementally innovative rms but also in their nancingand ownership patterns (Hollingsworth 1997) During the past thirtyyears these were characteristics of many rms in Germany and JapanSignicantly in both of these countries venture capital markets werenot highly developed nancing tended to be based more on long-termbank loans than on the equity markets and the relationships of rmstended to be relatively stable And these stable institutional arrange-ments also contributed to a high degree of innovativeness ndash butincremental rather than radical innovationsMeantime we should also pursue Soskicersquos ideas (1999) about how

rmsrsquo relationships with their competitors and collaborators shape theirstyles of innovativeness In those advanced industrial societies in whichbusiness rms are highly mobilized into business associations theretends to be a high degree of standardized quality controls over technicalnorms and a high degree of reliance on legal sanctions as monitoringinstruments vis-agrave-vis association members As a result of this high degreeof integration of rms into associations member rms are more likelyto be associated with incremental innovations than is likely to be the case with rms in those societies where associations are less welldeveloped In the latter type of society radical innovations are morelikely to occur In those advanced industrial societies where rms havelong-term stable relationships with collaborators in product design andproduction the style of innovation is likely to be incremental And it is in those societies in which there is a weaker tradition of strongcollaboration among rms that radical innovations are more likely tooccurObviously even in a country where there may be a great deal of

radical innovativeness most innovations will be of an incremental typefor there are simply fewer radical innovations But Germany and therest of Europe not only have had relatively few radical innovations incomparison with the United States but they also have lagged behindthe USA in developing relatively totally new industries Thus there hasbeen much less innovative activity in Europe in high-technology andnewly emerging industries than in the USAIn complementary research Audretsch and Feldman (1995) have

demonstrated that the USA has been more successful in spawning rmsengaged in the early stages of the industry life cycle but relative toGermany had an environment unfavorable to innovative activity byestablished rms especially in more mature industries German rmshave tended to be associated with a routinized technological regime ndashwhere established enterprises have had strong incentives to engage inincremental innovations but new rms have tended to have fundamental

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

628

innovative disadvantages relative to the Americans (see Audretsch 1995Winter 1982 Casper 2000)On both sides of the Atlantic universities engage in basic and applied

research But signicantly the institutional environment in which univer-sities are embedded inuences the degree to which the research is likelyto represent major breakthroughs or is likely to have more modestimplications Recent research in biomedical science demonstrates that inthose societies in which university professors andor graduate studentshave the following characteristics there are likely to be relatively majorresearch breakthroughs graduate students have a long training periodhigh dependency on a major professor low autonomy in dening thenature of a research project restricted job options and a university envi-ronment where there is a very modest or nonexistent entrepreneurialculture For example in Germany ndash a country with relatively few radicalbreakthroughs ndash the scholar who nishes a habilitation will have spentmany years on a research project before qualifying for a permanent acad-emic position will have been in a high-dependency position with a seniorprofessor will have had limited autonomy to dene a research projectand will be constrained from pursuing a high-risk research project Theyoung German will have an extraordinarily high level of training butwill have faced a number of disincentives to undertake high-risk researchAnd as soon as heshe has completed the habilitation and becomes aprofessor administrative responsibilities constitute another disincentivefrom engaging in high-risk or long-term research projects Moreover theGerman professor is a civil servant with a more stable and secure incomethan hisher American counterpart In a very different institutionalenvironment American professors in large research universities tend tobecome much more entrepreneurial Today many American professorsbehave as though they were operating their own rms within univer-sities and thus devote considerable time to generating funds for researchprojects assistants secretaries supplies etc And the variation in thisentrepreneurial spirit across countries is associated with variation in thetaking of high-risk research strategies and the making of major researchbreakthroughsIn sum societies vary in terms of the institutional environment in

which their organizations are embedded and this variation inuencesthe degree to which they have industrial activities in the early or latestages of the industry life cycle successful new or mature industriesand radical vs incremental innovations And it is this variation whichis encouraging rms in countries with relatively few new industries and with few or no radical innovations to invest in those countries with many new industries and numerous radical innovations ndash as astrategy to be able to transfer new technologies from the latter institu-tional environment to the former

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

629

Table 4 summarizes the way that the norms rules and values of a society are associated with Americarsquos and Germanyrsquos institutionalarrangements their social system of production and their styles of inno-vativeness

CONCLUSION

Several themes deserve to be reemphasized Most importantly institu-tional analysis is emerging as a dominant issue in several social sciencedisciplines The study of various institutional components discussedabove is a complex subject requiring multiple levels of analysis and aknowledge of multiple sectors within societies It requires a theoreticalperspective that evolves from an empirical analysis in different societiesIf we are to advance the study of institutional analysis we need cross-disciplinary collaboration and team research which are not readilyachieved given the structure of most of our research universitiesAt present scholars in each academic discipline have their own distinc-

tive strategies for studying institutional analysis Thus economists tendto use several approaches (North 1990 Williamson 1985 Hodgson 19881998 1999) anthropologists another (Geertz 1995) and sociologistspolitical scientists and historians employ other strategies (Campbell etal 1991 Powell and DiMaggio 1991 Katznelson 1998 Hall and Taylor1996) With practitioners in various disciplines pursuing their distinctiveapproaches to institutional analysis there is low potential for collabo-rative institutional analysis across the social sciences The lessons fromthe biological sciences over the last 40 years are very instructive oncepractitioners from various biological disciplines began to cooperate instudying biological phenomena by using the same concepts at the mole-cular level theoretical advance was very rapid Comparable phenomenatook place in the development of other hybrid elds eg biochemistrybiophysics (Judson 1979 Olby 1979 Kohler 1982 1994) Similarly weresocial scientists in various disciplines to work collaboratively in the areaof institutional analysis there would very likely be an acceleration intheir theoretical knowledgeThis is not a plea that everyone engaged in institutional analysis should

do the same kind of research Indeed just as biologists working at themolecular level study many different kinds of problems similarly socialscientists engaged in institutional analysis would also work on manyseparate problems but in the spirit of a collective enterprise Some wouldconcentrate their attention on the study of rules norms habits conven-tions and values while others would study how these are associatedwith congurations of various institutional arrangements (types ofmarkets hierarchies networks associations communities clans statesetc) Others would work on specic institutional sectors (eg education

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

630

Table 4

Rules norms hab

its an

d values associated w

ith sp

ecic co

ngu

ration

s of coo

rdinating mecha

nism

s soc

ial sy

stem

s of

prod

uction

an

d nationa

l styles of inno

vation

USA

Germany

Norms rules values which provide incentives for

Norms rules values which provide incentives for

short-term horizons

long-term horizons

Con

gu

rative

forms of

Marke

ts c

orpo

rate hierarchies reg

ulatory state

Assoc

iation

s (busine

ss assoc

iation

s la

bor union

s)

coordinating

corp

oratist-type

networks

marke

ts c

orporatist-

mecha

nism

s (listed in

type

state

order of im

portanc

e in

the particu

lar social

system

)

Social sys

tem of

prod

uction

Bus

iness system

Firm

struc

tures

Con

glom

erates b

ut with mov

emen

t toward

Firm

more co

ngru

ent with pa

rticular prod

ucts

elim

ination of produc

tion

in som

e sectors

emph

asis on long

-term strateg

ies rm

s are well

emph

asis on sh

ort-term

strateg

ies high

integrated

differentiation

between divisions

of rms

Firm

s mob

ilized in

Low

High

and in

tegrated

into busine

ss

associations

Indu

strial relations

Low

deg

ree of jo

b secu

rity rigid in

ternal lab

or

High deg

ree of jo

b secu

rity

exible interna

l lab

or

system

marke

t ex

ible externa

l labo

r marke

tsmarke

t rigid externa

l lab

or m

arke

tPatterns of

Dispersed owne

rship frequ

ent turnov

erMore stab

le owne

rship patterns

owne

rship

Table 4

continue

d

USA

Germany

Norms rules values which provide incentives for

Norms rules values which provide incentives for

short-term horizons

long-term horizons

Social system of

Prod

uction

Bus

iness system

Training for labo

r an

dHigh deg

ree of voluntary decision mak

ing by

Collective an

d m

ore co

mpulso

ry d

ecisions

abo

ut

man

agem

ent

individuals as to ho

w m

uch

tim

e an

d ene

rgy

how m

uch ed

ucation

is ne

eded

he

avy em

pha

sis

to inv

est in educ

ation heavy

emph

asis on

on techn

ical training (eg en

gine

ering)

busine

ss sch

ools training

in marke

ting

and

sales accou

nting

Fina

ncial marke

tsEqu

ity an

d ven

ture cap

ital m

arke

ts highly

Equ

ity an

d ven

ture cap

ital m

arke

ts w

eakly

dev

elop

ed na

ncing from

cen

tral gov

ernm

ent

deve

lope

d he

avy em

pha

sis on

corporate loan

sto priva

te rm

s for dev

elop

men

t of new

tech

nologies related

to na

tion

al security

Unive

rsity research

High individual auton

omy high poten

tial for

Low

ind

ividua

l au

tono

my until almost 40

yea

rs of

system

individual creativity u

nive

rsity pr

ofesso

rs

age poten

tial for ind

ividual creativity high

ly

high

ly entreprene

urial

cons

traine

d unive

rsity pr

ofessors not so

entrep

rene

urial

Inno

vative

styles

Highly su

ccessful in m

ajor break

through

s in

High pe

rforman

ce in proc

ess an

d inc

remen

tal

basic an

d app

lied scien

ce go

od in dev

elop

ing

inno

vation

s in existing produ

cts not so su

ccessful

totally

new

lines of pr

oducts an

d in m

ajor break

through

s in basic scien

ce or in

tech

nologies high

deg

ree of entrepr

eneu

rship

deve

loping

totally new

lines of prod

ucts of

amon

g research

ers

tech

nologies low d

egree of entrepr

eneu

rship

amon

g research

ers

Table 4

continue

d

USA

Germany

Norms rules values which provide incentives for

Norms rules values which provide incentives for

short-term horizons

long-term horizons

Prod

ucts in w

hich

Pr

oducts with sh

ort liv

es prod

ucts in w

hich

Older produ

cts with long

history machine

ry pap

er

society excels

tech

nology

is bo

th highly co

mplex

and

pr

inting

textile

s m

otor veh

icles m

achine

too

lsch

ange

s rapidly en

tertainm

ent indus

try

pha

rmaceu

ticals biotechn

olog

y softw

are

aerosp

ace

business systems nancial markets and systems of research) Distinctivewould be the recognition of how these separate studies are linkedtogether how each institutional sector is linked to a societyrsquos normsrules values etc and its conguration of institutional arrangementsSimilarly other institutional analysts might focus on the study of orga-nizations with a major concern for how the institutional environmentof organizations (eg norms rules values congurations of institutionalarrangements institutional sectors) inuenced the structure culture andoutcomes of organizations And nally others might study how all ofthe aforementioned aspects of institutional analysis inuenced a societyrsquosoverall performance Nelson and Sampat (1998) have appropriatelyreminded us that we should engage in institutional analysis in order toaddress specic problems but the view developed in this article is thatwe rst need some mapping of the terrain of institutional analysis inorder for our diverse enterprises to become a meaningful collective enter-prise especially if we are to understand how institutions inuence theinnovative processThe argument developed here is not that all social scientists should engage

in institutional analysis Obviously there are numerous other importantareas of research which lie beyond the eld of institutional analysis Nordoes this article argue that institutional analysis is the most importantsubject for social scientistsSince the structure and culture of our research universities reect an

enormous amount of disciplinary fragmentation what are the prospectsfor promoting the kind of collective and interdisciplinary research agendaproposed above In one respect the prospects are not encouragingDuring the past half century the efforts of American universities to have effective communication across the biological sciences have variedgreatly in their success At the University of California-Berkeley it was not until the 1980s that there was a serious effort to integrate thebiological sciences and at that time the campus abolished seventeendepartments to promote an integrated biological program But at theUniversity of Wisconsin the biological sciences are still very fragmentedinto dozens of different departments in botany plant pathology zoologyand physiology with two departments of biochemistry and two depart-ments of geneticsThere have been other approaches which have worked very well histor-

ically for promoting scientic integration amidst scientic diversity Forexample several summer schools and special institutes were organizedin the United States during the 1940s in an effort to overcome the frag-mentation in the biological sciences These programs made it possiblefor scientists in diverse academic disciplines and from different univer-sities and countries to come together to engage in intense and frequentinteraction Perhaps the most effective program was that held in the

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

634

summers at Cold Spring Harbor New York where a research agendafor the study of molecular biology was developed (Fischer and Lipson1988 Stent and Watson 1966) Scientists went there summer aftersummer and when they returned to their universities many began therestructuring of the biological sciences in their universitiesIn our own day we need summer institutes and other similar programs

so that scholars from diverse backgrounds can come together anddevelop a collective agenda for doing institutional analysis and forstudying the impact of institutions on the performance of societies ndashinnovativeness being one such subject If we can develop such instituteswe will increase the prospects for transforming our universities Thereis reason to be optimistic that once an eclectic group of scholars workwith a common set of concepts we can have genuine theoretical advanceif researchers can bring their diverse expertise to bear on a common setof problems There is considerable evidence that institutional analysisoffers high potential for advancing our knowledge of innovations andtechnological change (Hollingsworth and Hollingsworth 2000) If we aregoing to make substantial theoretical advance it is not likely to occurfrom the vantage point of a multiplicity of disciplines Rather it is likelyto occur as we develop a new intellectual framework and our own acad-emic journals This is a tall agenda which requires us to reassess theway that knowledge is presently produced in our societies

NOTE

This article is part of a much larger agenda involving the study of institutionsorganizations and innovations in which I am involved My debts are many andvaried The rst version was written while I was in residence in the ResearchUnit for Institutional Change and European Integration of the Austrian Academyof Sciences in Vienna in the summer of 1998 I am especially grateful to ProfessorsEgon Matzner and Sonja Puntscher-Riekmann for providing the stimulating envi-ronment in which to think through the ideas developed here Egon Matzner hasengaged in very stimulating discussions with me about the importance of insti-tutions in the general eld of socio-economics The nal version of the articlewas completed during my residence as a fellow at the Neurosciences Institutein La Jolla California I also wish to thank my colleagues in the InnovationTheme Project at the Netherlands Institute for Advanced Study (NIAS) during1998ndash9 and Edgar Grande and Karl Muumlller for stimulating conversations aboutthe interaction of institutions and organizations in facilitating particular typesof innovation I am especially grateful to Jerald Hage for helping me to developa number of issues in the article and for his useful comments on the rst draftFrans van Waarden early on helped me to work through the idea of multiplelevels of institutional analysis Others who have helped to develop the ideas inthe article have been my colleagues in the research project lsquoComparative socialsystems of productionrsquo especially Tom Burns Christel Lane Yoshitaka OkadaWolfgang Streeck and Richard Whitley My colleagues Robert Boyer SteveCasper and David Soskice have taught me much about how incentives providedby the institutional environments of rms inuence individual decision making

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

635

But my most important debt is to Ellen Jane Hollingsworth who has repeatedlyquestioned every single assumption of this article Without her rich and contin-uous tutoring this article would be even less developed than it is at presentFinally without the insightful assistance of David Gear in many phases of thisarticle it would not have been completed

REFERENCES

Aoki Masahiko (1990) lsquoTowards an economic model of the Japanese rmrsquo Journalof Economic Literature March 1ndash27

Arora A Landau Ralph and Rosenberg Nathan (eds) (1998) Chemicals and Long-Term Economic Growth Insights from the Chemical Industry New York JohnWiley amp Sons

Arrow Kenneth (1974) The Limits of Organization New York NortonArthur W B (1988a) lsquoSelf-reinforcing mechanisms in economicsrsquo in P W

Anderson K J Arrow and D Pines (eds) The Economy as an Evolving ComplexSystem Redwood City Calif Addison-Wesley

mdashmdash(1988b) lsquoCompeting technologies an overviewrsquo in G Dosi C Freeman RNelson G Silverberg and L Soete (eds) Technical Change and Economic TheoryLondon Pinter Publishers pp 590ndash607

Audretsch David B (1995) Innovation and Industry Evolution Cambridge MassMIT Press

Audretsch David B and Feldman Maryann P (1995) lsquoInnovative clusters andthe industry life cyclersquo Discussion Paper Wissenschaftszentrum BerlinMarket Processes and Corporate Development Research Unit FS-IV-95-7

Baum Joel A C and Oliver Christine (1992) lsquoInstitutional embeddedness andthe dynamics of organizational populationsrsquo American Sociological Review 57540ndash59

Baum Joel A C and Singh Jitendra V (1994a) lsquoOrganizationndashenvironment coevo-lutionrsquo in Joel A C Baum and Jitendra V Singh (eds) The EvolutionaryDynamics of Organizations New York Oxford University Press pp 379ndash402

mdashmdash(eds) (1994b) The Evolutionary Dynamics of Organizations New York OxfordUniversity Press

Berger Suzanne and Dore Ronald (eds) (1996) National Diversity and GlobalCapitalism Ithaca NY Cornell University Press

Boyer Robert (1997) lsquoThe variety and unequal performance of really existingmarkets farewell to Doctor Panglossrsquo in J Rogers Hollingsworth and RobertBoyer (eds) Contemporary Capitalism the Embeddedness of InstitutionsCambridge and New York Cambridge University Press pp 55ndash93

Burns Tom R and Dietz Thomas (2001 forthcoming) lsquoSocial rule system theorysocial action institutional arrangements and evolutionary processesrsquo in JRogers Hollingsworth and Karl Muumlller (eds) Advances in SocioeconomicsLanham Md Rowman amp Littleeld

Burns Tom R and Flam Helena (1987) The Shaping of Social Organization BeverlyHills Calif Sage

Calvert Randall L (1995) lsquoThe rational choice theory of social institutionsrsquo inJeffrey S Banks and Eric A Hanuskek (eds) Modern Political Economy NewYork Cambridge University Press pp 216ndash66

Camic Charles (1986) lsquoThe matter of habitrsquo American Journal of Sociology 911039ndash87

Campbell John L (1997) lsquoMechanisms of evolutionary change in economicsgovernance interaction interpretation and bricolagersquo in Lars Magnusson

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

636

to the institutional environment in which they are embedded but mustmodify their environment in order to maintain competitive advantagesAs these comments suggest the work of Nelson and his colleagues is

very promising for understanding why societies vary in their capacityto excel in different industries However to extend and complementtheir work we need a greater understanding of and consensus on themeaning of institutions and the institutional environment of rms Hencethis article is written in the spirit of complementing the agenda of Nelsonand his colleagues One of its goals is to develop a map of what can becalled the terrain of institutional analysis It assumes that until we havea map of this terrain scholars working in the broad eld of institutionalanalysis will not comprehend how their work interrelatesInstitutional analysis is now rather high on the research agenda of

the social sciences However we need to be aware of the obstaclesconfronting us as we attempt to advance an agenda of institutionalanalysis There is no consensus as to what is meant by institutions orby institutional analysis These terms are very widely used but they areused with different conceptualizations and the scholars who use themshare little common ground Until scholars have some consensus aboutthe meaning of the concepts they use their potential to bring about effec-tive advancement of knowledge is somewhat limited Thus thewidespread interest in several academic disciplines in the concepts lsquoinsti-tutionsrsquo and lsquoinstitutional analysisrsquo may well promise more than it candeliver given the organizational and disciplinary fragmentation ofcontemporary universitiesThere are many different approaches to the study of institutions

(Nelson and Sampat 1998) there are the new and the old institution-alisms (Stinchcombe 1997 Hodgson 1998 Langlois 1986 1989) thereis historical institutionalism (Steinmo et al 1992 Zysman 1994

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

598

ORGANIZATIONALCAPABILITIES INNOVATIVENESS

INSTITUTIONS INSTITUTIONAL ARRANGEMENTSAND INSTITUTIONAL SECTORS

Figure 1 Institutional environments organizations and innovativeness

Immergut 1998 Katznelson 1998) and several of the social scienceshave their own distinctive approaches to the study of institutions (Halland Taylor 1996 Hodgson 1988 Eggertsson 1990 Finnemore 1996Scott 1994 Calvert 1995 Hechter and Kanazawa 1997)The following comments reect some of the confusion in utilizing the

concept lsquoinstitutionrsquo Nobel laureate Douglass North in his bookInstitutions Institutional Change and Economic Performance (1990 3) denesinstitutions as lsquorules of the game in a societyrsquo To North institutions areconstraints which shape human interaction and the way that societiesevolve through time On the other hand Andrew Schotter argues thatinstitutions lsquoare not rules of the gamersquo Rather institutions are thebehavior that follows from rules Briey he is concerned with whatactors do with rules but not with what the rules are (Schotter 1981155) Many other examples might be given to illustrate the heterogeneityof approach to institutions and institutional analysis Even if scholarswere to agree with North that rules and norms are institutions theywould not necessarily agree on what a rule is Shimanoff for examplehas identied more than 100 synonyms for the concept lsquorulersquo (Shimanoff1980 57 Ostrom 1986 5)Another critical issue in the institutional literature is the relationship

between institutions and organizations North (1990) following from hisdenition of institutions argues that institutions and organizations aredistinct entities Which organizations come into being and how theyevolve through time is inuenced by a societyrsquos rules and norms thatis by its lsquoinstitutionsrsquo On the other hand a number of recent organiza-tional sociologists see very little difference between institutions andorganizations For them rules and norms are institutions which unfoldin tandem with organizational structures and processes and changes inorganizational forms internalize and reect changes in the societyrsquos rulesand norms Using this perspective a whole subdiscipline within soci-ology called the lsquonew institutionalismrsquo in organizational analysis hasemerged The lsquoinstitutionalistrsquo perspective on organizations assumes thatthe kinds of organizations which actors create are dictated by the culturalnorms and rules in which they are embedded (Powell and DiMaggio1991 Zucker 1988 1991)The importance of this disagreement about institutions is obvious If

institutions are so critical for understanding our societies it is impor-tant that we come to some systematic consensus as to what institutionsare and how they inuence social actors and the organizations that theycreate If we cannot do so we risk talking past one another and losingthe opportunities for cumulative knowledge based on articulation ofwidely shared theoretical understandingWe need not only conceptual clarication as to what institutions are

but also greater consensus as to how to study institutions No scientic

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

599

eld can advance very far if the practitioners do not share a commonunderstanding of the key concepts used in their analysis (Ostrom 19864) But with our universities and academic associations so fragmentedinto different disciplines and into various subspecialties within disci-plines it is difcult to advance the theoretical agenda of institutionalanalysis within the academy Indeed the disciplinary fragmentation ofthe modern university is a major barrier to the theoretical advancementof the study of institutions and innovations as well as most other hybrid elds of research (Hollingsworth 1984 Hollingsworth andHollingsworth 2000) And it will be only as a result of effective commu-nication across diverse elds of knowledge that our study of institutionsand innovativeness will be effectively advanced

MULTIPLE LEVELS OF INSTITUTIONAL ANALYSISAND INSTITUTIONAL CHANGE

There are innumerable signs that we are living in a time of great insti-tutional change the demise of the Soviet empire the processes ofEuropean political and economic integration rapid transformation inparts of the global economy the disintegration of the family structurethe weakening of voluntary associations and the decline in politicalparticipation in a number of advanced capitalist societies the weakeningof welfare states European law is superseding national law and is evenchanging complete national legal systems The list could go on and onEven though scholars discuss institutional change at length their

ability to measure the rate of institutional change is very limited Andmore crucial than the limited ability of scholars to measure institutionalchange is their very limited understanding of how to build new insti-tutions One of the reasons for these shortcomings is that the socialsciences are decient in a theory of institutions The building of newinstitutions and redressing the decline of some of the most importantinstitutions of our societies are among the most important problems ofour time If we are to advance in the development of a theory of insti-tutions we need to work collaboratively across the social sciences andwe need to dene the parameters of institutional analysis

First level of analysis

This article attempts to make some modest contribution to outlining theparameters of institutional analysis At the outset we need to recognizethat when we engage in institutional analysis we must be sensitive to multiple levels of reality As suggested above most scholars whoengage in institutional analysis do not participate in any coordinatedactivity with each other and their activity is fragmented into a variety

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

600

of disciplines and subdisciplines To establish some coherence to the eldof institutional analysis we need a map of the eld so that those workingin one area can see where their research ts in relation to other areasand other practitioners on the mapTable 1 presents such a map with multiple levels at which institu-

tional analysis occurs Theoretically each of these areas on the map isinterrelated with each other level However the various areascompo-nents on the map are arranged in descending order of stability orpermanence Those components at higher levels of reality are morepermanent and durable while those at lower levels change more rapidlyWere scholars doing institutional analysis able to reach some consensus

about where their own work ts in relation to all other practitioners inthe eld there would be increased potential for all practitioners tocommunicate with each other By analogy once geneticists crystallo-graphers biochemists etc had a good understanding of how theirinvestigations were related to each area of molecular biology the eldquickly was able to make theoretical advances (Judson 1979)At the rst level there are the basic norms rules conventions habits and

values of a society These are the most fundamental properties of insti-tutions and are the most enduring and resistant to change Rules normsconventions etc are institutions but are only one component of whatconstitutes institutional analysis As Burns and Dietz (2001 forthcoming)point out most human activity is organized and regulated by normsand rules and systems of rules These concepts are extremely importantfor institutional analysis as they exert the greatest inuence on the natureof the components of institutional analysis at the next four levels which

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

601

Table 1 Components of institutional analysis

1 Institutions = norms rules conventions habits and values (see North1990 Burns and Flam 1987)

2 Institutional arrangements = markets states corporate hierarchies networksassociations communities (Hollingsworth and Lindberg 1985 Campbell et al 1991 Hollingsworth et al 1994 Hollingsworth and Boyer 1997)

3 Institutional sectors = nancial system system of education businesssystem system of research (Hollingsworth 1997)

4 Organizations (Powell and DiMaggio 1991)5 Outputs and performance = statutes administrative decisions the nature

quantity and the quality of industrial products (Hollingsworth 1991 1997)sectoral and societal performance (Hollingsworth and Streeck 1994Hollingsworth et al 1990 Hollingsworth and Hanneman 1982)

Note The ve components in this table are arranged in descending order of permanenceand stability That is norms conventions etc are more enduring and persistent than eachof the other components of institutional analysis Each component is interrelated withevery other component and changes in one are highly likely to have some effect in bringingabout change in each of the other components

are depicted in Table 1 In most forms of social analysis it is extremelyimportant that we understand the social and cognitive conditions thatlead to compliance or non-compliance with rules and the conditionswhich lead to changes in rulesThe approach to the study of institutions employed here argues that

norms rules habits conventions and values both reect and shape thepreferences of actors Norms rules habits conventions and values inu-ence who and what are included in different types of decision makinghow information is processed and structured what action is taken(Shepsle 1986 1989) It is through norms and rules that behavior isjudged to be democratic fair or egalitarianBurns and Flam (1987) point out that in any society there are multiple

rule systems Within a family there are rules for decision making oftenquite different from rules and norms for decision making for a professorin a classroom or for the customer in a bank Despite the heterogeneityof rule systems there are meta rules and norms which encompass lesserrule systems Otherwise there would be such contradictory rule systemsthat a society would be paralyzed The existence of meta rule systemspermits different rule systems to intersect with each other so that ambi-guities can be resolved The greater the pluralism and complexity of asociety the more ambiguity there is about meta rules and norms in asociety and of course all ambiguities never completely disappearOverall the degree to which separate rule systems are interlinked is anempirical problem There are different sectors groups and interestspursuing their own action logic but through higher-order meta princi-ples and rules there can be order consensus and coherence in a societyIt is through a set of meta rules that class and ethnic conict in societiesare contained (For elaboration see Burns and Flam 1987)In many respects our understanding about norms rules habits con-

ventions and values inuences our perspective on how societies are con-structed and how they change lsquoNew institutionalistsrsquo (Posner 1992Schotter 1981 Williamson 1975 1985) tend to assume that at one timethere was a state of nature and that there was a movement from individ-uals to institutions ndash an approach often called methodological individual-ism (Popper 1961 Hodgson 1998 1999) And of course there areinnumerable instances which methodological individualists cite to demon-strate that individuals create new rules of behavior For example it is pos-sible for actors to change the rules of driving so that instead of drivingon the left side of the road drivers adopt a new rule and drive on the rightThis article however tends to equate social habits and institutions

As Hodgson and others remind us (Hodgson 1988 1989 1997 1999Grafstein 1992 Camic 1986 Johnson 1992 Nelson and Winter 1982Veblen 1899) social habits are the results of earlier choices and are ameans of avoiding endless deliberation Because cognitive frameworks

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

602

are learned through habit individuals rely on the acquisition of suchcognitive habits before reason communication choice or action arepossibleWhereas Schotter (1981) and other game theorists take the individual

as an agent unencumbered by previous habits Field (1984) and othershave stressed that there can be no games without prior norms and rulesand thus a set of norms and rules must be presumed at the start Thosewho attempt to explain institutions from individual behavior alone areusing a bad strategy (Hodgson 1998)The position here ndash heavily inuenced by Hodgson (1998 1997 1988)

ndash is that individuals are embedded in a complex institutional environ-ment and that institutions not only constrain but also shape individuals(also see Hollingsworth and Boyer 1997) It would be a mistake howeverto get involved in an innite regress in order to determine which camerst ndash individuals or institutions Of course institutions are formed andchanged by individuals just as individuals are shaped and constrainedby institutions But at a macro level it is institutions that provide acognitive framework whereby individuals can cope with their reality Inthis sense the micro and macro worlds are intertwined At the macrolevel there is considerable stability but at the micro level individualshave a signicant level of autonomy and there can be widespread diver-sity As Hodgson (1988) reminds us most institutions in a temporal senseexist prior to the individuals in any given societyIt would be a serious mistake to downplay the importance of indi-

viduals and micro level analysis as we study institutions In the nalanalysis it is at the level of individuals that norms rules habits conven-tions and values exist An individual is born into and socialized intogroups and a society and this is how one early in life acquires a senseof appropriate forms of behavior Because of the way that individualsare socialized into a world of rules norms habits conventions andvalues it is unnecessary for individuals to restructure the world anewevery day (Douglas 1987 Elster 1989) Every action does not have tobe seriously reected upon For this reason institutions provide cogni-tive frameworks for individuals make their environments predictableprovide the information for coping with complex problems and envi-ronments In the words of Johnson (1992 26) lsquoInstitutions reduceuncertainty coordinate the use of knowledge mediate conicts andprovide incentive systems By serving these functions institutions providethe stability necessary for the reproduction of societyrsquo However eachsociety has different forms of habits rules and norms and hence differentincentives and disincentive systems for learning and forgetting forprocessing information But because individuals have varying degreesof autonomy individuals and groups can deviate from the prescribedforms of behavior in a society And of course these changes at the level

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

603

of individuals become important in understanding processes of societalchangeThese views are not meant to imply that the type of institutional

analysis proposed herein approximates a general theory of society Thisis clearly not the case However it is intended to represent the rst stepsin a mapping exercise of the boundaries of institutional analysis and tosuggest a few methodological insights for studying institutionsOne should try to see norms and rules as continuous and not as

dichotomous entities to recognize that they come in varying strengthsLegro (1997) has suggested that we assess the robustness of norms andrules with three criteria their simplicity their durability and their concor-dance Simplicity refers to how well actors understand norms and ruleshow well they can be applied within a specic situation Some normsand rules can be so complex that actors can have considerable difcultyin applying them in specic cases Durability addresses the issue of howlong norms and rules have been in effect ndash in short in order to assesstheir level of legitimacy While the position of this article is that normsrules and values are quite durable they do vary in this respectConcordance refers to how widely applied a norm or rule is Thisaddresses the degree to which a rule is a meta rule the degree to whichit incorporates the heterogeneity of other norms and rules In sum theclearer the norms and rules of a society the longer they have been inexistence and the more widely applicable they are the greater theirimpact on a society Hence the more robust the norms and rules thegreater their impact on a society and the less their robustness the greatertheir exibility and the less their effect on shaping a societyrsquos outcomesand performancesBecause norms rules and values are quite durable they play an impor-

tant role in shaping the history of societies thus contributing to a greatdeal of path-dependency Actors attempt to adjust to their contempo-rary environment but since they are products of the past the historicallegacy of norms rules and values inuences the decisions they makeAlthough actors have some capacity to alter the course of their historythey are constrained by their past and the degree to which they canmove beyond their past is limited As Lanzara (1998) Johnson (1992)and others have argued societal inertia is a basic feature of institutionsThey provide the basic stability necessary for change Degrees of historyare continuously reproduced by the way in which the inhabitants ofsocieties are socialized History matters but at critical points in historythere is punctuated equilibrium (Somit and Peterson 1992) During mostperiods of history there is considerable stability in the norms rules andvalues of a society but at critical moments norms rules and values canquickly and dramatically be redened At all times norms rules habitsconventions and values are inuencing each of the other components in

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

604

the institutional framework discussed below but these other institutionalcomponents feed back and can modify rules norms conventions etc(Murmann 1998)

Second level of analysis

The norms rules habits conventions and values of a society lead to thenext level of analysis ndash the institutional arrangements which are involvedin the coordination of various economic actors producers and suppliersof raw materials knowledge etc processors of raw materials infor-mation workers customers of raw materials nished productsinformation etc nanciers governmental and other types of regulatorsThese actors regularly engage in contests to resolve various economicproblems in virtually all sectors of society how are prices to be setWhat quantity of various products is to be produced How are stan-dards of various products processes etc to be set How is the qualityof products and processes to be determined How are various societalprocesses to be nanced In order to confront these problems and toaddress the conicting positions of economic actors as they address these problems societies develop various institutional (governance)arrangements for coordinating different actors These consist of marketsvarious types of hierarchies and networks associations the state commu-nities and clans (see Hollingsworth and Boyer 1997 Campbell et al1991 ch 1)Each of these particular kinds of coordinating mechanism has many

different types For example there are many types of states (eg theregulatory state developmental state authoritarian state welfare state)on which there is an extensive literature (Kim 1997) Similarly there aredifferent types of markets networks different kinds of associations etc(Boyer 1997 Hage and Alter 1997 Schneiberg and Hollingsworth 1990)When we do institutional analysis we must engage in congurativeanalysis recognizing that actors are not coordinated or governed by asingle type of institutional arrangement Some of the literature discussesindustrial sectors as though they were coordinated or governed by asingle institutional arrangement whereas in fact each sector of aneconomy is coordinated by a conguration of institutional arrangementsSome congurations coordinate actors in certain problem areas whileother congurations of institutional arrangements coordinate actors inaddressing other problems The types of congurations which are domi-nant in a society are somewhat stable and tend to persist over timewithin a society (Hodgson 1999)When one mode of coordination is dominant in a society it will

inuence the role which other coordinating modes will play Hence thestrong role of the state in the Soviet Union inuenced the role of markets

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

605

associations etc in the governance of the Soviet economy Similarly theprominence of particular modes of coordination in a society inuencesits style of innovationMuch of the literature on institutional arrangements (eg forms of eco-

nomic coordination) remains fragmented and unintegrated It is helpfulto array these various forms of coordination in a two-dimensional tax-onomy as in Figure 2 On the vertical dimension the economistrsquos viewof a self-interested agent is contrasted with a more sociological perspec-tive according to which social rules obligation and compliance shapehuman actions On the horizontal dimension there is another continuum(ie the distribution of power) At one extreme of the dimension onends many and relatively equal agents interacting (eg as in a well-organized spot market) At the other extreme inequality in power resultsin a hierarchical form of coordination which structures the interactionbetween principals and agents or between leaders and followers Wherea society falls along the distribution of power is inuenced by the rulesnorms values etc which are dominant in a society at a particular momentin timeInstitutional arrangements can be visualized on two dimensions the

nature of action motive on the one hand and the distribution of poweron the other Markets (cell 1) combine self-interest with horizontalcoordination transactors and they reect sensitivity to concerns aboutsupply and demand thus providing ex post an unintended equilibriumParadoxically the more pure and perfect the market competition thegreater the need for codied rules of the games for coordinating economictransactions Thus collective associations (cell 6) andor various formsof state intervention (cell 4) are required to develop and enforce rulesfor transacting partners (Schneiberg and Hollingsworth 1990 Streeckand Schmitter 1985a) This is an example of how the norms and rulesand the institutional arrangements of a society are intertwined asreected in a conguration of institutional or governance arrangementsAlong the horizontal axis actors can be joined in an organization or

a rm a hierarchy is the generic terms for this institutional arrangement(cell 2) Along the horizontal line one recognizes the difference betweentransactions in a market and transactions within a rm The well-knownworks of Coase (1960 1981) and Williamson (1975 1985) utilize theconcept of transaction costs in explaining the emergence of corporatehierarchiesThere are also various types of networks (cell 5) which exhibit mixes

of self-interest and social obligation with some actors being formallyindependent and equal Yet in some networks (the large rms and theirsubcontractors) there is unequal power and initiative Networks maycomprise all kinds of actors some consist only of rms but others includeassociations and the state (Hage and Alter 1997)

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

606

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

607

DISTRIBUTION OF POWER

VERTICALHORIZONTAL

1 MARKETS 2 HIERARCHY

6ASSOCIATIONS

5NETWORKS

4STATE

3COMMUNITIES

SELF INTEREST

SOCIAL OBLIGATION

ACTION MOTIVE

Figure 2 A general taxonomy of institutional arrangements

The vertical axis deals with action motives Toward the upper part ofFigure 2 actors are engaged in individualistically oriented behaviorwhereas toward the lower part actors are more engaged in collectivebehavior and strive to cope with problems of common interest Cell 3 ndashcommunities and clans ndash consists of institutional arrangements based ontrust reciprocity or obligation and thus not derived from the pure selshcomputation of pleasures and pains This is an unconventional form ofcoordination for most neoclassical economists (however see Arrow1974) but not for many anthropologists political scientists and sociolo-gists (Streeck and Schmitter 1985a Polanyi 1944 Gambetta 1988Fukuyama 1995 Sabel 1992 Putnam 1993)In the neoclassical paradigm theorists argue that actors engage in

forms of exchanges that best promote their individual interests If somestructural conditions are fullled (absence of increasing returns to scalethe reversibility of transactions absence of uncertainty and completecontingent markets with no collusion between actors) then the invis-ible hand theorem applies and market-type activity functions quite welland also provides the optimum for society thereby combining efciencyharmony and order However an excess of market activity may welllead to ruinous competition and excessive conict There is variation inthe extent to which ruinous competition occurs depending on the socialcontext in which transactions take place Thus it is important that webe sensitive to the institutional context in which transactions areembedded and that we understand the degree to which social bondsexist at both the micro and macro levels of analysis Micro bonds facil-itate exchanges in a society but at the societal level social bonds existat the level of the collective ndash in the community or region and amongmembers of racial religious and ethnic groups All other things beingequal the more powerful the social bonds among transacting partnersthe more economic competition is likely to be restrained Thus mosttransactions occur not simply in an impersonal calculative system ofautonomous actors unrestrained by social ties ndash as implied by theneoclassical paradigm ndash but in the context of social ties variation in the strength of which leads to variation in levels of trust and transac-tion costs (Etzioni 1988 211 Granovetter 1985 Streeck and Schmitter1985a Hollingsworth and Boyer 1997 Hodgson 1988 1999 Schneibergand Hollingsworth 1990)Another form of multilateral institutional arrangement is various

types of associations (cell 6) Unlike networks clans and communitiesassociations are more formal organizations Whereas markets corporatehierarchies and networks tend to coordinate economic activity amongdifferent types of actors (eg producers with suppliers capital withlabor) associations typically coordinate actors engaged in the same orsimilar kinds of activities Business associations and labor unions are

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

608

some of the most common forms of associations for coordinatingeconomic activity in capitalist economies (Schneiberg and Hollingsworth1990 Streeck and Schmitter 1985a)Finally there is the state (cell 4) which is an institutional arrangement

quite unlike any of the others It is the state that sanctions and regu-lates the various non-state coordinating mechanisms that is the ultimateenforcer of rules of the various mechanisms that denes and enforcesproperty rights and that manipulates scal and monetary policy Overallit is the state that inuences the total incentive system of a society(Lindberg and Campbell 1991 Johnson 1992 40) At the same time thestate may also be an economic actor by engaging directly in productionand exchange relationsThe choices of institutional arrangements in Figure 2 are constrained

by the social context in which they are embedded Depending on thenature of that embeddedness there is variation in the collective formsof governance some of which are specied in the lower part of thegure Some modes of coordination in the bottom and upper levels ofthe typology are often mixed together though one coordinating modeis likely to be more prominent than another But actors are often engagedin complex congurations involving several kinds of institutionalarrangementsEach of these various institutional arrangements has its own logic ndash

its own rules its own procedures for enforcing compliance its ownnorms and ideologies which help to reduce the costs of enforcementThese are summarized in Table 2 which provides further elaborationabout the various coordinating mechanisms one nds in almost everycapitalist society Table 2 lists the organizational arrangements and theirstructures rules of exchange and means of enforcing compliance asso-ciated with each type of coordinating mechanism While each type ofinstitutional arrangement has various positive features each institutionalarrangement also has failures and these are featured in Table 3 It is thecontest between those who support and those who oppose these variousinstitutional arrangements that tends to lead to transformations in insti-tutional arrangements over time (Campbell et al 1991 Campbell 1997)Again it is difcult to conceive of pure institutional arrangements ndash

either exclusively markets or exclusively hierarchies ndash since it has beenwell known since Adam Smithrsquos Wealth of Nations that the division oflabor within the rm cannot be disentangled from the existence andextent of the market Institutional arrangements have their own distinc-tive form of efciency (good static efciency for the market dynamicefciency for rms) and inefciency often leading to considerableinequality Neither networks nor communities are panaceas for economiccoordination without being congured with other types of institutionalarrangements Networks and communities may solve certain issues but

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

609

Table 2

Log

ics of ins

titution

al arran

gemen

ts

Coordination

Organizational

Rules of

Individual means

Collective means

mechanisms

structure

exchange

of compliance

of compliance

Marke

tsEasy en

try an

d exit

Voluntary sp

ot excha

nge

Leg

al enforcemen

t of

Norm of pr

ivate

Bila

teral ex

chan

ge or

control

prope

rty

marke

t site (W

all Street)

Reg

ulations

to

Leg

itim

acy of m

arke

t en

force contracts

men

tality

Com

mun

ities

Inform

al m

embe

rship

Voluntary exch

ange

based

So

cial norms an

d m

oral

Highly institutiona

lized

ev

olve

s ov

er lon

g period

on social solid

arity an

d

princ

iples im

pose

norm

s an

d rules requ

ire

of tim

ehigh

deg

ree of tru

stob

ligations

mem

bers to accept

Kno

wledg

e of others an

d

lsquocorpo

ratersquoo

bligations

recipr

ocity ov

er tim

e

Networks

Semi-form

al m

embe

rship

Voluntary exch

ange

ove

r a

Con

tractual bon

dsPersona

l relation

sBila

teral or m

ultilateral

time pe

riod

Resou

rce de

pend

ence

Tru

st built outside

the

ex

chan

geecon

omic arena

Associatio

nsFo

rmal m

embe

rship

Restricted to mem

bers

Self-interest

Some deg

ree of

Multilateral excha

nge

Emph

asis on insider

Rep

utation

effects

compu

lsion

outsider or we

they

Priva

te interest type

of

men

tality

gove

rnan

ce

Private

Com

plex organ

izations

Restricted to mem

bers

Rew

ards to ind

ividuals

Highly institutiona

lized

hierarch

ies

which

ten

d to becom

e ex

chan

ge based

on

Asymmetric po

wer threat

rules

bureau

cratic hierarchies

asym

metric po

wer

of san

ctions

Mem

bers soc

ializ

ed into

bureau

cratic rules

corp

orate cu

lture us

e of

sanc

tions

State

Pub

lic hierarchy

Unilateral action

Exit vo

ice (vote lob

bying)

Coe

rcion

De jure

and impo

sed

Indirect political an

d

loya

ltyNorms an

d pu

blic rules

mem

bership

econ

omic excha

nge

Table 3

Failu

res of coo

rdinating mecha

nism

s

Coordinating mechanisms

Type of failure

Associations

Private hierarchies

State

Market

Communities

Networks

Enforcem

ent

Usu

ally relies on

Might enh

ance

Needs co

ntrol

Needs an

internal

Needs

tru

st and

Need an external

the state as an

oppo

rtun

istic

extern

al to state

enforcem

ent

loya

lty often

en

forcem

ent

enforcer

beha

vior

bureau

cracy

authority

coming from

au

thority

Resem

bles

The

ideal o

f (ju

dge

s

Facilitates

outside (fam

ily May

facilitate

enforcem

ent

internal m

arke

ts

parliamen

t collu

sion

and

relig

ion

cartelization an

d mecha

nism

might hur

t marke

t) to

impe

rfect

ethn

icity)

mon

opoly

of cartels

incu

mbe

nt

correct state

compe

titio

nCom

patible with

worke

rsab

uses

variou

s types of

Lob

bies can

co

mpe

tition

captur

e pu

blic

interest goa

ls

Public goo

d

Useful for

Gov

erna

nce co

sts

Can

provide

Can

not provide

Can

internalize

Useful for

and

establishing

might exceed

public goo

ds

colle

ctive go

ods

some co

llective

enha

ncing qu

ality

externality

stan

dards an

d

the be

nets of

but ha

s or deal w

ith

good

s (qua

lity

and training bu

t qu

ality for

internal division

difc

ultie

s in

externalities

training

) bu

t no

t no

t ve

ry goo

d in

setting ru

les of

of la

bor

prov

iding them

Inad

equate

othe

rs (welfare

prov

iding for

compe

titio

n in

Slow

to react to

in precise

mon

itoring of

gene

ral p

ublic

societal gen

eral

the indu

stry

chan

ges in the

am

ounts

tech

nical ch

ange

go

ods)

welfare

Useful for

environm

ent

Might fail in

and inno

vation

Mem

bers tightly

Weak in the

prov

iding

induc

ing

integrated

into

prov

ision of

man

y go

ods

tech

nical

community

colle

ctive go

als

colle

ctively

chan

geha

ve limite

d

that in

dividua

lcapa

city for

mem

bers can

not

inno

vatio

nsprov

ide for

them

selves

Table 3

continue

d

Coordinating mechanisms

Type of failure

Associations

Private hierarchies

State

Market

Communities

Networks

Efcien

cyFa

cilitates

De

cien

t in

Can

be high

ly

Some ba

sic social So

me go

ods

Slow

to en

hanc

e co

operation an

d

coop

eration an

d

bureau

cratic

relation

s cann

ot

cann

ot be

efc

ienc

y an

d

X-efcien

cy but

X-efcien

cyan

d can

not

be provided

by

delivered

at

speed of

not allocative

easily deliver

pure m

arke

t su

fcien

tly

adap

tive

ness

efcien

cy

good

s at

mecha

nism

slow costs

except in

low cost

indus

tries whe

re

tech

nology

is

complex an

d rapidly cha

nging

Equity

Narrow

Excessive

Might enh

ance

Facilitates

Might lead to

Whe

n widely

encompa

ssing

multiplication

ineq

uality

ineq

uality in

retard

ed

dev

elop

ed in

to

associationa

l of con

trollers

(pow

er

income an

d dev

elop

men

tindus

trial

stru

ctures lead

(fru

stration

and

an

d privilege

)wealth

districts

to in

come

ineq

uality)

netw

orks m

ay

equa

lity

facilitate greater

equality an

d

inco

me

distribution

whe

n wea

kly

dev

elop

ed

netw

orks ten

d to

increase soc

ial

ineq

uality

Note For allo

cativ

e efc

ienc

y an

d X

-efcien

cy see Leibe

nstein (19

66 19

76)

they raise other no less severe problems It is important to recognizethe imperfection of any single institutional arrangement in order tocomprehend the origin and transformation of any other institutionalarrangement (Hodgson 1999 ch 3)As suggested above each institutional arrangement is congured with

other institutional arrangements Usually one particular institutionalarrangement is more dominant in a particular conguration but becauseeach type of institutional arrangement has its own strengths andweaknesses there is no simple structural logic in the governance or coor-dination of a society Each institutional arrangement constrains the otherbut the inherent tension among the various institutional arrangementswithin a conguration contributes to changes in the governance orcoordination of a society The routines and logics of each of these cong-urations of institutional arrangements provide constraints and incentivesfor actors The way that actors perceive the incentives and constraintsof these governance congurations leads to a particular market logicand it is the specic market logic of a society which inuences its speciccapacities and weaknesses The inherent strain among the different logicsin a conguration of coordination helps to provide the exibility for asociety to adapt to new circumstancesIf there were a society with pure markets or a society coordinated

only by the state there would be too much rigidity and too little diver-sity to cope with the vast uncertainty in the global environment A societywith very little diversity in its coordination mechanisms would havelittle capacity to adapt to new circumstances Congurations with consid-erable diversity of institutional arrangements provide for a certainamount of incoherence in governance but they also provide for thecapacity to adapt to new circumstances While the Soviet Union wasdominated by the state as a coordinating mechanism there were alwaysfunctioning markets in the system Moreover the feudal society of thetwelfth and thirteenth centuries consisted not only of hierarchical rela-tionships between serfs and masters but also of urban guilds clericalhierarchies and markets (Johnson 1992 38) In sum the robustness ofinstitutions often depends on multiple and diverse principles and logicsof action on the inconsistency of principles and procedures on patternedforms of disorder (Lanzara 1998 Orren and Skowronek 1991 320 329)

Third level of analysis

The next level consists of the institutional sectors of a society The rulesnorms and values of a society inuence the array of institutional arrange-ments and levels one and two inuence the nature of and the relation-ships among various institutional sectors and all three levels areintricately linked together to form a social system of production Together

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

613

all of these components inuence the performance of economic sectorswithin a society as well as the performance of the total society A socialsystem of production is the way that a societyrsquos institutions (see rstlevel of analysis above) its institutional arrangements (see second levelof analysis above) and its institutional sectors are integrated into a socialcongurationAn institutional sector includes all organizations in a society which

supply a given service or product along with their associated focal orga-nizations (eg major suppliers funding sources regulators and so forth(Scott et al 1994 108 117 Campbell et al 1991 Hollingsworth et al1994) Institutional sectors include but are not limited to the societyrsquossystem of education system of research business system nancialmarkets legal system and the state The structure of the nancial marketssystem of training and education industrial relations system system ofresearch and state and legal systems are distinct and idiosyncratic ineach society In short they are system specicAll of these institutional sectors tend to cohere with each other

although they vary in the degree to which they are tightly coupled intoa full-edged system While each of the institutional sectors has someautonomy and may have some goals that are contradictory to the goalsof other institutional sectors with which it is linked an institutional logicin each society leads institutions to coalesce into a complex social con-guration (Hollingsworth 1991) This occurs because the institutionalsectors are embedded in a culture in which their logics are symbolicallygrounded organizationally structured technically and materially con-strained and politically defended The conguration of institutionalsectors usually exhibits some degree of adaptability to new challengesbut continues to evolve within an existing style But under new circum-stances or unprecedented disturbances these institutional congurationsare exposed to sharp historical limits as to what they may or may notdo (Schumpeter 1983 David 1988 Arthur 1988a 1988b Haringkanssen andLundgren 1997)Why do all of these different institutions coalesce into a complex

social conguration which is labeled here as a social system of pro-duction The literature suggests two contrasting interpretations Part ofthe answer ndash indeed a controversial one ndash is that these institutionalsectors are functionally determined by the requirements of the practiceof capitalism in each time and place (Habermas 1975) Another expla-nation emphasizes the genesis of the actual conguration via a trial and error process according to which the survival of rms regions orcountries is the outcome of complex evolutionary mechanisms (Maynard-Smith 1982 Nelson and Winter 1982) However the problem is evenmore complex Markets and other mechanisms for coordinating rela-tionships among economic actors place constraints on the means and

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

614

ends of economic activity to be achieved in any society The othercoordinating mechanisms include different kinds of hierarchies varioustypes of networks and associations ndash trade unions employers and busi-ness associations (Hollingsworth and Lindberg 1985 Campbell et al1991) The logic of these various institutional arrangements providesactors with vocabularies and logics for pursuing their goals for deningwhat is valued and for shaping the norms and rules by which theyabide (see Tables 2 and 3) In short in contrast to the logic of theneoclassical paradigm the argument here is that the dominant type ofinstitutional arrangements places severe constraints on the denition of needs preferences and choices of economic actors Whereas the neo-classical paradigm assumes that individuals and rms are sovereign thisarticle is based on the assumption that rms are inuenced by the hold that the institutional congurations making up a social system ofproduction have on individual decision making (Campbell et al 1991Etzioni 1988 Streeck and Schmitter 1985a Hollingsworth et al 1994Hollingsworth and Boyer 1997 Magnusson and Ottosson 1997 North1990 Hodgson 1999)Standard neoclassical economic theory has tended to downplay the

role of production and consequently of rms Even transaction cost theo-rists who are concerned with analyzing the rm as a coordinatingmechanism have been relatively unconcerned with the various compo-nents of a social system of production Indeed as long as there waswidespread optimism about the efcacy of Keynesian economics therewas relatively little concern among neoclassical economists with thesupply side of the economy Even in the opinion of most Keynesians agroup of experts should ideally be able to shape the size of aggregatedemand while the supply side of the economy would be left to the twominimalist institutions of neoclassical economics ndash markets and manage-rial hierarchies For several decades however it has become increasinglyobvious that some of the most competitive and successful patterns ofindustrial output and industrial production in capitalist economies donot derive from the neoclassical prescription of unregulated markets andcorporate hierarchies complemented by a neoliberal democratic stateIndeed it is now well understood that certain highly successful produc-tion patterns require for their emergence and survival institutionalarrangements and environments the very opposite of the prescriptionsfound in the neoclassical paradigm (see especially Streeck 1992 but alsoHollingsworth and Streeck 1994) Thus social scientists have come tounderstand that if they are to comprehend the behavior and performanceof contemporary economies concerns about social systems of produc-tion must be brought into the pictureBecause production involves more than technology a number of social

scientists have an increasing concern with social systems of production

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

615

The same equipment is frequently operated quite differently in the samesectors in different countries even when rms are competing in the same market (Maurice et al 1980 Sorge and Streeck 1988 Sorge 1989Hollingsworth et al 1994) Variations in production and process tech-nologies are inuenced partly by variations in the social environmentsin which they are embedded In other words rms are embedded intocomplex environments which among other things place constraints ontheir behavior Thus a social system of production is of major impor-tance in understanding the behavior and performance of an economyHow the state and other institutional arrangements (eg marketsnetworks private hierarchies associations) coalesce and are related toparticular social systems of production are important determinants ofeconomic performanceThe congurations of institutional arrangements that coordinate or

govern the behavior of actors in one society and its structure of specicinstitutional sectors cannot easily be transferred to another society forthey are embedded into a social system of production that is societallydistinct (Hollingsworth 1997) Societies borrow selected principles offoreign management styles and work practices but the effectiveness ofsuch borrowing is generally limited Economic behavior and performanceare shaped by the entire social system of production in which actors areembedded and not simply by specic principles of particular manage-ment styles and work practices Moreover a societyrsquos social system ofproduction tends to limit the kind of goods that it can produce and withwhich it can compete successfully in international marketsSome scholars (Kenney and Florida 1993 Oliver and Wilkinson 1988)

have assumed that the diffusion of particular forms of management stylesand work practices across societies could lead to system convergenceReferring to the Japanese production system Kenney and Florida haveargued that it consists of organizational practices whose fundamentallsquogenetic codersquo can successfully be inserted into another society and beginsuccessfully to reproduce its behavior in the new environment (1993 8)Their position is in the intellectual tradition of Antonio Gramsci (1971)who contended decades ago that the American system of mass produc-tion would most certainly diffuse to Europe over timeHowever the argument of this article is that even though British

French and American rms may adopt certain aspects of Japanesemanagement styles (eg just-in-time production self-managing teamsquality circles the use of statistical process controls) or some variant ofthe German vocational system their social systems of production willnot converge Moreover the overall conguration of a societyrsquos socialsystem of production inuences its sectoral and overall national economicperformance This helps to explain why societies have different systemsof innovation and why they vary in the clusters of industries in which

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

616

they are highly competitive in the international markets (Nelson 19931999 Hage and Hollingsworth 2000 Berger and Dore 1996)In order to understand how and why a societyrsquos economy performs

as it does it is necessary to understand its entire social system of produc-tion If a society is to modify substantially the performance of itsinstitutional sectors and its entire economy it cannot adapt only someof the management and work practices of its foreign competitors Ratherit must alter its entire social system of production Because a societyrsquosmodes of economic governance and coordination and its institutionalsectors develop according to particular logics and are system specicthere are serious limitations to the extent to which a society may mimicthe institutions (eg the rules and norms) institutional arrangements andinstitutional sectors of other societiesIn the history of modern societies there are logics by which institu-

tions coalesce into a social system of production (Hollingsworth 1991)Though institutions are constantly changing there are sharp limits tothe type and the direction of change that any particular institution canundergo because of its linkages with institutional arrangements and insti-tutional sectors Thus a societyrsquos business rms educational systemnancial markets industrial relations system etc can engage in seriousrestructuring only if most of the other institutional sectors also changeIn social systems of production there are pressures toward consistencyin the norms rules and values across institutional sectors though in anycomplex society social systems are obviously imperfectly integratedIndeed the degree to which the institutional norms and rules makingup a social system of production are loosely or tightly coupled is a vari-able of considerable importance In general the institutional sectorsmaking up a social system of production are interdependent and changesin one generally result in changes in the others Since each institutionallevel is dependent on the others for various types of resources there isinterdependence among the differing institutional spheres Moreovereach society has its norms moral principles rules and laws and recipesfor action as well as its own idiosyncratic customs traditions and prin-ciples of justice (Burns and Flam 1987)There are also other inherent obstacles to convergence among social

systems of production for where a system is at any one point in timeis inuenced by its initial state Systems with quite different initial statesare unlikely to converge with one anotherrsquos institutional practicesExisting institutional arrangements block certain institutional innova-tions and facilitate others (Roland 1990) Thus the institutions makingup a social system of production provide continuity even though insti-tutional arrangements are always changing but with a particular logicWhile Williamson (1975 1985) suggests that actors tend to choose theinstitutional arrangements which are most efcient North (1990) is

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

617

much closer to the mark in his argument that most societal institutionalarrangements exist as a result of custom and habit and are thereforeinefcient At any moment in time the world often appears to its actorsas very complex and uncertain For this reason actors often engage incontradictory forms of behavior pursuing different strategies as hedgesagainst a very uncertain world (Lanzara 1998) And their hedging andcontradictory forms of behavior may lead in somewhat different soci-etal directions all constrained by the institutional fabric within whichthe actors are embedded This kind of contradictory behavior occurs inpart because of the contradictions inherent in the dominant congura-tions of institutional arrangements with their contradictory logics (seelsquoSecond level of analysisrsquo above)Despite the emphasis on the logic of institutional continuity this is

not an argument that systems change along some predetermined pathThere are critical turning points in the history of highly industrializedsocieties but with the choices limited by the existing institutional terrainBeing path-dependent a social system of production continues along aparticular logic until or unless a fundamental societal crisis intervenes(Milgrom et al 1991 Krugman 1991 Durlauf 1991 Hollingsworth 1991Pred 1966 David 1988 Hodgson 1998)At this point it is important to confront the question of how social

systems of production evolve And this question gets to the heart of theproblem of building complex societal structures Certainly social systemsof production did not emerge from some process of social engineeringMoreover the various component parts of each social system of produc-tion have often not been designed to be part of a social system ofproduction The component parts of each social system of productionhave emerged more often than not as unintended by-products of goalswhich various actors had in mind at earlier moments in time It is usuallythe case that actors at time lsquotrsquo fail to comprehend the long-term conse-quences of their actions Wolfgang Streeck the author of several brilliantpapers on social systems of production argues that institutional sectordesigns created for one purpose generally address that goal but overtime those designs have quite unintended consequences (Streeck 1997a1997b Wright 1998) Describing the conguration of the German socialsystem of production of the 1970s and 1980s Streeck points out that itwas the unintended by-product of multiple points in history Someelements were pre-Wilhelmian others were introduced by the Alliedpowers after 1945 and others emerged during the years of the GermanFederal Republic All component parts of the German social system ofproduction

were and continue to be changing for their own reasons as wellas in reaction to each other and certainly there can be no presump-tion of a pre-established t between them even though one might

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

618

want to allow for some reinforcement effects of the modelrsquos histor-ically contingent social and economic success

(Streeck 1997b 54)

In sum the emergence of social systems of production is a long-termevolutionary process with each part interacting with its environment andresulting in a congurative whole but with no previous design by eithera single actor or a collectivity of actors As suggested above actors do not recreate their world anew At any moment in history the totalinstitutional complex of a society necessarily contains the resources and legacies of its past Preexisting institutional complexes are nevercompletely wiped out Societies consist of multiple layers of history withtheir diverse logics of action However the cumulative effect of smallperipheral changes in altering particular institutional sectors and insti-tutional arrangements can be substantial (Murmann 1998 Lanzara 1998)lsquoInstitution building is affected more by the ways in which people havecodied the past than by how they envision the future There are manymore potential resources in pre-existing institutional arrangements thanit is usually assumed or suspectedrsquo (Lanzara 1998 30)

Fourth level of analysis

The next level of institutional analysis organizational structures is some-what more controversial As suggested above Douglass North (1990)draws a sharp distinction between institutions and organizations Morerecently many organizational theorists (Powell 1991 Powell andDiMaggio 1991 DiMaggio and Powell 1983 Baum and Oliver 1992Kondra and Hinings 1998 Townley 1997) argue that institutional rulesnorms and conventions unfold in tandem with organizational structuresand this is my positionThe literature which focuses on how institutions inuence organiza-

tions is quite different from two other theoretical literatures which alsoare concerned with how organizational environments shape the behaviorof organizations There are resource dependency theorists (Pfeffer andSalancik 1978) who emphasize the role of environmental resources inshaping organizations while Hannan and Freeman (1977 1984) and otherpopulation ecologists emphasize the survival of organizations givencertain kinds of organizational conditions (Orru et al 1991)The institutionalist perspective as implied above emphasizes the

normative environment in which organizations are embedded It is a perspective which focuses on the way in which organizations in their behavior tend to conform to the institutional rules and normswhich are dominant in the organizational environment However allthree perspectives ndash the institutionalist the population ecology and

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

619

the resource dependency ndash emphasize how the environment inuencesorganizations and how organizations which are subject to the sameenvironment tend to converge in their behavior to have what DiMaggioand Powell have labeled organizational isomorphism (DiMaggio andPowell 1983)There has emerged a vast literature which demonstrates that within

each society there are scal political judicial and other regulatory normswhich limit and shape the culture and structure of organizationalbehavior The normative institutional environment of organizations limits the options of what organizations do in a particular society and inuences the patterns of ownership relations with suppliers andcustomers In short it is the normative environment of organizationswhich denes within a particular society what is socially acceptablebehavior for organizations (Hamilton and Biggart 1988 Hollingsworthand Hollingsworth 2000 Meyer and Rowan 1991 Zucker 1987 1988Orru et al 1991 Townley 1997)Thus far most of the analyses on isomorphic pressures operating on

organizations have been cross-sectional in nature However the entireconcept of isomorphism implies that there are strong environmental pres-sures exerted on organizations and in order to observe this phenomenonit is important that we have longitudinal studies which assess howchanges in an organizationrsquos institutional environment inuence changesin the structure and culture of organizations Moreover the expectationhere is that when one analyzes the patterns of organizational changeover time the isomorphism will be holistic ndash that is within a particularsociety there will be a tendency for the internal structure of particulartypes of organizations to converge in large part because of the pres-sures to conform to the changes in the external norms rules and valuesA close reading of Aoki (1990) and Slack and Hinings (1994) nds thatchanges in the national institutional environment of organizations haveinuenced changes in the internal structure of organizations within aspecic industry of a particular societyHowever it is inappropriate to view organizations as changing only

in relation to exogenous environmental change There is a kind of co-evolution which occurs between organizations and their institutionalenvironments Co-evolution implies nonlinear feedback between orga-nizations and their institutional environments This is of course quitedifferent from the more common way of studying organizations bymodeling relations between independent and dependent variables Ofcourse the more common type of analysis is quite appropriate whenthere are simple relationships which do not involve complex feedbackprocesses But in the strategy for institutional analysis which is proposedherein it is less useful to separate independent from dependent vari-ables and more useful to understand the interacting and co-evolutionary

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

620

processes among all of the levels of institutional analysis discussed inthe various sections above (Baum and Singh 1994a 379ndash80 Mowery andNelson 1999 Murmann 1998)The logic of the perspective herein suggests that within every

society there is variation in the structure and culture of business rmsuniversities and other complex organizations (Kondra and Hinings1998) However this variation takes place within parameters which aresystem specic For example every German rm and every Germanuniversity is different from every other German university and rm but there are a set of cultural and structural characteristics which distin-guish German rms and universities from those in America Similarlyeach Japanese rm and university is unique yet they have a set ofcultural and structural characteristics which are system specic andwhich differentiate them from business rms and universities in other countriesIn all societies each organization has its own distinctive organizational

rules norms and conventions which are subordinate to the meta norms and rules of the larger society within which they are embeddedHowever the strength of the institutional environment within whichorganizations are embedded varies from society to society Some societies have multiple institutional environments and there is hetero-geneity in terms of what constitutes appropriate behavior oforganizations In such societies there can be conict over whatinstitutional logics should regulate specic organizational functions(Friedland and Alford 1991 Townley 1997 262ndash3) In those societies inwhich the institutional norms habits and rules are most developed and in which the institutional pressures to conform are greatest thereis less variation in the structure and culture of business rms and variouskinds of research organizations In such societies the connectednessbetween research organizations and their institutional environments is sufciently strong for organizations to have low autonomy to pursue independent strategies and goals in these societies there is a great deal of organizational isomorphism Conversely the weaker the institutional environment in which research organizations areembedded the greater the variation in the structure and culture of business rms and research organizations Moreover where the institu-tional environments are more weakly developed organizations havegreater autonomy and exibility to respond to the development of new knowledge and to be highly innovative Hence it is in those societies where the institutional environments are most developed and most rigid and when there is less organizational autonomy and exibility that fewer radical innovations in basic and applied science as well as in totally new products and industrial sectors havebeen made

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

621

Fifth level of analysis

We come next to the outputs and performance of the various institutionalcomponents of a society It is at this level that institutional componentsare more pragmatic and exible For example in the legal sector thereare specic statutes and court rulings in the state sector there are specicpolicies in the business sector there are new products new technolo-gies and market strategies It is at this nal level of analysis thatinstitutional spheres are most open and susceptible to change and cross-national mimicry is easiest and most commonThrough the outputs of the society we can obtain some assessment as

to how well a society is performing Moreover we can assess how inno-vative it is how egalitarian it is in the distribution of its resources howegalitarian it is in terms of levels of health education etc Just as it iscomplicated to measure how egalitarian a society is similarly it is dif-cult to assess how innovative it is For example societies vary in makingincremental and radical breakthroughs in basic and applied science in developing totally new products and new kinds of organizations inincremental and process innovations in existing products and organiza-tions in developing new and different forms of marketing ndash bothdomestically and globally (Hage and Hollingsworth 2000) Of coursesocieties vary in their rates of economic growth in their rates of economicproductivity in their quality of life (eg rates of crime life expectancyetc) All of these performance criteria feed back and inuence each ofthe levels of institutional analysis discussed above ndash rules norms valuesetc institutional arrangements institutional sectors the structure andform of organizations Moreover performance measures may inuenceeach other For example if a society has very low rates of growth itmay not be very innovative in some of the types of innovative activitiesmentioned above Good or poor performances of certain kinds inuenceother performance indicators (Hollingsworth et al 1990)Different institutional arrangements and different social systems of

production result in different types of economic performance Hence solong as societies have different social systems of production there areserious constraints on the degree to which they can converge in theirinnovative styles Different social systems of production tend to maxi-mize in a more or less explicit manner different performance criteriausually mixing considerations about static and dynamic efciency protsecurity social peace and economic andor political power In short incontrast to the implications of neoclassical economic theory in real-worldeconomies there are no universal standards all economically rationalactors attempt to maximize Economic history provides numerous exam-ples of how a great variety of principles of rationality are implementedin different societies

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

622

A critical question in institutional analysis is whether an existing socialsystem of production which supports a set of routines for a particularkind of technology and industry can shift from old practices to newones There are numerous historical accounts of how the social systemof production in a particular society which worked so well for a numberof years could not adapt to new technologies (Lazonick 1990 Veblen1915 Nelson 1994) Schumpeter (1983) and Freeman (1991) have devel-oped the idea of a techno-economic paradigm which suggests thatdifferent technologies require specic forms of organizational arrange-ments ndash or what I called social systems of production Because oflsquoinstitutional inertiarsquo the social system of production of a society maynot be able to adapt to a new techno-economic paradigm (Nelson 199458 Lanzara 1998)Whether a social system of production can sustain its performance

standards depends not only on its intrinsic economic rationality but alsoon where it ts into a larger system If a particular social system ofproduction is immune from the innovativeness and competition of analternative system survival can be long lasting But if different socialsystems of production with diverging criteria of good economic perfor-mance meet in the world arena the arbitrariness of nationally imposedconstructed performance standards may be superseded by alternativeperformance criteria as a result of international competitivenessThe world economy is also socially constructed just as are national

economies Even if different social systems of production are competingin the international arena it is not always possible to determine whichis more competitively effective at any moment in time Hegemonicnation-states in the short run shape the rules of trade to favor their indus-trial sectors and rms But the history of hegemonic powers suggeststhat in the longer run such social systems of production sustainedlargely by military and political power eventually give way to moreinnovative and competitive social systems of production In our ownday as nation-states are increasingly integrated into a world economyeven hegemonic powers lose their innovative and competitive advan-tage and their share of world output decreases Such a country mayattempt to restructure its institutional arrangements and to readjust itsperformance preferences But to restructure its system of innovationgenerally calls for a major redistribution of power within a societyLargely for this reason societies have historically had limited capacityto reconstruct a system of innovation in the image of their major competi-tors Each countryrsquos social system of production is a conguration of ahost of norms rules and values as well as of institutional arrangementsEach system is constantly changing and is open to inuence from othersystems And indeed many technologies and practices diffuse from onesociety to another but a societyrsquos capacity to be innovative is constrained

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

623

by the existing social system of production Thus the same technologymay exist in numerous countries but how it is employed and how itinuences societal outcomes and performances varies from one institu-tional conguration or society to another (Hollingsworth 1998)

PERSPECTIVES ON INSTITUTIONAL CHANGE

A fundamental problem which remains unresolved in institutionalanalysis involves the nature of institutional change There is much confu-sion and miscommunication in the social sciences about institutionalchange in large part because we do not even know how to measure therate of institutional change Of course we know that there are bothexternal and internal focuses for change and we have had a good bitof scholarship on this subject What is not clearly understood is thatinstitutions and institutional arrangements within societies are histori-cally rooted that there is a great deal of path-dependency to the waythat institutions evolve Moreover the more intricately linked each sectoris with each other sector and with a societyrsquos rules and norms the lesschoice actors have to devise new institutions and institutional arrange-ments On the other hand if a sector or actor is too isolated it may betoo weak within a society to bring about any kind of effective changeThese considerations are very critical to the problem of how muchfreedom actors have to create new institutions afresh to what extentinstitutions diffuse from one society to another (Murmann 1998 Lanzara1998) This is a fundamental question for students of institutionalanalysis as we attempt to understand how the different institutionallevels are linked togetherThere are a variety of reasons why there is confusion in studying insti-

tutional change There is disagreement over the extent to which actorshave the freedom to build new institutional arrangements and if theydo to what extent the new arrangements and organizational patternsmay depart from past practices Among some institutional economists(Williamson 1985) there is a basic assumption that actors build efcientinstitutions by pursuing their self-interest and by promoting theirstrategic goals Moreover there is the assumption that a mixing of fullrationality with market competition tends to produce all of the optimalinstitutions that are needed in order to coordinate a complex capitalisteconomy (Hollingsworth and Boyer 1997 ch 14 Hodgson 1999) Incontrast the argument developed here is much more complexSocial institutions are historically rooted and there is a great deal of

path-dependency in the way that various institutional componentsevolve The shape of institutional congurations at any moment limitsthe type of options for change Because there is a great deal of institu-tional inertia radical change in the institutional components of a society

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

624

is uncommon (Lanzara 1998) Recognizing the constraints existinginstitutions exert on actors we note that existing congurations of insti-tutional arrangements institutional sectors and organizational propertiescan limit but also inuence the degree to which norms and rules canchange Many scholars (North 1990 Ostrom 1990) have emphasized therole of norms rules ideologies and values in limiting the ability of actorsto develop new institutional arrangements which radically depart fromexisting ones However there is a two-way process of interaction aprocess of co-evolution (Baum and Singh 1994a Nelson 1994) Thus asFriedland and Alford (1991 244) argue actors develop new institutionalarrangements by recognizing organizational failures and low perfor-mance but these new arrangements do not necessarily change a societyrsquosnorms rules habits and its underlying value systemCampbell (1997) advances our understanding of how existing institu-

tions enable actors to construct radically new institutions He recognizesthat most institutions and institutional arrangements embody a degreeof inertia from which it is difcult for actors to depart Moreover thedifferential power relations among actors make it unlikely that lesspowerful actors can change the existing power structure of a societyMost change in the institutional components of a society evolves througha process of constrained selection which reects to a considerable degreethe existing arrangements and power relations This becomes the inter-pretative frame for social actors with institutionalized scripts and ritualswhich tend to be taken for granted and appear to be quite ordinary andnatural This institutionalized scripting of the social world is an extremelyimportant source of social stability and inertiaCampbell has developed a very fruitful explanation of change in terms

of the various institutional components outlined here He distinguishesbetween incremental and radical institutional change and argues thatradical institutional change occurs when social actors with widelydiffering norms cultural scripts and rituals for action engage in intenseinteraction with each other (see also Knight 1992) The more diverse theinteractions the greater the potential for institutional change He arguesthat as shifts in the composition of interaction among social actors occurchanges in the interpretation of problems and interests will follow Ifthere are only minor changes in the extent to which diverse actorsinteract there will be only modest institutional change But the morefundamental the changes in the interaction of diverse social actors thegreater the change in the way that actors interpret their world And thegreater the change in the way that actors interpret their world the morelikely that radical changes in various institutional components willfollow In other words diverse decision making tends to create a widerrange of interpretative frameworks of the social world than is likely tobe the case if social actors operate in isolation

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

625

This perspective of diverse interactions among social actors facilitatesour ability to move beyond the traditional view that existing institutionsconstrain the range of institutional alternatives which actors face Theolder view has tended to suggest that actors engage in institutionalchange by extending existing institutional principles habits and conven-tions Specifying the conditions which facilitate fundamental institutionalchange permits a much richer perspective for institutional analysisCampbellrsquos perspective about the conditions under which fundamentalor radical change in institutional components occurs is consistent withmy own research about the conditions under which major discoveriesor fundamental new knowledge occur the more scholars with diversebackgrounds interact with intensity and frequency the greater the like-lihood that they will develop new and alternative ways of thinking(Hollingsworth and Hollingsworth 2000)These views of institutional change are also consistent with the recent

work of Nelson and his colleagues (Mowery and Nelson 1999 Murmann1998) They report the research on a variety of industrial sectors in anumber of different countries revealing that networks of actors in partic-ular industries business and professional associations universityresearch departments and institutes are frequently able to use their collec-tive power to modify a societyrsquos property rights research system andgovernmental policies It is the intense and frequent interaction of thesediverse complex sets of actors ndash sometimes constituting a technologysystem ndash that often leads to incremental changes in a societyrsquos total socialsystem of production It is this kind of collective action that brings aboutorganizational change change in institutional sectors institutionalarrangements and even in the rules and norms of a society And if weare to understand how the style of innovativeness of a society changeswe must rst understand the changes in its institutional makeup It ishighly unlikely that a society will develop a new style of innovativenesswithout changing its institutional structure

FIRMS THEIR INSTITUTIONAL ENVIRONMENT ANDNATIONAL SYSTEMS OF INNOVATION

Variation in the innovative style of societies is shaped by their insti-tutional makeup ndash the various levels which are discussed aboveinstitutions institutional arrangements the structure and cohesion of institutional sectors which constitute a societyrsquos social system of pro-duction and the structure and culture of its organizations ndash especiallyits business rms and other research organizations By undertaking aninstitutional analysis of a society one can begin to understand in whatkinds of organizations the production of specic kinds of knowledgetakes place and how this is linked to particular kinds of innovativeness

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

626

With such a perspective one can gain a rich understanding of why somesocieties excel in the production of radical breakthroughs in basic andapplied science and in developing radically new products and why othersocieties excel in more incremental innovationsReecting on these issues one cannot help being impressed with the

contrasts between the innovative styles of German and American orga-nizations For example Germany has continued to be very successful in making incremental innovations in many industries in which it wasvery competitive well before World War II paper printing materialsmachinery electrotechnical products motor vehicles chemicals textileyarns and fabrics But the Germans have been much less successful indeveloping totally new products or related innovations in biotechnologyelectronics telecommunications aircraft as well as other newer indus-trial areas In contrast the Americans have been very innovative in these and other industries with products having very short lives andwith technologies which change very rapidly and are very complex(Hollingsworth 1991 Soskice 1997 1999)The importance of institutions and innovativeness for international

competitiveness has led to the recognition that there are national systemsof innovation (Lundvall 1992 Nelson 1993 Edquist 1997 Hage andHollingsworth 2000) But even before the recent studies of nationalsystems of innovation Landes (1969) was a proponent of these viewswhen he made the argument that Germany had developed a nationalsystem of innovation by building on its education and scientic researchsystems which were lacking in Britain and elsewhere However the concept of a national system of innovation immediately poses theproblem of what kind of innovations should be consideredDavid Soskice an economist at the Wissenschaftszentrum Berlin has

addressed a set of problems somewhat complementary to the perspec-tive raised here His work (Soskice 1990 1997 1999) suggests that weare sensitive to the institutional environment in which the followingactors are embedded the employees of rms the owners and nanciersof rms the competitors of rms and nally their collaborators ndash aperspective complementary to Hollingsworthrsquos work on social systemsof production (Hollingsworth 1997 1998) When one reects on the insti-tutional contextsocial system of production within which these actorsare embedded one becomes aware that those countries which have some-what inexible labor markets external to rms but a business system in which rms have long-term commitments to their employees arestrongly associated with incremental but hardly at all with radical inno-vations In the rms of such countries (eg Germany and Japan) thereis much more consensus decision making than in societies where a lsquohireand rersquo set of practices is quite pervasive And this kind of consensusdecision making ndash while highly conducive to incremental innovativeness

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

627

ndash limits the capacity for radical innovations to occur This kind of lsquolongtermismrsquo set of norms and rules occurs not only in the employmentpractices of incrementally innovative rms but also in their nancingand ownership patterns (Hollingsworth 1997) During the past thirtyyears these were characteristics of many rms in Germany and JapanSignicantly in both of these countries venture capital markets werenot highly developed nancing tended to be based more on long-termbank loans than on the equity markets and the relationships of rmstended to be relatively stable And these stable institutional arrange-ments also contributed to a high degree of innovativeness ndash butincremental rather than radical innovationsMeantime we should also pursue Soskicersquos ideas (1999) about how

rmsrsquo relationships with their competitors and collaborators shape theirstyles of innovativeness In those advanced industrial societies in whichbusiness rms are highly mobilized into business associations theretends to be a high degree of standardized quality controls over technicalnorms and a high degree of reliance on legal sanctions as monitoringinstruments vis-agrave-vis association members As a result of this high degreeof integration of rms into associations member rms are more likelyto be associated with incremental innovations than is likely to be the case with rms in those societies where associations are less welldeveloped In the latter type of society radical innovations are morelikely to occur In those advanced industrial societies where rms havelong-term stable relationships with collaborators in product design andproduction the style of innovation is likely to be incremental And it is in those societies in which there is a weaker tradition of strongcollaboration among rms that radical innovations are more likely tooccurObviously even in a country where there may be a great deal of

radical innovativeness most innovations will be of an incremental typefor there are simply fewer radical innovations But Germany and therest of Europe not only have had relatively few radical innovations incomparison with the United States but they also have lagged behindthe USA in developing relatively totally new industries Thus there hasbeen much less innovative activity in Europe in high-technology andnewly emerging industries than in the USAIn complementary research Audretsch and Feldman (1995) have

demonstrated that the USA has been more successful in spawning rmsengaged in the early stages of the industry life cycle but relative toGermany had an environment unfavorable to innovative activity byestablished rms especially in more mature industries German rmshave tended to be associated with a routinized technological regime ndashwhere established enterprises have had strong incentives to engage inincremental innovations but new rms have tended to have fundamental

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

628

innovative disadvantages relative to the Americans (see Audretsch 1995Winter 1982 Casper 2000)On both sides of the Atlantic universities engage in basic and applied

research But signicantly the institutional environment in which univer-sities are embedded inuences the degree to which the research is likelyto represent major breakthroughs or is likely to have more modestimplications Recent research in biomedical science demonstrates that inthose societies in which university professors andor graduate studentshave the following characteristics there are likely to be relatively majorresearch breakthroughs graduate students have a long training periodhigh dependency on a major professor low autonomy in dening thenature of a research project restricted job options and a university envi-ronment where there is a very modest or nonexistent entrepreneurialculture For example in Germany ndash a country with relatively few radicalbreakthroughs ndash the scholar who nishes a habilitation will have spentmany years on a research project before qualifying for a permanent acad-emic position will have been in a high-dependency position with a seniorprofessor will have had limited autonomy to dene a research projectand will be constrained from pursuing a high-risk research project Theyoung German will have an extraordinarily high level of training butwill have faced a number of disincentives to undertake high-risk researchAnd as soon as heshe has completed the habilitation and becomes aprofessor administrative responsibilities constitute another disincentivefrom engaging in high-risk or long-term research projects Moreover theGerman professor is a civil servant with a more stable and secure incomethan hisher American counterpart In a very different institutionalenvironment American professors in large research universities tend tobecome much more entrepreneurial Today many American professorsbehave as though they were operating their own rms within univer-sities and thus devote considerable time to generating funds for researchprojects assistants secretaries supplies etc And the variation in thisentrepreneurial spirit across countries is associated with variation in thetaking of high-risk research strategies and the making of major researchbreakthroughsIn sum societies vary in terms of the institutional environment in

which their organizations are embedded and this variation inuencesthe degree to which they have industrial activities in the early or latestages of the industry life cycle successful new or mature industriesand radical vs incremental innovations And it is this variation whichis encouraging rms in countries with relatively few new industries and with few or no radical innovations to invest in those countries with many new industries and numerous radical innovations ndash as astrategy to be able to transfer new technologies from the latter institu-tional environment to the former

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

629

Table 4 summarizes the way that the norms rules and values of a society are associated with Americarsquos and Germanyrsquos institutionalarrangements their social system of production and their styles of inno-vativeness

CONCLUSION

Several themes deserve to be reemphasized Most importantly institu-tional analysis is emerging as a dominant issue in several social sciencedisciplines The study of various institutional components discussedabove is a complex subject requiring multiple levels of analysis and aknowledge of multiple sectors within societies It requires a theoreticalperspective that evolves from an empirical analysis in different societiesIf we are to advance the study of institutional analysis we need cross-disciplinary collaboration and team research which are not readilyachieved given the structure of most of our research universitiesAt present scholars in each academic discipline have their own distinc-

tive strategies for studying institutional analysis Thus economists tendto use several approaches (North 1990 Williamson 1985 Hodgson 19881998 1999) anthropologists another (Geertz 1995) and sociologistspolitical scientists and historians employ other strategies (Campbell etal 1991 Powell and DiMaggio 1991 Katznelson 1998 Hall and Taylor1996) With practitioners in various disciplines pursuing their distinctiveapproaches to institutional analysis there is low potential for collabo-rative institutional analysis across the social sciences The lessons fromthe biological sciences over the last 40 years are very instructive oncepractitioners from various biological disciplines began to cooperate instudying biological phenomena by using the same concepts at the mole-cular level theoretical advance was very rapid Comparable phenomenatook place in the development of other hybrid elds eg biochemistrybiophysics (Judson 1979 Olby 1979 Kohler 1982 1994) Similarly weresocial scientists in various disciplines to work collaboratively in the areaof institutional analysis there would very likely be an acceleration intheir theoretical knowledgeThis is not a plea that everyone engaged in institutional analysis should

do the same kind of research Indeed just as biologists working at themolecular level study many different kinds of problems similarly socialscientists engaged in institutional analysis would also work on manyseparate problems but in the spirit of a collective enterprise Some wouldconcentrate their attention on the study of rules norms habits conven-tions and values while others would study how these are associatedwith congurations of various institutional arrangements (types ofmarkets hierarchies networks associations communities clans statesetc) Others would work on specic institutional sectors (eg education

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

630

Table 4

Rules norms hab

its an

d values associated w

ith sp

ecic co

ngu

ration

s of coo

rdinating mecha

nism

s soc

ial sy

stem

s of

prod

uction

an

d nationa

l styles of inno

vation

USA

Germany

Norms rules values which provide incentives for

Norms rules values which provide incentives for

short-term horizons

long-term horizons

Con

gu

rative

forms of

Marke

ts c

orpo

rate hierarchies reg

ulatory state

Assoc

iation

s (busine

ss assoc

iation

s la

bor union

s)

coordinating

corp

oratist-type

networks

marke

ts c

orporatist-

mecha

nism

s (listed in

type

state

order of im

portanc

e in

the particu

lar social

system

)

Social sys

tem of

prod

uction

Bus

iness system

Firm

struc

tures

Con

glom

erates b

ut with mov

emen

t toward

Firm

more co

ngru

ent with pa

rticular prod

ucts

elim

ination of produc

tion

in som

e sectors

emph

asis on long

-term strateg

ies rm

s are well

emph

asis on sh

ort-term

strateg

ies high

integrated

differentiation

between divisions

of rms

Firm

s mob

ilized in

Low

High

and in

tegrated

into busine

ss

associations

Indu

strial relations

Low

deg

ree of jo

b secu

rity rigid in

ternal lab

or

High deg

ree of jo

b secu

rity

exible interna

l lab

or

system

marke

t ex

ible externa

l labo

r marke

tsmarke

t rigid externa

l lab

or m

arke

tPatterns of

Dispersed owne

rship frequ

ent turnov

erMore stab

le owne

rship patterns

owne

rship

Table 4

continue

d

USA

Germany

Norms rules values which provide incentives for

Norms rules values which provide incentives for

short-term horizons

long-term horizons

Social system of

Prod

uction

Bus

iness system

Training for labo

r an

dHigh deg

ree of voluntary decision mak

ing by

Collective an

d m

ore co

mpulso

ry d

ecisions

abo

ut

man

agem

ent

individuals as to ho

w m

uch

tim

e an

d ene

rgy

how m

uch ed

ucation

is ne

eded

he

avy em

pha

sis

to inv

est in educ

ation heavy

emph

asis on

on techn

ical training (eg en

gine

ering)

busine

ss sch

ools training

in marke

ting

and

sales accou

nting

Fina

ncial marke

tsEqu

ity an

d ven

ture cap

ital m

arke

ts highly

Equ

ity an

d ven

ture cap

ital m

arke

ts w

eakly

dev

elop

ed na

ncing from

cen

tral gov

ernm

ent

deve

lope

d he

avy em

pha

sis on

corporate loan

sto priva

te rm

s for dev

elop

men

t of new

tech

nologies related

to na

tion

al security

Unive

rsity research

High individual auton

omy high poten

tial for

Low

ind

ividua

l au

tono

my until almost 40

yea

rs of

system

individual creativity u

nive

rsity pr

ofesso

rs

age poten

tial for ind

ividual creativity high

ly

high

ly entreprene

urial

cons

traine

d unive

rsity pr

ofessors not so

entrep

rene

urial

Inno

vative

styles

Highly su

ccessful in m

ajor break

through

s in

High pe

rforman

ce in proc

ess an

d inc

remen

tal

basic an

d app

lied scien

ce go

od in dev

elop

ing

inno

vation

s in existing produ

cts not so su

ccessful

totally

new

lines of pr

oducts an

d in m

ajor break

through

s in basic scien

ce or in

tech

nologies high

deg

ree of entrepr

eneu

rship

deve

loping

totally new

lines of prod

ucts of

amon

g research

ers

tech

nologies low d

egree of entrepr

eneu

rship

amon

g research

ers

Table 4

continue

d

USA

Germany

Norms rules values which provide incentives for

Norms rules values which provide incentives for

short-term horizons

long-term horizons

Prod

ucts in w

hich

Pr

oducts with sh

ort liv

es prod

ucts in w

hich

Older produ

cts with long

history machine

ry pap

er

society excels

tech

nology

is bo

th highly co

mplex

and

pr

inting

textile

s m

otor veh

icles m

achine

too

lsch

ange

s rapidly en

tertainm

ent indus

try

pha

rmaceu

ticals biotechn

olog

y softw

are

aerosp

ace

business systems nancial markets and systems of research) Distinctivewould be the recognition of how these separate studies are linkedtogether how each institutional sector is linked to a societyrsquos normsrules values etc and its conguration of institutional arrangementsSimilarly other institutional analysts might focus on the study of orga-nizations with a major concern for how the institutional environmentof organizations (eg norms rules values congurations of institutionalarrangements institutional sectors) inuenced the structure culture andoutcomes of organizations And nally others might study how all ofthe aforementioned aspects of institutional analysis inuenced a societyrsquosoverall performance Nelson and Sampat (1998) have appropriatelyreminded us that we should engage in institutional analysis in order toaddress specic problems but the view developed in this article is thatwe rst need some mapping of the terrain of institutional analysis inorder for our diverse enterprises to become a meaningful collective enter-prise especially if we are to understand how institutions inuence theinnovative processThe argument developed here is not that all social scientists should engage

in institutional analysis Obviously there are numerous other importantareas of research which lie beyond the eld of institutional analysis Nordoes this article argue that institutional analysis is the most importantsubject for social scientistsSince the structure and culture of our research universities reect an

enormous amount of disciplinary fragmentation what are the prospectsfor promoting the kind of collective and interdisciplinary research agendaproposed above In one respect the prospects are not encouragingDuring the past half century the efforts of American universities to have effective communication across the biological sciences have variedgreatly in their success At the University of California-Berkeley it was not until the 1980s that there was a serious effort to integrate thebiological sciences and at that time the campus abolished seventeendepartments to promote an integrated biological program But at theUniversity of Wisconsin the biological sciences are still very fragmentedinto dozens of different departments in botany plant pathology zoologyand physiology with two departments of biochemistry and two depart-ments of geneticsThere have been other approaches which have worked very well histor-

ically for promoting scientic integration amidst scientic diversity Forexample several summer schools and special institutes were organizedin the United States during the 1940s in an effort to overcome the frag-mentation in the biological sciences These programs made it possiblefor scientists in diverse academic disciplines and from different univer-sities and countries to come together to engage in intense and frequentinteraction Perhaps the most effective program was that held in the

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

634

summers at Cold Spring Harbor New York where a research agendafor the study of molecular biology was developed (Fischer and Lipson1988 Stent and Watson 1966) Scientists went there summer aftersummer and when they returned to their universities many began therestructuring of the biological sciences in their universitiesIn our own day we need summer institutes and other similar programs

so that scholars from diverse backgrounds can come together anddevelop a collective agenda for doing institutional analysis and forstudying the impact of institutions on the performance of societies ndashinnovativeness being one such subject If we can develop such instituteswe will increase the prospects for transforming our universities Thereis reason to be optimistic that once an eclectic group of scholars workwith a common set of concepts we can have genuine theoretical advanceif researchers can bring their diverse expertise to bear on a common setof problems There is considerable evidence that institutional analysisoffers high potential for advancing our knowledge of innovations andtechnological change (Hollingsworth and Hollingsworth 2000) If we aregoing to make substantial theoretical advance it is not likely to occurfrom the vantage point of a multiplicity of disciplines Rather it is likelyto occur as we develop a new intellectual framework and our own acad-emic journals This is a tall agenda which requires us to reassess theway that knowledge is presently produced in our societies

NOTE

This article is part of a much larger agenda involving the study of institutionsorganizations and innovations in which I am involved My debts are many andvaried The rst version was written while I was in residence in the ResearchUnit for Institutional Change and European Integration of the Austrian Academyof Sciences in Vienna in the summer of 1998 I am especially grateful to ProfessorsEgon Matzner and Sonja Puntscher-Riekmann for providing the stimulating envi-ronment in which to think through the ideas developed here Egon Matzner hasengaged in very stimulating discussions with me about the importance of insti-tutions in the general eld of socio-economics The nal version of the articlewas completed during my residence as a fellow at the Neurosciences Institutein La Jolla California I also wish to thank my colleagues in the InnovationTheme Project at the Netherlands Institute for Advanced Study (NIAS) during1998ndash9 and Edgar Grande and Karl Muumlller for stimulating conversations aboutthe interaction of institutions and organizations in facilitating particular typesof innovation I am especially grateful to Jerald Hage for helping me to developa number of issues in the article and for his useful comments on the rst draftFrans van Waarden early on helped me to work through the idea of multiplelevels of institutional analysis Others who have helped to develop the ideas inthe article have been my colleagues in the research project lsquoComparative socialsystems of productionrsquo especially Tom Burns Christel Lane Yoshitaka OkadaWolfgang Streeck and Richard Whitley My colleagues Robert Boyer SteveCasper and David Soskice have taught me much about how incentives providedby the institutional environments of rms inuence individual decision making

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

635

But my most important debt is to Ellen Jane Hollingsworth who has repeatedlyquestioned every single assumption of this article Without her rich and contin-uous tutoring this article would be even less developed than it is at presentFinally without the insightful assistance of David Gear in many phases of thisarticle it would not have been completed

REFERENCES

Aoki Masahiko (1990) lsquoTowards an economic model of the Japanese rmrsquo Journalof Economic Literature March 1ndash27

Arora A Landau Ralph and Rosenberg Nathan (eds) (1998) Chemicals and Long-Term Economic Growth Insights from the Chemical Industry New York JohnWiley amp Sons

Arrow Kenneth (1974) The Limits of Organization New York NortonArthur W B (1988a) lsquoSelf-reinforcing mechanisms in economicsrsquo in P W

Anderson K J Arrow and D Pines (eds) The Economy as an Evolving ComplexSystem Redwood City Calif Addison-Wesley

mdashmdash(1988b) lsquoCompeting technologies an overviewrsquo in G Dosi C Freeman RNelson G Silverberg and L Soete (eds) Technical Change and Economic TheoryLondon Pinter Publishers pp 590ndash607

Audretsch David B (1995) Innovation and Industry Evolution Cambridge MassMIT Press

Audretsch David B and Feldman Maryann P (1995) lsquoInnovative clusters andthe industry life cyclersquo Discussion Paper Wissenschaftszentrum BerlinMarket Processes and Corporate Development Research Unit FS-IV-95-7

Baum Joel A C and Oliver Christine (1992) lsquoInstitutional embeddedness andthe dynamics of organizational populationsrsquo American Sociological Review 57540ndash59

Baum Joel A C and Singh Jitendra V (1994a) lsquoOrganizationndashenvironment coevo-lutionrsquo in Joel A C Baum and Jitendra V Singh (eds) The EvolutionaryDynamics of Organizations New York Oxford University Press pp 379ndash402

mdashmdash(eds) (1994b) The Evolutionary Dynamics of Organizations New York OxfordUniversity Press

Berger Suzanne and Dore Ronald (eds) (1996) National Diversity and GlobalCapitalism Ithaca NY Cornell University Press

Boyer Robert (1997) lsquoThe variety and unequal performance of really existingmarkets farewell to Doctor Panglossrsquo in J Rogers Hollingsworth and RobertBoyer (eds) Contemporary Capitalism the Embeddedness of InstitutionsCambridge and New York Cambridge University Press pp 55ndash93

Burns Tom R and Dietz Thomas (2001 forthcoming) lsquoSocial rule system theorysocial action institutional arrangements and evolutionary processesrsquo in JRogers Hollingsworth and Karl Muumlller (eds) Advances in SocioeconomicsLanham Md Rowman amp Littleeld

Burns Tom R and Flam Helena (1987) The Shaping of Social Organization BeverlyHills Calif Sage

Calvert Randall L (1995) lsquoThe rational choice theory of social institutionsrsquo inJeffrey S Banks and Eric A Hanuskek (eds) Modern Political Economy NewYork Cambridge University Press pp 216ndash66

Camic Charles (1986) lsquoThe matter of habitrsquo American Journal of Sociology 911039ndash87

Campbell John L (1997) lsquoMechanisms of evolutionary change in economicsgovernance interaction interpretation and bricolagersquo in Lars Magnusson

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

636

Immergut 1998 Katznelson 1998) and several of the social scienceshave their own distinctive approaches to the study of institutions (Halland Taylor 1996 Hodgson 1988 Eggertsson 1990 Finnemore 1996Scott 1994 Calvert 1995 Hechter and Kanazawa 1997)The following comments reect some of the confusion in utilizing the

concept lsquoinstitutionrsquo Nobel laureate Douglass North in his bookInstitutions Institutional Change and Economic Performance (1990 3) denesinstitutions as lsquorules of the game in a societyrsquo To North institutions areconstraints which shape human interaction and the way that societiesevolve through time On the other hand Andrew Schotter argues thatinstitutions lsquoare not rules of the gamersquo Rather institutions are thebehavior that follows from rules Briey he is concerned with whatactors do with rules but not with what the rules are (Schotter 1981155) Many other examples might be given to illustrate the heterogeneityof approach to institutions and institutional analysis Even if scholarswere to agree with North that rules and norms are institutions theywould not necessarily agree on what a rule is Shimanoff for examplehas identied more than 100 synonyms for the concept lsquorulersquo (Shimanoff1980 57 Ostrom 1986 5)Another critical issue in the institutional literature is the relationship

between institutions and organizations North (1990) following from hisdenition of institutions argues that institutions and organizations aredistinct entities Which organizations come into being and how theyevolve through time is inuenced by a societyrsquos rules and norms thatis by its lsquoinstitutionsrsquo On the other hand a number of recent organiza-tional sociologists see very little difference between institutions andorganizations For them rules and norms are institutions which unfoldin tandem with organizational structures and processes and changes inorganizational forms internalize and reect changes in the societyrsquos rulesand norms Using this perspective a whole subdiscipline within soci-ology called the lsquonew institutionalismrsquo in organizational analysis hasemerged The lsquoinstitutionalistrsquo perspective on organizations assumes thatthe kinds of organizations which actors create are dictated by the culturalnorms and rules in which they are embedded (Powell and DiMaggio1991 Zucker 1988 1991)The importance of this disagreement about institutions is obvious If

institutions are so critical for understanding our societies it is impor-tant that we come to some systematic consensus as to what institutionsare and how they inuence social actors and the organizations that theycreate If we cannot do so we risk talking past one another and losingthe opportunities for cumulative knowledge based on articulation ofwidely shared theoretical understandingWe need not only conceptual clarication as to what institutions are

but also greater consensus as to how to study institutions No scientic

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

599

eld can advance very far if the practitioners do not share a commonunderstanding of the key concepts used in their analysis (Ostrom 19864) But with our universities and academic associations so fragmentedinto different disciplines and into various subspecialties within disci-plines it is difcult to advance the theoretical agenda of institutionalanalysis within the academy Indeed the disciplinary fragmentation ofthe modern university is a major barrier to the theoretical advancementof the study of institutions and innovations as well as most other hybrid elds of research (Hollingsworth 1984 Hollingsworth andHollingsworth 2000) And it will be only as a result of effective commu-nication across diverse elds of knowledge that our study of institutionsand innovativeness will be effectively advanced

MULTIPLE LEVELS OF INSTITUTIONAL ANALYSISAND INSTITUTIONAL CHANGE

There are innumerable signs that we are living in a time of great insti-tutional change the demise of the Soviet empire the processes ofEuropean political and economic integration rapid transformation inparts of the global economy the disintegration of the family structurethe weakening of voluntary associations and the decline in politicalparticipation in a number of advanced capitalist societies the weakeningof welfare states European law is superseding national law and is evenchanging complete national legal systems The list could go on and onEven though scholars discuss institutional change at length their

ability to measure the rate of institutional change is very limited Andmore crucial than the limited ability of scholars to measure institutionalchange is their very limited understanding of how to build new insti-tutions One of the reasons for these shortcomings is that the socialsciences are decient in a theory of institutions The building of newinstitutions and redressing the decline of some of the most importantinstitutions of our societies are among the most important problems ofour time If we are to advance in the development of a theory of insti-tutions we need to work collaboratively across the social sciences andwe need to dene the parameters of institutional analysis

First level of analysis

This article attempts to make some modest contribution to outlining theparameters of institutional analysis At the outset we need to recognizethat when we engage in institutional analysis we must be sensitive to multiple levels of reality As suggested above most scholars whoengage in institutional analysis do not participate in any coordinatedactivity with each other and their activity is fragmented into a variety

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

600

of disciplines and subdisciplines To establish some coherence to the eldof institutional analysis we need a map of the eld so that those workingin one area can see where their research ts in relation to other areasand other practitioners on the mapTable 1 presents such a map with multiple levels at which institu-

tional analysis occurs Theoretically each of these areas on the map isinterrelated with each other level However the various areascompo-nents on the map are arranged in descending order of stability orpermanence Those components at higher levels of reality are morepermanent and durable while those at lower levels change more rapidlyWere scholars doing institutional analysis able to reach some consensus

about where their own work ts in relation to all other practitioners inthe eld there would be increased potential for all practitioners tocommunicate with each other By analogy once geneticists crystallo-graphers biochemists etc had a good understanding of how theirinvestigations were related to each area of molecular biology the eldquickly was able to make theoretical advances (Judson 1979)At the rst level there are the basic norms rules conventions habits and

values of a society These are the most fundamental properties of insti-tutions and are the most enduring and resistant to change Rules normsconventions etc are institutions but are only one component of whatconstitutes institutional analysis As Burns and Dietz (2001 forthcoming)point out most human activity is organized and regulated by normsand rules and systems of rules These concepts are extremely importantfor institutional analysis as they exert the greatest inuence on the natureof the components of institutional analysis at the next four levels which

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

601

Table 1 Components of institutional analysis

1 Institutions = norms rules conventions habits and values (see North1990 Burns and Flam 1987)

2 Institutional arrangements = markets states corporate hierarchies networksassociations communities (Hollingsworth and Lindberg 1985 Campbell et al 1991 Hollingsworth et al 1994 Hollingsworth and Boyer 1997)

3 Institutional sectors = nancial system system of education businesssystem system of research (Hollingsworth 1997)

4 Organizations (Powell and DiMaggio 1991)5 Outputs and performance = statutes administrative decisions the nature

quantity and the quality of industrial products (Hollingsworth 1991 1997)sectoral and societal performance (Hollingsworth and Streeck 1994Hollingsworth et al 1990 Hollingsworth and Hanneman 1982)

Note The ve components in this table are arranged in descending order of permanenceand stability That is norms conventions etc are more enduring and persistent than eachof the other components of institutional analysis Each component is interrelated withevery other component and changes in one are highly likely to have some effect in bringingabout change in each of the other components

are depicted in Table 1 In most forms of social analysis it is extremelyimportant that we understand the social and cognitive conditions thatlead to compliance or non-compliance with rules and the conditionswhich lead to changes in rulesThe approach to the study of institutions employed here argues that

norms rules habits conventions and values both reect and shape thepreferences of actors Norms rules habits conventions and values inu-ence who and what are included in different types of decision makinghow information is processed and structured what action is taken(Shepsle 1986 1989) It is through norms and rules that behavior isjudged to be democratic fair or egalitarianBurns and Flam (1987) point out that in any society there are multiple

rule systems Within a family there are rules for decision making oftenquite different from rules and norms for decision making for a professorin a classroom or for the customer in a bank Despite the heterogeneityof rule systems there are meta rules and norms which encompass lesserrule systems Otherwise there would be such contradictory rule systemsthat a society would be paralyzed The existence of meta rule systemspermits different rule systems to intersect with each other so that ambi-guities can be resolved The greater the pluralism and complexity of asociety the more ambiguity there is about meta rules and norms in asociety and of course all ambiguities never completely disappearOverall the degree to which separate rule systems are interlinked is anempirical problem There are different sectors groups and interestspursuing their own action logic but through higher-order meta princi-ples and rules there can be order consensus and coherence in a societyIt is through a set of meta rules that class and ethnic conict in societiesare contained (For elaboration see Burns and Flam 1987)In many respects our understanding about norms rules habits con-

ventions and values inuences our perspective on how societies are con-structed and how they change lsquoNew institutionalistsrsquo (Posner 1992Schotter 1981 Williamson 1975 1985) tend to assume that at one timethere was a state of nature and that there was a movement from individ-uals to institutions ndash an approach often called methodological individual-ism (Popper 1961 Hodgson 1998 1999) And of course there areinnumerable instances which methodological individualists cite to demon-strate that individuals create new rules of behavior For example it is pos-sible for actors to change the rules of driving so that instead of drivingon the left side of the road drivers adopt a new rule and drive on the rightThis article however tends to equate social habits and institutions

As Hodgson and others remind us (Hodgson 1988 1989 1997 1999Grafstein 1992 Camic 1986 Johnson 1992 Nelson and Winter 1982Veblen 1899) social habits are the results of earlier choices and are ameans of avoiding endless deliberation Because cognitive frameworks

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

602

are learned through habit individuals rely on the acquisition of suchcognitive habits before reason communication choice or action arepossibleWhereas Schotter (1981) and other game theorists take the individual

as an agent unencumbered by previous habits Field (1984) and othershave stressed that there can be no games without prior norms and rulesand thus a set of norms and rules must be presumed at the start Thosewho attempt to explain institutions from individual behavior alone areusing a bad strategy (Hodgson 1998)The position here ndash heavily inuenced by Hodgson (1998 1997 1988)

ndash is that individuals are embedded in a complex institutional environ-ment and that institutions not only constrain but also shape individuals(also see Hollingsworth and Boyer 1997) It would be a mistake howeverto get involved in an innite regress in order to determine which camerst ndash individuals or institutions Of course institutions are formed andchanged by individuals just as individuals are shaped and constrainedby institutions But at a macro level it is institutions that provide acognitive framework whereby individuals can cope with their reality Inthis sense the micro and macro worlds are intertwined At the macrolevel there is considerable stability but at the micro level individualshave a signicant level of autonomy and there can be widespread diver-sity As Hodgson (1988) reminds us most institutions in a temporal senseexist prior to the individuals in any given societyIt would be a serious mistake to downplay the importance of indi-

viduals and micro level analysis as we study institutions In the nalanalysis it is at the level of individuals that norms rules habits conven-tions and values exist An individual is born into and socialized intogroups and a society and this is how one early in life acquires a senseof appropriate forms of behavior Because of the way that individualsare socialized into a world of rules norms habits conventions andvalues it is unnecessary for individuals to restructure the world anewevery day (Douglas 1987 Elster 1989) Every action does not have tobe seriously reected upon For this reason institutions provide cogni-tive frameworks for individuals make their environments predictableprovide the information for coping with complex problems and envi-ronments In the words of Johnson (1992 26) lsquoInstitutions reduceuncertainty coordinate the use of knowledge mediate conicts andprovide incentive systems By serving these functions institutions providethe stability necessary for the reproduction of societyrsquo However eachsociety has different forms of habits rules and norms and hence differentincentives and disincentive systems for learning and forgetting forprocessing information But because individuals have varying degreesof autonomy individuals and groups can deviate from the prescribedforms of behavior in a society And of course these changes at the level

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

603

of individuals become important in understanding processes of societalchangeThese views are not meant to imply that the type of institutional

analysis proposed herein approximates a general theory of society Thisis clearly not the case However it is intended to represent the rst stepsin a mapping exercise of the boundaries of institutional analysis and tosuggest a few methodological insights for studying institutionsOne should try to see norms and rules as continuous and not as

dichotomous entities to recognize that they come in varying strengthsLegro (1997) has suggested that we assess the robustness of norms andrules with three criteria their simplicity their durability and their concor-dance Simplicity refers to how well actors understand norms and ruleshow well they can be applied within a specic situation Some normsand rules can be so complex that actors can have considerable difcultyin applying them in specic cases Durability addresses the issue of howlong norms and rules have been in effect ndash in short in order to assesstheir level of legitimacy While the position of this article is that normsrules and values are quite durable they do vary in this respectConcordance refers to how widely applied a norm or rule is Thisaddresses the degree to which a rule is a meta rule the degree to whichit incorporates the heterogeneity of other norms and rules In sum theclearer the norms and rules of a society the longer they have been inexistence and the more widely applicable they are the greater theirimpact on a society Hence the more robust the norms and rules thegreater their impact on a society and the less their robustness the greatertheir exibility and the less their effect on shaping a societyrsquos outcomesand performancesBecause norms rules and values are quite durable they play an impor-

tant role in shaping the history of societies thus contributing to a greatdeal of path-dependency Actors attempt to adjust to their contempo-rary environment but since they are products of the past the historicallegacy of norms rules and values inuences the decisions they makeAlthough actors have some capacity to alter the course of their historythey are constrained by their past and the degree to which they canmove beyond their past is limited As Lanzara (1998) Johnson (1992)and others have argued societal inertia is a basic feature of institutionsThey provide the basic stability necessary for change Degrees of historyare continuously reproduced by the way in which the inhabitants ofsocieties are socialized History matters but at critical points in historythere is punctuated equilibrium (Somit and Peterson 1992) During mostperiods of history there is considerable stability in the norms rules andvalues of a society but at critical moments norms rules and values canquickly and dramatically be redened At all times norms rules habitsconventions and values are inuencing each of the other components in

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

604

the institutional framework discussed below but these other institutionalcomponents feed back and can modify rules norms conventions etc(Murmann 1998)

Second level of analysis

The norms rules habits conventions and values of a society lead to thenext level of analysis ndash the institutional arrangements which are involvedin the coordination of various economic actors producers and suppliersof raw materials knowledge etc processors of raw materials infor-mation workers customers of raw materials nished productsinformation etc nanciers governmental and other types of regulatorsThese actors regularly engage in contests to resolve various economicproblems in virtually all sectors of society how are prices to be setWhat quantity of various products is to be produced How are stan-dards of various products processes etc to be set How is the qualityof products and processes to be determined How are various societalprocesses to be nanced In order to confront these problems and toaddress the conicting positions of economic actors as they address these problems societies develop various institutional (governance)arrangements for coordinating different actors These consist of marketsvarious types of hierarchies and networks associations the state commu-nities and clans (see Hollingsworth and Boyer 1997 Campbell et al1991 ch 1)Each of these particular kinds of coordinating mechanism has many

different types For example there are many types of states (eg theregulatory state developmental state authoritarian state welfare state)on which there is an extensive literature (Kim 1997) Similarly there aredifferent types of markets networks different kinds of associations etc(Boyer 1997 Hage and Alter 1997 Schneiberg and Hollingsworth 1990)When we do institutional analysis we must engage in congurativeanalysis recognizing that actors are not coordinated or governed by asingle type of institutional arrangement Some of the literature discussesindustrial sectors as though they were coordinated or governed by asingle institutional arrangement whereas in fact each sector of aneconomy is coordinated by a conguration of institutional arrangementsSome congurations coordinate actors in certain problem areas whileother congurations of institutional arrangements coordinate actors inaddressing other problems The types of congurations which are domi-nant in a society are somewhat stable and tend to persist over timewithin a society (Hodgson 1999)When one mode of coordination is dominant in a society it will

inuence the role which other coordinating modes will play Hence thestrong role of the state in the Soviet Union inuenced the role of markets

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

605

associations etc in the governance of the Soviet economy Similarly theprominence of particular modes of coordination in a society inuencesits style of innovationMuch of the literature on institutional arrangements (eg forms of eco-

nomic coordination) remains fragmented and unintegrated It is helpfulto array these various forms of coordination in a two-dimensional tax-onomy as in Figure 2 On the vertical dimension the economistrsquos viewof a self-interested agent is contrasted with a more sociological perspec-tive according to which social rules obligation and compliance shapehuman actions On the horizontal dimension there is another continuum(ie the distribution of power) At one extreme of the dimension onends many and relatively equal agents interacting (eg as in a well-organized spot market) At the other extreme inequality in power resultsin a hierarchical form of coordination which structures the interactionbetween principals and agents or between leaders and followers Wherea society falls along the distribution of power is inuenced by the rulesnorms values etc which are dominant in a society at a particular momentin timeInstitutional arrangements can be visualized on two dimensions the

nature of action motive on the one hand and the distribution of poweron the other Markets (cell 1) combine self-interest with horizontalcoordination transactors and they reect sensitivity to concerns aboutsupply and demand thus providing ex post an unintended equilibriumParadoxically the more pure and perfect the market competition thegreater the need for codied rules of the games for coordinating economictransactions Thus collective associations (cell 6) andor various formsof state intervention (cell 4) are required to develop and enforce rulesfor transacting partners (Schneiberg and Hollingsworth 1990 Streeckand Schmitter 1985a) This is an example of how the norms and rulesand the institutional arrangements of a society are intertwined asreected in a conguration of institutional or governance arrangementsAlong the horizontal axis actors can be joined in an organization or

a rm a hierarchy is the generic terms for this institutional arrangement(cell 2) Along the horizontal line one recognizes the difference betweentransactions in a market and transactions within a rm The well-knownworks of Coase (1960 1981) and Williamson (1975 1985) utilize theconcept of transaction costs in explaining the emergence of corporatehierarchiesThere are also various types of networks (cell 5) which exhibit mixes

of self-interest and social obligation with some actors being formallyindependent and equal Yet in some networks (the large rms and theirsubcontractors) there is unequal power and initiative Networks maycomprise all kinds of actors some consist only of rms but others includeassociations and the state (Hage and Alter 1997)

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

606

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

607

DISTRIBUTION OF POWER

VERTICALHORIZONTAL

1 MARKETS 2 HIERARCHY

6ASSOCIATIONS

5NETWORKS

4STATE

3COMMUNITIES

SELF INTEREST

SOCIAL OBLIGATION

ACTION MOTIVE

Figure 2 A general taxonomy of institutional arrangements

The vertical axis deals with action motives Toward the upper part ofFigure 2 actors are engaged in individualistically oriented behaviorwhereas toward the lower part actors are more engaged in collectivebehavior and strive to cope with problems of common interest Cell 3 ndashcommunities and clans ndash consists of institutional arrangements based ontrust reciprocity or obligation and thus not derived from the pure selshcomputation of pleasures and pains This is an unconventional form ofcoordination for most neoclassical economists (however see Arrow1974) but not for many anthropologists political scientists and sociolo-gists (Streeck and Schmitter 1985a Polanyi 1944 Gambetta 1988Fukuyama 1995 Sabel 1992 Putnam 1993)In the neoclassical paradigm theorists argue that actors engage in

forms of exchanges that best promote their individual interests If somestructural conditions are fullled (absence of increasing returns to scalethe reversibility of transactions absence of uncertainty and completecontingent markets with no collusion between actors) then the invis-ible hand theorem applies and market-type activity functions quite welland also provides the optimum for society thereby combining efciencyharmony and order However an excess of market activity may welllead to ruinous competition and excessive conict There is variation inthe extent to which ruinous competition occurs depending on the socialcontext in which transactions take place Thus it is important that webe sensitive to the institutional context in which transactions areembedded and that we understand the degree to which social bondsexist at both the micro and macro levels of analysis Micro bonds facil-itate exchanges in a society but at the societal level social bonds existat the level of the collective ndash in the community or region and amongmembers of racial religious and ethnic groups All other things beingequal the more powerful the social bonds among transacting partnersthe more economic competition is likely to be restrained Thus mosttransactions occur not simply in an impersonal calculative system ofautonomous actors unrestrained by social ties ndash as implied by theneoclassical paradigm ndash but in the context of social ties variation in the strength of which leads to variation in levels of trust and transac-tion costs (Etzioni 1988 211 Granovetter 1985 Streeck and Schmitter1985a Hollingsworth and Boyer 1997 Hodgson 1988 1999 Schneibergand Hollingsworth 1990)Another form of multilateral institutional arrangement is various

types of associations (cell 6) Unlike networks clans and communitiesassociations are more formal organizations Whereas markets corporatehierarchies and networks tend to coordinate economic activity amongdifferent types of actors (eg producers with suppliers capital withlabor) associations typically coordinate actors engaged in the same orsimilar kinds of activities Business associations and labor unions are

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

608

some of the most common forms of associations for coordinatingeconomic activity in capitalist economies (Schneiberg and Hollingsworth1990 Streeck and Schmitter 1985a)Finally there is the state (cell 4) which is an institutional arrangement

quite unlike any of the others It is the state that sanctions and regu-lates the various non-state coordinating mechanisms that is the ultimateenforcer of rules of the various mechanisms that denes and enforcesproperty rights and that manipulates scal and monetary policy Overallit is the state that inuences the total incentive system of a society(Lindberg and Campbell 1991 Johnson 1992 40) At the same time thestate may also be an economic actor by engaging directly in productionand exchange relationsThe choices of institutional arrangements in Figure 2 are constrained

by the social context in which they are embedded Depending on thenature of that embeddedness there is variation in the collective formsof governance some of which are specied in the lower part of thegure Some modes of coordination in the bottom and upper levels ofthe typology are often mixed together though one coordinating modeis likely to be more prominent than another But actors are often engagedin complex congurations involving several kinds of institutionalarrangementsEach of these various institutional arrangements has its own logic ndash

its own rules its own procedures for enforcing compliance its ownnorms and ideologies which help to reduce the costs of enforcementThese are summarized in Table 2 which provides further elaborationabout the various coordinating mechanisms one nds in almost everycapitalist society Table 2 lists the organizational arrangements and theirstructures rules of exchange and means of enforcing compliance asso-ciated with each type of coordinating mechanism While each type ofinstitutional arrangement has various positive features each institutionalarrangement also has failures and these are featured in Table 3 It is thecontest between those who support and those who oppose these variousinstitutional arrangements that tends to lead to transformations in insti-tutional arrangements over time (Campbell et al 1991 Campbell 1997)Again it is difcult to conceive of pure institutional arrangements ndash

either exclusively markets or exclusively hierarchies ndash since it has beenwell known since Adam Smithrsquos Wealth of Nations that the division oflabor within the rm cannot be disentangled from the existence andextent of the market Institutional arrangements have their own distinc-tive form of efciency (good static efciency for the market dynamicefciency for rms) and inefciency often leading to considerableinequality Neither networks nor communities are panaceas for economiccoordination without being congured with other types of institutionalarrangements Networks and communities may solve certain issues but

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

609

Table 2

Log

ics of ins

titution

al arran

gemen

ts

Coordination

Organizational

Rules of

Individual means

Collective means

mechanisms

structure

exchange

of compliance

of compliance

Marke

tsEasy en

try an

d exit

Voluntary sp

ot excha

nge

Leg

al enforcemen

t of

Norm of pr

ivate

Bila

teral ex

chan

ge or

control

prope

rty

marke

t site (W

all Street)

Reg

ulations

to

Leg

itim

acy of m

arke

t en

force contracts

men

tality

Com

mun

ities

Inform

al m

embe

rship

Voluntary exch

ange

based

So

cial norms an

d m

oral

Highly institutiona

lized

ev

olve

s ov

er lon

g period

on social solid

arity an

d

princ

iples im

pose

norm

s an

d rules requ

ire

of tim

ehigh

deg

ree of tru

stob

ligations

mem

bers to accept

Kno

wledg

e of others an

d

lsquocorpo

ratersquoo

bligations

recipr

ocity ov

er tim

e

Networks

Semi-form

al m

embe

rship

Voluntary exch

ange

ove

r a

Con

tractual bon

dsPersona

l relation

sBila

teral or m

ultilateral

time pe

riod

Resou

rce de

pend

ence

Tru

st built outside

the

ex

chan

geecon

omic arena

Associatio

nsFo

rmal m

embe

rship

Restricted to mem

bers

Self-interest

Some deg

ree of

Multilateral excha

nge

Emph

asis on insider

Rep

utation

effects

compu

lsion

outsider or we

they

Priva

te interest type

of

men

tality

gove

rnan

ce

Private

Com

plex organ

izations

Restricted to mem

bers

Rew

ards to ind

ividuals

Highly institutiona

lized

hierarch

ies

which

ten

d to becom

e ex

chan

ge based

on

Asymmetric po

wer threat

rules

bureau

cratic hierarchies

asym

metric po

wer

of san

ctions

Mem

bers soc

ializ

ed into

bureau

cratic rules

corp

orate cu

lture us

e of

sanc

tions

State

Pub

lic hierarchy

Unilateral action

Exit vo

ice (vote lob

bying)

Coe

rcion

De jure

and impo

sed

Indirect political an

d

loya

ltyNorms an

d pu

blic rules

mem

bership

econ

omic excha

nge

Table 3

Failu

res of coo

rdinating mecha

nism

s

Coordinating mechanisms

Type of failure

Associations

Private hierarchies

State

Market

Communities

Networks

Enforcem

ent

Usu

ally relies on

Might enh

ance

Needs co

ntrol

Needs an

internal

Needs

tru

st and

Need an external

the state as an

oppo

rtun

istic

extern

al to state

enforcem

ent

loya

lty often

en

forcem

ent

enforcer

beha

vior

bureau

cracy

authority

coming from

au

thority

Resem

bles

The

ideal o

f (ju

dge

s

Facilitates

outside (fam

ily May

facilitate

enforcem

ent

internal m

arke

ts

parliamen

t collu

sion

and

relig

ion

cartelization an

d mecha

nism

might hur

t marke

t) to

impe

rfect

ethn

icity)

mon

opoly

of cartels

incu

mbe

nt

correct state

compe

titio

nCom

patible with

worke

rsab

uses

variou

s types of

Lob

bies can

co

mpe

tition

captur

e pu

blic

interest goa

ls

Public goo

d

Useful for

Gov

erna

nce co

sts

Can

provide

Can

not provide

Can

internalize

Useful for

and

establishing

might exceed

public goo

ds

colle

ctive go

ods

some co

llective

enha

ncing qu

ality

externality

stan

dards an

d

the be

nets of

but ha

s or deal w

ith

good

s (qua

lity

and training bu

t qu

ality for

internal division

difc

ultie

s in

externalities

training

) bu

t no

t no

t ve

ry goo

d in

setting ru

les of

of la

bor

prov

iding them

Inad

equate

othe

rs (welfare

prov

iding for

compe

titio

n in

Slow

to react to

in precise

mon

itoring of

gene

ral p

ublic

societal gen

eral

the indu

stry

chan

ges in the

am

ounts

tech

nical ch

ange

go

ods)

welfare

Useful for

environm

ent

Might fail in

and inno

vation

Mem

bers tightly

Weak in the

prov

iding

induc

ing

integrated

into

prov

ision of

man

y go

ods

tech

nical

community

colle

ctive go

als

colle

ctively

chan

geha

ve limite

d

that in

dividua

lcapa

city for

mem

bers can

not

inno

vatio

nsprov

ide for

them

selves

Table 3

continue

d

Coordinating mechanisms

Type of failure

Associations

Private hierarchies

State

Market

Communities

Networks

Efcien

cyFa

cilitates

De

cien

t in

Can

be high

ly

Some ba

sic social So

me go

ods

Slow

to en

hanc

e co

operation an

d

coop

eration an

d

bureau

cratic

relation

s cann

ot

cann

ot be

efc

ienc

y an

d

X-efcien

cy but

X-efcien

cyan

d can

not

be provided

by

delivered

at

speed of

not allocative

easily deliver

pure m

arke

t su

fcien

tly

adap

tive

ness

efcien

cy

good

s at

mecha

nism

slow costs

except in

low cost

indus

tries whe

re

tech

nology

is

complex an

d rapidly cha

nging

Equity

Narrow

Excessive

Might enh

ance

Facilitates

Might lead to

Whe

n widely

encompa

ssing

multiplication

ineq

uality

ineq

uality in

retard

ed

dev

elop

ed in

to

associationa

l of con

trollers

(pow

er

income an

d dev

elop

men

tindus

trial

stru

ctures lead

(fru

stration

and

an

d privilege

)wealth

districts

to in

come

ineq

uality)

netw

orks m

ay

equa

lity

facilitate greater

equality an

d

inco

me

distribution

whe

n wea

kly

dev

elop

ed

netw

orks ten

d to

increase soc

ial

ineq

uality

Note For allo

cativ

e efc

ienc

y an

d X

-efcien

cy see Leibe

nstein (19

66 19

76)

they raise other no less severe problems It is important to recognizethe imperfection of any single institutional arrangement in order tocomprehend the origin and transformation of any other institutionalarrangement (Hodgson 1999 ch 3)As suggested above each institutional arrangement is congured with

other institutional arrangements Usually one particular institutionalarrangement is more dominant in a particular conguration but becauseeach type of institutional arrangement has its own strengths andweaknesses there is no simple structural logic in the governance or coor-dination of a society Each institutional arrangement constrains the otherbut the inherent tension among the various institutional arrangementswithin a conguration contributes to changes in the governance orcoordination of a society The routines and logics of each of these cong-urations of institutional arrangements provide constraints and incentivesfor actors The way that actors perceive the incentives and constraintsof these governance congurations leads to a particular market logicand it is the specic market logic of a society which inuences its speciccapacities and weaknesses The inherent strain among the different logicsin a conguration of coordination helps to provide the exibility for asociety to adapt to new circumstancesIf there were a society with pure markets or a society coordinated

only by the state there would be too much rigidity and too little diver-sity to cope with the vast uncertainty in the global environment A societywith very little diversity in its coordination mechanisms would havelittle capacity to adapt to new circumstances Congurations with consid-erable diversity of institutional arrangements provide for a certainamount of incoherence in governance but they also provide for thecapacity to adapt to new circumstances While the Soviet Union wasdominated by the state as a coordinating mechanism there were alwaysfunctioning markets in the system Moreover the feudal society of thetwelfth and thirteenth centuries consisted not only of hierarchical rela-tionships between serfs and masters but also of urban guilds clericalhierarchies and markets (Johnson 1992 38) In sum the robustness ofinstitutions often depends on multiple and diverse principles and logicsof action on the inconsistency of principles and procedures on patternedforms of disorder (Lanzara 1998 Orren and Skowronek 1991 320 329)

Third level of analysis

The next level consists of the institutional sectors of a society The rulesnorms and values of a society inuence the array of institutional arrange-ments and levels one and two inuence the nature of and the relation-ships among various institutional sectors and all three levels areintricately linked together to form a social system of production Together

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

613

all of these components inuence the performance of economic sectorswithin a society as well as the performance of the total society A socialsystem of production is the way that a societyrsquos institutions (see rstlevel of analysis above) its institutional arrangements (see second levelof analysis above) and its institutional sectors are integrated into a socialcongurationAn institutional sector includes all organizations in a society which

supply a given service or product along with their associated focal orga-nizations (eg major suppliers funding sources regulators and so forth(Scott et al 1994 108 117 Campbell et al 1991 Hollingsworth et al1994) Institutional sectors include but are not limited to the societyrsquossystem of education system of research business system nancialmarkets legal system and the state The structure of the nancial marketssystem of training and education industrial relations system system ofresearch and state and legal systems are distinct and idiosyncratic ineach society In short they are system specicAll of these institutional sectors tend to cohere with each other

although they vary in the degree to which they are tightly coupled intoa full-edged system While each of the institutional sectors has someautonomy and may have some goals that are contradictory to the goalsof other institutional sectors with which it is linked an institutional logicin each society leads institutions to coalesce into a complex social con-guration (Hollingsworth 1991) This occurs because the institutionalsectors are embedded in a culture in which their logics are symbolicallygrounded organizationally structured technically and materially con-strained and politically defended The conguration of institutionalsectors usually exhibits some degree of adaptability to new challengesbut continues to evolve within an existing style But under new circum-stances or unprecedented disturbances these institutional congurationsare exposed to sharp historical limits as to what they may or may notdo (Schumpeter 1983 David 1988 Arthur 1988a 1988b Haringkanssen andLundgren 1997)Why do all of these different institutions coalesce into a complex

social conguration which is labeled here as a social system of pro-duction The literature suggests two contrasting interpretations Part ofthe answer ndash indeed a controversial one ndash is that these institutionalsectors are functionally determined by the requirements of the practiceof capitalism in each time and place (Habermas 1975) Another expla-nation emphasizes the genesis of the actual conguration via a trial and error process according to which the survival of rms regions orcountries is the outcome of complex evolutionary mechanisms (Maynard-Smith 1982 Nelson and Winter 1982) However the problem is evenmore complex Markets and other mechanisms for coordinating rela-tionships among economic actors place constraints on the means and

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

614

ends of economic activity to be achieved in any society The othercoordinating mechanisms include different kinds of hierarchies varioustypes of networks and associations ndash trade unions employers and busi-ness associations (Hollingsworth and Lindberg 1985 Campbell et al1991) The logic of these various institutional arrangements providesactors with vocabularies and logics for pursuing their goals for deningwhat is valued and for shaping the norms and rules by which theyabide (see Tables 2 and 3) In short in contrast to the logic of theneoclassical paradigm the argument here is that the dominant type ofinstitutional arrangements places severe constraints on the denition of needs preferences and choices of economic actors Whereas the neo-classical paradigm assumes that individuals and rms are sovereign thisarticle is based on the assumption that rms are inuenced by the hold that the institutional congurations making up a social system ofproduction have on individual decision making (Campbell et al 1991Etzioni 1988 Streeck and Schmitter 1985a Hollingsworth et al 1994Hollingsworth and Boyer 1997 Magnusson and Ottosson 1997 North1990 Hodgson 1999)Standard neoclassical economic theory has tended to downplay the

role of production and consequently of rms Even transaction cost theo-rists who are concerned with analyzing the rm as a coordinatingmechanism have been relatively unconcerned with the various compo-nents of a social system of production Indeed as long as there waswidespread optimism about the efcacy of Keynesian economics therewas relatively little concern among neoclassical economists with thesupply side of the economy Even in the opinion of most Keynesians agroup of experts should ideally be able to shape the size of aggregatedemand while the supply side of the economy would be left to the twominimalist institutions of neoclassical economics ndash markets and manage-rial hierarchies For several decades however it has become increasinglyobvious that some of the most competitive and successful patterns ofindustrial output and industrial production in capitalist economies donot derive from the neoclassical prescription of unregulated markets andcorporate hierarchies complemented by a neoliberal democratic stateIndeed it is now well understood that certain highly successful produc-tion patterns require for their emergence and survival institutionalarrangements and environments the very opposite of the prescriptionsfound in the neoclassical paradigm (see especially Streeck 1992 but alsoHollingsworth and Streeck 1994) Thus social scientists have come tounderstand that if they are to comprehend the behavior and performanceof contemporary economies concerns about social systems of produc-tion must be brought into the pictureBecause production involves more than technology a number of social

scientists have an increasing concern with social systems of production

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

615

The same equipment is frequently operated quite differently in the samesectors in different countries even when rms are competing in the same market (Maurice et al 1980 Sorge and Streeck 1988 Sorge 1989Hollingsworth et al 1994) Variations in production and process tech-nologies are inuenced partly by variations in the social environmentsin which they are embedded In other words rms are embedded intocomplex environments which among other things place constraints ontheir behavior Thus a social system of production is of major impor-tance in understanding the behavior and performance of an economyHow the state and other institutional arrangements (eg marketsnetworks private hierarchies associations) coalesce and are related toparticular social systems of production are important determinants ofeconomic performanceThe congurations of institutional arrangements that coordinate or

govern the behavior of actors in one society and its structure of specicinstitutional sectors cannot easily be transferred to another society forthey are embedded into a social system of production that is societallydistinct (Hollingsworth 1997) Societies borrow selected principles offoreign management styles and work practices but the effectiveness ofsuch borrowing is generally limited Economic behavior and performanceare shaped by the entire social system of production in which actors areembedded and not simply by specic principles of particular manage-ment styles and work practices Moreover a societyrsquos social system ofproduction tends to limit the kind of goods that it can produce and withwhich it can compete successfully in international marketsSome scholars (Kenney and Florida 1993 Oliver and Wilkinson 1988)

have assumed that the diffusion of particular forms of management stylesand work practices across societies could lead to system convergenceReferring to the Japanese production system Kenney and Florida haveargued that it consists of organizational practices whose fundamentallsquogenetic codersquo can successfully be inserted into another society and beginsuccessfully to reproduce its behavior in the new environment (1993 8)Their position is in the intellectual tradition of Antonio Gramsci (1971)who contended decades ago that the American system of mass produc-tion would most certainly diffuse to Europe over timeHowever the argument of this article is that even though British

French and American rms may adopt certain aspects of Japanesemanagement styles (eg just-in-time production self-managing teamsquality circles the use of statistical process controls) or some variant ofthe German vocational system their social systems of production willnot converge Moreover the overall conguration of a societyrsquos socialsystem of production inuences its sectoral and overall national economicperformance This helps to explain why societies have different systemsof innovation and why they vary in the clusters of industries in which

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

616

they are highly competitive in the international markets (Nelson 19931999 Hage and Hollingsworth 2000 Berger and Dore 1996)In order to understand how and why a societyrsquos economy performs

as it does it is necessary to understand its entire social system of produc-tion If a society is to modify substantially the performance of itsinstitutional sectors and its entire economy it cannot adapt only someof the management and work practices of its foreign competitors Ratherit must alter its entire social system of production Because a societyrsquosmodes of economic governance and coordination and its institutionalsectors develop according to particular logics and are system specicthere are serious limitations to the extent to which a society may mimicthe institutions (eg the rules and norms) institutional arrangements andinstitutional sectors of other societiesIn the history of modern societies there are logics by which institu-

tions coalesce into a social system of production (Hollingsworth 1991)Though institutions are constantly changing there are sharp limits tothe type and the direction of change that any particular institution canundergo because of its linkages with institutional arrangements and insti-tutional sectors Thus a societyrsquos business rms educational systemnancial markets industrial relations system etc can engage in seriousrestructuring only if most of the other institutional sectors also changeIn social systems of production there are pressures toward consistencyin the norms rules and values across institutional sectors though in anycomplex society social systems are obviously imperfectly integratedIndeed the degree to which the institutional norms and rules makingup a social system of production are loosely or tightly coupled is a vari-able of considerable importance In general the institutional sectorsmaking up a social system of production are interdependent and changesin one generally result in changes in the others Since each institutionallevel is dependent on the others for various types of resources there isinterdependence among the differing institutional spheres Moreovereach society has its norms moral principles rules and laws and recipesfor action as well as its own idiosyncratic customs traditions and prin-ciples of justice (Burns and Flam 1987)There are also other inherent obstacles to convergence among social

systems of production for where a system is at any one point in timeis inuenced by its initial state Systems with quite different initial statesare unlikely to converge with one anotherrsquos institutional practicesExisting institutional arrangements block certain institutional innova-tions and facilitate others (Roland 1990) Thus the institutions makingup a social system of production provide continuity even though insti-tutional arrangements are always changing but with a particular logicWhile Williamson (1975 1985) suggests that actors tend to choose theinstitutional arrangements which are most efcient North (1990) is

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

617

much closer to the mark in his argument that most societal institutionalarrangements exist as a result of custom and habit and are thereforeinefcient At any moment in time the world often appears to its actorsas very complex and uncertain For this reason actors often engage incontradictory forms of behavior pursuing different strategies as hedgesagainst a very uncertain world (Lanzara 1998) And their hedging andcontradictory forms of behavior may lead in somewhat different soci-etal directions all constrained by the institutional fabric within whichthe actors are embedded This kind of contradictory behavior occurs inpart because of the contradictions inherent in the dominant congura-tions of institutional arrangements with their contradictory logics (seelsquoSecond level of analysisrsquo above)Despite the emphasis on the logic of institutional continuity this is

not an argument that systems change along some predetermined pathThere are critical turning points in the history of highly industrializedsocieties but with the choices limited by the existing institutional terrainBeing path-dependent a social system of production continues along aparticular logic until or unless a fundamental societal crisis intervenes(Milgrom et al 1991 Krugman 1991 Durlauf 1991 Hollingsworth 1991Pred 1966 David 1988 Hodgson 1998)At this point it is important to confront the question of how social

systems of production evolve And this question gets to the heart of theproblem of building complex societal structures Certainly social systemsof production did not emerge from some process of social engineeringMoreover the various component parts of each social system of produc-tion have often not been designed to be part of a social system ofproduction The component parts of each social system of productionhave emerged more often than not as unintended by-products of goalswhich various actors had in mind at earlier moments in time It is usuallythe case that actors at time lsquotrsquo fail to comprehend the long-term conse-quences of their actions Wolfgang Streeck the author of several brilliantpapers on social systems of production argues that institutional sectordesigns created for one purpose generally address that goal but overtime those designs have quite unintended consequences (Streeck 1997a1997b Wright 1998) Describing the conguration of the German socialsystem of production of the 1970s and 1980s Streeck points out that itwas the unintended by-product of multiple points in history Someelements were pre-Wilhelmian others were introduced by the Alliedpowers after 1945 and others emerged during the years of the GermanFederal Republic All component parts of the German social system ofproduction

were and continue to be changing for their own reasons as wellas in reaction to each other and certainly there can be no presump-tion of a pre-established t between them even though one might

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

618

want to allow for some reinforcement effects of the modelrsquos histor-ically contingent social and economic success

(Streeck 1997b 54)

In sum the emergence of social systems of production is a long-termevolutionary process with each part interacting with its environment andresulting in a congurative whole but with no previous design by eithera single actor or a collectivity of actors As suggested above actors do not recreate their world anew At any moment in history the totalinstitutional complex of a society necessarily contains the resources and legacies of its past Preexisting institutional complexes are nevercompletely wiped out Societies consist of multiple layers of history withtheir diverse logics of action However the cumulative effect of smallperipheral changes in altering particular institutional sectors and insti-tutional arrangements can be substantial (Murmann 1998 Lanzara 1998)lsquoInstitution building is affected more by the ways in which people havecodied the past than by how they envision the future There are manymore potential resources in pre-existing institutional arrangements thanit is usually assumed or suspectedrsquo (Lanzara 1998 30)

Fourth level of analysis

The next level of institutional analysis organizational structures is some-what more controversial As suggested above Douglass North (1990)draws a sharp distinction between institutions and organizations Morerecently many organizational theorists (Powell 1991 Powell andDiMaggio 1991 DiMaggio and Powell 1983 Baum and Oliver 1992Kondra and Hinings 1998 Townley 1997) argue that institutional rulesnorms and conventions unfold in tandem with organizational structuresand this is my positionThe literature which focuses on how institutions inuence organiza-

tions is quite different from two other theoretical literatures which alsoare concerned with how organizational environments shape the behaviorof organizations There are resource dependency theorists (Pfeffer andSalancik 1978) who emphasize the role of environmental resources inshaping organizations while Hannan and Freeman (1977 1984) and otherpopulation ecologists emphasize the survival of organizations givencertain kinds of organizational conditions (Orru et al 1991)The institutionalist perspective as implied above emphasizes the

normative environment in which organizations are embedded It is a perspective which focuses on the way in which organizations in their behavior tend to conform to the institutional rules and normswhich are dominant in the organizational environment However allthree perspectives ndash the institutionalist the population ecology and

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

619

the resource dependency ndash emphasize how the environment inuencesorganizations and how organizations which are subject to the sameenvironment tend to converge in their behavior to have what DiMaggioand Powell have labeled organizational isomorphism (DiMaggio andPowell 1983)There has emerged a vast literature which demonstrates that within

each society there are scal political judicial and other regulatory normswhich limit and shape the culture and structure of organizationalbehavior The normative institutional environment of organizations limits the options of what organizations do in a particular society and inuences the patterns of ownership relations with suppliers andcustomers In short it is the normative environment of organizationswhich denes within a particular society what is socially acceptablebehavior for organizations (Hamilton and Biggart 1988 Hollingsworthand Hollingsworth 2000 Meyer and Rowan 1991 Zucker 1987 1988Orru et al 1991 Townley 1997)Thus far most of the analyses on isomorphic pressures operating on

organizations have been cross-sectional in nature However the entireconcept of isomorphism implies that there are strong environmental pres-sures exerted on organizations and in order to observe this phenomenonit is important that we have longitudinal studies which assess howchanges in an organizationrsquos institutional environment inuence changesin the structure and culture of organizations Moreover the expectationhere is that when one analyzes the patterns of organizational changeover time the isomorphism will be holistic ndash that is within a particularsociety there will be a tendency for the internal structure of particulartypes of organizations to converge in large part because of the pres-sures to conform to the changes in the external norms rules and valuesA close reading of Aoki (1990) and Slack and Hinings (1994) nds thatchanges in the national institutional environment of organizations haveinuenced changes in the internal structure of organizations within aspecic industry of a particular societyHowever it is inappropriate to view organizations as changing only

in relation to exogenous environmental change There is a kind of co-evolution which occurs between organizations and their institutionalenvironments Co-evolution implies nonlinear feedback between orga-nizations and their institutional environments This is of course quitedifferent from the more common way of studying organizations bymodeling relations between independent and dependent variables Ofcourse the more common type of analysis is quite appropriate whenthere are simple relationships which do not involve complex feedbackprocesses But in the strategy for institutional analysis which is proposedherein it is less useful to separate independent from dependent vari-ables and more useful to understand the interacting and co-evolutionary

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

620

processes among all of the levels of institutional analysis discussed inthe various sections above (Baum and Singh 1994a 379ndash80 Mowery andNelson 1999 Murmann 1998)The logic of the perspective herein suggests that within every

society there is variation in the structure and culture of business rmsuniversities and other complex organizations (Kondra and Hinings1998) However this variation takes place within parameters which aresystem specic For example every German rm and every Germanuniversity is different from every other German university and rm but there are a set of cultural and structural characteristics which distin-guish German rms and universities from those in America Similarlyeach Japanese rm and university is unique yet they have a set ofcultural and structural characteristics which are system specic andwhich differentiate them from business rms and universities in other countriesIn all societies each organization has its own distinctive organizational

rules norms and conventions which are subordinate to the meta norms and rules of the larger society within which they are embeddedHowever the strength of the institutional environment within whichorganizations are embedded varies from society to society Some societies have multiple institutional environments and there is hetero-geneity in terms of what constitutes appropriate behavior oforganizations In such societies there can be conict over whatinstitutional logics should regulate specic organizational functions(Friedland and Alford 1991 Townley 1997 262ndash3) In those societies inwhich the institutional norms habits and rules are most developed and in which the institutional pressures to conform are greatest thereis less variation in the structure and culture of business rms and variouskinds of research organizations In such societies the connectednessbetween research organizations and their institutional environments is sufciently strong for organizations to have low autonomy to pursue independent strategies and goals in these societies there is a great deal of organizational isomorphism Conversely the weaker the institutional environment in which research organizations areembedded the greater the variation in the structure and culture of business rms and research organizations Moreover where the institu-tional environments are more weakly developed organizations havegreater autonomy and exibility to respond to the development of new knowledge and to be highly innovative Hence it is in those societies where the institutional environments are most developed and most rigid and when there is less organizational autonomy and exibility that fewer radical innovations in basic and applied science as well as in totally new products and industrial sectors havebeen made

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

621

Fifth level of analysis

We come next to the outputs and performance of the various institutionalcomponents of a society It is at this level that institutional componentsare more pragmatic and exible For example in the legal sector thereare specic statutes and court rulings in the state sector there are specicpolicies in the business sector there are new products new technolo-gies and market strategies It is at this nal level of analysis thatinstitutional spheres are most open and susceptible to change and cross-national mimicry is easiest and most commonThrough the outputs of the society we can obtain some assessment as

to how well a society is performing Moreover we can assess how inno-vative it is how egalitarian it is in the distribution of its resources howegalitarian it is in terms of levels of health education etc Just as it iscomplicated to measure how egalitarian a society is similarly it is dif-cult to assess how innovative it is For example societies vary in makingincremental and radical breakthroughs in basic and applied science in developing totally new products and new kinds of organizations inincremental and process innovations in existing products and organiza-tions in developing new and different forms of marketing ndash bothdomestically and globally (Hage and Hollingsworth 2000) Of coursesocieties vary in their rates of economic growth in their rates of economicproductivity in their quality of life (eg rates of crime life expectancyetc) All of these performance criteria feed back and inuence each ofthe levels of institutional analysis discussed above ndash rules norms valuesetc institutional arrangements institutional sectors the structure andform of organizations Moreover performance measures may inuenceeach other For example if a society has very low rates of growth itmay not be very innovative in some of the types of innovative activitiesmentioned above Good or poor performances of certain kinds inuenceother performance indicators (Hollingsworth et al 1990)Different institutional arrangements and different social systems of

production result in different types of economic performance Hence solong as societies have different social systems of production there areserious constraints on the degree to which they can converge in theirinnovative styles Different social systems of production tend to maxi-mize in a more or less explicit manner different performance criteriausually mixing considerations about static and dynamic efciency protsecurity social peace and economic andor political power In short incontrast to the implications of neoclassical economic theory in real-worldeconomies there are no universal standards all economically rationalactors attempt to maximize Economic history provides numerous exam-ples of how a great variety of principles of rationality are implementedin different societies

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

622

A critical question in institutional analysis is whether an existing socialsystem of production which supports a set of routines for a particularkind of technology and industry can shift from old practices to newones There are numerous historical accounts of how the social systemof production in a particular society which worked so well for a numberof years could not adapt to new technologies (Lazonick 1990 Veblen1915 Nelson 1994) Schumpeter (1983) and Freeman (1991) have devel-oped the idea of a techno-economic paradigm which suggests thatdifferent technologies require specic forms of organizational arrange-ments ndash or what I called social systems of production Because oflsquoinstitutional inertiarsquo the social system of production of a society maynot be able to adapt to a new techno-economic paradigm (Nelson 199458 Lanzara 1998)Whether a social system of production can sustain its performance

standards depends not only on its intrinsic economic rationality but alsoon where it ts into a larger system If a particular social system ofproduction is immune from the innovativeness and competition of analternative system survival can be long lasting But if different socialsystems of production with diverging criteria of good economic perfor-mance meet in the world arena the arbitrariness of nationally imposedconstructed performance standards may be superseded by alternativeperformance criteria as a result of international competitivenessThe world economy is also socially constructed just as are national

economies Even if different social systems of production are competingin the international arena it is not always possible to determine whichis more competitively effective at any moment in time Hegemonicnation-states in the short run shape the rules of trade to favor their indus-trial sectors and rms But the history of hegemonic powers suggeststhat in the longer run such social systems of production sustainedlargely by military and political power eventually give way to moreinnovative and competitive social systems of production In our ownday as nation-states are increasingly integrated into a world economyeven hegemonic powers lose their innovative and competitive advan-tage and their share of world output decreases Such a country mayattempt to restructure its institutional arrangements and to readjust itsperformance preferences But to restructure its system of innovationgenerally calls for a major redistribution of power within a societyLargely for this reason societies have historically had limited capacityto reconstruct a system of innovation in the image of their major competi-tors Each countryrsquos social system of production is a conguration of ahost of norms rules and values as well as of institutional arrangementsEach system is constantly changing and is open to inuence from othersystems And indeed many technologies and practices diffuse from onesociety to another but a societyrsquos capacity to be innovative is constrained

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

623

by the existing social system of production Thus the same technologymay exist in numerous countries but how it is employed and how itinuences societal outcomes and performances varies from one institu-tional conguration or society to another (Hollingsworth 1998)

PERSPECTIVES ON INSTITUTIONAL CHANGE

A fundamental problem which remains unresolved in institutionalanalysis involves the nature of institutional change There is much confu-sion and miscommunication in the social sciences about institutionalchange in large part because we do not even know how to measure therate of institutional change Of course we know that there are bothexternal and internal focuses for change and we have had a good bitof scholarship on this subject What is not clearly understood is thatinstitutions and institutional arrangements within societies are histori-cally rooted that there is a great deal of path-dependency to the waythat institutions evolve Moreover the more intricately linked each sectoris with each other sector and with a societyrsquos rules and norms the lesschoice actors have to devise new institutions and institutional arrange-ments On the other hand if a sector or actor is too isolated it may betoo weak within a society to bring about any kind of effective changeThese considerations are very critical to the problem of how muchfreedom actors have to create new institutions afresh to what extentinstitutions diffuse from one society to another (Murmann 1998 Lanzara1998) This is a fundamental question for students of institutionalanalysis as we attempt to understand how the different institutionallevels are linked togetherThere are a variety of reasons why there is confusion in studying insti-

tutional change There is disagreement over the extent to which actorshave the freedom to build new institutional arrangements and if theydo to what extent the new arrangements and organizational patternsmay depart from past practices Among some institutional economists(Williamson 1985) there is a basic assumption that actors build efcientinstitutions by pursuing their self-interest and by promoting theirstrategic goals Moreover there is the assumption that a mixing of fullrationality with market competition tends to produce all of the optimalinstitutions that are needed in order to coordinate a complex capitalisteconomy (Hollingsworth and Boyer 1997 ch 14 Hodgson 1999) Incontrast the argument developed here is much more complexSocial institutions are historically rooted and there is a great deal of

path-dependency in the way that various institutional componentsevolve The shape of institutional congurations at any moment limitsthe type of options for change Because there is a great deal of institu-tional inertia radical change in the institutional components of a society

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

624

is uncommon (Lanzara 1998) Recognizing the constraints existinginstitutions exert on actors we note that existing congurations of insti-tutional arrangements institutional sectors and organizational propertiescan limit but also inuence the degree to which norms and rules canchange Many scholars (North 1990 Ostrom 1990) have emphasized therole of norms rules ideologies and values in limiting the ability of actorsto develop new institutional arrangements which radically depart fromexisting ones However there is a two-way process of interaction aprocess of co-evolution (Baum and Singh 1994a Nelson 1994) Thus asFriedland and Alford (1991 244) argue actors develop new institutionalarrangements by recognizing organizational failures and low perfor-mance but these new arrangements do not necessarily change a societyrsquosnorms rules habits and its underlying value systemCampbell (1997) advances our understanding of how existing institu-

tions enable actors to construct radically new institutions He recognizesthat most institutions and institutional arrangements embody a degreeof inertia from which it is difcult for actors to depart Moreover thedifferential power relations among actors make it unlikely that lesspowerful actors can change the existing power structure of a societyMost change in the institutional components of a society evolves througha process of constrained selection which reects to a considerable degreethe existing arrangements and power relations This becomes the inter-pretative frame for social actors with institutionalized scripts and ritualswhich tend to be taken for granted and appear to be quite ordinary andnatural This institutionalized scripting of the social world is an extremelyimportant source of social stability and inertiaCampbell has developed a very fruitful explanation of change in terms

of the various institutional components outlined here He distinguishesbetween incremental and radical institutional change and argues thatradical institutional change occurs when social actors with widelydiffering norms cultural scripts and rituals for action engage in intenseinteraction with each other (see also Knight 1992) The more diverse theinteractions the greater the potential for institutional change He arguesthat as shifts in the composition of interaction among social actors occurchanges in the interpretation of problems and interests will follow Ifthere are only minor changes in the extent to which diverse actorsinteract there will be only modest institutional change But the morefundamental the changes in the interaction of diverse social actors thegreater the change in the way that actors interpret their world And thegreater the change in the way that actors interpret their world the morelikely that radical changes in various institutional components willfollow In other words diverse decision making tends to create a widerrange of interpretative frameworks of the social world than is likely tobe the case if social actors operate in isolation

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

625

This perspective of diverse interactions among social actors facilitatesour ability to move beyond the traditional view that existing institutionsconstrain the range of institutional alternatives which actors face Theolder view has tended to suggest that actors engage in institutionalchange by extending existing institutional principles habits and conven-tions Specifying the conditions which facilitate fundamental institutionalchange permits a much richer perspective for institutional analysisCampbellrsquos perspective about the conditions under which fundamentalor radical change in institutional components occurs is consistent withmy own research about the conditions under which major discoveriesor fundamental new knowledge occur the more scholars with diversebackgrounds interact with intensity and frequency the greater the like-lihood that they will develop new and alternative ways of thinking(Hollingsworth and Hollingsworth 2000)These views of institutional change are also consistent with the recent

work of Nelson and his colleagues (Mowery and Nelson 1999 Murmann1998) They report the research on a variety of industrial sectors in anumber of different countries revealing that networks of actors in partic-ular industries business and professional associations universityresearch departments and institutes are frequently able to use their collec-tive power to modify a societyrsquos property rights research system andgovernmental policies It is the intense and frequent interaction of thesediverse complex sets of actors ndash sometimes constituting a technologysystem ndash that often leads to incremental changes in a societyrsquos total socialsystem of production It is this kind of collective action that brings aboutorganizational change change in institutional sectors institutionalarrangements and even in the rules and norms of a society And if weare to understand how the style of innovativeness of a society changeswe must rst understand the changes in its institutional makeup It ishighly unlikely that a society will develop a new style of innovativenesswithout changing its institutional structure

FIRMS THEIR INSTITUTIONAL ENVIRONMENT ANDNATIONAL SYSTEMS OF INNOVATION

Variation in the innovative style of societies is shaped by their insti-tutional makeup ndash the various levels which are discussed aboveinstitutions institutional arrangements the structure and cohesion of institutional sectors which constitute a societyrsquos social system of pro-duction and the structure and culture of its organizations ndash especiallyits business rms and other research organizations By undertaking aninstitutional analysis of a society one can begin to understand in whatkinds of organizations the production of specic kinds of knowledgetakes place and how this is linked to particular kinds of innovativeness

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

626

With such a perspective one can gain a rich understanding of why somesocieties excel in the production of radical breakthroughs in basic andapplied science and in developing radically new products and why othersocieties excel in more incremental innovationsReecting on these issues one cannot help being impressed with the

contrasts between the innovative styles of German and American orga-nizations For example Germany has continued to be very successful in making incremental innovations in many industries in which it wasvery competitive well before World War II paper printing materialsmachinery electrotechnical products motor vehicles chemicals textileyarns and fabrics But the Germans have been much less successful indeveloping totally new products or related innovations in biotechnologyelectronics telecommunications aircraft as well as other newer indus-trial areas In contrast the Americans have been very innovative in these and other industries with products having very short lives andwith technologies which change very rapidly and are very complex(Hollingsworth 1991 Soskice 1997 1999)The importance of institutions and innovativeness for international

competitiveness has led to the recognition that there are national systemsof innovation (Lundvall 1992 Nelson 1993 Edquist 1997 Hage andHollingsworth 2000) But even before the recent studies of nationalsystems of innovation Landes (1969) was a proponent of these viewswhen he made the argument that Germany had developed a nationalsystem of innovation by building on its education and scientic researchsystems which were lacking in Britain and elsewhere However the concept of a national system of innovation immediately poses theproblem of what kind of innovations should be consideredDavid Soskice an economist at the Wissenschaftszentrum Berlin has

addressed a set of problems somewhat complementary to the perspec-tive raised here His work (Soskice 1990 1997 1999) suggests that weare sensitive to the institutional environment in which the followingactors are embedded the employees of rms the owners and nanciersof rms the competitors of rms and nally their collaborators ndash aperspective complementary to Hollingsworthrsquos work on social systemsof production (Hollingsworth 1997 1998) When one reects on the insti-tutional contextsocial system of production within which these actorsare embedded one becomes aware that those countries which have some-what inexible labor markets external to rms but a business system in which rms have long-term commitments to their employees arestrongly associated with incremental but hardly at all with radical inno-vations In the rms of such countries (eg Germany and Japan) thereis much more consensus decision making than in societies where a lsquohireand rersquo set of practices is quite pervasive And this kind of consensusdecision making ndash while highly conducive to incremental innovativeness

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

627

ndash limits the capacity for radical innovations to occur This kind of lsquolongtermismrsquo set of norms and rules occurs not only in the employmentpractices of incrementally innovative rms but also in their nancingand ownership patterns (Hollingsworth 1997) During the past thirtyyears these were characteristics of many rms in Germany and JapanSignicantly in both of these countries venture capital markets werenot highly developed nancing tended to be based more on long-termbank loans than on the equity markets and the relationships of rmstended to be relatively stable And these stable institutional arrange-ments also contributed to a high degree of innovativeness ndash butincremental rather than radical innovationsMeantime we should also pursue Soskicersquos ideas (1999) about how

rmsrsquo relationships with their competitors and collaborators shape theirstyles of innovativeness In those advanced industrial societies in whichbusiness rms are highly mobilized into business associations theretends to be a high degree of standardized quality controls over technicalnorms and a high degree of reliance on legal sanctions as monitoringinstruments vis-agrave-vis association members As a result of this high degreeof integration of rms into associations member rms are more likelyto be associated with incremental innovations than is likely to be the case with rms in those societies where associations are less welldeveloped In the latter type of society radical innovations are morelikely to occur In those advanced industrial societies where rms havelong-term stable relationships with collaborators in product design andproduction the style of innovation is likely to be incremental And it is in those societies in which there is a weaker tradition of strongcollaboration among rms that radical innovations are more likely tooccurObviously even in a country where there may be a great deal of

radical innovativeness most innovations will be of an incremental typefor there are simply fewer radical innovations But Germany and therest of Europe not only have had relatively few radical innovations incomparison with the United States but they also have lagged behindthe USA in developing relatively totally new industries Thus there hasbeen much less innovative activity in Europe in high-technology andnewly emerging industries than in the USAIn complementary research Audretsch and Feldman (1995) have

demonstrated that the USA has been more successful in spawning rmsengaged in the early stages of the industry life cycle but relative toGermany had an environment unfavorable to innovative activity byestablished rms especially in more mature industries German rmshave tended to be associated with a routinized technological regime ndashwhere established enterprises have had strong incentives to engage inincremental innovations but new rms have tended to have fundamental

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

628

innovative disadvantages relative to the Americans (see Audretsch 1995Winter 1982 Casper 2000)On both sides of the Atlantic universities engage in basic and applied

research But signicantly the institutional environment in which univer-sities are embedded inuences the degree to which the research is likelyto represent major breakthroughs or is likely to have more modestimplications Recent research in biomedical science demonstrates that inthose societies in which university professors andor graduate studentshave the following characteristics there are likely to be relatively majorresearch breakthroughs graduate students have a long training periodhigh dependency on a major professor low autonomy in dening thenature of a research project restricted job options and a university envi-ronment where there is a very modest or nonexistent entrepreneurialculture For example in Germany ndash a country with relatively few radicalbreakthroughs ndash the scholar who nishes a habilitation will have spentmany years on a research project before qualifying for a permanent acad-emic position will have been in a high-dependency position with a seniorprofessor will have had limited autonomy to dene a research projectand will be constrained from pursuing a high-risk research project Theyoung German will have an extraordinarily high level of training butwill have faced a number of disincentives to undertake high-risk researchAnd as soon as heshe has completed the habilitation and becomes aprofessor administrative responsibilities constitute another disincentivefrom engaging in high-risk or long-term research projects Moreover theGerman professor is a civil servant with a more stable and secure incomethan hisher American counterpart In a very different institutionalenvironment American professors in large research universities tend tobecome much more entrepreneurial Today many American professorsbehave as though they were operating their own rms within univer-sities and thus devote considerable time to generating funds for researchprojects assistants secretaries supplies etc And the variation in thisentrepreneurial spirit across countries is associated with variation in thetaking of high-risk research strategies and the making of major researchbreakthroughsIn sum societies vary in terms of the institutional environment in

which their organizations are embedded and this variation inuencesthe degree to which they have industrial activities in the early or latestages of the industry life cycle successful new or mature industriesand radical vs incremental innovations And it is this variation whichis encouraging rms in countries with relatively few new industries and with few or no radical innovations to invest in those countries with many new industries and numerous radical innovations ndash as astrategy to be able to transfer new technologies from the latter institu-tional environment to the former

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

629

Table 4 summarizes the way that the norms rules and values of a society are associated with Americarsquos and Germanyrsquos institutionalarrangements their social system of production and their styles of inno-vativeness

CONCLUSION

Several themes deserve to be reemphasized Most importantly institu-tional analysis is emerging as a dominant issue in several social sciencedisciplines The study of various institutional components discussedabove is a complex subject requiring multiple levels of analysis and aknowledge of multiple sectors within societies It requires a theoreticalperspective that evolves from an empirical analysis in different societiesIf we are to advance the study of institutional analysis we need cross-disciplinary collaboration and team research which are not readilyachieved given the structure of most of our research universitiesAt present scholars in each academic discipline have their own distinc-

tive strategies for studying institutional analysis Thus economists tendto use several approaches (North 1990 Williamson 1985 Hodgson 19881998 1999) anthropologists another (Geertz 1995) and sociologistspolitical scientists and historians employ other strategies (Campbell etal 1991 Powell and DiMaggio 1991 Katznelson 1998 Hall and Taylor1996) With practitioners in various disciplines pursuing their distinctiveapproaches to institutional analysis there is low potential for collabo-rative institutional analysis across the social sciences The lessons fromthe biological sciences over the last 40 years are very instructive oncepractitioners from various biological disciplines began to cooperate instudying biological phenomena by using the same concepts at the mole-cular level theoretical advance was very rapid Comparable phenomenatook place in the development of other hybrid elds eg biochemistrybiophysics (Judson 1979 Olby 1979 Kohler 1982 1994) Similarly weresocial scientists in various disciplines to work collaboratively in the areaof institutional analysis there would very likely be an acceleration intheir theoretical knowledgeThis is not a plea that everyone engaged in institutional analysis should

do the same kind of research Indeed just as biologists working at themolecular level study many different kinds of problems similarly socialscientists engaged in institutional analysis would also work on manyseparate problems but in the spirit of a collective enterprise Some wouldconcentrate their attention on the study of rules norms habits conven-tions and values while others would study how these are associatedwith congurations of various institutional arrangements (types ofmarkets hierarchies networks associations communities clans statesetc) Others would work on specic institutional sectors (eg education

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

630

Table 4

Rules norms hab

its an

d values associated w

ith sp

ecic co

ngu

ration

s of coo

rdinating mecha

nism

s soc

ial sy

stem

s of

prod

uction

an

d nationa

l styles of inno

vation

USA

Germany

Norms rules values which provide incentives for

Norms rules values which provide incentives for

short-term horizons

long-term horizons

Con

gu

rative

forms of

Marke

ts c

orpo

rate hierarchies reg

ulatory state

Assoc

iation

s (busine

ss assoc

iation

s la

bor union

s)

coordinating

corp

oratist-type

networks

marke

ts c

orporatist-

mecha

nism

s (listed in

type

state

order of im

portanc

e in

the particu

lar social

system

)

Social sys

tem of

prod

uction

Bus

iness system

Firm

struc

tures

Con

glom

erates b

ut with mov

emen

t toward

Firm

more co

ngru

ent with pa

rticular prod

ucts

elim

ination of produc

tion

in som

e sectors

emph

asis on long

-term strateg

ies rm

s are well

emph

asis on sh

ort-term

strateg

ies high

integrated

differentiation

between divisions

of rms

Firm

s mob

ilized in

Low

High

and in

tegrated

into busine

ss

associations

Indu

strial relations

Low

deg

ree of jo

b secu

rity rigid in

ternal lab

or

High deg

ree of jo

b secu

rity

exible interna

l lab

or

system

marke

t ex

ible externa

l labo

r marke

tsmarke

t rigid externa

l lab

or m

arke

tPatterns of

Dispersed owne

rship frequ

ent turnov

erMore stab

le owne

rship patterns

owne

rship

Table 4

continue

d

USA

Germany

Norms rules values which provide incentives for

Norms rules values which provide incentives for

short-term horizons

long-term horizons

Social system of

Prod

uction

Bus

iness system

Training for labo

r an

dHigh deg

ree of voluntary decision mak

ing by

Collective an

d m

ore co

mpulso

ry d

ecisions

abo

ut

man

agem

ent

individuals as to ho

w m

uch

tim

e an

d ene

rgy

how m

uch ed

ucation

is ne

eded

he

avy em

pha

sis

to inv

est in educ

ation heavy

emph

asis on

on techn

ical training (eg en

gine

ering)

busine

ss sch

ools training

in marke

ting

and

sales accou

nting

Fina

ncial marke

tsEqu

ity an

d ven

ture cap

ital m

arke

ts highly

Equ

ity an

d ven

ture cap

ital m

arke

ts w

eakly

dev

elop

ed na

ncing from

cen

tral gov

ernm

ent

deve

lope

d he

avy em

pha

sis on

corporate loan

sto priva

te rm

s for dev

elop

men

t of new

tech

nologies related

to na

tion

al security

Unive

rsity research

High individual auton

omy high poten

tial for

Low

ind

ividua

l au

tono

my until almost 40

yea

rs of

system

individual creativity u

nive

rsity pr

ofesso

rs

age poten

tial for ind

ividual creativity high

ly

high

ly entreprene

urial

cons

traine

d unive

rsity pr

ofessors not so

entrep

rene

urial

Inno

vative

styles

Highly su

ccessful in m

ajor break

through

s in

High pe

rforman

ce in proc

ess an

d inc

remen

tal

basic an

d app

lied scien

ce go

od in dev

elop

ing

inno

vation

s in existing produ

cts not so su

ccessful

totally

new

lines of pr

oducts an

d in m

ajor break

through

s in basic scien

ce or in

tech

nologies high

deg

ree of entrepr

eneu

rship

deve

loping

totally new

lines of prod

ucts of

amon

g research

ers

tech

nologies low d

egree of entrepr

eneu

rship

amon

g research

ers

Table 4

continue

d

USA

Germany

Norms rules values which provide incentives for

Norms rules values which provide incentives for

short-term horizons

long-term horizons

Prod

ucts in w

hich

Pr

oducts with sh

ort liv

es prod

ucts in w

hich

Older produ

cts with long

history machine

ry pap

er

society excels

tech

nology

is bo

th highly co

mplex

and

pr

inting

textile

s m

otor veh

icles m

achine

too

lsch

ange

s rapidly en

tertainm

ent indus

try

pha

rmaceu

ticals biotechn

olog

y softw

are

aerosp

ace

business systems nancial markets and systems of research) Distinctivewould be the recognition of how these separate studies are linkedtogether how each institutional sector is linked to a societyrsquos normsrules values etc and its conguration of institutional arrangementsSimilarly other institutional analysts might focus on the study of orga-nizations with a major concern for how the institutional environmentof organizations (eg norms rules values congurations of institutionalarrangements institutional sectors) inuenced the structure culture andoutcomes of organizations And nally others might study how all ofthe aforementioned aspects of institutional analysis inuenced a societyrsquosoverall performance Nelson and Sampat (1998) have appropriatelyreminded us that we should engage in institutional analysis in order toaddress specic problems but the view developed in this article is thatwe rst need some mapping of the terrain of institutional analysis inorder for our diverse enterprises to become a meaningful collective enter-prise especially if we are to understand how institutions inuence theinnovative processThe argument developed here is not that all social scientists should engage

in institutional analysis Obviously there are numerous other importantareas of research which lie beyond the eld of institutional analysis Nordoes this article argue that institutional analysis is the most importantsubject for social scientistsSince the structure and culture of our research universities reect an

enormous amount of disciplinary fragmentation what are the prospectsfor promoting the kind of collective and interdisciplinary research agendaproposed above In one respect the prospects are not encouragingDuring the past half century the efforts of American universities to have effective communication across the biological sciences have variedgreatly in their success At the University of California-Berkeley it was not until the 1980s that there was a serious effort to integrate thebiological sciences and at that time the campus abolished seventeendepartments to promote an integrated biological program But at theUniversity of Wisconsin the biological sciences are still very fragmentedinto dozens of different departments in botany plant pathology zoologyand physiology with two departments of biochemistry and two depart-ments of geneticsThere have been other approaches which have worked very well histor-

ically for promoting scientic integration amidst scientic diversity Forexample several summer schools and special institutes were organizedin the United States during the 1940s in an effort to overcome the frag-mentation in the biological sciences These programs made it possiblefor scientists in diverse academic disciplines and from different univer-sities and countries to come together to engage in intense and frequentinteraction Perhaps the most effective program was that held in the

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

634

summers at Cold Spring Harbor New York where a research agendafor the study of molecular biology was developed (Fischer and Lipson1988 Stent and Watson 1966) Scientists went there summer aftersummer and when they returned to their universities many began therestructuring of the biological sciences in their universitiesIn our own day we need summer institutes and other similar programs

so that scholars from diverse backgrounds can come together anddevelop a collective agenda for doing institutional analysis and forstudying the impact of institutions on the performance of societies ndashinnovativeness being one such subject If we can develop such instituteswe will increase the prospects for transforming our universities Thereis reason to be optimistic that once an eclectic group of scholars workwith a common set of concepts we can have genuine theoretical advanceif researchers can bring their diverse expertise to bear on a common setof problems There is considerable evidence that institutional analysisoffers high potential for advancing our knowledge of innovations andtechnological change (Hollingsworth and Hollingsworth 2000) If we aregoing to make substantial theoretical advance it is not likely to occurfrom the vantage point of a multiplicity of disciplines Rather it is likelyto occur as we develop a new intellectual framework and our own acad-emic journals This is a tall agenda which requires us to reassess theway that knowledge is presently produced in our societies

NOTE

This article is part of a much larger agenda involving the study of institutionsorganizations and innovations in which I am involved My debts are many andvaried The rst version was written while I was in residence in the ResearchUnit for Institutional Change and European Integration of the Austrian Academyof Sciences in Vienna in the summer of 1998 I am especially grateful to ProfessorsEgon Matzner and Sonja Puntscher-Riekmann for providing the stimulating envi-ronment in which to think through the ideas developed here Egon Matzner hasengaged in very stimulating discussions with me about the importance of insti-tutions in the general eld of socio-economics The nal version of the articlewas completed during my residence as a fellow at the Neurosciences Institutein La Jolla California I also wish to thank my colleagues in the InnovationTheme Project at the Netherlands Institute for Advanced Study (NIAS) during1998ndash9 and Edgar Grande and Karl Muumlller for stimulating conversations aboutthe interaction of institutions and organizations in facilitating particular typesof innovation I am especially grateful to Jerald Hage for helping me to developa number of issues in the article and for his useful comments on the rst draftFrans van Waarden early on helped me to work through the idea of multiplelevels of institutional analysis Others who have helped to develop the ideas inthe article have been my colleagues in the research project lsquoComparative socialsystems of productionrsquo especially Tom Burns Christel Lane Yoshitaka OkadaWolfgang Streeck and Richard Whitley My colleagues Robert Boyer SteveCasper and David Soskice have taught me much about how incentives providedby the institutional environments of rms inuence individual decision making

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

635

But my most important debt is to Ellen Jane Hollingsworth who has repeatedlyquestioned every single assumption of this article Without her rich and contin-uous tutoring this article would be even less developed than it is at presentFinally without the insightful assistance of David Gear in many phases of thisarticle it would not have been completed

REFERENCES

Aoki Masahiko (1990) lsquoTowards an economic model of the Japanese rmrsquo Journalof Economic Literature March 1ndash27

Arora A Landau Ralph and Rosenberg Nathan (eds) (1998) Chemicals and Long-Term Economic Growth Insights from the Chemical Industry New York JohnWiley amp Sons

Arrow Kenneth (1974) The Limits of Organization New York NortonArthur W B (1988a) lsquoSelf-reinforcing mechanisms in economicsrsquo in P W

Anderson K J Arrow and D Pines (eds) The Economy as an Evolving ComplexSystem Redwood City Calif Addison-Wesley

mdashmdash(1988b) lsquoCompeting technologies an overviewrsquo in G Dosi C Freeman RNelson G Silverberg and L Soete (eds) Technical Change and Economic TheoryLondon Pinter Publishers pp 590ndash607

Audretsch David B (1995) Innovation and Industry Evolution Cambridge MassMIT Press

Audretsch David B and Feldman Maryann P (1995) lsquoInnovative clusters andthe industry life cyclersquo Discussion Paper Wissenschaftszentrum BerlinMarket Processes and Corporate Development Research Unit FS-IV-95-7

Baum Joel A C and Oliver Christine (1992) lsquoInstitutional embeddedness andthe dynamics of organizational populationsrsquo American Sociological Review 57540ndash59

Baum Joel A C and Singh Jitendra V (1994a) lsquoOrganizationndashenvironment coevo-lutionrsquo in Joel A C Baum and Jitendra V Singh (eds) The EvolutionaryDynamics of Organizations New York Oxford University Press pp 379ndash402

mdashmdash(eds) (1994b) The Evolutionary Dynamics of Organizations New York OxfordUniversity Press

Berger Suzanne and Dore Ronald (eds) (1996) National Diversity and GlobalCapitalism Ithaca NY Cornell University Press

Boyer Robert (1997) lsquoThe variety and unequal performance of really existingmarkets farewell to Doctor Panglossrsquo in J Rogers Hollingsworth and RobertBoyer (eds) Contemporary Capitalism the Embeddedness of InstitutionsCambridge and New York Cambridge University Press pp 55ndash93

Burns Tom R and Dietz Thomas (2001 forthcoming) lsquoSocial rule system theorysocial action institutional arrangements and evolutionary processesrsquo in JRogers Hollingsworth and Karl Muumlller (eds) Advances in SocioeconomicsLanham Md Rowman amp Littleeld

Burns Tom R and Flam Helena (1987) The Shaping of Social Organization BeverlyHills Calif Sage

Calvert Randall L (1995) lsquoThe rational choice theory of social institutionsrsquo inJeffrey S Banks and Eric A Hanuskek (eds) Modern Political Economy NewYork Cambridge University Press pp 216ndash66

Camic Charles (1986) lsquoThe matter of habitrsquo American Journal of Sociology 911039ndash87

Campbell John L (1997) lsquoMechanisms of evolutionary change in economicsgovernance interaction interpretation and bricolagersquo in Lars Magnusson

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

636

eld can advance very far if the practitioners do not share a commonunderstanding of the key concepts used in their analysis (Ostrom 19864) But with our universities and academic associations so fragmentedinto different disciplines and into various subspecialties within disci-plines it is difcult to advance the theoretical agenda of institutionalanalysis within the academy Indeed the disciplinary fragmentation ofthe modern university is a major barrier to the theoretical advancementof the study of institutions and innovations as well as most other hybrid elds of research (Hollingsworth 1984 Hollingsworth andHollingsworth 2000) And it will be only as a result of effective commu-nication across diverse elds of knowledge that our study of institutionsand innovativeness will be effectively advanced

MULTIPLE LEVELS OF INSTITUTIONAL ANALYSISAND INSTITUTIONAL CHANGE

There are innumerable signs that we are living in a time of great insti-tutional change the demise of the Soviet empire the processes ofEuropean political and economic integration rapid transformation inparts of the global economy the disintegration of the family structurethe weakening of voluntary associations and the decline in politicalparticipation in a number of advanced capitalist societies the weakeningof welfare states European law is superseding national law and is evenchanging complete national legal systems The list could go on and onEven though scholars discuss institutional change at length their

ability to measure the rate of institutional change is very limited Andmore crucial than the limited ability of scholars to measure institutionalchange is their very limited understanding of how to build new insti-tutions One of the reasons for these shortcomings is that the socialsciences are decient in a theory of institutions The building of newinstitutions and redressing the decline of some of the most importantinstitutions of our societies are among the most important problems ofour time If we are to advance in the development of a theory of insti-tutions we need to work collaboratively across the social sciences andwe need to dene the parameters of institutional analysis

First level of analysis

This article attempts to make some modest contribution to outlining theparameters of institutional analysis At the outset we need to recognizethat when we engage in institutional analysis we must be sensitive to multiple levels of reality As suggested above most scholars whoengage in institutional analysis do not participate in any coordinatedactivity with each other and their activity is fragmented into a variety

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

600

of disciplines and subdisciplines To establish some coherence to the eldof institutional analysis we need a map of the eld so that those workingin one area can see where their research ts in relation to other areasand other practitioners on the mapTable 1 presents such a map with multiple levels at which institu-

tional analysis occurs Theoretically each of these areas on the map isinterrelated with each other level However the various areascompo-nents on the map are arranged in descending order of stability orpermanence Those components at higher levels of reality are morepermanent and durable while those at lower levels change more rapidlyWere scholars doing institutional analysis able to reach some consensus

about where their own work ts in relation to all other practitioners inthe eld there would be increased potential for all practitioners tocommunicate with each other By analogy once geneticists crystallo-graphers biochemists etc had a good understanding of how theirinvestigations were related to each area of molecular biology the eldquickly was able to make theoretical advances (Judson 1979)At the rst level there are the basic norms rules conventions habits and

values of a society These are the most fundamental properties of insti-tutions and are the most enduring and resistant to change Rules normsconventions etc are institutions but are only one component of whatconstitutes institutional analysis As Burns and Dietz (2001 forthcoming)point out most human activity is organized and regulated by normsand rules and systems of rules These concepts are extremely importantfor institutional analysis as they exert the greatest inuence on the natureof the components of institutional analysis at the next four levels which

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

601

Table 1 Components of institutional analysis

1 Institutions = norms rules conventions habits and values (see North1990 Burns and Flam 1987)

2 Institutional arrangements = markets states corporate hierarchies networksassociations communities (Hollingsworth and Lindberg 1985 Campbell et al 1991 Hollingsworth et al 1994 Hollingsworth and Boyer 1997)

3 Institutional sectors = nancial system system of education businesssystem system of research (Hollingsworth 1997)

4 Organizations (Powell and DiMaggio 1991)5 Outputs and performance = statutes administrative decisions the nature

quantity and the quality of industrial products (Hollingsworth 1991 1997)sectoral and societal performance (Hollingsworth and Streeck 1994Hollingsworth et al 1990 Hollingsworth and Hanneman 1982)

Note The ve components in this table are arranged in descending order of permanenceand stability That is norms conventions etc are more enduring and persistent than eachof the other components of institutional analysis Each component is interrelated withevery other component and changes in one are highly likely to have some effect in bringingabout change in each of the other components

are depicted in Table 1 In most forms of social analysis it is extremelyimportant that we understand the social and cognitive conditions thatlead to compliance or non-compliance with rules and the conditionswhich lead to changes in rulesThe approach to the study of institutions employed here argues that

norms rules habits conventions and values both reect and shape thepreferences of actors Norms rules habits conventions and values inu-ence who and what are included in different types of decision makinghow information is processed and structured what action is taken(Shepsle 1986 1989) It is through norms and rules that behavior isjudged to be democratic fair or egalitarianBurns and Flam (1987) point out that in any society there are multiple

rule systems Within a family there are rules for decision making oftenquite different from rules and norms for decision making for a professorin a classroom or for the customer in a bank Despite the heterogeneityof rule systems there are meta rules and norms which encompass lesserrule systems Otherwise there would be such contradictory rule systemsthat a society would be paralyzed The existence of meta rule systemspermits different rule systems to intersect with each other so that ambi-guities can be resolved The greater the pluralism and complexity of asociety the more ambiguity there is about meta rules and norms in asociety and of course all ambiguities never completely disappearOverall the degree to which separate rule systems are interlinked is anempirical problem There are different sectors groups and interestspursuing their own action logic but through higher-order meta princi-ples and rules there can be order consensus and coherence in a societyIt is through a set of meta rules that class and ethnic conict in societiesare contained (For elaboration see Burns and Flam 1987)In many respects our understanding about norms rules habits con-

ventions and values inuences our perspective on how societies are con-structed and how they change lsquoNew institutionalistsrsquo (Posner 1992Schotter 1981 Williamson 1975 1985) tend to assume that at one timethere was a state of nature and that there was a movement from individ-uals to institutions ndash an approach often called methodological individual-ism (Popper 1961 Hodgson 1998 1999) And of course there areinnumerable instances which methodological individualists cite to demon-strate that individuals create new rules of behavior For example it is pos-sible for actors to change the rules of driving so that instead of drivingon the left side of the road drivers adopt a new rule and drive on the rightThis article however tends to equate social habits and institutions

As Hodgson and others remind us (Hodgson 1988 1989 1997 1999Grafstein 1992 Camic 1986 Johnson 1992 Nelson and Winter 1982Veblen 1899) social habits are the results of earlier choices and are ameans of avoiding endless deliberation Because cognitive frameworks

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

602

are learned through habit individuals rely on the acquisition of suchcognitive habits before reason communication choice or action arepossibleWhereas Schotter (1981) and other game theorists take the individual

as an agent unencumbered by previous habits Field (1984) and othershave stressed that there can be no games without prior norms and rulesand thus a set of norms and rules must be presumed at the start Thosewho attempt to explain institutions from individual behavior alone areusing a bad strategy (Hodgson 1998)The position here ndash heavily inuenced by Hodgson (1998 1997 1988)

ndash is that individuals are embedded in a complex institutional environ-ment and that institutions not only constrain but also shape individuals(also see Hollingsworth and Boyer 1997) It would be a mistake howeverto get involved in an innite regress in order to determine which camerst ndash individuals or institutions Of course institutions are formed andchanged by individuals just as individuals are shaped and constrainedby institutions But at a macro level it is institutions that provide acognitive framework whereby individuals can cope with their reality Inthis sense the micro and macro worlds are intertwined At the macrolevel there is considerable stability but at the micro level individualshave a signicant level of autonomy and there can be widespread diver-sity As Hodgson (1988) reminds us most institutions in a temporal senseexist prior to the individuals in any given societyIt would be a serious mistake to downplay the importance of indi-

viduals and micro level analysis as we study institutions In the nalanalysis it is at the level of individuals that norms rules habits conven-tions and values exist An individual is born into and socialized intogroups and a society and this is how one early in life acquires a senseof appropriate forms of behavior Because of the way that individualsare socialized into a world of rules norms habits conventions andvalues it is unnecessary for individuals to restructure the world anewevery day (Douglas 1987 Elster 1989) Every action does not have tobe seriously reected upon For this reason institutions provide cogni-tive frameworks for individuals make their environments predictableprovide the information for coping with complex problems and envi-ronments In the words of Johnson (1992 26) lsquoInstitutions reduceuncertainty coordinate the use of knowledge mediate conicts andprovide incentive systems By serving these functions institutions providethe stability necessary for the reproduction of societyrsquo However eachsociety has different forms of habits rules and norms and hence differentincentives and disincentive systems for learning and forgetting forprocessing information But because individuals have varying degreesof autonomy individuals and groups can deviate from the prescribedforms of behavior in a society And of course these changes at the level

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

603

of individuals become important in understanding processes of societalchangeThese views are not meant to imply that the type of institutional

analysis proposed herein approximates a general theory of society Thisis clearly not the case However it is intended to represent the rst stepsin a mapping exercise of the boundaries of institutional analysis and tosuggest a few methodological insights for studying institutionsOne should try to see norms and rules as continuous and not as

dichotomous entities to recognize that they come in varying strengthsLegro (1997) has suggested that we assess the robustness of norms andrules with three criteria their simplicity their durability and their concor-dance Simplicity refers to how well actors understand norms and ruleshow well they can be applied within a specic situation Some normsand rules can be so complex that actors can have considerable difcultyin applying them in specic cases Durability addresses the issue of howlong norms and rules have been in effect ndash in short in order to assesstheir level of legitimacy While the position of this article is that normsrules and values are quite durable they do vary in this respectConcordance refers to how widely applied a norm or rule is Thisaddresses the degree to which a rule is a meta rule the degree to whichit incorporates the heterogeneity of other norms and rules In sum theclearer the norms and rules of a society the longer they have been inexistence and the more widely applicable they are the greater theirimpact on a society Hence the more robust the norms and rules thegreater their impact on a society and the less their robustness the greatertheir exibility and the less their effect on shaping a societyrsquos outcomesand performancesBecause norms rules and values are quite durable they play an impor-

tant role in shaping the history of societies thus contributing to a greatdeal of path-dependency Actors attempt to adjust to their contempo-rary environment but since they are products of the past the historicallegacy of norms rules and values inuences the decisions they makeAlthough actors have some capacity to alter the course of their historythey are constrained by their past and the degree to which they canmove beyond their past is limited As Lanzara (1998) Johnson (1992)and others have argued societal inertia is a basic feature of institutionsThey provide the basic stability necessary for change Degrees of historyare continuously reproduced by the way in which the inhabitants ofsocieties are socialized History matters but at critical points in historythere is punctuated equilibrium (Somit and Peterson 1992) During mostperiods of history there is considerable stability in the norms rules andvalues of a society but at critical moments norms rules and values canquickly and dramatically be redened At all times norms rules habitsconventions and values are inuencing each of the other components in

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

604

the institutional framework discussed below but these other institutionalcomponents feed back and can modify rules norms conventions etc(Murmann 1998)

Second level of analysis

The norms rules habits conventions and values of a society lead to thenext level of analysis ndash the institutional arrangements which are involvedin the coordination of various economic actors producers and suppliersof raw materials knowledge etc processors of raw materials infor-mation workers customers of raw materials nished productsinformation etc nanciers governmental and other types of regulatorsThese actors regularly engage in contests to resolve various economicproblems in virtually all sectors of society how are prices to be setWhat quantity of various products is to be produced How are stan-dards of various products processes etc to be set How is the qualityof products and processes to be determined How are various societalprocesses to be nanced In order to confront these problems and toaddress the conicting positions of economic actors as they address these problems societies develop various institutional (governance)arrangements for coordinating different actors These consist of marketsvarious types of hierarchies and networks associations the state commu-nities and clans (see Hollingsworth and Boyer 1997 Campbell et al1991 ch 1)Each of these particular kinds of coordinating mechanism has many

different types For example there are many types of states (eg theregulatory state developmental state authoritarian state welfare state)on which there is an extensive literature (Kim 1997) Similarly there aredifferent types of markets networks different kinds of associations etc(Boyer 1997 Hage and Alter 1997 Schneiberg and Hollingsworth 1990)When we do institutional analysis we must engage in congurativeanalysis recognizing that actors are not coordinated or governed by asingle type of institutional arrangement Some of the literature discussesindustrial sectors as though they were coordinated or governed by asingle institutional arrangement whereas in fact each sector of aneconomy is coordinated by a conguration of institutional arrangementsSome congurations coordinate actors in certain problem areas whileother congurations of institutional arrangements coordinate actors inaddressing other problems The types of congurations which are domi-nant in a society are somewhat stable and tend to persist over timewithin a society (Hodgson 1999)When one mode of coordination is dominant in a society it will

inuence the role which other coordinating modes will play Hence thestrong role of the state in the Soviet Union inuenced the role of markets

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

605

associations etc in the governance of the Soviet economy Similarly theprominence of particular modes of coordination in a society inuencesits style of innovationMuch of the literature on institutional arrangements (eg forms of eco-

nomic coordination) remains fragmented and unintegrated It is helpfulto array these various forms of coordination in a two-dimensional tax-onomy as in Figure 2 On the vertical dimension the economistrsquos viewof a self-interested agent is contrasted with a more sociological perspec-tive according to which social rules obligation and compliance shapehuman actions On the horizontal dimension there is another continuum(ie the distribution of power) At one extreme of the dimension onends many and relatively equal agents interacting (eg as in a well-organized spot market) At the other extreme inequality in power resultsin a hierarchical form of coordination which structures the interactionbetween principals and agents or between leaders and followers Wherea society falls along the distribution of power is inuenced by the rulesnorms values etc which are dominant in a society at a particular momentin timeInstitutional arrangements can be visualized on two dimensions the

nature of action motive on the one hand and the distribution of poweron the other Markets (cell 1) combine self-interest with horizontalcoordination transactors and they reect sensitivity to concerns aboutsupply and demand thus providing ex post an unintended equilibriumParadoxically the more pure and perfect the market competition thegreater the need for codied rules of the games for coordinating economictransactions Thus collective associations (cell 6) andor various formsof state intervention (cell 4) are required to develop and enforce rulesfor transacting partners (Schneiberg and Hollingsworth 1990 Streeckand Schmitter 1985a) This is an example of how the norms and rulesand the institutional arrangements of a society are intertwined asreected in a conguration of institutional or governance arrangementsAlong the horizontal axis actors can be joined in an organization or

a rm a hierarchy is the generic terms for this institutional arrangement(cell 2) Along the horizontal line one recognizes the difference betweentransactions in a market and transactions within a rm The well-knownworks of Coase (1960 1981) and Williamson (1975 1985) utilize theconcept of transaction costs in explaining the emergence of corporatehierarchiesThere are also various types of networks (cell 5) which exhibit mixes

of self-interest and social obligation with some actors being formallyindependent and equal Yet in some networks (the large rms and theirsubcontractors) there is unequal power and initiative Networks maycomprise all kinds of actors some consist only of rms but others includeassociations and the state (Hage and Alter 1997)

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

606

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

607

DISTRIBUTION OF POWER

VERTICALHORIZONTAL

1 MARKETS 2 HIERARCHY

6ASSOCIATIONS

5NETWORKS

4STATE

3COMMUNITIES

SELF INTEREST

SOCIAL OBLIGATION

ACTION MOTIVE

Figure 2 A general taxonomy of institutional arrangements

The vertical axis deals with action motives Toward the upper part ofFigure 2 actors are engaged in individualistically oriented behaviorwhereas toward the lower part actors are more engaged in collectivebehavior and strive to cope with problems of common interest Cell 3 ndashcommunities and clans ndash consists of institutional arrangements based ontrust reciprocity or obligation and thus not derived from the pure selshcomputation of pleasures and pains This is an unconventional form ofcoordination for most neoclassical economists (however see Arrow1974) but not for many anthropologists political scientists and sociolo-gists (Streeck and Schmitter 1985a Polanyi 1944 Gambetta 1988Fukuyama 1995 Sabel 1992 Putnam 1993)In the neoclassical paradigm theorists argue that actors engage in

forms of exchanges that best promote their individual interests If somestructural conditions are fullled (absence of increasing returns to scalethe reversibility of transactions absence of uncertainty and completecontingent markets with no collusion between actors) then the invis-ible hand theorem applies and market-type activity functions quite welland also provides the optimum for society thereby combining efciencyharmony and order However an excess of market activity may welllead to ruinous competition and excessive conict There is variation inthe extent to which ruinous competition occurs depending on the socialcontext in which transactions take place Thus it is important that webe sensitive to the institutional context in which transactions areembedded and that we understand the degree to which social bondsexist at both the micro and macro levels of analysis Micro bonds facil-itate exchanges in a society but at the societal level social bonds existat the level of the collective ndash in the community or region and amongmembers of racial religious and ethnic groups All other things beingequal the more powerful the social bonds among transacting partnersthe more economic competition is likely to be restrained Thus mosttransactions occur not simply in an impersonal calculative system ofautonomous actors unrestrained by social ties ndash as implied by theneoclassical paradigm ndash but in the context of social ties variation in the strength of which leads to variation in levels of trust and transac-tion costs (Etzioni 1988 211 Granovetter 1985 Streeck and Schmitter1985a Hollingsworth and Boyer 1997 Hodgson 1988 1999 Schneibergand Hollingsworth 1990)Another form of multilateral institutional arrangement is various

types of associations (cell 6) Unlike networks clans and communitiesassociations are more formal organizations Whereas markets corporatehierarchies and networks tend to coordinate economic activity amongdifferent types of actors (eg producers with suppliers capital withlabor) associations typically coordinate actors engaged in the same orsimilar kinds of activities Business associations and labor unions are

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

608

some of the most common forms of associations for coordinatingeconomic activity in capitalist economies (Schneiberg and Hollingsworth1990 Streeck and Schmitter 1985a)Finally there is the state (cell 4) which is an institutional arrangement

quite unlike any of the others It is the state that sanctions and regu-lates the various non-state coordinating mechanisms that is the ultimateenforcer of rules of the various mechanisms that denes and enforcesproperty rights and that manipulates scal and monetary policy Overallit is the state that inuences the total incentive system of a society(Lindberg and Campbell 1991 Johnson 1992 40) At the same time thestate may also be an economic actor by engaging directly in productionand exchange relationsThe choices of institutional arrangements in Figure 2 are constrained

by the social context in which they are embedded Depending on thenature of that embeddedness there is variation in the collective formsof governance some of which are specied in the lower part of thegure Some modes of coordination in the bottom and upper levels ofthe typology are often mixed together though one coordinating modeis likely to be more prominent than another But actors are often engagedin complex congurations involving several kinds of institutionalarrangementsEach of these various institutional arrangements has its own logic ndash

its own rules its own procedures for enforcing compliance its ownnorms and ideologies which help to reduce the costs of enforcementThese are summarized in Table 2 which provides further elaborationabout the various coordinating mechanisms one nds in almost everycapitalist society Table 2 lists the organizational arrangements and theirstructures rules of exchange and means of enforcing compliance asso-ciated with each type of coordinating mechanism While each type ofinstitutional arrangement has various positive features each institutionalarrangement also has failures and these are featured in Table 3 It is thecontest between those who support and those who oppose these variousinstitutional arrangements that tends to lead to transformations in insti-tutional arrangements over time (Campbell et al 1991 Campbell 1997)Again it is difcult to conceive of pure institutional arrangements ndash

either exclusively markets or exclusively hierarchies ndash since it has beenwell known since Adam Smithrsquos Wealth of Nations that the division oflabor within the rm cannot be disentangled from the existence andextent of the market Institutional arrangements have their own distinc-tive form of efciency (good static efciency for the market dynamicefciency for rms) and inefciency often leading to considerableinequality Neither networks nor communities are panaceas for economiccoordination without being congured with other types of institutionalarrangements Networks and communities may solve certain issues but

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

609

Table 2

Log

ics of ins

titution

al arran

gemen

ts

Coordination

Organizational

Rules of

Individual means

Collective means

mechanisms

structure

exchange

of compliance

of compliance

Marke

tsEasy en

try an

d exit

Voluntary sp

ot excha

nge

Leg

al enforcemen

t of

Norm of pr

ivate

Bila

teral ex

chan

ge or

control

prope

rty

marke

t site (W

all Street)

Reg

ulations

to

Leg

itim

acy of m

arke

t en

force contracts

men

tality

Com

mun

ities

Inform

al m

embe

rship

Voluntary exch

ange

based

So

cial norms an

d m

oral

Highly institutiona

lized

ev

olve

s ov

er lon

g period

on social solid

arity an

d

princ

iples im

pose

norm

s an

d rules requ

ire

of tim

ehigh

deg

ree of tru

stob

ligations

mem

bers to accept

Kno

wledg

e of others an

d

lsquocorpo

ratersquoo

bligations

recipr

ocity ov

er tim

e

Networks

Semi-form

al m

embe

rship

Voluntary exch

ange

ove

r a

Con

tractual bon

dsPersona

l relation

sBila

teral or m

ultilateral

time pe

riod

Resou

rce de

pend

ence

Tru

st built outside

the

ex

chan

geecon

omic arena

Associatio

nsFo

rmal m

embe

rship

Restricted to mem

bers

Self-interest

Some deg

ree of

Multilateral excha

nge

Emph

asis on insider

Rep

utation

effects

compu

lsion

outsider or we

they

Priva

te interest type

of

men

tality

gove

rnan

ce

Private

Com

plex organ

izations

Restricted to mem

bers

Rew

ards to ind

ividuals

Highly institutiona

lized

hierarch

ies

which

ten

d to becom

e ex

chan

ge based

on

Asymmetric po

wer threat

rules

bureau

cratic hierarchies

asym

metric po

wer

of san

ctions

Mem

bers soc

ializ

ed into

bureau

cratic rules

corp

orate cu

lture us

e of

sanc

tions

State

Pub

lic hierarchy

Unilateral action

Exit vo

ice (vote lob

bying)

Coe

rcion

De jure

and impo

sed

Indirect political an

d

loya

ltyNorms an

d pu

blic rules

mem

bership

econ

omic excha

nge

Table 3

Failu

res of coo

rdinating mecha

nism

s

Coordinating mechanisms

Type of failure

Associations

Private hierarchies

State

Market

Communities

Networks

Enforcem

ent

Usu

ally relies on

Might enh

ance

Needs co

ntrol

Needs an

internal

Needs

tru

st and

Need an external

the state as an

oppo

rtun

istic

extern

al to state

enforcem

ent

loya

lty often

en

forcem

ent

enforcer

beha

vior

bureau

cracy

authority

coming from

au

thority

Resem

bles

The

ideal o

f (ju

dge

s

Facilitates

outside (fam

ily May

facilitate

enforcem

ent

internal m

arke

ts

parliamen

t collu

sion

and

relig

ion

cartelization an

d mecha

nism

might hur

t marke

t) to

impe

rfect

ethn

icity)

mon

opoly

of cartels

incu

mbe

nt

correct state

compe

titio

nCom

patible with

worke

rsab

uses

variou

s types of

Lob

bies can

co

mpe

tition

captur

e pu

blic

interest goa

ls

Public goo

d

Useful for

Gov

erna

nce co

sts

Can

provide

Can

not provide

Can

internalize

Useful for

and

establishing

might exceed

public goo

ds

colle

ctive go

ods

some co

llective

enha

ncing qu

ality

externality

stan

dards an

d

the be

nets of

but ha

s or deal w

ith

good

s (qua

lity

and training bu

t qu

ality for

internal division

difc

ultie

s in

externalities

training

) bu

t no

t no

t ve

ry goo

d in

setting ru

les of

of la

bor

prov

iding them

Inad

equate

othe

rs (welfare

prov

iding for

compe

titio

n in

Slow

to react to

in precise

mon

itoring of

gene

ral p

ublic

societal gen

eral

the indu

stry

chan

ges in the

am

ounts

tech

nical ch

ange

go

ods)

welfare

Useful for

environm

ent

Might fail in

and inno

vation

Mem

bers tightly

Weak in the

prov

iding

induc

ing

integrated

into

prov

ision of

man

y go

ods

tech

nical

community

colle

ctive go

als

colle

ctively

chan

geha

ve limite

d

that in

dividua

lcapa

city for

mem

bers can

not

inno

vatio

nsprov

ide for

them

selves

Table 3

continue

d

Coordinating mechanisms

Type of failure

Associations

Private hierarchies

State

Market

Communities

Networks

Efcien

cyFa

cilitates

De

cien

t in

Can

be high

ly

Some ba

sic social So

me go

ods

Slow

to en

hanc

e co

operation an

d

coop

eration an

d

bureau

cratic

relation

s cann

ot

cann

ot be

efc

ienc

y an

d

X-efcien

cy but

X-efcien

cyan

d can

not

be provided

by

delivered

at

speed of

not allocative

easily deliver

pure m

arke

t su

fcien

tly

adap

tive

ness

efcien

cy

good

s at

mecha

nism

slow costs

except in

low cost

indus

tries whe

re

tech

nology

is

complex an

d rapidly cha

nging

Equity

Narrow

Excessive

Might enh

ance

Facilitates

Might lead to

Whe

n widely

encompa

ssing

multiplication

ineq

uality

ineq

uality in

retard

ed

dev

elop

ed in

to

associationa

l of con

trollers

(pow

er

income an

d dev

elop

men

tindus

trial

stru

ctures lead

(fru

stration

and

an

d privilege

)wealth

districts

to in

come

ineq

uality)

netw

orks m

ay

equa

lity

facilitate greater

equality an

d

inco

me

distribution

whe

n wea

kly

dev

elop

ed

netw

orks ten

d to

increase soc

ial

ineq

uality

Note For allo

cativ

e efc

ienc

y an

d X

-efcien

cy see Leibe

nstein (19

66 19

76)

they raise other no less severe problems It is important to recognizethe imperfection of any single institutional arrangement in order tocomprehend the origin and transformation of any other institutionalarrangement (Hodgson 1999 ch 3)As suggested above each institutional arrangement is congured with

other institutional arrangements Usually one particular institutionalarrangement is more dominant in a particular conguration but becauseeach type of institutional arrangement has its own strengths andweaknesses there is no simple structural logic in the governance or coor-dination of a society Each institutional arrangement constrains the otherbut the inherent tension among the various institutional arrangementswithin a conguration contributes to changes in the governance orcoordination of a society The routines and logics of each of these cong-urations of institutional arrangements provide constraints and incentivesfor actors The way that actors perceive the incentives and constraintsof these governance congurations leads to a particular market logicand it is the specic market logic of a society which inuences its speciccapacities and weaknesses The inherent strain among the different logicsin a conguration of coordination helps to provide the exibility for asociety to adapt to new circumstancesIf there were a society with pure markets or a society coordinated

only by the state there would be too much rigidity and too little diver-sity to cope with the vast uncertainty in the global environment A societywith very little diversity in its coordination mechanisms would havelittle capacity to adapt to new circumstances Congurations with consid-erable diversity of institutional arrangements provide for a certainamount of incoherence in governance but they also provide for thecapacity to adapt to new circumstances While the Soviet Union wasdominated by the state as a coordinating mechanism there were alwaysfunctioning markets in the system Moreover the feudal society of thetwelfth and thirteenth centuries consisted not only of hierarchical rela-tionships between serfs and masters but also of urban guilds clericalhierarchies and markets (Johnson 1992 38) In sum the robustness ofinstitutions often depends on multiple and diverse principles and logicsof action on the inconsistency of principles and procedures on patternedforms of disorder (Lanzara 1998 Orren and Skowronek 1991 320 329)

Third level of analysis

The next level consists of the institutional sectors of a society The rulesnorms and values of a society inuence the array of institutional arrange-ments and levels one and two inuence the nature of and the relation-ships among various institutional sectors and all three levels areintricately linked together to form a social system of production Together

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

613

all of these components inuence the performance of economic sectorswithin a society as well as the performance of the total society A socialsystem of production is the way that a societyrsquos institutions (see rstlevel of analysis above) its institutional arrangements (see second levelof analysis above) and its institutional sectors are integrated into a socialcongurationAn institutional sector includes all organizations in a society which

supply a given service or product along with their associated focal orga-nizations (eg major suppliers funding sources regulators and so forth(Scott et al 1994 108 117 Campbell et al 1991 Hollingsworth et al1994) Institutional sectors include but are not limited to the societyrsquossystem of education system of research business system nancialmarkets legal system and the state The structure of the nancial marketssystem of training and education industrial relations system system ofresearch and state and legal systems are distinct and idiosyncratic ineach society In short they are system specicAll of these institutional sectors tend to cohere with each other

although they vary in the degree to which they are tightly coupled intoa full-edged system While each of the institutional sectors has someautonomy and may have some goals that are contradictory to the goalsof other institutional sectors with which it is linked an institutional logicin each society leads institutions to coalesce into a complex social con-guration (Hollingsworth 1991) This occurs because the institutionalsectors are embedded in a culture in which their logics are symbolicallygrounded organizationally structured technically and materially con-strained and politically defended The conguration of institutionalsectors usually exhibits some degree of adaptability to new challengesbut continues to evolve within an existing style But under new circum-stances or unprecedented disturbances these institutional congurationsare exposed to sharp historical limits as to what they may or may notdo (Schumpeter 1983 David 1988 Arthur 1988a 1988b Haringkanssen andLundgren 1997)Why do all of these different institutions coalesce into a complex

social conguration which is labeled here as a social system of pro-duction The literature suggests two contrasting interpretations Part ofthe answer ndash indeed a controversial one ndash is that these institutionalsectors are functionally determined by the requirements of the practiceof capitalism in each time and place (Habermas 1975) Another expla-nation emphasizes the genesis of the actual conguration via a trial and error process according to which the survival of rms regions orcountries is the outcome of complex evolutionary mechanisms (Maynard-Smith 1982 Nelson and Winter 1982) However the problem is evenmore complex Markets and other mechanisms for coordinating rela-tionships among economic actors place constraints on the means and

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

614

ends of economic activity to be achieved in any society The othercoordinating mechanisms include different kinds of hierarchies varioustypes of networks and associations ndash trade unions employers and busi-ness associations (Hollingsworth and Lindberg 1985 Campbell et al1991) The logic of these various institutional arrangements providesactors with vocabularies and logics for pursuing their goals for deningwhat is valued and for shaping the norms and rules by which theyabide (see Tables 2 and 3) In short in contrast to the logic of theneoclassical paradigm the argument here is that the dominant type ofinstitutional arrangements places severe constraints on the denition of needs preferences and choices of economic actors Whereas the neo-classical paradigm assumes that individuals and rms are sovereign thisarticle is based on the assumption that rms are inuenced by the hold that the institutional congurations making up a social system ofproduction have on individual decision making (Campbell et al 1991Etzioni 1988 Streeck and Schmitter 1985a Hollingsworth et al 1994Hollingsworth and Boyer 1997 Magnusson and Ottosson 1997 North1990 Hodgson 1999)Standard neoclassical economic theory has tended to downplay the

role of production and consequently of rms Even transaction cost theo-rists who are concerned with analyzing the rm as a coordinatingmechanism have been relatively unconcerned with the various compo-nents of a social system of production Indeed as long as there waswidespread optimism about the efcacy of Keynesian economics therewas relatively little concern among neoclassical economists with thesupply side of the economy Even in the opinion of most Keynesians agroup of experts should ideally be able to shape the size of aggregatedemand while the supply side of the economy would be left to the twominimalist institutions of neoclassical economics ndash markets and manage-rial hierarchies For several decades however it has become increasinglyobvious that some of the most competitive and successful patterns ofindustrial output and industrial production in capitalist economies donot derive from the neoclassical prescription of unregulated markets andcorporate hierarchies complemented by a neoliberal democratic stateIndeed it is now well understood that certain highly successful produc-tion patterns require for their emergence and survival institutionalarrangements and environments the very opposite of the prescriptionsfound in the neoclassical paradigm (see especially Streeck 1992 but alsoHollingsworth and Streeck 1994) Thus social scientists have come tounderstand that if they are to comprehend the behavior and performanceof contemporary economies concerns about social systems of produc-tion must be brought into the pictureBecause production involves more than technology a number of social

scientists have an increasing concern with social systems of production

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

615

The same equipment is frequently operated quite differently in the samesectors in different countries even when rms are competing in the same market (Maurice et al 1980 Sorge and Streeck 1988 Sorge 1989Hollingsworth et al 1994) Variations in production and process tech-nologies are inuenced partly by variations in the social environmentsin which they are embedded In other words rms are embedded intocomplex environments which among other things place constraints ontheir behavior Thus a social system of production is of major impor-tance in understanding the behavior and performance of an economyHow the state and other institutional arrangements (eg marketsnetworks private hierarchies associations) coalesce and are related toparticular social systems of production are important determinants ofeconomic performanceThe congurations of institutional arrangements that coordinate or

govern the behavior of actors in one society and its structure of specicinstitutional sectors cannot easily be transferred to another society forthey are embedded into a social system of production that is societallydistinct (Hollingsworth 1997) Societies borrow selected principles offoreign management styles and work practices but the effectiveness ofsuch borrowing is generally limited Economic behavior and performanceare shaped by the entire social system of production in which actors areembedded and not simply by specic principles of particular manage-ment styles and work practices Moreover a societyrsquos social system ofproduction tends to limit the kind of goods that it can produce and withwhich it can compete successfully in international marketsSome scholars (Kenney and Florida 1993 Oliver and Wilkinson 1988)

have assumed that the diffusion of particular forms of management stylesand work practices across societies could lead to system convergenceReferring to the Japanese production system Kenney and Florida haveargued that it consists of organizational practices whose fundamentallsquogenetic codersquo can successfully be inserted into another society and beginsuccessfully to reproduce its behavior in the new environment (1993 8)Their position is in the intellectual tradition of Antonio Gramsci (1971)who contended decades ago that the American system of mass produc-tion would most certainly diffuse to Europe over timeHowever the argument of this article is that even though British

French and American rms may adopt certain aspects of Japanesemanagement styles (eg just-in-time production self-managing teamsquality circles the use of statistical process controls) or some variant ofthe German vocational system their social systems of production willnot converge Moreover the overall conguration of a societyrsquos socialsystem of production inuences its sectoral and overall national economicperformance This helps to explain why societies have different systemsof innovation and why they vary in the clusters of industries in which

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

616

they are highly competitive in the international markets (Nelson 19931999 Hage and Hollingsworth 2000 Berger and Dore 1996)In order to understand how and why a societyrsquos economy performs

as it does it is necessary to understand its entire social system of produc-tion If a society is to modify substantially the performance of itsinstitutional sectors and its entire economy it cannot adapt only someof the management and work practices of its foreign competitors Ratherit must alter its entire social system of production Because a societyrsquosmodes of economic governance and coordination and its institutionalsectors develop according to particular logics and are system specicthere are serious limitations to the extent to which a society may mimicthe institutions (eg the rules and norms) institutional arrangements andinstitutional sectors of other societiesIn the history of modern societies there are logics by which institu-

tions coalesce into a social system of production (Hollingsworth 1991)Though institutions are constantly changing there are sharp limits tothe type and the direction of change that any particular institution canundergo because of its linkages with institutional arrangements and insti-tutional sectors Thus a societyrsquos business rms educational systemnancial markets industrial relations system etc can engage in seriousrestructuring only if most of the other institutional sectors also changeIn social systems of production there are pressures toward consistencyin the norms rules and values across institutional sectors though in anycomplex society social systems are obviously imperfectly integratedIndeed the degree to which the institutional norms and rules makingup a social system of production are loosely or tightly coupled is a vari-able of considerable importance In general the institutional sectorsmaking up a social system of production are interdependent and changesin one generally result in changes in the others Since each institutionallevel is dependent on the others for various types of resources there isinterdependence among the differing institutional spheres Moreovereach society has its norms moral principles rules and laws and recipesfor action as well as its own idiosyncratic customs traditions and prin-ciples of justice (Burns and Flam 1987)There are also other inherent obstacles to convergence among social

systems of production for where a system is at any one point in timeis inuenced by its initial state Systems with quite different initial statesare unlikely to converge with one anotherrsquos institutional practicesExisting institutional arrangements block certain institutional innova-tions and facilitate others (Roland 1990) Thus the institutions makingup a social system of production provide continuity even though insti-tutional arrangements are always changing but with a particular logicWhile Williamson (1975 1985) suggests that actors tend to choose theinstitutional arrangements which are most efcient North (1990) is

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

617

much closer to the mark in his argument that most societal institutionalarrangements exist as a result of custom and habit and are thereforeinefcient At any moment in time the world often appears to its actorsas very complex and uncertain For this reason actors often engage incontradictory forms of behavior pursuing different strategies as hedgesagainst a very uncertain world (Lanzara 1998) And their hedging andcontradictory forms of behavior may lead in somewhat different soci-etal directions all constrained by the institutional fabric within whichthe actors are embedded This kind of contradictory behavior occurs inpart because of the contradictions inherent in the dominant congura-tions of institutional arrangements with their contradictory logics (seelsquoSecond level of analysisrsquo above)Despite the emphasis on the logic of institutional continuity this is

not an argument that systems change along some predetermined pathThere are critical turning points in the history of highly industrializedsocieties but with the choices limited by the existing institutional terrainBeing path-dependent a social system of production continues along aparticular logic until or unless a fundamental societal crisis intervenes(Milgrom et al 1991 Krugman 1991 Durlauf 1991 Hollingsworth 1991Pred 1966 David 1988 Hodgson 1998)At this point it is important to confront the question of how social

systems of production evolve And this question gets to the heart of theproblem of building complex societal structures Certainly social systemsof production did not emerge from some process of social engineeringMoreover the various component parts of each social system of produc-tion have often not been designed to be part of a social system ofproduction The component parts of each social system of productionhave emerged more often than not as unintended by-products of goalswhich various actors had in mind at earlier moments in time It is usuallythe case that actors at time lsquotrsquo fail to comprehend the long-term conse-quences of their actions Wolfgang Streeck the author of several brilliantpapers on social systems of production argues that institutional sectordesigns created for one purpose generally address that goal but overtime those designs have quite unintended consequences (Streeck 1997a1997b Wright 1998) Describing the conguration of the German socialsystem of production of the 1970s and 1980s Streeck points out that itwas the unintended by-product of multiple points in history Someelements were pre-Wilhelmian others were introduced by the Alliedpowers after 1945 and others emerged during the years of the GermanFederal Republic All component parts of the German social system ofproduction

were and continue to be changing for their own reasons as wellas in reaction to each other and certainly there can be no presump-tion of a pre-established t between them even though one might

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

618

want to allow for some reinforcement effects of the modelrsquos histor-ically contingent social and economic success

(Streeck 1997b 54)

In sum the emergence of social systems of production is a long-termevolutionary process with each part interacting with its environment andresulting in a congurative whole but with no previous design by eithera single actor or a collectivity of actors As suggested above actors do not recreate their world anew At any moment in history the totalinstitutional complex of a society necessarily contains the resources and legacies of its past Preexisting institutional complexes are nevercompletely wiped out Societies consist of multiple layers of history withtheir diverse logics of action However the cumulative effect of smallperipheral changes in altering particular institutional sectors and insti-tutional arrangements can be substantial (Murmann 1998 Lanzara 1998)lsquoInstitution building is affected more by the ways in which people havecodied the past than by how they envision the future There are manymore potential resources in pre-existing institutional arrangements thanit is usually assumed or suspectedrsquo (Lanzara 1998 30)

Fourth level of analysis

The next level of institutional analysis organizational structures is some-what more controversial As suggested above Douglass North (1990)draws a sharp distinction between institutions and organizations Morerecently many organizational theorists (Powell 1991 Powell andDiMaggio 1991 DiMaggio and Powell 1983 Baum and Oliver 1992Kondra and Hinings 1998 Townley 1997) argue that institutional rulesnorms and conventions unfold in tandem with organizational structuresand this is my positionThe literature which focuses on how institutions inuence organiza-

tions is quite different from two other theoretical literatures which alsoare concerned with how organizational environments shape the behaviorof organizations There are resource dependency theorists (Pfeffer andSalancik 1978) who emphasize the role of environmental resources inshaping organizations while Hannan and Freeman (1977 1984) and otherpopulation ecologists emphasize the survival of organizations givencertain kinds of organizational conditions (Orru et al 1991)The institutionalist perspective as implied above emphasizes the

normative environment in which organizations are embedded It is a perspective which focuses on the way in which organizations in their behavior tend to conform to the institutional rules and normswhich are dominant in the organizational environment However allthree perspectives ndash the institutionalist the population ecology and

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

619

the resource dependency ndash emphasize how the environment inuencesorganizations and how organizations which are subject to the sameenvironment tend to converge in their behavior to have what DiMaggioand Powell have labeled organizational isomorphism (DiMaggio andPowell 1983)There has emerged a vast literature which demonstrates that within

each society there are scal political judicial and other regulatory normswhich limit and shape the culture and structure of organizationalbehavior The normative institutional environment of organizations limits the options of what organizations do in a particular society and inuences the patterns of ownership relations with suppliers andcustomers In short it is the normative environment of organizationswhich denes within a particular society what is socially acceptablebehavior for organizations (Hamilton and Biggart 1988 Hollingsworthand Hollingsworth 2000 Meyer and Rowan 1991 Zucker 1987 1988Orru et al 1991 Townley 1997)Thus far most of the analyses on isomorphic pressures operating on

organizations have been cross-sectional in nature However the entireconcept of isomorphism implies that there are strong environmental pres-sures exerted on organizations and in order to observe this phenomenonit is important that we have longitudinal studies which assess howchanges in an organizationrsquos institutional environment inuence changesin the structure and culture of organizations Moreover the expectationhere is that when one analyzes the patterns of organizational changeover time the isomorphism will be holistic ndash that is within a particularsociety there will be a tendency for the internal structure of particulartypes of organizations to converge in large part because of the pres-sures to conform to the changes in the external norms rules and valuesA close reading of Aoki (1990) and Slack and Hinings (1994) nds thatchanges in the national institutional environment of organizations haveinuenced changes in the internal structure of organizations within aspecic industry of a particular societyHowever it is inappropriate to view organizations as changing only

in relation to exogenous environmental change There is a kind of co-evolution which occurs between organizations and their institutionalenvironments Co-evolution implies nonlinear feedback between orga-nizations and their institutional environments This is of course quitedifferent from the more common way of studying organizations bymodeling relations between independent and dependent variables Ofcourse the more common type of analysis is quite appropriate whenthere are simple relationships which do not involve complex feedbackprocesses But in the strategy for institutional analysis which is proposedherein it is less useful to separate independent from dependent vari-ables and more useful to understand the interacting and co-evolutionary

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

620

processes among all of the levels of institutional analysis discussed inthe various sections above (Baum and Singh 1994a 379ndash80 Mowery andNelson 1999 Murmann 1998)The logic of the perspective herein suggests that within every

society there is variation in the structure and culture of business rmsuniversities and other complex organizations (Kondra and Hinings1998) However this variation takes place within parameters which aresystem specic For example every German rm and every Germanuniversity is different from every other German university and rm but there are a set of cultural and structural characteristics which distin-guish German rms and universities from those in America Similarlyeach Japanese rm and university is unique yet they have a set ofcultural and structural characteristics which are system specic andwhich differentiate them from business rms and universities in other countriesIn all societies each organization has its own distinctive organizational

rules norms and conventions which are subordinate to the meta norms and rules of the larger society within which they are embeddedHowever the strength of the institutional environment within whichorganizations are embedded varies from society to society Some societies have multiple institutional environments and there is hetero-geneity in terms of what constitutes appropriate behavior oforganizations In such societies there can be conict over whatinstitutional logics should regulate specic organizational functions(Friedland and Alford 1991 Townley 1997 262ndash3) In those societies inwhich the institutional norms habits and rules are most developed and in which the institutional pressures to conform are greatest thereis less variation in the structure and culture of business rms and variouskinds of research organizations In such societies the connectednessbetween research organizations and their institutional environments is sufciently strong for organizations to have low autonomy to pursue independent strategies and goals in these societies there is a great deal of organizational isomorphism Conversely the weaker the institutional environment in which research organizations areembedded the greater the variation in the structure and culture of business rms and research organizations Moreover where the institu-tional environments are more weakly developed organizations havegreater autonomy and exibility to respond to the development of new knowledge and to be highly innovative Hence it is in those societies where the institutional environments are most developed and most rigid and when there is less organizational autonomy and exibility that fewer radical innovations in basic and applied science as well as in totally new products and industrial sectors havebeen made

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

621

Fifth level of analysis

We come next to the outputs and performance of the various institutionalcomponents of a society It is at this level that institutional componentsare more pragmatic and exible For example in the legal sector thereare specic statutes and court rulings in the state sector there are specicpolicies in the business sector there are new products new technolo-gies and market strategies It is at this nal level of analysis thatinstitutional spheres are most open and susceptible to change and cross-national mimicry is easiest and most commonThrough the outputs of the society we can obtain some assessment as

to how well a society is performing Moreover we can assess how inno-vative it is how egalitarian it is in the distribution of its resources howegalitarian it is in terms of levels of health education etc Just as it iscomplicated to measure how egalitarian a society is similarly it is dif-cult to assess how innovative it is For example societies vary in makingincremental and radical breakthroughs in basic and applied science in developing totally new products and new kinds of organizations inincremental and process innovations in existing products and organiza-tions in developing new and different forms of marketing ndash bothdomestically and globally (Hage and Hollingsworth 2000) Of coursesocieties vary in their rates of economic growth in their rates of economicproductivity in their quality of life (eg rates of crime life expectancyetc) All of these performance criteria feed back and inuence each ofthe levels of institutional analysis discussed above ndash rules norms valuesetc institutional arrangements institutional sectors the structure andform of organizations Moreover performance measures may inuenceeach other For example if a society has very low rates of growth itmay not be very innovative in some of the types of innovative activitiesmentioned above Good or poor performances of certain kinds inuenceother performance indicators (Hollingsworth et al 1990)Different institutional arrangements and different social systems of

production result in different types of economic performance Hence solong as societies have different social systems of production there areserious constraints on the degree to which they can converge in theirinnovative styles Different social systems of production tend to maxi-mize in a more or less explicit manner different performance criteriausually mixing considerations about static and dynamic efciency protsecurity social peace and economic andor political power In short incontrast to the implications of neoclassical economic theory in real-worldeconomies there are no universal standards all economically rationalactors attempt to maximize Economic history provides numerous exam-ples of how a great variety of principles of rationality are implementedin different societies

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

622

A critical question in institutional analysis is whether an existing socialsystem of production which supports a set of routines for a particularkind of technology and industry can shift from old practices to newones There are numerous historical accounts of how the social systemof production in a particular society which worked so well for a numberof years could not adapt to new technologies (Lazonick 1990 Veblen1915 Nelson 1994) Schumpeter (1983) and Freeman (1991) have devel-oped the idea of a techno-economic paradigm which suggests thatdifferent technologies require specic forms of organizational arrange-ments ndash or what I called social systems of production Because oflsquoinstitutional inertiarsquo the social system of production of a society maynot be able to adapt to a new techno-economic paradigm (Nelson 199458 Lanzara 1998)Whether a social system of production can sustain its performance

standards depends not only on its intrinsic economic rationality but alsoon where it ts into a larger system If a particular social system ofproduction is immune from the innovativeness and competition of analternative system survival can be long lasting But if different socialsystems of production with diverging criteria of good economic perfor-mance meet in the world arena the arbitrariness of nationally imposedconstructed performance standards may be superseded by alternativeperformance criteria as a result of international competitivenessThe world economy is also socially constructed just as are national

economies Even if different social systems of production are competingin the international arena it is not always possible to determine whichis more competitively effective at any moment in time Hegemonicnation-states in the short run shape the rules of trade to favor their indus-trial sectors and rms But the history of hegemonic powers suggeststhat in the longer run such social systems of production sustainedlargely by military and political power eventually give way to moreinnovative and competitive social systems of production In our ownday as nation-states are increasingly integrated into a world economyeven hegemonic powers lose their innovative and competitive advan-tage and their share of world output decreases Such a country mayattempt to restructure its institutional arrangements and to readjust itsperformance preferences But to restructure its system of innovationgenerally calls for a major redistribution of power within a societyLargely for this reason societies have historically had limited capacityto reconstruct a system of innovation in the image of their major competi-tors Each countryrsquos social system of production is a conguration of ahost of norms rules and values as well as of institutional arrangementsEach system is constantly changing and is open to inuence from othersystems And indeed many technologies and practices diffuse from onesociety to another but a societyrsquos capacity to be innovative is constrained

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

623

by the existing social system of production Thus the same technologymay exist in numerous countries but how it is employed and how itinuences societal outcomes and performances varies from one institu-tional conguration or society to another (Hollingsworth 1998)

PERSPECTIVES ON INSTITUTIONAL CHANGE

A fundamental problem which remains unresolved in institutionalanalysis involves the nature of institutional change There is much confu-sion and miscommunication in the social sciences about institutionalchange in large part because we do not even know how to measure therate of institutional change Of course we know that there are bothexternal and internal focuses for change and we have had a good bitof scholarship on this subject What is not clearly understood is thatinstitutions and institutional arrangements within societies are histori-cally rooted that there is a great deal of path-dependency to the waythat institutions evolve Moreover the more intricately linked each sectoris with each other sector and with a societyrsquos rules and norms the lesschoice actors have to devise new institutions and institutional arrange-ments On the other hand if a sector or actor is too isolated it may betoo weak within a society to bring about any kind of effective changeThese considerations are very critical to the problem of how muchfreedom actors have to create new institutions afresh to what extentinstitutions diffuse from one society to another (Murmann 1998 Lanzara1998) This is a fundamental question for students of institutionalanalysis as we attempt to understand how the different institutionallevels are linked togetherThere are a variety of reasons why there is confusion in studying insti-

tutional change There is disagreement over the extent to which actorshave the freedom to build new institutional arrangements and if theydo to what extent the new arrangements and organizational patternsmay depart from past practices Among some institutional economists(Williamson 1985) there is a basic assumption that actors build efcientinstitutions by pursuing their self-interest and by promoting theirstrategic goals Moreover there is the assumption that a mixing of fullrationality with market competition tends to produce all of the optimalinstitutions that are needed in order to coordinate a complex capitalisteconomy (Hollingsworth and Boyer 1997 ch 14 Hodgson 1999) Incontrast the argument developed here is much more complexSocial institutions are historically rooted and there is a great deal of

path-dependency in the way that various institutional componentsevolve The shape of institutional congurations at any moment limitsthe type of options for change Because there is a great deal of institu-tional inertia radical change in the institutional components of a society

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

624

is uncommon (Lanzara 1998) Recognizing the constraints existinginstitutions exert on actors we note that existing congurations of insti-tutional arrangements institutional sectors and organizational propertiescan limit but also inuence the degree to which norms and rules canchange Many scholars (North 1990 Ostrom 1990) have emphasized therole of norms rules ideologies and values in limiting the ability of actorsto develop new institutional arrangements which radically depart fromexisting ones However there is a two-way process of interaction aprocess of co-evolution (Baum and Singh 1994a Nelson 1994) Thus asFriedland and Alford (1991 244) argue actors develop new institutionalarrangements by recognizing organizational failures and low perfor-mance but these new arrangements do not necessarily change a societyrsquosnorms rules habits and its underlying value systemCampbell (1997) advances our understanding of how existing institu-

tions enable actors to construct radically new institutions He recognizesthat most institutions and institutional arrangements embody a degreeof inertia from which it is difcult for actors to depart Moreover thedifferential power relations among actors make it unlikely that lesspowerful actors can change the existing power structure of a societyMost change in the institutional components of a society evolves througha process of constrained selection which reects to a considerable degreethe existing arrangements and power relations This becomes the inter-pretative frame for social actors with institutionalized scripts and ritualswhich tend to be taken for granted and appear to be quite ordinary andnatural This institutionalized scripting of the social world is an extremelyimportant source of social stability and inertiaCampbell has developed a very fruitful explanation of change in terms

of the various institutional components outlined here He distinguishesbetween incremental and radical institutional change and argues thatradical institutional change occurs when social actors with widelydiffering norms cultural scripts and rituals for action engage in intenseinteraction with each other (see also Knight 1992) The more diverse theinteractions the greater the potential for institutional change He arguesthat as shifts in the composition of interaction among social actors occurchanges in the interpretation of problems and interests will follow Ifthere are only minor changes in the extent to which diverse actorsinteract there will be only modest institutional change But the morefundamental the changes in the interaction of diverse social actors thegreater the change in the way that actors interpret their world And thegreater the change in the way that actors interpret their world the morelikely that radical changes in various institutional components willfollow In other words diverse decision making tends to create a widerrange of interpretative frameworks of the social world than is likely tobe the case if social actors operate in isolation

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

625

This perspective of diverse interactions among social actors facilitatesour ability to move beyond the traditional view that existing institutionsconstrain the range of institutional alternatives which actors face Theolder view has tended to suggest that actors engage in institutionalchange by extending existing institutional principles habits and conven-tions Specifying the conditions which facilitate fundamental institutionalchange permits a much richer perspective for institutional analysisCampbellrsquos perspective about the conditions under which fundamentalor radical change in institutional components occurs is consistent withmy own research about the conditions under which major discoveriesor fundamental new knowledge occur the more scholars with diversebackgrounds interact with intensity and frequency the greater the like-lihood that they will develop new and alternative ways of thinking(Hollingsworth and Hollingsworth 2000)These views of institutional change are also consistent with the recent

work of Nelson and his colleagues (Mowery and Nelson 1999 Murmann1998) They report the research on a variety of industrial sectors in anumber of different countries revealing that networks of actors in partic-ular industries business and professional associations universityresearch departments and institutes are frequently able to use their collec-tive power to modify a societyrsquos property rights research system andgovernmental policies It is the intense and frequent interaction of thesediverse complex sets of actors ndash sometimes constituting a technologysystem ndash that often leads to incremental changes in a societyrsquos total socialsystem of production It is this kind of collective action that brings aboutorganizational change change in institutional sectors institutionalarrangements and even in the rules and norms of a society And if weare to understand how the style of innovativeness of a society changeswe must rst understand the changes in its institutional makeup It ishighly unlikely that a society will develop a new style of innovativenesswithout changing its institutional structure

FIRMS THEIR INSTITUTIONAL ENVIRONMENT ANDNATIONAL SYSTEMS OF INNOVATION

Variation in the innovative style of societies is shaped by their insti-tutional makeup ndash the various levels which are discussed aboveinstitutions institutional arrangements the structure and cohesion of institutional sectors which constitute a societyrsquos social system of pro-duction and the structure and culture of its organizations ndash especiallyits business rms and other research organizations By undertaking aninstitutional analysis of a society one can begin to understand in whatkinds of organizations the production of specic kinds of knowledgetakes place and how this is linked to particular kinds of innovativeness

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

626

With such a perspective one can gain a rich understanding of why somesocieties excel in the production of radical breakthroughs in basic andapplied science and in developing radically new products and why othersocieties excel in more incremental innovationsReecting on these issues one cannot help being impressed with the

contrasts between the innovative styles of German and American orga-nizations For example Germany has continued to be very successful in making incremental innovations in many industries in which it wasvery competitive well before World War II paper printing materialsmachinery electrotechnical products motor vehicles chemicals textileyarns and fabrics But the Germans have been much less successful indeveloping totally new products or related innovations in biotechnologyelectronics telecommunications aircraft as well as other newer indus-trial areas In contrast the Americans have been very innovative in these and other industries with products having very short lives andwith technologies which change very rapidly and are very complex(Hollingsworth 1991 Soskice 1997 1999)The importance of institutions and innovativeness for international

competitiveness has led to the recognition that there are national systemsof innovation (Lundvall 1992 Nelson 1993 Edquist 1997 Hage andHollingsworth 2000) But even before the recent studies of nationalsystems of innovation Landes (1969) was a proponent of these viewswhen he made the argument that Germany had developed a nationalsystem of innovation by building on its education and scientic researchsystems which were lacking in Britain and elsewhere However the concept of a national system of innovation immediately poses theproblem of what kind of innovations should be consideredDavid Soskice an economist at the Wissenschaftszentrum Berlin has

addressed a set of problems somewhat complementary to the perspec-tive raised here His work (Soskice 1990 1997 1999) suggests that weare sensitive to the institutional environment in which the followingactors are embedded the employees of rms the owners and nanciersof rms the competitors of rms and nally their collaborators ndash aperspective complementary to Hollingsworthrsquos work on social systemsof production (Hollingsworth 1997 1998) When one reects on the insti-tutional contextsocial system of production within which these actorsare embedded one becomes aware that those countries which have some-what inexible labor markets external to rms but a business system in which rms have long-term commitments to their employees arestrongly associated with incremental but hardly at all with radical inno-vations In the rms of such countries (eg Germany and Japan) thereis much more consensus decision making than in societies where a lsquohireand rersquo set of practices is quite pervasive And this kind of consensusdecision making ndash while highly conducive to incremental innovativeness

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

627

ndash limits the capacity for radical innovations to occur This kind of lsquolongtermismrsquo set of norms and rules occurs not only in the employmentpractices of incrementally innovative rms but also in their nancingand ownership patterns (Hollingsworth 1997) During the past thirtyyears these were characteristics of many rms in Germany and JapanSignicantly in both of these countries venture capital markets werenot highly developed nancing tended to be based more on long-termbank loans than on the equity markets and the relationships of rmstended to be relatively stable And these stable institutional arrange-ments also contributed to a high degree of innovativeness ndash butincremental rather than radical innovationsMeantime we should also pursue Soskicersquos ideas (1999) about how

rmsrsquo relationships with their competitors and collaborators shape theirstyles of innovativeness In those advanced industrial societies in whichbusiness rms are highly mobilized into business associations theretends to be a high degree of standardized quality controls over technicalnorms and a high degree of reliance on legal sanctions as monitoringinstruments vis-agrave-vis association members As a result of this high degreeof integration of rms into associations member rms are more likelyto be associated with incremental innovations than is likely to be the case with rms in those societies where associations are less welldeveloped In the latter type of society radical innovations are morelikely to occur In those advanced industrial societies where rms havelong-term stable relationships with collaborators in product design andproduction the style of innovation is likely to be incremental And it is in those societies in which there is a weaker tradition of strongcollaboration among rms that radical innovations are more likely tooccurObviously even in a country where there may be a great deal of

radical innovativeness most innovations will be of an incremental typefor there are simply fewer radical innovations But Germany and therest of Europe not only have had relatively few radical innovations incomparison with the United States but they also have lagged behindthe USA in developing relatively totally new industries Thus there hasbeen much less innovative activity in Europe in high-technology andnewly emerging industries than in the USAIn complementary research Audretsch and Feldman (1995) have

demonstrated that the USA has been more successful in spawning rmsengaged in the early stages of the industry life cycle but relative toGermany had an environment unfavorable to innovative activity byestablished rms especially in more mature industries German rmshave tended to be associated with a routinized technological regime ndashwhere established enterprises have had strong incentives to engage inincremental innovations but new rms have tended to have fundamental

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

628

innovative disadvantages relative to the Americans (see Audretsch 1995Winter 1982 Casper 2000)On both sides of the Atlantic universities engage in basic and applied

research But signicantly the institutional environment in which univer-sities are embedded inuences the degree to which the research is likelyto represent major breakthroughs or is likely to have more modestimplications Recent research in biomedical science demonstrates that inthose societies in which university professors andor graduate studentshave the following characteristics there are likely to be relatively majorresearch breakthroughs graduate students have a long training periodhigh dependency on a major professor low autonomy in dening thenature of a research project restricted job options and a university envi-ronment where there is a very modest or nonexistent entrepreneurialculture For example in Germany ndash a country with relatively few radicalbreakthroughs ndash the scholar who nishes a habilitation will have spentmany years on a research project before qualifying for a permanent acad-emic position will have been in a high-dependency position with a seniorprofessor will have had limited autonomy to dene a research projectand will be constrained from pursuing a high-risk research project Theyoung German will have an extraordinarily high level of training butwill have faced a number of disincentives to undertake high-risk researchAnd as soon as heshe has completed the habilitation and becomes aprofessor administrative responsibilities constitute another disincentivefrom engaging in high-risk or long-term research projects Moreover theGerman professor is a civil servant with a more stable and secure incomethan hisher American counterpart In a very different institutionalenvironment American professors in large research universities tend tobecome much more entrepreneurial Today many American professorsbehave as though they were operating their own rms within univer-sities and thus devote considerable time to generating funds for researchprojects assistants secretaries supplies etc And the variation in thisentrepreneurial spirit across countries is associated with variation in thetaking of high-risk research strategies and the making of major researchbreakthroughsIn sum societies vary in terms of the institutional environment in

which their organizations are embedded and this variation inuencesthe degree to which they have industrial activities in the early or latestages of the industry life cycle successful new or mature industriesand radical vs incremental innovations And it is this variation whichis encouraging rms in countries with relatively few new industries and with few or no radical innovations to invest in those countries with many new industries and numerous radical innovations ndash as astrategy to be able to transfer new technologies from the latter institu-tional environment to the former

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

629

Table 4 summarizes the way that the norms rules and values of a society are associated with Americarsquos and Germanyrsquos institutionalarrangements their social system of production and their styles of inno-vativeness

CONCLUSION

Several themes deserve to be reemphasized Most importantly institu-tional analysis is emerging as a dominant issue in several social sciencedisciplines The study of various institutional components discussedabove is a complex subject requiring multiple levels of analysis and aknowledge of multiple sectors within societies It requires a theoreticalperspective that evolves from an empirical analysis in different societiesIf we are to advance the study of institutional analysis we need cross-disciplinary collaboration and team research which are not readilyachieved given the structure of most of our research universitiesAt present scholars in each academic discipline have their own distinc-

tive strategies for studying institutional analysis Thus economists tendto use several approaches (North 1990 Williamson 1985 Hodgson 19881998 1999) anthropologists another (Geertz 1995) and sociologistspolitical scientists and historians employ other strategies (Campbell etal 1991 Powell and DiMaggio 1991 Katznelson 1998 Hall and Taylor1996) With practitioners in various disciplines pursuing their distinctiveapproaches to institutional analysis there is low potential for collabo-rative institutional analysis across the social sciences The lessons fromthe biological sciences over the last 40 years are very instructive oncepractitioners from various biological disciplines began to cooperate instudying biological phenomena by using the same concepts at the mole-cular level theoretical advance was very rapid Comparable phenomenatook place in the development of other hybrid elds eg biochemistrybiophysics (Judson 1979 Olby 1979 Kohler 1982 1994) Similarly weresocial scientists in various disciplines to work collaboratively in the areaof institutional analysis there would very likely be an acceleration intheir theoretical knowledgeThis is not a plea that everyone engaged in institutional analysis should

do the same kind of research Indeed just as biologists working at themolecular level study many different kinds of problems similarly socialscientists engaged in institutional analysis would also work on manyseparate problems but in the spirit of a collective enterprise Some wouldconcentrate their attention on the study of rules norms habits conven-tions and values while others would study how these are associatedwith congurations of various institutional arrangements (types ofmarkets hierarchies networks associations communities clans statesetc) Others would work on specic institutional sectors (eg education

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

630

Table 4

Rules norms hab

its an

d values associated w

ith sp

ecic co

ngu

ration

s of coo

rdinating mecha

nism

s soc

ial sy

stem

s of

prod

uction

an

d nationa

l styles of inno

vation

USA

Germany

Norms rules values which provide incentives for

Norms rules values which provide incentives for

short-term horizons

long-term horizons

Con

gu

rative

forms of

Marke

ts c

orpo

rate hierarchies reg

ulatory state

Assoc

iation

s (busine

ss assoc

iation

s la

bor union

s)

coordinating

corp

oratist-type

networks

marke

ts c

orporatist-

mecha

nism

s (listed in

type

state

order of im

portanc

e in

the particu

lar social

system

)

Social sys

tem of

prod

uction

Bus

iness system

Firm

struc

tures

Con

glom

erates b

ut with mov

emen

t toward

Firm

more co

ngru

ent with pa

rticular prod

ucts

elim

ination of produc

tion

in som

e sectors

emph

asis on long

-term strateg

ies rm

s are well

emph

asis on sh

ort-term

strateg

ies high

integrated

differentiation

between divisions

of rms

Firm

s mob

ilized in

Low

High

and in

tegrated

into busine

ss

associations

Indu

strial relations

Low

deg

ree of jo

b secu

rity rigid in

ternal lab

or

High deg

ree of jo

b secu

rity

exible interna

l lab

or

system

marke

t ex

ible externa

l labo

r marke

tsmarke

t rigid externa

l lab

or m

arke

tPatterns of

Dispersed owne

rship frequ

ent turnov

erMore stab

le owne

rship patterns

owne

rship

Table 4

continue

d

USA

Germany

Norms rules values which provide incentives for

Norms rules values which provide incentives for

short-term horizons

long-term horizons

Social system of

Prod

uction

Bus

iness system

Training for labo

r an

dHigh deg

ree of voluntary decision mak

ing by

Collective an

d m

ore co

mpulso

ry d

ecisions

abo

ut

man

agem

ent

individuals as to ho

w m

uch

tim

e an

d ene

rgy

how m

uch ed

ucation

is ne

eded

he

avy em

pha

sis

to inv

est in educ

ation heavy

emph

asis on

on techn

ical training (eg en

gine

ering)

busine

ss sch

ools training

in marke

ting

and

sales accou

nting

Fina

ncial marke

tsEqu

ity an

d ven

ture cap

ital m

arke

ts highly

Equ

ity an

d ven

ture cap

ital m

arke

ts w

eakly

dev

elop

ed na

ncing from

cen

tral gov

ernm

ent

deve

lope

d he

avy em

pha

sis on

corporate loan

sto priva

te rm

s for dev

elop

men

t of new

tech

nologies related

to na

tion

al security

Unive

rsity research

High individual auton

omy high poten

tial for

Low

ind

ividua

l au

tono

my until almost 40

yea

rs of

system

individual creativity u

nive

rsity pr

ofesso

rs

age poten

tial for ind

ividual creativity high

ly

high

ly entreprene

urial

cons

traine

d unive

rsity pr

ofessors not so

entrep

rene

urial

Inno

vative

styles

Highly su

ccessful in m

ajor break

through

s in

High pe

rforman

ce in proc

ess an

d inc

remen

tal

basic an

d app

lied scien

ce go

od in dev

elop

ing

inno

vation

s in existing produ

cts not so su

ccessful

totally

new

lines of pr

oducts an

d in m

ajor break

through

s in basic scien

ce or in

tech

nologies high

deg

ree of entrepr

eneu

rship

deve

loping

totally new

lines of prod

ucts of

amon

g research

ers

tech

nologies low d

egree of entrepr

eneu

rship

amon

g research

ers

Table 4

continue

d

USA

Germany

Norms rules values which provide incentives for

Norms rules values which provide incentives for

short-term horizons

long-term horizons

Prod

ucts in w

hich

Pr

oducts with sh

ort liv

es prod

ucts in w

hich

Older produ

cts with long

history machine

ry pap

er

society excels

tech

nology

is bo

th highly co

mplex

and

pr

inting

textile

s m

otor veh

icles m

achine

too

lsch

ange

s rapidly en

tertainm

ent indus

try

pha

rmaceu

ticals biotechn

olog

y softw

are

aerosp

ace

business systems nancial markets and systems of research) Distinctivewould be the recognition of how these separate studies are linkedtogether how each institutional sector is linked to a societyrsquos normsrules values etc and its conguration of institutional arrangementsSimilarly other institutional analysts might focus on the study of orga-nizations with a major concern for how the institutional environmentof organizations (eg norms rules values congurations of institutionalarrangements institutional sectors) inuenced the structure culture andoutcomes of organizations And nally others might study how all ofthe aforementioned aspects of institutional analysis inuenced a societyrsquosoverall performance Nelson and Sampat (1998) have appropriatelyreminded us that we should engage in institutional analysis in order toaddress specic problems but the view developed in this article is thatwe rst need some mapping of the terrain of institutional analysis inorder for our diverse enterprises to become a meaningful collective enter-prise especially if we are to understand how institutions inuence theinnovative processThe argument developed here is not that all social scientists should engage

in institutional analysis Obviously there are numerous other importantareas of research which lie beyond the eld of institutional analysis Nordoes this article argue that institutional analysis is the most importantsubject for social scientistsSince the structure and culture of our research universities reect an

enormous amount of disciplinary fragmentation what are the prospectsfor promoting the kind of collective and interdisciplinary research agendaproposed above In one respect the prospects are not encouragingDuring the past half century the efforts of American universities to have effective communication across the biological sciences have variedgreatly in their success At the University of California-Berkeley it was not until the 1980s that there was a serious effort to integrate thebiological sciences and at that time the campus abolished seventeendepartments to promote an integrated biological program But at theUniversity of Wisconsin the biological sciences are still very fragmentedinto dozens of different departments in botany plant pathology zoologyand physiology with two departments of biochemistry and two depart-ments of geneticsThere have been other approaches which have worked very well histor-

ically for promoting scientic integration amidst scientic diversity Forexample several summer schools and special institutes were organizedin the United States during the 1940s in an effort to overcome the frag-mentation in the biological sciences These programs made it possiblefor scientists in diverse academic disciplines and from different univer-sities and countries to come together to engage in intense and frequentinteraction Perhaps the most effective program was that held in the

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

634

summers at Cold Spring Harbor New York where a research agendafor the study of molecular biology was developed (Fischer and Lipson1988 Stent and Watson 1966) Scientists went there summer aftersummer and when they returned to their universities many began therestructuring of the biological sciences in their universitiesIn our own day we need summer institutes and other similar programs

so that scholars from diverse backgrounds can come together anddevelop a collective agenda for doing institutional analysis and forstudying the impact of institutions on the performance of societies ndashinnovativeness being one such subject If we can develop such instituteswe will increase the prospects for transforming our universities Thereis reason to be optimistic that once an eclectic group of scholars workwith a common set of concepts we can have genuine theoretical advanceif researchers can bring their diverse expertise to bear on a common setof problems There is considerable evidence that institutional analysisoffers high potential for advancing our knowledge of innovations andtechnological change (Hollingsworth and Hollingsworth 2000) If we aregoing to make substantial theoretical advance it is not likely to occurfrom the vantage point of a multiplicity of disciplines Rather it is likelyto occur as we develop a new intellectual framework and our own acad-emic journals This is a tall agenda which requires us to reassess theway that knowledge is presently produced in our societies

NOTE

This article is part of a much larger agenda involving the study of institutionsorganizations and innovations in which I am involved My debts are many andvaried The rst version was written while I was in residence in the ResearchUnit for Institutional Change and European Integration of the Austrian Academyof Sciences in Vienna in the summer of 1998 I am especially grateful to ProfessorsEgon Matzner and Sonja Puntscher-Riekmann for providing the stimulating envi-ronment in which to think through the ideas developed here Egon Matzner hasengaged in very stimulating discussions with me about the importance of insti-tutions in the general eld of socio-economics The nal version of the articlewas completed during my residence as a fellow at the Neurosciences Institutein La Jolla California I also wish to thank my colleagues in the InnovationTheme Project at the Netherlands Institute for Advanced Study (NIAS) during1998ndash9 and Edgar Grande and Karl Muumlller for stimulating conversations aboutthe interaction of institutions and organizations in facilitating particular typesof innovation I am especially grateful to Jerald Hage for helping me to developa number of issues in the article and for his useful comments on the rst draftFrans van Waarden early on helped me to work through the idea of multiplelevels of institutional analysis Others who have helped to develop the ideas inthe article have been my colleagues in the research project lsquoComparative socialsystems of productionrsquo especially Tom Burns Christel Lane Yoshitaka OkadaWolfgang Streeck and Richard Whitley My colleagues Robert Boyer SteveCasper and David Soskice have taught me much about how incentives providedby the institutional environments of rms inuence individual decision making

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

635

But my most important debt is to Ellen Jane Hollingsworth who has repeatedlyquestioned every single assumption of this article Without her rich and contin-uous tutoring this article would be even less developed than it is at presentFinally without the insightful assistance of David Gear in many phases of thisarticle it would not have been completed

REFERENCES

Aoki Masahiko (1990) lsquoTowards an economic model of the Japanese rmrsquo Journalof Economic Literature March 1ndash27

Arora A Landau Ralph and Rosenberg Nathan (eds) (1998) Chemicals and Long-Term Economic Growth Insights from the Chemical Industry New York JohnWiley amp Sons

Arrow Kenneth (1974) The Limits of Organization New York NortonArthur W B (1988a) lsquoSelf-reinforcing mechanisms in economicsrsquo in P W

Anderson K J Arrow and D Pines (eds) The Economy as an Evolving ComplexSystem Redwood City Calif Addison-Wesley

mdashmdash(1988b) lsquoCompeting technologies an overviewrsquo in G Dosi C Freeman RNelson G Silverberg and L Soete (eds) Technical Change and Economic TheoryLondon Pinter Publishers pp 590ndash607

Audretsch David B (1995) Innovation and Industry Evolution Cambridge MassMIT Press

Audretsch David B and Feldman Maryann P (1995) lsquoInnovative clusters andthe industry life cyclersquo Discussion Paper Wissenschaftszentrum BerlinMarket Processes and Corporate Development Research Unit FS-IV-95-7

Baum Joel A C and Oliver Christine (1992) lsquoInstitutional embeddedness andthe dynamics of organizational populationsrsquo American Sociological Review 57540ndash59

Baum Joel A C and Singh Jitendra V (1994a) lsquoOrganizationndashenvironment coevo-lutionrsquo in Joel A C Baum and Jitendra V Singh (eds) The EvolutionaryDynamics of Organizations New York Oxford University Press pp 379ndash402

mdashmdash(eds) (1994b) The Evolutionary Dynamics of Organizations New York OxfordUniversity Press

Berger Suzanne and Dore Ronald (eds) (1996) National Diversity and GlobalCapitalism Ithaca NY Cornell University Press

Boyer Robert (1997) lsquoThe variety and unequal performance of really existingmarkets farewell to Doctor Panglossrsquo in J Rogers Hollingsworth and RobertBoyer (eds) Contemporary Capitalism the Embeddedness of InstitutionsCambridge and New York Cambridge University Press pp 55ndash93

Burns Tom R and Dietz Thomas (2001 forthcoming) lsquoSocial rule system theorysocial action institutional arrangements and evolutionary processesrsquo in JRogers Hollingsworth and Karl Muumlller (eds) Advances in SocioeconomicsLanham Md Rowman amp Littleeld

Burns Tom R and Flam Helena (1987) The Shaping of Social Organization BeverlyHills Calif Sage

Calvert Randall L (1995) lsquoThe rational choice theory of social institutionsrsquo inJeffrey S Banks and Eric A Hanuskek (eds) Modern Political Economy NewYork Cambridge University Press pp 216ndash66

Camic Charles (1986) lsquoThe matter of habitrsquo American Journal of Sociology 911039ndash87

Campbell John L (1997) lsquoMechanisms of evolutionary change in economicsgovernance interaction interpretation and bricolagersquo in Lars Magnusson

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

636

of disciplines and subdisciplines To establish some coherence to the eldof institutional analysis we need a map of the eld so that those workingin one area can see where their research ts in relation to other areasand other practitioners on the mapTable 1 presents such a map with multiple levels at which institu-

tional analysis occurs Theoretically each of these areas on the map isinterrelated with each other level However the various areascompo-nents on the map are arranged in descending order of stability orpermanence Those components at higher levels of reality are morepermanent and durable while those at lower levels change more rapidlyWere scholars doing institutional analysis able to reach some consensus

about where their own work ts in relation to all other practitioners inthe eld there would be increased potential for all practitioners tocommunicate with each other By analogy once geneticists crystallo-graphers biochemists etc had a good understanding of how theirinvestigations were related to each area of molecular biology the eldquickly was able to make theoretical advances (Judson 1979)At the rst level there are the basic norms rules conventions habits and

values of a society These are the most fundamental properties of insti-tutions and are the most enduring and resistant to change Rules normsconventions etc are institutions but are only one component of whatconstitutes institutional analysis As Burns and Dietz (2001 forthcoming)point out most human activity is organized and regulated by normsand rules and systems of rules These concepts are extremely importantfor institutional analysis as they exert the greatest inuence on the natureof the components of institutional analysis at the next four levels which

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

601

Table 1 Components of institutional analysis

1 Institutions = norms rules conventions habits and values (see North1990 Burns and Flam 1987)

2 Institutional arrangements = markets states corporate hierarchies networksassociations communities (Hollingsworth and Lindberg 1985 Campbell et al 1991 Hollingsworth et al 1994 Hollingsworth and Boyer 1997)

3 Institutional sectors = nancial system system of education businesssystem system of research (Hollingsworth 1997)

4 Organizations (Powell and DiMaggio 1991)5 Outputs and performance = statutes administrative decisions the nature

quantity and the quality of industrial products (Hollingsworth 1991 1997)sectoral and societal performance (Hollingsworth and Streeck 1994Hollingsworth et al 1990 Hollingsworth and Hanneman 1982)

Note The ve components in this table are arranged in descending order of permanenceand stability That is norms conventions etc are more enduring and persistent than eachof the other components of institutional analysis Each component is interrelated withevery other component and changes in one are highly likely to have some effect in bringingabout change in each of the other components

are depicted in Table 1 In most forms of social analysis it is extremelyimportant that we understand the social and cognitive conditions thatlead to compliance or non-compliance with rules and the conditionswhich lead to changes in rulesThe approach to the study of institutions employed here argues that

norms rules habits conventions and values both reect and shape thepreferences of actors Norms rules habits conventions and values inu-ence who and what are included in different types of decision makinghow information is processed and structured what action is taken(Shepsle 1986 1989) It is through norms and rules that behavior isjudged to be democratic fair or egalitarianBurns and Flam (1987) point out that in any society there are multiple

rule systems Within a family there are rules for decision making oftenquite different from rules and norms for decision making for a professorin a classroom or for the customer in a bank Despite the heterogeneityof rule systems there are meta rules and norms which encompass lesserrule systems Otherwise there would be such contradictory rule systemsthat a society would be paralyzed The existence of meta rule systemspermits different rule systems to intersect with each other so that ambi-guities can be resolved The greater the pluralism and complexity of asociety the more ambiguity there is about meta rules and norms in asociety and of course all ambiguities never completely disappearOverall the degree to which separate rule systems are interlinked is anempirical problem There are different sectors groups and interestspursuing their own action logic but through higher-order meta princi-ples and rules there can be order consensus and coherence in a societyIt is through a set of meta rules that class and ethnic conict in societiesare contained (For elaboration see Burns and Flam 1987)In many respects our understanding about norms rules habits con-

ventions and values inuences our perspective on how societies are con-structed and how they change lsquoNew institutionalistsrsquo (Posner 1992Schotter 1981 Williamson 1975 1985) tend to assume that at one timethere was a state of nature and that there was a movement from individ-uals to institutions ndash an approach often called methodological individual-ism (Popper 1961 Hodgson 1998 1999) And of course there areinnumerable instances which methodological individualists cite to demon-strate that individuals create new rules of behavior For example it is pos-sible for actors to change the rules of driving so that instead of drivingon the left side of the road drivers adopt a new rule and drive on the rightThis article however tends to equate social habits and institutions

As Hodgson and others remind us (Hodgson 1988 1989 1997 1999Grafstein 1992 Camic 1986 Johnson 1992 Nelson and Winter 1982Veblen 1899) social habits are the results of earlier choices and are ameans of avoiding endless deliberation Because cognitive frameworks

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

602

are learned through habit individuals rely on the acquisition of suchcognitive habits before reason communication choice or action arepossibleWhereas Schotter (1981) and other game theorists take the individual

as an agent unencumbered by previous habits Field (1984) and othershave stressed that there can be no games without prior norms and rulesand thus a set of norms and rules must be presumed at the start Thosewho attempt to explain institutions from individual behavior alone areusing a bad strategy (Hodgson 1998)The position here ndash heavily inuenced by Hodgson (1998 1997 1988)

ndash is that individuals are embedded in a complex institutional environ-ment and that institutions not only constrain but also shape individuals(also see Hollingsworth and Boyer 1997) It would be a mistake howeverto get involved in an innite regress in order to determine which camerst ndash individuals or institutions Of course institutions are formed andchanged by individuals just as individuals are shaped and constrainedby institutions But at a macro level it is institutions that provide acognitive framework whereby individuals can cope with their reality Inthis sense the micro and macro worlds are intertwined At the macrolevel there is considerable stability but at the micro level individualshave a signicant level of autonomy and there can be widespread diver-sity As Hodgson (1988) reminds us most institutions in a temporal senseexist prior to the individuals in any given societyIt would be a serious mistake to downplay the importance of indi-

viduals and micro level analysis as we study institutions In the nalanalysis it is at the level of individuals that norms rules habits conven-tions and values exist An individual is born into and socialized intogroups and a society and this is how one early in life acquires a senseof appropriate forms of behavior Because of the way that individualsare socialized into a world of rules norms habits conventions andvalues it is unnecessary for individuals to restructure the world anewevery day (Douglas 1987 Elster 1989) Every action does not have tobe seriously reected upon For this reason institutions provide cogni-tive frameworks for individuals make their environments predictableprovide the information for coping with complex problems and envi-ronments In the words of Johnson (1992 26) lsquoInstitutions reduceuncertainty coordinate the use of knowledge mediate conicts andprovide incentive systems By serving these functions institutions providethe stability necessary for the reproduction of societyrsquo However eachsociety has different forms of habits rules and norms and hence differentincentives and disincentive systems for learning and forgetting forprocessing information But because individuals have varying degreesof autonomy individuals and groups can deviate from the prescribedforms of behavior in a society And of course these changes at the level

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

603

of individuals become important in understanding processes of societalchangeThese views are not meant to imply that the type of institutional

analysis proposed herein approximates a general theory of society Thisis clearly not the case However it is intended to represent the rst stepsin a mapping exercise of the boundaries of institutional analysis and tosuggest a few methodological insights for studying institutionsOne should try to see norms and rules as continuous and not as

dichotomous entities to recognize that they come in varying strengthsLegro (1997) has suggested that we assess the robustness of norms andrules with three criteria their simplicity their durability and their concor-dance Simplicity refers to how well actors understand norms and ruleshow well they can be applied within a specic situation Some normsand rules can be so complex that actors can have considerable difcultyin applying them in specic cases Durability addresses the issue of howlong norms and rules have been in effect ndash in short in order to assesstheir level of legitimacy While the position of this article is that normsrules and values are quite durable they do vary in this respectConcordance refers to how widely applied a norm or rule is Thisaddresses the degree to which a rule is a meta rule the degree to whichit incorporates the heterogeneity of other norms and rules In sum theclearer the norms and rules of a society the longer they have been inexistence and the more widely applicable they are the greater theirimpact on a society Hence the more robust the norms and rules thegreater their impact on a society and the less their robustness the greatertheir exibility and the less their effect on shaping a societyrsquos outcomesand performancesBecause norms rules and values are quite durable they play an impor-

tant role in shaping the history of societies thus contributing to a greatdeal of path-dependency Actors attempt to adjust to their contempo-rary environment but since they are products of the past the historicallegacy of norms rules and values inuences the decisions they makeAlthough actors have some capacity to alter the course of their historythey are constrained by their past and the degree to which they canmove beyond their past is limited As Lanzara (1998) Johnson (1992)and others have argued societal inertia is a basic feature of institutionsThey provide the basic stability necessary for change Degrees of historyare continuously reproduced by the way in which the inhabitants ofsocieties are socialized History matters but at critical points in historythere is punctuated equilibrium (Somit and Peterson 1992) During mostperiods of history there is considerable stability in the norms rules andvalues of a society but at critical moments norms rules and values canquickly and dramatically be redened At all times norms rules habitsconventions and values are inuencing each of the other components in

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

604

the institutional framework discussed below but these other institutionalcomponents feed back and can modify rules norms conventions etc(Murmann 1998)

Second level of analysis

The norms rules habits conventions and values of a society lead to thenext level of analysis ndash the institutional arrangements which are involvedin the coordination of various economic actors producers and suppliersof raw materials knowledge etc processors of raw materials infor-mation workers customers of raw materials nished productsinformation etc nanciers governmental and other types of regulatorsThese actors regularly engage in contests to resolve various economicproblems in virtually all sectors of society how are prices to be setWhat quantity of various products is to be produced How are stan-dards of various products processes etc to be set How is the qualityof products and processes to be determined How are various societalprocesses to be nanced In order to confront these problems and toaddress the conicting positions of economic actors as they address these problems societies develop various institutional (governance)arrangements for coordinating different actors These consist of marketsvarious types of hierarchies and networks associations the state commu-nities and clans (see Hollingsworth and Boyer 1997 Campbell et al1991 ch 1)Each of these particular kinds of coordinating mechanism has many

different types For example there are many types of states (eg theregulatory state developmental state authoritarian state welfare state)on which there is an extensive literature (Kim 1997) Similarly there aredifferent types of markets networks different kinds of associations etc(Boyer 1997 Hage and Alter 1997 Schneiberg and Hollingsworth 1990)When we do institutional analysis we must engage in congurativeanalysis recognizing that actors are not coordinated or governed by asingle type of institutional arrangement Some of the literature discussesindustrial sectors as though they were coordinated or governed by asingle institutional arrangement whereas in fact each sector of aneconomy is coordinated by a conguration of institutional arrangementsSome congurations coordinate actors in certain problem areas whileother congurations of institutional arrangements coordinate actors inaddressing other problems The types of congurations which are domi-nant in a society are somewhat stable and tend to persist over timewithin a society (Hodgson 1999)When one mode of coordination is dominant in a society it will

inuence the role which other coordinating modes will play Hence thestrong role of the state in the Soviet Union inuenced the role of markets

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

605

associations etc in the governance of the Soviet economy Similarly theprominence of particular modes of coordination in a society inuencesits style of innovationMuch of the literature on institutional arrangements (eg forms of eco-

nomic coordination) remains fragmented and unintegrated It is helpfulto array these various forms of coordination in a two-dimensional tax-onomy as in Figure 2 On the vertical dimension the economistrsquos viewof a self-interested agent is contrasted with a more sociological perspec-tive according to which social rules obligation and compliance shapehuman actions On the horizontal dimension there is another continuum(ie the distribution of power) At one extreme of the dimension onends many and relatively equal agents interacting (eg as in a well-organized spot market) At the other extreme inequality in power resultsin a hierarchical form of coordination which structures the interactionbetween principals and agents or between leaders and followers Wherea society falls along the distribution of power is inuenced by the rulesnorms values etc which are dominant in a society at a particular momentin timeInstitutional arrangements can be visualized on two dimensions the

nature of action motive on the one hand and the distribution of poweron the other Markets (cell 1) combine self-interest with horizontalcoordination transactors and they reect sensitivity to concerns aboutsupply and demand thus providing ex post an unintended equilibriumParadoxically the more pure and perfect the market competition thegreater the need for codied rules of the games for coordinating economictransactions Thus collective associations (cell 6) andor various formsof state intervention (cell 4) are required to develop and enforce rulesfor transacting partners (Schneiberg and Hollingsworth 1990 Streeckand Schmitter 1985a) This is an example of how the norms and rulesand the institutional arrangements of a society are intertwined asreected in a conguration of institutional or governance arrangementsAlong the horizontal axis actors can be joined in an organization or

a rm a hierarchy is the generic terms for this institutional arrangement(cell 2) Along the horizontal line one recognizes the difference betweentransactions in a market and transactions within a rm The well-knownworks of Coase (1960 1981) and Williamson (1975 1985) utilize theconcept of transaction costs in explaining the emergence of corporatehierarchiesThere are also various types of networks (cell 5) which exhibit mixes

of self-interest and social obligation with some actors being formallyindependent and equal Yet in some networks (the large rms and theirsubcontractors) there is unequal power and initiative Networks maycomprise all kinds of actors some consist only of rms but others includeassociations and the state (Hage and Alter 1997)

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

606

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

607

DISTRIBUTION OF POWER

VERTICALHORIZONTAL

1 MARKETS 2 HIERARCHY

6ASSOCIATIONS

5NETWORKS

4STATE

3COMMUNITIES

SELF INTEREST

SOCIAL OBLIGATION

ACTION MOTIVE

Figure 2 A general taxonomy of institutional arrangements

The vertical axis deals with action motives Toward the upper part ofFigure 2 actors are engaged in individualistically oriented behaviorwhereas toward the lower part actors are more engaged in collectivebehavior and strive to cope with problems of common interest Cell 3 ndashcommunities and clans ndash consists of institutional arrangements based ontrust reciprocity or obligation and thus not derived from the pure selshcomputation of pleasures and pains This is an unconventional form ofcoordination for most neoclassical economists (however see Arrow1974) but not for many anthropologists political scientists and sociolo-gists (Streeck and Schmitter 1985a Polanyi 1944 Gambetta 1988Fukuyama 1995 Sabel 1992 Putnam 1993)In the neoclassical paradigm theorists argue that actors engage in

forms of exchanges that best promote their individual interests If somestructural conditions are fullled (absence of increasing returns to scalethe reversibility of transactions absence of uncertainty and completecontingent markets with no collusion between actors) then the invis-ible hand theorem applies and market-type activity functions quite welland also provides the optimum for society thereby combining efciencyharmony and order However an excess of market activity may welllead to ruinous competition and excessive conict There is variation inthe extent to which ruinous competition occurs depending on the socialcontext in which transactions take place Thus it is important that webe sensitive to the institutional context in which transactions areembedded and that we understand the degree to which social bondsexist at both the micro and macro levels of analysis Micro bonds facil-itate exchanges in a society but at the societal level social bonds existat the level of the collective ndash in the community or region and amongmembers of racial religious and ethnic groups All other things beingequal the more powerful the social bonds among transacting partnersthe more economic competition is likely to be restrained Thus mosttransactions occur not simply in an impersonal calculative system ofautonomous actors unrestrained by social ties ndash as implied by theneoclassical paradigm ndash but in the context of social ties variation in the strength of which leads to variation in levels of trust and transac-tion costs (Etzioni 1988 211 Granovetter 1985 Streeck and Schmitter1985a Hollingsworth and Boyer 1997 Hodgson 1988 1999 Schneibergand Hollingsworth 1990)Another form of multilateral institutional arrangement is various

types of associations (cell 6) Unlike networks clans and communitiesassociations are more formal organizations Whereas markets corporatehierarchies and networks tend to coordinate economic activity amongdifferent types of actors (eg producers with suppliers capital withlabor) associations typically coordinate actors engaged in the same orsimilar kinds of activities Business associations and labor unions are

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

608

some of the most common forms of associations for coordinatingeconomic activity in capitalist economies (Schneiberg and Hollingsworth1990 Streeck and Schmitter 1985a)Finally there is the state (cell 4) which is an institutional arrangement

quite unlike any of the others It is the state that sanctions and regu-lates the various non-state coordinating mechanisms that is the ultimateenforcer of rules of the various mechanisms that denes and enforcesproperty rights and that manipulates scal and monetary policy Overallit is the state that inuences the total incentive system of a society(Lindberg and Campbell 1991 Johnson 1992 40) At the same time thestate may also be an economic actor by engaging directly in productionand exchange relationsThe choices of institutional arrangements in Figure 2 are constrained

by the social context in which they are embedded Depending on thenature of that embeddedness there is variation in the collective formsof governance some of which are specied in the lower part of thegure Some modes of coordination in the bottom and upper levels ofthe typology are often mixed together though one coordinating modeis likely to be more prominent than another But actors are often engagedin complex congurations involving several kinds of institutionalarrangementsEach of these various institutional arrangements has its own logic ndash

its own rules its own procedures for enforcing compliance its ownnorms and ideologies which help to reduce the costs of enforcementThese are summarized in Table 2 which provides further elaborationabout the various coordinating mechanisms one nds in almost everycapitalist society Table 2 lists the organizational arrangements and theirstructures rules of exchange and means of enforcing compliance asso-ciated with each type of coordinating mechanism While each type ofinstitutional arrangement has various positive features each institutionalarrangement also has failures and these are featured in Table 3 It is thecontest between those who support and those who oppose these variousinstitutional arrangements that tends to lead to transformations in insti-tutional arrangements over time (Campbell et al 1991 Campbell 1997)Again it is difcult to conceive of pure institutional arrangements ndash

either exclusively markets or exclusively hierarchies ndash since it has beenwell known since Adam Smithrsquos Wealth of Nations that the division oflabor within the rm cannot be disentangled from the existence andextent of the market Institutional arrangements have their own distinc-tive form of efciency (good static efciency for the market dynamicefciency for rms) and inefciency often leading to considerableinequality Neither networks nor communities are panaceas for economiccoordination without being congured with other types of institutionalarrangements Networks and communities may solve certain issues but

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

609

Table 2

Log

ics of ins

titution

al arran

gemen

ts

Coordination

Organizational

Rules of

Individual means

Collective means

mechanisms

structure

exchange

of compliance

of compliance

Marke

tsEasy en

try an

d exit

Voluntary sp

ot excha

nge

Leg

al enforcemen

t of

Norm of pr

ivate

Bila

teral ex

chan

ge or

control

prope

rty

marke

t site (W

all Street)

Reg

ulations

to

Leg

itim

acy of m

arke

t en

force contracts

men

tality

Com

mun

ities

Inform

al m

embe

rship

Voluntary exch

ange

based

So

cial norms an

d m

oral

Highly institutiona

lized

ev

olve

s ov

er lon

g period

on social solid

arity an

d

princ

iples im

pose

norm

s an

d rules requ

ire

of tim

ehigh

deg

ree of tru

stob

ligations

mem

bers to accept

Kno

wledg

e of others an

d

lsquocorpo

ratersquoo

bligations

recipr

ocity ov

er tim

e

Networks

Semi-form

al m

embe

rship

Voluntary exch

ange

ove

r a

Con

tractual bon

dsPersona

l relation

sBila

teral or m

ultilateral

time pe

riod

Resou

rce de

pend

ence

Tru

st built outside

the

ex

chan

geecon

omic arena

Associatio

nsFo

rmal m

embe

rship

Restricted to mem

bers

Self-interest

Some deg

ree of

Multilateral excha

nge

Emph

asis on insider

Rep

utation

effects

compu

lsion

outsider or we

they

Priva

te interest type

of

men

tality

gove

rnan

ce

Private

Com

plex organ

izations

Restricted to mem

bers

Rew

ards to ind

ividuals

Highly institutiona

lized

hierarch

ies

which

ten

d to becom

e ex

chan

ge based

on

Asymmetric po

wer threat

rules

bureau

cratic hierarchies

asym

metric po

wer

of san

ctions

Mem

bers soc

ializ

ed into

bureau

cratic rules

corp

orate cu

lture us

e of

sanc

tions

State

Pub

lic hierarchy

Unilateral action

Exit vo

ice (vote lob

bying)

Coe

rcion

De jure

and impo

sed

Indirect political an

d

loya

ltyNorms an

d pu

blic rules

mem

bership

econ

omic excha

nge

Table 3

Failu

res of coo

rdinating mecha

nism

s

Coordinating mechanisms

Type of failure

Associations

Private hierarchies

State

Market

Communities

Networks

Enforcem

ent

Usu

ally relies on

Might enh

ance

Needs co

ntrol

Needs an

internal

Needs

tru

st and

Need an external

the state as an

oppo

rtun

istic

extern

al to state

enforcem

ent

loya

lty often

en

forcem

ent

enforcer

beha

vior

bureau

cracy

authority

coming from

au

thority

Resem

bles

The

ideal o

f (ju

dge

s

Facilitates

outside (fam

ily May

facilitate

enforcem

ent

internal m

arke

ts

parliamen

t collu

sion

and

relig

ion

cartelization an

d mecha

nism

might hur

t marke

t) to

impe

rfect

ethn

icity)

mon

opoly

of cartels

incu

mbe

nt

correct state

compe

titio

nCom

patible with

worke

rsab

uses

variou

s types of

Lob

bies can

co

mpe

tition

captur

e pu

blic

interest goa

ls

Public goo

d

Useful for

Gov

erna

nce co

sts

Can

provide

Can

not provide

Can

internalize

Useful for

and

establishing

might exceed

public goo

ds

colle

ctive go

ods

some co

llective

enha

ncing qu

ality

externality

stan

dards an

d

the be

nets of

but ha

s or deal w

ith

good

s (qua

lity

and training bu

t qu

ality for

internal division

difc

ultie

s in

externalities

training

) bu

t no

t no

t ve

ry goo

d in

setting ru

les of

of la

bor

prov

iding them

Inad

equate

othe

rs (welfare

prov

iding for

compe

titio

n in

Slow

to react to

in precise

mon

itoring of

gene

ral p

ublic

societal gen

eral

the indu

stry

chan

ges in the

am

ounts

tech

nical ch

ange

go

ods)

welfare

Useful for

environm

ent

Might fail in

and inno

vation

Mem

bers tightly

Weak in the

prov

iding

induc

ing

integrated

into

prov

ision of

man

y go

ods

tech

nical

community

colle

ctive go

als

colle

ctively

chan

geha

ve limite

d

that in

dividua

lcapa

city for

mem

bers can

not

inno

vatio

nsprov

ide for

them

selves

Table 3

continue

d

Coordinating mechanisms

Type of failure

Associations

Private hierarchies

State

Market

Communities

Networks

Efcien

cyFa

cilitates

De

cien

t in

Can

be high

ly

Some ba

sic social So

me go

ods

Slow

to en

hanc

e co

operation an

d

coop

eration an

d

bureau

cratic

relation

s cann

ot

cann

ot be

efc

ienc

y an

d

X-efcien

cy but

X-efcien

cyan

d can

not

be provided

by

delivered

at

speed of

not allocative

easily deliver

pure m

arke

t su

fcien

tly

adap

tive

ness

efcien

cy

good

s at

mecha

nism

slow costs

except in

low cost

indus

tries whe

re

tech

nology

is

complex an

d rapidly cha

nging

Equity

Narrow

Excessive

Might enh

ance

Facilitates

Might lead to

Whe

n widely

encompa

ssing

multiplication

ineq

uality

ineq

uality in

retard

ed

dev

elop

ed in

to

associationa

l of con

trollers

(pow

er

income an

d dev

elop

men

tindus

trial

stru

ctures lead

(fru

stration

and

an

d privilege

)wealth

districts

to in

come

ineq

uality)

netw

orks m

ay

equa

lity

facilitate greater

equality an

d

inco

me

distribution

whe

n wea

kly

dev

elop

ed

netw

orks ten

d to

increase soc

ial

ineq

uality

Note For allo

cativ

e efc

ienc

y an

d X

-efcien

cy see Leibe

nstein (19

66 19

76)

they raise other no less severe problems It is important to recognizethe imperfection of any single institutional arrangement in order tocomprehend the origin and transformation of any other institutionalarrangement (Hodgson 1999 ch 3)As suggested above each institutional arrangement is congured with

other institutional arrangements Usually one particular institutionalarrangement is more dominant in a particular conguration but becauseeach type of institutional arrangement has its own strengths andweaknesses there is no simple structural logic in the governance or coor-dination of a society Each institutional arrangement constrains the otherbut the inherent tension among the various institutional arrangementswithin a conguration contributes to changes in the governance orcoordination of a society The routines and logics of each of these cong-urations of institutional arrangements provide constraints and incentivesfor actors The way that actors perceive the incentives and constraintsof these governance congurations leads to a particular market logicand it is the specic market logic of a society which inuences its speciccapacities and weaknesses The inherent strain among the different logicsin a conguration of coordination helps to provide the exibility for asociety to adapt to new circumstancesIf there were a society with pure markets or a society coordinated

only by the state there would be too much rigidity and too little diver-sity to cope with the vast uncertainty in the global environment A societywith very little diversity in its coordination mechanisms would havelittle capacity to adapt to new circumstances Congurations with consid-erable diversity of institutional arrangements provide for a certainamount of incoherence in governance but they also provide for thecapacity to adapt to new circumstances While the Soviet Union wasdominated by the state as a coordinating mechanism there were alwaysfunctioning markets in the system Moreover the feudal society of thetwelfth and thirteenth centuries consisted not only of hierarchical rela-tionships between serfs and masters but also of urban guilds clericalhierarchies and markets (Johnson 1992 38) In sum the robustness ofinstitutions often depends on multiple and diverse principles and logicsof action on the inconsistency of principles and procedures on patternedforms of disorder (Lanzara 1998 Orren and Skowronek 1991 320 329)

Third level of analysis

The next level consists of the institutional sectors of a society The rulesnorms and values of a society inuence the array of institutional arrange-ments and levels one and two inuence the nature of and the relation-ships among various institutional sectors and all three levels areintricately linked together to form a social system of production Together

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

613

all of these components inuence the performance of economic sectorswithin a society as well as the performance of the total society A socialsystem of production is the way that a societyrsquos institutions (see rstlevel of analysis above) its institutional arrangements (see second levelof analysis above) and its institutional sectors are integrated into a socialcongurationAn institutional sector includes all organizations in a society which

supply a given service or product along with their associated focal orga-nizations (eg major suppliers funding sources regulators and so forth(Scott et al 1994 108 117 Campbell et al 1991 Hollingsworth et al1994) Institutional sectors include but are not limited to the societyrsquossystem of education system of research business system nancialmarkets legal system and the state The structure of the nancial marketssystem of training and education industrial relations system system ofresearch and state and legal systems are distinct and idiosyncratic ineach society In short they are system specicAll of these institutional sectors tend to cohere with each other

although they vary in the degree to which they are tightly coupled intoa full-edged system While each of the institutional sectors has someautonomy and may have some goals that are contradictory to the goalsof other institutional sectors with which it is linked an institutional logicin each society leads institutions to coalesce into a complex social con-guration (Hollingsworth 1991) This occurs because the institutionalsectors are embedded in a culture in which their logics are symbolicallygrounded organizationally structured technically and materially con-strained and politically defended The conguration of institutionalsectors usually exhibits some degree of adaptability to new challengesbut continues to evolve within an existing style But under new circum-stances or unprecedented disturbances these institutional congurationsare exposed to sharp historical limits as to what they may or may notdo (Schumpeter 1983 David 1988 Arthur 1988a 1988b Haringkanssen andLundgren 1997)Why do all of these different institutions coalesce into a complex

social conguration which is labeled here as a social system of pro-duction The literature suggests two contrasting interpretations Part ofthe answer ndash indeed a controversial one ndash is that these institutionalsectors are functionally determined by the requirements of the practiceof capitalism in each time and place (Habermas 1975) Another expla-nation emphasizes the genesis of the actual conguration via a trial and error process according to which the survival of rms regions orcountries is the outcome of complex evolutionary mechanisms (Maynard-Smith 1982 Nelson and Winter 1982) However the problem is evenmore complex Markets and other mechanisms for coordinating rela-tionships among economic actors place constraints on the means and

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

614

ends of economic activity to be achieved in any society The othercoordinating mechanisms include different kinds of hierarchies varioustypes of networks and associations ndash trade unions employers and busi-ness associations (Hollingsworth and Lindberg 1985 Campbell et al1991) The logic of these various institutional arrangements providesactors with vocabularies and logics for pursuing their goals for deningwhat is valued and for shaping the norms and rules by which theyabide (see Tables 2 and 3) In short in contrast to the logic of theneoclassical paradigm the argument here is that the dominant type ofinstitutional arrangements places severe constraints on the denition of needs preferences and choices of economic actors Whereas the neo-classical paradigm assumes that individuals and rms are sovereign thisarticle is based on the assumption that rms are inuenced by the hold that the institutional congurations making up a social system ofproduction have on individual decision making (Campbell et al 1991Etzioni 1988 Streeck and Schmitter 1985a Hollingsworth et al 1994Hollingsworth and Boyer 1997 Magnusson and Ottosson 1997 North1990 Hodgson 1999)Standard neoclassical economic theory has tended to downplay the

role of production and consequently of rms Even transaction cost theo-rists who are concerned with analyzing the rm as a coordinatingmechanism have been relatively unconcerned with the various compo-nents of a social system of production Indeed as long as there waswidespread optimism about the efcacy of Keynesian economics therewas relatively little concern among neoclassical economists with thesupply side of the economy Even in the opinion of most Keynesians agroup of experts should ideally be able to shape the size of aggregatedemand while the supply side of the economy would be left to the twominimalist institutions of neoclassical economics ndash markets and manage-rial hierarchies For several decades however it has become increasinglyobvious that some of the most competitive and successful patterns ofindustrial output and industrial production in capitalist economies donot derive from the neoclassical prescription of unregulated markets andcorporate hierarchies complemented by a neoliberal democratic stateIndeed it is now well understood that certain highly successful produc-tion patterns require for their emergence and survival institutionalarrangements and environments the very opposite of the prescriptionsfound in the neoclassical paradigm (see especially Streeck 1992 but alsoHollingsworth and Streeck 1994) Thus social scientists have come tounderstand that if they are to comprehend the behavior and performanceof contemporary economies concerns about social systems of produc-tion must be brought into the pictureBecause production involves more than technology a number of social

scientists have an increasing concern with social systems of production

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

615

The same equipment is frequently operated quite differently in the samesectors in different countries even when rms are competing in the same market (Maurice et al 1980 Sorge and Streeck 1988 Sorge 1989Hollingsworth et al 1994) Variations in production and process tech-nologies are inuenced partly by variations in the social environmentsin which they are embedded In other words rms are embedded intocomplex environments which among other things place constraints ontheir behavior Thus a social system of production is of major impor-tance in understanding the behavior and performance of an economyHow the state and other institutional arrangements (eg marketsnetworks private hierarchies associations) coalesce and are related toparticular social systems of production are important determinants ofeconomic performanceThe congurations of institutional arrangements that coordinate or

govern the behavior of actors in one society and its structure of specicinstitutional sectors cannot easily be transferred to another society forthey are embedded into a social system of production that is societallydistinct (Hollingsworth 1997) Societies borrow selected principles offoreign management styles and work practices but the effectiveness ofsuch borrowing is generally limited Economic behavior and performanceare shaped by the entire social system of production in which actors areembedded and not simply by specic principles of particular manage-ment styles and work practices Moreover a societyrsquos social system ofproduction tends to limit the kind of goods that it can produce and withwhich it can compete successfully in international marketsSome scholars (Kenney and Florida 1993 Oliver and Wilkinson 1988)

have assumed that the diffusion of particular forms of management stylesand work practices across societies could lead to system convergenceReferring to the Japanese production system Kenney and Florida haveargued that it consists of organizational practices whose fundamentallsquogenetic codersquo can successfully be inserted into another society and beginsuccessfully to reproduce its behavior in the new environment (1993 8)Their position is in the intellectual tradition of Antonio Gramsci (1971)who contended decades ago that the American system of mass produc-tion would most certainly diffuse to Europe over timeHowever the argument of this article is that even though British

French and American rms may adopt certain aspects of Japanesemanagement styles (eg just-in-time production self-managing teamsquality circles the use of statistical process controls) or some variant ofthe German vocational system their social systems of production willnot converge Moreover the overall conguration of a societyrsquos socialsystem of production inuences its sectoral and overall national economicperformance This helps to explain why societies have different systemsof innovation and why they vary in the clusters of industries in which

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

616

they are highly competitive in the international markets (Nelson 19931999 Hage and Hollingsworth 2000 Berger and Dore 1996)In order to understand how and why a societyrsquos economy performs

as it does it is necessary to understand its entire social system of produc-tion If a society is to modify substantially the performance of itsinstitutional sectors and its entire economy it cannot adapt only someof the management and work practices of its foreign competitors Ratherit must alter its entire social system of production Because a societyrsquosmodes of economic governance and coordination and its institutionalsectors develop according to particular logics and are system specicthere are serious limitations to the extent to which a society may mimicthe institutions (eg the rules and norms) institutional arrangements andinstitutional sectors of other societiesIn the history of modern societies there are logics by which institu-

tions coalesce into a social system of production (Hollingsworth 1991)Though institutions are constantly changing there are sharp limits tothe type and the direction of change that any particular institution canundergo because of its linkages with institutional arrangements and insti-tutional sectors Thus a societyrsquos business rms educational systemnancial markets industrial relations system etc can engage in seriousrestructuring only if most of the other institutional sectors also changeIn social systems of production there are pressures toward consistencyin the norms rules and values across institutional sectors though in anycomplex society social systems are obviously imperfectly integratedIndeed the degree to which the institutional norms and rules makingup a social system of production are loosely or tightly coupled is a vari-able of considerable importance In general the institutional sectorsmaking up a social system of production are interdependent and changesin one generally result in changes in the others Since each institutionallevel is dependent on the others for various types of resources there isinterdependence among the differing institutional spheres Moreovereach society has its norms moral principles rules and laws and recipesfor action as well as its own idiosyncratic customs traditions and prin-ciples of justice (Burns and Flam 1987)There are also other inherent obstacles to convergence among social

systems of production for where a system is at any one point in timeis inuenced by its initial state Systems with quite different initial statesare unlikely to converge with one anotherrsquos institutional practicesExisting institutional arrangements block certain institutional innova-tions and facilitate others (Roland 1990) Thus the institutions makingup a social system of production provide continuity even though insti-tutional arrangements are always changing but with a particular logicWhile Williamson (1975 1985) suggests that actors tend to choose theinstitutional arrangements which are most efcient North (1990) is

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

617

much closer to the mark in his argument that most societal institutionalarrangements exist as a result of custom and habit and are thereforeinefcient At any moment in time the world often appears to its actorsas very complex and uncertain For this reason actors often engage incontradictory forms of behavior pursuing different strategies as hedgesagainst a very uncertain world (Lanzara 1998) And their hedging andcontradictory forms of behavior may lead in somewhat different soci-etal directions all constrained by the institutional fabric within whichthe actors are embedded This kind of contradictory behavior occurs inpart because of the contradictions inherent in the dominant congura-tions of institutional arrangements with their contradictory logics (seelsquoSecond level of analysisrsquo above)Despite the emphasis on the logic of institutional continuity this is

not an argument that systems change along some predetermined pathThere are critical turning points in the history of highly industrializedsocieties but with the choices limited by the existing institutional terrainBeing path-dependent a social system of production continues along aparticular logic until or unless a fundamental societal crisis intervenes(Milgrom et al 1991 Krugman 1991 Durlauf 1991 Hollingsworth 1991Pred 1966 David 1988 Hodgson 1998)At this point it is important to confront the question of how social

systems of production evolve And this question gets to the heart of theproblem of building complex societal structures Certainly social systemsof production did not emerge from some process of social engineeringMoreover the various component parts of each social system of produc-tion have often not been designed to be part of a social system ofproduction The component parts of each social system of productionhave emerged more often than not as unintended by-products of goalswhich various actors had in mind at earlier moments in time It is usuallythe case that actors at time lsquotrsquo fail to comprehend the long-term conse-quences of their actions Wolfgang Streeck the author of several brilliantpapers on social systems of production argues that institutional sectordesigns created for one purpose generally address that goal but overtime those designs have quite unintended consequences (Streeck 1997a1997b Wright 1998) Describing the conguration of the German socialsystem of production of the 1970s and 1980s Streeck points out that itwas the unintended by-product of multiple points in history Someelements were pre-Wilhelmian others were introduced by the Alliedpowers after 1945 and others emerged during the years of the GermanFederal Republic All component parts of the German social system ofproduction

were and continue to be changing for their own reasons as wellas in reaction to each other and certainly there can be no presump-tion of a pre-established t between them even though one might

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

618

want to allow for some reinforcement effects of the modelrsquos histor-ically contingent social and economic success

(Streeck 1997b 54)

In sum the emergence of social systems of production is a long-termevolutionary process with each part interacting with its environment andresulting in a congurative whole but with no previous design by eithera single actor or a collectivity of actors As suggested above actors do not recreate their world anew At any moment in history the totalinstitutional complex of a society necessarily contains the resources and legacies of its past Preexisting institutional complexes are nevercompletely wiped out Societies consist of multiple layers of history withtheir diverse logics of action However the cumulative effect of smallperipheral changes in altering particular institutional sectors and insti-tutional arrangements can be substantial (Murmann 1998 Lanzara 1998)lsquoInstitution building is affected more by the ways in which people havecodied the past than by how they envision the future There are manymore potential resources in pre-existing institutional arrangements thanit is usually assumed or suspectedrsquo (Lanzara 1998 30)

Fourth level of analysis

The next level of institutional analysis organizational structures is some-what more controversial As suggested above Douglass North (1990)draws a sharp distinction between institutions and organizations Morerecently many organizational theorists (Powell 1991 Powell andDiMaggio 1991 DiMaggio and Powell 1983 Baum and Oliver 1992Kondra and Hinings 1998 Townley 1997) argue that institutional rulesnorms and conventions unfold in tandem with organizational structuresand this is my positionThe literature which focuses on how institutions inuence organiza-

tions is quite different from two other theoretical literatures which alsoare concerned with how organizational environments shape the behaviorof organizations There are resource dependency theorists (Pfeffer andSalancik 1978) who emphasize the role of environmental resources inshaping organizations while Hannan and Freeman (1977 1984) and otherpopulation ecologists emphasize the survival of organizations givencertain kinds of organizational conditions (Orru et al 1991)The institutionalist perspective as implied above emphasizes the

normative environment in which organizations are embedded It is a perspective which focuses on the way in which organizations in their behavior tend to conform to the institutional rules and normswhich are dominant in the organizational environment However allthree perspectives ndash the institutionalist the population ecology and

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

619

the resource dependency ndash emphasize how the environment inuencesorganizations and how organizations which are subject to the sameenvironment tend to converge in their behavior to have what DiMaggioand Powell have labeled organizational isomorphism (DiMaggio andPowell 1983)There has emerged a vast literature which demonstrates that within

each society there are scal political judicial and other regulatory normswhich limit and shape the culture and structure of organizationalbehavior The normative institutional environment of organizations limits the options of what organizations do in a particular society and inuences the patterns of ownership relations with suppliers andcustomers In short it is the normative environment of organizationswhich denes within a particular society what is socially acceptablebehavior for organizations (Hamilton and Biggart 1988 Hollingsworthand Hollingsworth 2000 Meyer and Rowan 1991 Zucker 1987 1988Orru et al 1991 Townley 1997)Thus far most of the analyses on isomorphic pressures operating on

organizations have been cross-sectional in nature However the entireconcept of isomorphism implies that there are strong environmental pres-sures exerted on organizations and in order to observe this phenomenonit is important that we have longitudinal studies which assess howchanges in an organizationrsquos institutional environment inuence changesin the structure and culture of organizations Moreover the expectationhere is that when one analyzes the patterns of organizational changeover time the isomorphism will be holistic ndash that is within a particularsociety there will be a tendency for the internal structure of particulartypes of organizations to converge in large part because of the pres-sures to conform to the changes in the external norms rules and valuesA close reading of Aoki (1990) and Slack and Hinings (1994) nds thatchanges in the national institutional environment of organizations haveinuenced changes in the internal structure of organizations within aspecic industry of a particular societyHowever it is inappropriate to view organizations as changing only

in relation to exogenous environmental change There is a kind of co-evolution which occurs between organizations and their institutionalenvironments Co-evolution implies nonlinear feedback between orga-nizations and their institutional environments This is of course quitedifferent from the more common way of studying organizations bymodeling relations between independent and dependent variables Ofcourse the more common type of analysis is quite appropriate whenthere are simple relationships which do not involve complex feedbackprocesses But in the strategy for institutional analysis which is proposedherein it is less useful to separate independent from dependent vari-ables and more useful to understand the interacting and co-evolutionary

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

620

processes among all of the levels of institutional analysis discussed inthe various sections above (Baum and Singh 1994a 379ndash80 Mowery andNelson 1999 Murmann 1998)The logic of the perspective herein suggests that within every

society there is variation in the structure and culture of business rmsuniversities and other complex organizations (Kondra and Hinings1998) However this variation takes place within parameters which aresystem specic For example every German rm and every Germanuniversity is different from every other German university and rm but there are a set of cultural and structural characteristics which distin-guish German rms and universities from those in America Similarlyeach Japanese rm and university is unique yet they have a set ofcultural and structural characteristics which are system specic andwhich differentiate them from business rms and universities in other countriesIn all societies each organization has its own distinctive organizational

rules norms and conventions which are subordinate to the meta norms and rules of the larger society within which they are embeddedHowever the strength of the institutional environment within whichorganizations are embedded varies from society to society Some societies have multiple institutional environments and there is hetero-geneity in terms of what constitutes appropriate behavior oforganizations In such societies there can be conict over whatinstitutional logics should regulate specic organizational functions(Friedland and Alford 1991 Townley 1997 262ndash3) In those societies inwhich the institutional norms habits and rules are most developed and in which the institutional pressures to conform are greatest thereis less variation in the structure and culture of business rms and variouskinds of research organizations In such societies the connectednessbetween research organizations and their institutional environments is sufciently strong for organizations to have low autonomy to pursue independent strategies and goals in these societies there is a great deal of organizational isomorphism Conversely the weaker the institutional environment in which research organizations areembedded the greater the variation in the structure and culture of business rms and research organizations Moreover where the institu-tional environments are more weakly developed organizations havegreater autonomy and exibility to respond to the development of new knowledge and to be highly innovative Hence it is in those societies where the institutional environments are most developed and most rigid and when there is less organizational autonomy and exibility that fewer radical innovations in basic and applied science as well as in totally new products and industrial sectors havebeen made

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

621

Fifth level of analysis

We come next to the outputs and performance of the various institutionalcomponents of a society It is at this level that institutional componentsare more pragmatic and exible For example in the legal sector thereare specic statutes and court rulings in the state sector there are specicpolicies in the business sector there are new products new technolo-gies and market strategies It is at this nal level of analysis thatinstitutional spheres are most open and susceptible to change and cross-national mimicry is easiest and most commonThrough the outputs of the society we can obtain some assessment as

to how well a society is performing Moreover we can assess how inno-vative it is how egalitarian it is in the distribution of its resources howegalitarian it is in terms of levels of health education etc Just as it iscomplicated to measure how egalitarian a society is similarly it is dif-cult to assess how innovative it is For example societies vary in makingincremental and radical breakthroughs in basic and applied science in developing totally new products and new kinds of organizations inincremental and process innovations in existing products and organiza-tions in developing new and different forms of marketing ndash bothdomestically and globally (Hage and Hollingsworth 2000) Of coursesocieties vary in their rates of economic growth in their rates of economicproductivity in their quality of life (eg rates of crime life expectancyetc) All of these performance criteria feed back and inuence each ofthe levels of institutional analysis discussed above ndash rules norms valuesetc institutional arrangements institutional sectors the structure andform of organizations Moreover performance measures may inuenceeach other For example if a society has very low rates of growth itmay not be very innovative in some of the types of innovative activitiesmentioned above Good or poor performances of certain kinds inuenceother performance indicators (Hollingsworth et al 1990)Different institutional arrangements and different social systems of

production result in different types of economic performance Hence solong as societies have different social systems of production there areserious constraints on the degree to which they can converge in theirinnovative styles Different social systems of production tend to maxi-mize in a more or less explicit manner different performance criteriausually mixing considerations about static and dynamic efciency protsecurity social peace and economic andor political power In short incontrast to the implications of neoclassical economic theory in real-worldeconomies there are no universal standards all economically rationalactors attempt to maximize Economic history provides numerous exam-ples of how a great variety of principles of rationality are implementedin different societies

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

622

A critical question in institutional analysis is whether an existing socialsystem of production which supports a set of routines for a particularkind of technology and industry can shift from old practices to newones There are numerous historical accounts of how the social systemof production in a particular society which worked so well for a numberof years could not adapt to new technologies (Lazonick 1990 Veblen1915 Nelson 1994) Schumpeter (1983) and Freeman (1991) have devel-oped the idea of a techno-economic paradigm which suggests thatdifferent technologies require specic forms of organizational arrange-ments ndash or what I called social systems of production Because oflsquoinstitutional inertiarsquo the social system of production of a society maynot be able to adapt to a new techno-economic paradigm (Nelson 199458 Lanzara 1998)Whether a social system of production can sustain its performance

standards depends not only on its intrinsic economic rationality but alsoon where it ts into a larger system If a particular social system ofproduction is immune from the innovativeness and competition of analternative system survival can be long lasting But if different socialsystems of production with diverging criteria of good economic perfor-mance meet in the world arena the arbitrariness of nationally imposedconstructed performance standards may be superseded by alternativeperformance criteria as a result of international competitivenessThe world economy is also socially constructed just as are national

economies Even if different social systems of production are competingin the international arena it is not always possible to determine whichis more competitively effective at any moment in time Hegemonicnation-states in the short run shape the rules of trade to favor their indus-trial sectors and rms But the history of hegemonic powers suggeststhat in the longer run such social systems of production sustainedlargely by military and political power eventually give way to moreinnovative and competitive social systems of production In our ownday as nation-states are increasingly integrated into a world economyeven hegemonic powers lose their innovative and competitive advan-tage and their share of world output decreases Such a country mayattempt to restructure its institutional arrangements and to readjust itsperformance preferences But to restructure its system of innovationgenerally calls for a major redistribution of power within a societyLargely for this reason societies have historically had limited capacityto reconstruct a system of innovation in the image of their major competi-tors Each countryrsquos social system of production is a conguration of ahost of norms rules and values as well as of institutional arrangementsEach system is constantly changing and is open to inuence from othersystems And indeed many technologies and practices diffuse from onesociety to another but a societyrsquos capacity to be innovative is constrained

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

623

by the existing social system of production Thus the same technologymay exist in numerous countries but how it is employed and how itinuences societal outcomes and performances varies from one institu-tional conguration or society to another (Hollingsworth 1998)

PERSPECTIVES ON INSTITUTIONAL CHANGE

A fundamental problem which remains unresolved in institutionalanalysis involves the nature of institutional change There is much confu-sion and miscommunication in the social sciences about institutionalchange in large part because we do not even know how to measure therate of institutional change Of course we know that there are bothexternal and internal focuses for change and we have had a good bitof scholarship on this subject What is not clearly understood is thatinstitutions and institutional arrangements within societies are histori-cally rooted that there is a great deal of path-dependency to the waythat institutions evolve Moreover the more intricately linked each sectoris with each other sector and with a societyrsquos rules and norms the lesschoice actors have to devise new institutions and institutional arrange-ments On the other hand if a sector or actor is too isolated it may betoo weak within a society to bring about any kind of effective changeThese considerations are very critical to the problem of how muchfreedom actors have to create new institutions afresh to what extentinstitutions diffuse from one society to another (Murmann 1998 Lanzara1998) This is a fundamental question for students of institutionalanalysis as we attempt to understand how the different institutionallevels are linked togetherThere are a variety of reasons why there is confusion in studying insti-

tutional change There is disagreement over the extent to which actorshave the freedom to build new institutional arrangements and if theydo to what extent the new arrangements and organizational patternsmay depart from past practices Among some institutional economists(Williamson 1985) there is a basic assumption that actors build efcientinstitutions by pursuing their self-interest and by promoting theirstrategic goals Moreover there is the assumption that a mixing of fullrationality with market competition tends to produce all of the optimalinstitutions that are needed in order to coordinate a complex capitalisteconomy (Hollingsworth and Boyer 1997 ch 14 Hodgson 1999) Incontrast the argument developed here is much more complexSocial institutions are historically rooted and there is a great deal of

path-dependency in the way that various institutional componentsevolve The shape of institutional congurations at any moment limitsthe type of options for change Because there is a great deal of institu-tional inertia radical change in the institutional components of a society

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

624

is uncommon (Lanzara 1998) Recognizing the constraints existinginstitutions exert on actors we note that existing congurations of insti-tutional arrangements institutional sectors and organizational propertiescan limit but also inuence the degree to which norms and rules canchange Many scholars (North 1990 Ostrom 1990) have emphasized therole of norms rules ideologies and values in limiting the ability of actorsto develop new institutional arrangements which radically depart fromexisting ones However there is a two-way process of interaction aprocess of co-evolution (Baum and Singh 1994a Nelson 1994) Thus asFriedland and Alford (1991 244) argue actors develop new institutionalarrangements by recognizing organizational failures and low perfor-mance but these new arrangements do not necessarily change a societyrsquosnorms rules habits and its underlying value systemCampbell (1997) advances our understanding of how existing institu-

tions enable actors to construct radically new institutions He recognizesthat most institutions and institutional arrangements embody a degreeof inertia from which it is difcult for actors to depart Moreover thedifferential power relations among actors make it unlikely that lesspowerful actors can change the existing power structure of a societyMost change in the institutional components of a society evolves througha process of constrained selection which reects to a considerable degreethe existing arrangements and power relations This becomes the inter-pretative frame for social actors with institutionalized scripts and ritualswhich tend to be taken for granted and appear to be quite ordinary andnatural This institutionalized scripting of the social world is an extremelyimportant source of social stability and inertiaCampbell has developed a very fruitful explanation of change in terms

of the various institutional components outlined here He distinguishesbetween incremental and radical institutional change and argues thatradical institutional change occurs when social actors with widelydiffering norms cultural scripts and rituals for action engage in intenseinteraction with each other (see also Knight 1992) The more diverse theinteractions the greater the potential for institutional change He arguesthat as shifts in the composition of interaction among social actors occurchanges in the interpretation of problems and interests will follow Ifthere are only minor changes in the extent to which diverse actorsinteract there will be only modest institutional change But the morefundamental the changes in the interaction of diverse social actors thegreater the change in the way that actors interpret their world And thegreater the change in the way that actors interpret their world the morelikely that radical changes in various institutional components willfollow In other words diverse decision making tends to create a widerrange of interpretative frameworks of the social world than is likely tobe the case if social actors operate in isolation

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

625

This perspective of diverse interactions among social actors facilitatesour ability to move beyond the traditional view that existing institutionsconstrain the range of institutional alternatives which actors face Theolder view has tended to suggest that actors engage in institutionalchange by extending existing institutional principles habits and conven-tions Specifying the conditions which facilitate fundamental institutionalchange permits a much richer perspective for institutional analysisCampbellrsquos perspective about the conditions under which fundamentalor radical change in institutional components occurs is consistent withmy own research about the conditions under which major discoveriesor fundamental new knowledge occur the more scholars with diversebackgrounds interact with intensity and frequency the greater the like-lihood that they will develop new and alternative ways of thinking(Hollingsworth and Hollingsworth 2000)These views of institutional change are also consistent with the recent

work of Nelson and his colleagues (Mowery and Nelson 1999 Murmann1998) They report the research on a variety of industrial sectors in anumber of different countries revealing that networks of actors in partic-ular industries business and professional associations universityresearch departments and institutes are frequently able to use their collec-tive power to modify a societyrsquos property rights research system andgovernmental policies It is the intense and frequent interaction of thesediverse complex sets of actors ndash sometimes constituting a technologysystem ndash that often leads to incremental changes in a societyrsquos total socialsystem of production It is this kind of collective action that brings aboutorganizational change change in institutional sectors institutionalarrangements and even in the rules and norms of a society And if weare to understand how the style of innovativeness of a society changeswe must rst understand the changes in its institutional makeup It ishighly unlikely that a society will develop a new style of innovativenesswithout changing its institutional structure

FIRMS THEIR INSTITUTIONAL ENVIRONMENT ANDNATIONAL SYSTEMS OF INNOVATION

Variation in the innovative style of societies is shaped by their insti-tutional makeup ndash the various levels which are discussed aboveinstitutions institutional arrangements the structure and cohesion of institutional sectors which constitute a societyrsquos social system of pro-duction and the structure and culture of its organizations ndash especiallyits business rms and other research organizations By undertaking aninstitutional analysis of a society one can begin to understand in whatkinds of organizations the production of specic kinds of knowledgetakes place and how this is linked to particular kinds of innovativeness

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

626

With such a perspective one can gain a rich understanding of why somesocieties excel in the production of radical breakthroughs in basic andapplied science and in developing radically new products and why othersocieties excel in more incremental innovationsReecting on these issues one cannot help being impressed with the

contrasts between the innovative styles of German and American orga-nizations For example Germany has continued to be very successful in making incremental innovations in many industries in which it wasvery competitive well before World War II paper printing materialsmachinery electrotechnical products motor vehicles chemicals textileyarns and fabrics But the Germans have been much less successful indeveloping totally new products or related innovations in biotechnologyelectronics telecommunications aircraft as well as other newer indus-trial areas In contrast the Americans have been very innovative in these and other industries with products having very short lives andwith technologies which change very rapidly and are very complex(Hollingsworth 1991 Soskice 1997 1999)The importance of institutions and innovativeness for international

competitiveness has led to the recognition that there are national systemsof innovation (Lundvall 1992 Nelson 1993 Edquist 1997 Hage andHollingsworth 2000) But even before the recent studies of nationalsystems of innovation Landes (1969) was a proponent of these viewswhen he made the argument that Germany had developed a nationalsystem of innovation by building on its education and scientic researchsystems which were lacking in Britain and elsewhere However the concept of a national system of innovation immediately poses theproblem of what kind of innovations should be consideredDavid Soskice an economist at the Wissenschaftszentrum Berlin has

addressed a set of problems somewhat complementary to the perspec-tive raised here His work (Soskice 1990 1997 1999) suggests that weare sensitive to the institutional environment in which the followingactors are embedded the employees of rms the owners and nanciersof rms the competitors of rms and nally their collaborators ndash aperspective complementary to Hollingsworthrsquos work on social systemsof production (Hollingsworth 1997 1998) When one reects on the insti-tutional contextsocial system of production within which these actorsare embedded one becomes aware that those countries which have some-what inexible labor markets external to rms but a business system in which rms have long-term commitments to their employees arestrongly associated with incremental but hardly at all with radical inno-vations In the rms of such countries (eg Germany and Japan) thereis much more consensus decision making than in societies where a lsquohireand rersquo set of practices is quite pervasive And this kind of consensusdecision making ndash while highly conducive to incremental innovativeness

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

627

ndash limits the capacity for radical innovations to occur This kind of lsquolongtermismrsquo set of norms and rules occurs not only in the employmentpractices of incrementally innovative rms but also in their nancingand ownership patterns (Hollingsworth 1997) During the past thirtyyears these were characteristics of many rms in Germany and JapanSignicantly in both of these countries venture capital markets werenot highly developed nancing tended to be based more on long-termbank loans than on the equity markets and the relationships of rmstended to be relatively stable And these stable institutional arrange-ments also contributed to a high degree of innovativeness ndash butincremental rather than radical innovationsMeantime we should also pursue Soskicersquos ideas (1999) about how

rmsrsquo relationships with their competitors and collaborators shape theirstyles of innovativeness In those advanced industrial societies in whichbusiness rms are highly mobilized into business associations theretends to be a high degree of standardized quality controls over technicalnorms and a high degree of reliance on legal sanctions as monitoringinstruments vis-agrave-vis association members As a result of this high degreeof integration of rms into associations member rms are more likelyto be associated with incremental innovations than is likely to be the case with rms in those societies where associations are less welldeveloped In the latter type of society radical innovations are morelikely to occur In those advanced industrial societies where rms havelong-term stable relationships with collaborators in product design andproduction the style of innovation is likely to be incremental And it is in those societies in which there is a weaker tradition of strongcollaboration among rms that radical innovations are more likely tooccurObviously even in a country where there may be a great deal of

radical innovativeness most innovations will be of an incremental typefor there are simply fewer radical innovations But Germany and therest of Europe not only have had relatively few radical innovations incomparison with the United States but they also have lagged behindthe USA in developing relatively totally new industries Thus there hasbeen much less innovative activity in Europe in high-technology andnewly emerging industries than in the USAIn complementary research Audretsch and Feldman (1995) have

demonstrated that the USA has been more successful in spawning rmsengaged in the early stages of the industry life cycle but relative toGermany had an environment unfavorable to innovative activity byestablished rms especially in more mature industries German rmshave tended to be associated with a routinized technological regime ndashwhere established enterprises have had strong incentives to engage inincremental innovations but new rms have tended to have fundamental

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

628

innovative disadvantages relative to the Americans (see Audretsch 1995Winter 1982 Casper 2000)On both sides of the Atlantic universities engage in basic and applied

research But signicantly the institutional environment in which univer-sities are embedded inuences the degree to which the research is likelyto represent major breakthroughs or is likely to have more modestimplications Recent research in biomedical science demonstrates that inthose societies in which university professors andor graduate studentshave the following characteristics there are likely to be relatively majorresearch breakthroughs graduate students have a long training periodhigh dependency on a major professor low autonomy in dening thenature of a research project restricted job options and a university envi-ronment where there is a very modest or nonexistent entrepreneurialculture For example in Germany ndash a country with relatively few radicalbreakthroughs ndash the scholar who nishes a habilitation will have spentmany years on a research project before qualifying for a permanent acad-emic position will have been in a high-dependency position with a seniorprofessor will have had limited autonomy to dene a research projectand will be constrained from pursuing a high-risk research project Theyoung German will have an extraordinarily high level of training butwill have faced a number of disincentives to undertake high-risk researchAnd as soon as heshe has completed the habilitation and becomes aprofessor administrative responsibilities constitute another disincentivefrom engaging in high-risk or long-term research projects Moreover theGerman professor is a civil servant with a more stable and secure incomethan hisher American counterpart In a very different institutionalenvironment American professors in large research universities tend tobecome much more entrepreneurial Today many American professorsbehave as though they were operating their own rms within univer-sities and thus devote considerable time to generating funds for researchprojects assistants secretaries supplies etc And the variation in thisentrepreneurial spirit across countries is associated with variation in thetaking of high-risk research strategies and the making of major researchbreakthroughsIn sum societies vary in terms of the institutional environment in

which their organizations are embedded and this variation inuencesthe degree to which they have industrial activities in the early or latestages of the industry life cycle successful new or mature industriesand radical vs incremental innovations And it is this variation whichis encouraging rms in countries with relatively few new industries and with few or no radical innovations to invest in those countries with many new industries and numerous radical innovations ndash as astrategy to be able to transfer new technologies from the latter institu-tional environment to the former

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

629

Table 4 summarizes the way that the norms rules and values of a society are associated with Americarsquos and Germanyrsquos institutionalarrangements their social system of production and their styles of inno-vativeness

CONCLUSION

Several themes deserve to be reemphasized Most importantly institu-tional analysis is emerging as a dominant issue in several social sciencedisciplines The study of various institutional components discussedabove is a complex subject requiring multiple levels of analysis and aknowledge of multiple sectors within societies It requires a theoreticalperspective that evolves from an empirical analysis in different societiesIf we are to advance the study of institutional analysis we need cross-disciplinary collaboration and team research which are not readilyachieved given the structure of most of our research universitiesAt present scholars in each academic discipline have their own distinc-

tive strategies for studying institutional analysis Thus economists tendto use several approaches (North 1990 Williamson 1985 Hodgson 19881998 1999) anthropologists another (Geertz 1995) and sociologistspolitical scientists and historians employ other strategies (Campbell etal 1991 Powell and DiMaggio 1991 Katznelson 1998 Hall and Taylor1996) With practitioners in various disciplines pursuing their distinctiveapproaches to institutional analysis there is low potential for collabo-rative institutional analysis across the social sciences The lessons fromthe biological sciences over the last 40 years are very instructive oncepractitioners from various biological disciplines began to cooperate instudying biological phenomena by using the same concepts at the mole-cular level theoretical advance was very rapid Comparable phenomenatook place in the development of other hybrid elds eg biochemistrybiophysics (Judson 1979 Olby 1979 Kohler 1982 1994) Similarly weresocial scientists in various disciplines to work collaboratively in the areaof institutional analysis there would very likely be an acceleration intheir theoretical knowledgeThis is not a plea that everyone engaged in institutional analysis should

do the same kind of research Indeed just as biologists working at themolecular level study many different kinds of problems similarly socialscientists engaged in institutional analysis would also work on manyseparate problems but in the spirit of a collective enterprise Some wouldconcentrate their attention on the study of rules norms habits conven-tions and values while others would study how these are associatedwith congurations of various institutional arrangements (types ofmarkets hierarchies networks associations communities clans statesetc) Others would work on specic institutional sectors (eg education

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

630

Table 4

Rules norms hab

its an

d values associated w

ith sp

ecic co

ngu

ration

s of coo

rdinating mecha

nism

s soc

ial sy

stem

s of

prod

uction

an

d nationa

l styles of inno

vation

USA

Germany

Norms rules values which provide incentives for

Norms rules values which provide incentives for

short-term horizons

long-term horizons

Con

gu

rative

forms of

Marke

ts c

orpo

rate hierarchies reg

ulatory state

Assoc

iation

s (busine

ss assoc

iation

s la

bor union

s)

coordinating

corp

oratist-type

networks

marke

ts c

orporatist-

mecha

nism

s (listed in

type

state

order of im

portanc

e in

the particu

lar social

system

)

Social sys

tem of

prod

uction

Bus

iness system

Firm

struc

tures

Con

glom

erates b

ut with mov

emen

t toward

Firm

more co

ngru

ent with pa

rticular prod

ucts

elim

ination of produc

tion

in som

e sectors

emph

asis on long

-term strateg

ies rm

s are well

emph

asis on sh

ort-term

strateg

ies high

integrated

differentiation

between divisions

of rms

Firm

s mob

ilized in

Low

High

and in

tegrated

into busine

ss

associations

Indu

strial relations

Low

deg

ree of jo

b secu

rity rigid in

ternal lab

or

High deg

ree of jo

b secu

rity

exible interna

l lab

or

system

marke

t ex

ible externa

l labo

r marke

tsmarke

t rigid externa

l lab

or m

arke

tPatterns of

Dispersed owne

rship frequ

ent turnov

erMore stab

le owne

rship patterns

owne

rship

Table 4

continue

d

USA

Germany

Norms rules values which provide incentives for

Norms rules values which provide incentives for

short-term horizons

long-term horizons

Social system of

Prod

uction

Bus

iness system

Training for labo

r an

dHigh deg

ree of voluntary decision mak

ing by

Collective an

d m

ore co

mpulso

ry d

ecisions

abo

ut

man

agem

ent

individuals as to ho

w m

uch

tim

e an

d ene

rgy

how m

uch ed

ucation

is ne

eded

he

avy em

pha

sis

to inv

est in educ

ation heavy

emph

asis on

on techn

ical training (eg en

gine

ering)

busine

ss sch

ools training

in marke

ting

and

sales accou

nting

Fina

ncial marke

tsEqu

ity an

d ven

ture cap

ital m

arke

ts highly

Equ

ity an

d ven

ture cap

ital m

arke

ts w

eakly

dev

elop

ed na

ncing from

cen

tral gov

ernm

ent

deve

lope

d he

avy em

pha

sis on

corporate loan

sto priva

te rm

s for dev

elop

men

t of new

tech

nologies related

to na

tion

al security

Unive

rsity research

High individual auton

omy high poten

tial for

Low

ind

ividua

l au

tono

my until almost 40

yea

rs of

system

individual creativity u

nive

rsity pr

ofesso

rs

age poten

tial for ind

ividual creativity high

ly

high

ly entreprene

urial

cons

traine

d unive

rsity pr

ofessors not so

entrep

rene

urial

Inno

vative

styles

Highly su

ccessful in m

ajor break

through

s in

High pe

rforman

ce in proc

ess an

d inc

remen

tal

basic an

d app

lied scien

ce go

od in dev

elop

ing

inno

vation

s in existing produ

cts not so su

ccessful

totally

new

lines of pr

oducts an

d in m

ajor break

through

s in basic scien

ce or in

tech

nologies high

deg

ree of entrepr

eneu

rship

deve

loping

totally new

lines of prod

ucts of

amon

g research

ers

tech

nologies low d

egree of entrepr

eneu

rship

amon

g research

ers

Table 4

continue

d

USA

Germany

Norms rules values which provide incentives for

Norms rules values which provide incentives for

short-term horizons

long-term horizons

Prod

ucts in w

hich

Pr

oducts with sh

ort liv

es prod

ucts in w

hich

Older produ

cts with long

history machine

ry pap

er

society excels

tech

nology

is bo

th highly co

mplex

and

pr

inting

textile

s m

otor veh

icles m

achine

too

lsch

ange

s rapidly en

tertainm

ent indus

try

pha

rmaceu

ticals biotechn

olog

y softw

are

aerosp

ace

business systems nancial markets and systems of research) Distinctivewould be the recognition of how these separate studies are linkedtogether how each institutional sector is linked to a societyrsquos normsrules values etc and its conguration of institutional arrangementsSimilarly other institutional analysts might focus on the study of orga-nizations with a major concern for how the institutional environmentof organizations (eg norms rules values congurations of institutionalarrangements institutional sectors) inuenced the structure culture andoutcomes of organizations And nally others might study how all ofthe aforementioned aspects of institutional analysis inuenced a societyrsquosoverall performance Nelson and Sampat (1998) have appropriatelyreminded us that we should engage in institutional analysis in order toaddress specic problems but the view developed in this article is thatwe rst need some mapping of the terrain of institutional analysis inorder for our diverse enterprises to become a meaningful collective enter-prise especially if we are to understand how institutions inuence theinnovative processThe argument developed here is not that all social scientists should engage

in institutional analysis Obviously there are numerous other importantareas of research which lie beyond the eld of institutional analysis Nordoes this article argue that institutional analysis is the most importantsubject for social scientistsSince the structure and culture of our research universities reect an

enormous amount of disciplinary fragmentation what are the prospectsfor promoting the kind of collective and interdisciplinary research agendaproposed above In one respect the prospects are not encouragingDuring the past half century the efforts of American universities to have effective communication across the biological sciences have variedgreatly in their success At the University of California-Berkeley it was not until the 1980s that there was a serious effort to integrate thebiological sciences and at that time the campus abolished seventeendepartments to promote an integrated biological program But at theUniversity of Wisconsin the biological sciences are still very fragmentedinto dozens of different departments in botany plant pathology zoologyand physiology with two departments of biochemistry and two depart-ments of geneticsThere have been other approaches which have worked very well histor-

ically for promoting scientic integration amidst scientic diversity Forexample several summer schools and special institutes were organizedin the United States during the 1940s in an effort to overcome the frag-mentation in the biological sciences These programs made it possiblefor scientists in diverse academic disciplines and from different univer-sities and countries to come together to engage in intense and frequentinteraction Perhaps the most effective program was that held in the

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

634

summers at Cold Spring Harbor New York where a research agendafor the study of molecular biology was developed (Fischer and Lipson1988 Stent and Watson 1966) Scientists went there summer aftersummer and when they returned to their universities many began therestructuring of the biological sciences in their universitiesIn our own day we need summer institutes and other similar programs

so that scholars from diverse backgrounds can come together anddevelop a collective agenda for doing institutional analysis and forstudying the impact of institutions on the performance of societies ndashinnovativeness being one such subject If we can develop such instituteswe will increase the prospects for transforming our universities Thereis reason to be optimistic that once an eclectic group of scholars workwith a common set of concepts we can have genuine theoretical advanceif researchers can bring their diverse expertise to bear on a common setof problems There is considerable evidence that institutional analysisoffers high potential for advancing our knowledge of innovations andtechnological change (Hollingsworth and Hollingsworth 2000) If we aregoing to make substantial theoretical advance it is not likely to occurfrom the vantage point of a multiplicity of disciplines Rather it is likelyto occur as we develop a new intellectual framework and our own acad-emic journals This is a tall agenda which requires us to reassess theway that knowledge is presently produced in our societies

NOTE

This article is part of a much larger agenda involving the study of institutionsorganizations and innovations in which I am involved My debts are many andvaried The rst version was written while I was in residence in the ResearchUnit for Institutional Change and European Integration of the Austrian Academyof Sciences in Vienna in the summer of 1998 I am especially grateful to ProfessorsEgon Matzner and Sonja Puntscher-Riekmann for providing the stimulating envi-ronment in which to think through the ideas developed here Egon Matzner hasengaged in very stimulating discussions with me about the importance of insti-tutions in the general eld of socio-economics The nal version of the articlewas completed during my residence as a fellow at the Neurosciences Institutein La Jolla California I also wish to thank my colleagues in the InnovationTheme Project at the Netherlands Institute for Advanced Study (NIAS) during1998ndash9 and Edgar Grande and Karl Muumlller for stimulating conversations aboutthe interaction of institutions and organizations in facilitating particular typesof innovation I am especially grateful to Jerald Hage for helping me to developa number of issues in the article and for his useful comments on the rst draftFrans van Waarden early on helped me to work through the idea of multiplelevels of institutional analysis Others who have helped to develop the ideas inthe article have been my colleagues in the research project lsquoComparative socialsystems of productionrsquo especially Tom Burns Christel Lane Yoshitaka OkadaWolfgang Streeck and Richard Whitley My colleagues Robert Boyer SteveCasper and David Soskice have taught me much about how incentives providedby the institutional environments of rms inuence individual decision making

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

635

But my most important debt is to Ellen Jane Hollingsworth who has repeatedlyquestioned every single assumption of this article Without her rich and contin-uous tutoring this article would be even less developed than it is at presentFinally without the insightful assistance of David Gear in many phases of thisarticle it would not have been completed

REFERENCES

Aoki Masahiko (1990) lsquoTowards an economic model of the Japanese rmrsquo Journalof Economic Literature March 1ndash27

Arora A Landau Ralph and Rosenberg Nathan (eds) (1998) Chemicals and Long-Term Economic Growth Insights from the Chemical Industry New York JohnWiley amp Sons

Arrow Kenneth (1974) The Limits of Organization New York NortonArthur W B (1988a) lsquoSelf-reinforcing mechanisms in economicsrsquo in P W

Anderson K J Arrow and D Pines (eds) The Economy as an Evolving ComplexSystem Redwood City Calif Addison-Wesley

mdashmdash(1988b) lsquoCompeting technologies an overviewrsquo in G Dosi C Freeman RNelson G Silverberg and L Soete (eds) Technical Change and Economic TheoryLondon Pinter Publishers pp 590ndash607

Audretsch David B (1995) Innovation and Industry Evolution Cambridge MassMIT Press

Audretsch David B and Feldman Maryann P (1995) lsquoInnovative clusters andthe industry life cyclersquo Discussion Paper Wissenschaftszentrum BerlinMarket Processes and Corporate Development Research Unit FS-IV-95-7

Baum Joel A C and Oliver Christine (1992) lsquoInstitutional embeddedness andthe dynamics of organizational populationsrsquo American Sociological Review 57540ndash59

Baum Joel A C and Singh Jitendra V (1994a) lsquoOrganizationndashenvironment coevo-lutionrsquo in Joel A C Baum and Jitendra V Singh (eds) The EvolutionaryDynamics of Organizations New York Oxford University Press pp 379ndash402

mdashmdash(eds) (1994b) The Evolutionary Dynamics of Organizations New York OxfordUniversity Press

Berger Suzanne and Dore Ronald (eds) (1996) National Diversity and GlobalCapitalism Ithaca NY Cornell University Press

Boyer Robert (1997) lsquoThe variety and unequal performance of really existingmarkets farewell to Doctor Panglossrsquo in J Rogers Hollingsworth and RobertBoyer (eds) Contemporary Capitalism the Embeddedness of InstitutionsCambridge and New York Cambridge University Press pp 55ndash93

Burns Tom R and Dietz Thomas (2001 forthcoming) lsquoSocial rule system theorysocial action institutional arrangements and evolutionary processesrsquo in JRogers Hollingsworth and Karl Muumlller (eds) Advances in SocioeconomicsLanham Md Rowman amp Littleeld

Burns Tom R and Flam Helena (1987) The Shaping of Social Organization BeverlyHills Calif Sage

Calvert Randall L (1995) lsquoThe rational choice theory of social institutionsrsquo inJeffrey S Banks and Eric A Hanuskek (eds) Modern Political Economy NewYork Cambridge University Press pp 216ndash66

Camic Charles (1986) lsquoThe matter of habitrsquo American Journal of Sociology 911039ndash87

Campbell John L (1997) lsquoMechanisms of evolutionary change in economicsgovernance interaction interpretation and bricolagersquo in Lars Magnusson

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

636

are depicted in Table 1 In most forms of social analysis it is extremelyimportant that we understand the social and cognitive conditions thatlead to compliance or non-compliance with rules and the conditionswhich lead to changes in rulesThe approach to the study of institutions employed here argues that

norms rules habits conventions and values both reect and shape thepreferences of actors Norms rules habits conventions and values inu-ence who and what are included in different types of decision makinghow information is processed and structured what action is taken(Shepsle 1986 1989) It is through norms and rules that behavior isjudged to be democratic fair or egalitarianBurns and Flam (1987) point out that in any society there are multiple

rule systems Within a family there are rules for decision making oftenquite different from rules and norms for decision making for a professorin a classroom or for the customer in a bank Despite the heterogeneityof rule systems there are meta rules and norms which encompass lesserrule systems Otherwise there would be such contradictory rule systemsthat a society would be paralyzed The existence of meta rule systemspermits different rule systems to intersect with each other so that ambi-guities can be resolved The greater the pluralism and complexity of asociety the more ambiguity there is about meta rules and norms in asociety and of course all ambiguities never completely disappearOverall the degree to which separate rule systems are interlinked is anempirical problem There are different sectors groups and interestspursuing their own action logic but through higher-order meta princi-ples and rules there can be order consensus and coherence in a societyIt is through a set of meta rules that class and ethnic conict in societiesare contained (For elaboration see Burns and Flam 1987)In many respects our understanding about norms rules habits con-

ventions and values inuences our perspective on how societies are con-structed and how they change lsquoNew institutionalistsrsquo (Posner 1992Schotter 1981 Williamson 1975 1985) tend to assume that at one timethere was a state of nature and that there was a movement from individ-uals to institutions ndash an approach often called methodological individual-ism (Popper 1961 Hodgson 1998 1999) And of course there areinnumerable instances which methodological individualists cite to demon-strate that individuals create new rules of behavior For example it is pos-sible for actors to change the rules of driving so that instead of drivingon the left side of the road drivers adopt a new rule and drive on the rightThis article however tends to equate social habits and institutions

As Hodgson and others remind us (Hodgson 1988 1989 1997 1999Grafstein 1992 Camic 1986 Johnson 1992 Nelson and Winter 1982Veblen 1899) social habits are the results of earlier choices and are ameans of avoiding endless deliberation Because cognitive frameworks

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

602

are learned through habit individuals rely on the acquisition of suchcognitive habits before reason communication choice or action arepossibleWhereas Schotter (1981) and other game theorists take the individual

as an agent unencumbered by previous habits Field (1984) and othershave stressed that there can be no games without prior norms and rulesand thus a set of norms and rules must be presumed at the start Thosewho attempt to explain institutions from individual behavior alone areusing a bad strategy (Hodgson 1998)The position here ndash heavily inuenced by Hodgson (1998 1997 1988)

ndash is that individuals are embedded in a complex institutional environ-ment and that institutions not only constrain but also shape individuals(also see Hollingsworth and Boyer 1997) It would be a mistake howeverto get involved in an innite regress in order to determine which camerst ndash individuals or institutions Of course institutions are formed andchanged by individuals just as individuals are shaped and constrainedby institutions But at a macro level it is institutions that provide acognitive framework whereby individuals can cope with their reality Inthis sense the micro and macro worlds are intertwined At the macrolevel there is considerable stability but at the micro level individualshave a signicant level of autonomy and there can be widespread diver-sity As Hodgson (1988) reminds us most institutions in a temporal senseexist prior to the individuals in any given societyIt would be a serious mistake to downplay the importance of indi-

viduals and micro level analysis as we study institutions In the nalanalysis it is at the level of individuals that norms rules habits conven-tions and values exist An individual is born into and socialized intogroups and a society and this is how one early in life acquires a senseof appropriate forms of behavior Because of the way that individualsare socialized into a world of rules norms habits conventions andvalues it is unnecessary for individuals to restructure the world anewevery day (Douglas 1987 Elster 1989) Every action does not have tobe seriously reected upon For this reason institutions provide cogni-tive frameworks for individuals make their environments predictableprovide the information for coping with complex problems and envi-ronments In the words of Johnson (1992 26) lsquoInstitutions reduceuncertainty coordinate the use of knowledge mediate conicts andprovide incentive systems By serving these functions institutions providethe stability necessary for the reproduction of societyrsquo However eachsociety has different forms of habits rules and norms and hence differentincentives and disincentive systems for learning and forgetting forprocessing information But because individuals have varying degreesof autonomy individuals and groups can deviate from the prescribedforms of behavior in a society And of course these changes at the level

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

603

of individuals become important in understanding processes of societalchangeThese views are not meant to imply that the type of institutional

analysis proposed herein approximates a general theory of society Thisis clearly not the case However it is intended to represent the rst stepsin a mapping exercise of the boundaries of institutional analysis and tosuggest a few methodological insights for studying institutionsOne should try to see norms and rules as continuous and not as

dichotomous entities to recognize that they come in varying strengthsLegro (1997) has suggested that we assess the robustness of norms andrules with three criteria their simplicity their durability and their concor-dance Simplicity refers to how well actors understand norms and ruleshow well they can be applied within a specic situation Some normsand rules can be so complex that actors can have considerable difcultyin applying them in specic cases Durability addresses the issue of howlong norms and rules have been in effect ndash in short in order to assesstheir level of legitimacy While the position of this article is that normsrules and values are quite durable they do vary in this respectConcordance refers to how widely applied a norm or rule is Thisaddresses the degree to which a rule is a meta rule the degree to whichit incorporates the heterogeneity of other norms and rules In sum theclearer the norms and rules of a society the longer they have been inexistence and the more widely applicable they are the greater theirimpact on a society Hence the more robust the norms and rules thegreater their impact on a society and the less their robustness the greatertheir exibility and the less their effect on shaping a societyrsquos outcomesand performancesBecause norms rules and values are quite durable they play an impor-

tant role in shaping the history of societies thus contributing to a greatdeal of path-dependency Actors attempt to adjust to their contempo-rary environment but since they are products of the past the historicallegacy of norms rules and values inuences the decisions they makeAlthough actors have some capacity to alter the course of their historythey are constrained by their past and the degree to which they canmove beyond their past is limited As Lanzara (1998) Johnson (1992)and others have argued societal inertia is a basic feature of institutionsThey provide the basic stability necessary for change Degrees of historyare continuously reproduced by the way in which the inhabitants ofsocieties are socialized History matters but at critical points in historythere is punctuated equilibrium (Somit and Peterson 1992) During mostperiods of history there is considerable stability in the norms rules andvalues of a society but at critical moments norms rules and values canquickly and dramatically be redened At all times norms rules habitsconventions and values are inuencing each of the other components in

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

604

the institutional framework discussed below but these other institutionalcomponents feed back and can modify rules norms conventions etc(Murmann 1998)

Second level of analysis

The norms rules habits conventions and values of a society lead to thenext level of analysis ndash the institutional arrangements which are involvedin the coordination of various economic actors producers and suppliersof raw materials knowledge etc processors of raw materials infor-mation workers customers of raw materials nished productsinformation etc nanciers governmental and other types of regulatorsThese actors regularly engage in contests to resolve various economicproblems in virtually all sectors of society how are prices to be setWhat quantity of various products is to be produced How are stan-dards of various products processes etc to be set How is the qualityof products and processes to be determined How are various societalprocesses to be nanced In order to confront these problems and toaddress the conicting positions of economic actors as they address these problems societies develop various institutional (governance)arrangements for coordinating different actors These consist of marketsvarious types of hierarchies and networks associations the state commu-nities and clans (see Hollingsworth and Boyer 1997 Campbell et al1991 ch 1)Each of these particular kinds of coordinating mechanism has many

different types For example there are many types of states (eg theregulatory state developmental state authoritarian state welfare state)on which there is an extensive literature (Kim 1997) Similarly there aredifferent types of markets networks different kinds of associations etc(Boyer 1997 Hage and Alter 1997 Schneiberg and Hollingsworth 1990)When we do institutional analysis we must engage in congurativeanalysis recognizing that actors are not coordinated or governed by asingle type of institutional arrangement Some of the literature discussesindustrial sectors as though they were coordinated or governed by asingle institutional arrangement whereas in fact each sector of aneconomy is coordinated by a conguration of institutional arrangementsSome congurations coordinate actors in certain problem areas whileother congurations of institutional arrangements coordinate actors inaddressing other problems The types of congurations which are domi-nant in a society are somewhat stable and tend to persist over timewithin a society (Hodgson 1999)When one mode of coordination is dominant in a society it will

inuence the role which other coordinating modes will play Hence thestrong role of the state in the Soviet Union inuenced the role of markets

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

605

associations etc in the governance of the Soviet economy Similarly theprominence of particular modes of coordination in a society inuencesits style of innovationMuch of the literature on institutional arrangements (eg forms of eco-

nomic coordination) remains fragmented and unintegrated It is helpfulto array these various forms of coordination in a two-dimensional tax-onomy as in Figure 2 On the vertical dimension the economistrsquos viewof a self-interested agent is contrasted with a more sociological perspec-tive according to which social rules obligation and compliance shapehuman actions On the horizontal dimension there is another continuum(ie the distribution of power) At one extreme of the dimension onends many and relatively equal agents interacting (eg as in a well-organized spot market) At the other extreme inequality in power resultsin a hierarchical form of coordination which structures the interactionbetween principals and agents or between leaders and followers Wherea society falls along the distribution of power is inuenced by the rulesnorms values etc which are dominant in a society at a particular momentin timeInstitutional arrangements can be visualized on two dimensions the

nature of action motive on the one hand and the distribution of poweron the other Markets (cell 1) combine self-interest with horizontalcoordination transactors and they reect sensitivity to concerns aboutsupply and demand thus providing ex post an unintended equilibriumParadoxically the more pure and perfect the market competition thegreater the need for codied rules of the games for coordinating economictransactions Thus collective associations (cell 6) andor various formsof state intervention (cell 4) are required to develop and enforce rulesfor transacting partners (Schneiberg and Hollingsworth 1990 Streeckand Schmitter 1985a) This is an example of how the norms and rulesand the institutional arrangements of a society are intertwined asreected in a conguration of institutional or governance arrangementsAlong the horizontal axis actors can be joined in an organization or

a rm a hierarchy is the generic terms for this institutional arrangement(cell 2) Along the horizontal line one recognizes the difference betweentransactions in a market and transactions within a rm The well-knownworks of Coase (1960 1981) and Williamson (1975 1985) utilize theconcept of transaction costs in explaining the emergence of corporatehierarchiesThere are also various types of networks (cell 5) which exhibit mixes

of self-interest and social obligation with some actors being formallyindependent and equal Yet in some networks (the large rms and theirsubcontractors) there is unequal power and initiative Networks maycomprise all kinds of actors some consist only of rms but others includeassociations and the state (Hage and Alter 1997)

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

606

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

607

DISTRIBUTION OF POWER

VERTICALHORIZONTAL

1 MARKETS 2 HIERARCHY

6ASSOCIATIONS

5NETWORKS

4STATE

3COMMUNITIES

SELF INTEREST

SOCIAL OBLIGATION

ACTION MOTIVE

Figure 2 A general taxonomy of institutional arrangements

The vertical axis deals with action motives Toward the upper part ofFigure 2 actors are engaged in individualistically oriented behaviorwhereas toward the lower part actors are more engaged in collectivebehavior and strive to cope with problems of common interest Cell 3 ndashcommunities and clans ndash consists of institutional arrangements based ontrust reciprocity or obligation and thus not derived from the pure selshcomputation of pleasures and pains This is an unconventional form ofcoordination for most neoclassical economists (however see Arrow1974) but not for many anthropologists political scientists and sociolo-gists (Streeck and Schmitter 1985a Polanyi 1944 Gambetta 1988Fukuyama 1995 Sabel 1992 Putnam 1993)In the neoclassical paradigm theorists argue that actors engage in

forms of exchanges that best promote their individual interests If somestructural conditions are fullled (absence of increasing returns to scalethe reversibility of transactions absence of uncertainty and completecontingent markets with no collusion between actors) then the invis-ible hand theorem applies and market-type activity functions quite welland also provides the optimum for society thereby combining efciencyharmony and order However an excess of market activity may welllead to ruinous competition and excessive conict There is variation inthe extent to which ruinous competition occurs depending on the socialcontext in which transactions take place Thus it is important that webe sensitive to the institutional context in which transactions areembedded and that we understand the degree to which social bondsexist at both the micro and macro levels of analysis Micro bonds facil-itate exchanges in a society but at the societal level social bonds existat the level of the collective ndash in the community or region and amongmembers of racial religious and ethnic groups All other things beingequal the more powerful the social bonds among transacting partnersthe more economic competition is likely to be restrained Thus mosttransactions occur not simply in an impersonal calculative system ofautonomous actors unrestrained by social ties ndash as implied by theneoclassical paradigm ndash but in the context of social ties variation in the strength of which leads to variation in levels of trust and transac-tion costs (Etzioni 1988 211 Granovetter 1985 Streeck and Schmitter1985a Hollingsworth and Boyer 1997 Hodgson 1988 1999 Schneibergand Hollingsworth 1990)Another form of multilateral institutional arrangement is various

types of associations (cell 6) Unlike networks clans and communitiesassociations are more formal organizations Whereas markets corporatehierarchies and networks tend to coordinate economic activity amongdifferent types of actors (eg producers with suppliers capital withlabor) associations typically coordinate actors engaged in the same orsimilar kinds of activities Business associations and labor unions are

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

608

some of the most common forms of associations for coordinatingeconomic activity in capitalist economies (Schneiberg and Hollingsworth1990 Streeck and Schmitter 1985a)Finally there is the state (cell 4) which is an institutional arrangement

quite unlike any of the others It is the state that sanctions and regu-lates the various non-state coordinating mechanisms that is the ultimateenforcer of rules of the various mechanisms that denes and enforcesproperty rights and that manipulates scal and monetary policy Overallit is the state that inuences the total incentive system of a society(Lindberg and Campbell 1991 Johnson 1992 40) At the same time thestate may also be an economic actor by engaging directly in productionand exchange relationsThe choices of institutional arrangements in Figure 2 are constrained

by the social context in which they are embedded Depending on thenature of that embeddedness there is variation in the collective formsof governance some of which are specied in the lower part of thegure Some modes of coordination in the bottom and upper levels ofthe typology are often mixed together though one coordinating modeis likely to be more prominent than another But actors are often engagedin complex congurations involving several kinds of institutionalarrangementsEach of these various institutional arrangements has its own logic ndash

its own rules its own procedures for enforcing compliance its ownnorms and ideologies which help to reduce the costs of enforcementThese are summarized in Table 2 which provides further elaborationabout the various coordinating mechanisms one nds in almost everycapitalist society Table 2 lists the organizational arrangements and theirstructures rules of exchange and means of enforcing compliance asso-ciated with each type of coordinating mechanism While each type ofinstitutional arrangement has various positive features each institutionalarrangement also has failures and these are featured in Table 3 It is thecontest between those who support and those who oppose these variousinstitutional arrangements that tends to lead to transformations in insti-tutional arrangements over time (Campbell et al 1991 Campbell 1997)Again it is difcult to conceive of pure institutional arrangements ndash

either exclusively markets or exclusively hierarchies ndash since it has beenwell known since Adam Smithrsquos Wealth of Nations that the division oflabor within the rm cannot be disentangled from the existence andextent of the market Institutional arrangements have their own distinc-tive form of efciency (good static efciency for the market dynamicefciency for rms) and inefciency often leading to considerableinequality Neither networks nor communities are panaceas for economiccoordination without being congured with other types of institutionalarrangements Networks and communities may solve certain issues but

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

609

Table 2

Log

ics of ins

titution

al arran

gemen

ts

Coordination

Organizational

Rules of

Individual means

Collective means

mechanisms

structure

exchange

of compliance

of compliance

Marke

tsEasy en

try an

d exit

Voluntary sp

ot excha

nge

Leg

al enforcemen

t of

Norm of pr

ivate

Bila

teral ex

chan

ge or

control

prope

rty

marke

t site (W

all Street)

Reg

ulations

to

Leg

itim

acy of m

arke

t en

force contracts

men

tality

Com

mun

ities

Inform

al m

embe

rship

Voluntary exch

ange

based

So

cial norms an

d m

oral

Highly institutiona

lized

ev

olve

s ov

er lon

g period

on social solid

arity an

d

princ

iples im

pose

norm

s an

d rules requ

ire

of tim

ehigh

deg

ree of tru

stob

ligations

mem

bers to accept

Kno

wledg

e of others an

d

lsquocorpo

ratersquoo

bligations

recipr

ocity ov

er tim

e

Networks

Semi-form

al m

embe

rship

Voluntary exch

ange

ove

r a

Con

tractual bon

dsPersona

l relation

sBila

teral or m

ultilateral

time pe

riod

Resou

rce de

pend

ence

Tru

st built outside

the

ex

chan

geecon

omic arena

Associatio

nsFo

rmal m

embe

rship

Restricted to mem

bers

Self-interest

Some deg

ree of

Multilateral excha

nge

Emph

asis on insider

Rep

utation

effects

compu

lsion

outsider or we

they

Priva

te interest type

of

men

tality

gove

rnan

ce

Private

Com

plex organ

izations

Restricted to mem

bers

Rew

ards to ind

ividuals

Highly institutiona

lized

hierarch

ies

which

ten

d to becom

e ex

chan

ge based

on

Asymmetric po

wer threat

rules

bureau

cratic hierarchies

asym

metric po

wer

of san

ctions

Mem

bers soc

ializ

ed into

bureau

cratic rules

corp

orate cu

lture us

e of

sanc

tions

State

Pub

lic hierarchy

Unilateral action

Exit vo

ice (vote lob

bying)

Coe

rcion

De jure

and impo

sed

Indirect political an

d

loya

ltyNorms an

d pu

blic rules

mem

bership

econ

omic excha

nge

Table 3

Failu

res of coo

rdinating mecha

nism

s

Coordinating mechanisms

Type of failure

Associations

Private hierarchies

State

Market

Communities

Networks

Enforcem

ent

Usu

ally relies on

Might enh

ance

Needs co

ntrol

Needs an

internal

Needs

tru

st and

Need an external

the state as an

oppo

rtun

istic

extern

al to state

enforcem

ent

loya

lty often

en

forcem

ent

enforcer

beha

vior

bureau

cracy

authority

coming from

au

thority

Resem

bles

The

ideal o

f (ju

dge

s

Facilitates

outside (fam

ily May

facilitate

enforcem

ent

internal m

arke

ts

parliamen

t collu

sion

and

relig

ion

cartelization an

d mecha

nism

might hur

t marke

t) to

impe

rfect

ethn

icity)

mon

opoly

of cartels

incu

mbe

nt

correct state

compe

titio

nCom

patible with

worke

rsab

uses

variou

s types of

Lob

bies can

co

mpe

tition

captur

e pu

blic

interest goa

ls

Public goo

d

Useful for

Gov

erna

nce co

sts

Can

provide

Can

not provide

Can

internalize

Useful for

and

establishing

might exceed

public goo

ds

colle

ctive go

ods

some co

llective

enha

ncing qu

ality

externality

stan

dards an

d

the be

nets of

but ha

s or deal w

ith

good

s (qua

lity

and training bu

t qu

ality for

internal division

difc

ultie

s in

externalities

training

) bu

t no

t no

t ve

ry goo

d in

setting ru

les of

of la

bor

prov

iding them

Inad

equate

othe

rs (welfare

prov

iding for

compe

titio

n in

Slow

to react to

in precise

mon

itoring of

gene

ral p

ublic

societal gen

eral

the indu

stry

chan

ges in the

am

ounts

tech

nical ch

ange

go

ods)

welfare

Useful for

environm

ent

Might fail in

and inno

vation

Mem

bers tightly

Weak in the

prov

iding

induc

ing

integrated

into

prov

ision of

man

y go

ods

tech

nical

community

colle

ctive go

als

colle

ctively

chan

geha

ve limite

d

that in

dividua

lcapa

city for

mem

bers can

not

inno

vatio

nsprov

ide for

them

selves

Table 3

continue

d

Coordinating mechanisms

Type of failure

Associations

Private hierarchies

State

Market

Communities

Networks

Efcien

cyFa

cilitates

De

cien

t in

Can

be high

ly

Some ba

sic social So

me go

ods

Slow

to en

hanc

e co

operation an

d

coop

eration an

d

bureau

cratic

relation

s cann

ot

cann

ot be

efc

ienc

y an

d

X-efcien

cy but

X-efcien

cyan

d can

not

be provided

by

delivered

at

speed of

not allocative

easily deliver

pure m

arke

t su

fcien

tly

adap

tive

ness

efcien

cy

good

s at

mecha

nism

slow costs

except in

low cost

indus

tries whe

re

tech

nology

is

complex an

d rapidly cha

nging

Equity

Narrow

Excessive

Might enh

ance

Facilitates

Might lead to

Whe

n widely

encompa

ssing

multiplication

ineq

uality

ineq

uality in

retard

ed

dev

elop

ed in

to

associationa

l of con

trollers

(pow

er

income an

d dev

elop

men

tindus

trial

stru

ctures lead

(fru

stration

and

an

d privilege

)wealth

districts

to in

come

ineq

uality)

netw

orks m

ay

equa

lity

facilitate greater

equality an

d

inco

me

distribution

whe

n wea

kly

dev

elop

ed

netw

orks ten

d to

increase soc

ial

ineq

uality

Note For allo

cativ

e efc

ienc

y an

d X

-efcien

cy see Leibe

nstein (19

66 19

76)

they raise other no less severe problems It is important to recognizethe imperfection of any single institutional arrangement in order tocomprehend the origin and transformation of any other institutionalarrangement (Hodgson 1999 ch 3)As suggested above each institutional arrangement is congured with

other institutional arrangements Usually one particular institutionalarrangement is more dominant in a particular conguration but becauseeach type of institutional arrangement has its own strengths andweaknesses there is no simple structural logic in the governance or coor-dination of a society Each institutional arrangement constrains the otherbut the inherent tension among the various institutional arrangementswithin a conguration contributes to changes in the governance orcoordination of a society The routines and logics of each of these cong-urations of institutional arrangements provide constraints and incentivesfor actors The way that actors perceive the incentives and constraintsof these governance congurations leads to a particular market logicand it is the specic market logic of a society which inuences its speciccapacities and weaknesses The inherent strain among the different logicsin a conguration of coordination helps to provide the exibility for asociety to adapt to new circumstancesIf there were a society with pure markets or a society coordinated

only by the state there would be too much rigidity and too little diver-sity to cope with the vast uncertainty in the global environment A societywith very little diversity in its coordination mechanisms would havelittle capacity to adapt to new circumstances Congurations with consid-erable diversity of institutional arrangements provide for a certainamount of incoherence in governance but they also provide for thecapacity to adapt to new circumstances While the Soviet Union wasdominated by the state as a coordinating mechanism there were alwaysfunctioning markets in the system Moreover the feudal society of thetwelfth and thirteenth centuries consisted not only of hierarchical rela-tionships between serfs and masters but also of urban guilds clericalhierarchies and markets (Johnson 1992 38) In sum the robustness ofinstitutions often depends on multiple and diverse principles and logicsof action on the inconsistency of principles and procedures on patternedforms of disorder (Lanzara 1998 Orren and Skowronek 1991 320 329)

Third level of analysis

The next level consists of the institutional sectors of a society The rulesnorms and values of a society inuence the array of institutional arrange-ments and levels one and two inuence the nature of and the relation-ships among various institutional sectors and all three levels areintricately linked together to form a social system of production Together

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

613

all of these components inuence the performance of economic sectorswithin a society as well as the performance of the total society A socialsystem of production is the way that a societyrsquos institutions (see rstlevel of analysis above) its institutional arrangements (see second levelof analysis above) and its institutional sectors are integrated into a socialcongurationAn institutional sector includes all organizations in a society which

supply a given service or product along with their associated focal orga-nizations (eg major suppliers funding sources regulators and so forth(Scott et al 1994 108 117 Campbell et al 1991 Hollingsworth et al1994) Institutional sectors include but are not limited to the societyrsquossystem of education system of research business system nancialmarkets legal system and the state The structure of the nancial marketssystem of training and education industrial relations system system ofresearch and state and legal systems are distinct and idiosyncratic ineach society In short they are system specicAll of these institutional sectors tend to cohere with each other

although they vary in the degree to which they are tightly coupled intoa full-edged system While each of the institutional sectors has someautonomy and may have some goals that are contradictory to the goalsof other institutional sectors with which it is linked an institutional logicin each society leads institutions to coalesce into a complex social con-guration (Hollingsworth 1991) This occurs because the institutionalsectors are embedded in a culture in which their logics are symbolicallygrounded organizationally structured technically and materially con-strained and politically defended The conguration of institutionalsectors usually exhibits some degree of adaptability to new challengesbut continues to evolve within an existing style But under new circum-stances or unprecedented disturbances these institutional congurationsare exposed to sharp historical limits as to what they may or may notdo (Schumpeter 1983 David 1988 Arthur 1988a 1988b Haringkanssen andLundgren 1997)Why do all of these different institutions coalesce into a complex

social conguration which is labeled here as a social system of pro-duction The literature suggests two contrasting interpretations Part ofthe answer ndash indeed a controversial one ndash is that these institutionalsectors are functionally determined by the requirements of the practiceof capitalism in each time and place (Habermas 1975) Another expla-nation emphasizes the genesis of the actual conguration via a trial and error process according to which the survival of rms regions orcountries is the outcome of complex evolutionary mechanisms (Maynard-Smith 1982 Nelson and Winter 1982) However the problem is evenmore complex Markets and other mechanisms for coordinating rela-tionships among economic actors place constraints on the means and

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

614

ends of economic activity to be achieved in any society The othercoordinating mechanisms include different kinds of hierarchies varioustypes of networks and associations ndash trade unions employers and busi-ness associations (Hollingsworth and Lindberg 1985 Campbell et al1991) The logic of these various institutional arrangements providesactors with vocabularies and logics for pursuing their goals for deningwhat is valued and for shaping the norms and rules by which theyabide (see Tables 2 and 3) In short in contrast to the logic of theneoclassical paradigm the argument here is that the dominant type ofinstitutional arrangements places severe constraints on the denition of needs preferences and choices of economic actors Whereas the neo-classical paradigm assumes that individuals and rms are sovereign thisarticle is based on the assumption that rms are inuenced by the hold that the institutional congurations making up a social system ofproduction have on individual decision making (Campbell et al 1991Etzioni 1988 Streeck and Schmitter 1985a Hollingsworth et al 1994Hollingsworth and Boyer 1997 Magnusson and Ottosson 1997 North1990 Hodgson 1999)Standard neoclassical economic theory has tended to downplay the

role of production and consequently of rms Even transaction cost theo-rists who are concerned with analyzing the rm as a coordinatingmechanism have been relatively unconcerned with the various compo-nents of a social system of production Indeed as long as there waswidespread optimism about the efcacy of Keynesian economics therewas relatively little concern among neoclassical economists with thesupply side of the economy Even in the opinion of most Keynesians agroup of experts should ideally be able to shape the size of aggregatedemand while the supply side of the economy would be left to the twominimalist institutions of neoclassical economics ndash markets and manage-rial hierarchies For several decades however it has become increasinglyobvious that some of the most competitive and successful patterns ofindustrial output and industrial production in capitalist economies donot derive from the neoclassical prescription of unregulated markets andcorporate hierarchies complemented by a neoliberal democratic stateIndeed it is now well understood that certain highly successful produc-tion patterns require for their emergence and survival institutionalarrangements and environments the very opposite of the prescriptionsfound in the neoclassical paradigm (see especially Streeck 1992 but alsoHollingsworth and Streeck 1994) Thus social scientists have come tounderstand that if they are to comprehend the behavior and performanceof contemporary economies concerns about social systems of produc-tion must be brought into the pictureBecause production involves more than technology a number of social

scientists have an increasing concern with social systems of production

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

615

The same equipment is frequently operated quite differently in the samesectors in different countries even when rms are competing in the same market (Maurice et al 1980 Sorge and Streeck 1988 Sorge 1989Hollingsworth et al 1994) Variations in production and process tech-nologies are inuenced partly by variations in the social environmentsin which they are embedded In other words rms are embedded intocomplex environments which among other things place constraints ontheir behavior Thus a social system of production is of major impor-tance in understanding the behavior and performance of an economyHow the state and other institutional arrangements (eg marketsnetworks private hierarchies associations) coalesce and are related toparticular social systems of production are important determinants ofeconomic performanceThe congurations of institutional arrangements that coordinate or

govern the behavior of actors in one society and its structure of specicinstitutional sectors cannot easily be transferred to another society forthey are embedded into a social system of production that is societallydistinct (Hollingsworth 1997) Societies borrow selected principles offoreign management styles and work practices but the effectiveness ofsuch borrowing is generally limited Economic behavior and performanceare shaped by the entire social system of production in which actors areembedded and not simply by specic principles of particular manage-ment styles and work practices Moreover a societyrsquos social system ofproduction tends to limit the kind of goods that it can produce and withwhich it can compete successfully in international marketsSome scholars (Kenney and Florida 1993 Oliver and Wilkinson 1988)

have assumed that the diffusion of particular forms of management stylesand work practices across societies could lead to system convergenceReferring to the Japanese production system Kenney and Florida haveargued that it consists of organizational practices whose fundamentallsquogenetic codersquo can successfully be inserted into another society and beginsuccessfully to reproduce its behavior in the new environment (1993 8)Their position is in the intellectual tradition of Antonio Gramsci (1971)who contended decades ago that the American system of mass produc-tion would most certainly diffuse to Europe over timeHowever the argument of this article is that even though British

French and American rms may adopt certain aspects of Japanesemanagement styles (eg just-in-time production self-managing teamsquality circles the use of statistical process controls) or some variant ofthe German vocational system their social systems of production willnot converge Moreover the overall conguration of a societyrsquos socialsystem of production inuences its sectoral and overall national economicperformance This helps to explain why societies have different systemsof innovation and why they vary in the clusters of industries in which

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

616

they are highly competitive in the international markets (Nelson 19931999 Hage and Hollingsworth 2000 Berger and Dore 1996)In order to understand how and why a societyrsquos economy performs

as it does it is necessary to understand its entire social system of produc-tion If a society is to modify substantially the performance of itsinstitutional sectors and its entire economy it cannot adapt only someof the management and work practices of its foreign competitors Ratherit must alter its entire social system of production Because a societyrsquosmodes of economic governance and coordination and its institutionalsectors develop according to particular logics and are system specicthere are serious limitations to the extent to which a society may mimicthe institutions (eg the rules and norms) institutional arrangements andinstitutional sectors of other societiesIn the history of modern societies there are logics by which institu-

tions coalesce into a social system of production (Hollingsworth 1991)Though institutions are constantly changing there are sharp limits tothe type and the direction of change that any particular institution canundergo because of its linkages with institutional arrangements and insti-tutional sectors Thus a societyrsquos business rms educational systemnancial markets industrial relations system etc can engage in seriousrestructuring only if most of the other institutional sectors also changeIn social systems of production there are pressures toward consistencyin the norms rules and values across institutional sectors though in anycomplex society social systems are obviously imperfectly integratedIndeed the degree to which the institutional norms and rules makingup a social system of production are loosely or tightly coupled is a vari-able of considerable importance In general the institutional sectorsmaking up a social system of production are interdependent and changesin one generally result in changes in the others Since each institutionallevel is dependent on the others for various types of resources there isinterdependence among the differing institutional spheres Moreovereach society has its norms moral principles rules and laws and recipesfor action as well as its own idiosyncratic customs traditions and prin-ciples of justice (Burns and Flam 1987)There are also other inherent obstacles to convergence among social

systems of production for where a system is at any one point in timeis inuenced by its initial state Systems with quite different initial statesare unlikely to converge with one anotherrsquos institutional practicesExisting institutional arrangements block certain institutional innova-tions and facilitate others (Roland 1990) Thus the institutions makingup a social system of production provide continuity even though insti-tutional arrangements are always changing but with a particular logicWhile Williamson (1975 1985) suggests that actors tend to choose theinstitutional arrangements which are most efcient North (1990) is

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

617

much closer to the mark in his argument that most societal institutionalarrangements exist as a result of custom and habit and are thereforeinefcient At any moment in time the world often appears to its actorsas very complex and uncertain For this reason actors often engage incontradictory forms of behavior pursuing different strategies as hedgesagainst a very uncertain world (Lanzara 1998) And their hedging andcontradictory forms of behavior may lead in somewhat different soci-etal directions all constrained by the institutional fabric within whichthe actors are embedded This kind of contradictory behavior occurs inpart because of the contradictions inherent in the dominant congura-tions of institutional arrangements with their contradictory logics (seelsquoSecond level of analysisrsquo above)Despite the emphasis on the logic of institutional continuity this is

not an argument that systems change along some predetermined pathThere are critical turning points in the history of highly industrializedsocieties but with the choices limited by the existing institutional terrainBeing path-dependent a social system of production continues along aparticular logic until or unless a fundamental societal crisis intervenes(Milgrom et al 1991 Krugman 1991 Durlauf 1991 Hollingsworth 1991Pred 1966 David 1988 Hodgson 1998)At this point it is important to confront the question of how social

systems of production evolve And this question gets to the heart of theproblem of building complex societal structures Certainly social systemsof production did not emerge from some process of social engineeringMoreover the various component parts of each social system of produc-tion have often not been designed to be part of a social system ofproduction The component parts of each social system of productionhave emerged more often than not as unintended by-products of goalswhich various actors had in mind at earlier moments in time It is usuallythe case that actors at time lsquotrsquo fail to comprehend the long-term conse-quences of their actions Wolfgang Streeck the author of several brilliantpapers on social systems of production argues that institutional sectordesigns created for one purpose generally address that goal but overtime those designs have quite unintended consequences (Streeck 1997a1997b Wright 1998) Describing the conguration of the German socialsystem of production of the 1970s and 1980s Streeck points out that itwas the unintended by-product of multiple points in history Someelements were pre-Wilhelmian others were introduced by the Alliedpowers after 1945 and others emerged during the years of the GermanFederal Republic All component parts of the German social system ofproduction

were and continue to be changing for their own reasons as wellas in reaction to each other and certainly there can be no presump-tion of a pre-established t between them even though one might

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

618

want to allow for some reinforcement effects of the modelrsquos histor-ically contingent social and economic success

(Streeck 1997b 54)

In sum the emergence of social systems of production is a long-termevolutionary process with each part interacting with its environment andresulting in a congurative whole but with no previous design by eithera single actor or a collectivity of actors As suggested above actors do not recreate their world anew At any moment in history the totalinstitutional complex of a society necessarily contains the resources and legacies of its past Preexisting institutional complexes are nevercompletely wiped out Societies consist of multiple layers of history withtheir diverse logics of action However the cumulative effect of smallperipheral changes in altering particular institutional sectors and insti-tutional arrangements can be substantial (Murmann 1998 Lanzara 1998)lsquoInstitution building is affected more by the ways in which people havecodied the past than by how they envision the future There are manymore potential resources in pre-existing institutional arrangements thanit is usually assumed or suspectedrsquo (Lanzara 1998 30)

Fourth level of analysis

The next level of institutional analysis organizational structures is some-what more controversial As suggested above Douglass North (1990)draws a sharp distinction between institutions and organizations Morerecently many organizational theorists (Powell 1991 Powell andDiMaggio 1991 DiMaggio and Powell 1983 Baum and Oliver 1992Kondra and Hinings 1998 Townley 1997) argue that institutional rulesnorms and conventions unfold in tandem with organizational structuresand this is my positionThe literature which focuses on how institutions inuence organiza-

tions is quite different from two other theoretical literatures which alsoare concerned with how organizational environments shape the behaviorof organizations There are resource dependency theorists (Pfeffer andSalancik 1978) who emphasize the role of environmental resources inshaping organizations while Hannan and Freeman (1977 1984) and otherpopulation ecologists emphasize the survival of organizations givencertain kinds of organizational conditions (Orru et al 1991)The institutionalist perspective as implied above emphasizes the

normative environment in which organizations are embedded It is a perspective which focuses on the way in which organizations in their behavior tend to conform to the institutional rules and normswhich are dominant in the organizational environment However allthree perspectives ndash the institutionalist the population ecology and

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

619

the resource dependency ndash emphasize how the environment inuencesorganizations and how organizations which are subject to the sameenvironment tend to converge in their behavior to have what DiMaggioand Powell have labeled organizational isomorphism (DiMaggio andPowell 1983)There has emerged a vast literature which demonstrates that within

each society there are scal political judicial and other regulatory normswhich limit and shape the culture and structure of organizationalbehavior The normative institutional environment of organizations limits the options of what organizations do in a particular society and inuences the patterns of ownership relations with suppliers andcustomers In short it is the normative environment of organizationswhich denes within a particular society what is socially acceptablebehavior for organizations (Hamilton and Biggart 1988 Hollingsworthand Hollingsworth 2000 Meyer and Rowan 1991 Zucker 1987 1988Orru et al 1991 Townley 1997)Thus far most of the analyses on isomorphic pressures operating on

organizations have been cross-sectional in nature However the entireconcept of isomorphism implies that there are strong environmental pres-sures exerted on organizations and in order to observe this phenomenonit is important that we have longitudinal studies which assess howchanges in an organizationrsquos institutional environment inuence changesin the structure and culture of organizations Moreover the expectationhere is that when one analyzes the patterns of organizational changeover time the isomorphism will be holistic ndash that is within a particularsociety there will be a tendency for the internal structure of particulartypes of organizations to converge in large part because of the pres-sures to conform to the changes in the external norms rules and valuesA close reading of Aoki (1990) and Slack and Hinings (1994) nds thatchanges in the national institutional environment of organizations haveinuenced changes in the internal structure of organizations within aspecic industry of a particular societyHowever it is inappropriate to view organizations as changing only

in relation to exogenous environmental change There is a kind of co-evolution which occurs between organizations and their institutionalenvironments Co-evolution implies nonlinear feedback between orga-nizations and their institutional environments This is of course quitedifferent from the more common way of studying organizations bymodeling relations between independent and dependent variables Ofcourse the more common type of analysis is quite appropriate whenthere are simple relationships which do not involve complex feedbackprocesses But in the strategy for institutional analysis which is proposedherein it is less useful to separate independent from dependent vari-ables and more useful to understand the interacting and co-evolutionary

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

620

processes among all of the levels of institutional analysis discussed inthe various sections above (Baum and Singh 1994a 379ndash80 Mowery andNelson 1999 Murmann 1998)The logic of the perspective herein suggests that within every

society there is variation in the structure and culture of business rmsuniversities and other complex organizations (Kondra and Hinings1998) However this variation takes place within parameters which aresystem specic For example every German rm and every Germanuniversity is different from every other German university and rm but there are a set of cultural and structural characteristics which distin-guish German rms and universities from those in America Similarlyeach Japanese rm and university is unique yet they have a set ofcultural and structural characteristics which are system specic andwhich differentiate them from business rms and universities in other countriesIn all societies each organization has its own distinctive organizational

rules norms and conventions which are subordinate to the meta norms and rules of the larger society within which they are embeddedHowever the strength of the institutional environment within whichorganizations are embedded varies from society to society Some societies have multiple institutional environments and there is hetero-geneity in terms of what constitutes appropriate behavior oforganizations In such societies there can be conict over whatinstitutional logics should regulate specic organizational functions(Friedland and Alford 1991 Townley 1997 262ndash3) In those societies inwhich the institutional norms habits and rules are most developed and in which the institutional pressures to conform are greatest thereis less variation in the structure and culture of business rms and variouskinds of research organizations In such societies the connectednessbetween research organizations and their institutional environments is sufciently strong for organizations to have low autonomy to pursue independent strategies and goals in these societies there is a great deal of organizational isomorphism Conversely the weaker the institutional environment in which research organizations areembedded the greater the variation in the structure and culture of business rms and research organizations Moreover where the institu-tional environments are more weakly developed organizations havegreater autonomy and exibility to respond to the development of new knowledge and to be highly innovative Hence it is in those societies where the institutional environments are most developed and most rigid and when there is less organizational autonomy and exibility that fewer radical innovations in basic and applied science as well as in totally new products and industrial sectors havebeen made

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

621

Fifth level of analysis

We come next to the outputs and performance of the various institutionalcomponents of a society It is at this level that institutional componentsare more pragmatic and exible For example in the legal sector thereare specic statutes and court rulings in the state sector there are specicpolicies in the business sector there are new products new technolo-gies and market strategies It is at this nal level of analysis thatinstitutional spheres are most open and susceptible to change and cross-national mimicry is easiest and most commonThrough the outputs of the society we can obtain some assessment as

to how well a society is performing Moreover we can assess how inno-vative it is how egalitarian it is in the distribution of its resources howegalitarian it is in terms of levels of health education etc Just as it iscomplicated to measure how egalitarian a society is similarly it is dif-cult to assess how innovative it is For example societies vary in makingincremental and radical breakthroughs in basic and applied science in developing totally new products and new kinds of organizations inincremental and process innovations in existing products and organiza-tions in developing new and different forms of marketing ndash bothdomestically and globally (Hage and Hollingsworth 2000) Of coursesocieties vary in their rates of economic growth in their rates of economicproductivity in their quality of life (eg rates of crime life expectancyetc) All of these performance criteria feed back and inuence each ofthe levels of institutional analysis discussed above ndash rules norms valuesetc institutional arrangements institutional sectors the structure andform of organizations Moreover performance measures may inuenceeach other For example if a society has very low rates of growth itmay not be very innovative in some of the types of innovative activitiesmentioned above Good or poor performances of certain kinds inuenceother performance indicators (Hollingsworth et al 1990)Different institutional arrangements and different social systems of

production result in different types of economic performance Hence solong as societies have different social systems of production there areserious constraints on the degree to which they can converge in theirinnovative styles Different social systems of production tend to maxi-mize in a more or less explicit manner different performance criteriausually mixing considerations about static and dynamic efciency protsecurity social peace and economic andor political power In short incontrast to the implications of neoclassical economic theory in real-worldeconomies there are no universal standards all economically rationalactors attempt to maximize Economic history provides numerous exam-ples of how a great variety of principles of rationality are implementedin different societies

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

622

A critical question in institutional analysis is whether an existing socialsystem of production which supports a set of routines for a particularkind of technology and industry can shift from old practices to newones There are numerous historical accounts of how the social systemof production in a particular society which worked so well for a numberof years could not adapt to new technologies (Lazonick 1990 Veblen1915 Nelson 1994) Schumpeter (1983) and Freeman (1991) have devel-oped the idea of a techno-economic paradigm which suggests thatdifferent technologies require specic forms of organizational arrange-ments ndash or what I called social systems of production Because oflsquoinstitutional inertiarsquo the social system of production of a society maynot be able to adapt to a new techno-economic paradigm (Nelson 199458 Lanzara 1998)Whether a social system of production can sustain its performance

standards depends not only on its intrinsic economic rationality but alsoon where it ts into a larger system If a particular social system ofproduction is immune from the innovativeness and competition of analternative system survival can be long lasting But if different socialsystems of production with diverging criteria of good economic perfor-mance meet in the world arena the arbitrariness of nationally imposedconstructed performance standards may be superseded by alternativeperformance criteria as a result of international competitivenessThe world economy is also socially constructed just as are national

economies Even if different social systems of production are competingin the international arena it is not always possible to determine whichis more competitively effective at any moment in time Hegemonicnation-states in the short run shape the rules of trade to favor their indus-trial sectors and rms But the history of hegemonic powers suggeststhat in the longer run such social systems of production sustainedlargely by military and political power eventually give way to moreinnovative and competitive social systems of production In our ownday as nation-states are increasingly integrated into a world economyeven hegemonic powers lose their innovative and competitive advan-tage and their share of world output decreases Such a country mayattempt to restructure its institutional arrangements and to readjust itsperformance preferences But to restructure its system of innovationgenerally calls for a major redistribution of power within a societyLargely for this reason societies have historically had limited capacityto reconstruct a system of innovation in the image of their major competi-tors Each countryrsquos social system of production is a conguration of ahost of norms rules and values as well as of institutional arrangementsEach system is constantly changing and is open to inuence from othersystems And indeed many technologies and practices diffuse from onesociety to another but a societyrsquos capacity to be innovative is constrained

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

623

by the existing social system of production Thus the same technologymay exist in numerous countries but how it is employed and how itinuences societal outcomes and performances varies from one institu-tional conguration or society to another (Hollingsworth 1998)

PERSPECTIVES ON INSTITUTIONAL CHANGE

A fundamental problem which remains unresolved in institutionalanalysis involves the nature of institutional change There is much confu-sion and miscommunication in the social sciences about institutionalchange in large part because we do not even know how to measure therate of institutional change Of course we know that there are bothexternal and internal focuses for change and we have had a good bitof scholarship on this subject What is not clearly understood is thatinstitutions and institutional arrangements within societies are histori-cally rooted that there is a great deal of path-dependency to the waythat institutions evolve Moreover the more intricately linked each sectoris with each other sector and with a societyrsquos rules and norms the lesschoice actors have to devise new institutions and institutional arrange-ments On the other hand if a sector or actor is too isolated it may betoo weak within a society to bring about any kind of effective changeThese considerations are very critical to the problem of how muchfreedom actors have to create new institutions afresh to what extentinstitutions diffuse from one society to another (Murmann 1998 Lanzara1998) This is a fundamental question for students of institutionalanalysis as we attempt to understand how the different institutionallevels are linked togetherThere are a variety of reasons why there is confusion in studying insti-

tutional change There is disagreement over the extent to which actorshave the freedom to build new institutional arrangements and if theydo to what extent the new arrangements and organizational patternsmay depart from past practices Among some institutional economists(Williamson 1985) there is a basic assumption that actors build efcientinstitutions by pursuing their self-interest and by promoting theirstrategic goals Moreover there is the assumption that a mixing of fullrationality with market competition tends to produce all of the optimalinstitutions that are needed in order to coordinate a complex capitalisteconomy (Hollingsworth and Boyer 1997 ch 14 Hodgson 1999) Incontrast the argument developed here is much more complexSocial institutions are historically rooted and there is a great deal of

path-dependency in the way that various institutional componentsevolve The shape of institutional congurations at any moment limitsthe type of options for change Because there is a great deal of institu-tional inertia radical change in the institutional components of a society

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

624

is uncommon (Lanzara 1998) Recognizing the constraints existinginstitutions exert on actors we note that existing congurations of insti-tutional arrangements institutional sectors and organizational propertiescan limit but also inuence the degree to which norms and rules canchange Many scholars (North 1990 Ostrom 1990) have emphasized therole of norms rules ideologies and values in limiting the ability of actorsto develop new institutional arrangements which radically depart fromexisting ones However there is a two-way process of interaction aprocess of co-evolution (Baum and Singh 1994a Nelson 1994) Thus asFriedland and Alford (1991 244) argue actors develop new institutionalarrangements by recognizing organizational failures and low perfor-mance but these new arrangements do not necessarily change a societyrsquosnorms rules habits and its underlying value systemCampbell (1997) advances our understanding of how existing institu-

tions enable actors to construct radically new institutions He recognizesthat most institutions and institutional arrangements embody a degreeof inertia from which it is difcult for actors to depart Moreover thedifferential power relations among actors make it unlikely that lesspowerful actors can change the existing power structure of a societyMost change in the institutional components of a society evolves througha process of constrained selection which reects to a considerable degreethe existing arrangements and power relations This becomes the inter-pretative frame for social actors with institutionalized scripts and ritualswhich tend to be taken for granted and appear to be quite ordinary andnatural This institutionalized scripting of the social world is an extremelyimportant source of social stability and inertiaCampbell has developed a very fruitful explanation of change in terms

of the various institutional components outlined here He distinguishesbetween incremental and radical institutional change and argues thatradical institutional change occurs when social actors with widelydiffering norms cultural scripts and rituals for action engage in intenseinteraction with each other (see also Knight 1992) The more diverse theinteractions the greater the potential for institutional change He arguesthat as shifts in the composition of interaction among social actors occurchanges in the interpretation of problems and interests will follow Ifthere are only minor changes in the extent to which diverse actorsinteract there will be only modest institutional change But the morefundamental the changes in the interaction of diverse social actors thegreater the change in the way that actors interpret their world And thegreater the change in the way that actors interpret their world the morelikely that radical changes in various institutional components willfollow In other words diverse decision making tends to create a widerrange of interpretative frameworks of the social world than is likely tobe the case if social actors operate in isolation

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

625

This perspective of diverse interactions among social actors facilitatesour ability to move beyond the traditional view that existing institutionsconstrain the range of institutional alternatives which actors face Theolder view has tended to suggest that actors engage in institutionalchange by extending existing institutional principles habits and conven-tions Specifying the conditions which facilitate fundamental institutionalchange permits a much richer perspective for institutional analysisCampbellrsquos perspective about the conditions under which fundamentalor radical change in institutional components occurs is consistent withmy own research about the conditions under which major discoveriesor fundamental new knowledge occur the more scholars with diversebackgrounds interact with intensity and frequency the greater the like-lihood that they will develop new and alternative ways of thinking(Hollingsworth and Hollingsworth 2000)These views of institutional change are also consistent with the recent

work of Nelson and his colleagues (Mowery and Nelson 1999 Murmann1998) They report the research on a variety of industrial sectors in anumber of different countries revealing that networks of actors in partic-ular industries business and professional associations universityresearch departments and institutes are frequently able to use their collec-tive power to modify a societyrsquos property rights research system andgovernmental policies It is the intense and frequent interaction of thesediverse complex sets of actors ndash sometimes constituting a technologysystem ndash that often leads to incremental changes in a societyrsquos total socialsystem of production It is this kind of collective action that brings aboutorganizational change change in institutional sectors institutionalarrangements and even in the rules and norms of a society And if weare to understand how the style of innovativeness of a society changeswe must rst understand the changes in its institutional makeup It ishighly unlikely that a society will develop a new style of innovativenesswithout changing its institutional structure

FIRMS THEIR INSTITUTIONAL ENVIRONMENT ANDNATIONAL SYSTEMS OF INNOVATION

Variation in the innovative style of societies is shaped by their insti-tutional makeup ndash the various levels which are discussed aboveinstitutions institutional arrangements the structure and cohesion of institutional sectors which constitute a societyrsquos social system of pro-duction and the structure and culture of its organizations ndash especiallyits business rms and other research organizations By undertaking aninstitutional analysis of a society one can begin to understand in whatkinds of organizations the production of specic kinds of knowledgetakes place and how this is linked to particular kinds of innovativeness

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

626

With such a perspective one can gain a rich understanding of why somesocieties excel in the production of radical breakthroughs in basic andapplied science and in developing radically new products and why othersocieties excel in more incremental innovationsReecting on these issues one cannot help being impressed with the

contrasts between the innovative styles of German and American orga-nizations For example Germany has continued to be very successful in making incremental innovations in many industries in which it wasvery competitive well before World War II paper printing materialsmachinery electrotechnical products motor vehicles chemicals textileyarns and fabrics But the Germans have been much less successful indeveloping totally new products or related innovations in biotechnologyelectronics telecommunications aircraft as well as other newer indus-trial areas In contrast the Americans have been very innovative in these and other industries with products having very short lives andwith technologies which change very rapidly and are very complex(Hollingsworth 1991 Soskice 1997 1999)The importance of institutions and innovativeness for international

competitiveness has led to the recognition that there are national systemsof innovation (Lundvall 1992 Nelson 1993 Edquist 1997 Hage andHollingsworth 2000) But even before the recent studies of nationalsystems of innovation Landes (1969) was a proponent of these viewswhen he made the argument that Germany had developed a nationalsystem of innovation by building on its education and scientic researchsystems which were lacking in Britain and elsewhere However the concept of a national system of innovation immediately poses theproblem of what kind of innovations should be consideredDavid Soskice an economist at the Wissenschaftszentrum Berlin has

addressed a set of problems somewhat complementary to the perspec-tive raised here His work (Soskice 1990 1997 1999) suggests that weare sensitive to the institutional environment in which the followingactors are embedded the employees of rms the owners and nanciersof rms the competitors of rms and nally their collaborators ndash aperspective complementary to Hollingsworthrsquos work on social systemsof production (Hollingsworth 1997 1998) When one reects on the insti-tutional contextsocial system of production within which these actorsare embedded one becomes aware that those countries which have some-what inexible labor markets external to rms but a business system in which rms have long-term commitments to their employees arestrongly associated with incremental but hardly at all with radical inno-vations In the rms of such countries (eg Germany and Japan) thereis much more consensus decision making than in societies where a lsquohireand rersquo set of practices is quite pervasive And this kind of consensusdecision making ndash while highly conducive to incremental innovativeness

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

627

ndash limits the capacity for radical innovations to occur This kind of lsquolongtermismrsquo set of norms and rules occurs not only in the employmentpractices of incrementally innovative rms but also in their nancingand ownership patterns (Hollingsworth 1997) During the past thirtyyears these were characteristics of many rms in Germany and JapanSignicantly in both of these countries venture capital markets werenot highly developed nancing tended to be based more on long-termbank loans than on the equity markets and the relationships of rmstended to be relatively stable And these stable institutional arrange-ments also contributed to a high degree of innovativeness ndash butincremental rather than radical innovationsMeantime we should also pursue Soskicersquos ideas (1999) about how

rmsrsquo relationships with their competitors and collaborators shape theirstyles of innovativeness In those advanced industrial societies in whichbusiness rms are highly mobilized into business associations theretends to be a high degree of standardized quality controls over technicalnorms and a high degree of reliance on legal sanctions as monitoringinstruments vis-agrave-vis association members As a result of this high degreeof integration of rms into associations member rms are more likelyto be associated with incremental innovations than is likely to be the case with rms in those societies where associations are less welldeveloped In the latter type of society radical innovations are morelikely to occur In those advanced industrial societies where rms havelong-term stable relationships with collaborators in product design andproduction the style of innovation is likely to be incremental And it is in those societies in which there is a weaker tradition of strongcollaboration among rms that radical innovations are more likely tooccurObviously even in a country where there may be a great deal of

radical innovativeness most innovations will be of an incremental typefor there are simply fewer radical innovations But Germany and therest of Europe not only have had relatively few radical innovations incomparison with the United States but they also have lagged behindthe USA in developing relatively totally new industries Thus there hasbeen much less innovative activity in Europe in high-technology andnewly emerging industries than in the USAIn complementary research Audretsch and Feldman (1995) have

demonstrated that the USA has been more successful in spawning rmsengaged in the early stages of the industry life cycle but relative toGermany had an environment unfavorable to innovative activity byestablished rms especially in more mature industries German rmshave tended to be associated with a routinized technological regime ndashwhere established enterprises have had strong incentives to engage inincremental innovations but new rms have tended to have fundamental

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

628

innovative disadvantages relative to the Americans (see Audretsch 1995Winter 1982 Casper 2000)On both sides of the Atlantic universities engage in basic and applied

research But signicantly the institutional environment in which univer-sities are embedded inuences the degree to which the research is likelyto represent major breakthroughs or is likely to have more modestimplications Recent research in biomedical science demonstrates that inthose societies in which university professors andor graduate studentshave the following characteristics there are likely to be relatively majorresearch breakthroughs graduate students have a long training periodhigh dependency on a major professor low autonomy in dening thenature of a research project restricted job options and a university envi-ronment where there is a very modest or nonexistent entrepreneurialculture For example in Germany ndash a country with relatively few radicalbreakthroughs ndash the scholar who nishes a habilitation will have spentmany years on a research project before qualifying for a permanent acad-emic position will have been in a high-dependency position with a seniorprofessor will have had limited autonomy to dene a research projectand will be constrained from pursuing a high-risk research project Theyoung German will have an extraordinarily high level of training butwill have faced a number of disincentives to undertake high-risk researchAnd as soon as heshe has completed the habilitation and becomes aprofessor administrative responsibilities constitute another disincentivefrom engaging in high-risk or long-term research projects Moreover theGerman professor is a civil servant with a more stable and secure incomethan hisher American counterpart In a very different institutionalenvironment American professors in large research universities tend tobecome much more entrepreneurial Today many American professorsbehave as though they were operating their own rms within univer-sities and thus devote considerable time to generating funds for researchprojects assistants secretaries supplies etc And the variation in thisentrepreneurial spirit across countries is associated with variation in thetaking of high-risk research strategies and the making of major researchbreakthroughsIn sum societies vary in terms of the institutional environment in

which their organizations are embedded and this variation inuencesthe degree to which they have industrial activities in the early or latestages of the industry life cycle successful new or mature industriesand radical vs incremental innovations And it is this variation whichis encouraging rms in countries with relatively few new industries and with few or no radical innovations to invest in those countries with many new industries and numerous radical innovations ndash as astrategy to be able to transfer new technologies from the latter institu-tional environment to the former

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

629

Table 4 summarizes the way that the norms rules and values of a society are associated with Americarsquos and Germanyrsquos institutionalarrangements their social system of production and their styles of inno-vativeness

CONCLUSION

Several themes deserve to be reemphasized Most importantly institu-tional analysis is emerging as a dominant issue in several social sciencedisciplines The study of various institutional components discussedabove is a complex subject requiring multiple levels of analysis and aknowledge of multiple sectors within societies It requires a theoreticalperspective that evolves from an empirical analysis in different societiesIf we are to advance the study of institutional analysis we need cross-disciplinary collaboration and team research which are not readilyachieved given the structure of most of our research universitiesAt present scholars in each academic discipline have their own distinc-

tive strategies for studying institutional analysis Thus economists tendto use several approaches (North 1990 Williamson 1985 Hodgson 19881998 1999) anthropologists another (Geertz 1995) and sociologistspolitical scientists and historians employ other strategies (Campbell etal 1991 Powell and DiMaggio 1991 Katznelson 1998 Hall and Taylor1996) With practitioners in various disciplines pursuing their distinctiveapproaches to institutional analysis there is low potential for collabo-rative institutional analysis across the social sciences The lessons fromthe biological sciences over the last 40 years are very instructive oncepractitioners from various biological disciplines began to cooperate instudying biological phenomena by using the same concepts at the mole-cular level theoretical advance was very rapid Comparable phenomenatook place in the development of other hybrid elds eg biochemistrybiophysics (Judson 1979 Olby 1979 Kohler 1982 1994) Similarly weresocial scientists in various disciplines to work collaboratively in the areaof institutional analysis there would very likely be an acceleration intheir theoretical knowledgeThis is not a plea that everyone engaged in institutional analysis should

do the same kind of research Indeed just as biologists working at themolecular level study many different kinds of problems similarly socialscientists engaged in institutional analysis would also work on manyseparate problems but in the spirit of a collective enterprise Some wouldconcentrate their attention on the study of rules norms habits conven-tions and values while others would study how these are associatedwith congurations of various institutional arrangements (types ofmarkets hierarchies networks associations communities clans statesetc) Others would work on specic institutional sectors (eg education

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

630

Table 4

Rules norms hab

its an

d values associated w

ith sp

ecic co

ngu

ration

s of coo

rdinating mecha

nism

s soc

ial sy

stem

s of

prod

uction

an

d nationa

l styles of inno

vation

USA

Germany

Norms rules values which provide incentives for

Norms rules values which provide incentives for

short-term horizons

long-term horizons

Con

gu

rative

forms of

Marke

ts c

orpo

rate hierarchies reg

ulatory state

Assoc

iation

s (busine

ss assoc

iation

s la

bor union

s)

coordinating

corp

oratist-type

networks

marke

ts c

orporatist-

mecha

nism

s (listed in

type

state

order of im

portanc

e in

the particu

lar social

system

)

Social sys

tem of

prod

uction

Bus

iness system

Firm

struc

tures

Con

glom

erates b

ut with mov

emen

t toward

Firm

more co

ngru

ent with pa

rticular prod

ucts

elim

ination of produc

tion

in som

e sectors

emph

asis on long

-term strateg

ies rm

s are well

emph

asis on sh

ort-term

strateg

ies high

integrated

differentiation

between divisions

of rms

Firm

s mob

ilized in

Low

High

and in

tegrated

into busine

ss

associations

Indu

strial relations

Low

deg

ree of jo

b secu

rity rigid in

ternal lab

or

High deg

ree of jo

b secu

rity

exible interna

l lab

or

system

marke

t ex

ible externa

l labo

r marke

tsmarke

t rigid externa

l lab

or m

arke

tPatterns of

Dispersed owne

rship frequ

ent turnov

erMore stab

le owne

rship patterns

owne

rship

Table 4

continue

d

USA

Germany

Norms rules values which provide incentives for

Norms rules values which provide incentives for

short-term horizons

long-term horizons

Social system of

Prod

uction

Bus

iness system

Training for labo

r an

dHigh deg

ree of voluntary decision mak

ing by

Collective an

d m

ore co

mpulso

ry d

ecisions

abo

ut

man

agem

ent

individuals as to ho

w m

uch

tim

e an

d ene

rgy

how m

uch ed

ucation

is ne

eded

he

avy em

pha

sis

to inv

est in educ

ation heavy

emph

asis on

on techn

ical training (eg en

gine

ering)

busine

ss sch

ools training

in marke

ting

and

sales accou

nting

Fina

ncial marke

tsEqu

ity an

d ven

ture cap

ital m

arke

ts highly

Equ

ity an

d ven

ture cap

ital m

arke

ts w

eakly

dev

elop

ed na

ncing from

cen

tral gov

ernm

ent

deve

lope

d he

avy em

pha

sis on

corporate loan

sto priva

te rm

s for dev

elop

men

t of new

tech

nologies related

to na

tion

al security

Unive

rsity research

High individual auton

omy high poten

tial for

Low

ind

ividua

l au

tono

my until almost 40

yea

rs of

system

individual creativity u

nive

rsity pr

ofesso

rs

age poten

tial for ind

ividual creativity high

ly

high

ly entreprene

urial

cons

traine

d unive

rsity pr

ofessors not so

entrep

rene

urial

Inno

vative

styles

Highly su

ccessful in m

ajor break

through

s in

High pe

rforman

ce in proc

ess an

d inc

remen

tal

basic an

d app

lied scien

ce go

od in dev

elop

ing

inno

vation

s in existing produ

cts not so su

ccessful

totally

new

lines of pr

oducts an

d in m

ajor break

through

s in basic scien

ce or in

tech

nologies high

deg

ree of entrepr

eneu

rship

deve

loping

totally new

lines of prod

ucts of

amon

g research

ers

tech

nologies low d

egree of entrepr

eneu

rship

amon

g research

ers

Table 4

continue

d

USA

Germany

Norms rules values which provide incentives for

Norms rules values which provide incentives for

short-term horizons

long-term horizons

Prod

ucts in w

hich

Pr

oducts with sh

ort liv

es prod

ucts in w

hich

Older produ

cts with long

history machine

ry pap

er

society excels

tech

nology

is bo

th highly co

mplex

and

pr

inting

textile

s m

otor veh

icles m

achine

too

lsch

ange

s rapidly en

tertainm

ent indus

try

pha

rmaceu

ticals biotechn

olog

y softw

are

aerosp

ace

business systems nancial markets and systems of research) Distinctivewould be the recognition of how these separate studies are linkedtogether how each institutional sector is linked to a societyrsquos normsrules values etc and its conguration of institutional arrangementsSimilarly other institutional analysts might focus on the study of orga-nizations with a major concern for how the institutional environmentof organizations (eg norms rules values congurations of institutionalarrangements institutional sectors) inuenced the structure culture andoutcomes of organizations And nally others might study how all ofthe aforementioned aspects of institutional analysis inuenced a societyrsquosoverall performance Nelson and Sampat (1998) have appropriatelyreminded us that we should engage in institutional analysis in order toaddress specic problems but the view developed in this article is thatwe rst need some mapping of the terrain of institutional analysis inorder for our diverse enterprises to become a meaningful collective enter-prise especially if we are to understand how institutions inuence theinnovative processThe argument developed here is not that all social scientists should engage

in institutional analysis Obviously there are numerous other importantareas of research which lie beyond the eld of institutional analysis Nordoes this article argue that institutional analysis is the most importantsubject for social scientistsSince the structure and culture of our research universities reect an

enormous amount of disciplinary fragmentation what are the prospectsfor promoting the kind of collective and interdisciplinary research agendaproposed above In one respect the prospects are not encouragingDuring the past half century the efforts of American universities to have effective communication across the biological sciences have variedgreatly in their success At the University of California-Berkeley it was not until the 1980s that there was a serious effort to integrate thebiological sciences and at that time the campus abolished seventeendepartments to promote an integrated biological program But at theUniversity of Wisconsin the biological sciences are still very fragmentedinto dozens of different departments in botany plant pathology zoologyand physiology with two departments of biochemistry and two depart-ments of geneticsThere have been other approaches which have worked very well histor-

ically for promoting scientic integration amidst scientic diversity Forexample several summer schools and special institutes were organizedin the United States during the 1940s in an effort to overcome the frag-mentation in the biological sciences These programs made it possiblefor scientists in diverse academic disciplines and from different univer-sities and countries to come together to engage in intense and frequentinteraction Perhaps the most effective program was that held in the

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

634

summers at Cold Spring Harbor New York where a research agendafor the study of molecular biology was developed (Fischer and Lipson1988 Stent and Watson 1966) Scientists went there summer aftersummer and when they returned to their universities many began therestructuring of the biological sciences in their universitiesIn our own day we need summer institutes and other similar programs

so that scholars from diverse backgrounds can come together anddevelop a collective agenda for doing institutional analysis and forstudying the impact of institutions on the performance of societies ndashinnovativeness being one such subject If we can develop such instituteswe will increase the prospects for transforming our universities Thereis reason to be optimistic that once an eclectic group of scholars workwith a common set of concepts we can have genuine theoretical advanceif researchers can bring their diverse expertise to bear on a common setof problems There is considerable evidence that institutional analysisoffers high potential for advancing our knowledge of innovations andtechnological change (Hollingsworth and Hollingsworth 2000) If we aregoing to make substantial theoretical advance it is not likely to occurfrom the vantage point of a multiplicity of disciplines Rather it is likelyto occur as we develop a new intellectual framework and our own acad-emic journals This is a tall agenda which requires us to reassess theway that knowledge is presently produced in our societies

NOTE

This article is part of a much larger agenda involving the study of institutionsorganizations and innovations in which I am involved My debts are many andvaried The rst version was written while I was in residence in the ResearchUnit for Institutional Change and European Integration of the Austrian Academyof Sciences in Vienna in the summer of 1998 I am especially grateful to ProfessorsEgon Matzner and Sonja Puntscher-Riekmann for providing the stimulating envi-ronment in which to think through the ideas developed here Egon Matzner hasengaged in very stimulating discussions with me about the importance of insti-tutions in the general eld of socio-economics The nal version of the articlewas completed during my residence as a fellow at the Neurosciences Institutein La Jolla California I also wish to thank my colleagues in the InnovationTheme Project at the Netherlands Institute for Advanced Study (NIAS) during1998ndash9 and Edgar Grande and Karl Muumlller for stimulating conversations aboutthe interaction of institutions and organizations in facilitating particular typesof innovation I am especially grateful to Jerald Hage for helping me to developa number of issues in the article and for his useful comments on the rst draftFrans van Waarden early on helped me to work through the idea of multiplelevels of institutional analysis Others who have helped to develop the ideas inthe article have been my colleagues in the research project lsquoComparative socialsystems of productionrsquo especially Tom Burns Christel Lane Yoshitaka OkadaWolfgang Streeck and Richard Whitley My colleagues Robert Boyer SteveCasper and David Soskice have taught me much about how incentives providedby the institutional environments of rms inuence individual decision making

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

635

But my most important debt is to Ellen Jane Hollingsworth who has repeatedlyquestioned every single assumption of this article Without her rich and contin-uous tutoring this article would be even less developed than it is at presentFinally without the insightful assistance of David Gear in many phases of thisarticle it would not have been completed

REFERENCES

Aoki Masahiko (1990) lsquoTowards an economic model of the Japanese rmrsquo Journalof Economic Literature March 1ndash27

Arora A Landau Ralph and Rosenberg Nathan (eds) (1998) Chemicals and Long-Term Economic Growth Insights from the Chemical Industry New York JohnWiley amp Sons

Arrow Kenneth (1974) The Limits of Organization New York NortonArthur W B (1988a) lsquoSelf-reinforcing mechanisms in economicsrsquo in P W

Anderson K J Arrow and D Pines (eds) The Economy as an Evolving ComplexSystem Redwood City Calif Addison-Wesley

mdashmdash(1988b) lsquoCompeting technologies an overviewrsquo in G Dosi C Freeman RNelson G Silverberg and L Soete (eds) Technical Change and Economic TheoryLondon Pinter Publishers pp 590ndash607

Audretsch David B (1995) Innovation and Industry Evolution Cambridge MassMIT Press

Audretsch David B and Feldman Maryann P (1995) lsquoInnovative clusters andthe industry life cyclersquo Discussion Paper Wissenschaftszentrum BerlinMarket Processes and Corporate Development Research Unit FS-IV-95-7

Baum Joel A C and Oliver Christine (1992) lsquoInstitutional embeddedness andthe dynamics of organizational populationsrsquo American Sociological Review 57540ndash59

Baum Joel A C and Singh Jitendra V (1994a) lsquoOrganizationndashenvironment coevo-lutionrsquo in Joel A C Baum and Jitendra V Singh (eds) The EvolutionaryDynamics of Organizations New York Oxford University Press pp 379ndash402

mdashmdash(eds) (1994b) The Evolutionary Dynamics of Organizations New York OxfordUniversity Press

Berger Suzanne and Dore Ronald (eds) (1996) National Diversity and GlobalCapitalism Ithaca NY Cornell University Press

Boyer Robert (1997) lsquoThe variety and unequal performance of really existingmarkets farewell to Doctor Panglossrsquo in J Rogers Hollingsworth and RobertBoyer (eds) Contemporary Capitalism the Embeddedness of InstitutionsCambridge and New York Cambridge University Press pp 55ndash93

Burns Tom R and Dietz Thomas (2001 forthcoming) lsquoSocial rule system theorysocial action institutional arrangements and evolutionary processesrsquo in JRogers Hollingsworth and Karl Muumlller (eds) Advances in SocioeconomicsLanham Md Rowman amp Littleeld

Burns Tom R and Flam Helena (1987) The Shaping of Social Organization BeverlyHills Calif Sage

Calvert Randall L (1995) lsquoThe rational choice theory of social institutionsrsquo inJeffrey S Banks and Eric A Hanuskek (eds) Modern Political Economy NewYork Cambridge University Press pp 216ndash66

Camic Charles (1986) lsquoThe matter of habitrsquo American Journal of Sociology 911039ndash87

Campbell John L (1997) lsquoMechanisms of evolutionary change in economicsgovernance interaction interpretation and bricolagersquo in Lars Magnusson

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

636

are learned through habit individuals rely on the acquisition of suchcognitive habits before reason communication choice or action arepossibleWhereas Schotter (1981) and other game theorists take the individual

as an agent unencumbered by previous habits Field (1984) and othershave stressed that there can be no games without prior norms and rulesand thus a set of norms and rules must be presumed at the start Thosewho attempt to explain institutions from individual behavior alone areusing a bad strategy (Hodgson 1998)The position here ndash heavily inuenced by Hodgson (1998 1997 1988)

ndash is that individuals are embedded in a complex institutional environ-ment and that institutions not only constrain but also shape individuals(also see Hollingsworth and Boyer 1997) It would be a mistake howeverto get involved in an innite regress in order to determine which camerst ndash individuals or institutions Of course institutions are formed andchanged by individuals just as individuals are shaped and constrainedby institutions But at a macro level it is institutions that provide acognitive framework whereby individuals can cope with their reality Inthis sense the micro and macro worlds are intertwined At the macrolevel there is considerable stability but at the micro level individualshave a signicant level of autonomy and there can be widespread diver-sity As Hodgson (1988) reminds us most institutions in a temporal senseexist prior to the individuals in any given societyIt would be a serious mistake to downplay the importance of indi-

viduals and micro level analysis as we study institutions In the nalanalysis it is at the level of individuals that norms rules habits conven-tions and values exist An individual is born into and socialized intogroups and a society and this is how one early in life acquires a senseof appropriate forms of behavior Because of the way that individualsare socialized into a world of rules norms habits conventions andvalues it is unnecessary for individuals to restructure the world anewevery day (Douglas 1987 Elster 1989) Every action does not have tobe seriously reected upon For this reason institutions provide cogni-tive frameworks for individuals make their environments predictableprovide the information for coping with complex problems and envi-ronments In the words of Johnson (1992 26) lsquoInstitutions reduceuncertainty coordinate the use of knowledge mediate conicts andprovide incentive systems By serving these functions institutions providethe stability necessary for the reproduction of societyrsquo However eachsociety has different forms of habits rules and norms and hence differentincentives and disincentive systems for learning and forgetting forprocessing information But because individuals have varying degreesof autonomy individuals and groups can deviate from the prescribedforms of behavior in a society And of course these changes at the level

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

603

of individuals become important in understanding processes of societalchangeThese views are not meant to imply that the type of institutional

analysis proposed herein approximates a general theory of society Thisis clearly not the case However it is intended to represent the rst stepsin a mapping exercise of the boundaries of institutional analysis and tosuggest a few methodological insights for studying institutionsOne should try to see norms and rules as continuous and not as

dichotomous entities to recognize that they come in varying strengthsLegro (1997) has suggested that we assess the robustness of norms andrules with three criteria their simplicity their durability and their concor-dance Simplicity refers to how well actors understand norms and ruleshow well they can be applied within a specic situation Some normsand rules can be so complex that actors can have considerable difcultyin applying them in specic cases Durability addresses the issue of howlong norms and rules have been in effect ndash in short in order to assesstheir level of legitimacy While the position of this article is that normsrules and values are quite durable they do vary in this respectConcordance refers to how widely applied a norm or rule is Thisaddresses the degree to which a rule is a meta rule the degree to whichit incorporates the heterogeneity of other norms and rules In sum theclearer the norms and rules of a society the longer they have been inexistence and the more widely applicable they are the greater theirimpact on a society Hence the more robust the norms and rules thegreater their impact on a society and the less their robustness the greatertheir exibility and the less their effect on shaping a societyrsquos outcomesand performancesBecause norms rules and values are quite durable they play an impor-

tant role in shaping the history of societies thus contributing to a greatdeal of path-dependency Actors attempt to adjust to their contempo-rary environment but since they are products of the past the historicallegacy of norms rules and values inuences the decisions they makeAlthough actors have some capacity to alter the course of their historythey are constrained by their past and the degree to which they canmove beyond their past is limited As Lanzara (1998) Johnson (1992)and others have argued societal inertia is a basic feature of institutionsThey provide the basic stability necessary for change Degrees of historyare continuously reproduced by the way in which the inhabitants ofsocieties are socialized History matters but at critical points in historythere is punctuated equilibrium (Somit and Peterson 1992) During mostperiods of history there is considerable stability in the norms rules andvalues of a society but at critical moments norms rules and values canquickly and dramatically be redened At all times norms rules habitsconventions and values are inuencing each of the other components in

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

604

the institutional framework discussed below but these other institutionalcomponents feed back and can modify rules norms conventions etc(Murmann 1998)

Second level of analysis

The norms rules habits conventions and values of a society lead to thenext level of analysis ndash the institutional arrangements which are involvedin the coordination of various economic actors producers and suppliersof raw materials knowledge etc processors of raw materials infor-mation workers customers of raw materials nished productsinformation etc nanciers governmental and other types of regulatorsThese actors regularly engage in contests to resolve various economicproblems in virtually all sectors of society how are prices to be setWhat quantity of various products is to be produced How are stan-dards of various products processes etc to be set How is the qualityof products and processes to be determined How are various societalprocesses to be nanced In order to confront these problems and toaddress the conicting positions of economic actors as they address these problems societies develop various institutional (governance)arrangements for coordinating different actors These consist of marketsvarious types of hierarchies and networks associations the state commu-nities and clans (see Hollingsworth and Boyer 1997 Campbell et al1991 ch 1)Each of these particular kinds of coordinating mechanism has many

different types For example there are many types of states (eg theregulatory state developmental state authoritarian state welfare state)on which there is an extensive literature (Kim 1997) Similarly there aredifferent types of markets networks different kinds of associations etc(Boyer 1997 Hage and Alter 1997 Schneiberg and Hollingsworth 1990)When we do institutional analysis we must engage in congurativeanalysis recognizing that actors are not coordinated or governed by asingle type of institutional arrangement Some of the literature discussesindustrial sectors as though they were coordinated or governed by asingle institutional arrangement whereas in fact each sector of aneconomy is coordinated by a conguration of institutional arrangementsSome congurations coordinate actors in certain problem areas whileother congurations of institutional arrangements coordinate actors inaddressing other problems The types of congurations which are domi-nant in a society are somewhat stable and tend to persist over timewithin a society (Hodgson 1999)When one mode of coordination is dominant in a society it will

inuence the role which other coordinating modes will play Hence thestrong role of the state in the Soviet Union inuenced the role of markets

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

605

associations etc in the governance of the Soviet economy Similarly theprominence of particular modes of coordination in a society inuencesits style of innovationMuch of the literature on institutional arrangements (eg forms of eco-

nomic coordination) remains fragmented and unintegrated It is helpfulto array these various forms of coordination in a two-dimensional tax-onomy as in Figure 2 On the vertical dimension the economistrsquos viewof a self-interested agent is contrasted with a more sociological perspec-tive according to which social rules obligation and compliance shapehuman actions On the horizontal dimension there is another continuum(ie the distribution of power) At one extreme of the dimension onends many and relatively equal agents interacting (eg as in a well-organized spot market) At the other extreme inequality in power resultsin a hierarchical form of coordination which structures the interactionbetween principals and agents or between leaders and followers Wherea society falls along the distribution of power is inuenced by the rulesnorms values etc which are dominant in a society at a particular momentin timeInstitutional arrangements can be visualized on two dimensions the

nature of action motive on the one hand and the distribution of poweron the other Markets (cell 1) combine self-interest with horizontalcoordination transactors and they reect sensitivity to concerns aboutsupply and demand thus providing ex post an unintended equilibriumParadoxically the more pure and perfect the market competition thegreater the need for codied rules of the games for coordinating economictransactions Thus collective associations (cell 6) andor various formsof state intervention (cell 4) are required to develop and enforce rulesfor transacting partners (Schneiberg and Hollingsworth 1990 Streeckand Schmitter 1985a) This is an example of how the norms and rulesand the institutional arrangements of a society are intertwined asreected in a conguration of institutional or governance arrangementsAlong the horizontal axis actors can be joined in an organization or

a rm a hierarchy is the generic terms for this institutional arrangement(cell 2) Along the horizontal line one recognizes the difference betweentransactions in a market and transactions within a rm The well-knownworks of Coase (1960 1981) and Williamson (1975 1985) utilize theconcept of transaction costs in explaining the emergence of corporatehierarchiesThere are also various types of networks (cell 5) which exhibit mixes

of self-interest and social obligation with some actors being formallyindependent and equal Yet in some networks (the large rms and theirsubcontractors) there is unequal power and initiative Networks maycomprise all kinds of actors some consist only of rms but others includeassociations and the state (Hage and Alter 1997)

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

606

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

607

DISTRIBUTION OF POWER

VERTICALHORIZONTAL

1 MARKETS 2 HIERARCHY

6ASSOCIATIONS

5NETWORKS

4STATE

3COMMUNITIES

SELF INTEREST

SOCIAL OBLIGATION

ACTION MOTIVE

Figure 2 A general taxonomy of institutional arrangements

The vertical axis deals with action motives Toward the upper part ofFigure 2 actors are engaged in individualistically oriented behaviorwhereas toward the lower part actors are more engaged in collectivebehavior and strive to cope with problems of common interest Cell 3 ndashcommunities and clans ndash consists of institutional arrangements based ontrust reciprocity or obligation and thus not derived from the pure selshcomputation of pleasures and pains This is an unconventional form ofcoordination for most neoclassical economists (however see Arrow1974) but not for many anthropologists political scientists and sociolo-gists (Streeck and Schmitter 1985a Polanyi 1944 Gambetta 1988Fukuyama 1995 Sabel 1992 Putnam 1993)In the neoclassical paradigm theorists argue that actors engage in

forms of exchanges that best promote their individual interests If somestructural conditions are fullled (absence of increasing returns to scalethe reversibility of transactions absence of uncertainty and completecontingent markets with no collusion between actors) then the invis-ible hand theorem applies and market-type activity functions quite welland also provides the optimum for society thereby combining efciencyharmony and order However an excess of market activity may welllead to ruinous competition and excessive conict There is variation inthe extent to which ruinous competition occurs depending on the socialcontext in which transactions take place Thus it is important that webe sensitive to the institutional context in which transactions areembedded and that we understand the degree to which social bondsexist at both the micro and macro levels of analysis Micro bonds facil-itate exchanges in a society but at the societal level social bonds existat the level of the collective ndash in the community or region and amongmembers of racial religious and ethnic groups All other things beingequal the more powerful the social bonds among transacting partnersthe more economic competition is likely to be restrained Thus mosttransactions occur not simply in an impersonal calculative system ofautonomous actors unrestrained by social ties ndash as implied by theneoclassical paradigm ndash but in the context of social ties variation in the strength of which leads to variation in levels of trust and transac-tion costs (Etzioni 1988 211 Granovetter 1985 Streeck and Schmitter1985a Hollingsworth and Boyer 1997 Hodgson 1988 1999 Schneibergand Hollingsworth 1990)Another form of multilateral institutional arrangement is various

types of associations (cell 6) Unlike networks clans and communitiesassociations are more formal organizations Whereas markets corporatehierarchies and networks tend to coordinate economic activity amongdifferent types of actors (eg producers with suppliers capital withlabor) associations typically coordinate actors engaged in the same orsimilar kinds of activities Business associations and labor unions are

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

608

some of the most common forms of associations for coordinatingeconomic activity in capitalist economies (Schneiberg and Hollingsworth1990 Streeck and Schmitter 1985a)Finally there is the state (cell 4) which is an institutional arrangement

quite unlike any of the others It is the state that sanctions and regu-lates the various non-state coordinating mechanisms that is the ultimateenforcer of rules of the various mechanisms that denes and enforcesproperty rights and that manipulates scal and monetary policy Overallit is the state that inuences the total incentive system of a society(Lindberg and Campbell 1991 Johnson 1992 40) At the same time thestate may also be an economic actor by engaging directly in productionand exchange relationsThe choices of institutional arrangements in Figure 2 are constrained

by the social context in which they are embedded Depending on thenature of that embeddedness there is variation in the collective formsof governance some of which are specied in the lower part of thegure Some modes of coordination in the bottom and upper levels ofthe typology are often mixed together though one coordinating modeis likely to be more prominent than another But actors are often engagedin complex congurations involving several kinds of institutionalarrangementsEach of these various institutional arrangements has its own logic ndash

its own rules its own procedures for enforcing compliance its ownnorms and ideologies which help to reduce the costs of enforcementThese are summarized in Table 2 which provides further elaborationabout the various coordinating mechanisms one nds in almost everycapitalist society Table 2 lists the organizational arrangements and theirstructures rules of exchange and means of enforcing compliance asso-ciated with each type of coordinating mechanism While each type ofinstitutional arrangement has various positive features each institutionalarrangement also has failures and these are featured in Table 3 It is thecontest between those who support and those who oppose these variousinstitutional arrangements that tends to lead to transformations in insti-tutional arrangements over time (Campbell et al 1991 Campbell 1997)Again it is difcult to conceive of pure institutional arrangements ndash

either exclusively markets or exclusively hierarchies ndash since it has beenwell known since Adam Smithrsquos Wealth of Nations that the division oflabor within the rm cannot be disentangled from the existence andextent of the market Institutional arrangements have their own distinc-tive form of efciency (good static efciency for the market dynamicefciency for rms) and inefciency often leading to considerableinequality Neither networks nor communities are panaceas for economiccoordination without being congured with other types of institutionalarrangements Networks and communities may solve certain issues but

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

609

Table 2

Log

ics of ins

titution

al arran

gemen

ts

Coordination

Organizational

Rules of

Individual means

Collective means

mechanisms

structure

exchange

of compliance

of compliance

Marke

tsEasy en

try an

d exit

Voluntary sp

ot excha

nge

Leg

al enforcemen

t of

Norm of pr

ivate

Bila

teral ex

chan

ge or

control

prope

rty

marke

t site (W

all Street)

Reg

ulations

to

Leg

itim

acy of m

arke

t en

force contracts

men

tality

Com

mun

ities

Inform

al m

embe

rship

Voluntary exch

ange

based

So

cial norms an

d m

oral

Highly institutiona

lized

ev

olve

s ov

er lon

g period

on social solid

arity an

d

princ

iples im

pose

norm

s an

d rules requ

ire

of tim

ehigh

deg

ree of tru

stob

ligations

mem

bers to accept

Kno

wledg

e of others an

d

lsquocorpo

ratersquoo

bligations

recipr

ocity ov

er tim

e

Networks

Semi-form

al m

embe

rship

Voluntary exch

ange

ove

r a

Con

tractual bon

dsPersona

l relation

sBila

teral or m

ultilateral

time pe

riod

Resou

rce de

pend

ence

Tru

st built outside

the

ex

chan

geecon

omic arena

Associatio

nsFo

rmal m

embe

rship

Restricted to mem

bers

Self-interest

Some deg

ree of

Multilateral excha

nge

Emph

asis on insider

Rep

utation

effects

compu

lsion

outsider or we

they

Priva

te interest type

of

men

tality

gove

rnan

ce

Private

Com

plex organ

izations

Restricted to mem

bers

Rew

ards to ind

ividuals

Highly institutiona

lized

hierarch

ies

which

ten

d to becom

e ex

chan

ge based

on

Asymmetric po

wer threat

rules

bureau

cratic hierarchies

asym

metric po

wer

of san

ctions

Mem

bers soc

ializ

ed into

bureau

cratic rules

corp

orate cu

lture us

e of

sanc

tions

State

Pub

lic hierarchy

Unilateral action

Exit vo

ice (vote lob

bying)

Coe

rcion

De jure

and impo

sed

Indirect political an

d

loya

ltyNorms an

d pu

blic rules

mem

bership

econ

omic excha

nge

Table 3

Failu

res of coo

rdinating mecha

nism

s

Coordinating mechanisms

Type of failure

Associations

Private hierarchies

State

Market

Communities

Networks

Enforcem

ent

Usu

ally relies on

Might enh

ance

Needs co

ntrol

Needs an

internal

Needs

tru

st and

Need an external

the state as an

oppo

rtun

istic

extern

al to state

enforcem

ent

loya

lty often

en

forcem

ent

enforcer

beha

vior

bureau

cracy

authority

coming from

au

thority

Resem

bles

The

ideal o

f (ju

dge

s

Facilitates

outside (fam

ily May

facilitate

enforcem

ent

internal m

arke

ts

parliamen

t collu

sion

and

relig

ion

cartelization an

d mecha

nism

might hur

t marke

t) to

impe

rfect

ethn

icity)

mon

opoly

of cartels

incu

mbe

nt

correct state

compe

titio

nCom

patible with

worke

rsab

uses

variou

s types of

Lob

bies can

co

mpe

tition

captur

e pu

blic

interest goa

ls

Public goo

d

Useful for

Gov

erna

nce co

sts

Can

provide

Can

not provide

Can

internalize

Useful for

and

establishing

might exceed

public goo

ds

colle

ctive go

ods

some co

llective

enha

ncing qu

ality

externality

stan

dards an

d

the be

nets of

but ha

s or deal w

ith

good

s (qua

lity

and training bu

t qu

ality for

internal division

difc

ultie

s in

externalities

training

) bu

t no

t no

t ve

ry goo

d in

setting ru

les of

of la

bor

prov

iding them

Inad

equate

othe

rs (welfare

prov

iding for

compe

titio

n in

Slow

to react to

in precise

mon

itoring of

gene

ral p

ublic

societal gen

eral

the indu

stry

chan

ges in the

am

ounts

tech

nical ch

ange

go

ods)

welfare

Useful for

environm

ent

Might fail in

and inno

vation

Mem

bers tightly

Weak in the

prov

iding

induc

ing

integrated

into

prov

ision of

man

y go

ods

tech

nical

community

colle

ctive go

als

colle

ctively

chan

geha

ve limite

d

that in

dividua

lcapa

city for

mem

bers can

not

inno

vatio

nsprov

ide for

them

selves

Table 3

continue

d

Coordinating mechanisms

Type of failure

Associations

Private hierarchies

State

Market

Communities

Networks

Efcien

cyFa

cilitates

De

cien

t in

Can

be high

ly

Some ba

sic social So

me go

ods

Slow

to en

hanc

e co

operation an

d

coop

eration an

d

bureau

cratic

relation

s cann

ot

cann

ot be

efc

ienc

y an

d

X-efcien

cy but

X-efcien

cyan

d can

not

be provided

by

delivered

at

speed of

not allocative

easily deliver

pure m

arke

t su

fcien

tly

adap

tive

ness

efcien

cy

good

s at

mecha

nism

slow costs

except in

low cost

indus

tries whe

re

tech

nology

is

complex an

d rapidly cha

nging

Equity

Narrow

Excessive

Might enh

ance

Facilitates

Might lead to

Whe

n widely

encompa

ssing

multiplication

ineq

uality

ineq

uality in

retard

ed

dev

elop

ed in

to

associationa

l of con

trollers

(pow

er

income an

d dev

elop

men

tindus

trial

stru

ctures lead

(fru

stration

and

an

d privilege

)wealth

districts

to in

come

ineq

uality)

netw

orks m

ay

equa

lity

facilitate greater

equality an

d

inco

me

distribution

whe

n wea

kly

dev

elop

ed

netw

orks ten

d to

increase soc

ial

ineq

uality

Note For allo

cativ

e efc

ienc

y an

d X

-efcien

cy see Leibe

nstein (19

66 19

76)

they raise other no less severe problems It is important to recognizethe imperfection of any single institutional arrangement in order tocomprehend the origin and transformation of any other institutionalarrangement (Hodgson 1999 ch 3)As suggested above each institutional arrangement is congured with

other institutional arrangements Usually one particular institutionalarrangement is more dominant in a particular conguration but becauseeach type of institutional arrangement has its own strengths andweaknesses there is no simple structural logic in the governance or coor-dination of a society Each institutional arrangement constrains the otherbut the inherent tension among the various institutional arrangementswithin a conguration contributes to changes in the governance orcoordination of a society The routines and logics of each of these cong-urations of institutional arrangements provide constraints and incentivesfor actors The way that actors perceive the incentives and constraintsof these governance congurations leads to a particular market logicand it is the specic market logic of a society which inuences its speciccapacities and weaknesses The inherent strain among the different logicsin a conguration of coordination helps to provide the exibility for asociety to adapt to new circumstancesIf there were a society with pure markets or a society coordinated

only by the state there would be too much rigidity and too little diver-sity to cope with the vast uncertainty in the global environment A societywith very little diversity in its coordination mechanisms would havelittle capacity to adapt to new circumstances Congurations with consid-erable diversity of institutional arrangements provide for a certainamount of incoherence in governance but they also provide for thecapacity to adapt to new circumstances While the Soviet Union wasdominated by the state as a coordinating mechanism there were alwaysfunctioning markets in the system Moreover the feudal society of thetwelfth and thirteenth centuries consisted not only of hierarchical rela-tionships between serfs and masters but also of urban guilds clericalhierarchies and markets (Johnson 1992 38) In sum the robustness ofinstitutions often depends on multiple and diverse principles and logicsof action on the inconsistency of principles and procedures on patternedforms of disorder (Lanzara 1998 Orren and Skowronek 1991 320 329)

Third level of analysis

The next level consists of the institutional sectors of a society The rulesnorms and values of a society inuence the array of institutional arrange-ments and levels one and two inuence the nature of and the relation-ships among various institutional sectors and all three levels areintricately linked together to form a social system of production Together

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

613

all of these components inuence the performance of economic sectorswithin a society as well as the performance of the total society A socialsystem of production is the way that a societyrsquos institutions (see rstlevel of analysis above) its institutional arrangements (see second levelof analysis above) and its institutional sectors are integrated into a socialcongurationAn institutional sector includes all organizations in a society which

supply a given service or product along with their associated focal orga-nizations (eg major suppliers funding sources regulators and so forth(Scott et al 1994 108 117 Campbell et al 1991 Hollingsworth et al1994) Institutional sectors include but are not limited to the societyrsquossystem of education system of research business system nancialmarkets legal system and the state The structure of the nancial marketssystem of training and education industrial relations system system ofresearch and state and legal systems are distinct and idiosyncratic ineach society In short they are system specicAll of these institutional sectors tend to cohere with each other

although they vary in the degree to which they are tightly coupled intoa full-edged system While each of the institutional sectors has someautonomy and may have some goals that are contradictory to the goalsof other institutional sectors with which it is linked an institutional logicin each society leads institutions to coalesce into a complex social con-guration (Hollingsworth 1991) This occurs because the institutionalsectors are embedded in a culture in which their logics are symbolicallygrounded organizationally structured technically and materially con-strained and politically defended The conguration of institutionalsectors usually exhibits some degree of adaptability to new challengesbut continues to evolve within an existing style But under new circum-stances or unprecedented disturbances these institutional congurationsare exposed to sharp historical limits as to what they may or may notdo (Schumpeter 1983 David 1988 Arthur 1988a 1988b Haringkanssen andLundgren 1997)Why do all of these different institutions coalesce into a complex

social conguration which is labeled here as a social system of pro-duction The literature suggests two contrasting interpretations Part ofthe answer ndash indeed a controversial one ndash is that these institutionalsectors are functionally determined by the requirements of the practiceof capitalism in each time and place (Habermas 1975) Another expla-nation emphasizes the genesis of the actual conguration via a trial and error process according to which the survival of rms regions orcountries is the outcome of complex evolutionary mechanisms (Maynard-Smith 1982 Nelson and Winter 1982) However the problem is evenmore complex Markets and other mechanisms for coordinating rela-tionships among economic actors place constraints on the means and

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

614

ends of economic activity to be achieved in any society The othercoordinating mechanisms include different kinds of hierarchies varioustypes of networks and associations ndash trade unions employers and busi-ness associations (Hollingsworth and Lindberg 1985 Campbell et al1991) The logic of these various institutional arrangements providesactors with vocabularies and logics for pursuing their goals for deningwhat is valued and for shaping the norms and rules by which theyabide (see Tables 2 and 3) In short in contrast to the logic of theneoclassical paradigm the argument here is that the dominant type ofinstitutional arrangements places severe constraints on the denition of needs preferences and choices of economic actors Whereas the neo-classical paradigm assumes that individuals and rms are sovereign thisarticle is based on the assumption that rms are inuenced by the hold that the institutional congurations making up a social system ofproduction have on individual decision making (Campbell et al 1991Etzioni 1988 Streeck and Schmitter 1985a Hollingsworth et al 1994Hollingsworth and Boyer 1997 Magnusson and Ottosson 1997 North1990 Hodgson 1999)Standard neoclassical economic theory has tended to downplay the

role of production and consequently of rms Even transaction cost theo-rists who are concerned with analyzing the rm as a coordinatingmechanism have been relatively unconcerned with the various compo-nents of a social system of production Indeed as long as there waswidespread optimism about the efcacy of Keynesian economics therewas relatively little concern among neoclassical economists with thesupply side of the economy Even in the opinion of most Keynesians agroup of experts should ideally be able to shape the size of aggregatedemand while the supply side of the economy would be left to the twominimalist institutions of neoclassical economics ndash markets and manage-rial hierarchies For several decades however it has become increasinglyobvious that some of the most competitive and successful patterns ofindustrial output and industrial production in capitalist economies donot derive from the neoclassical prescription of unregulated markets andcorporate hierarchies complemented by a neoliberal democratic stateIndeed it is now well understood that certain highly successful produc-tion patterns require for their emergence and survival institutionalarrangements and environments the very opposite of the prescriptionsfound in the neoclassical paradigm (see especially Streeck 1992 but alsoHollingsworth and Streeck 1994) Thus social scientists have come tounderstand that if they are to comprehend the behavior and performanceof contemporary economies concerns about social systems of produc-tion must be brought into the pictureBecause production involves more than technology a number of social

scientists have an increasing concern with social systems of production

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

615

The same equipment is frequently operated quite differently in the samesectors in different countries even when rms are competing in the same market (Maurice et al 1980 Sorge and Streeck 1988 Sorge 1989Hollingsworth et al 1994) Variations in production and process tech-nologies are inuenced partly by variations in the social environmentsin which they are embedded In other words rms are embedded intocomplex environments which among other things place constraints ontheir behavior Thus a social system of production is of major impor-tance in understanding the behavior and performance of an economyHow the state and other institutional arrangements (eg marketsnetworks private hierarchies associations) coalesce and are related toparticular social systems of production are important determinants ofeconomic performanceThe congurations of institutional arrangements that coordinate or

govern the behavior of actors in one society and its structure of specicinstitutional sectors cannot easily be transferred to another society forthey are embedded into a social system of production that is societallydistinct (Hollingsworth 1997) Societies borrow selected principles offoreign management styles and work practices but the effectiveness ofsuch borrowing is generally limited Economic behavior and performanceare shaped by the entire social system of production in which actors areembedded and not simply by specic principles of particular manage-ment styles and work practices Moreover a societyrsquos social system ofproduction tends to limit the kind of goods that it can produce and withwhich it can compete successfully in international marketsSome scholars (Kenney and Florida 1993 Oliver and Wilkinson 1988)

have assumed that the diffusion of particular forms of management stylesand work practices across societies could lead to system convergenceReferring to the Japanese production system Kenney and Florida haveargued that it consists of organizational practices whose fundamentallsquogenetic codersquo can successfully be inserted into another society and beginsuccessfully to reproduce its behavior in the new environment (1993 8)Their position is in the intellectual tradition of Antonio Gramsci (1971)who contended decades ago that the American system of mass produc-tion would most certainly diffuse to Europe over timeHowever the argument of this article is that even though British

French and American rms may adopt certain aspects of Japanesemanagement styles (eg just-in-time production self-managing teamsquality circles the use of statistical process controls) or some variant ofthe German vocational system their social systems of production willnot converge Moreover the overall conguration of a societyrsquos socialsystem of production inuences its sectoral and overall national economicperformance This helps to explain why societies have different systemsof innovation and why they vary in the clusters of industries in which

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

616

they are highly competitive in the international markets (Nelson 19931999 Hage and Hollingsworth 2000 Berger and Dore 1996)In order to understand how and why a societyrsquos economy performs

as it does it is necessary to understand its entire social system of produc-tion If a society is to modify substantially the performance of itsinstitutional sectors and its entire economy it cannot adapt only someof the management and work practices of its foreign competitors Ratherit must alter its entire social system of production Because a societyrsquosmodes of economic governance and coordination and its institutionalsectors develop according to particular logics and are system specicthere are serious limitations to the extent to which a society may mimicthe institutions (eg the rules and norms) institutional arrangements andinstitutional sectors of other societiesIn the history of modern societies there are logics by which institu-

tions coalesce into a social system of production (Hollingsworth 1991)Though institutions are constantly changing there are sharp limits tothe type and the direction of change that any particular institution canundergo because of its linkages with institutional arrangements and insti-tutional sectors Thus a societyrsquos business rms educational systemnancial markets industrial relations system etc can engage in seriousrestructuring only if most of the other institutional sectors also changeIn social systems of production there are pressures toward consistencyin the norms rules and values across institutional sectors though in anycomplex society social systems are obviously imperfectly integratedIndeed the degree to which the institutional norms and rules makingup a social system of production are loosely or tightly coupled is a vari-able of considerable importance In general the institutional sectorsmaking up a social system of production are interdependent and changesin one generally result in changes in the others Since each institutionallevel is dependent on the others for various types of resources there isinterdependence among the differing institutional spheres Moreovereach society has its norms moral principles rules and laws and recipesfor action as well as its own idiosyncratic customs traditions and prin-ciples of justice (Burns and Flam 1987)There are also other inherent obstacles to convergence among social

systems of production for where a system is at any one point in timeis inuenced by its initial state Systems with quite different initial statesare unlikely to converge with one anotherrsquos institutional practicesExisting institutional arrangements block certain institutional innova-tions and facilitate others (Roland 1990) Thus the institutions makingup a social system of production provide continuity even though insti-tutional arrangements are always changing but with a particular logicWhile Williamson (1975 1985) suggests that actors tend to choose theinstitutional arrangements which are most efcient North (1990) is

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

617

much closer to the mark in his argument that most societal institutionalarrangements exist as a result of custom and habit and are thereforeinefcient At any moment in time the world often appears to its actorsas very complex and uncertain For this reason actors often engage incontradictory forms of behavior pursuing different strategies as hedgesagainst a very uncertain world (Lanzara 1998) And their hedging andcontradictory forms of behavior may lead in somewhat different soci-etal directions all constrained by the institutional fabric within whichthe actors are embedded This kind of contradictory behavior occurs inpart because of the contradictions inherent in the dominant congura-tions of institutional arrangements with their contradictory logics (seelsquoSecond level of analysisrsquo above)Despite the emphasis on the logic of institutional continuity this is

not an argument that systems change along some predetermined pathThere are critical turning points in the history of highly industrializedsocieties but with the choices limited by the existing institutional terrainBeing path-dependent a social system of production continues along aparticular logic until or unless a fundamental societal crisis intervenes(Milgrom et al 1991 Krugman 1991 Durlauf 1991 Hollingsworth 1991Pred 1966 David 1988 Hodgson 1998)At this point it is important to confront the question of how social

systems of production evolve And this question gets to the heart of theproblem of building complex societal structures Certainly social systemsof production did not emerge from some process of social engineeringMoreover the various component parts of each social system of produc-tion have often not been designed to be part of a social system ofproduction The component parts of each social system of productionhave emerged more often than not as unintended by-products of goalswhich various actors had in mind at earlier moments in time It is usuallythe case that actors at time lsquotrsquo fail to comprehend the long-term conse-quences of their actions Wolfgang Streeck the author of several brilliantpapers on social systems of production argues that institutional sectordesigns created for one purpose generally address that goal but overtime those designs have quite unintended consequences (Streeck 1997a1997b Wright 1998) Describing the conguration of the German socialsystem of production of the 1970s and 1980s Streeck points out that itwas the unintended by-product of multiple points in history Someelements were pre-Wilhelmian others were introduced by the Alliedpowers after 1945 and others emerged during the years of the GermanFederal Republic All component parts of the German social system ofproduction

were and continue to be changing for their own reasons as wellas in reaction to each other and certainly there can be no presump-tion of a pre-established t between them even though one might

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

618

want to allow for some reinforcement effects of the modelrsquos histor-ically contingent social and economic success

(Streeck 1997b 54)

In sum the emergence of social systems of production is a long-termevolutionary process with each part interacting with its environment andresulting in a congurative whole but with no previous design by eithera single actor or a collectivity of actors As suggested above actors do not recreate their world anew At any moment in history the totalinstitutional complex of a society necessarily contains the resources and legacies of its past Preexisting institutional complexes are nevercompletely wiped out Societies consist of multiple layers of history withtheir diverse logics of action However the cumulative effect of smallperipheral changes in altering particular institutional sectors and insti-tutional arrangements can be substantial (Murmann 1998 Lanzara 1998)lsquoInstitution building is affected more by the ways in which people havecodied the past than by how they envision the future There are manymore potential resources in pre-existing institutional arrangements thanit is usually assumed or suspectedrsquo (Lanzara 1998 30)

Fourth level of analysis

The next level of institutional analysis organizational structures is some-what more controversial As suggested above Douglass North (1990)draws a sharp distinction between institutions and organizations Morerecently many organizational theorists (Powell 1991 Powell andDiMaggio 1991 DiMaggio and Powell 1983 Baum and Oliver 1992Kondra and Hinings 1998 Townley 1997) argue that institutional rulesnorms and conventions unfold in tandem with organizational structuresand this is my positionThe literature which focuses on how institutions inuence organiza-

tions is quite different from two other theoretical literatures which alsoare concerned with how organizational environments shape the behaviorof organizations There are resource dependency theorists (Pfeffer andSalancik 1978) who emphasize the role of environmental resources inshaping organizations while Hannan and Freeman (1977 1984) and otherpopulation ecologists emphasize the survival of organizations givencertain kinds of organizational conditions (Orru et al 1991)The institutionalist perspective as implied above emphasizes the

normative environment in which organizations are embedded It is a perspective which focuses on the way in which organizations in their behavior tend to conform to the institutional rules and normswhich are dominant in the organizational environment However allthree perspectives ndash the institutionalist the population ecology and

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

619

the resource dependency ndash emphasize how the environment inuencesorganizations and how organizations which are subject to the sameenvironment tend to converge in their behavior to have what DiMaggioand Powell have labeled organizational isomorphism (DiMaggio andPowell 1983)There has emerged a vast literature which demonstrates that within

each society there are scal political judicial and other regulatory normswhich limit and shape the culture and structure of organizationalbehavior The normative institutional environment of organizations limits the options of what organizations do in a particular society and inuences the patterns of ownership relations with suppliers andcustomers In short it is the normative environment of organizationswhich denes within a particular society what is socially acceptablebehavior for organizations (Hamilton and Biggart 1988 Hollingsworthand Hollingsworth 2000 Meyer and Rowan 1991 Zucker 1987 1988Orru et al 1991 Townley 1997)Thus far most of the analyses on isomorphic pressures operating on

organizations have been cross-sectional in nature However the entireconcept of isomorphism implies that there are strong environmental pres-sures exerted on organizations and in order to observe this phenomenonit is important that we have longitudinal studies which assess howchanges in an organizationrsquos institutional environment inuence changesin the structure and culture of organizations Moreover the expectationhere is that when one analyzes the patterns of organizational changeover time the isomorphism will be holistic ndash that is within a particularsociety there will be a tendency for the internal structure of particulartypes of organizations to converge in large part because of the pres-sures to conform to the changes in the external norms rules and valuesA close reading of Aoki (1990) and Slack and Hinings (1994) nds thatchanges in the national institutional environment of organizations haveinuenced changes in the internal structure of organizations within aspecic industry of a particular societyHowever it is inappropriate to view organizations as changing only

in relation to exogenous environmental change There is a kind of co-evolution which occurs between organizations and their institutionalenvironments Co-evolution implies nonlinear feedback between orga-nizations and their institutional environments This is of course quitedifferent from the more common way of studying organizations bymodeling relations between independent and dependent variables Ofcourse the more common type of analysis is quite appropriate whenthere are simple relationships which do not involve complex feedbackprocesses But in the strategy for institutional analysis which is proposedherein it is less useful to separate independent from dependent vari-ables and more useful to understand the interacting and co-evolutionary

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

620

processes among all of the levels of institutional analysis discussed inthe various sections above (Baum and Singh 1994a 379ndash80 Mowery andNelson 1999 Murmann 1998)The logic of the perspective herein suggests that within every

society there is variation in the structure and culture of business rmsuniversities and other complex organizations (Kondra and Hinings1998) However this variation takes place within parameters which aresystem specic For example every German rm and every Germanuniversity is different from every other German university and rm but there are a set of cultural and structural characteristics which distin-guish German rms and universities from those in America Similarlyeach Japanese rm and university is unique yet they have a set ofcultural and structural characteristics which are system specic andwhich differentiate them from business rms and universities in other countriesIn all societies each organization has its own distinctive organizational

rules norms and conventions which are subordinate to the meta norms and rules of the larger society within which they are embeddedHowever the strength of the institutional environment within whichorganizations are embedded varies from society to society Some societies have multiple institutional environments and there is hetero-geneity in terms of what constitutes appropriate behavior oforganizations In such societies there can be conict over whatinstitutional logics should regulate specic organizational functions(Friedland and Alford 1991 Townley 1997 262ndash3) In those societies inwhich the institutional norms habits and rules are most developed and in which the institutional pressures to conform are greatest thereis less variation in the structure and culture of business rms and variouskinds of research organizations In such societies the connectednessbetween research organizations and their institutional environments is sufciently strong for organizations to have low autonomy to pursue independent strategies and goals in these societies there is a great deal of organizational isomorphism Conversely the weaker the institutional environment in which research organizations areembedded the greater the variation in the structure and culture of business rms and research organizations Moreover where the institu-tional environments are more weakly developed organizations havegreater autonomy and exibility to respond to the development of new knowledge and to be highly innovative Hence it is in those societies where the institutional environments are most developed and most rigid and when there is less organizational autonomy and exibility that fewer radical innovations in basic and applied science as well as in totally new products and industrial sectors havebeen made

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

621

Fifth level of analysis

We come next to the outputs and performance of the various institutionalcomponents of a society It is at this level that institutional componentsare more pragmatic and exible For example in the legal sector thereare specic statutes and court rulings in the state sector there are specicpolicies in the business sector there are new products new technolo-gies and market strategies It is at this nal level of analysis thatinstitutional spheres are most open and susceptible to change and cross-national mimicry is easiest and most commonThrough the outputs of the society we can obtain some assessment as

to how well a society is performing Moreover we can assess how inno-vative it is how egalitarian it is in the distribution of its resources howegalitarian it is in terms of levels of health education etc Just as it iscomplicated to measure how egalitarian a society is similarly it is dif-cult to assess how innovative it is For example societies vary in makingincremental and radical breakthroughs in basic and applied science in developing totally new products and new kinds of organizations inincremental and process innovations in existing products and organiza-tions in developing new and different forms of marketing ndash bothdomestically and globally (Hage and Hollingsworth 2000) Of coursesocieties vary in their rates of economic growth in their rates of economicproductivity in their quality of life (eg rates of crime life expectancyetc) All of these performance criteria feed back and inuence each ofthe levels of institutional analysis discussed above ndash rules norms valuesetc institutional arrangements institutional sectors the structure andform of organizations Moreover performance measures may inuenceeach other For example if a society has very low rates of growth itmay not be very innovative in some of the types of innovative activitiesmentioned above Good or poor performances of certain kinds inuenceother performance indicators (Hollingsworth et al 1990)Different institutional arrangements and different social systems of

production result in different types of economic performance Hence solong as societies have different social systems of production there areserious constraints on the degree to which they can converge in theirinnovative styles Different social systems of production tend to maxi-mize in a more or less explicit manner different performance criteriausually mixing considerations about static and dynamic efciency protsecurity social peace and economic andor political power In short incontrast to the implications of neoclassical economic theory in real-worldeconomies there are no universal standards all economically rationalactors attempt to maximize Economic history provides numerous exam-ples of how a great variety of principles of rationality are implementedin different societies

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

622

A critical question in institutional analysis is whether an existing socialsystem of production which supports a set of routines for a particularkind of technology and industry can shift from old practices to newones There are numerous historical accounts of how the social systemof production in a particular society which worked so well for a numberof years could not adapt to new technologies (Lazonick 1990 Veblen1915 Nelson 1994) Schumpeter (1983) and Freeman (1991) have devel-oped the idea of a techno-economic paradigm which suggests thatdifferent technologies require specic forms of organizational arrange-ments ndash or what I called social systems of production Because oflsquoinstitutional inertiarsquo the social system of production of a society maynot be able to adapt to a new techno-economic paradigm (Nelson 199458 Lanzara 1998)Whether a social system of production can sustain its performance

standards depends not only on its intrinsic economic rationality but alsoon where it ts into a larger system If a particular social system ofproduction is immune from the innovativeness and competition of analternative system survival can be long lasting But if different socialsystems of production with diverging criteria of good economic perfor-mance meet in the world arena the arbitrariness of nationally imposedconstructed performance standards may be superseded by alternativeperformance criteria as a result of international competitivenessThe world economy is also socially constructed just as are national

economies Even if different social systems of production are competingin the international arena it is not always possible to determine whichis more competitively effective at any moment in time Hegemonicnation-states in the short run shape the rules of trade to favor their indus-trial sectors and rms But the history of hegemonic powers suggeststhat in the longer run such social systems of production sustainedlargely by military and political power eventually give way to moreinnovative and competitive social systems of production In our ownday as nation-states are increasingly integrated into a world economyeven hegemonic powers lose their innovative and competitive advan-tage and their share of world output decreases Such a country mayattempt to restructure its institutional arrangements and to readjust itsperformance preferences But to restructure its system of innovationgenerally calls for a major redistribution of power within a societyLargely for this reason societies have historically had limited capacityto reconstruct a system of innovation in the image of their major competi-tors Each countryrsquos social system of production is a conguration of ahost of norms rules and values as well as of institutional arrangementsEach system is constantly changing and is open to inuence from othersystems And indeed many technologies and practices diffuse from onesociety to another but a societyrsquos capacity to be innovative is constrained

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

623

by the existing social system of production Thus the same technologymay exist in numerous countries but how it is employed and how itinuences societal outcomes and performances varies from one institu-tional conguration or society to another (Hollingsworth 1998)

PERSPECTIVES ON INSTITUTIONAL CHANGE

A fundamental problem which remains unresolved in institutionalanalysis involves the nature of institutional change There is much confu-sion and miscommunication in the social sciences about institutionalchange in large part because we do not even know how to measure therate of institutional change Of course we know that there are bothexternal and internal focuses for change and we have had a good bitof scholarship on this subject What is not clearly understood is thatinstitutions and institutional arrangements within societies are histori-cally rooted that there is a great deal of path-dependency to the waythat institutions evolve Moreover the more intricately linked each sectoris with each other sector and with a societyrsquos rules and norms the lesschoice actors have to devise new institutions and institutional arrange-ments On the other hand if a sector or actor is too isolated it may betoo weak within a society to bring about any kind of effective changeThese considerations are very critical to the problem of how muchfreedom actors have to create new institutions afresh to what extentinstitutions diffuse from one society to another (Murmann 1998 Lanzara1998) This is a fundamental question for students of institutionalanalysis as we attempt to understand how the different institutionallevels are linked togetherThere are a variety of reasons why there is confusion in studying insti-

tutional change There is disagreement over the extent to which actorshave the freedom to build new institutional arrangements and if theydo to what extent the new arrangements and organizational patternsmay depart from past practices Among some institutional economists(Williamson 1985) there is a basic assumption that actors build efcientinstitutions by pursuing their self-interest and by promoting theirstrategic goals Moreover there is the assumption that a mixing of fullrationality with market competition tends to produce all of the optimalinstitutions that are needed in order to coordinate a complex capitalisteconomy (Hollingsworth and Boyer 1997 ch 14 Hodgson 1999) Incontrast the argument developed here is much more complexSocial institutions are historically rooted and there is a great deal of

path-dependency in the way that various institutional componentsevolve The shape of institutional congurations at any moment limitsthe type of options for change Because there is a great deal of institu-tional inertia radical change in the institutional components of a society

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

624

is uncommon (Lanzara 1998) Recognizing the constraints existinginstitutions exert on actors we note that existing congurations of insti-tutional arrangements institutional sectors and organizational propertiescan limit but also inuence the degree to which norms and rules canchange Many scholars (North 1990 Ostrom 1990) have emphasized therole of norms rules ideologies and values in limiting the ability of actorsto develop new institutional arrangements which radically depart fromexisting ones However there is a two-way process of interaction aprocess of co-evolution (Baum and Singh 1994a Nelson 1994) Thus asFriedland and Alford (1991 244) argue actors develop new institutionalarrangements by recognizing organizational failures and low perfor-mance but these new arrangements do not necessarily change a societyrsquosnorms rules habits and its underlying value systemCampbell (1997) advances our understanding of how existing institu-

tions enable actors to construct radically new institutions He recognizesthat most institutions and institutional arrangements embody a degreeof inertia from which it is difcult for actors to depart Moreover thedifferential power relations among actors make it unlikely that lesspowerful actors can change the existing power structure of a societyMost change in the institutional components of a society evolves througha process of constrained selection which reects to a considerable degreethe existing arrangements and power relations This becomes the inter-pretative frame for social actors with institutionalized scripts and ritualswhich tend to be taken for granted and appear to be quite ordinary andnatural This institutionalized scripting of the social world is an extremelyimportant source of social stability and inertiaCampbell has developed a very fruitful explanation of change in terms

of the various institutional components outlined here He distinguishesbetween incremental and radical institutional change and argues thatradical institutional change occurs when social actors with widelydiffering norms cultural scripts and rituals for action engage in intenseinteraction with each other (see also Knight 1992) The more diverse theinteractions the greater the potential for institutional change He arguesthat as shifts in the composition of interaction among social actors occurchanges in the interpretation of problems and interests will follow Ifthere are only minor changes in the extent to which diverse actorsinteract there will be only modest institutional change But the morefundamental the changes in the interaction of diverse social actors thegreater the change in the way that actors interpret their world And thegreater the change in the way that actors interpret their world the morelikely that radical changes in various institutional components willfollow In other words diverse decision making tends to create a widerrange of interpretative frameworks of the social world than is likely tobe the case if social actors operate in isolation

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

625

This perspective of diverse interactions among social actors facilitatesour ability to move beyond the traditional view that existing institutionsconstrain the range of institutional alternatives which actors face Theolder view has tended to suggest that actors engage in institutionalchange by extending existing institutional principles habits and conven-tions Specifying the conditions which facilitate fundamental institutionalchange permits a much richer perspective for institutional analysisCampbellrsquos perspective about the conditions under which fundamentalor radical change in institutional components occurs is consistent withmy own research about the conditions under which major discoveriesor fundamental new knowledge occur the more scholars with diversebackgrounds interact with intensity and frequency the greater the like-lihood that they will develop new and alternative ways of thinking(Hollingsworth and Hollingsworth 2000)These views of institutional change are also consistent with the recent

work of Nelson and his colleagues (Mowery and Nelson 1999 Murmann1998) They report the research on a variety of industrial sectors in anumber of different countries revealing that networks of actors in partic-ular industries business and professional associations universityresearch departments and institutes are frequently able to use their collec-tive power to modify a societyrsquos property rights research system andgovernmental policies It is the intense and frequent interaction of thesediverse complex sets of actors ndash sometimes constituting a technologysystem ndash that often leads to incremental changes in a societyrsquos total socialsystem of production It is this kind of collective action that brings aboutorganizational change change in institutional sectors institutionalarrangements and even in the rules and norms of a society And if weare to understand how the style of innovativeness of a society changeswe must rst understand the changes in its institutional makeup It ishighly unlikely that a society will develop a new style of innovativenesswithout changing its institutional structure

FIRMS THEIR INSTITUTIONAL ENVIRONMENT ANDNATIONAL SYSTEMS OF INNOVATION

Variation in the innovative style of societies is shaped by their insti-tutional makeup ndash the various levels which are discussed aboveinstitutions institutional arrangements the structure and cohesion of institutional sectors which constitute a societyrsquos social system of pro-duction and the structure and culture of its organizations ndash especiallyits business rms and other research organizations By undertaking aninstitutional analysis of a society one can begin to understand in whatkinds of organizations the production of specic kinds of knowledgetakes place and how this is linked to particular kinds of innovativeness

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

626

With such a perspective one can gain a rich understanding of why somesocieties excel in the production of radical breakthroughs in basic andapplied science and in developing radically new products and why othersocieties excel in more incremental innovationsReecting on these issues one cannot help being impressed with the

contrasts between the innovative styles of German and American orga-nizations For example Germany has continued to be very successful in making incremental innovations in many industries in which it wasvery competitive well before World War II paper printing materialsmachinery electrotechnical products motor vehicles chemicals textileyarns and fabrics But the Germans have been much less successful indeveloping totally new products or related innovations in biotechnologyelectronics telecommunications aircraft as well as other newer indus-trial areas In contrast the Americans have been very innovative in these and other industries with products having very short lives andwith technologies which change very rapidly and are very complex(Hollingsworth 1991 Soskice 1997 1999)The importance of institutions and innovativeness for international

competitiveness has led to the recognition that there are national systemsof innovation (Lundvall 1992 Nelson 1993 Edquist 1997 Hage andHollingsworth 2000) But even before the recent studies of nationalsystems of innovation Landes (1969) was a proponent of these viewswhen he made the argument that Germany had developed a nationalsystem of innovation by building on its education and scientic researchsystems which were lacking in Britain and elsewhere However the concept of a national system of innovation immediately poses theproblem of what kind of innovations should be consideredDavid Soskice an economist at the Wissenschaftszentrum Berlin has

addressed a set of problems somewhat complementary to the perspec-tive raised here His work (Soskice 1990 1997 1999) suggests that weare sensitive to the institutional environment in which the followingactors are embedded the employees of rms the owners and nanciersof rms the competitors of rms and nally their collaborators ndash aperspective complementary to Hollingsworthrsquos work on social systemsof production (Hollingsworth 1997 1998) When one reects on the insti-tutional contextsocial system of production within which these actorsare embedded one becomes aware that those countries which have some-what inexible labor markets external to rms but a business system in which rms have long-term commitments to their employees arestrongly associated with incremental but hardly at all with radical inno-vations In the rms of such countries (eg Germany and Japan) thereis much more consensus decision making than in societies where a lsquohireand rersquo set of practices is quite pervasive And this kind of consensusdecision making ndash while highly conducive to incremental innovativeness

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

627

ndash limits the capacity for radical innovations to occur This kind of lsquolongtermismrsquo set of norms and rules occurs not only in the employmentpractices of incrementally innovative rms but also in their nancingand ownership patterns (Hollingsworth 1997) During the past thirtyyears these were characteristics of many rms in Germany and JapanSignicantly in both of these countries venture capital markets werenot highly developed nancing tended to be based more on long-termbank loans than on the equity markets and the relationships of rmstended to be relatively stable And these stable institutional arrange-ments also contributed to a high degree of innovativeness ndash butincremental rather than radical innovationsMeantime we should also pursue Soskicersquos ideas (1999) about how

rmsrsquo relationships with their competitors and collaborators shape theirstyles of innovativeness In those advanced industrial societies in whichbusiness rms are highly mobilized into business associations theretends to be a high degree of standardized quality controls over technicalnorms and a high degree of reliance on legal sanctions as monitoringinstruments vis-agrave-vis association members As a result of this high degreeof integration of rms into associations member rms are more likelyto be associated with incremental innovations than is likely to be the case with rms in those societies where associations are less welldeveloped In the latter type of society radical innovations are morelikely to occur In those advanced industrial societies where rms havelong-term stable relationships with collaborators in product design andproduction the style of innovation is likely to be incremental And it is in those societies in which there is a weaker tradition of strongcollaboration among rms that radical innovations are more likely tooccurObviously even in a country where there may be a great deal of

radical innovativeness most innovations will be of an incremental typefor there are simply fewer radical innovations But Germany and therest of Europe not only have had relatively few radical innovations incomparison with the United States but they also have lagged behindthe USA in developing relatively totally new industries Thus there hasbeen much less innovative activity in Europe in high-technology andnewly emerging industries than in the USAIn complementary research Audretsch and Feldman (1995) have

demonstrated that the USA has been more successful in spawning rmsengaged in the early stages of the industry life cycle but relative toGermany had an environment unfavorable to innovative activity byestablished rms especially in more mature industries German rmshave tended to be associated with a routinized technological regime ndashwhere established enterprises have had strong incentives to engage inincremental innovations but new rms have tended to have fundamental

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

628

innovative disadvantages relative to the Americans (see Audretsch 1995Winter 1982 Casper 2000)On both sides of the Atlantic universities engage in basic and applied

research But signicantly the institutional environment in which univer-sities are embedded inuences the degree to which the research is likelyto represent major breakthroughs or is likely to have more modestimplications Recent research in biomedical science demonstrates that inthose societies in which university professors andor graduate studentshave the following characteristics there are likely to be relatively majorresearch breakthroughs graduate students have a long training periodhigh dependency on a major professor low autonomy in dening thenature of a research project restricted job options and a university envi-ronment where there is a very modest or nonexistent entrepreneurialculture For example in Germany ndash a country with relatively few radicalbreakthroughs ndash the scholar who nishes a habilitation will have spentmany years on a research project before qualifying for a permanent acad-emic position will have been in a high-dependency position with a seniorprofessor will have had limited autonomy to dene a research projectand will be constrained from pursuing a high-risk research project Theyoung German will have an extraordinarily high level of training butwill have faced a number of disincentives to undertake high-risk researchAnd as soon as heshe has completed the habilitation and becomes aprofessor administrative responsibilities constitute another disincentivefrom engaging in high-risk or long-term research projects Moreover theGerman professor is a civil servant with a more stable and secure incomethan hisher American counterpart In a very different institutionalenvironment American professors in large research universities tend tobecome much more entrepreneurial Today many American professorsbehave as though they were operating their own rms within univer-sities and thus devote considerable time to generating funds for researchprojects assistants secretaries supplies etc And the variation in thisentrepreneurial spirit across countries is associated with variation in thetaking of high-risk research strategies and the making of major researchbreakthroughsIn sum societies vary in terms of the institutional environment in

which their organizations are embedded and this variation inuencesthe degree to which they have industrial activities in the early or latestages of the industry life cycle successful new or mature industriesand radical vs incremental innovations And it is this variation whichis encouraging rms in countries with relatively few new industries and with few or no radical innovations to invest in those countries with many new industries and numerous radical innovations ndash as astrategy to be able to transfer new technologies from the latter institu-tional environment to the former

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

629

Table 4 summarizes the way that the norms rules and values of a society are associated with Americarsquos and Germanyrsquos institutionalarrangements their social system of production and their styles of inno-vativeness

CONCLUSION

Several themes deserve to be reemphasized Most importantly institu-tional analysis is emerging as a dominant issue in several social sciencedisciplines The study of various institutional components discussedabove is a complex subject requiring multiple levels of analysis and aknowledge of multiple sectors within societies It requires a theoreticalperspective that evolves from an empirical analysis in different societiesIf we are to advance the study of institutional analysis we need cross-disciplinary collaboration and team research which are not readilyachieved given the structure of most of our research universitiesAt present scholars in each academic discipline have their own distinc-

tive strategies for studying institutional analysis Thus economists tendto use several approaches (North 1990 Williamson 1985 Hodgson 19881998 1999) anthropologists another (Geertz 1995) and sociologistspolitical scientists and historians employ other strategies (Campbell etal 1991 Powell and DiMaggio 1991 Katznelson 1998 Hall and Taylor1996) With practitioners in various disciplines pursuing their distinctiveapproaches to institutional analysis there is low potential for collabo-rative institutional analysis across the social sciences The lessons fromthe biological sciences over the last 40 years are very instructive oncepractitioners from various biological disciplines began to cooperate instudying biological phenomena by using the same concepts at the mole-cular level theoretical advance was very rapid Comparable phenomenatook place in the development of other hybrid elds eg biochemistrybiophysics (Judson 1979 Olby 1979 Kohler 1982 1994) Similarly weresocial scientists in various disciplines to work collaboratively in the areaof institutional analysis there would very likely be an acceleration intheir theoretical knowledgeThis is not a plea that everyone engaged in institutional analysis should

do the same kind of research Indeed just as biologists working at themolecular level study many different kinds of problems similarly socialscientists engaged in institutional analysis would also work on manyseparate problems but in the spirit of a collective enterprise Some wouldconcentrate their attention on the study of rules norms habits conven-tions and values while others would study how these are associatedwith congurations of various institutional arrangements (types ofmarkets hierarchies networks associations communities clans statesetc) Others would work on specic institutional sectors (eg education

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

630

Table 4

Rules norms hab

its an

d values associated w

ith sp

ecic co

ngu

ration

s of coo

rdinating mecha

nism

s soc

ial sy

stem

s of

prod

uction

an

d nationa

l styles of inno

vation

USA

Germany

Norms rules values which provide incentives for

Norms rules values which provide incentives for

short-term horizons

long-term horizons

Con

gu

rative

forms of

Marke

ts c

orpo

rate hierarchies reg

ulatory state

Assoc

iation

s (busine

ss assoc

iation

s la

bor union

s)

coordinating

corp

oratist-type

networks

marke

ts c

orporatist-

mecha

nism

s (listed in

type

state

order of im

portanc

e in

the particu

lar social

system

)

Social sys

tem of

prod

uction

Bus

iness system

Firm

struc

tures

Con

glom

erates b

ut with mov

emen

t toward

Firm

more co

ngru

ent with pa

rticular prod

ucts

elim

ination of produc

tion

in som

e sectors

emph

asis on long

-term strateg

ies rm

s are well

emph

asis on sh

ort-term

strateg

ies high

integrated

differentiation

between divisions

of rms

Firm

s mob

ilized in

Low

High

and in

tegrated

into busine

ss

associations

Indu

strial relations

Low

deg

ree of jo

b secu

rity rigid in

ternal lab

or

High deg

ree of jo

b secu

rity

exible interna

l lab

or

system

marke

t ex

ible externa

l labo

r marke

tsmarke

t rigid externa

l lab

or m

arke

tPatterns of

Dispersed owne

rship frequ

ent turnov

erMore stab

le owne

rship patterns

owne

rship

Table 4

continue

d

USA

Germany

Norms rules values which provide incentives for

Norms rules values which provide incentives for

short-term horizons

long-term horizons

Social system of

Prod

uction

Bus

iness system

Training for labo

r an

dHigh deg

ree of voluntary decision mak

ing by

Collective an

d m

ore co

mpulso

ry d

ecisions

abo

ut

man

agem

ent

individuals as to ho

w m

uch

tim

e an

d ene

rgy

how m

uch ed

ucation

is ne

eded

he

avy em

pha

sis

to inv

est in educ

ation heavy

emph

asis on

on techn

ical training (eg en

gine

ering)

busine

ss sch

ools training

in marke

ting

and

sales accou

nting

Fina

ncial marke

tsEqu

ity an

d ven

ture cap

ital m

arke

ts highly

Equ

ity an

d ven

ture cap

ital m

arke

ts w

eakly

dev

elop

ed na

ncing from

cen

tral gov

ernm

ent

deve

lope

d he

avy em

pha

sis on

corporate loan

sto priva

te rm

s for dev

elop

men

t of new

tech

nologies related

to na

tion

al security

Unive

rsity research

High individual auton

omy high poten

tial for

Low

ind

ividua

l au

tono

my until almost 40

yea

rs of

system

individual creativity u

nive

rsity pr

ofesso

rs

age poten

tial for ind

ividual creativity high

ly

high

ly entreprene

urial

cons

traine

d unive

rsity pr

ofessors not so

entrep

rene

urial

Inno

vative

styles

Highly su

ccessful in m

ajor break

through

s in

High pe

rforman

ce in proc

ess an

d inc

remen

tal

basic an

d app

lied scien

ce go

od in dev

elop

ing

inno

vation

s in existing produ

cts not so su

ccessful

totally

new

lines of pr

oducts an

d in m

ajor break

through

s in basic scien

ce or in

tech

nologies high

deg

ree of entrepr

eneu

rship

deve

loping

totally new

lines of prod

ucts of

amon

g research

ers

tech

nologies low d

egree of entrepr

eneu

rship

amon

g research

ers

Table 4

continue

d

USA

Germany

Norms rules values which provide incentives for

Norms rules values which provide incentives for

short-term horizons

long-term horizons

Prod

ucts in w

hich

Pr

oducts with sh

ort liv

es prod

ucts in w

hich

Older produ

cts with long

history machine

ry pap

er

society excels

tech

nology

is bo

th highly co

mplex

and

pr

inting

textile

s m

otor veh

icles m

achine

too

lsch

ange

s rapidly en

tertainm

ent indus

try

pha

rmaceu

ticals biotechn

olog

y softw

are

aerosp

ace

business systems nancial markets and systems of research) Distinctivewould be the recognition of how these separate studies are linkedtogether how each institutional sector is linked to a societyrsquos normsrules values etc and its conguration of institutional arrangementsSimilarly other institutional analysts might focus on the study of orga-nizations with a major concern for how the institutional environmentof organizations (eg norms rules values congurations of institutionalarrangements institutional sectors) inuenced the structure culture andoutcomes of organizations And nally others might study how all ofthe aforementioned aspects of institutional analysis inuenced a societyrsquosoverall performance Nelson and Sampat (1998) have appropriatelyreminded us that we should engage in institutional analysis in order toaddress specic problems but the view developed in this article is thatwe rst need some mapping of the terrain of institutional analysis inorder for our diverse enterprises to become a meaningful collective enter-prise especially if we are to understand how institutions inuence theinnovative processThe argument developed here is not that all social scientists should engage

in institutional analysis Obviously there are numerous other importantareas of research which lie beyond the eld of institutional analysis Nordoes this article argue that institutional analysis is the most importantsubject for social scientistsSince the structure and culture of our research universities reect an

enormous amount of disciplinary fragmentation what are the prospectsfor promoting the kind of collective and interdisciplinary research agendaproposed above In one respect the prospects are not encouragingDuring the past half century the efforts of American universities to have effective communication across the biological sciences have variedgreatly in their success At the University of California-Berkeley it was not until the 1980s that there was a serious effort to integrate thebiological sciences and at that time the campus abolished seventeendepartments to promote an integrated biological program But at theUniversity of Wisconsin the biological sciences are still very fragmentedinto dozens of different departments in botany plant pathology zoologyand physiology with two departments of biochemistry and two depart-ments of geneticsThere have been other approaches which have worked very well histor-

ically for promoting scientic integration amidst scientic diversity Forexample several summer schools and special institutes were organizedin the United States during the 1940s in an effort to overcome the frag-mentation in the biological sciences These programs made it possiblefor scientists in diverse academic disciplines and from different univer-sities and countries to come together to engage in intense and frequentinteraction Perhaps the most effective program was that held in the

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

634

summers at Cold Spring Harbor New York where a research agendafor the study of molecular biology was developed (Fischer and Lipson1988 Stent and Watson 1966) Scientists went there summer aftersummer and when they returned to their universities many began therestructuring of the biological sciences in their universitiesIn our own day we need summer institutes and other similar programs

so that scholars from diverse backgrounds can come together anddevelop a collective agenda for doing institutional analysis and forstudying the impact of institutions on the performance of societies ndashinnovativeness being one such subject If we can develop such instituteswe will increase the prospects for transforming our universities Thereis reason to be optimistic that once an eclectic group of scholars workwith a common set of concepts we can have genuine theoretical advanceif researchers can bring their diverse expertise to bear on a common setof problems There is considerable evidence that institutional analysisoffers high potential for advancing our knowledge of innovations andtechnological change (Hollingsworth and Hollingsworth 2000) If we aregoing to make substantial theoretical advance it is not likely to occurfrom the vantage point of a multiplicity of disciplines Rather it is likelyto occur as we develop a new intellectual framework and our own acad-emic journals This is a tall agenda which requires us to reassess theway that knowledge is presently produced in our societies

NOTE

This article is part of a much larger agenda involving the study of institutionsorganizations and innovations in which I am involved My debts are many andvaried The rst version was written while I was in residence in the ResearchUnit for Institutional Change and European Integration of the Austrian Academyof Sciences in Vienna in the summer of 1998 I am especially grateful to ProfessorsEgon Matzner and Sonja Puntscher-Riekmann for providing the stimulating envi-ronment in which to think through the ideas developed here Egon Matzner hasengaged in very stimulating discussions with me about the importance of insti-tutions in the general eld of socio-economics The nal version of the articlewas completed during my residence as a fellow at the Neurosciences Institutein La Jolla California I also wish to thank my colleagues in the InnovationTheme Project at the Netherlands Institute for Advanced Study (NIAS) during1998ndash9 and Edgar Grande and Karl Muumlller for stimulating conversations aboutthe interaction of institutions and organizations in facilitating particular typesof innovation I am especially grateful to Jerald Hage for helping me to developa number of issues in the article and for his useful comments on the rst draftFrans van Waarden early on helped me to work through the idea of multiplelevels of institutional analysis Others who have helped to develop the ideas inthe article have been my colleagues in the research project lsquoComparative socialsystems of productionrsquo especially Tom Burns Christel Lane Yoshitaka OkadaWolfgang Streeck and Richard Whitley My colleagues Robert Boyer SteveCasper and David Soskice have taught me much about how incentives providedby the institutional environments of rms inuence individual decision making

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

635

But my most important debt is to Ellen Jane Hollingsworth who has repeatedlyquestioned every single assumption of this article Without her rich and contin-uous tutoring this article would be even less developed than it is at presentFinally without the insightful assistance of David Gear in many phases of thisarticle it would not have been completed

REFERENCES

Aoki Masahiko (1990) lsquoTowards an economic model of the Japanese rmrsquo Journalof Economic Literature March 1ndash27

Arora A Landau Ralph and Rosenberg Nathan (eds) (1998) Chemicals and Long-Term Economic Growth Insights from the Chemical Industry New York JohnWiley amp Sons

Arrow Kenneth (1974) The Limits of Organization New York NortonArthur W B (1988a) lsquoSelf-reinforcing mechanisms in economicsrsquo in P W

Anderson K J Arrow and D Pines (eds) The Economy as an Evolving ComplexSystem Redwood City Calif Addison-Wesley

mdashmdash(1988b) lsquoCompeting technologies an overviewrsquo in G Dosi C Freeman RNelson G Silverberg and L Soete (eds) Technical Change and Economic TheoryLondon Pinter Publishers pp 590ndash607

Audretsch David B (1995) Innovation and Industry Evolution Cambridge MassMIT Press

Audretsch David B and Feldman Maryann P (1995) lsquoInnovative clusters andthe industry life cyclersquo Discussion Paper Wissenschaftszentrum BerlinMarket Processes and Corporate Development Research Unit FS-IV-95-7

Baum Joel A C and Oliver Christine (1992) lsquoInstitutional embeddedness andthe dynamics of organizational populationsrsquo American Sociological Review 57540ndash59

Baum Joel A C and Singh Jitendra V (1994a) lsquoOrganizationndashenvironment coevo-lutionrsquo in Joel A C Baum and Jitendra V Singh (eds) The EvolutionaryDynamics of Organizations New York Oxford University Press pp 379ndash402

mdashmdash(eds) (1994b) The Evolutionary Dynamics of Organizations New York OxfordUniversity Press

Berger Suzanne and Dore Ronald (eds) (1996) National Diversity and GlobalCapitalism Ithaca NY Cornell University Press

Boyer Robert (1997) lsquoThe variety and unequal performance of really existingmarkets farewell to Doctor Panglossrsquo in J Rogers Hollingsworth and RobertBoyer (eds) Contemporary Capitalism the Embeddedness of InstitutionsCambridge and New York Cambridge University Press pp 55ndash93

Burns Tom R and Dietz Thomas (2001 forthcoming) lsquoSocial rule system theorysocial action institutional arrangements and evolutionary processesrsquo in JRogers Hollingsworth and Karl Muumlller (eds) Advances in SocioeconomicsLanham Md Rowman amp Littleeld

Burns Tom R and Flam Helena (1987) The Shaping of Social Organization BeverlyHills Calif Sage

Calvert Randall L (1995) lsquoThe rational choice theory of social institutionsrsquo inJeffrey S Banks and Eric A Hanuskek (eds) Modern Political Economy NewYork Cambridge University Press pp 216ndash66

Camic Charles (1986) lsquoThe matter of habitrsquo American Journal of Sociology 911039ndash87

Campbell John L (1997) lsquoMechanisms of evolutionary change in economicsgovernance interaction interpretation and bricolagersquo in Lars Magnusson

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

636

of individuals become important in understanding processes of societalchangeThese views are not meant to imply that the type of institutional

analysis proposed herein approximates a general theory of society Thisis clearly not the case However it is intended to represent the rst stepsin a mapping exercise of the boundaries of institutional analysis and tosuggest a few methodological insights for studying institutionsOne should try to see norms and rules as continuous and not as

dichotomous entities to recognize that they come in varying strengthsLegro (1997) has suggested that we assess the robustness of norms andrules with three criteria their simplicity their durability and their concor-dance Simplicity refers to how well actors understand norms and ruleshow well they can be applied within a specic situation Some normsand rules can be so complex that actors can have considerable difcultyin applying them in specic cases Durability addresses the issue of howlong norms and rules have been in effect ndash in short in order to assesstheir level of legitimacy While the position of this article is that normsrules and values are quite durable they do vary in this respectConcordance refers to how widely applied a norm or rule is Thisaddresses the degree to which a rule is a meta rule the degree to whichit incorporates the heterogeneity of other norms and rules In sum theclearer the norms and rules of a society the longer they have been inexistence and the more widely applicable they are the greater theirimpact on a society Hence the more robust the norms and rules thegreater their impact on a society and the less their robustness the greatertheir exibility and the less their effect on shaping a societyrsquos outcomesand performancesBecause norms rules and values are quite durable they play an impor-

tant role in shaping the history of societies thus contributing to a greatdeal of path-dependency Actors attempt to adjust to their contempo-rary environment but since they are products of the past the historicallegacy of norms rules and values inuences the decisions they makeAlthough actors have some capacity to alter the course of their historythey are constrained by their past and the degree to which they canmove beyond their past is limited As Lanzara (1998) Johnson (1992)and others have argued societal inertia is a basic feature of institutionsThey provide the basic stability necessary for change Degrees of historyare continuously reproduced by the way in which the inhabitants ofsocieties are socialized History matters but at critical points in historythere is punctuated equilibrium (Somit and Peterson 1992) During mostperiods of history there is considerable stability in the norms rules andvalues of a society but at critical moments norms rules and values canquickly and dramatically be redened At all times norms rules habitsconventions and values are inuencing each of the other components in

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

604

the institutional framework discussed below but these other institutionalcomponents feed back and can modify rules norms conventions etc(Murmann 1998)

Second level of analysis

The norms rules habits conventions and values of a society lead to thenext level of analysis ndash the institutional arrangements which are involvedin the coordination of various economic actors producers and suppliersof raw materials knowledge etc processors of raw materials infor-mation workers customers of raw materials nished productsinformation etc nanciers governmental and other types of regulatorsThese actors regularly engage in contests to resolve various economicproblems in virtually all sectors of society how are prices to be setWhat quantity of various products is to be produced How are stan-dards of various products processes etc to be set How is the qualityof products and processes to be determined How are various societalprocesses to be nanced In order to confront these problems and toaddress the conicting positions of economic actors as they address these problems societies develop various institutional (governance)arrangements for coordinating different actors These consist of marketsvarious types of hierarchies and networks associations the state commu-nities and clans (see Hollingsworth and Boyer 1997 Campbell et al1991 ch 1)Each of these particular kinds of coordinating mechanism has many

different types For example there are many types of states (eg theregulatory state developmental state authoritarian state welfare state)on which there is an extensive literature (Kim 1997) Similarly there aredifferent types of markets networks different kinds of associations etc(Boyer 1997 Hage and Alter 1997 Schneiberg and Hollingsworth 1990)When we do institutional analysis we must engage in congurativeanalysis recognizing that actors are not coordinated or governed by asingle type of institutional arrangement Some of the literature discussesindustrial sectors as though they were coordinated or governed by asingle institutional arrangement whereas in fact each sector of aneconomy is coordinated by a conguration of institutional arrangementsSome congurations coordinate actors in certain problem areas whileother congurations of institutional arrangements coordinate actors inaddressing other problems The types of congurations which are domi-nant in a society are somewhat stable and tend to persist over timewithin a society (Hodgson 1999)When one mode of coordination is dominant in a society it will

inuence the role which other coordinating modes will play Hence thestrong role of the state in the Soviet Union inuenced the role of markets

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

605

associations etc in the governance of the Soviet economy Similarly theprominence of particular modes of coordination in a society inuencesits style of innovationMuch of the literature on institutional arrangements (eg forms of eco-

nomic coordination) remains fragmented and unintegrated It is helpfulto array these various forms of coordination in a two-dimensional tax-onomy as in Figure 2 On the vertical dimension the economistrsquos viewof a self-interested agent is contrasted with a more sociological perspec-tive according to which social rules obligation and compliance shapehuman actions On the horizontal dimension there is another continuum(ie the distribution of power) At one extreme of the dimension onends many and relatively equal agents interacting (eg as in a well-organized spot market) At the other extreme inequality in power resultsin a hierarchical form of coordination which structures the interactionbetween principals and agents or between leaders and followers Wherea society falls along the distribution of power is inuenced by the rulesnorms values etc which are dominant in a society at a particular momentin timeInstitutional arrangements can be visualized on two dimensions the

nature of action motive on the one hand and the distribution of poweron the other Markets (cell 1) combine self-interest with horizontalcoordination transactors and they reect sensitivity to concerns aboutsupply and demand thus providing ex post an unintended equilibriumParadoxically the more pure and perfect the market competition thegreater the need for codied rules of the games for coordinating economictransactions Thus collective associations (cell 6) andor various formsof state intervention (cell 4) are required to develop and enforce rulesfor transacting partners (Schneiberg and Hollingsworth 1990 Streeckand Schmitter 1985a) This is an example of how the norms and rulesand the institutional arrangements of a society are intertwined asreected in a conguration of institutional or governance arrangementsAlong the horizontal axis actors can be joined in an organization or

a rm a hierarchy is the generic terms for this institutional arrangement(cell 2) Along the horizontal line one recognizes the difference betweentransactions in a market and transactions within a rm The well-knownworks of Coase (1960 1981) and Williamson (1975 1985) utilize theconcept of transaction costs in explaining the emergence of corporatehierarchiesThere are also various types of networks (cell 5) which exhibit mixes

of self-interest and social obligation with some actors being formallyindependent and equal Yet in some networks (the large rms and theirsubcontractors) there is unequal power and initiative Networks maycomprise all kinds of actors some consist only of rms but others includeassociations and the state (Hage and Alter 1997)

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

606

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

607

DISTRIBUTION OF POWER

VERTICALHORIZONTAL

1 MARKETS 2 HIERARCHY

6ASSOCIATIONS

5NETWORKS

4STATE

3COMMUNITIES

SELF INTEREST

SOCIAL OBLIGATION

ACTION MOTIVE

Figure 2 A general taxonomy of institutional arrangements

The vertical axis deals with action motives Toward the upper part ofFigure 2 actors are engaged in individualistically oriented behaviorwhereas toward the lower part actors are more engaged in collectivebehavior and strive to cope with problems of common interest Cell 3 ndashcommunities and clans ndash consists of institutional arrangements based ontrust reciprocity or obligation and thus not derived from the pure selshcomputation of pleasures and pains This is an unconventional form ofcoordination for most neoclassical economists (however see Arrow1974) but not for many anthropologists political scientists and sociolo-gists (Streeck and Schmitter 1985a Polanyi 1944 Gambetta 1988Fukuyama 1995 Sabel 1992 Putnam 1993)In the neoclassical paradigm theorists argue that actors engage in

forms of exchanges that best promote their individual interests If somestructural conditions are fullled (absence of increasing returns to scalethe reversibility of transactions absence of uncertainty and completecontingent markets with no collusion between actors) then the invis-ible hand theorem applies and market-type activity functions quite welland also provides the optimum for society thereby combining efciencyharmony and order However an excess of market activity may welllead to ruinous competition and excessive conict There is variation inthe extent to which ruinous competition occurs depending on the socialcontext in which transactions take place Thus it is important that webe sensitive to the institutional context in which transactions areembedded and that we understand the degree to which social bondsexist at both the micro and macro levels of analysis Micro bonds facil-itate exchanges in a society but at the societal level social bonds existat the level of the collective ndash in the community or region and amongmembers of racial religious and ethnic groups All other things beingequal the more powerful the social bonds among transacting partnersthe more economic competition is likely to be restrained Thus mosttransactions occur not simply in an impersonal calculative system ofautonomous actors unrestrained by social ties ndash as implied by theneoclassical paradigm ndash but in the context of social ties variation in the strength of which leads to variation in levels of trust and transac-tion costs (Etzioni 1988 211 Granovetter 1985 Streeck and Schmitter1985a Hollingsworth and Boyer 1997 Hodgson 1988 1999 Schneibergand Hollingsworth 1990)Another form of multilateral institutional arrangement is various

types of associations (cell 6) Unlike networks clans and communitiesassociations are more formal organizations Whereas markets corporatehierarchies and networks tend to coordinate economic activity amongdifferent types of actors (eg producers with suppliers capital withlabor) associations typically coordinate actors engaged in the same orsimilar kinds of activities Business associations and labor unions are

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

608

some of the most common forms of associations for coordinatingeconomic activity in capitalist economies (Schneiberg and Hollingsworth1990 Streeck and Schmitter 1985a)Finally there is the state (cell 4) which is an institutional arrangement

quite unlike any of the others It is the state that sanctions and regu-lates the various non-state coordinating mechanisms that is the ultimateenforcer of rules of the various mechanisms that denes and enforcesproperty rights and that manipulates scal and monetary policy Overallit is the state that inuences the total incentive system of a society(Lindberg and Campbell 1991 Johnson 1992 40) At the same time thestate may also be an economic actor by engaging directly in productionand exchange relationsThe choices of institutional arrangements in Figure 2 are constrained

by the social context in which they are embedded Depending on thenature of that embeddedness there is variation in the collective formsof governance some of which are specied in the lower part of thegure Some modes of coordination in the bottom and upper levels ofthe typology are often mixed together though one coordinating modeis likely to be more prominent than another But actors are often engagedin complex congurations involving several kinds of institutionalarrangementsEach of these various institutional arrangements has its own logic ndash

its own rules its own procedures for enforcing compliance its ownnorms and ideologies which help to reduce the costs of enforcementThese are summarized in Table 2 which provides further elaborationabout the various coordinating mechanisms one nds in almost everycapitalist society Table 2 lists the organizational arrangements and theirstructures rules of exchange and means of enforcing compliance asso-ciated with each type of coordinating mechanism While each type ofinstitutional arrangement has various positive features each institutionalarrangement also has failures and these are featured in Table 3 It is thecontest between those who support and those who oppose these variousinstitutional arrangements that tends to lead to transformations in insti-tutional arrangements over time (Campbell et al 1991 Campbell 1997)Again it is difcult to conceive of pure institutional arrangements ndash

either exclusively markets or exclusively hierarchies ndash since it has beenwell known since Adam Smithrsquos Wealth of Nations that the division oflabor within the rm cannot be disentangled from the existence andextent of the market Institutional arrangements have their own distinc-tive form of efciency (good static efciency for the market dynamicefciency for rms) and inefciency often leading to considerableinequality Neither networks nor communities are panaceas for economiccoordination without being congured with other types of institutionalarrangements Networks and communities may solve certain issues but

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

609

Table 2

Log

ics of ins

titution

al arran

gemen

ts

Coordination

Organizational

Rules of

Individual means

Collective means

mechanisms

structure

exchange

of compliance

of compliance

Marke

tsEasy en

try an

d exit

Voluntary sp

ot excha

nge

Leg

al enforcemen

t of

Norm of pr

ivate

Bila

teral ex

chan

ge or

control

prope

rty

marke

t site (W

all Street)

Reg

ulations

to

Leg

itim

acy of m

arke

t en

force contracts

men

tality

Com

mun

ities

Inform

al m

embe

rship

Voluntary exch

ange

based

So

cial norms an

d m

oral

Highly institutiona

lized

ev

olve

s ov

er lon

g period

on social solid

arity an

d

princ

iples im

pose

norm

s an

d rules requ

ire

of tim

ehigh

deg

ree of tru

stob

ligations

mem

bers to accept

Kno

wledg

e of others an

d

lsquocorpo

ratersquoo

bligations

recipr

ocity ov

er tim

e

Networks

Semi-form

al m

embe

rship

Voluntary exch

ange

ove

r a

Con

tractual bon

dsPersona

l relation

sBila

teral or m

ultilateral

time pe

riod

Resou

rce de

pend

ence

Tru

st built outside

the

ex

chan

geecon

omic arena

Associatio

nsFo

rmal m

embe

rship

Restricted to mem

bers

Self-interest

Some deg

ree of

Multilateral excha

nge

Emph

asis on insider

Rep

utation

effects

compu

lsion

outsider or we

they

Priva

te interest type

of

men

tality

gove

rnan

ce

Private

Com

plex organ

izations

Restricted to mem

bers

Rew

ards to ind

ividuals

Highly institutiona

lized

hierarch

ies

which

ten

d to becom

e ex

chan

ge based

on

Asymmetric po

wer threat

rules

bureau

cratic hierarchies

asym

metric po

wer

of san

ctions

Mem

bers soc

ializ

ed into

bureau

cratic rules

corp

orate cu

lture us

e of

sanc

tions

State

Pub

lic hierarchy

Unilateral action

Exit vo

ice (vote lob

bying)

Coe

rcion

De jure

and impo

sed

Indirect political an

d

loya

ltyNorms an

d pu

blic rules

mem

bership

econ

omic excha

nge

Table 3

Failu

res of coo

rdinating mecha

nism

s

Coordinating mechanisms

Type of failure

Associations

Private hierarchies

State

Market

Communities

Networks

Enforcem

ent

Usu

ally relies on

Might enh

ance

Needs co

ntrol

Needs an

internal

Needs

tru

st and

Need an external

the state as an

oppo

rtun

istic

extern

al to state

enforcem

ent

loya

lty often

en

forcem

ent

enforcer

beha

vior

bureau

cracy

authority

coming from

au

thority

Resem

bles

The

ideal o

f (ju

dge

s

Facilitates

outside (fam

ily May

facilitate

enforcem

ent

internal m

arke

ts

parliamen

t collu

sion

and

relig

ion

cartelization an

d mecha

nism

might hur

t marke

t) to

impe

rfect

ethn

icity)

mon

opoly

of cartels

incu

mbe

nt

correct state

compe

titio

nCom

patible with

worke

rsab

uses

variou

s types of

Lob

bies can

co

mpe

tition

captur

e pu

blic

interest goa

ls

Public goo

d

Useful for

Gov

erna

nce co

sts

Can

provide

Can

not provide

Can

internalize

Useful for

and

establishing

might exceed

public goo

ds

colle

ctive go

ods

some co

llective

enha

ncing qu

ality

externality

stan

dards an

d

the be

nets of

but ha

s or deal w

ith

good

s (qua

lity

and training bu

t qu

ality for

internal division

difc

ultie

s in

externalities

training

) bu

t no

t no

t ve

ry goo

d in

setting ru

les of

of la

bor

prov

iding them

Inad

equate

othe

rs (welfare

prov

iding for

compe

titio

n in

Slow

to react to

in precise

mon

itoring of

gene

ral p

ublic

societal gen

eral

the indu

stry

chan

ges in the

am

ounts

tech

nical ch

ange

go

ods)

welfare

Useful for

environm

ent

Might fail in

and inno

vation

Mem

bers tightly

Weak in the

prov

iding

induc

ing

integrated

into

prov

ision of

man

y go

ods

tech

nical

community

colle

ctive go

als

colle

ctively

chan

geha

ve limite

d

that in

dividua

lcapa

city for

mem

bers can

not

inno

vatio

nsprov

ide for

them

selves

Table 3

continue

d

Coordinating mechanisms

Type of failure

Associations

Private hierarchies

State

Market

Communities

Networks

Efcien

cyFa

cilitates

De

cien

t in

Can

be high

ly

Some ba

sic social So

me go

ods

Slow

to en

hanc

e co

operation an

d

coop

eration an

d

bureau

cratic

relation

s cann

ot

cann

ot be

efc

ienc

y an

d

X-efcien

cy but

X-efcien

cyan

d can

not

be provided

by

delivered

at

speed of

not allocative

easily deliver

pure m

arke

t su

fcien

tly

adap

tive

ness

efcien

cy

good

s at

mecha

nism

slow costs

except in

low cost

indus

tries whe

re

tech

nology

is

complex an

d rapidly cha

nging

Equity

Narrow

Excessive

Might enh

ance

Facilitates

Might lead to

Whe

n widely

encompa

ssing

multiplication

ineq

uality

ineq

uality in

retard

ed

dev

elop

ed in

to

associationa

l of con

trollers

(pow

er

income an

d dev

elop

men

tindus

trial

stru

ctures lead

(fru

stration

and

an

d privilege

)wealth

districts

to in

come

ineq

uality)

netw

orks m

ay

equa

lity

facilitate greater

equality an

d

inco

me

distribution

whe

n wea

kly

dev

elop

ed

netw

orks ten

d to

increase soc

ial

ineq

uality

Note For allo

cativ

e efc

ienc

y an

d X

-efcien

cy see Leibe

nstein (19

66 19

76)

they raise other no less severe problems It is important to recognizethe imperfection of any single institutional arrangement in order tocomprehend the origin and transformation of any other institutionalarrangement (Hodgson 1999 ch 3)As suggested above each institutional arrangement is congured with

other institutional arrangements Usually one particular institutionalarrangement is more dominant in a particular conguration but becauseeach type of institutional arrangement has its own strengths andweaknesses there is no simple structural logic in the governance or coor-dination of a society Each institutional arrangement constrains the otherbut the inherent tension among the various institutional arrangementswithin a conguration contributes to changes in the governance orcoordination of a society The routines and logics of each of these cong-urations of institutional arrangements provide constraints and incentivesfor actors The way that actors perceive the incentives and constraintsof these governance congurations leads to a particular market logicand it is the specic market logic of a society which inuences its speciccapacities and weaknesses The inherent strain among the different logicsin a conguration of coordination helps to provide the exibility for asociety to adapt to new circumstancesIf there were a society with pure markets or a society coordinated

only by the state there would be too much rigidity and too little diver-sity to cope with the vast uncertainty in the global environment A societywith very little diversity in its coordination mechanisms would havelittle capacity to adapt to new circumstances Congurations with consid-erable diversity of institutional arrangements provide for a certainamount of incoherence in governance but they also provide for thecapacity to adapt to new circumstances While the Soviet Union wasdominated by the state as a coordinating mechanism there were alwaysfunctioning markets in the system Moreover the feudal society of thetwelfth and thirteenth centuries consisted not only of hierarchical rela-tionships between serfs and masters but also of urban guilds clericalhierarchies and markets (Johnson 1992 38) In sum the robustness ofinstitutions often depends on multiple and diverse principles and logicsof action on the inconsistency of principles and procedures on patternedforms of disorder (Lanzara 1998 Orren and Skowronek 1991 320 329)

Third level of analysis

The next level consists of the institutional sectors of a society The rulesnorms and values of a society inuence the array of institutional arrange-ments and levels one and two inuence the nature of and the relation-ships among various institutional sectors and all three levels areintricately linked together to form a social system of production Together

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

613

all of these components inuence the performance of economic sectorswithin a society as well as the performance of the total society A socialsystem of production is the way that a societyrsquos institutions (see rstlevel of analysis above) its institutional arrangements (see second levelof analysis above) and its institutional sectors are integrated into a socialcongurationAn institutional sector includes all organizations in a society which

supply a given service or product along with their associated focal orga-nizations (eg major suppliers funding sources regulators and so forth(Scott et al 1994 108 117 Campbell et al 1991 Hollingsworth et al1994) Institutional sectors include but are not limited to the societyrsquossystem of education system of research business system nancialmarkets legal system and the state The structure of the nancial marketssystem of training and education industrial relations system system ofresearch and state and legal systems are distinct and idiosyncratic ineach society In short they are system specicAll of these institutional sectors tend to cohere with each other

although they vary in the degree to which they are tightly coupled intoa full-edged system While each of the institutional sectors has someautonomy and may have some goals that are contradictory to the goalsof other institutional sectors with which it is linked an institutional logicin each society leads institutions to coalesce into a complex social con-guration (Hollingsworth 1991) This occurs because the institutionalsectors are embedded in a culture in which their logics are symbolicallygrounded organizationally structured technically and materially con-strained and politically defended The conguration of institutionalsectors usually exhibits some degree of adaptability to new challengesbut continues to evolve within an existing style But under new circum-stances or unprecedented disturbances these institutional congurationsare exposed to sharp historical limits as to what they may or may notdo (Schumpeter 1983 David 1988 Arthur 1988a 1988b Haringkanssen andLundgren 1997)Why do all of these different institutions coalesce into a complex

social conguration which is labeled here as a social system of pro-duction The literature suggests two contrasting interpretations Part ofthe answer ndash indeed a controversial one ndash is that these institutionalsectors are functionally determined by the requirements of the practiceof capitalism in each time and place (Habermas 1975) Another expla-nation emphasizes the genesis of the actual conguration via a trial and error process according to which the survival of rms regions orcountries is the outcome of complex evolutionary mechanisms (Maynard-Smith 1982 Nelson and Winter 1982) However the problem is evenmore complex Markets and other mechanisms for coordinating rela-tionships among economic actors place constraints on the means and

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

614

ends of economic activity to be achieved in any society The othercoordinating mechanisms include different kinds of hierarchies varioustypes of networks and associations ndash trade unions employers and busi-ness associations (Hollingsworth and Lindberg 1985 Campbell et al1991) The logic of these various institutional arrangements providesactors with vocabularies and logics for pursuing their goals for deningwhat is valued and for shaping the norms and rules by which theyabide (see Tables 2 and 3) In short in contrast to the logic of theneoclassical paradigm the argument here is that the dominant type ofinstitutional arrangements places severe constraints on the denition of needs preferences and choices of economic actors Whereas the neo-classical paradigm assumes that individuals and rms are sovereign thisarticle is based on the assumption that rms are inuenced by the hold that the institutional congurations making up a social system ofproduction have on individual decision making (Campbell et al 1991Etzioni 1988 Streeck and Schmitter 1985a Hollingsworth et al 1994Hollingsworth and Boyer 1997 Magnusson and Ottosson 1997 North1990 Hodgson 1999)Standard neoclassical economic theory has tended to downplay the

role of production and consequently of rms Even transaction cost theo-rists who are concerned with analyzing the rm as a coordinatingmechanism have been relatively unconcerned with the various compo-nents of a social system of production Indeed as long as there waswidespread optimism about the efcacy of Keynesian economics therewas relatively little concern among neoclassical economists with thesupply side of the economy Even in the opinion of most Keynesians agroup of experts should ideally be able to shape the size of aggregatedemand while the supply side of the economy would be left to the twominimalist institutions of neoclassical economics ndash markets and manage-rial hierarchies For several decades however it has become increasinglyobvious that some of the most competitive and successful patterns ofindustrial output and industrial production in capitalist economies donot derive from the neoclassical prescription of unregulated markets andcorporate hierarchies complemented by a neoliberal democratic stateIndeed it is now well understood that certain highly successful produc-tion patterns require for their emergence and survival institutionalarrangements and environments the very opposite of the prescriptionsfound in the neoclassical paradigm (see especially Streeck 1992 but alsoHollingsworth and Streeck 1994) Thus social scientists have come tounderstand that if they are to comprehend the behavior and performanceof contemporary economies concerns about social systems of produc-tion must be brought into the pictureBecause production involves more than technology a number of social

scientists have an increasing concern with social systems of production

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

615

The same equipment is frequently operated quite differently in the samesectors in different countries even when rms are competing in the same market (Maurice et al 1980 Sorge and Streeck 1988 Sorge 1989Hollingsworth et al 1994) Variations in production and process tech-nologies are inuenced partly by variations in the social environmentsin which they are embedded In other words rms are embedded intocomplex environments which among other things place constraints ontheir behavior Thus a social system of production is of major impor-tance in understanding the behavior and performance of an economyHow the state and other institutional arrangements (eg marketsnetworks private hierarchies associations) coalesce and are related toparticular social systems of production are important determinants ofeconomic performanceThe congurations of institutional arrangements that coordinate or

govern the behavior of actors in one society and its structure of specicinstitutional sectors cannot easily be transferred to another society forthey are embedded into a social system of production that is societallydistinct (Hollingsworth 1997) Societies borrow selected principles offoreign management styles and work practices but the effectiveness ofsuch borrowing is generally limited Economic behavior and performanceare shaped by the entire social system of production in which actors areembedded and not simply by specic principles of particular manage-ment styles and work practices Moreover a societyrsquos social system ofproduction tends to limit the kind of goods that it can produce and withwhich it can compete successfully in international marketsSome scholars (Kenney and Florida 1993 Oliver and Wilkinson 1988)

have assumed that the diffusion of particular forms of management stylesand work practices across societies could lead to system convergenceReferring to the Japanese production system Kenney and Florida haveargued that it consists of organizational practices whose fundamentallsquogenetic codersquo can successfully be inserted into another society and beginsuccessfully to reproduce its behavior in the new environment (1993 8)Their position is in the intellectual tradition of Antonio Gramsci (1971)who contended decades ago that the American system of mass produc-tion would most certainly diffuse to Europe over timeHowever the argument of this article is that even though British

French and American rms may adopt certain aspects of Japanesemanagement styles (eg just-in-time production self-managing teamsquality circles the use of statistical process controls) or some variant ofthe German vocational system their social systems of production willnot converge Moreover the overall conguration of a societyrsquos socialsystem of production inuences its sectoral and overall national economicperformance This helps to explain why societies have different systemsof innovation and why they vary in the clusters of industries in which

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

616

they are highly competitive in the international markets (Nelson 19931999 Hage and Hollingsworth 2000 Berger and Dore 1996)In order to understand how and why a societyrsquos economy performs

as it does it is necessary to understand its entire social system of produc-tion If a society is to modify substantially the performance of itsinstitutional sectors and its entire economy it cannot adapt only someof the management and work practices of its foreign competitors Ratherit must alter its entire social system of production Because a societyrsquosmodes of economic governance and coordination and its institutionalsectors develop according to particular logics and are system specicthere are serious limitations to the extent to which a society may mimicthe institutions (eg the rules and norms) institutional arrangements andinstitutional sectors of other societiesIn the history of modern societies there are logics by which institu-

tions coalesce into a social system of production (Hollingsworth 1991)Though institutions are constantly changing there are sharp limits tothe type and the direction of change that any particular institution canundergo because of its linkages with institutional arrangements and insti-tutional sectors Thus a societyrsquos business rms educational systemnancial markets industrial relations system etc can engage in seriousrestructuring only if most of the other institutional sectors also changeIn social systems of production there are pressures toward consistencyin the norms rules and values across institutional sectors though in anycomplex society social systems are obviously imperfectly integratedIndeed the degree to which the institutional norms and rules makingup a social system of production are loosely or tightly coupled is a vari-able of considerable importance In general the institutional sectorsmaking up a social system of production are interdependent and changesin one generally result in changes in the others Since each institutionallevel is dependent on the others for various types of resources there isinterdependence among the differing institutional spheres Moreovereach society has its norms moral principles rules and laws and recipesfor action as well as its own idiosyncratic customs traditions and prin-ciples of justice (Burns and Flam 1987)There are also other inherent obstacles to convergence among social

systems of production for where a system is at any one point in timeis inuenced by its initial state Systems with quite different initial statesare unlikely to converge with one anotherrsquos institutional practicesExisting institutional arrangements block certain institutional innova-tions and facilitate others (Roland 1990) Thus the institutions makingup a social system of production provide continuity even though insti-tutional arrangements are always changing but with a particular logicWhile Williamson (1975 1985) suggests that actors tend to choose theinstitutional arrangements which are most efcient North (1990) is

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

617

much closer to the mark in his argument that most societal institutionalarrangements exist as a result of custom and habit and are thereforeinefcient At any moment in time the world often appears to its actorsas very complex and uncertain For this reason actors often engage incontradictory forms of behavior pursuing different strategies as hedgesagainst a very uncertain world (Lanzara 1998) And their hedging andcontradictory forms of behavior may lead in somewhat different soci-etal directions all constrained by the institutional fabric within whichthe actors are embedded This kind of contradictory behavior occurs inpart because of the contradictions inherent in the dominant congura-tions of institutional arrangements with their contradictory logics (seelsquoSecond level of analysisrsquo above)Despite the emphasis on the logic of institutional continuity this is

not an argument that systems change along some predetermined pathThere are critical turning points in the history of highly industrializedsocieties but with the choices limited by the existing institutional terrainBeing path-dependent a social system of production continues along aparticular logic until or unless a fundamental societal crisis intervenes(Milgrom et al 1991 Krugman 1991 Durlauf 1991 Hollingsworth 1991Pred 1966 David 1988 Hodgson 1998)At this point it is important to confront the question of how social

systems of production evolve And this question gets to the heart of theproblem of building complex societal structures Certainly social systemsof production did not emerge from some process of social engineeringMoreover the various component parts of each social system of produc-tion have often not been designed to be part of a social system ofproduction The component parts of each social system of productionhave emerged more often than not as unintended by-products of goalswhich various actors had in mind at earlier moments in time It is usuallythe case that actors at time lsquotrsquo fail to comprehend the long-term conse-quences of their actions Wolfgang Streeck the author of several brilliantpapers on social systems of production argues that institutional sectordesigns created for one purpose generally address that goal but overtime those designs have quite unintended consequences (Streeck 1997a1997b Wright 1998) Describing the conguration of the German socialsystem of production of the 1970s and 1980s Streeck points out that itwas the unintended by-product of multiple points in history Someelements were pre-Wilhelmian others were introduced by the Alliedpowers after 1945 and others emerged during the years of the GermanFederal Republic All component parts of the German social system ofproduction

were and continue to be changing for their own reasons as wellas in reaction to each other and certainly there can be no presump-tion of a pre-established t between them even though one might

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

618

want to allow for some reinforcement effects of the modelrsquos histor-ically contingent social and economic success

(Streeck 1997b 54)

In sum the emergence of social systems of production is a long-termevolutionary process with each part interacting with its environment andresulting in a congurative whole but with no previous design by eithera single actor or a collectivity of actors As suggested above actors do not recreate their world anew At any moment in history the totalinstitutional complex of a society necessarily contains the resources and legacies of its past Preexisting institutional complexes are nevercompletely wiped out Societies consist of multiple layers of history withtheir diverse logics of action However the cumulative effect of smallperipheral changes in altering particular institutional sectors and insti-tutional arrangements can be substantial (Murmann 1998 Lanzara 1998)lsquoInstitution building is affected more by the ways in which people havecodied the past than by how they envision the future There are manymore potential resources in pre-existing institutional arrangements thanit is usually assumed or suspectedrsquo (Lanzara 1998 30)

Fourth level of analysis

The next level of institutional analysis organizational structures is some-what more controversial As suggested above Douglass North (1990)draws a sharp distinction between institutions and organizations Morerecently many organizational theorists (Powell 1991 Powell andDiMaggio 1991 DiMaggio and Powell 1983 Baum and Oliver 1992Kondra and Hinings 1998 Townley 1997) argue that institutional rulesnorms and conventions unfold in tandem with organizational structuresand this is my positionThe literature which focuses on how institutions inuence organiza-

tions is quite different from two other theoretical literatures which alsoare concerned with how organizational environments shape the behaviorof organizations There are resource dependency theorists (Pfeffer andSalancik 1978) who emphasize the role of environmental resources inshaping organizations while Hannan and Freeman (1977 1984) and otherpopulation ecologists emphasize the survival of organizations givencertain kinds of organizational conditions (Orru et al 1991)The institutionalist perspective as implied above emphasizes the

normative environment in which organizations are embedded It is a perspective which focuses on the way in which organizations in their behavior tend to conform to the institutional rules and normswhich are dominant in the organizational environment However allthree perspectives ndash the institutionalist the population ecology and

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

619

the resource dependency ndash emphasize how the environment inuencesorganizations and how organizations which are subject to the sameenvironment tend to converge in their behavior to have what DiMaggioand Powell have labeled organizational isomorphism (DiMaggio andPowell 1983)There has emerged a vast literature which demonstrates that within

each society there are scal political judicial and other regulatory normswhich limit and shape the culture and structure of organizationalbehavior The normative institutional environment of organizations limits the options of what organizations do in a particular society and inuences the patterns of ownership relations with suppliers andcustomers In short it is the normative environment of organizationswhich denes within a particular society what is socially acceptablebehavior for organizations (Hamilton and Biggart 1988 Hollingsworthand Hollingsworth 2000 Meyer and Rowan 1991 Zucker 1987 1988Orru et al 1991 Townley 1997)Thus far most of the analyses on isomorphic pressures operating on

organizations have been cross-sectional in nature However the entireconcept of isomorphism implies that there are strong environmental pres-sures exerted on organizations and in order to observe this phenomenonit is important that we have longitudinal studies which assess howchanges in an organizationrsquos institutional environment inuence changesin the structure and culture of organizations Moreover the expectationhere is that when one analyzes the patterns of organizational changeover time the isomorphism will be holistic ndash that is within a particularsociety there will be a tendency for the internal structure of particulartypes of organizations to converge in large part because of the pres-sures to conform to the changes in the external norms rules and valuesA close reading of Aoki (1990) and Slack and Hinings (1994) nds thatchanges in the national institutional environment of organizations haveinuenced changes in the internal structure of organizations within aspecic industry of a particular societyHowever it is inappropriate to view organizations as changing only

in relation to exogenous environmental change There is a kind of co-evolution which occurs between organizations and their institutionalenvironments Co-evolution implies nonlinear feedback between orga-nizations and their institutional environments This is of course quitedifferent from the more common way of studying organizations bymodeling relations between independent and dependent variables Ofcourse the more common type of analysis is quite appropriate whenthere are simple relationships which do not involve complex feedbackprocesses But in the strategy for institutional analysis which is proposedherein it is less useful to separate independent from dependent vari-ables and more useful to understand the interacting and co-evolutionary

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

620

processes among all of the levels of institutional analysis discussed inthe various sections above (Baum and Singh 1994a 379ndash80 Mowery andNelson 1999 Murmann 1998)The logic of the perspective herein suggests that within every

society there is variation in the structure and culture of business rmsuniversities and other complex organizations (Kondra and Hinings1998) However this variation takes place within parameters which aresystem specic For example every German rm and every Germanuniversity is different from every other German university and rm but there are a set of cultural and structural characteristics which distin-guish German rms and universities from those in America Similarlyeach Japanese rm and university is unique yet they have a set ofcultural and structural characteristics which are system specic andwhich differentiate them from business rms and universities in other countriesIn all societies each organization has its own distinctive organizational

rules norms and conventions which are subordinate to the meta norms and rules of the larger society within which they are embeddedHowever the strength of the institutional environment within whichorganizations are embedded varies from society to society Some societies have multiple institutional environments and there is hetero-geneity in terms of what constitutes appropriate behavior oforganizations In such societies there can be conict over whatinstitutional logics should regulate specic organizational functions(Friedland and Alford 1991 Townley 1997 262ndash3) In those societies inwhich the institutional norms habits and rules are most developed and in which the institutional pressures to conform are greatest thereis less variation in the structure and culture of business rms and variouskinds of research organizations In such societies the connectednessbetween research organizations and their institutional environments is sufciently strong for organizations to have low autonomy to pursue independent strategies and goals in these societies there is a great deal of organizational isomorphism Conversely the weaker the institutional environment in which research organizations areembedded the greater the variation in the structure and culture of business rms and research organizations Moreover where the institu-tional environments are more weakly developed organizations havegreater autonomy and exibility to respond to the development of new knowledge and to be highly innovative Hence it is in those societies where the institutional environments are most developed and most rigid and when there is less organizational autonomy and exibility that fewer radical innovations in basic and applied science as well as in totally new products and industrial sectors havebeen made

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

621

Fifth level of analysis

We come next to the outputs and performance of the various institutionalcomponents of a society It is at this level that institutional componentsare more pragmatic and exible For example in the legal sector thereare specic statutes and court rulings in the state sector there are specicpolicies in the business sector there are new products new technolo-gies and market strategies It is at this nal level of analysis thatinstitutional spheres are most open and susceptible to change and cross-national mimicry is easiest and most commonThrough the outputs of the society we can obtain some assessment as

to how well a society is performing Moreover we can assess how inno-vative it is how egalitarian it is in the distribution of its resources howegalitarian it is in terms of levels of health education etc Just as it iscomplicated to measure how egalitarian a society is similarly it is dif-cult to assess how innovative it is For example societies vary in makingincremental and radical breakthroughs in basic and applied science in developing totally new products and new kinds of organizations inincremental and process innovations in existing products and organiza-tions in developing new and different forms of marketing ndash bothdomestically and globally (Hage and Hollingsworth 2000) Of coursesocieties vary in their rates of economic growth in their rates of economicproductivity in their quality of life (eg rates of crime life expectancyetc) All of these performance criteria feed back and inuence each ofthe levels of institutional analysis discussed above ndash rules norms valuesetc institutional arrangements institutional sectors the structure andform of organizations Moreover performance measures may inuenceeach other For example if a society has very low rates of growth itmay not be very innovative in some of the types of innovative activitiesmentioned above Good or poor performances of certain kinds inuenceother performance indicators (Hollingsworth et al 1990)Different institutional arrangements and different social systems of

production result in different types of economic performance Hence solong as societies have different social systems of production there areserious constraints on the degree to which they can converge in theirinnovative styles Different social systems of production tend to maxi-mize in a more or less explicit manner different performance criteriausually mixing considerations about static and dynamic efciency protsecurity social peace and economic andor political power In short incontrast to the implications of neoclassical economic theory in real-worldeconomies there are no universal standards all economically rationalactors attempt to maximize Economic history provides numerous exam-ples of how a great variety of principles of rationality are implementedin different societies

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

622

A critical question in institutional analysis is whether an existing socialsystem of production which supports a set of routines for a particularkind of technology and industry can shift from old practices to newones There are numerous historical accounts of how the social systemof production in a particular society which worked so well for a numberof years could not adapt to new technologies (Lazonick 1990 Veblen1915 Nelson 1994) Schumpeter (1983) and Freeman (1991) have devel-oped the idea of a techno-economic paradigm which suggests thatdifferent technologies require specic forms of organizational arrange-ments ndash or what I called social systems of production Because oflsquoinstitutional inertiarsquo the social system of production of a society maynot be able to adapt to a new techno-economic paradigm (Nelson 199458 Lanzara 1998)Whether a social system of production can sustain its performance

standards depends not only on its intrinsic economic rationality but alsoon where it ts into a larger system If a particular social system ofproduction is immune from the innovativeness and competition of analternative system survival can be long lasting But if different socialsystems of production with diverging criteria of good economic perfor-mance meet in the world arena the arbitrariness of nationally imposedconstructed performance standards may be superseded by alternativeperformance criteria as a result of international competitivenessThe world economy is also socially constructed just as are national

economies Even if different social systems of production are competingin the international arena it is not always possible to determine whichis more competitively effective at any moment in time Hegemonicnation-states in the short run shape the rules of trade to favor their indus-trial sectors and rms But the history of hegemonic powers suggeststhat in the longer run such social systems of production sustainedlargely by military and political power eventually give way to moreinnovative and competitive social systems of production In our ownday as nation-states are increasingly integrated into a world economyeven hegemonic powers lose their innovative and competitive advan-tage and their share of world output decreases Such a country mayattempt to restructure its institutional arrangements and to readjust itsperformance preferences But to restructure its system of innovationgenerally calls for a major redistribution of power within a societyLargely for this reason societies have historically had limited capacityto reconstruct a system of innovation in the image of their major competi-tors Each countryrsquos social system of production is a conguration of ahost of norms rules and values as well as of institutional arrangementsEach system is constantly changing and is open to inuence from othersystems And indeed many technologies and practices diffuse from onesociety to another but a societyrsquos capacity to be innovative is constrained

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

623

by the existing social system of production Thus the same technologymay exist in numerous countries but how it is employed and how itinuences societal outcomes and performances varies from one institu-tional conguration or society to another (Hollingsworth 1998)

PERSPECTIVES ON INSTITUTIONAL CHANGE

A fundamental problem which remains unresolved in institutionalanalysis involves the nature of institutional change There is much confu-sion and miscommunication in the social sciences about institutionalchange in large part because we do not even know how to measure therate of institutional change Of course we know that there are bothexternal and internal focuses for change and we have had a good bitof scholarship on this subject What is not clearly understood is thatinstitutions and institutional arrangements within societies are histori-cally rooted that there is a great deal of path-dependency to the waythat institutions evolve Moreover the more intricately linked each sectoris with each other sector and with a societyrsquos rules and norms the lesschoice actors have to devise new institutions and institutional arrange-ments On the other hand if a sector or actor is too isolated it may betoo weak within a society to bring about any kind of effective changeThese considerations are very critical to the problem of how muchfreedom actors have to create new institutions afresh to what extentinstitutions diffuse from one society to another (Murmann 1998 Lanzara1998) This is a fundamental question for students of institutionalanalysis as we attempt to understand how the different institutionallevels are linked togetherThere are a variety of reasons why there is confusion in studying insti-

tutional change There is disagreement over the extent to which actorshave the freedom to build new institutional arrangements and if theydo to what extent the new arrangements and organizational patternsmay depart from past practices Among some institutional economists(Williamson 1985) there is a basic assumption that actors build efcientinstitutions by pursuing their self-interest and by promoting theirstrategic goals Moreover there is the assumption that a mixing of fullrationality with market competition tends to produce all of the optimalinstitutions that are needed in order to coordinate a complex capitalisteconomy (Hollingsworth and Boyer 1997 ch 14 Hodgson 1999) Incontrast the argument developed here is much more complexSocial institutions are historically rooted and there is a great deal of

path-dependency in the way that various institutional componentsevolve The shape of institutional congurations at any moment limitsthe type of options for change Because there is a great deal of institu-tional inertia radical change in the institutional components of a society

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

624

is uncommon (Lanzara 1998) Recognizing the constraints existinginstitutions exert on actors we note that existing congurations of insti-tutional arrangements institutional sectors and organizational propertiescan limit but also inuence the degree to which norms and rules canchange Many scholars (North 1990 Ostrom 1990) have emphasized therole of norms rules ideologies and values in limiting the ability of actorsto develop new institutional arrangements which radically depart fromexisting ones However there is a two-way process of interaction aprocess of co-evolution (Baum and Singh 1994a Nelson 1994) Thus asFriedland and Alford (1991 244) argue actors develop new institutionalarrangements by recognizing organizational failures and low perfor-mance but these new arrangements do not necessarily change a societyrsquosnorms rules habits and its underlying value systemCampbell (1997) advances our understanding of how existing institu-

tions enable actors to construct radically new institutions He recognizesthat most institutions and institutional arrangements embody a degreeof inertia from which it is difcult for actors to depart Moreover thedifferential power relations among actors make it unlikely that lesspowerful actors can change the existing power structure of a societyMost change in the institutional components of a society evolves througha process of constrained selection which reects to a considerable degreethe existing arrangements and power relations This becomes the inter-pretative frame for social actors with institutionalized scripts and ritualswhich tend to be taken for granted and appear to be quite ordinary andnatural This institutionalized scripting of the social world is an extremelyimportant source of social stability and inertiaCampbell has developed a very fruitful explanation of change in terms

of the various institutional components outlined here He distinguishesbetween incremental and radical institutional change and argues thatradical institutional change occurs when social actors with widelydiffering norms cultural scripts and rituals for action engage in intenseinteraction with each other (see also Knight 1992) The more diverse theinteractions the greater the potential for institutional change He arguesthat as shifts in the composition of interaction among social actors occurchanges in the interpretation of problems and interests will follow Ifthere are only minor changes in the extent to which diverse actorsinteract there will be only modest institutional change But the morefundamental the changes in the interaction of diverse social actors thegreater the change in the way that actors interpret their world And thegreater the change in the way that actors interpret their world the morelikely that radical changes in various institutional components willfollow In other words diverse decision making tends to create a widerrange of interpretative frameworks of the social world than is likely tobe the case if social actors operate in isolation

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

625

This perspective of diverse interactions among social actors facilitatesour ability to move beyond the traditional view that existing institutionsconstrain the range of institutional alternatives which actors face Theolder view has tended to suggest that actors engage in institutionalchange by extending existing institutional principles habits and conven-tions Specifying the conditions which facilitate fundamental institutionalchange permits a much richer perspective for institutional analysisCampbellrsquos perspective about the conditions under which fundamentalor radical change in institutional components occurs is consistent withmy own research about the conditions under which major discoveriesor fundamental new knowledge occur the more scholars with diversebackgrounds interact with intensity and frequency the greater the like-lihood that they will develop new and alternative ways of thinking(Hollingsworth and Hollingsworth 2000)These views of institutional change are also consistent with the recent

work of Nelson and his colleagues (Mowery and Nelson 1999 Murmann1998) They report the research on a variety of industrial sectors in anumber of different countries revealing that networks of actors in partic-ular industries business and professional associations universityresearch departments and institutes are frequently able to use their collec-tive power to modify a societyrsquos property rights research system andgovernmental policies It is the intense and frequent interaction of thesediverse complex sets of actors ndash sometimes constituting a technologysystem ndash that often leads to incremental changes in a societyrsquos total socialsystem of production It is this kind of collective action that brings aboutorganizational change change in institutional sectors institutionalarrangements and even in the rules and norms of a society And if weare to understand how the style of innovativeness of a society changeswe must rst understand the changes in its institutional makeup It ishighly unlikely that a society will develop a new style of innovativenesswithout changing its institutional structure

FIRMS THEIR INSTITUTIONAL ENVIRONMENT ANDNATIONAL SYSTEMS OF INNOVATION

Variation in the innovative style of societies is shaped by their insti-tutional makeup ndash the various levels which are discussed aboveinstitutions institutional arrangements the structure and cohesion of institutional sectors which constitute a societyrsquos social system of pro-duction and the structure and culture of its organizations ndash especiallyits business rms and other research organizations By undertaking aninstitutional analysis of a society one can begin to understand in whatkinds of organizations the production of specic kinds of knowledgetakes place and how this is linked to particular kinds of innovativeness

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

626

With such a perspective one can gain a rich understanding of why somesocieties excel in the production of radical breakthroughs in basic andapplied science and in developing radically new products and why othersocieties excel in more incremental innovationsReecting on these issues one cannot help being impressed with the

contrasts between the innovative styles of German and American orga-nizations For example Germany has continued to be very successful in making incremental innovations in many industries in which it wasvery competitive well before World War II paper printing materialsmachinery electrotechnical products motor vehicles chemicals textileyarns and fabrics But the Germans have been much less successful indeveloping totally new products or related innovations in biotechnologyelectronics telecommunications aircraft as well as other newer indus-trial areas In contrast the Americans have been very innovative in these and other industries with products having very short lives andwith technologies which change very rapidly and are very complex(Hollingsworth 1991 Soskice 1997 1999)The importance of institutions and innovativeness for international

competitiveness has led to the recognition that there are national systemsof innovation (Lundvall 1992 Nelson 1993 Edquist 1997 Hage andHollingsworth 2000) But even before the recent studies of nationalsystems of innovation Landes (1969) was a proponent of these viewswhen he made the argument that Germany had developed a nationalsystem of innovation by building on its education and scientic researchsystems which were lacking in Britain and elsewhere However the concept of a national system of innovation immediately poses theproblem of what kind of innovations should be consideredDavid Soskice an economist at the Wissenschaftszentrum Berlin has

addressed a set of problems somewhat complementary to the perspec-tive raised here His work (Soskice 1990 1997 1999) suggests that weare sensitive to the institutional environment in which the followingactors are embedded the employees of rms the owners and nanciersof rms the competitors of rms and nally their collaborators ndash aperspective complementary to Hollingsworthrsquos work on social systemsof production (Hollingsworth 1997 1998) When one reects on the insti-tutional contextsocial system of production within which these actorsare embedded one becomes aware that those countries which have some-what inexible labor markets external to rms but a business system in which rms have long-term commitments to their employees arestrongly associated with incremental but hardly at all with radical inno-vations In the rms of such countries (eg Germany and Japan) thereis much more consensus decision making than in societies where a lsquohireand rersquo set of practices is quite pervasive And this kind of consensusdecision making ndash while highly conducive to incremental innovativeness

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

627

ndash limits the capacity for radical innovations to occur This kind of lsquolongtermismrsquo set of norms and rules occurs not only in the employmentpractices of incrementally innovative rms but also in their nancingand ownership patterns (Hollingsworth 1997) During the past thirtyyears these were characteristics of many rms in Germany and JapanSignicantly in both of these countries venture capital markets werenot highly developed nancing tended to be based more on long-termbank loans than on the equity markets and the relationships of rmstended to be relatively stable And these stable institutional arrange-ments also contributed to a high degree of innovativeness ndash butincremental rather than radical innovationsMeantime we should also pursue Soskicersquos ideas (1999) about how

rmsrsquo relationships with their competitors and collaborators shape theirstyles of innovativeness In those advanced industrial societies in whichbusiness rms are highly mobilized into business associations theretends to be a high degree of standardized quality controls over technicalnorms and a high degree of reliance on legal sanctions as monitoringinstruments vis-agrave-vis association members As a result of this high degreeof integration of rms into associations member rms are more likelyto be associated with incremental innovations than is likely to be the case with rms in those societies where associations are less welldeveloped In the latter type of society radical innovations are morelikely to occur In those advanced industrial societies where rms havelong-term stable relationships with collaborators in product design andproduction the style of innovation is likely to be incremental And it is in those societies in which there is a weaker tradition of strongcollaboration among rms that radical innovations are more likely tooccurObviously even in a country where there may be a great deal of

radical innovativeness most innovations will be of an incremental typefor there are simply fewer radical innovations But Germany and therest of Europe not only have had relatively few radical innovations incomparison with the United States but they also have lagged behindthe USA in developing relatively totally new industries Thus there hasbeen much less innovative activity in Europe in high-technology andnewly emerging industries than in the USAIn complementary research Audretsch and Feldman (1995) have

demonstrated that the USA has been more successful in spawning rmsengaged in the early stages of the industry life cycle but relative toGermany had an environment unfavorable to innovative activity byestablished rms especially in more mature industries German rmshave tended to be associated with a routinized technological regime ndashwhere established enterprises have had strong incentives to engage inincremental innovations but new rms have tended to have fundamental

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

628

innovative disadvantages relative to the Americans (see Audretsch 1995Winter 1982 Casper 2000)On both sides of the Atlantic universities engage in basic and applied

research But signicantly the institutional environment in which univer-sities are embedded inuences the degree to which the research is likelyto represent major breakthroughs or is likely to have more modestimplications Recent research in biomedical science demonstrates that inthose societies in which university professors andor graduate studentshave the following characteristics there are likely to be relatively majorresearch breakthroughs graduate students have a long training periodhigh dependency on a major professor low autonomy in dening thenature of a research project restricted job options and a university envi-ronment where there is a very modest or nonexistent entrepreneurialculture For example in Germany ndash a country with relatively few radicalbreakthroughs ndash the scholar who nishes a habilitation will have spentmany years on a research project before qualifying for a permanent acad-emic position will have been in a high-dependency position with a seniorprofessor will have had limited autonomy to dene a research projectand will be constrained from pursuing a high-risk research project Theyoung German will have an extraordinarily high level of training butwill have faced a number of disincentives to undertake high-risk researchAnd as soon as heshe has completed the habilitation and becomes aprofessor administrative responsibilities constitute another disincentivefrom engaging in high-risk or long-term research projects Moreover theGerman professor is a civil servant with a more stable and secure incomethan hisher American counterpart In a very different institutionalenvironment American professors in large research universities tend tobecome much more entrepreneurial Today many American professorsbehave as though they were operating their own rms within univer-sities and thus devote considerable time to generating funds for researchprojects assistants secretaries supplies etc And the variation in thisentrepreneurial spirit across countries is associated with variation in thetaking of high-risk research strategies and the making of major researchbreakthroughsIn sum societies vary in terms of the institutional environment in

which their organizations are embedded and this variation inuencesthe degree to which they have industrial activities in the early or latestages of the industry life cycle successful new or mature industriesand radical vs incremental innovations And it is this variation whichis encouraging rms in countries with relatively few new industries and with few or no radical innovations to invest in those countries with many new industries and numerous radical innovations ndash as astrategy to be able to transfer new technologies from the latter institu-tional environment to the former

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

629

Table 4 summarizes the way that the norms rules and values of a society are associated with Americarsquos and Germanyrsquos institutionalarrangements their social system of production and their styles of inno-vativeness

CONCLUSION

Several themes deserve to be reemphasized Most importantly institu-tional analysis is emerging as a dominant issue in several social sciencedisciplines The study of various institutional components discussedabove is a complex subject requiring multiple levels of analysis and aknowledge of multiple sectors within societies It requires a theoreticalperspective that evolves from an empirical analysis in different societiesIf we are to advance the study of institutional analysis we need cross-disciplinary collaboration and team research which are not readilyachieved given the structure of most of our research universitiesAt present scholars in each academic discipline have their own distinc-

tive strategies for studying institutional analysis Thus economists tendto use several approaches (North 1990 Williamson 1985 Hodgson 19881998 1999) anthropologists another (Geertz 1995) and sociologistspolitical scientists and historians employ other strategies (Campbell etal 1991 Powell and DiMaggio 1991 Katznelson 1998 Hall and Taylor1996) With practitioners in various disciplines pursuing their distinctiveapproaches to institutional analysis there is low potential for collabo-rative institutional analysis across the social sciences The lessons fromthe biological sciences over the last 40 years are very instructive oncepractitioners from various biological disciplines began to cooperate instudying biological phenomena by using the same concepts at the mole-cular level theoretical advance was very rapid Comparable phenomenatook place in the development of other hybrid elds eg biochemistrybiophysics (Judson 1979 Olby 1979 Kohler 1982 1994) Similarly weresocial scientists in various disciplines to work collaboratively in the areaof institutional analysis there would very likely be an acceleration intheir theoretical knowledgeThis is not a plea that everyone engaged in institutional analysis should

do the same kind of research Indeed just as biologists working at themolecular level study many different kinds of problems similarly socialscientists engaged in institutional analysis would also work on manyseparate problems but in the spirit of a collective enterprise Some wouldconcentrate their attention on the study of rules norms habits conven-tions and values while others would study how these are associatedwith congurations of various institutional arrangements (types ofmarkets hierarchies networks associations communities clans statesetc) Others would work on specic institutional sectors (eg education

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

630

Table 4

Rules norms hab

its an

d values associated w

ith sp

ecic co

ngu

ration

s of coo

rdinating mecha

nism

s soc

ial sy

stem

s of

prod

uction

an

d nationa

l styles of inno

vation

USA

Germany

Norms rules values which provide incentives for

Norms rules values which provide incentives for

short-term horizons

long-term horizons

Con

gu

rative

forms of

Marke

ts c

orpo

rate hierarchies reg

ulatory state

Assoc

iation

s (busine

ss assoc

iation

s la

bor union

s)

coordinating

corp

oratist-type

networks

marke

ts c

orporatist-

mecha

nism

s (listed in

type

state

order of im

portanc

e in

the particu

lar social

system

)

Social sys

tem of

prod

uction

Bus

iness system

Firm

struc

tures

Con

glom

erates b

ut with mov

emen

t toward

Firm

more co

ngru

ent with pa

rticular prod

ucts

elim

ination of produc

tion

in som

e sectors

emph

asis on long

-term strateg

ies rm

s are well

emph

asis on sh

ort-term

strateg

ies high

integrated

differentiation

between divisions

of rms

Firm

s mob

ilized in

Low

High

and in

tegrated

into busine

ss

associations

Indu

strial relations

Low

deg

ree of jo

b secu

rity rigid in

ternal lab

or

High deg

ree of jo

b secu

rity

exible interna

l lab

or

system

marke

t ex

ible externa

l labo

r marke

tsmarke

t rigid externa

l lab

or m

arke

tPatterns of

Dispersed owne

rship frequ

ent turnov

erMore stab

le owne

rship patterns

owne

rship

Table 4

continue

d

USA

Germany

Norms rules values which provide incentives for

Norms rules values which provide incentives for

short-term horizons

long-term horizons

Social system of

Prod

uction

Bus

iness system

Training for labo

r an

dHigh deg

ree of voluntary decision mak

ing by

Collective an

d m

ore co

mpulso

ry d

ecisions

abo

ut

man

agem

ent

individuals as to ho

w m

uch

tim

e an

d ene

rgy

how m

uch ed

ucation

is ne

eded

he

avy em

pha

sis

to inv

est in educ

ation heavy

emph

asis on

on techn

ical training (eg en

gine

ering)

busine

ss sch

ools training

in marke

ting

and

sales accou

nting

Fina

ncial marke

tsEqu

ity an

d ven

ture cap

ital m

arke

ts highly

Equ

ity an

d ven

ture cap

ital m

arke

ts w

eakly

dev

elop

ed na

ncing from

cen

tral gov

ernm

ent

deve

lope

d he

avy em

pha

sis on

corporate loan

sto priva

te rm

s for dev

elop

men

t of new

tech

nologies related

to na

tion

al security

Unive

rsity research

High individual auton

omy high poten

tial for

Low

ind

ividua

l au

tono

my until almost 40

yea

rs of

system

individual creativity u

nive

rsity pr

ofesso

rs

age poten

tial for ind

ividual creativity high

ly

high

ly entreprene

urial

cons

traine

d unive

rsity pr

ofessors not so

entrep

rene

urial

Inno

vative

styles

Highly su

ccessful in m

ajor break

through

s in

High pe

rforman

ce in proc

ess an

d inc

remen

tal

basic an

d app

lied scien

ce go

od in dev

elop

ing

inno

vation

s in existing produ

cts not so su

ccessful

totally

new

lines of pr

oducts an

d in m

ajor break

through

s in basic scien

ce or in

tech

nologies high

deg

ree of entrepr

eneu

rship

deve

loping

totally new

lines of prod

ucts of

amon

g research

ers

tech

nologies low d

egree of entrepr

eneu

rship

amon

g research

ers

Table 4

continue

d

USA

Germany

Norms rules values which provide incentives for

Norms rules values which provide incentives for

short-term horizons

long-term horizons

Prod

ucts in w

hich

Pr

oducts with sh

ort liv

es prod

ucts in w

hich

Older produ

cts with long

history machine

ry pap

er

society excels

tech

nology

is bo

th highly co

mplex

and

pr

inting

textile

s m

otor veh

icles m

achine

too

lsch

ange

s rapidly en

tertainm

ent indus

try

pha

rmaceu

ticals biotechn

olog

y softw

are

aerosp

ace

business systems nancial markets and systems of research) Distinctivewould be the recognition of how these separate studies are linkedtogether how each institutional sector is linked to a societyrsquos normsrules values etc and its conguration of institutional arrangementsSimilarly other institutional analysts might focus on the study of orga-nizations with a major concern for how the institutional environmentof organizations (eg norms rules values congurations of institutionalarrangements institutional sectors) inuenced the structure culture andoutcomes of organizations And nally others might study how all ofthe aforementioned aspects of institutional analysis inuenced a societyrsquosoverall performance Nelson and Sampat (1998) have appropriatelyreminded us that we should engage in institutional analysis in order toaddress specic problems but the view developed in this article is thatwe rst need some mapping of the terrain of institutional analysis inorder for our diverse enterprises to become a meaningful collective enter-prise especially if we are to understand how institutions inuence theinnovative processThe argument developed here is not that all social scientists should engage

in institutional analysis Obviously there are numerous other importantareas of research which lie beyond the eld of institutional analysis Nordoes this article argue that institutional analysis is the most importantsubject for social scientistsSince the structure and culture of our research universities reect an

enormous amount of disciplinary fragmentation what are the prospectsfor promoting the kind of collective and interdisciplinary research agendaproposed above In one respect the prospects are not encouragingDuring the past half century the efforts of American universities to have effective communication across the biological sciences have variedgreatly in their success At the University of California-Berkeley it was not until the 1980s that there was a serious effort to integrate thebiological sciences and at that time the campus abolished seventeendepartments to promote an integrated biological program But at theUniversity of Wisconsin the biological sciences are still very fragmentedinto dozens of different departments in botany plant pathology zoologyand physiology with two departments of biochemistry and two depart-ments of geneticsThere have been other approaches which have worked very well histor-

ically for promoting scientic integration amidst scientic diversity Forexample several summer schools and special institutes were organizedin the United States during the 1940s in an effort to overcome the frag-mentation in the biological sciences These programs made it possiblefor scientists in diverse academic disciplines and from different univer-sities and countries to come together to engage in intense and frequentinteraction Perhaps the most effective program was that held in the

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

634

summers at Cold Spring Harbor New York where a research agendafor the study of molecular biology was developed (Fischer and Lipson1988 Stent and Watson 1966) Scientists went there summer aftersummer and when they returned to their universities many began therestructuring of the biological sciences in their universitiesIn our own day we need summer institutes and other similar programs

so that scholars from diverse backgrounds can come together anddevelop a collective agenda for doing institutional analysis and forstudying the impact of institutions on the performance of societies ndashinnovativeness being one such subject If we can develop such instituteswe will increase the prospects for transforming our universities Thereis reason to be optimistic that once an eclectic group of scholars workwith a common set of concepts we can have genuine theoretical advanceif researchers can bring their diverse expertise to bear on a common setof problems There is considerable evidence that institutional analysisoffers high potential for advancing our knowledge of innovations andtechnological change (Hollingsworth and Hollingsworth 2000) If we aregoing to make substantial theoretical advance it is not likely to occurfrom the vantage point of a multiplicity of disciplines Rather it is likelyto occur as we develop a new intellectual framework and our own acad-emic journals This is a tall agenda which requires us to reassess theway that knowledge is presently produced in our societies

NOTE

This article is part of a much larger agenda involving the study of institutionsorganizations and innovations in which I am involved My debts are many andvaried The rst version was written while I was in residence in the ResearchUnit for Institutional Change and European Integration of the Austrian Academyof Sciences in Vienna in the summer of 1998 I am especially grateful to ProfessorsEgon Matzner and Sonja Puntscher-Riekmann for providing the stimulating envi-ronment in which to think through the ideas developed here Egon Matzner hasengaged in very stimulating discussions with me about the importance of insti-tutions in the general eld of socio-economics The nal version of the articlewas completed during my residence as a fellow at the Neurosciences Institutein La Jolla California I also wish to thank my colleagues in the InnovationTheme Project at the Netherlands Institute for Advanced Study (NIAS) during1998ndash9 and Edgar Grande and Karl Muumlller for stimulating conversations aboutthe interaction of institutions and organizations in facilitating particular typesof innovation I am especially grateful to Jerald Hage for helping me to developa number of issues in the article and for his useful comments on the rst draftFrans van Waarden early on helped me to work through the idea of multiplelevels of institutional analysis Others who have helped to develop the ideas inthe article have been my colleagues in the research project lsquoComparative socialsystems of productionrsquo especially Tom Burns Christel Lane Yoshitaka OkadaWolfgang Streeck and Richard Whitley My colleagues Robert Boyer SteveCasper and David Soskice have taught me much about how incentives providedby the institutional environments of rms inuence individual decision making

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

635

But my most important debt is to Ellen Jane Hollingsworth who has repeatedlyquestioned every single assumption of this article Without her rich and contin-uous tutoring this article would be even less developed than it is at presentFinally without the insightful assistance of David Gear in many phases of thisarticle it would not have been completed

REFERENCES

Aoki Masahiko (1990) lsquoTowards an economic model of the Japanese rmrsquo Journalof Economic Literature March 1ndash27

Arora A Landau Ralph and Rosenberg Nathan (eds) (1998) Chemicals and Long-Term Economic Growth Insights from the Chemical Industry New York JohnWiley amp Sons

Arrow Kenneth (1974) The Limits of Organization New York NortonArthur W B (1988a) lsquoSelf-reinforcing mechanisms in economicsrsquo in P W

Anderson K J Arrow and D Pines (eds) The Economy as an Evolving ComplexSystem Redwood City Calif Addison-Wesley

mdashmdash(1988b) lsquoCompeting technologies an overviewrsquo in G Dosi C Freeman RNelson G Silverberg and L Soete (eds) Technical Change and Economic TheoryLondon Pinter Publishers pp 590ndash607

Audretsch David B (1995) Innovation and Industry Evolution Cambridge MassMIT Press

Audretsch David B and Feldman Maryann P (1995) lsquoInnovative clusters andthe industry life cyclersquo Discussion Paper Wissenschaftszentrum BerlinMarket Processes and Corporate Development Research Unit FS-IV-95-7

Baum Joel A C and Oliver Christine (1992) lsquoInstitutional embeddedness andthe dynamics of organizational populationsrsquo American Sociological Review 57540ndash59

Baum Joel A C and Singh Jitendra V (1994a) lsquoOrganizationndashenvironment coevo-lutionrsquo in Joel A C Baum and Jitendra V Singh (eds) The EvolutionaryDynamics of Organizations New York Oxford University Press pp 379ndash402

mdashmdash(eds) (1994b) The Evolutionary Dynamics of Organizations New York OxfordUniversity Press

Berger Suzanne and Dore Ronald (eds) (1996) National Diversity and GlobalCapitalism Ithaca NY Cornell University Press

Boyer Robert (1997) lsquoThe variety and unequal performance of really existingmarkets farewell to Doctor Panglossrsquo in J Rogers Hollingsworth and RobertBoyer (eds) Contemporary Capitalism the Embeddedness of InstitutionsCambridge and New York Cambridge University Press pp 55ndash93

Burns Tom R and Dietz Thomas (2001 forthcoming) lsquoSocial rule system theorysocial action institutional arrangements and evolutionary processesrsquo in JRogers Hollingsworth and Karl Muumlller (eds) Advances in SocioeconomicsLanham Md Rowman amp Littleeld

Burns Tom R and Flam Helena (1987) The Shaping of Social Organization BeverlyHills Calif Sage

Calvert Randall L (1995) lsquoThe rational choice theory of social institutionsrsquo inJeffrey S Banks and Eric A Hanuskek (eds) Modern Political Economy NewYork Cambridge University Press pp 216ndash66

Camic Charles (1986) lsquoThe matter of habitrsquo American Journal of Sociology 911039ndash87

Campbell John L (1997) lsquoMechanisms of evolutionary change in economicsgovernance interaction interpretation and bricolagersquo in Lars Magnusson

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

636

the institutional framework discussed below but these other institutionalcomponents feed back and can modify rules norms conventions etc(Murmann 1998)

Second level of analysis

The norms rules habits conventions and values of a society lead to thenext level of analysis ndash the institutional arrangements which are involvedin the coordination of various economic actors producers and suppliersof raw materials knowledge etc processors of raw materials infor-mation workers customers of raw materials nished productsinformation etc nanciers governmental and other types of regulatorsThese actors regularly engage in contests to resolve various economicproblems in virtually all sectors of society how are prices to be setWhat quantity of various products is to be produced How are stan-dards of various products processes etc to be set How is the qualityof products and processes to be determined How are various societalprocesses to be nanced In order to confront these problems and toaddress the conicting positions of economic actors as they address these problems societies develop various institutional (governance)arrangements for coordinating different actors These consist of marketsvarious types of hierarchies and networks associations the state commu-nities and clans (see Hollingsworth and Boyer 1997 Campbell et al1991 ch 1)Each of these particular kinds of coordinating mechanism has many

different types For example there are many types of states (eg theregulatory state developmental state authoritarian state welfare state)on which there is an extensive literature (Kim 1997) Similarly there aredifferent types of markets networks different kinds of associations etc(Boyer 1997 Hage and Alter 1997 Schneiberg and Hollingsworth 1990)When we do institutional analysis we must engage in congurativeanalysis recognizing that actors are not coordinated or governed by asingle type of institutional arrangement Some of the literature discussesindustrial sectors as though they were coordinated or governed by asingle institutional arrangement whereas in fact each sector of aneconomy is coordinated by a conguration of institutional arrangementsSome congurations coordinate actors in certain problem areas whileother congurations of institutional arrangements coordinate actors inaddressing other problems The types of congurations which are domi-nant in a society are somewhat stable and tend to persist over timewithin a society (Hodgson 1999)When one mode of coordination is dominant in a society it will

inuence the role which other coordinating modes will play Hence thestrong role of the state in the Soviet Union inuenced the role of markets

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

605

associations etc in the governance of the Soviet economy Similarly theprominence of particular modes of coordination in a society inuencesits style of innovationMuch of the literature on institutional arrangements (eg forms of eco-

nomic coordination) remains fragmented and unintegrated It is helpfulto array these various forms of coordination in a two-dimensional tax-onomy as in Figure 2 On the vertical dimension the economistrsquos viewof a self-interested agent is contrasted with a more sociological perspec-tive according to which social rules obligation and compliance shapehuman actions On the horizontal dimension there is another continuum(ie the distribution of power) At one extreme of the dimension onends many and relatively equal agents interacting (eg as in a well-organized spot market) At the other extreme inequality in power resultsin a hierarchical form of coordination which structures the interactionbetween principals and agents or between leaders and followers Wherea society falls along the distribution of power is inuenced by the rulesnorms values etc which are dominant in a society at a particular momentin timeInstitutional arrangements can be visualized on two dimensions the

nature of action motive on the one hand and the distribution of poweron the other Markets (cell 1) combine self-interest with horizontalcoordination transactors and they reect sensitivity to concerns aboutsupply and demand thus providing ex post an unintended equilibriumParadoxically the more pure and perfect the market competition thegreater the need for codied rules of the games for coordinating economictransactions Thus collective associations (cell 6) andor various formsof state intervention (cell 4) are required to develop and enforce rulesfor transacting partners (Schneiberg and Hollingsworth 1990 Streeckand Schmitter 1985a) This is an example of how the norms and rulesand the institutional arrangements of a society are intertwined asreected in a conguration of institutional or governance arrangementsAlong the horizontal axis actors can be joined in an organization or

a rm a hierarchy is the generic terms for this institutional arrangement(cell 2) Along the horizontal line one recognizes the difference betweentransactions in a market and transactions within a rm The well-knownworks of Coase (1960 1981) and Williamson (1975 1985) utilize theconcept of transaction costs in explaining the emergence of corporatehierarchiesThere are also various types of networks (cell 5) which exhibit mixes

of self-interest and social obligation with some actors being formallyindependent and equal Yet in some networks (the large rms and theirsubcontractors) there is unequal power and initiative Networks maycomprise all kinds of actors some consist only of rms but others includeassociations and the state (Hage and Alter 1997)

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

606

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

607

DISTRIBUTION OF POWER

VERTICALHORIZONTAL

1 MARKETS 2 HIERARCHY

6ASSOCIATIONS

5NETWORKS

4STATE

3COMMUNITIES

SELF INTEREST

SOCIAL OBLIGATION

ACTION MOTIVE

Figure 2 A general taxonomy of institutional arrangements

The vertical axis deals with action motives Toward the upper part ofFigure 2 actors are engaged in individualistically oriented behaviorwhereas toward the lower part actors are more engaged in collectivebehavior and strive to cope with problems of common interest Cell 3 ndashcommunities and clans ndash consists of institutional arrangements based ontrust reciprocity or obligation and thus not derived from the pure selshcomputation of pleasures and pains This is an unconventional form ofcoordination for most neoclassical economists (however see Arrow1974) but not for many anthropologists political scientists and sociolo-gists (Streeck and Schmitter 1985a Polanyi 1944 Gambetta 1988Fukuyama 1995 Sabel 1992 Putnam 1993)In the neoclassical paradigm theorists argue that actors engage in

forms of exchanges that best promote their individual interests If somestructural conditions are fullled (absence of increasing returns to scalethe reversibility of transactions absence of uncertainty and completecontingent markets with no collusion between actors) then the invis-ible hand theorem applies and market-type activity functions quite welland also provides the optimum for society thereby combining efciencyharmony and order However an excess of market activity may welllead to ruinous competition and excessive conict There is variation inthe extent to which ruinous competition occurs depending on the socialcontext in which transactions take place Thus it is important that webe sensitive to the institutional context in which transactions areembedded and that we understand the degree to which social bondsexist at both the micro and macro levels of analysis Micro bonds facil-itate exchanges in a society but at the societal level social bonds existat the level of the collective ndash in the community or region and amongmembers of racial religious and ethnic groups All other things beingequal the more powerful the social bonds among transacting partnersthe more economic competition is likely to be restrained Thus mosttransactions occur not simply in an impersonal calculative system ofautonomous actors unrestrained by social ties ndash as implied by theneoclassical paradigm ndash but in the context of social ties variation in the strength of which leads to variation in levels of trust and transac-tion costs (Etzioni 1988 211 Granovetter 1985 Streeck and Schmitter1985a Hollingsworth and Boyer 1997 Hodgson 1988 1999 Schneibergand Hollingsworth 1990)Another form of multilateral institutional arrangement is various

types of associations (cell 6) Unlike networks clans and communitiesassociations are more formal organizations Whereas markets corporatehierarchies and networks tend to coordinate economic activity amongdifferent types of actors (eg producers with suppliers capital withlabor) associations typically coordinate actors engaged in the same orsimilar kinds of activities Business associations and labor unions are

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

608

some of the most common forms of associations for coordinatingeconomic activity in capitalist economies (Schneiberg and Hollingsworth1990 Streeck and Schmitter 1985a)Finally there is the state (cell 4) which is an institutional arrangement

quite unlike any of the others It is the state that sanctions and regu-lates the various non-state coordinating mechanisms that is the ultimateenforcer of rules of the various mechanisms that denes and enforcesproperty rights and that manipulates scal and monetary policy Overallit is the state that inuences the total incentive system of a society(Lindberg and Campbell 1991 Johnson 1992 40) At the same time thestate may also be an economic actor by engaging directly in productionand exchange relationsThe choices of institutional arrangements in Figure 2 are constrained

by the social context in which they are embedded Depending on thenature of that embeddedness there is variation in the collective formsof governance some of which are specied in the lower part of thegure Some modes of coordination in the bottom and upper levels ofthe typology are often mixed together though one coordinating modeis likely to be more prominent than another But actors are often engagedin complex congurations involving several kinds of institutionalarrangementsEach of these various institutional arrangements has its own logic ndash

its own rules its own procedures for enforcing compliance its ownnorms and ideologies which help to reduce the costs of enforcementThese are summarized in Table 2 which provides further elaborationabout the various coordinating mechanisms one nds in almost everycapitalist society Table 2 lists the organizational arrangements and theirstructures rules of exchange and means of enforcing compliance asso-ciated with each type of coordinating mechanism While each type ofinstitutional arrangement has various positive features each institutionalarrangement also has failures and these are featured in Table 3 It is thecontest between those who support and those who oppose these variousinstitutional arrangements that tends to lead to transformations in insti-tutional arrangements over time (Campbell et al 1991 Campbell 1997)Again it is difcult to conceive of pure institutional arrangements ndash

either exclusively markets or exclusively hierarchies ndash since it has beenwell known since Adam Smithrsquos Wealth of Nations that the division oflabor within the rm cannot be disentangled from the existence andextent of the market Institutional arrangements have their own distinc-tive form of efciency (good static efciency for the market dynamicefciency for rms) and inefciency often leading to considerableinequality Neither networks nor communities are panaceas for economiccoordination without being congured with other types of institutionalarrangements Networks and communities may solve certain issues but

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

609

Table 2

Log

ics of ins

titution

al arran

gemen

ts

Coordination

Organizational

Rules of

Individual means

Collective means

mechanisms

structure

exchange

of compliance

of compliance

Marke

tsEasy en

try an

d exit

Voluntary sp

ot excha

nge

Leg

al enforcemen

t of

Norm of pr

ivate

Bila

teral ex

chan

ge or

control

prope

rty

marke

t site (W

all Street)

Reg

ulations

to

Leg

itim

acy of m

arke

t en

force contracts

men

tality

Com

mun

ities

Inform

al m

embe

rship

Voluntary exch

ange

based

So

cial norms an

d m

oral

Highly institutiona

lized

ev

olve

s ov

er lon

g period

on social solid

arity an

d

princ

iples im

pose

norm

s an

d rules requ

ire

of tim

ehigh

deg

ree of tru

stob

ligations

mem

bers to accept

Kno

wledg

e of others an

d

lsquocorpo

ratersquoo

bligations

recipr

ocity ov

er tim

e

Networks

Semi-form

al m

embe

rship

Voluntary exch

ange

ove

r a

Con

tractual bon

dsPersona

l relation

sBila

teral or m

ultilateral

time pe

riod

Resou

rce de

pend

ence

Tru

st built outside

the

ex

chan

geecon

omic arena

Associatio

nsFo

rmal m

embe

rship

Restricted to mem

bers

Self-interest

Some deg

ree of

Multilateral excha

nge

Emph

asis on insider

Rep

utation

effects

compu

lsion

outsider or we

they

Priva

te interest type

of

men

tality

gove

rnan

ce

Private

Com

plex organ

izations

Restricted to mem

bers

Rew

ards to ind

ividuals

Highly institutiona

lized

hierarch

ies

which

ten

d to becom

e ex

chan

ge based

on

Asymmetric po

wer threat

rules

bureau

cratic hierarchies

asym

metric po

wer

of san

ctions

Mem

bers soc

ializ

ed into

bureau

cratic rules

corp

orate cu

lture us

e of

sanc

tions

State

Pub

lic hierarchy

Unilateral action

Exit vo

ice (vote lob

bying)

Coe

rcion

De jure

and impo

sed

Indirect political an

d

loya

ltyNorms an

d pu

blic rules

mem

bership

econ

omic excha

nge

Table 3

Failu

res of coo

rdinating mecha

nism

s

Coordinating mechanisms

Type of failure

Associations

Private hierarchies

State

Market

Communities

Networks

Enforcem

ent

Usu

ally relies on

Might enh

ance

Needs co

ntrol

Needs an

internal

Needs

tru

st and

Need an external

the state as an

oppo

rtun

istic

extern

al to state

enforcem

ent

loya

lty often

en

forcem

ent

enforcer

beha

vior

bureau

cracy

authority

coming from

au

thority

Resem

bles

The

ideal o

f (ju

dge

s

Facilitates

outside (fam

ily May

facilitate

enforcem

ent

internal m

arke

ts

parliamen

t collu

sion

and

relig

ion

cartelization an

d mecha

nism

might hur

t marke

t) to

impe

rfect

ethn

icity)

mon

opoly

of cartels

incu

mbe

nt

correct state

compe

titio

nCom

patible with

worke

rsab

uses

variou

s types of

Lob

bies can

co

mpe

tition

captur

e pu

blic

interest goa

ls

Public goo

d

Useful for

Gov

erna

nce co

sts

Can

provide

Can

not provide

Can

internalize

Useful for

and

establishing

might exceed

public goo

ds

colle

ctive go

ods

some co

llective

enha

ncing qu

ality

externality

stan

dards an

d

the be

nets of

but ha

s or deal w

ith

good

s (qua

lity

and training bu

t qu

ality for

internal division

difc

ultie

s in

externalities

training

) bu

t no

t no

t ve

ry goo

d in

setting ru

les of

of la

bor

prov

iding them

Inad

equate

othe

rs (welfare

prov

iding for

compe

titio

n in

Slow

to react to

in precise

mon

itoring of

gene

ral p

ublic

societal gen

eral

the indu

stry

chan

ges in the

am

ounts

tech

nical ch

ange

go

ods)

welfare

Useful for

environm

ent

Might fail in

and inno

vation

Mem

bers tightly

Weak in the

prov

iding

induc

ing

integrated

into

prov

ision of

man

y go

ods

tech

nical

community

colle

ctive go

als

colle

ctively

chan

geha

ve limite

d

that in

dividua

lcapa

city for

mem

bers can

not

inno

vatio

nsprov

ide for

them

selves

Table 3

continue

d

Coordinating mechanisms

Type of failure

Associations

Private hierarchies

State

Market

Communities

Networks

Efcien

cyFa

cilitates

De

cien

t in

Can

be high

ly

Some ba

sic social So

me go

ods

Slow

to en

hanc

e co

operation an

d

coop

eration an

d

bureau

cratic

relation

s cann

ot

cann

ot be

efc

ienc

y an

d

X-efcien

cy but

X-efcien

cyan

d can

not

be provided

by

delivered

at

speed of

not allocative

easily deliver

pure m

arke

t su

fcien

tly

adap

tive

ness

efcien

cy

good

s at

mecha

nism

slow costs

except in

low cost

indus

tries whe

re

tech

nology

is

complex an

d rapidly cha

nging

Equity

Narrow

Excessive

Might enh

ance

Facilitates

Might lead to

Whe

n widely

encompa

ssing

multiplication

ineq

uality

ineq

uality in

retard

ed

dev

elop

ed in

to

associationa

l of con

trollers

(pow

er

income an

d dev

elop

men

tindus

trial

stru

ctures lead

(fru

stration

and

an

d privilege

)wealth

districts

to in

come

ineq

uality)

netw

orks m

ay

equa

lity

facilitate greater

equality an

d

inco

me

distribution

whe

n wea

kly

dev

elop

ed

netw

orks ten

d to

increase soc

ial

ineq

uality

Note For allo

cativ

e efc

ienc

y an

d X

-efcien

cy see Leibe

nstein (19

66 19

76)

they raise other no less severe problems It is important to recognizethe imperfection of any single institutional arrangement in order tocomprehend the origin and transformation of any other institutionalarrangement (Hodgson 1999 ch 3)As suggested above each institutional arrangement is congured with

other institutional arrangements Usually one particular institutionalarrangement is more dominant in a particular conguration but becauseeach type of institutional arrangement has its own strengths andweaknesses there is no simple structural logic in the governance or coor-dination of a society Each institutional arrangement constrains the otherbut the inherent tension among the various institutional arrangementswithin a conguration contributes to changes in the governance orcoordination of a society The routines and logics of each of these cong-urations of institutional arrangements provide constraints and incentivesfor actors The way that actors perceive the incentives and constraintsof these governance congurations leads to a particular market logicand it is the specic market logic of a society which inuences its speciccapacities and weaknesses The inherent strain among the different logicsin a conguration of coordination helps to provide the exibility for asociety to adapt to new circumstancesIf there were a society with pure markets or a society coordinated

only by the state there would be too much rigidity and too little diver-sity to cope with the vast uncertainty in the global environment A societywith very little diversity in its coordination mechanisms would havelittle capacity to adapt to new circumstances Congurations with consid-erable diversity of institutional arrangements provide for a certainamount of incoherence in governance but they also provide for thecapacity to adapt to new circumstances While the Soviet Union wasdominated by the state as a coordinating mechanism there were alwaysfunctioning markets in the system Moreover the feudal society of thetwelfth and thirteenth centuries consisted not only of hierarchical rela-tionships between serfs and masters but also of urban guilds clericalhierarchies and markets (Johnson 1992 38) In sum the robustness ofinstitutions often depends on multiple and diverse principles and logicsof action on the inconsistency of principles and procedures on patternedforms of disorder (Lanzara 1998 Orren and Skowronek 1991 320 329)

Third level of analysis

The next level consists of the institutional sectors of a society The rulesnorms and values of a society inuence the array of institutional arrange-ments and levels one and two inuence the nature of and the relation-ships among various institutional sectors and all three levels areintricately linked together to form a social system of production Together

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

613

all of these components inuence the performance of economic sectorswithin a society as well as the performance of the total society A socialsystem of production is the way that a societyrsquos institutions (see rstlevel of analysis above) its institutional arrangements (see second levelof analysis above) and its institutional sectors are integrated into a socialcongurationAn institutional sector includes all organizations in a society which

supply a given service or product along with their associated focal orga-nizations (eg major suppliers funding sources regulators and so forth(Scott et al 1994 108 117 Campbell et al 1991 Hollingsworth et al1994) Institutional sectors include but are not limited to the societyrsquossystem of education system of research business system nancialmarkets legal system and the state The structure of the nancial marketssystem of training and education industrial relations system system ofresearch and state and legal systems are distinct and idiosyncratic ineach society In short they are system specicAll of these institutional sectors tend to cohere with each other

although they vary in the degree to which they are tightly coupled intoa full-edged system While each of the institutional sectors has someautonomy and may have some goals that are contradictory to the goalsof other institutional sectors with which it is linked an institutional logicin each society leads institutions to coalesce into a complex social con-guration (Hollingsworth 1991) This occurs because the institutionalsectors are embedded in a culture in which their logics are symbolicallygrounded organizationally structured technically and materially con-strained and politically defended The conguration of institutionalsectors usually exhibits some degree of adaptability to new challengesbut continues to evolve within an existing style But under new circum-stances or unprecedented disturbances these institutional congurationsare exposed to sharp historical limits as to what they may or may notdo (Schumpeter 1983 David 1988 Arthur 1988a 1988b Haringkanssen andLundgren 1997)Why do all of these different institutions coalesce into a complex

social conguration which is labeled here as a social system of pro-duction The literature suggests two contrasting interpretations Part ofthe answer ndash indeed a controversial one ndash is that these institutionalsectors are functionally determined by the requirements of the practiceof capitalism in each time and place (Habermas 1975) Another expla-nation emphasizes the genesis of the actual conguration via a trial and error process according to which the survival of rms regions orcountries is the outcome of complex evolutionary mechanisms (Maynard-Smith 1982 Nelson and Winter 1982) However the problem is evenmore complex Markets and other mechanisms for coordinating rela-tionships among economic actors place constraints on the means and

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

614

ends of economic activity to be achieved in any society The othercoordinating mechanisms include different kinds of hierarchies varioustypes of networks and associations ndash trade unions employers and busi-ness associations (Hollingsworth and Lindberg 1985 Campbell et al1991) The logic of these various institutional arrangements providesactors with vocabularies and logics for pursuing their goals for deningwhat is valued and for shaping the norms and rules by which theyabide (see Tables 2 and 3) In short in contrast to the logic of theneoclassical paradigm the argument here is that the dominant type ofinstitutional arrangements places severe constraints on the denition of needs preferences and choices of economic actors Whereas the neo-classical paradigm assumes that individuals and rms are sovereign thisarticle is based on the assumption that rms are inuenced by the hold that the institutional congurations making up a social system ofproduction have on individual decision making (Campbell et al 1991Etzioni 1988 Streeck and Schmitter 1985a Hollingsworth et al 1994Hollingsworth and Boyer 1997 Magnusson and Ottosson 1997 North1990 Hodgson 1999)Standard neoclassical economic theory has tended to downplay the

role of production and consequently of rms Even transaction cost theo-rists who are concerned with analyzing the rm as a coordinatingmechanism have been relatively unconcerned with the various compo-nents of a social system of production Indeed as long as there waswidespread optimism about the efcacy of Keynesian economics therewas relatively little concern among neoclassical economists with thesupply side of the economy Even in the opinion of most Keynesians agroup of experts should ideally be able to shape the size of aggregatedemand while the supply side of the economy would be left to the twominimalist institutions of neoclassical economics ndash markets and manage-rial hierarchies For several decades however it has become increasinglyobvious that some of the most competitive and successful patterns ofindustrial output and industrial production in capitalist economies donot derive from the neoclassical prescription of unregulated markets andcorporate hierarchies complemented by a neoliberal democratic stateIndeed it is now well understood that certain highly successful produc-tion patterns require for their emergence and survival institutionalarrangements and environments the very opposite of the prescriptionsfound in the neoclassical paradigm (see especially Streeck 1992 but alsoHollingsworth and Streeck 1994) Thus social scientists have come tounderstand that if they are to comprehend the behavior and performanceof contemporary economies concerns about social systems of produc-tion must be brought into the pictureBecause production involves more than technology a number of social

scientists have an increasing concern with social systems of production

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

615

The same equipment is frequently operated quite differently in the samesectors in different countries even when rms are competing in the same market (Maurice et al 1980 Sorge and Streeck 1988 Sorge 1989Hollingsworth et al 1994) Variations in production and process tech-nologies are inuenced partly by variations in the social environmentsin which they are embedded In other words rms are embedded intocomplex environments which among other things place constraints ontheir behavior Thus a social system of production is of major impor-tance in understanding the behavior and performance of an economyHow the state and other institutional arrangements (eg marketsnetworks private hierarchies associations) coalesce and are related toparticular social systems of production are important determinants ofeconomic performanceThe congurations of institutional arrangements that coordinate or

govern the behavior of actors in one society and its structure of specicinstitutional sectors cannot easily be transferred to another society forthey are embedded into a social system of production that is societallydistinct (Hollingsworth 1997) Societies borrow selected principles offoreign management styles and work practices but the effectiveness ofsuch borrowing is generally limited Economic behavior and performanceare shaped by the entire social system of production in which actors areembedded and not simply by specic principles of particular manage-ment styles and work practices Moreover a societyrsquos social system ofproduction tends to limit the kind of goods that it can produce and withwhich it can compete successfully in international marketsSome scholars (Kenney and Florida 1993 Oliver and Wilkinson 1988)

have assumed that the diffusion of particular forms of management stylesand work practices across societies could lead to system convergenceReferring to the Japanese production system Kenney and Florida haveargued that it consists of organizational practices whose fundamentallsquogenetic codersquo can successfully be inserted into another society and beginsuccessfully to reproduce its behavior in the new environment (1993 8)Their position is in the intellectual tradition of Antonio Gramsci (1971)who contended decades ago that the American system of mass produc-tion would most certainly diffuse to Europe over timeHowever the argument of this article is that even though British

French and American rms may adopt certain aspects of Japanesemanagement styles (eg just-in-time production self-managing teamsquality circles the use of statistical process controls) or some variant ofthe German vocational system their social systems of production willnot converge Moreover the overall conguration of a societyrsquos socialsystem of production inuences its sectoral and overall national economicperformance This helps to explain why societies have different systemsof innovation and why they vary in the clusters of industries in which

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

616

they are highly competitive in the international markets (Nelson 19931999 Hage and Hollingsworth 2000 Berger and Dore 1996)In order to understand how and why a societyrsquos economy performs

as it does it is necessary to understand its entire social system of produc-tion If a society is to modify substantially the performance of itsinstitutional sectors and its entire economy it cannot adapt only someof the management and work practices of its foreign competitors Ratherit must alter its entire social system of production Because a societyrsquosmodes of economic governance and coordination and its institutionalsectors develop according to particular logics and are system specicthere are serious limitations to the extent to which a society may mimicthe institutions (eg the rules and norms) institutional arrangements andinstitutional sectors of other societiesIn the history of modern societies there are logics by which institu-

tions coalesce into a social system of production (Hollingsworth 1991)Though institutions are constantly changing there are sharp limits tothe type and the direction of change that any particular institution canundergo because of its linkages with institutional arrangements and insti-tutional sectors Thus a societyrsquos business rms educational systemnancial markets industrial relations system etc can engage in seriousrestructuring only if most of the other institutional sectors also changeIn social systems of production there are pressures toward consistencyin the norms rules and values across institutional sectors though in anycomplex society social systems are obviously imperfectly integratedIndeed the degree to which the institutional norms and rules makingup a social system of production are loosely or tightly coupled is a vari-able of considerable importance In general the institutional sectorsmaking up a social system of production are interdependent and changesin one generally result in changes in the others Since each institutionallevel is dependent on the others for various types of resources there isinterdependence among the differing institutional spheres Moreovereach society has its norms moral principles rules and laws and recipesfor action as well as its own idiosyncratic customs traditions and prin-ciples of justice (Burns and Flam 1987)There are also other inherent obstacles to convergence among social

systems of production for where a system is at any one point in timeis inuenced by its initial state Systems with quite different initial statesare unlikely to converge with one anotherrsquos institutional practicesExisting institutional arrangements block certain institutional innova-tions and facilitate others (Roland 1990) Thus the institutions makingup a social system of production provide continuity even though insti-tutional arrangements are always changing but with a particular logicWhile Williamson (1975 1985) suggests that actors tend to choose theinstitutional arrangements which are most efcient North (1990) is

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

617

much closer to the mark in his argument that most societal institutionalarrangements exist as a result of custom and habit and are thereforeinefcient At any moment in time the world often appears to its actorsas very complex and uncertain For this reason actors often engage incontradictory forms of behavior pursuing different strategies as hedgesagainst a very uncertain world (Lanzara 1998) And their hedging andcontradictory forms of behavior may lead in somewhat different soci-etal directions all constrained by the institutional fabric within whichthe actors are embedded This kind of contradictory behavior occurs inpart because of the contradictions inherent in the dominant congura-tions of institutional arrangements with their contradictory logics (seelsquoSecond level of analysisrsquo above)Despite the emphasis on the logic of institutional continuity this is

not an argument that systems change along some predetermined pathThere are critical turning points in the history of highly industrializedsocieties but with the choices limited by the existing institutional terrainBeing path-dependent a social system of production continues along aparticular logic until or unless a fundamental societal crisis intervenes(Milgrom et al 1991 Krugman 1991 Durlauf 1991 Hollingsworth 1991Pred 1966 David 1988 Hodgson 1998)At this point it is important to confront the question of how social

systems of production evolve And this question gets to the heart of theproblem of building complex societal structures Certainly social systemsof production did not emerge from some process of social engineeringMoreover the various component parts of each social system of produc-tion have often not been designed to be part of a social system ofproduction The component parts of each social system of productionhave emerged more often than not as unintended by-products of goalswhich various actors had in mind at earlier moments in time It is usuallythe case that actors at time lsquotrsquo fail to comprehend the long-term conse-quences of their actions Wolfgang Streeck the author of several brilliantpapers on social systems of production argues that institutional sectordesigns created for one purpose generally address that goal but overtime those designs have quite unintended consequences (Streeck 1997a1997b Wright 1998) Describing the conguration of the German socialsystem of production of the 1970s and 1980s Streeck points out that itwas the unintended by-product of multiple points in history Someelements were pre-Wilhelmian others were introduced by the Alliedpowers after 1945 and others emerged during the years of the GermanFederal Republic All component parts of the German social system ofproduction

were and continue to be changing for their own reasons as wellas in reaction to each other and certainly there can be no presump-tion of a pre-established t between them even though one might

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

618

want to allow for some reinforcement effects of the modelrsquos histor-ically contingent social and economic success

(Streeck 1997b 54)

In sum the emergence of social systems of production is a long-termevolutionary process with each part interacting with its environment andresulting in a congurative whole but with no previous design by eithera single actor or a collectivity of actors As suggested above actors do not recreate their world anew At any moment in history the totalinstitutional complex of a society necessarily contains the resources and legacies of its past Preexisting institutional complexes are nevercompletely wiped out Societies consist of multiple layers of history withtheir diverse logics of action However the cumulative effect of smallperipheral changes in altering particular institutional sectors and insti-tutional arrangements can be substantial (Murmann 1998 Lanzara 1998)lsquoInstitution building is affected more by the ways in which people havecodied the past than by how they envision the future There are manymore potential resources in pre-existing institutional arrangements thanit is usually assumed or suspectedrsquo (Lanzara 1998 30)

Fourth level of analysis

The next level of institutional analysis organizational structures is some-what more controversial As suggested above Douglass North (1990)draws a sharp distinction between institutions and organizations Morerecently many organizational theorists (Powell 1991 Powell andDiMaggio 1991 DiMaggio and Powell 1983 Baum and Oliver 1992Kondra and Hinings 1998 Townley 1997) argue that institutional rulesnorms and conventions unfold in tandem with organizational structuresand this is my positionThe literature which focuses on how institutions inuence organiza-

tions is quite different from two other theoretical literatures which alsoare concerned with how organizational environments shape the behaviorof organizations There are resource dependency theorists (Pfeffer andSalancik 1978) who emphasize the role of environmental resources inshaping organizations while Hannan and Freeman (1977 1984) and otherpopulation ecologists emphasize the survival of organizations givencertain kinds of organizational conditions (Orru et al 1991)The institutionalist perspective as implied above emphasizes the

normative environment in which organizations are embedded It is a perspective which focuses on the way in which organizations in their behavior tend to conform to the institutional rules and normswhich are dominant in the organizational environment However allthree perspectives ndash the institutionalist the population ecology and

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

619

the resource dependency ndash emphasize how the environment inuencesorganizations and how organizations which are subject to the sameenvironment tend to converge in their behavior to have what DiMaggioand Powell have labeled organizational isomorphism (DiMaggio andPowell 1983)There has emerged a vast literature which demonstrates that within

each society there are scal political judicial and other regulatory normswhich limit and shape the culture and structure of organizationalbehavior The normative institutional environment of organizations limits the options of what organizations do in a particular society and inuences the patterns of ownership relations with suppliers andcustomers In short it is the normative environment of organizationswhich denes within a particular society what is socially acceptablebehavior for organizations (Hamilton and Biggart 1988 Hollingsworthand Hollingsworth 2000 Meyer and Rowan 1991 Zucker 1987 1988Orru et al 1991 Townley 1997)Thus far most of the analyses on isomorphic pressures operating on

organizations have been cross-sectional in nature However the entireconcept of isomorphism implies that there are strong environmental pres-sures exerted on organizations and in order to observe this phenomenonit is important that we have longitudinal studies which assess howchanges in an organizationrsquos institutional environment inuence changesin the structure and culture of organizations Moreover the expectationhere is that when one analyzes the patterns of organizational changeover time the isomorphism will be holistic ndash that is within a particularsociety there will be a tendency for the internal structure of particulartypes of organizations to converge in large part because of the pres-sures to conform to the changes in the external norms rules and valuesA close reading of Aoki (1990) and Slack and Hinings (1994) nds thatchanges in the national institutional environment of organizations haveinuenced changes in the internal structure of organizations within aspecic industry of a particular societyHowever it is inappropriate to view organizations as changing only

in relation to exogenous environmental change There is a kind of co-evolution which occurs between organizations and their institutionalenvironments Co-evolution implies nonlinear feedback between orga-nizations and their institutional environments This is of course quitedifferent from the more common way of studying organizations bymodeling relations between independent and dependent variables Ofcourse the more common type of analysis is quite appropriate whenthere are simple relationships which do not involve complex feedbackprocesses But in the strategy for institutional analysis which is proposedherein it is less useful to separate independent from dependent vari-ables and more useful to understand the interacting and co-evolutionary

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

620

processes among all of the levels of institutional analysis discussed inthe various sections above (Baum and Singh 1994a 379ndash80 Mowery andNelson 1999 Murmann 1998)The logic of the perspective herein suggests that within every

society there is variation in the structure and culture of business rmsuniversities and other complex organizations (Kondra and Hinings1998) However this variation takes place within parameters which aresystem specic For example every German rm and every Germanuniversity is different from every other German university and rm but there are a set of cultural and structural characteristics which distin-guish German rms and universities from those in America Similarlyeach Japanese rm and university is unique yet they have a set ofcultural and structural characteristics which are system specic andwhich differentiate them from business rms and universities in other countriesIn all societies each organization has its own distinctive organizational

rules norms and conventions which are subordinate to the meta norms and rules of the larger society within which they are embeddedHowever the strength of the institutional environment within whichorganizations are embedded varies from society to society Some societies have multiple institutional environments and there is hetero-geneity in terms of what constitutes appropriate behavior oforganizations In such societies there can be conict over whatinstitutional logics should regulate specic organizational functions(Friedland and Alford 1991 Townley 1997 262ndash3) In those societies inwhich the institutional norms habits and rules are most developed and in which the institutional pressures to conform are greatest thereis less variation in the structure and culture of business rms and variouskinds of research organizations In such societies the connectednessbetween research organizations and their institutional environments is sufciently strong for organizations to have low autonomy to pursue independent strategies and goals in these societies there is a great deal of organizational isomorphism Conversely the weaker the institutional environment in which research organizations areembedded the greater the variation in the structure and culture of business rms and research organizations Moreover where the institu-tional environments are more weakly developed organizations havegreater autonomy and exibility to respond to the development of new knowledge and to be highly innovative Hence it is in those societies where the institutional environments are most developed and most rigid and when there is less organizational autonomy and exibility that fewer radical innovations in basic and applied science as well as in totally new products and industrial sectors havebeen made

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

621

Fifth level of analysis

We come next to the outputs and performance of the various institutionalcomponents of a society It is at this level that institutional componentsare more pragmatic and exible For example in the legal sector thereare specic statutes and court rulings in the state sector there are specicpolicies in the business sector there are new products new technolo-gies and market strategies It is at this nal level of analysis thatinstitutional spheres are most open and susceptible to change and cross-national mimicry is easiest and most commonThrough the outputs of the society we can obtain some assessment as

to how well a society is performing Moreover we can assess how inno-vative it is how egalitarian it is in the distribution of its resources howegalitarian it is in terms of levels of health education etc Just as it iscomplicated to measure how egalitarian a society is similarly it is dif-cult to assess how innovative it is For example societies vary in makingincremental and radical breakthroughs in basic and applied science in developing totally new products and new kinds of organizations inincremental and process innovations in existing products and organiza-tions in developing new and different forms of marketing ndash bothdomestically and globally (Hage and Hollingsworth 2000) Of coursesocieties vary in their rates of economic growth in their rates of economicproductivity in their quality of life (eg rates of crime life expectancyetc) All of these performance criteria feed back and inuence each ofthe levels of institutional analysis discussed above ndash rules norms valuesetc institutional arrangements institutional sectors the structure andform of organizations Moreover performance measures may inuenceeach other For example if a society has very low rates of growth itmay not be very innovative in some of the types of innovative activitiesmentioned above Good or poor performances of certain kinds inuenceother performance indicators (Hollingsworth et al 1990)Different institutional arrangements and different social systems of

production result in different types of economic performance Hence solong as societies have different social systems of production there areserious constraints on the degree to which they can converge in theirinnovative styles Different social systems of production tend to maxi-mize in a more or less explicit manner different performance criteriausually mixing considerations about static and dynamic efciency protsecurity social peace and economic andor political power In short incontrast to the implications of neoclassical economic theory in real-worldeconomies there are no universal standards all economically rationalactors attempt to maximize Economic history provides numerous exam-ples of how a great variety of principles of rationality are implementedin different societies

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

622

A critical question in institutional analysis is whether an existing socialsystem of production which supports a set of routines for a particularkind of technology and industry can shift from old practices to newones There are numerous historical accounts of how the social systemof production in a particular society which worked so well for a numberof years could not adapt to new technologies (Lazonick 1990 Veblen1915 Nelson 1994) Schumpeter (1983) and Freeman (1991) have devel-oped the idea of a techno-economic paradigm which suggests thatdifferent technologies require specic forms of organizational arrange-ments ndash or what I called social systems of production Because oflsquoinstitutional inertiarsquo the social system of production of a society maynot be able to adapt to a new techno-economic paradigm (Nelson 199458 Lanzara 1998)Whether a social system of production can sustain its performance

standards depends not only on its intrinsic economic rationality but alsoon where it ts into a larger system If a particular social system ofproduction is immune from the innovativeness and competition of analternative system survival can be long lasting But if different socialsystems of production with diverging criteria of good economic perfor-mance meet in the world arena the arbitrariness of nationally imposedconstructed performance standards may be superseded by alternativeperformance criteria as a result of international competitivenessThe world economy is also socially constructed just as are national

economies Even if different social systems of production are competingin the international arena it is not always possible to determine whichis more competitively effective at any moment in time Hegemonicnation-states in the short run shape the rules of trade to favor their indus-trial sectors and rms But the history of hegemonic powers suggeststhat in the longer run such social systems of production sustainedlargely by military and political power eventually give way to moreinnovative and competitive social systems of production In our ownday as nation-states are increasingly integrated into a world economyeven hegemonic powers lose their innovative and competitive advan-tage and their share of world output decreases Such a country mayattempt to restructure its institutional arrangements and to readjust itsperformance preferences But to restructure its system of innovationgenerally calls for a major redistribution of power within a societyLargely for this reason societies have historically had limited capacityto reconstruct a system of innovation in the image of their major competi-tors Each countryrsquos social system of production is a conguration of ahost of norms rules and values as well as of institutional arrangementsEach system is constantly changing and is open to inuence from othersystems And indeed many technologies and practices diffuse from onesociety to another but a societyrsquos capacity to be innovative is constrained

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

623

by the existing social system of production Thus the same technologymay exist in numerous countries but how it is employed and how itinuences societal outcomes and performances varies from one institu-tional conguration or society to another (Hollingsworth 1998)

PERSPECTIVES ON INSTITUTIONAL CHANGE

A fundamental problem which remains unresolved in institutionalanalysis involves the nature of institutional change There is much confu-sion and miscommunication in the social sciences about institutionalchange in large part because we do not even know how to measure therate of institutional change Of course we know that there are bothexternal and internal focuses for change and we have had a good bitof scholarship on this subject What is not clearly understood is thatinstitutions and institutional arrangements within societies are histori-cally rooted that there is a great deal of path-dependency to the waythat institutions evolve Moreover the more intricately linked each sectoris with each other sector and with a societyrsquos rules and norms the lesschoice actors have to devise new institutions and institutional arrange-ments On the other hand if a sector or actor is too isolated it may betoo weak within a society to bring about any kind of effective changeThese considerations are very critical to the problem of how muchfreedom actors have to create new institutions afresh to what extentinstitutions diffuse from one society to another (Murmann 1998 Lanzara1998) This is a fundamental question for students of institutionalanalysis as we attempt to understand how the different institutionallevels are linked togetherThere are a variety of reasons why there is confusion in studying insti-

tutional change There is disagreement over the extent to which actorshave the freedom to build new institutional arrangements and if theydo to what extent the new arrangements and organizational patternsmay depart from past practices Among some institutional economists(Williamson 1985) there is a basic assumption that actors build efcientinstitutions by pursuing their self-interest and by promoting theirstrategic goals Moreover there is the assumption that a mixing of fullrationality with market competition tends to produce all of the optimalinstitutions that are needed in order to coordinate a complex capitalisteconomy (Hollingsworth and Boyer 1997 ch 14 Hodgson 1999) Incontrast the argument developed here is much more complexSocial institutions are historically rooted and there is a great deal of

path-dependency in the way that various institutional componentsevolve The shape of institutional congurations at any moment limitsthe type of options for change Because there is a great deal of institu-tional inertia radical change in the institutional components of a society

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

624

is uncommon (Lanzara 1998) Recognizing the constraints existinginstitutions exert on actors we note that existing congurations of insti-tutional arrangements institutional sectors and organizational propertiescan limit but also inuence the degree to which norms and rules canchange Many scholars (North 1990 Ostrom 1990) have emphasized therole of norms rules ideologies and values in limiting the ability of actorsto develop new institutional arrangements which radically depart fromexisting ones However there is a two-way process of interaction aprocess of co-evolution (Baum and Singh 1994a Nelson 1994) Thus asFriedland and Alford (1991 244) argue actors develop new institutionalarrangements by recognizing organizational failures and low perfor-mance but these new arrangements do not necessarily change a societyrsquosnorms rules habits and its underlying value systemCampbell (1997) advances our understanding of how existing institu-

tions enable actors to construct radically new institutions He recognizesthat most institutions and institutional arrangements embody a degreeof inertia from which it is difcult for actors to depart Moreover thedifferential power relations among actors make it unlikely that lesspowerful actors can change the existing power structure of a societyMost change in the institutional components of a society evolves througha process of constrained selection which reects to a considerable degreethe existing arrangements and power relations This becomes the inter-pretative frame for social actors with institutionalized scripts and ritualswhich tend to be taken for granted and appear to be quite ordinary andnatural This institutionalized scripting of the social world is an extremelyimportant source of social stability and inertiaCampbell has developed a very fruitful explanation of change in terms

of the various institutional components outlined here He distinguishesbetween incremental and radical institutional change and argues thatradical institutional change occurs when social actors with widelydiffering norms cultural scripts and rituals for action engage in intenseinteraction with each other (see also Knight 1992) The more diverse theinteractions the greater the potential for institutional change He arguesthat as shifts in the composition of interaction among social actors occurchanges in the interpretation of problems and interests will follow Ifthere are only minor changes in the extent to which diverse actorsinteract there will be only modest institutional change But the morefundamental the changes in the interaction of diverse social actors thegreater the change in the way that actors interpret their world And thegreater the change in the way that actors interpret their world the morelikely that radical changes in various institutional components willfollow In other words diverse decision making tends to create a widerrange of interpretative frameworks of the social world than is likely tobe the case if social actors operate in isolation

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

625

This perspective of diverse interactions among social actors facilitatesour ability to move beyond the traditional view that existing institutionsconstrain the range of institutional alternatives which actors face Theolder view has tended to suggest that actors engage in institutionalchange by extending existing institutional principles habits and conven-tions Specifying the conditions which facilitate fundamental institutionalchange permits a much richer perspective for institutional analysisCampbellrsquos perspective about the conditions under which fundamentalor radical change in institutional components occurs is consistent withmy own research about the conditions under which major discoveriesor fundamental new knowledge occur the more scholars with diversebackgrounds interact with intensity and frequency the greater the like-lihood that they will develop new and alternative ways of thinking(Hollingsworth and Hollingsworth 2000)These views of institutional change are also consistent with the recent

work of Nelson and his colleagues (Mowery and Nelson 1999 Murmann1998) They report the research on a variety of industrial sectors in anumber of different countries revealing that networks of actors in partic-ular industries business and professional associations universityresearch departments and institutes are frequently able to use their collec-tive power to modify a societyrsquos property rights research system andgovernmental policies It is the intense and frequent interaction of thesediverse complex sets of actors ndash sometimes constituting a technologysystem ndash that often leads to incremental changes in a societyrsquos total socialsystem of production It is this kind of collective action that brings aboutorganizational change change in institutional sectors institutionalarrangements and even in the rules and norms of a society And if weare to understand how the style of innovativeness of a society changeswe must rst understand the changes in its institutional makeup It ishighly unlikely that a society will develop a new style of innovativenesswithout changing its institutional structure

FIRMS THEIR INSTITUTIONAL ENVIRONMENT ANDNATIONAL SYSTEMS OF INNOVATION

Variation in the innovative style of societies is shaped by their insti-tutional makeup ndash the various levels which are discussed aboveinstitutions institutional arrangements the structure and cohesion of institutional sectors which constitute a societyrsquos social system of pro-duction and the structure and culture of its organizations ndash especiallyits business rms and other research organizations By undertaking aninstitutional analysis of a society one can begin to understand in whatkinds of organizations the production of specic kinds of knowledgetakes place and how this is linked to particular kinds of innovativeness

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

626

With such a perspective one can gain a rich understanding of why somesocieties excel in the production of radical breakthroughs in basic andapplied science and in developing radically new products and why othersocieties excel in more incremental innovationsReecting on these issues one cannot help being impressed with the

contrasts between the innovative styles of German and American orga-nizations For example Germany has continued to be very successful in making incremental innovations in many industries in which it wasvery competitive well before World War II paper printing materialsmachinery electrotechnical products motor vehicles chemicals textileyarns and fabrics But the Germans have been much less successful indeveloping totally new products or related innovations in biotechnologyelectronics telecommunications aircraft as well as other newer indus-trial areas In contrast the Americans have been very innovative in these and other industries with products having very short lives andwith technologies which change very rapidly and are very complex(Hollingsworth 1991 Soskice 1997 1999)The importance of institutions and innovativeness for international

competitiveness has led to the recognition that there are national systemsof innovation (Lundvall 1992 Nelson 1993 Edquist 1997 Hage andHollingsworth 2000) But even before the recent studies of nationalsystems of innovation Landes (1969) was a proponent of these viewswhen he made the argument that Germany had developed a nationalsystem of innovation by building on its education and scientic researchsystems which were lacking in Britain and elsewhere However the concept of a national system of innovation immediately poses theproblem of what kind of innovations should be consideredDavid Soskice an economist at the Wissenschaftszentrum Berlin has

addressed a set of problems somewhat complementary to the perspec-tive raised here His work (Soskice 1990 1997 1999) suggests that weare sensitive to the institutional environment in which the followingactors are embedded the employees of rms the owners and nanciersof rms the competitors of rms and nally their collaborators ndash aperspective complementary to Hollingsworthrsquos work on social systemsof production (Hollingsworth 1997 1998) When one reects on the insti-tutional contextsocial system of production within which these actorsare embedded one becomes aware that those countries which have some-what inexible labor markets external to rms but a business system in which rms have long-term commitments to their employees arestrongly associated with incremental but hardly at all with radical inno-vations In the rms of such countries (eg Germany and Japan) thereis much more consensus decision making than in societies where a lsquohireand rersquo set of practices is quite pervasive And this kind of consensusdecision making ndash while highly conducive to incremental innovativeness

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

627

ndash limits the capacity for radical innovations to occur This kind of lsquolongtermismrsquo set of norms and rules occurs not only in the employmentpractices of incrementally innovative rms but also in their nancingand ownership patterns (Hollingsworth 1997) During the past thirtyyears these were characteristics of many rms in Germany and JapanSignicantly in both of these countries venture capital markets werenot highly developed nancing tended to be based more on long-termbank loans than on the equity markets and the relationships of rmstended to be relatively stable And these stable institutional arrange-ments also contributed to a high degree of innovativeness ndash butincremental rather than radical innovationsMeantime we should also pursue Soskicersquos ideas (1999) about how

rmsrsquo relationships with their competitors and collaborators shape theirstyles of innovativeness In those advanced industrial societies in whichbusiness rms are highly mobilized into business associations theretends to be a high degree of standardized quality controls over technicalnorms and a high degree of reliance on legal sanctions as monitoringinstruments vis-agrave-vis association members As a result of this high degreeof integration of rms into associations member rms are more likelyto be associated with incremental innovations than is likely to be the case with rms in those societies where associations are less welldeveloped In the latter type of society radical innovations are morelikely to occur In those advanced industrial societies where rms havelong-term stable relationships with collaborators in product design andproduction the style of innovation is likely to be incremental And it is in those societies in which there is a weaker tradition of strongcollaboration among rms that radical innovations are more likely tooccurObviously even in a country where there may be a great deal of

radical innovativeness most innovations will be of an incremental typefor there are simply fewer radical innovations But Germany and therest of Europe not only have had relatively few radical innovations incomparison with the United States but they also have lagged behindthe USA in developing relatively totally new industries Thus there hasbeen much less innovative activity in Europe in high-technology andnewly emerging industries than in the USAIn complementary research Audretsch and Feldman (1995) have

demonstrated that the USA has been more successful in spawning rmsengaged in the early stages of the industry life cycle but relative toGermany had an environment unfavorable to innovative activity byestablished rms especially in more mature industries German rmshave tended to be associated with a routinized technological regime ndashwhere established enterprises have had strong incentives to engage inincremental innovations but new rms have tended to have fundamental

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

628

innovative disadvantages relative to the Americans (see Audretsch 1995Winter 1982 Casper 2000)On both sides of the Atlantic universities engage in basic and applied

research But signicantly the institutional environment in which univer-sities are embedded inuences the degree to which the research is likelyto represent major breakthroughs or is likely to have more modestimplications Recent research in biomedical science demonstrates that inthose societies in which university professors andor graduate studentshave the following characteristics there are likely to be relatively majorresearch breakthroughs graduate students have a long training periodhigh dependency on a major professor low autonomy in dening thenature of a research project restricted job options and a university envi-ronment where there is a very modest or nonexistent entrepreneurialculture For example in Germany ndash a country with relatively few radicalbreakthroughs ndash the scholar who nishes a habilitation will have spentmany years on a research project before qualifying for a permanent acad-emic position will have been in a high-dependency position with a seniorprofessor will have had limited autonomy to dene a research projectand will be constrained from pursuing a high-risk research project Theyoung German will have an extraordinarily high level of training butwill have faced a number of disincentives to undertake high-risk researchAnd as soon as heshe has completed the habilitation and becomes aprofessor administrative responsibilities constitute another disincentivefrom engaging in high-risk or long-term research projects Moreover theGerman professor is a civil servant with a more stable and secure incomethan hisher American counterpart In a very different institutionalenvironment American professors in large research universities tend tobecome much more entrepreneurial Today many American professorsbehave as though they were operating their own rms within univer-sities and thus devote considerable time to generating funds for researchprojects assistants secretaries supplies etc And the variation in thisentrepreneurial spirit across countries is associated with variation in thetaking of high-risk research strategies and the making of major researchbreakthroughsIn sum societies vary in terms of the institutional environment in

which their organizations are embedded and this variation inuencesthe degree to which they have industrial activities in the early or latestages of the industry life cycle successful new or mature industriesand radical vs incremental innovations And it is this variation whichis encouraging rms in countries with relatively few new industries and with few or no radical innovations to invest in those countries with many new industries and numerous radical innovations ndash as astrategy to be able to transfer new technologies from the latter institu-tional environment to the former

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

629

Table 4 summarizes the way that the norms rules and values of a society are associated with Americarsquos and Germanyrsquos institutionalarrangements their social system of production and their styles of inno-vativeness

CONCLUSION

Several themes deserve to be reemphasized Most importantly institu-tional analysis is emerging as a dominant issue in several social sciencedisciplines The study of various institutional components discussedabove is a complex subject requiring multiple levels of analysis and aknowledge of multiple sectors within societies It requires a theoreticalperspective that evolves from an empirical analysis in different societiesIf we are to advance the study of institutional analysis we need cross-disciplinary collaboration and team research which are not readilyachieved given the structure of most of our research universitiesAt present scholars in each academic discipline have their own distinc-

tive strategies for studying institutional analysis Thus economists tendto use several approaches (North 1990 Williamson 1985 Hodgson 19881998 1999) anthropologists another (Geertz 1995) and sociologistspolitical scientists and historians employ other strategies (Campbell etal 1991 Powell and DiMaggio 1991 Katznelson 1998 Hall and Taylor1996) With practitioners in various disciplines pursuing their distinctiveapproaches to institutional analysis there is low potential for collabo-rative institutional analysis across the social sciences The lessons fromthe biological sciences over the last 40 years are very instructive oncepractitioners from various biological disciplines began to cooperate instudying biological phenomena by using the same concepts at the mole-cular level theoretical advance was very rapid Comparable phenomenatook place in the development of other hybrid elds eg biochemistrybiophysics (Judson 1979 Olby 1979 Kohler 1982 1994) Similarly weresocial scientists in various disciplines to work collaboratively in the areaof institutional analysis there would very likely be an acceleration intheir theoretical knowledgeThis is not a plea that everyone engaged in institutional analysis should

do the same kind of research Indeed just as biologists working at themolecular level study many different kinds of problems similarly socialscientists engaged in institutional analysis would also work on manyseparate problems but in the spirit of a collective enterprise Some wouldconcentrate their attention on the study of rules norms habits conven-tions and values while others would study how these are associatedwith congurations of various institutional arrangements (types ofmarkets hierarchies networks associations communities clans statesetc) Others would work on specic institutional sectors (eg education

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

630

Table 4

Rules norms hab

its an

d values associated w

ith sp

ecic co

ngu

ration

s of coo

rdinating mecha

nism

s soc

ial sy

stem

s of

prod

uction

an

d nationa

l styles of inno

vation

USA

Germany

Norms rules values which provide incentives for

Norms rules values which provide incentives for

short-term horizons

long-term horizons

Con

gu

rative

forms of

Marke

ts c

orpo

rate hierarchies reg

ulatory state

Assoc

iation

s (busine

ss assoc

iation

s la

bor union

s)

coordinating

corp

oratist-type

networks

marke

ts c

orporatist-

mecha

nism

s (listed in

type

state

order of im

portanc

e in

the particu

lar social

system

)

Social sys

tem of

prod

uction

Bus

iness system

Firm

struc

tures

Con

glom

erates b

ut with mov

emen

t toward

Firm

more co

ngru

ent with pa

rticular prod

ucts

elim

ination of produc

tion

in som

e sectors

emph

asis on long

-term strateg

ies rm

s are well

emph

asis on sh

ort-term

strateg

ies high

integrated

differentiation

between divisions

of rms

Firm

s mob

ilized in

Low

High

and in

tegrated

into busine

ss

associations

Indu

strial relations

Low

deg

ree of jo

b secu

rity rigid in

ternal lab

or

High deg

ree of jo

b secu

rity

exible interna

l lab

or

system

marke

t ex

ible externa

l labo

r marke

tsmarke

t rigid externa

l lab

or m

arke

tPatterns of

Dispersed owne

rship frequ

ent turnov

erMore stab

le owne

rship patterns

owne

rship

Table 4

continue

d

USA

Germany

Norms rules values which provide incentives for

Norms rules values which provide incentives for

short-term horizons

long-term horizons

Social system of

Prod

uction

Bus

iness system

Training for labo

r an

dHigh deg

ree of voluntary decision mak

ing by

Collective an

d m

ore co

mpulso

ry d

ecisions

abo

ut

man

agem

ent

individuals as to ho

w m

uch

tim

e an

d ene

rgy

how m

uch ed

ucation

is ne

eded

he

avy em

pha

sis

to inv

est in educ

ation heavy

emph

asis on

on techn

ical training (eg en

gine

ering)

busine

ss sch

ools training

in marke

ting

and

sales accou

nting

Fina

ncial marke

tsEqu

ity an

d ven

ture cap

ital m

arke

ts highly

Equ

ity an

d ven

ture cap

ital m

arke

ts w

eakly

dev

elop

ed na

ncing from

cen

tral gov

ernm

ent

deve

lope

d he

avy em

pha

sis on

corporate loan

sto priva

te rm

s for dev

elop

men

t of new

tech

nologies related

to na

tion

al security

Unive

rsity research

High individual auton

omy high poten

tial for

Low

ind

ividua

l au

tono

my until almost 40

yea

rs of

system

individual creativity u

nive

rsity pr

ofesso

rs

age poten

tial for ind

ividual creativity high

ly

high

ly entreprene

urial

cons

traine

d unive

rsity pr

ofessors not so

entrep

rene

urial

Inno

vative

styles

Highly su

ccessful in m

ajor break

through

s in

High pe

rforman

ce in proc

ess an

d inc

remen

tal

basic an

d app

lied scien

ce go

od in dev

elop

ing

inno

vation

s in existing produ

cts not so su

ccessful

totally

new

lines of pr

oducts an

d in m

ajor break

through

s in basic scien

ce or in

tech

nologies high

deg

ree of entrepr

eneu

rship

deve

loping

totally new

lines of prod

ucts of

amon

g research

ers

tech

nologies low d

egree of entrepr

eneu

rship

amon

g research

ers

Table 4

continue

d

USA

Germany

Norms rules values which provide incentives for

Norms rules values which provide incentives for

short-term horizons

long-term horizons

Prod

ucts in w

hich

Pr

oducts with sh

ort liv

es prod

ucts in w

hich

Older produ

cts with long

history machine

ry pap

er

society excels

tech

nology

is bo

th highly co

mplex

and

pr

inting

textile

s m

otor veh

icles m

achine

too

lsch

ange

s rapidly en

tertainm

ent indus

try

pha

rmaceu

ticals biotechn

olog

y softw

are

aerosp

ace

business systems nancial markets and systems of research) Distinctivewould be the recognition of how these separate studies are linkedtogether how each institutional sector is linked to a societyrsquos normsrules values etc and its conguration of institutional arrangementsSimilarly other institutional analysts might focus on the study of orga-nizations with a major concern for how the institutional environmentof organizations (eg norms rules values congurations of institutionalarrangements institutional sectors) inuenced the structure culture andoutcomes of organizations And nally others might study how all ofthe aforementioned aspects of institutional analysis inuenced a societyrsquosoverall performance Nelson and Sampat (1998) have appropriatelyreminded us that we should engage in institutional analysis in order toaddress specic problems but the view developed in this article is thatwe rst need some mapping of the terrain of institutional analysis inorder for our diverse enterprises to become a meaningful collective enter-prise especially if we are to understand how institutions inuence theinnovative processThe argument developed here is not that all social scientists should engage

in institutional analysis Obviously there are numerous other importantareas of research which lie beyond the eld of institutional analysis Nordoes this article argue that institutional analysis is the most importantsubject for social scientistsSince the structure and culture of our research universities reect an

enormous amount of disciplinary fragmentation what are the prospectsfor promoting the kind of collective and interdisciplinary research agendaproposed above In one respect the prospects are not encouragingDuring the past half century the efforts of American universities to have effective communication across the biological sciences have variedgreatly in their success At the University of California-Berkeley it was not until the 1980s that there was a serious effort to integrate thebiological sciences and at that time the campus abolished seventeendepartments to promote an integrated biological program But at theUniversity of Wisconsin the biological sciences are still very fragmentedinto dozens of different departments in botany plant pathology zoologyand physiology with two departments of biochemistry and two depart-ments of geneticsThere have been other approaches which have worked very well histor-

ically for promoting scientic integration amidst scientic diversity Forexample several summer schools and special institutes were organizedin the United States during the 1940s in an effort to overcome the frag-mentation in the biological sciences These programs made it possiblefor scientists in diverse academic disciplines and from different univer-sities and countries to come together to engage in intense and frequentinteraction Perhaps the most effective program was that held in the

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

634

summers at Cold Spring Harbor New York where a research agendafor the study of molecular biology was developed (Fischer and Lipson1988 Stent and Watson 1966) Scientists went there summer aftersummer and when they returned to their universities many began therestructuring of the biological sciences in their universitiesIn our own day we need summer institutes and other similar programs

so that scholars from diverse backgrounds can come together anddevelop a collective agenda for doing institutional analysis and forstudying the impact of institutions on the performance of societies ndashinnovativeness being one such subject If we can develop such instituteswe will increase the prospects for transforming our universities Thereis reason to be optimistic that once an eclectic group of scholars workwith a common set of concepts we can have genuine theoretical advanceif researchers can bring their diverse expertise to bear on a common setof problems There is considerable evidence that institutional analysisoffers high potential for advancing our knowledge of innovations andtechnological change (Hollingsworth and Hollingsworth 2000) If we aregoing to make substantial theoretical advance it is not likely to occurfrom the vantage point of a multiplicity of disciplines Rather it is likelyto occur as we develop a new intellectual framework and our own acad-emic journals This is a tall agenda which requires us to reassess theway that knowledge is presently produced in our societies

NOTE

This article is part of a much larger agenda involving the study of institutionsorganizations and innovations in which I am involved My debts are many andvaried The rst version was written while I was in residence in the ResearchUnit for Institutional Change and European Integration of the Austrian Academyof Sciences in Vienna in the summer of 1998 I am especially grateful to ProfessorsEgon Matzner and Sonja Puntscher-Riekmann for providing the stimulating envi-ronment in which to think through the ideas developed here Egon Matzner hasengaged in very stimulating discussions with me about the importance of insti-tutions in the general eld of socio-economics The nal version of the articlewas completed during my residence as a fellow at the Neurosciences Institutein La Jolla California I also wish to thank my colleagues in the InnovationTheme Project at the Netherlands Institute for Advanced Study (NIAS) during1998ndash9 and Edgar Grande and Karl Muumlller for stimulating conversations aboutthe interaction of institutions and organizations in facilitating particular typesof innovation I am especially grateful to Jerald Hage for helping me to developa number of issues in the article and for his useful comments on the rst draftFrans van Waarden early on helped me to work through the idea of multiplelevels of institutional analysis Others who have helped to develop the ideas inthe article have been my colleagues in the research project lsquoComparative socialsystems of productionrsquo especially Tom Burns Christel Lane Yoshitaka OkadaWolfgang Streeck and Richard Whitley My colleagues Robert Boyer SteveCasper and David Soskice have taught me much about how incentives providedby the institutional environments of rms inuence individual decision making

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

635

But my most important debt is to Ellen Jane Hollingsworth who has repeatedlyquestioned every single assumption of this article Without her rich and contin-uous tutoring this article would be even less developed than it is at presentFinally without the insightful assistance of David Gear in many phases of thisarticle it would not have been completed

REFERENCES

Aoki Masahiko (1990) lsquoTowards an economic model of the Japanese rmrsquo Journalof Economic Literature March 1ndash27

Arora A Landau Ralph and Rosenberg Nathan (eds) (1998) Chemicals and Long-Term Economic Growth Insights from the Chemical Industry New York JohnWiley amp Sons

Arrow Kenneth (1974) The Limits of Organization New York NortonArthur W B (1988a) lsquoSelf-reinforcing mechanisms in economicsrsquo in P W

Anderson K J Arrow and D Pines (eds) The Economy as an Evolving ComplexSystem Redwood City Calif Addison-Wesley

mdashmdash(1988b) lsquoCompeting technologies an overviewrsquo in G Dosi C Freeman RNelson G Silverberg and L Soete (eds) Technical Change and Economic TheoryLondon Pinter Publishers pp 590ndash607

Audretsch David B (1995) Innovation and Industry Evolution Cambridge MassMIT Press

Audretsch David B and Feldman Maryann P (1995) lsquoInnovative clusters andthe industry life cyclersquo Discussion Paper Wissenschaftszentrum BerlinMarket Processes and Corporate Development Research Unit FS-IV-95-7

Baum Joel A C and Oliver Christine (1992) lsquoInstitutional embeddedness andthe dynamics of organizational populationsrsquo American Sociological Review 57540ndash59

Baum Joel A C and Singh Jitendra V (1994a) lsquoOrganizationndashenvironment coevo-lutionrsquo in Joel A C Baum and Jitendra V Singh (eds) The EvolutionaryDynamics of Organizations New York Oxford University Press pp 379ndash402

mdashmdash(eds) (1994b) The Evolutionary Dynamics of Organizations New York OxfordUniversity Press

Berger Suzanne and Dore Ronald (eds) (1996) National Diversity and GlobalCapitalism Ithaca NY Cornell University Press

Boyer Robert (1997) lsquoThe variety and unequal performance of really existingmarkets farewell to Doctor Panglossrsquo in J Rogers Hollingsworth and RobertBoyer (eds) Contemporary Capitalism the Embeddedness of InstitutionsCambridge and New York Cambridge University Press pp 55ndash93

Burns Tom R and Dietz Thomas (2001 forthcoming) lsquoSocial rule system theorysocial action institutional arrangements and evolutionary processesrsquo in JRogers Hollingsworth and Karl Muumlller (eds) Advances in SocioeconomicsLanham Md Rowman amp Littleeld

Burns Tom R and Flam Helena (1987) The Shaping of Social Organization BeverlyHills Calif Sage

Calvert Randall L (1995) lsquoThe rational choice theory of social institutionsrsquo inJeffrey S Banks and Eric A Hanuskek (eds) Modern Political Economy NewYork Cambridge University Press pp 216ndash66

Camic Charles (1986) lsquoThe matter of habitrsquo American Journal of Sociology 911039ndash87

Campbell John L (1997) lsquoMechanisms of evolutionary change in economicsgovernance interaction interpretation and bricolagersquo in Lars Magnusson

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

636

associations etc in the governance of the Soviet economy Similarly theprominence of particular modes of coordination in a society inuencesits style of innovationMuch of the literature on institutional arrangements (eg forms of eco-

nomic coordination) remains fragmented and unintegrated It is helpfulto array these various forms of coordination in a two-dimensional tax-onomy as in Figure 2 On the vertical dimension the economistrsquos viewof a self-interested agent is contrasted with a more sociological perspec-tive according to which social rules obligation and compliance shapehuman actions On the horizontal dimension there is another continuum(ie the distribution of power) At one extreme of the dimension onends many and relatively equal agents interacting (eg as in a well-organized spot market) At the other extreme inequality in power resultsin a hierarchical form of coordination which structures the interactionbetween principals and agents or between leaders and followers Wherea society falls along the distribution of power is inuenced by the rulesnorms values etc which are dominant in a society at a particular momentin timeInstitutional arrangements can be visualized on two dimensions the

nature of action motive on the one hand and the distribution of poweron the other Markets (cell 1) combine self-interest with horizontalcoordination transactors and they reect sensitivity to concerns aboutsupply and demand thus providing ex post an unintended equilibriumParadoxically the more pure and perfect the market competition thegreater the need for codied rules of the games for coordinating economictransactions Thus collective associations (cell 6) andor various formsof state intervention (cell 4) are required to develop and enforce rulesfor transacting partners (Schneiberg and Hollingsworth 1990 Streeckand Schmitter 1985a) This is an example of how the norms and rulesand the institutional arrangements of a society are intertwined asreected in a conguration of institutional or governance arrangementsAlong the horizontal axis actors can be joined in an organization or

a rm a hierarchy is the generic terms for this institutional arrangement(cell 2) Along the horizontal line one recognizes the difference betweentransactions in a market and transactions within a rm The well-knownworks of Coase (1960 1981) and Williamson (1975 1985) utilize theconcept of transaction costs in explaining the emergence of corporatehierarchiesThere are also various types of networks (cell 5) which exhibit mixes

of self-interest and social obligation with some actors being formallyindependent and equal Yet in some networks (the large rms and theirsubcontractors) there is unequal power and initiative Networks maycomprise all kinds of actors some consist only of rms but others includeassociations and the state (Hage and Alter 1997)

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

606

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

607

DISTRIBUTION OF POWER

VERTICALHORIZONTAL

1 MARKETS 2 HIERARCHY

6ASSOCIATIONS

5NETWORKS

4STATE

3COMMUNITIES

SELF INTEREST

SOCIAL OBLIGATION

ACTION MOTIVE

Figure 2 A general taxonomy of institutional arrangements

The vertical axis deals with action motives Toward the upper part ofFigure 2 actors are engaged in individualistically oriented behaviorwhereas toward the lower part actors are more engaged in collectivebehavior and strive to cope with problems of common interest Cell 3 ndashcommunities and clans ndash consists of institutional arrangements based ontrust reciprocity or obligation and thus not derived from the pure selshcomputation of pleasures and pains This is an unconventional form ofcoordination for most neoclassical economists (however see Arrow1974) but not for many anthropologists political scientists and sociolo-gists (Streeck and Schmitter 1985a Polanyi 1944 Gambetta 1988Fukuyama 1995 Sabel 1992 Putnam 1993)In the neoclassical paradigm theorists argue that actors engage in

forms of exchanges that best promote their individual interests If somestructural conditions are fullled (absence of increasing returns to scalethe reversibility of transactions absence of uncertainty and completecontingent markets with no collusion between actors) then the invis-ible hand theorem applies and market-type activity functions quite welland also provides the optimum for society thereby combining efciencyharmony and order However an excess of market activity may welllead to ruinous competition and excessive conict There is variation inthe extent to which ruinous competition occurs depending on the socialcontext in which transactions take place Thus it is important that webe sensitive to the institutional context in which transactions areembedded and that we understand the degree to which social bondsexist at both the micro and macro levels of analysis Micro bonds facil-itate exchanges in a society but at the societal level social bonds existat the level of the collective ndash in the community or region and amongmembers of racial religious and ethnic groups All other things beingequal the more powerful the social bonds among transacting partnersthe more economic competition is likely to be restrained Thus mosttransactions occur not simply in an impersonal calculative system ofautonomous actors unrestrained by social ties ndash as implied by theneoclassical paradigm ndash but in the context of social ties variation in the strength of which leads to variation in levels of trust and transac-tion costs (Etzioni 1988 211 Granovetter 1985 Streeck and Schmitter1985a Hollingsworth and Boyer 1997 Hodgson 1988 1999 Schneibergand Hollingsworth 1990)Another form of multilateral institutional arrangement is various

types of associations (cell 6) Unlike networks clans and communitiesassociations are more formal organizations Whereas markets corporatehierarchies and networks tend to coordinate economic activity amongdifferent types of actors (eg producers with suppliers capital withlabor) associations typically coordinate actors engaged in the same orsimilar kinds of activities Business associations and labor unions are

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

608

some of the most common forms of associations for coordinatingeconomic activity in capitalist economies (Schneiberg and Hollingsworth1990 Streeck and Schmitter 1985a)Finally there is the state (cell 4) which is an institutional arrangement

quite unlike any of the others It is the state that sanctions and regu-lates the various non-state coordinating mechanisms that is the ultimateenforcer of rules of the various mechanisms that denes and enforcesproperty rights and that manipulates scal and monetary policy Overallit is the state that inuences the total incentive system of a society(Lindberg and Campbell 1991 Johnson 1992 40) At the same time thestate may also be an economic actor by engaging directly in productionand exchange relationsThe choices of institutional arrangements in Figure 2 are constrained

by the social context in which they are embedded Depending on thenature of that embeddedness there is variation in the collective formsof governance some of which are specied in the lower part of thegure Some modes of coordination in the bottom and upper levels ofthe typology are often mixed together though one coordinating modeis likely to be more prominent than another But actors are often engagedin complex congurations involving several kinds of institutionalarrangementsEach of these various institutional arrangements has its own logic ndash

its own rules its own procedures for enforcing compliance its ownnorms and ideologies which help to reduce the costs of enforcementThese are summarized in Table 2 which provides further elaborationabout the various coordinating mechanisms one nds in almost everycapitalist society Table 2 lists the organizational arrangements and theirstructures rules of exchange and means of enforcing compliance asso-ciated with each type of coordinating mechanism While each type ofinstitutional arrangement has various positive features each institutionalarrangement also has failures and these are featured in Table 3 It is thecontest between those who support and those who oppose these variousinstitutional arrangements that tends to lead to transformations in insti-tutional arrangements over time (Campbell et al 1991 Campbell 1997)Again it is difcult to conceive of pure institutional arrangements ndash

either exclusively markets or exclusively hierarchies ndash since it has beenwell known since Adam Smithrsquos Wealth of Nations that the division oflabor within the rm cannot be disentangled from the existence andextent of the market Institutional arrangements have their own distinc-tive form of efciency (good static efciency for the market dynamicefciency for rms) and inefciency often leading to considerableinequality Neither networks nor communities are panaceas for economiccoordination without being congured with other types of institutionalarrangements Networks and communities may solve certain issues but

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

609

Table 2

Log

ics of ins

titution

al arran

gemen

ts

Coordination

Organizational

Rules of

Individual means

Collective means

mechanisms

structure

exchange

of compliance

of compliance

Marke

tsEasy en

try an

d exit

Voluntary sp

ot excha

nge

Leg

al enforcemen

t of

Norm of pr

ivate

Bila

teral ex

chan

ge or

control

prope

rty

marke

t site (W

all Street)

Reg

ulations

to

Leg

itim

acy of m

arke

t en

force contracts

men

tality

Com

mun

ities

Inform

al m

embe

rship

Voluntary exch

ange

based

So

cial norms an

d m

oral

Highly institutiona

lized

ev

olve

s ov

er lon

g period

on social solid

arity an

d

princ

iples im

pose

norm

s an

d rules requ

ire

of tim

ehigh

deg

ree of tru

stob

ligations

mem

bers to accept

Kno

wledg

e of others an

d

lsquocorpo

ratersquoo

bligations

recipr

ocity ov

er tim

e

Networks

Semi-form

al m

embe

rship

Voluntary exch

ange

ove

r a

Con

tractual bon

dsPersona

l relation

sBila

teral or m

ultilateral

time pe

riod

Resou

rce de

pend

ence

Tru

st built outside

the

ex

chan

geecon

omic arena

Associatio

nsFo

rmal m

embe

rship

Restricted to mem

bers

Self-interest

Some deg

ree of

Multilateral excha

nge

Emph

asis on insider

Rep

utation

effects

compu

lsion

outsider or we

they

Priva

te interest type

of

men

tality

gove

rnan

ce

Private

Com

plex organ

izations

Restricted to mem

bers

Rew

ards to ind

ividuals

Highly institutiona

lized

hierarch

ies

which

ten

d to becom

e ex

chan

ge based

on

Asymmetric po

wer threat

rules

bureau

cratic hierarchies

asym

metric po

wer

of san

ctions

Mem

bers soc

ializ

ed into

bureau

cratic rules

corp

orate cu

lture us

e of

sanc

tions

State

Pub

lic hierarchy

Unilateral action

Exit vo

ice (vote lob

bying)

Coe

rcion

De jure

and impo

sed

Indirect political an

d

loya

ltyNorms an

d pu

blic rules

mem

bership

econ

omic excha

nge

Table 3

Failu

res of coo

rdinating mecha

nism

s

Coordinating mechanisms

Type of failure

Associations

Private hierarchies

State

Market

Communities

Networks

Enforcem

ent

Usu

ally relies on

Might enh

ance

Needs co

ntrol

Needs an

internal

Needs

tru

st and

Need an external

the state as an

oppo

rtun

istic

extern

al to state

enforcem

ent

loya

lty often

en

forcem

ent

enforcer

beha

vior

bureau

cracy

authority

coming from

au

thority

Resem

bles

The

ideal o

f (ju

dge

s

Facilitates

outside (fam

ily May

facilitate

enforcem

ent

internal m

arke

ts

parliamen

t collu

sion

and

relig

ion

cartelization an

d mecha

nism

might hur

t marke

t) to

impe

rfect

ethn

icity)

mon

opoly

of cartels

incu

mbe

nt

correct state

compe

titio

nCom

patible with

worke

rsab

uses

variou

s types of

Lob

bies can

co

mpe

tition

captur

e pu

blic

interest goa

ls

Public goo

d

Useful for

Gov

erna

nce co

sts

Can

provide

Can

not provide

Can

internalize

Useful for

and

establishing

might exceed

public goo

ds

colle

ctive go

ods

some co

llective

enha

ncing qu

ality

externality

stan

dards an

d

the be

nets of

but ha

s or deal w

ith

good

s (qua

lity

and training bu

t qu

ality for

internal division

difc

ultie

s in

externalities

training

) bu

t no

t no

t ve

ry goo

d in

setting ru

les of

of la

bor

prov

iding them

Inad

equate

othe

rs (welfare

prov

iding for

compe

titio

n in

Slow

to react to

in precise

mon

itoring of

gene

ral p

ublic

societal gen

eral

the indu

stry

chan

ges in the

am

ounts

tech

nical ch

ange

go

ods)

welfare

Useful for

environm

ent

Might fail in

and inno

vation

Mem

bers tightly

Weak in the

prov

iding

induc

ing

integrated

into

prov

ision of

man

y go

ods

tech

nical

community

colle

ctive go

als

colle

ctively

chan

geha

ve limite

d

that in

dividua

lcapa

city for

mem

bers can

not

inno

vatio

nsprov

ide for

them

selves

Table 3

continue

d

Coordinating mechanisms

Type of failure

Associations

Private hierarchies

State

Market

Communities

Networks

Efcien

cyFa

cilitates

De

cien

t in

Can

be high

ly

Some ba

sic social So

me go

ods

Slow

to en

hanc

e co

operation an

d

coop

eration an

d

bureau

cratic

relation

s cann

ot

cann

ot be

efc

ienc

y an

d

X-efcien

cy but

X-efcien

cyan

d can

not

be provided

by

delivered

at

speed of

not allocative

easily deliver

pure m

arke

t su

fcien

tly

adap

tive

ness

efcien

cy

good

s at

mecha

nism

slow costs

except in

low cost

indus

tries whe

re

tech

nology

is

complex an

d rapidly cha

nging

Equity

Narrow

Excessive

Might enh

ance

Facilitates

Might lead to

Whe

n widely

encompa

ssing

multiplication

ineq

uality

ineq

uality in

retard

ed

dev

elop

ed in

to

associationa

l of con

trollers

(pow

er

income an

d dev

elop

men

tindus

trial

stru

ctures lead

(fru

stration

and

an

d privilege

)wealth

districts

to in

come

ineq

uality)

netw

orks m

ay

equa

lity

facilitate greater

equality an

d

inco

me

distribution

whe

n wea

kly

dev

elop

ed

netw

orks ten

d to

increase soc

ial

ineq

uality

Note For allo

cativ

e efc

ienc

y an

d X

-efcien

cy see Leibe

nstein (19

66 19

76)

they raise other no less severe problems It is important to recognizethe imperfection of any single institutional arrangement in order tocomprehend the origin and transformation of any other institutionalarrangement (Hodgson 1999 ch 3)As suggested above each institutional arrangement is congured with

other institutional arrangements Usually one particular institutionalarrangement is more dominant in a particular conguration but becauseeach type of institutional arrangement has its own strengths andweaknesses there is no simple structural logic in the governance or coor-dination of a society Each institutional arrangement constrains the otherbut the inherent tension among the various institutional arrangementswithin a conguration contributes to changes in the governance orcoordination of a society The routines and logics of each of these cong-urations of institutional arrangements provide constraints and incentivesfor actors The way that actors perceive the incentives and constraintsof these governance congurations leads to a particular market logicand it is the specic market logic of a society which inuences its speciccapacities and weaknesses The inherent strain among the different logicsin a conguration of coordination helps to provide the exibility for asociety to adapt to new circumstancesIf there were a society with pure markets or a society coordinated

only by the state there would be too much rigidity and too little diver-sity to cope with the vast uncertainty in the global environment A societywith very little diversity in its coordination mechanisms would havelittle capacity to adapt to new circumstances Congurations with consid-erable diversity of institutional arrangements provide for a certainamount of incoherence in governance but they also provide for thecapacity to adapt to new circumstances While the Soviet Union wasdominated by the state as a coordinating mechanism there were alwaysfunctioning markets in the system Moreover the feudal society of thetwelfth and thirteenth centuries consisted not only of hierarchical rela-tionships between serfs and masters but also of urban guilds clericalhierarchies and markets (Johnson 1992 38) In sum the robustness ofinstitutions often depends on multiple and diverse principles and logicsof action on the inconsistency of principles and procedures on patternedforms of disorder (Lanzara 1998 Orren and Skowronek 1991 320 329)

Third level of analysis

The next level consists of the institutional sectors of a society The rulesnorms and values of a society inuence the array of institutional arrange-ments and levels one and two inuence the nature of and the relation-ships among various institutional sectors and all three levels areintricately linked together to form a social system of production Together

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

613

all of these components inuence the performance of economic sectorswithin a society as well as the performance of the total society A socialsystem of production is the way that a societyrsquos institutions (see rstlevel of analysis above) its institutional arrangements (see second levelof analysis above) and its institutional sectors are integrated into a socialcongurationAn institutional sector includes all organizations in a society which

supply a given service or product along with their associated focal orga-nizations (eg major suppliers funding sources regulators and so forth(Scott et al 1994 108 117 Campbell et al 1991 Hollingsworth et al1994) Institutional sectors include but are not limited to the societyrsquossystem of education system of research business system nancialmarkets legal system and the state The structure of the nancial marketssystem of training and education industrial relations system system ofresearch and state and legal systems are distinct and idiosyncratic ineach society In short they are system specicAll of these institutional sectors tend to cohere with each other

although they vary in the degree to which they are tightly coupled intoa full-edged system While each of the institutional sectors has someautonomy and may have some goals that are contradictory to the goalsof other institutional sectors with which it is linked an institutional logicin each society leads institutions to coalesce into a complex social con-guration (Hollingsworth 1991) This occurs because the institutionalsectors are embedded in a culture in which their logics are symbolicallygrounded organizationally structured technically and materially con-strained and politically defended The conguration of institutionalsectors usually exhibits some degree of adaptability to new challengesbut continues to evolve within an existing style But under new circum-stances or unprecedented disturbances these institutional congurationsare exposed to sharp historical limits as to what they may or may notdo (Schumpeter 1983 David 1988 Arthur 1988a 1988b Haringkanssen andLundgren 1997)Why do all of these different institutions coalesce into a complex

social conguration which is labeled here as a social system of pro-duction The literature suggests two contrasting interpretations Part ofthe answer ndash indeed a controversial one ndash is that these institutionalsectors are functionally determined by the requirements of the practiceof capitalism in each time and place (Habermas 1975) Another expla-nation emphasizes the genesis of the actual conguration via a trial and error process according to which the survival of rms regions orcountries is the outcome of complex evolutionary mechanisms (Maynard-Smith 1982 Nelson and Winter 1982) However the problem is evenmore complex Markets and other mechanisms for coordinating rela-tionships among economic actors place constraints on the means and

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

614

ends of economic activity to be achieved in any society The othercoordinating mechanisms include different kinds of hierarchies varioustypes of networks and associations ndash trade unions employers and busi-ness associations (Hollingsworth and Lindberg 1985 Campbell et al1991) The logic of these various institutional arrangements providesactors with vocabularies and logics for pursuing their goals for deningwhat is valued and for shaping the norms and rules by which theyabide (see Tables 2 and 3) In short in contrast to the logic of theneoclassical paradigm the argument here is that the dominant type ofinstitutional arrangements places severe constraints on the denition of needs preferences and choices of economic actors Whereas the neo-classical paradigm assumes that individuals and rms are sovereign thisarticle is based on the assumption that rms are inuenced by the hold that the institutional congurations making up a social system ofproduction have on individual decision making (Campbell et al 1991Etzioni 1988 Streeck and Schmitter 1985a Hollingsworth et al 1994Hollingsworth and Boyer 1997 Magnusson and Ottosson 1997 North1990 Hodgson 1999)Standard neoclassical economic theory has tended to downplay the

role of production and consequently of rms Even transaction cost theo-rists who are concerned with analyzing the rm as a coordinatingmechanism have been relatively unconcerned with the various compo-nents of a social system of production Indeed as long as there waswidespread optimism about the efcacy of Keynesian economics therewas relatively little concern among neoclassical economists with thesupply side of the economy Even in the opinion of most Keynesians agroup of experts should ideally be able to shape the size of aggregatedemand while the supply side of the economy would be left to the twominimalist institutions of neoclassical economics ndash markets and manage-rial hierarchies For several decades however it has become increasinglyobvious that some of the most competitive and successful patterns ofindustrial output and industrial production in capitalist economies donot derive from the neoclassical prescription of unregulated markets andcorporate hierarchies complemented by a neoliberal democratic stateIndeed it is now well understood that certain highly successful produc-tion patterns require for their emergence and survival institutionalarrangements and environments the very opposite of the prescriptionsfound in the neoclassical paradigm (see especially Streeck 1992 but alsoHollingsworth and Streeck 1994) Thus social scientists have come tounderstand that if they are to comprehend the behavior and performanceof contemporary economies concerns about social systems of produc-tion must be brought into the pictureBecause production involves more than technology a number of social

scientists have an increasing concern with social systems of production

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

615

The same equipment is frequently operated quite differently in the samesectors in different countries even when rms are competing in the same market (Maurice et al 1980 Sorge and Streeck 1988 Sorge 1989Hollingsworth et al 1994) Variations in production and process tech-nologies are inuenced partly by variations in the social environmentsin which they are embedded In other words rms are embedded intocomplex environments which among other things place constraints ontheir behavior Thus a social system of production is of major impor-tance in understanding the behavior and performance of an economyHow the state and other institutional arrangements (eg marketsnetworks private hierarchies associations) coalesce and are related toparticular social systems of production are important determinants ofeconomic performanceThe congurations of institutional arrangements that coordinate or

govern the behavior of actors in one society and its structure of specicinstitutional sectors cannot easily be transferred to another society forthey are embedded into a social system of production that is societallydistinct (Hollingsworth 1997) Societies borrow selected principles offoreign management styles and work practices but the effectiveness ofsuch borrowing is generally limited Economic behavior and performanceare shaped by the entire social system of production in which actors areembedded and not simply by specic principles of particular manage-ment styles and work practices Moreover a societyrsquos social system ofproduction tends to limit the kind of goods that it can produce and withwhich it can compete successfully in international marketsSome scholars (Kenney and Florida 1993 Oliver and Wilkinson 1988)

have assumed that the diffusion of particular forms of management stylesand work practices across societies could lead to system convergenceReferring to the Japanese production system Kenney and Florida haveargued that it consists of organizational practices whose fundamentallsquogenetic codersquo can successfully be inserted into another society and beginsuccessfully to reproduce its behavior in the new environment (1993 8)Their position is in the intellectual tradition of Antonio Gramsci (1971)who contended decades ago that the American system of mass produc-tion would most certainly diffuse to Europe over timeHowever the argument of this article is that even though British

French and American rms may adopt certain aspects of Japanesemanagement styles (eg just-in-time production self-managing teamsquality circles the use of statistical process controls) or some variant ofthe German vocational system their social systems of production willnot converge Moreover the overall conguration of a societyrsquos socialsystem of production inuences its sectoral and overall national economicperformance This helps to explain why societies have different systemsof innovation and why they vary in the clusters of industries in which

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

616

they are highly competitive in the international markets (Nelson 19931999 Hage and Hollingsworth 2000 Berger and Dore 1996)In order to understand how and why a societyrsquos economy performs

as it does it is necessary to understand its entire social system of produc-tion If a society is to modify substantially the performance of itsinstitutional sectors and its entire economy it cannot adapt only someof the management and work practices of its foreign competitors Ratherit must alter its entire social system of production Because a societyrsquosmodes of economic governance and coordination and its institutionalsectors develop according to particular logics and are system specicthere are serious limitations to the extent to which a society may mimicthe institutions (eg the rules and norms) institutional arrangements andinstitutional sectors of other societiesIn the history of modern societies there are logics by which institu-

tions coalesce into a social system of production (Hollingsworth 1991)Though institutions are constantly changing there are sharp limits tothe type and the direction of change that any particular institution canundergo because of its linkages with institutional arrangements and insti-tutional sectors Thus a societyrsquos business rms educational systemnancial markets industrial relations system etc can engage in seriousrestructuring only if most of the other institutional sectors also changeIn social systems of production there are pressures toward consistencyin the norms rules and values across institutional sectors though in anycomplex society social systems are obviously imperfectly integratedIndeed the degree to which the institutional norms and rules makingup a social system of production are loosely or tightly coupled is a vari-able of considerable importance In general the institutional sectorsmaking up a social system of production are interdependent and changesin one generally result in changes in the others Since each institutionallevel is dependent on the others for various types of resources there isinterdependence among the differing institutional spheres Moreovereach society has its norms moral principles rules and laws and recipesfor action as well as its own idiosyncratic customs traditions and prin-ciples of justice (Burns and Flam 1987)There are also other inherent obstacles to convergence among social

systems of production for where a system is at any one point in timeis inuenced by its initial state Systems with quite different initial statesare unlikely to converge with one anotherrsquos institutional practicesExisting institutional arrangements block certain institutional innova-tions and facilitate others (Roland 1990) Thus the institutions makingup a social system of production provide continuity even though insti-tutional arrangements are always changing but with a particular logicWhile Williamson (1975 1985) suggests that actors tend to choose theinstitutional arrangements which are most efcient North (1990) is

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

617

much closer to the mark in his argument that most societal institutionalarrangements exist as a result of custom and habit and are thereforeinefcient At any moment in time the world often appears to its actorsas very complex and uncertain For this reason actors often engage incontradictory forms of behavior pursuing different strategies as hedgesagainst a very uncertain world (Lanzara 1998) And their hedging andcontradictory forms of behavior may lead in somewhat different soci-etal directions all constrained by the institutional fabric within whichthe actors are embedded This kind of contradictory behavior occurs inpart because of the contradictions inherent in the dominant congura-tions of institutional arrangements with their contradictory logics (seelsquoSecond level of analysisrsquo above)Despite the emphasis on the logic of institutional continuity this is

not an argument that systems change along some predetermined pathThere are critical turning points in the history of highly industrializedsocieties but with the choices limited by the existing institutional terrainBeing path-dependent a social system of production continues along aparticular logic until or unless a fundamental societal crisis intervenes(Milgrom et al 1991 Krugman 1991 Durlauf 1991 Hollingsworth 1991Pred 1966 David 1988 Hodgson 1998)At this point it is important to confront the question of how social

systems of production evolve And this question gets to the heart of theproblem of building complex societal structures Certainly social systemsof production did not emerge from some process of social engineeringMoreover the various component parts of each social system of produc-tion have often not been designed to be part of a social system ofproduction The component parts of each social system of productionhave emerged more often than not as unintended by-products of goalswhich various actors had in mind at earlier moments in time It is usuallythe case that actors at time lsquotrsquo fail to comprehend the long-term conse-quences of their actions Wolfgang Streeck the author of several brilliantpapers on social systems of production argues that institutional sectordesigns created for one purpose generally address that goal but overtime those designs have quite unintended consequences (Streeck 1997a1997b Wright 1998) Describing the conguration of the German socialsystem of production of the 1970s and 1980s Streeck points out that itwas the unintended by-product of multiple points in history Someelements were pre-Wilhelmian others were introduced by the Alliedpowers after 1945 and others emerged during the years of the GermanFederal Republic All component parts of the German social system ofproduction

were and continue to be changing for their own reasons as wellas in reaction to each other and certainly there can be no presump-tion of a pre-established t between them even though one might

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

618

want to allow for some reinforcement effects of the modelrsquos histor-ically contingent social and economic success

(Streeck 1997b 54)

In sum the emergence of social systems of production is a long-termevolutionary process with each part interacting with its environment andresulting in a congurative whole but with no previous design by eithera single actor or a collectivity of actors As suggested above actors do not recreate their world anew At any moment in history the totalinstitutional complex of a society necessarily contains the resources and legacies of its past Preexisting institutional complexes are nevercompletely wiped out Societies consist of multiple layers of history withtheir diverse logics of action However the cumulative effect of smallperipheral changes in altering particular institutional sectors and insti-tutional arrangements can be substantial (Murmann 1998 Lanzara 1998)lsquoInstitution building is affected more by the ways in which people havecodied the past than by how they envision the future There are manymore potential resources in pre-existing institutional arrangements thanit is usually assumed or suspectedrsquo (Lanzara 1998 30)

Fourth level of analysis

The next level of institutional analysis organizational structures is some-what more controversial As suggested above Douglass North (1990)draws a sharp distinction between institutions and organizations Morerecently many organizational theorists (Powell 1991 Powell andDiMaggio 1991 DiMaggio and Powell 1983 Baum and Oliver 1992Kondra and Hinings 1998 Townley 1997) argue that institutional rulesnorms and conventions unfold in tandem with organizational structuresand this is my positionThe literature which focuses on how institutions inuence organiza-

tions is quite different from two other theoretical literatures which alsoare concerned with how organizational environments shape the behaviorof organizations There are resource dependency theorists (Pfeffer andSalancik 1978) who emphasize the role of environmental resources inshaping organizations while Hannan and Freeman (1977 1984) and otherpopulation ecologists emphasize the survival of organizations givencertain kinds of organizational conditions (Orru et al 1991)The institutionalist perspective as implied above emphasizes the

normative environment in which organizations are embedded It is a perspective which focuses on the way in which organizations in their behavior tend to conform to the institutional rules and normswhich are dominant in the organizational environment However allthree perspectives ndash the institutionalist the population ecology and

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

619

the resource dependency ndash emphasize how the environment inuencesorganizations and how organizations which are subject to the sameenvironment tend to converge in their behavior to have what DiMaggioand Powell have labeled organizational isomorphism (DiMaggio andPowell 1983)There has emerged a vast literature which demonstrates that within

each society there are scal political judicial and other regulatory normswhich limit and shape the culture and structure of organizationalbehavior The normative institutional environment of organizations limits the options of what organizations do in a particular society and inuences the patterns of ownership relations with suppliers andcustomers In short it is the normative environment of organizationswhich denes within a particular society what is socially acceptablebehavior for organizations (Hamilton and Biggart 1988 Hollingsworthand Hollingsworth 2000 Meyer and Rowan 1991 Zucker 1987 1988Orru et al 1991 Townley 1997)Thus far most of the analyses on isomorphic pressures operating on

organizations have been cross-sectional in nature However the entireconcept of isomorphism implies that there are strong environmental pres-sures exerted on organizations and in order to observe this phenomenonit is important that we have longitudinal studies which assess howchanges in an organizationrsquos institutional environment inuence changesin the structure and culture of organizations Moreover the expectationhere is that when one analyzes the patterns of organizational changeover time the isomorphism will be holistic ndash that is within a particularsociety there will be a tendency for the internal structure of particulartypes of organizations to converge in large part because of the pres-sures to conform to the changes in the external norms rules and valuesA close reading of Aoki (1990) and Slack and Hinings (1994) nds thatchanges in the national institutional environment of organizations haveinuenced changes in the internal structure of organizations within aspecic industry of a particular societyHowever it is inappropriate to view organizations as changing only

in relation to exogenous environmental change There is a kind of co-evolution which occurs between organizations and their institutionalenvironments Co-evolution implies nonlinear feedback between orga-nizations and their institutional environments This is of course quitedifferent from the more common way of studying organizations bymodeling relations between independent and dependent variables Ofcourse the more common type of analysis is quite appropriate whenthere are simple relationships which do not involve complex feedbackprocesses But in the strategy for institutional analysis which is proposedherein it is less useful to separate independent from dependent vari-ables and more useful to understand the interacting and co-evolutionary

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

620

processes among all of the levels of institutional analysis discussed inthe various sections above (Baum and Singh 1994a 379ndash80 Mowery andNelson 1999 Murmann 1998)The logic of the perspective herein suggests that within every

society there is variation in the structure and culture of business rmsuniversities and other complex organizations (Kondra and Hinings1998) However this variation takes place within parameters which aresystem specic For example every German rm and every Germanuniversity is different from every other German university and rm but there are a set of cultural and structural characteristics which distin-guish German rms and universities from those in America Similarlyeach Japanese rm and university is unique yet they have a set ofcultural and structural characteristics which are system specic andwhich differentiate them from business rms and universities in other countriesIn all societies each organization has its own distinctive organizational

rules norms and conventions which are subordinate to the meta norms and rules of the larger society within which they are embeddedHowever the strength of the institutional environment within whichorganizations are embedded varies from society to society Some societies have multiple institutional environments and there is hetero-geneity in terms of what constitutes appropriate behavior oforganizations In such societies there can be conict over whatinstitutional logics should regulate specic organizational functions(Friedland and Alford 1991 Townley 1997 262ndash3) In those societies inwhich the institutional norms habits and rules are most developed and in which the institutional pressures to conform are greatest thereis less variation in the structure and culture of business rms and variouskinds of research organizations In such societies the connectednessbetween research organizations and their institutional environments is sufciently strong for organizations to have low autonomy to pursue independent strategies and goals in these societies there is a great deal of organizational isomorphism Conversely the weaker the institutional environment in which research organizations areembedded the greater the variation in the structure and culture of business rms and research organizations Moreover where the institu-tional environments are more weakly developed organizations havegreater autonomy and exibility to respond to the development of new knowledge and to be highly innovative Hence it is in those societies where the institutional environments are most developed and most rigid and when there is less organizational autonomy and exibility that fewer radical innovations in basic and applied science as well as in totally new products and industrial sectors havebeen made

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

621

Fifth level of analysis

We come next to the outputs and performance of the various institutionalcomponents of a society It is at this level that institutional componentsare more pragmatic and exible For example in the legal sector thereare specic statutes and court rulings in the state sector there are specicpolicies in the business sector there are new products new technolo-gies and market strategies It is at this nal level of analysis thatinstitutional spheres are most open and susceptible to change and cross-national mimicry is easiest and most commonThrough the outputs of the society we can obtain some assessment as

to how well a society is performing Moreover we can assess how inno-vative it is how egalitarian it is in the distribution of its resources howegalitarian it is in terms of levels of health education etc Just as it iscomplicated to measure how egalitarian a society is similarly it is dif-cult to assess how innovative it is For example societies vary in makingincremental and radical breakthroughs in basic and applied science in developing totally new products and new kinds of organizations inincremental and process innovations in existing products and organiza-tions in developing new and different forms of marketing ndash bothdomestically and globally (Hage and Hollingsworth 2000) Of coursesocieties vary in their rates of economic growth in their rates of economicproductivity in their quality of life (eg rates of crime life expectancyetc) All of these performance criteria feed back and inuence each ofthe levels of institutional analysis discussed above ndash rules norms valuesetc institutional arrangements institutional sectors the structure andform of organizations Moreover performance measures may inuenceeach other For example if a society has very low rates of growth itmay not be very innovative in some of the types of innovative activitiesmentioned above Good or poor performances of certain kinds inuenceother performance indicators (Hollingsworth et al 1990)Different institutional arrangements and different social systems of

production result in different types of economic performance Hence solong as societies have different social systems of production there areserious constraints on the degree to which they can converge in theirinnovative styles Different social systems of production tend to maxi-mize in a more or less explicit manner different performance criteriausually mixing considerations about static and dynamic efciency protsecurity social peace and economic andor political power In short incontrast to the implications of neoclassical economic theory in real-worldeconomies there are no universal standards all economically rationalactors attempt to maximize Economic history provides numerous exam-ples of how a great variety of principles of rationality are implementedin different societies

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

622

A critical question in institutional analysis is whether an existing socialsystem of production which supports a set of routines for a particularkind of technology and industry can shift from old practices to newones There are numerous historical accounts of how the social systemof production in a particular society which worked so well for a numberof years could not adapt to new technologies (Lazonick 1990 Veblen1915 Nelson 1994) Schumpeter (1983) and Freeman (1991) have devel-oped the idea of a techno-economic paradigm which suggests thatdifferent technologies require specic forms of organizational arrange-ments ndash or what I called social systems of production Because oflsquoinstitutional inertiarsquo the social system of production of a society maynot be able to adapt to a new techno-economic paradigm (Nelson 199458 Lanzara 1998)Whether a social system of production can sustain its performance

standards depends not only on its intrinsic economic rationality but alsoon where it ts into a larger system If a particular social system ofproduction is immune from the innovativeness and competition of analternative system survival can be long lasting But if different socialsystems of production with diverging criteria of good economic perfor-mance meet in the world arena the arbitrariness of nationally imposedconstructed performance standards may be superseded by alternativeperformance criteria as a result of international competitivenessThe world economy is also socially constructed just as are national

economies Even if different social systems of production are competingin the international arena it is not always possible to determine whichis more competitively effective at any moment in time Hegemonicnation-states in the short run shape the rules of trade to favor their indus-trial sectors and rms But the history of hegemonic powers suggeststhat in the longer run such social systems of production sustainedlargely by military and political power eventually give way to moreinnovative and competitive social systems of production In our ownday as nation-states are increasingly integrated into a world economyeven hegemonic powers lose their innovative and competitive advan-tage and their share of world output decreases Such a country mayattempt to restructure its institutional arrangements and to readjust itsperformance preferences But to restructure its system of innovationgenerally calls for a major redistribution of power within a societyLargely for this reason societies have historically had limited capacityto reconstruct a system of innovation in the image of their major competi-tors Each countryrsquos social system of production is a conguration of ahost of norms rules and values as well as of institutional arrangementsEach system is constantly changing and is open to inuence from othersystems And indeed many technologies and practices diffuse from onesociety to another but a societyrsquos capacity to be innovative is constrained

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

623

by the existing social system of production Thus the same technologymay exist in numerous countries but how it is employed and how itinuences societal outcomes and performances varies from one institu-tional conguration or society to another (Hollingsworth 1998)

PERSPECTIVES ON INSTITUTIONAL CHANGE

A fundamental problem which remains unresolved in institutionalanalysis involves the nature of institutional change There is much confu-sion and miscommunication in the social sciences about institutionalchange in large part because we do not even know how to measure therate of institutional change Of course we know that there are bothexternal and internal focuses for change and we have had a good bitof scholarship on this subject What is not clearly understood is thatinstitutions and institutional arrangements within societies are histori-cally rooted that there is a great deal of path-dependency to the waythat institutions evolve Moreover the more intricately linked each sectoris with each other sector and with a societyrsquos rules and norms the lesschoice actors have to devise new institutions and institutional arrange-ments On the other hand if a sector or actor is too isolated it may betoo weak within a society to bring about any kind of effective changeThese considerations are very critical to the problem of how muchfreedom actors have to create new institutions afresh to what extentinstitutions diffuse from one society to another (Murmann 1998 Lanzara1998) This is a fundamental question for students of institutionalanalysis as we attempt to understand how the different institutionallevels are linked togetherThere are a variety of reasons why there is confusion in studying insti-

tutional change There is disagreement over the extent to which actorshave the freedom to build new institutional arrangements and if theydo to what extent the new arrangements and organizational patternsmay depart from past practices Among some institutional economists(Williamson 1985) there is a basic assumption that actors build efcientinstitutions by pursuing their self-interest and by promoting theirstrategic goals Moreover there is the assumption that a mixing of fullrationality with market competition tends to produce all of the optimalinstitutions that are needed in order to coordinate a complex capitalisteconomy (Hollingsworth and Boyer 1997 ch 14 Hodgson 1999) Incontrast the argument developed here is much more complexSocial institutions are historically rooted and there is a great deal of

path-dependency in the way that various institutional componentsevolve The shape of institutional congurations at any moment limitsthe type of options for change Because there is a great deal of institu-tional inertia radical change in the institutional components of a society

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

624

is uncommon (Lanzara 1998) Recognizing the constraints existinginstitutions exert on actors we note that existing congurations of insti-tutional arrangements institutional sectors and organizational propertiescan limit but also inuence the degree to which norms and rules canchange Many scholars (North 1990 Ostrom 1990) have emphasized therole of norms rules ideologies and values in limiting the ability of actorsto develop new institutional arrangements which radically depart fromexisting ones However there is a two-way process of interaction aprocess of co-evolution (Baum and Singh 1994a Nelson 1994) Thus asFriedland and Alford (1991 244) argue actors develop new institutionalarrangements by recognizing organizational failures and low perfor-mance but these new arrangements do not necessarily change a societyrsquosnorms rules habits and its underlying value systemCampbell (1997) advances our understanding of how existing institu-

tions enable actors to construct radically new institutions He recognizesthat most institutions and institutional arrangements embody a degreeof inertia from which it is difcult for actors to depart Moreover thedifferential power relations among actors make it unlikely that lesspowerful actors can change the existing power structure of a societyMost change in the institutional components of a society evolves througha process of constrained selection which reects to a considerable degreethe existing arrangements and power relations This becomes the inter-pretative frame for social actors with institutionalized scripts and ritualswhich tend to be taken for granted and appear to be quite ordinary andnatural This institutionalized scripting of the social world is an extremelyimportant source of social stability and inertiaCampbell has developed a very fruitful explanation of change in terms

of the various institutional components outlined here He distinguishesbetween incremental and radical institutional change and argues thatradical institutional change occurs when social actors with widelydiffering norms cultural scripts and rituals for action engage in intenseinteraction with each other (see also Knight 1992) The more diverse theinteractions the greater the potential for institutional change He arguesthat as shifts in the composition of interaction among social actors occurchanges in the interpretation of problems and interests will follow Ifthere are only minor changes in the extent to which diverse actorsinteract there will be only modest institutional change But the morefundamental the changes in the interaction of diverse social actors thegreater the change in the way that actors interpret their world And thegreater the change in the way that actors interpret their world the morelikely that radical changes in various institutional components willfollow In other words diverse decision making tends to create a widerrange of interpretative frameworks of the social world than is likely tobe the case if social actors operate in isolation

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

625

This perspective of diverse interactions among social actors facilitatesour ability to move beyond the traditional view that existing institutionsconstrain the range of institutional alternatives which actors face Theolder view has tended to suggest that actors engage in institutionalchange by extending existing institutional principles habits and conven-tions Specifying the conditions which facilitate fundamental institutionalchange permits a much richer perspective for institutional analysisCampbellrsquos perspective about the conditions under which fundamentalor radical change in institutional components occurs is consistent withmy own research about the conditions under which major discoveriesor fundamental new knowledge occur the more scholars with diversebackgrounds interact with intensity and frequency the greater the like-lihood that they will develop new and alternative ways of thinking(Hollingsworth and Hollingsworth 2000)These views of institutional change are also consistent with the recent

work of Nelson and his colleagues (Mowery and Nelson 1999 Murmann1998) They report the research on a variety of industrial sectors in anumber of different countries revealing that networks of actors in partic-ular industries business and professional associations universityresearch departments and institutes are frequently able to use their collec-tive power to modify a societyrsquos property rights research system andgovernmental policies It is the intense and frequent interaction of thesediverse complex sets of actors ndash sometimes constituting a technologysystem ndash that often leads to incremental changes in a societyrsquos total socialsystem of production It is this kind of collective action that brings aboutorganizational change change in institutional sectors institutionalarrangements and even in the rules and norms of a society And if weare to understand how the style of innovativeness of a society changeswe must rst understand the changes in its institutional makeup It ishighly unlikely that a society will develop a new style of innovativenesswithout changing its institutional structure

FIRMS THEIR INSTITUTIONAL ENVIRONMENT ANDNATIONAL SYSTEMS OF INNOVATION

Variation in the innovative style of societies is shaped by their insti-tutional makeup ndash the various levels which are discussed aboveinstitutions institutional arrangements the structure and cohesion of institutional sectors which constitute a societyrsquos social system of pro-duction and the structure and culture of its organizations ndash especiallyits business rms and other research organizations By undertaking aninstitutional analysis of a society one can begin to understand in whatkinds of organizations the production of specic kinds of knowledgetakes place and how this is linked to particular kinds of innovativeness

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

626

With such a perspective one can gain a rich understanding of why somesocieties excel in the production of radical breakthroughs in basic andapplied science and in developing radically new products and why othersocieties excel in more incremental innovationsReecting on these issues one cannot help being impressed with the

contrasts between the innovative styles of German and American orga-nizations For example Germany has continued to be very successful in making incremental innovations in many industries in which it wasvery competitive well before World War II paper printing materialsmachinery electrotechnical products motor vehicles chemicals textileyarns and fabrics But the Germans have been much less successful indeveloping totally new products or related innovations in biotechnologyelectronics telecommunications aircraft as well as other newer indus-trial areas In contrast the Americans have been very innovative in these and other industries with products having very short lives andwith technologies which change very rapidly and are very complex(Hollingsworth 1991 Soskice 1997 1999)The importance of institutions and innovativeness for international

competitiveness has led to the recognition that there are national systemsof innovation (Lundvall 1992 Nelson 1993 Edquist 1997 Hage andHollingsworth 2000) But even before the recent studies of nationalsystems of innovation Landes (1969) was a proponent of these viewswhen he made the argument that Germany had developed a nationalsystem of innovation by building on its education and scientic researchsystems which were lacking in Britain and elsewhere However the concept of a national system of innovation immediately poses theproblem of what kind of innovations should be consideredDavid Soskice an economist at the Wissenschaftszentrum Berlin has

addressed a set of problems somewhat complementary to the perspec-tive raised here His work (Soskice 1990 1997 1999) suggests that weare sensitive to the institutional environment in which the followingactors are embedded the employees of rms the owners and nanciersof rms the competitors of rms and nally their collaborators ndash aperspective complementary to Hollingsworthrsquos work on social systemsof production (Hollingsworth 1997 1998) When one reects on the insti-tutional contextsocial system of production within which these actorsare embedded one becomes aware that those countries which have some-what inexible labor markets external to rms but a business system in which rms have long-term commitments to their employees arestrongly associated with incremental but hardly at all with radical inno-vations In the rms of such countries (eg Germany and Japan) thereis much more consensus decision making than in societies where a lsquohireand rersquo set of practices is quite pervasive And this kind of consensusdecision making ndash while highly conducive to incremental innovativeness

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

627

ndash limits the capacity for radical innovations to occur This kind of lsquolongtermismrsquo set of norms and rules occurs not only in the employmentpractices of incrementally innovative rms but also in their nancingand ownership patterns (Hollingsworth 1997) During the past thirtyyears these were characteristics of many rms in Germany and JapanSignicantly in both of these countries venture capital markets werenot highly developed nancing tended to be based more on long-termbank loans than on the equity markets and the relationships of rmstended to be relatively stable And these stable institutional arrange-ments also contributed to a high degree of innovativeness ndash butincremental rather than radical innovationsMeantime we should also pursue Soskicersquos ideas (1999) about how

rmsrsquo relationships with their competitors and collaborators shape theirstyles of innovativeness In those advanced industrial societies in whichbusiness rms are highly mobilized into business associations theretends to be a high degree of standardized quality controls over technicalnorms and a high degree of reliance on legal sanctions as monitoringinstruments vis-agrave-vis association members As a result of this high degreeof integration of rms into associations member rms are more likelyto be associated with incremental innovations than is likely to be the case with rms in those societies where associations are less welldeveloped In the latter type of society radical innovations are morelikely to occur In those advanced industrial societies where rms havelong-term stable relationships with collaborators in product design andproduction the style of innovation is likely to be incremental And it is in those societies in which there is a weaker tradition of strongcollaboration among rms that radical innovations are more likely tooccurObviously even in a country where there may be a great deal of

radical innovativeness most innovations will be of an incremental typefor there are simply fewer radical innovations But Germany and therest of Europe not only have had relatively few radical innovations incomparison with the United States but they also have lagged behindthe USA in developing relatively totally new industries Thus there hasbeen much less innovative activity in Europe in high-technology andnewly emerging industries than in the USAIn complementary research Audretsch and Feldman (1995) have

demonstrated that the USA has been more successful in spawning rmsengaged in the early stages of the industry life cycle but relative toGermany had an environment unfavorable to innovative activity byestablished rms especially in more mature industries German rmshave tended to be associated with a routinized technological regime ndashwhere established enterprises have had strong incentives to engage inincremental innovations but new rms have tended to have fundamental

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

628

innovative disadvantages relative to the Americans (see Audretsch 1995Winter 1982 Casper 2000)On both sides of the Atlantic universities engage in basic and applied

research But signicantly the institutional environment in which univer-sities are embedded inuences the degree to which the research is likelyto represent major breakthroughs or is likely to have more modestimplications Recent research in biomedical science demonstrates that inthose societies in which university professors andor graduate studentshave the following characteristics there are likely to be relatively majorresearch breakthroughs graduate students have a long training periodhigh dependency on a major professor low autonomy in dening thenature of a research project restricted job options and a university envi-ronment where there is a very modest or nonexistent entrepreneurialculture For example in Germany ndash a country with relatively few radicalbreakthroughs ndash the scholar who nishes a habilitation will have spentmany years on a research project before qualifying for a permanent acad-emic position will have been in a high-dependency position with a seniorprofessor will have had limited autonomy to dene a research projectand will be constrained from pursuing a high-risk research project Theyoung German will have an extraordinarily high level of training butwill have faced a number of disincentives to undertake high-risk researchAnd as soon as heshe has completed the habilitation and becomes aprofessor administrative responsibilities constitute another disincentivefrom engaging in high-risk or long-term research projects Moreover theGerman professor is a civil servant with a more stable and secure incomethan hisher American counterpart In a very different institutionalenvironment American professors in large research universities tend tobecome much more entrepreneurial Today many American professorsbehave as though they were operating their own rms within univer-sities and thus devote considerable time to generating funds for researchprojects assistants secretaries supplies etc And the variation in thisentrepreneurial spirit across countries is associated with variation in thetaking of high-risk research strategies and the making of major researchbreakthroughsIn sum societies vary in terms of the institutional environment in

which their organizations are embedded and this variation inuencesthe degree to which they have industrial activities in the early or latestages of the industry life cycle successful new or mature industriesand radical vs incremental innovations And it is this variation whichis encouraging rms in countries with relatively few new industries and with few or no radical innovations to invest in those countries with many new industries and numerous radical innovations ndash as astrategy to be able to transfer new technologies from the latter institu-tional environment to the former

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

629

Table 4 summarizes the way that the norms rules and values of a society are associated with Americarsquos and Germanyrsquos institutionalarrangements their social system of production and their styles of inno-vativeness

CONCLUSION

Several themes deserve to be reemphasized Most importantly institu-tional analysis is emerging as a dominant issue in several social sciencedisciplines The study of various institutional components discussedabove is a complex subject requiring multiple levels of analysis and aknowledge of multiple sectors within societies It requires a theoreticalperspective that evolves from an empirical analysis in different societiesIf we are to advance the study of institutional analysis we need cross-disciplinary collaboration and team research which are not readilyachieved given the structure of most of our research universitiesAt present scholars in each academic discipline have their own distinc-

tive strategies for studying institutional analysis Thus economists tendto use several approaches (North 1990 Williamson 1985 Hodgson 19881998 1999) anthropologists another (Geertz 1995) and sociologistspolitical scientists and historians employ other strategies (Campbell etal 1991 Powell and DiMaggio 1991 Katznelson 1998 Hall and Taylor1996) With practitioners in various disciplines pursuing their distinctiveapproaches to institutional analysis there is low potential for collabo-rative institutional analysis across the social sciences The lessons fromthe biological sciences over the last 40 years are very instructive oncepractitioners from various biological disciplines began to cooperate instudying biological phenomena by using the same concepts at the mole-cular level theoretical advance was very rapid Comparable phenomenatook place in the development of other hybrid elds eg biochemistrybiophysics (Judson 1979 Olby 1979 Kohler 1982 1994) Similarly weresocial scientists in various disciplines to work collaboratively in the areaof institutional analysis there would very likely be an acceleration intheir theoretical knowledgeThis is not a plea that everyone engaged in institutional analysis should

do the same kind of research Indeed just as biologists working at themolecular level study many different kinds of problems similarly socialscientists engaged in institutional analysis would also work on manyseparate problems but in the spirit of a collective enterprise Some wouldconcentrate their attention on the study of rules norms habits conven-tions and values while others would study how these are associatedwith congurations of various institutional arrangements (types ofmarkets hierarchies networks associations communities clans statesetc) Others would work on specic institutional sectors (eg education

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

630

Table 4

Rules norms hab

its an

d values associated w

ith sp

ecic co

ngu

ration

s of coo

rdinating mecha

nism

s soc

ial sy

stem

s of

prod

uction

an

d nationa

l styles of inno

vation

USA

Germany

Norms rules values which provide incentives for

Norms rules values which provide incentives for

short-term horizons

long-term horizons

Con

gu

rative

forms of

Marke

ts c

orpo

rate hierarchies reg

ulatory state

Assoc

iation

s (busine

ss assoc

iation

s la

bor union

s)

coordinating

corp

oratist-type

networks

marke

ts c

orporatist-

mecha

nism

s (listed in

type

state

order of im

portanc

e in

the particu

lar social

system

)

Social sys

tem of

prod

uction

Bus

iness system

Firm

struc

tures

Con

glom

erates b

ut with mov

emen

t toward

Firm

more co

ngru

ent with pa

rticular prod

ucts

elim

ination of produc

tion

in som

e sectors

emph

asis on long

-term strateg

ies rm

s are well

emph

asis on sh

ort-term

strateg

ies high

integrated

differentiation

between divisions

of rms

Firm

s mob

ilized in

Low

High

and in

tegrated

into busine

ss

associations

Indu

strial relations

Low

deg

ree of jo

b secu

rity rigid in

ternal lab

or

High deg

ree of jo

b secu

rity

exible interna

l lab

or

system

marke

t ex

ible externa

l labo

r marke

tsmarke

t rigid externa

l lab

or m

arke

tPatterns of

Dispersed owne

rship frequ

ent turnov

erMore stab

le owne

rship patterns

owne

rship

Table 4

continue

d

USA

Germany

Norms rules values which provide incentives for

Norms rules values which provide incentives for

short-term horizons

long-term horizons

Social system of

Prod

uction

Bus

iness system

Training for labo

r an

dHigh deg

ree of voluntary decision mak

ing by

Collective an

d m

ore co

mpulso

ry d

ecisions

abo

ut

man

agem

ent

individuals as to ho

w m

uch

tim

e an

d ene

rgy

how m

uch ed

ucation

is ne

eded

he

avy em

pha

sis

to inv

est in educ

ation heavy

emph

asis on

on techn

ical training (eg en

gine

ering)

busine

ss sch

ools training

in marke

ting

and

sales accou

nting

Fina

ncial marke

tsEqu

ity an

d ven

ture cap

ital m

arke

ts highly

Equ

ity an

d ven

ture cap

ital m

arke

ts w

eakly

dev

elop

ed na

ncing from

cen

tral gov

ernm

ent

deve

lope

d he

avy em

pha

sis on

corporate loan

sto priva

te rm

s for dev

elop

men

t of new

tech

nologies related

to na

tion

al security

Unive

rsity research

High individual auton

omy high poten

tial for

Low

ind

ividua

l au

tono

my until almost 40

yea

rs of

system

individual creativity u

nive

rsity pr

ofesso

rs

age poten

tial for ind

ividual creativity high

ly

high

ly entreprene

urial

cons

traine

d unive

rsity pr

ofessors not so

entrep

rene

urial

Inno

vative

styles

Highly su

ccessful in m

ajor break

through

s in

High pe

rforman

ce in proc

ess an

d inc

remen

tal

basic an

d app

lied scien

ce go

od in dev

elop

ing

inno

vation

s in existing produ

cts not so su

ccessful

totally

new

lines of pr

oducts an

d in m

ajor break

through

s in basic scien

ce or in

tech

nologies high

deg

ree of entrepr

eneu

rship

deve

loping

totally new

lines of prod

ucts of

amon

g research

ers

tech

nologies low d

egree of entrepr

eneu

rship

amon

g research

ers

Table 4

continue

d

USA

Germany

Norms rules values which provide incentives for

Norms rules values which provide incentives for

short-term horizons

long-term horizons

Prod

ucts in w

hich

Pr

oducts with sh

ort liv

es prod

ucts in w

hich

Older produ

cts with long

history machine

ry pap

er

society excels

tech

nology

is bo

th highly co

mplex

and

pr

inting

textile

s m

otor veh

icles m

achine

too

lsch

ange

s rapidly en

tertainm

ent indus

try

pha

rmaceu

ticals biotechn

olog

y softw

are

aerosp

ace

business systems nancial markets and systems of research) Distinctivewould be the recognition of how these separate studies are linkedtogether how each institutional sector is linked to a societyrsquos normsrules values etc and its conguration of institutional arrangementsSimilarly other institutional analysts might focus on the study of orga-nizations with a major concern for how the institutional environmentof organizations (eg norms rules values congurations of institutionalarrangements institutional sectors) inuenced the structure culture andoutcomes of organizations And nally others might study how all ofthe aforementioned aspects of institutional analysis inuenced a societyrsquosoverall performance Nelson and Sampat (1998) have appropriatelyreminded us that we should engage in institutional analysis in order toaddress specic problems but the view developed in this article is thatwe rst need some mapping of the terrain of institutional analysis inorder for our diverse enterprises to become a meaningful collective enter-prise especially if we are to understand how institutions inuence theinnovative processThe argument developed here is not that all social scientists should engage

in institutional analysis Obviously there are numerous other importantareas of research which lie beyond the eld of institutional analysis Nordoes this article argue that institutional analysis is the most importantsubject for social scientistsSince the structure and culture of our research universities reect an

enormous amount of disciplinary fragmentation what are the prospectsfor promoting the kind of collective and interdisciplinary research agendaproposed above In one respect the prospects are not encouragingDuring the past half century the efforts of American universities to have effective communication across the biological sciences have variedgreatly in their success At the University of California-Berkeley it was not until the 1980s that there was a serious effort to integrate thebiological sciences and at that time the campus abolished seventeendepartments to promote an integrated biological program But at theUniversity of Wisconsin the biological sciences are still very fragmentedinto dozens of different departments in botany plant pathology zoologyand physiology with two departments of biochemistry and two depart-ments of geneticsThere have been other approaches which have worked very well histor-

ically for promoting scientic integration amidst scientic diversity Forexample several summer schools and special institutes were organizedin the United States during the 1940s in an effort to overcome the frag-mentation in the biological sciences These programs made it possiblefor scientists in diverse academic disciplines and from different univer-sities and countries to come together to engage in intense and frequentinteraction Perhaps the most effective program was that held in the

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

634

summers at Cold Spring Harbor New York where a research agendafor the study of molecular biology was developed (Fischer and Lipson1988 Stent and Watson 1966) Scientists went there summer aftersummer and when they returned to their universities many began therestructuring of the biological sciences in their universitiesIn our own day we need summer institutes and other similar programs

so that scholars from diverse backgrounds can come together anddevelop a collective agenda for doing institutional analysis and forstudying the impact of institutions on the performance of societies ndashinnovativeness being one such subject If we can develop such instituteswe will increase the prospects for transforming our universities Thereis reason to be optimistic that once an eclectic group of scholars workwith a common set of concepts we can have genuine theoretical advanceif researchers can bring their diverse expertise to bear on a common setof problems There is considerable evidence that institutional analysisoffers high potential for advancing our knowledge of innovations andtechnological change (Hollingsworth and Hollingsworth 2000) If we aregoing to make substantial theoretical advance it is not likely to occurfrom the vantage point of a multiplicity of disciplines Rather it is likelyto occur as we develop a new intellectual framework and our own acad-emic journals This is a tall agenda which requires us to reassess theway that knowledge is presently produced in our societies

NOTE

This article is part of a much larger agenda involving the study of institutionsorganizations and innovations in which I am involved My debts are many andvaried The rst version was written while I was in residence in the ResearchUnit for Institutional Change and European Integration of the Austrian Academyof Sciences in Vienna in the summer of 1998 I am especially grateful to ProfessorsEgon Matzner and Sonja Puntscher-Riekmann for providing the stimulating envi-ronment in which to think through the ideas developed here Egon Matzner hasengaged in very stimulating discussions with me about the importance of insti-tutions in the general eld of socio-economics The nal version of the articlewas completed during my residence as a fellow at the Neurosciences Institutein La Jolla California I also wish to thank my colleagues in the InnovationTheme Project at the Netherlands Institute for Advanced Study (NIAS) during1998ndash9 and Edgar Grande and Karl Muumlller for stimulating conversations aboutthe interaction of institutions and organizations in facilitating particular typesof innovation I am especially grateful to Jerald Hage for helping me to developa number of issues in the article and for his useful comments on the rst draftFrans van Waarden early on helped me to work through the idea of multiplelevels of institutional analysis Others who have helped to develop the ideas inthe article have been my colleagues in the research project lsquoComparative socialsystems of productionrsquo especially Tom Burns Christel Lane Yoshitaka OkadaWolfgang Streeck and Richard Whitley My colleagues Robert Boyer SteveCasper and David Soskice have taught me much about how incentives providedby the institutional environments of rms inuence individual decision making

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

635

But my most important debt is to Ellen Jane Hollingsworth who has repeatedlyquestioned every single assumption of this article Without her rich and contin-uous tutoring this article would be even less developed than it is at presentFinally without the insightful assistance of David Gear in many phases of thisarticle it would not have been completed

REFERENCES

Aoki Masahiko (1990) lsquoTowards an economic model of the Japanese rmrsquo Journalof Economic Literature March 1ndash27

Arora A Landau Ralph and Rosenberg Nathan (eds) (1998) Chemicals and Long-Term Economic Growth Insights from the Chemical Industry New York JohnWiley amp Sons

Arrow Kenneth (1974) The Limits of Organization New York NortonArthur W B (1988a) lsquoSelf-reinforcing mechanisms in economicsrsquo in P W

Anderson K J Arrow and D Pines (eds) The Economy as an Evolving ComplexSystem Redwood City Calif Addison-Wesley

mdashmdash(1988b) lsquoCompeting technologies an overviewrsquo in G Dosi C Freeman RNelson G Silverberg and L Soete (eds) Technical Change and Economic TheoryLondon Pinter Publishers pp 590ndash607

Audretsch David B (1995) Innovation and Industry Evolution Cambridge MassMIT Press

Audretsch David B and Feldman Maryann P (1995) lsquoInnovative clusters andthe industry life cyclersquo Discussion Paper Wissenschaftszentrum BerlinMarket Processes and Corporate Development Research Unit FS-IV-95-7

Baum Joel A C and Oliver Christine (1992) lsquoInstitutional embeddedness andthe dynamics of organizational populationsrsquo American Sociological Review 57540ndash59

Baum Joel A C and Singh Jitendra V (1994a) lsquoOrganizationndashenvironment coevo-lutionrsquo in Joel A C Baum and Jitendra V Singh (eds) The EvolutionaryDynamics of Organizations New York Oxford University Press pp 379ndash402

mdashmdash(eds) (1994b) The Evolutionary Dynamics of Organizations New York OxfordUniversity Press

Berger Suzanne and Dore Ronald (eds) (1996) National Diversity and GlobalCapitalism Ithaca NY Cornell University Press

Boyer Robert (1997) lsquoThe variety and unequal performance of really existingmarkets farewell to Doctor Panglossrsquo in J Rogers Hollingsworth and RobertBoyer (eds) Contemporary Capitalism the Embeddedness of InstitutionsCambridge and New York Cambridge University Press pp 55ndash93

Burns Tom R and Dietz Thomas (2001 forthcoming) lsquoSocial rule system theorysocial action institutional arrangements and evolutionary processesrsquo in JRogers Hollingsworth and Karl Muumlller (eds) Advances in SocioeconomicsLanham Md Rowman amp Littleeld

Burns Tom R and Flam Helena (1987) The Shaping of Social Organization BeverlyHills Calif Sage

Calvert Randall L (1995) lsquoThe rational choice theory of social institutionsrsquo inJeffrey S Banks and Eric A Hanuskek (eds) Modern Political Economy NewYork Cambridge University Press pp 216ndash66

Camic Charles (1986) lsquoThe matter of habitrsquo American Journal of Sociology 911039ndash87

Campbell John L (1997) lsquoMechanisms of evolutionary change in economicsgovernance interaction interpretation and bricolagersquo in Lars Magnusson

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

636

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

607

DISTRIBUTION OF POWER

VERTICALHORIZONTAL

1 MARKETS 2 HIERARCHY

6ASSOCIATIONS

5NETWORKS

4STATE

3COMMUNITIES

SELF INTEREST

SOCIAL OBLIGATION

ACTION MOTIVE

Figure 2 A general taxonomy of institutional arrangements

The vertical axis deals with action motives Toward the upper part ofFigure 2 actors are engaged in individualistically oriented behaviorwhereas toward the lower part actors are more engaged in collectivebehavior and strive to cope with problems of common interest Cell 3 ndashcommunities and clans ndash consists of institutional arrangements based ontrust reciprocity or obligation and thus not derived from the pure selshcomputation of pleasures and pains This is an unconventional form ofcoordination for most neoclassical economists (however see Arrow1974) but not for many anthropologists political scientists and sociolo-gists (Streeck and Schmitter 1985a Polanyi 1944 Gambetta 1988Fukuyama 1995 Sabel 1992 Putnam 1993)In the neoclassical paradigm theorists argue that actors engage in

forms of exchanges that best promote their individual interests If somestructural conditions are fullled (absence of increasing returns to scalethe reversibility of transactions absence of uncertainty and completecontingent markets with no collusion between actors) then the invis-ible hand theorem applies and market-type activity functions quite welland also provides the optimum for society thereby combining efciencyharmony and order However an excess of market activity may welllead to ruinous competition and excessive conict There is variation inthe extent to which ruinous competition occurs depending on the socialcontext in which transactions take place Thus it is important that webe sensitive to the institutional context in which transactions areembedded and that we understand the degree to which social bondsexist at both the micro and macro levels of analysis Micro bonds facil-itate exchanges in a society but at the societal level social bonds existat the level of the collective ndash in the community or region and amongmembers of racial religious and ethnic groups All other things beingequal the more powerful the social bonds among transacting partnersthe more economic competition is likely to be restrained Thus mosttransactions occur not simply in an impersonal calculative system ofautonomous actors unrestrained by social ties ndash as implied by theneoclassical paradigm ndash but in the context of social ties variation in the strength of which leads to variation in levels of trust and transac-tion costs (Etzioni 1988 211 Granovetter 1985 Streeck and Schmitter1985a Hollingsworth and Boyer 1997 Hodgson 1988 1999 Schneibergand Hollingsworth 1990)Another form of multilateral institutional arrangement is various

types of associations (cell 6) Unlike networks clans and communitiesassociations are more formal organizations Whereas markets corporatehierarchies and networks tend to coordinate economic activity amongdifferent types of actors (eg producers with suppliers capital withlabor) associations typically coordinate actors engaged in the same orsimilar kinds of activities Business associations and labor unions are

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

608

some of the most common forms of associations for coordinatingeconomic activity in capitalist economies (Schneiberg and Hollingsworth1990 Streeck and Schmitter 1985a)Finally there is the state (cell 4) which is an institutional arrangement

quite unlike any of the others It is the state that sanctions and regu-lates the various non-state coordinating mechanisms that is the ultimateenforcer of rules of the various mechanisms that denes and enforcesproperty rights and that manipulates scal and monetary policy Overallit is the state that inuences the total incentive system of a society(Lindberg and Campbell 1991 Johnson 1992 40) At the same time thestate may also be an economic actor by engaging directly in productionand exchange relationsThe choices of institutional arrangements in Figure 2 are constrained

by the social context in which they are embedded Depending on thenature of that embeddedness there is variation in the collective formsof governance some of which are specied in the lower part of thegure Some modes of coordination in the bottom and upper levels ofthe typology are often mixed together though one coordinating modeis likely to be more prominent than another But actors are often engagedin complex congurations involving several kinds of institutionalarrangementsEach of these various institutional arrangements has its own logic ndash

its own rules its own procedures for enforcing compliance its ownnorms and ideologies which help to reduce the costs of enforcementThese are summarized in Table 2 which provides further elaborationabout the various coordinating mechanisms one nds in almost everycapitalist society Table 2 lists the organizational arrangements and theirstructures rules of exchange and means of enforcing compliance asso-ciated with each type of coordinating mechanism While each type ofinstitutional arrangement has various positive features each institutionalarrangement also has failures and these are featured in Table 3 It is thecontest between those who support and those who oppose these variousinstitutional arrangements that tends to lead to transformations in insti-tutional arrangements over time (Campbell et al 1991 Campbell 1997)Again it is difcult to conceive of pure institutional arrangements ndash

either exclusively markets or exclusively hierarchies ndash since it has beenwell known since Adam Smithrsquos Wealth of Nations that the division oflabor within the rm cannot be disentangled from the existence andextent of the market Institutional arrangements have their own distinc-tive form of efciency (good static efciency for the market dynamicefciency for rms) and inefciency often leading to considerableinequality Neither networks nor communities are panaceas for economiccoordination without being congured with other types of institutionalarrangements Networks and communities may solve certain issues but

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

609

Table 2

Log

ics of ins

titution

al arran

gemen

ts

Coordination

Organizational

Rules of

Individual means

Collective means

mechanisms

structure

exchange

of compliance

of compliance

Marke

tsEasy en

try an

d exit

Voluntary sp

ot excha

nge

Leg

al enforcemen

t of

Norm of pr

ivate

Bila

teral ex

chan

ge or

control

prope

rty

marke

t site (W

all Street)

Reg

ulations

to

Leg

itim

acy of m

arke

t en

force contracts

men

tality

Com

mun

ities

Inform

al m

embe

rship

Voluntary exch

ange

based

So

cial norms an

d m

oral

Highly institutiona

lized

ev

olve

s ov

er lon

g period

on social solid

arity an

d

princ

iples im

pose

norm

s an

d rules requ

ire

of tim

ehigh

deg

ree of tru

stob

ligations

mem

bers to accept

Kno

wledg

e of others an

d

lsquocorpo

ratersquoo

bligations

recipr

ocity ov

er tim

e

Networks

Semi-form

al m

embe

rship

Voluntary exch

ange

ove

r a

Con

tractual bon

dsPersona

l relation

sBila

teral or m

ultilateral

time pe

riod

Resou

rce de

pend

ence

Tru

st built outside

the

ex

chan

geecon

omic arena

Associatio

nsFo

rmal m

embe

rship

Restricted to mem

bers

Self-interest

Some deg

ree of

Multilateral excha

nge

Emph

asis on insider

Rep

utation

effects

compu

lsion

outsider or we

they

Priva

te interest type

of

men

tality

gove

rnan

ce

Private

Com

plex organ

izations

Restricted to mem

bers

Rew

ards to ind

ividuals

Highly institutiona

lized

hierarch

ies

which

ten

d to becom

e ex

chan

ge based

on

Asymmetric po

wer threat

rules

bureau

cratic hierarchies

asym

metric po

wer

of san

ctions

Mem

bers soc

ializ

ed into

bureau

cratic rules

corp

orate cu

lture us

e of

sanc

tions

State

Pub

lic hierarchy

Unilateral action

Exit vo

ice (vote lob

bying)

Coe

rcion

De jure

and impo

sed

Indirect political an

d

loya

ltyNorms an

d pu

blic rules

mem

bership

econ

omic excha

nge

Table 3

Failu

res of coo

rdinating mecha

nism

s

Coordinating mechanisms

Type of failure

Associations

Private hierarchies

State

Market

Communities

Networks

Enforcem

ent

Usu

ally relies on

Might enh

ance

Needs co

ntrol

Needs an

internal

Needs

tru

st and

Need an external

the state as an

oppo

rtun

istic

extern

al to state

enforcem

ent

loya

lty often

en

forcem

ent

enforcer

beha

vior

bureau

cracy

authority

coming from

au

thority

Resem

bles

The

ideal o

f (ju

dge

s

Facilitates

outside (fam

ily May

facilitate

enforcem

ent

internal m

arke

ts

parliamen

t collu

sion

and

relig

ion

cartelization an

d mecha

nism

might hur

t marke

t) to

impe

rfect

ethn

icity)

mon

opoly

of cartels

incu

mbe

nt

correct state

compe

titio

nCom

patible with

worke

rsab

uses

variou

s types of

Lob

bies can

co

mpe

tition

captur

e pu

blic

interest goa

ls

Public goo

d

Useful for

Gov

erna

nce co

sts

Can

provide

Can

not provide

Can

internalize

Useful for

and

establishing

might exceed

public goo

ds

colle

ctive go

ods

some co

llective

enha

ncing qu

ality

externality

stan

dards an

d

the be

nets of

but ha

s or deal w

ith

good

s (qua

lity

and training bu

t qu

ality for

internal division

difc

ultie

s in

externalities

training

) bu

t no

t no

t ve

ry goo

d in

setting ru

les of

of la

bor

prov

iding them

Inad

equate

othe

rs (welfare

prov

iding for

compe

titio

n in

Slow

to react to

in precise

mon

itoring of

gene

ral p

ublic

societal gen

eral

the indu

stry

chan

ges in the

am

ounts

tech

nical ch

ange

go

ods)

welfare

Useful for

environm

ent

Might fail in

and inno

vation

Mem

bers tightly

Weak in the

prov

iding

induc

ing

integrated

into

prov

ision of

man

y go

ods

tech

nical

community

colle

ctive go

als

colle

ctively

chan

geha

ve limite

d

that in

dividua

lcapa

city for

mem

bers can

not

inno

vatio

nsprov

ide for

them

selves

Table 3

continue

d

Coordinating mechanisms

Type of failure

Associations

Private hierarchies

State

Market

Communities

Networks

Efcien

cyFa

cilitates

De

cien

t in

Can

be high

ly

Some ba

sic social So

me go

ods

Slow

to en

hanc

e co

operation an

d

coop

eration an

d

bureau

cratic

relation

s cann

ot

cann

ot be

efc

ienc

y an

d

X-efcien

cy but

X-efcien

cyan

d can

not

be provided

by

delivered

at

speed of

not allocative

easily deliver

pure m

arke

t su

fcien

tly

adap

tive

ness

efcien

cy

good

s at

mecha

nism

slow costs

except in

low cost

indus

tries whe

re

tech

nology

is

complex an

d rapidly cha

nging

Equity

Narrow

Excessive

Might enh

ance

Facilitates

Might lead to

Whe

n widely

encompa

ssing

multiplication

ineq

uality

ineq

uality in

retard

ed

dev

elop

ed in

to

associationa

l of con

trollers

(pow

er

income an

d dev

elop

men

tindus

trial

stru

ctures lead

(fru

stration

and

an

d privilege

)wealth

districts

to in

come

ineq

uality)

netw

orks m

ay

equa

lity

facilitate greater

equality an

d

inco

me

distribution

whe

n wea

kly

dev

elop

ed

netw

orks ten

d to

increase soc

ial

ineq

uality

Note For allo

cativ

e efc

ienc

y an

d X

-efcien

cy see Leibe

nstein (19

66 19

76)

they raise other no less severe problems It is important to recognizethe imperfection of any single institutional arrangement in order tocomprehend the origin and transformation of any other institutionalarrangement (Hodgson 1999 ch 3)As suggested above each institutional arrangement is congured with

other institutional arrangements Usually one particular institutionalarrangement is more dominant in a particular conguration but becauseeach type of institutional arrangement has its own strengths andweaknesses there is no simple structural logic in the governance or coor-dination of a society Each institutional arrangement constrains the otherbut the inherent tension among the various institutional arrangementswithin a conguration contributes to changes in the governance orcoordination of a society The routines and logics of each of these cong-urations of institutional arrangements provide constraints and incentivesfor actors The way that actors perceive the incentives and constraintsof these governance congurations leads to a particular market logicand it is the specic market logic of a society which inuences its speciccapacities and weaknesses The inherent strain among the different logicsin a conguration of coordination helps to provide the exibility for asociety to adapt to new circumstancesIf there were a society with pure markets or a society coordinated

only by the state there would be too much rigidity and too little diver-sity to cope with the vast uncertainty in the global environment A societywith very little diversity in its coordination mechanisms would havelittle capacity to adapt to new circumstances Congurations with consid-erable diversity of institutional arrangements provide for a certainamount of incoherence in governance but they also provide for thecapacity to adapt to new circumstances While the Soviet Union wasdominated by the state as a coordinating mechanism there were alwaysfunctioning markets in the system Moreover the feudal society of thetwelfth and thirteenth centuries consisted not only of hierarchical rela-tionships between serfs and masters but also of urban guilds clericalhierarchies and markets (Johnson 1992 38) In sum the robustness ofinstitutions often depends on multiple and diverse principles and logicsof action on the inconsistency of principles and procedures on patternedforms of disorder (Lanzara 1998 Orren and Skowronek 1991 320 329)

Third level of analysis

The next level consists of the institutional sectors of a society The rulesnorms and values of a society inuence the array of institutional arrange-ments and levels one and two inuence the nature of and the relation-ships among various institutional sectors and all three levels areintricately linked together to form a social system of production Together

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

613

all of these components inuence the performance of economic sectorswithin a society as well as the performance of the total society A socialsystem of production is the way that a societyrsquos institutions (see rstlevel of analysis above) its institutional arrangements (see second levelof analysis above) and its institutional sectors are integrated into a socialcongurationAn institutional sector includes all organizations in a society which

supply a given service or product along with their associated focal orga-nizations (eg major suppliers funding sources regulators and so forth(Scott et al 1994 108 117 Campbell et al 1991 Hollingsworth et al1994) Institutional sectors include but are not limited to the societyrsquossystem of education system of research business system nancialmarkets legal system and the state The structure of the nancial marketssystem of training and education industrial relations system system ofresearch and state and legal systems are distinct and idiosyncratic ineach society In short they are system specicAll of these institutional sectors tend to cohere with each other

although they vary in the degree to which they are tightly coupled intoa full-edged system While each of the institutional sectors has someautonomy and may have some goals that are contradictory to the goalsof other institutional sectors with which it is linked an institutional logicin each society leads institutions to coalesce into a complex social con-guration (Hollingsworth 1991) This occurs because the institutionalsectors are embedded in a culture in which their logics are symbolicallygrounded organizationally structured technically and materially con-strained and politically defended The conguration of institutionalsectors usually exhibits some degree of adaptability to new challengesbut continues to evolve within an existing style But under new circum-stances or unprecedented disturbances these institutional congurationsare exposed to sharp historical limits as to what they may or may notdo (Schumpeter 1983 David 1988 Arthur 1988a 1988b Haringkanssen andLundgren 1997)Why do all of these different institutions coalesce into a complex

social conguration which is labeled here as a social system of pro-duction The literature suggests two contrasting interpretations Part ofthe answer ndash indeed a controversial one ndash is that these institutionalsectors are functionally determined by the requirements of the practiceof capitalism in each time and place (Habermas 1975) Another expla-nation emphasizes the genesis of the actual conguration via a trial and error process according to which the survival of rms regions orcountries is the outcome of complex evolutionary mechanisms (Maynard-Smith 1982 Nelson and Winter 1982) However the problem is evenmore complex Markets and other mechanisms for coordinating rela-tionships among economic actors place constraints on the means and

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

614

ends of economic activity to be achieved in any society The othercoordinating mechanisms include different kinds of hierarchies varioustypes of networks and associations ndash trade unions employers and busi-ness associations (Hollingsworth and Lindberg 1985 Campbell et al1991) The logic of these various institutional arrangements providesactors with vocabularies and logics for pursuing their goals for deningwhat is valued and for shaping the norms and rules by which theyabide (see Tables 2 and 3) In short in contrast to the logic of theneoclassical paradigm the argument here is that the dominant type ofinstitutional arrangements places severe constraints on the denition of needs preferences and choices of economic actors Whereas the neo-classical paradigm assumes that individuals and rms are sovereign thisarticle is based on the assumption that rms are inuenced by the hold that the institutional congurations making up a social system ofproduction have on individual decision making (Campbell et al 1991Etzioni 1988 Streeck and Schmitter 1985a Hollingsworth et al 1994Hollingsworth and Boyer 1997 Magnusson and Ottosson 1997 North1990 Hodgson 1999)Standard neoclassical economic theory has tended to downplay the

role of production and consequently of rms Even transaction cost theo-rists who are concerned with analyzing the rm as a coordinatingmechanism have been relatively unconcerned with the various compo-nents of a social system of production Indeed as long as there waswidespread optimism about the efcacy of Keynesian economics therewas relatively little concern among neoclassical economists with thesupply side of the economy Even in the opinion of most Keynesians agroup of experts should ideally be able to shape the size of aggregatedemand while the supply side of the economy would be left to the twominimalist institutions of neoclassical economics ndash markets and manage-rial hierarchies For several decades however it has become increasinglyobvious that some of the most competitive and successful patterns ofindustrial output and industrial production in capitalist economies donot derive from the neoclassical prescription of unregulated markets andcorporate hierarchies complemented by a neoliberal democratic stateIndeed it is now well understood that certain highly successful produc-tion patterns require for their emergence and survival institutionalarrangements and environments the very opposite of the prescriptionsfound in the neoclassical paradigm (see especially Streeck 1992 but alsoHollingsworth and Streeck 1994) Thus social scientists have come tounderstand that if they are to comprehend the behavior and performanceof contemporary economies concerns about social systems of produc-tion must be brought into the pictureBecause production involves more than technology a number of social

scientists have an increasing concern with social systems of production

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

615

The same equipment is frequently operated quite differently in the samesectors in different countries even when rms are competing in the same market (Maurice et al 1980 Sorge and Streeck 1988 Sorge 1989Hollingsworth et al 1994) Variations in production and process tech-nologies are inuenced partly by variations in the social environmentsin which they are embedded In other words rms are embedded intocomplex environments which among other things place constraints ontheir behavior Thus a social system of production is of major impor-tance in understanding the behavior and performance of an economyHow the state and other institutional arrangements (eg marketsnetworks private hierarchies associations) coalesce and are related toparticular social systems of production are important determinants ofeconomic performanceThe congurations of institutional arrangements that coordinate or

govern the behavior of actors in one society and its structure of specicinstitutional sectors cannot easily be transferred to another society forthey are embedded into a social system of production that is societallydistinct (Hollingsworth 1997) Societies borrow selected principles offoreign management styles and work practices but the effectiveness ofsuch borrowing is generally limited Economic behavior and performanceare shaped by the entire social system of production in which actors areembedded and not simply by specic principles of particular manage-ment styles and work practices Moreover a societyrsquos social system ofproduction tends to limit the kind of goods that it can produce and withwhich it can compete successfully in international marketsSome scholars (Kenney and Florida 1993 Oliver and Wilkinson 1988)

have assumed that the diffusion of particular forms of management stylesand work practices across societies could lead to system convergenceReferring to the Japanese production system Kenney and Florida haveargued that it consists of organizational practices whose fundamentallsquogenetic codersquo can successfully be inserted into another society and beginsuccessfully to reproduce its behavior in the new environment (1993 8)Their position is in the intellectual tradition of Antonio Gramsci (1971)who contended decades ago that the American system of mass produc-tion would most certainly diffuse to Europe over timeHowever the argument of this article is that even though British

French and American rms may adopt certain aspects of Japanesemanagement styles (eg just-in-time production self-managing teamsquality circles the use of statistical process controls) or some variant ofthe German vocational system their social systems of production willnot converge Moreover the overall conguration of a societyrsquos socialsystem of production inuences its sectoral and overall national economicperformance This helps to explain why societies have different systemsof innovation and why they vary in the clusters of industries in which

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

616

they are highly competitive in the international markets (Nelson 19931999 Hage and Hollingsworth 2000 Berger and Dore 1996)In order to understand how and why a societyrsquos economy performs

as it does it is necessary to understand its entire social system of produc-tion If a society is to modify substantially the performance of itsinstitutional sectors and its entire economy it cannot adapt only someof the management and work practices of its foreign competitors Ratherit must alter its entire social system of production Because a societyrsquosmodes of economic governance and coordination and its institutionalsectors develop according to particular logics and are system specicthere are serious limitations to the extent to which a society may mimicthe institutions (eg the rules and norms) institutional arrangements andinstitutional sectors of other societiesIn the history of modern societies there are logics by which institu-

tions coalesce into a social system of production (Hollingsworth 1991)Though institutions are constantly changing there are sharp limits tothe type and the direction of change that any particular institution canundergo because of its linkages with institutional arrangements and insti-tutional sectors Thus a societyrsquos business rms educational systemnancial markets industrial relations system etc can engage in seriousrestructuring only if most of the other institutional sectors also changeIn social systems of production there are pressures toward consistencyin the norms rules and values across institutional sectors though in anycomplex society social systems are obviously imperfectly integratedIndeed the degree to which the institutional norms and rules makingup a social system of production are loosely or tightly coupled is a vari-able of considerable importance In general the institutional sectorsmaking up a social system of production are interdependent and changesin one generally result in changes in the others Since each institutionallevel is dependent on the others for various types of resources there isinterdependence among the differing institutional spheres Moreovereach society has its norms moral principles rules and laws and recipesfor action as well as its own idiosyncratic customs traditions and prin-ciples of justice (Burns and Flam 1987)There are also other inherent obstacles to convergence among social

systems of production for where a system is at any one point in timeis inuenced by its initial state Systems with quite different initial statesare unlikely to converge with one anotherrsquos institutional practicesExisting institutional arrangements block certain institutional innova-tions and facilitate others (Roland 1990) Thus the institutions makingup a social system of production provide continuity even though insti-tutional arrangements are always changing but with a particular logicWhile Williamson (1975 1985) suggests that actors tend to choose theinstitutional arrangements which are most efcient North (1990) is

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

617

much closer to the mark in his argument that most societal institutionalarrangements exist as a result of custom and habit and are thereforeinefcient At any moment in time the world often appears to its actorsas very complex and uncertain For this reason actors often engage incontradictory forms of behavior pursuing different strategies as hedgesagainst a very uncertain world (Lanzara 1998) And their hedging andcontradictory forms of behavior may lead in somewhat different soci-etal directions all constrained by the institutional fabric within whichthe actors are embedded This kind of contradictory behavior occurs inpart because of the contradictions inherent in the dominant congura-tions of institutional arrangements with their contradictory logics (seelsquoSecond level of analysisrsquo above)Despite the emphasis on the logic of institutional continuity this is

not an argument that systems change along some predetermined pathThere are critical turning points in the history of highly industrializedsocieties but with the choices limited by the existing institutional terrainBeing path-dependent a social system of production continues along aparticular logic until or unless a fundamental societal crisis intervenes(Milgrom et al 1991 Krugman 1991 Durlauf 1991 Hollingsworth 1991Pred 1966 David 1988 Hodgson 1998)At this point it is important to confront the question of how social

systems of production evolve And this question gets to the heart of theproblem of building complex societal structures Certainly social systemsof production did not emerge from some process of social engineeringMoreover the various component parts of each social system of produc-tion have often not been designed to be part of a social system ofproduction The component parts of each social system of productionhave emerged more often than not as unintended by-products of goalswhich various actors had in mind at earlier moments in time It is usuallythe case that actors at time lsquotrsquo fail to comprehend the long-term conse-quences of their actions Wolfgang Streeck the author of several brilliantpapers on social systems of production argues that institutional sectordesigns created for one purpose generally address that goal but overtime those designs have quite unintended consequences (Streeck 1997a1997b Wright 1998) Describing the conguration of the German socialsystem of production of the 1970s and 1980s Streeck points out that itwas the unintended by-product of multiple points in history Someelements were pre-Wilhelmian others were introduced by the Alliedpowers after 1945 and others emerged during the years of the GermanFederal Republic All component parts of the German social system ofproduction

were and continue to be changing for their own reasons as wellas in reaction to each other and certainly there can be no presump-tion of a pre-established t between them even though one might

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

618

want to allow for some reinforcement effects of the modelrsquos histor-ically contingent social and economic success

(Streeck 1997b 54)

In sum the emergence of social systems of production is a long-termevolutionary process with each part interacting with its environment andresulting in a congurative whole but with no previous design by eithera single actor or a collectivity of actors As suggested above actors do not recreate their world anew At any moment in history the totalinstitutional complex of a society necessarily contains the resources and legacies of its past Preexisting institutional complexes are nevercompletely wiped out Societies consist of multiple layers of history withtheir diverse logics of action However the cumulative effect of smallperipheral changes in altering particular institutional sectors and insti-tutional arrangements can be substantial (Murmann 1998 Lanzara 1998)lsquoInstitution building is affected more by the ways in which people havecodied the past than by how they envision the future There are manymore potential resources in pre-existing institutional arrangements thanit is usually assumed or suspectedrsquo (Lanzara 1998 30)

Fourth level of analysis

The next level of institutional analysis organizational structures is some-what more controversial As suggested above Douglass North (1990)draws a sharp distinction between institutions and organizations Morerecently many organizational theorists (Powell 1991 Powell andDiMaggio 1991 DiMaggio and Powell 1983 Baum and Oliver 1992Kondra and Hinings 1998 Townley 1997) argue that institutional rulesnorms and conventions unfold in tandem with organizational structuresand this is my positionThe literature which focuses on how institutions inuence organiza-

tions is quite different from two other theoretical literatures which alsoare concerned with how organizational environments shape the behaviorof organizations There are resource dependency theorists (Pfeffer andSalancik 1978) who emphasize the role of environmental resources inshaping organizations while Hannan and Freeman (1977 1984) and otherpopulation ecologists emphasize the survival of organizations givencertain kinds of organizational conditions (Orru et al 1991)The institutionalist perspective as implied above emphasizes the

normative environment in which organizations are embedded It is a perspective which focuses on the way in which organizations in their behavior tend to conform to the institutional rules and normswhich are dominant in the organizational environment However allthree perspectives ndash the institutionalist the population ecology and

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

619

the resource dependency ndash emphasize how the environment inuencesorganizations and how organizations which are subject to the sameenvironment tend to converge in their behavior to have what DiMaggioand Powell have labeled organizational isomorphism (DiMaggio andPowell 1983)There has emerged a vast literature which demonstrates that within

each society there are scal political judicial and other regulatory normswhich limit and shape the culture and structure of organizationalbehavior The normative institutional environment of organizations limits the options of what organizations do in a particular society and inuences the patterns of ownership relations with suppliers andcustomers In short it is the normative environment of organizationswhich denes within a particular society what is socially acceptablebehavior for organizations (Hamilton and Biggart 1988 Hollingsworthand Hollingsworth 2000 Meyer and Rowan 1991 Zucker 1987 1988Orru et al 1991 Townley 1997)Thus far most of the analyses on isomorphic pressures operating on

organizations have been cross-sectional in nature However the entireconcept of isomorphism implies that there are strong environmental pres-sures exerted on organizations and in order to observe this phenomenonit is important that we have longitudinal studies which assess howchanges in an organizationrsquos institutional environment inuence changesin the structure and culture of organizations Moreover the expectationhere is that when one analyzes the patterns of organizational changeover time the isomorphism will be holistic ndash that is within a particularsociety there will be a tendency for the internal structure of particulartypes of organizations to converge in large part because of the pres-sures to conform to the changes in the external norms rules and valuesA close reading of Aoki (1990) and Slack and Hinings (1994) nds thatchanges in the national institutional environment of organizations haveinuenced changes in the internal structure of organizations within aspecic industry of a particular societyHowever it is inappropriate to view organizations as changing only

in relation to exogenous environmental change There is a kind of co-evolution which occurs between organizations and their institutionalenvironments Co-evolution implies nonlinear feedback between orga-nizations and their institutional environments This is of course quitedifferent from the more common way of studying organizations bymodeling relations between independent and dependent variables Ofcourse the more common type of analysis is quite appropriate whenthere are simple relationships which do not involve complex feedbackprocesses But in the strategy for institutional analysis which is proposedherein it is less useful to separate independent from dependent vari-ables and more useful to understand the interacting and co-evolutionary

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

620

processes among all of the levels of institutional analysis discussed inthe various sections above (Baum and Singh 1994a 379ndash80 Mowery andNelson 1999 Murmann 1998)The logic of the perspective herein suggests that within every

society there is variation in the structure and culture of business rmsuniversities and other complex organizations (Kondra and Hinings1998) However this variation takes place within parameters which aresystem specic For example every German rm and every Germanuniversity is different from every other German university and rm but there are a set of cultural and structural characteristics which distin-guish German rms and universities from those in America Similarlyeach Japanese rm and university is unique yet they have a set ofcultural and structural characteristics which are system specic andwhich differentiate them from business rms and universities in other countriesIn all societies each organization has its own distinctive organizational

rules norms and conventions which are subordinate to the meta norms and rules of the larger society within which they are embeddedHowever the strength of the institutional environment within whichorganizations are embedded varies from society to society Some societies have multiple institutional environments and there is hetero-geneity in terms of what constitutes appropriate behavior oforganizations In such societies there can be conict over whatinstitutional logics should regulate specic organizational functions(Friedland and Alford 1991 Townley 1997 262ndash3) In those societies inwhich the institutional norms habits and rules are most developed and in which the institutional pressures to conform are greatest thereis less variation in the structure and culture of business rms and variouskinds of research organizations In such societies the connectednessbetween research organizations and their institutional environments is sufciently strong for organizations to have low autonomy to pursue independent strategies and goals in these societies there is a great deal of organizational isomorphism Conversely the weaker the institutional environment in which research organizations areembedded the greater the variation in the structure and culture of business rms and research organizations Moreover where the institu-tional environments are more weakly developed organizations havegreater autonomy and exibility to respond to the development of new knowledge and to be highly innovative Hence it is in those societies where the institutional environments are most developed and most rigid and when there is less organizational autonomy and exibility that fewer radical innovations in basic and applied science as well as in totally new products and industrial sectors havebeen made

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

621

Fifth level of analysis

We come next to the outputs and performance of the various institutionalcomponents of a society It is at this level that institutional componentsare more pragmatic and exible For example in the legal sector thereare specic statutes and court rulings in the state sector there are specicpolicies in the business sector there are new products new technolo-gies and market strategies It is at this nal level of analysis thatinstitutional spheres are most open and susceptible to change and cross-national mimicry is easiest and most commonThrough the outputs of the society we can obtain some assessment as

to how well a society is performing Moreover we can assess how inno-vative it is how egalitarian it is in the distribution of its resources howegalitarian it is in terms of levels of health education etc Just as it iscomplicated to measure how egalitarian a society is similarly it is dif-cult to assess how innovative it is For example societies vary in makingincremental and radical breakthroughs in basic and applied science in developing totally new products and new kinds of organizations inincremental and process innovations in existing products and organiza-tions in developing new and different forms of marketing ndash bothdomestically and globally (Hage and Hollingsworth 2000) Of coursesocieties vary in their rates of economic growth in their rates of economicproductivity in their quality of life (eg rates of crime life expectancyetc) All of these performance criteria feed back and inuence each ofthe levels of institutional analysis discussed above ndash rules norms valuesetc institutional arrangements institutional sectors the structure andform of organizations Moreover performance measures may inuenceeach other For example if a society has very low rates of growth itmay not be very innovative in some of the types of innovative activitiesmentioned above Good or poor performances of certain kinds inuenceother performance indicators (Hollingsworth et al 1990)Different institutional arrangements and different social systems of

production result in different types of economic performance Hence solong as societies have different social systems of production there areserious constraints on the degree to which they can converge in theirinnovative styles Different social systems of production tend to maxi-mize in a more or less explicit manner different performance criteriausually mixing considerations about static and dynamic efciency protsecurity social peace and economic andor political power In short incontrast to the implications of neoclassical economic theory in real-worldeconomies there are no universal standards all economically rationalactors attempt to maximize Economic history provides numerous exam-ples of how a great variety of principles of rationality are implementedin different societies

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

622

A critical question in institutional analysis is whether an existing socialsystem of production which supports a set of routines for a particularkind of technology and industry can shift from old practices to newones There are numerous historical accounts of how the social systemof production in a particular society which worked so well for a numberof years could not adapt to new technologies (Lazonick 1990 Veblen1915 Nelson 1994) Schumpeter (1983) and Freeman (1991) have devel-oped the idea of a techno-economic paradigm which suggests thatdifferent technologies require specic forms of organizational arrange-ments ndash or what I called social systems of production Because oflsquoinstitutional inertiarsquo the social system of production of a society maynot be able to adapt to a new techno-economic paradigm (Nelson 199458 Lanzara 1998)Whether a social system of production can sustain its performance

standards depends not only on its intrinsic economic rationality but alsoon where it ts into a larger system If a particular social system ofproduction is immune from the innovativeness and competition of analternative system survival can be long lasting But if different socialsystems of production with diverging criteria of good economic perfor-mance meet in the world arena the arbitrariness of nationally imposedconstructed performance standards may be superseded by alternativeperformance criteria as a result of international competitivenessThe world economy is also socially constructed just as are national

economies Even if different social systems of production are competingin the international arena it is not always possible to determine whichis more competitively effective at any moment in time Hegemonicnation-states in the short run shape the rules of trade to favor their indus-trial sectors and rms But the history of hegemonic powers suggeststhat in the longer run such social systems of production sustainedlargely by military and political power eventually give way to moreinnovative and competitive social systems of production In our ownday as nation-states are increasingly integrated into a world economyeven hegemonic powers lose their innovative and competitive advan-tage and their share of world output decreases Such a country mayattempt to restructure its institutional arrangements and to readjust itsperformance preferences But to restructure its system of innovationgenerally calls for a major redistribution of power within a societyLargely for this reason societies have historically had limited capacityto reconstruct a system of innovation in the image of their major competi-tors Each countryrsquos social system of production is a conguration of ahost of norms rules and values as well as of institutional arrangementsEach system is constantly changing and is open to inuence from othersystems And indeed many technologies and practices diffuse from onesociety to another but a societyrsquos capacity to be innovative is constrained

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

623

by the existing social system of production Thus the same technologymay exist in numerous countries but how it is employed and how itinuences societal outcomes and performances varies from one institu-tional conguration or society to another (Hollingsworth 1998)

PERSPECTIVES ON INSTITUTIONAL CHANGE

A fundamental problem which remains unresolved in institutionalanalysis involves the nature of institutional change There is much confu-sion and miscommunication in the social sciences about institutionalchange in large part because we do not even know how to measure therate of institutional change Of course we know that there are bothexternal and internal focuses for change and we have had a good bitof scholarship on this subject What is not clearly understood is thatinstitutions and institutional arrangements within societies are histori-cally rooted that there is a great deal of path-dependency to the waythat institutions evolve Moreover the more intricately linked each sectoris with each other sector and with a societyrsquos rules and norms the lesschoice actors have to devise new institutions and institutional arrange-ments On the other hand if a sector or actor is too isolated it may betoo weak within a society to bring about any kind of effective changeThese considerations are very critical to the problem of how muchfreedom actors have to create new institutions afresh to what extentinstitutions diffuse from one society to another (Murmann 1998 Lanzara1998) This is a fundamental question for students of institutionalanalysis as we attempt to understand how the different institutionallevels are linked togetherThere are a variety of reasons why there is confusion in studying insti-

tutional change There is disagreement over the extent to which actorshave the freedom to build new institutional arrangements and if theydo to what extent the new arrangements and organizational patternsmay depart from past practices Among some institutional economists(Williamson 1985) there is a basic assumption that actors build efcientinstitutions by pursuing their self-interest and by promoting theirstrategic goals Moreover there is the assumption that a mixing of fullrationality with market competition tends to produce all of the optimalinstitutions that are needed in order to coordinate a complex capitalisteconomy (Hollingsworth and Boyer 1997 ch 14 Hodgson 1999) Incontrast the argument developed here is much more complexSocial institutions are historically rooted and there is a great deal of

path-dependency in the way that various institutional componentsevolve The shape of institutional congurations at any moment limitsthe type of options for change Because there is a great deal of institu-tional inertia radical change in the institutional components of a society

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

624

is uncommon (Lanzara 1998) Recognizing the constraints existinginstitutions exert on actors we note that existing congurations of insti-tutional arrangements institutional sectors and organizational propertiescan limit but also inuence the degree to which norms and rules canchange Many scholars (North 1990 Ostrom 1990) have emphasized therole of norms rules ideologies and values in limiting the ability of actorsto develop new institutional arrangements which radically depart fromexisting ones However there is a two-way process of interaction aprocess of co-evolution (Baum and Singh 1994a Nelson 1994) Thus asFriedland and Alford (1991 244) argue actors develop new institutionalarrangements by recognizing organizational failures and low perfor-mance but these new arrangements do not necessarily change a societyrsquosnorms rules habits and its underlying value systemCampbell (1997) advances our understanding of how existing institu-

tions enable actors to construct radically new institutions He recognizesthat most institutions and institutional arrangements embody a degreeof inertia from which it is difcult for actors to depart Moreover thedifferential power relations among actors make it unlikely that lesspowerful actors can change the existing power structure of a societyMost change in the institutional components of a society evolves througha process of constrained selection which reects to a considerable degreethe existing arrangements and power relations This becomes the inter-pretative frame for social actors with institutionalized scripts and ritualswhich tend to be taken for granted and appear to be quite ordinary andnatural This institutionalized scripting of the social world is an extremelyimportant source of social stability and inertiaCampbell has developed a very fruitful explanation of change in terms

of the various institutional components outlined here He distinguishesbetween incremental and radical institutional change and argues thatradical institutional change occurs when social actors with widelydiffering norms cultural scripts and rituals for action engage in intenseinteraction with each other (see also Knight 1992) The more diverse theinteractions the greater the potential for institutional change He arguesthat as shifts in the composition of interaction among social actors occurchanges in the interpretation of problems and interests will follow Ifthere are only minor changes in the extent to which diverse actorsinteract there will be only modest institutional change But the morefundamental the changes in the interaction of diverse social actors thegreater the change in the way that actors interpret their world And thegreater the change in the way that actors interpret their world the morelikely that radical changes in various institutional components willfollow In other words diverse decision making tends to create a widerrange of interpretative frameworks of the social world than is likely tobe the case if social actors operate in isolation

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

625

This perspective of diverse interactions among social actors facilitatesour ability to move beyond the traditional view that existing institutionsconstrain the range of institutional alternatives which actors face Theolder view has tended to suggest that actors engage in institutionalchange by extending existing institutional principles habits and conven-tions Specifying the conditions which facilitate fundamental institutionalchange permits a much richer perspective for institutional analysisCampbellrsquos perspective about the conditions under which fundamentalor radical change in institutional components occurs is consistent withmy own research about the conditions under which major discoveriesor fundamental new knowledge occur the more scholars with diversebackgrounds interact with intensity and frequency the greater the like-lihood that they will develop new and alternative ways of thinking(Hollingsworth and Hollingsworth 2000)These views of institutional change are also consistent with the recent

work of Nelson and his colleagues (Mowery and Nelson 1999 Murmann1998) They report the research on a variety of industrial sectors in anumber of different countries revealing that networks of actors in partic-ular industries business and professional associations universityresearch departments and institutes are frequently able to use their collec-tive power to modify a societyrsquos property rights research system andgovernmental policies It is the intense and frequent interaction of thesediverse complex sets of actors ndash sometimes constituting a technologysystem ndash that often leads to incremental changes in a societyrsquos total socialsystem of production It is this kind of collective action that brings aboutorganizational change change in institutional sectors institutionalarrangements and even in the rules and norms of a society And if weare to understand how the style of innovativeness of a society changeswe must rst understand the changes in its institutional makeup It ishighly unlikely that a society will develop a new style of innovativenesswithout changing its institutional structure

FIRMS THEIR INSTITUTIONAL ENVIRONMENT ANDNATIONAL SYSTEMS OF INNOVATION

Variation in the innovative style of societies is shaped by their insti-tutional makeup ndash the various levels which are discussed aboveinstitutions institutional arrangements the structure and cohesion of institutional sectors which constitute a societyrsquos social system of pro-duction and the structure and culture of its organizations ndash especiallyits business rms and other research organizations By undertaking aninstitutional analysis of a society one can begin to understand in whatkinds of organizations the production of specic kinds of knowledgetakes place and how this is linked to particular kinds of innovativeness

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

626

With such a perspective one can gain a rich understanding of why somesocieties excel in the production of radical breakthroughs in basic andapplied science and in developing radically new products and why othersocieties excel in more incremental innovationsReecting on these issues one cannot help being impressed with the

contrasts between the innovative styles of German and American orga-nizations For example Germany has continued to be very successful in making incremental innovations in many industries in which it wasvery competitive well before World War II paper printing materialsmachinery electrotechnical products motor vehicles chemicals textileyarns and fabrics But the Germans have been much less successful indeveloping totally new products or related innovations in biotechnologyelectronics telecommunications aircraft as well as other newer indus-trial areas In contrast the Americans have been very innovative in these and other industries with products having very short lives andwith technologies which change very rapidly and are very complex(Hollingsworth 1991 Soskice 1997 1999)The importance of institutions and innovativeness for international

competitiveness has led to the recognition that there are national systemsof innovation (Lundvall 1992 Nelson 1993 Edquist 1997 Hage andHollingsworth 2000) But even before the recent studies of nationalsystems of innovation Landes (1969) was a proponent of these viewswhen he made the argument that Germany had developed a nationalsystem of innovation by building on its education and scientic researchsystems which were lacking in Britain and elsewhere However the concept of a national system of innovation immediately poses theproblem of what kind of innovations should be consideredDavid Soskice an economist at the Wissenschaftszentrum Berlin has

addressed a set of problems somewhat complementary to the perspec-tive raised here His work (Soskice 1990 1997 1999) suggests that weare sensitive to the institutional environment in which the followingactors are embedded the employees of rms the owners and nanciersof rms the competitors of rms and nally their collaborators ndash aperspective complementary to Hollingsworthrsquos work on social systemsof production (Hollingsworth 1997 1998) When one reects on the insti-tutional contextsocial system of production within which these actorsare embedded one becomes aware that those countries which have some-what inexible labor markets external to rms but a business system in which rms have long-term commitments to their employees arestrongly associated with incremental but hardly at all with radical inno-vations In the rms of such countries (eg Germany and Japan) thereis much more consensus decision making than in societies where a lsquohireand rersquo set of practices is quite pervasive And this kind of consensusdecision making ndash while highly conducive to incremental innovativeness

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

627

ndash limits the capacity for radical innovations to occur This kind of lsquolongtermismrsquo set of norms and rules occurs not only in the employmentpractices of incrementally innovative rms but also in their nancingand ownership patterns (Hollingsworth 1997) During the past thirtyyears these were characteristics of many rms in Germany and JapanSignicantly in both of these countries venture capital markets werenot highly developed nancing tended to be based more on long-termbank loans than on the equity markets and the relationships of rmstended to be relatively stable And these stable institutional arrange-ments also contributed to a high degree of innovativeness ndash butincremental rather than radical innovationsMeantime we should also pursue Soskicersquos ideas (1999) about how

rmsrsquo relationships with their competitors and collaborators shape theirstyles of innovativeness In those advanced industrial societies in whichbusiness rms are highly mobilized into business associations theretends to be a high degree of standardized quality controls over technicalnorms and a high degree of reliance on legal sanctions as monitoringinstruments vis-agrave-vis association members As a result of this high degreeof integration of rms into associations member rms are more likelyto be associated with incremental innovations than is likely to be the case with rms in those societies where associations are less welldeveloped In the latter type of society radical innovations are morelikely to occur In those advanced industrial societies where rms havelong-term stable relationships with collaborators in product design andproduction the style of innovation is likely to be incremental And it is in those societies in which there is a weaker tradition of strongcollaboration among rms that radical innovations are more likely tooccurObviously even in a country where there may be a great deal of

radical innovativeness most innovations will be of an incremental typefor there are simply fewer radical innovations But Germany and therest of Europe not only have had relatively few radical innovations incomparison with the United States but they also have lagged behindthe USA in developing relatively totally new industries Thus there hasbeen much less innovative activity in Europe in high-technology andnewly emerging industries than in the USAIn complementary research Audretsch and Feldman (1995) have

demonstrated that the USA has been more successful in spawning rmsengaged in the early stages of the industry life cycle but relative toGermany had an environment unfavorable to innovative activity byestablished rms especially in more mature industries German rmshave tended to be associated with a routinized technological regime ndashwhere established enterprises have had strong incentives to engage inincremental innovations but new rms have tended to have fundamental

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

628

innovative disadvantages relative to the Americans (see Audretsch 1995Winter 1982 Casper 2000)On both sides of the Atlantic universities engage in basic and applied

research But signicantly the institutional environment in which univer-sities are embedded inuences the degree to which the research is likelyto represent major breakthroughs or is likely to have more modestimplications Recent research in biomedical science demonstrates that inthose societies in which university professors andor graduate studentshave the following characteristics there are likely to be relatively majorresearch breakthroughs graduate students have a long training periodhigh dependency on a major professor low autonomy in dening thenature of a research project restricted job options and a university envi-ronment where there is a very modest or nonexistent entrepreneurialculture For example in Germany ndash a country with relatively few radicalbreakthroughs ndash the scholar who nishes a habilitation will have spentmany years on a research project before qualifying for a permanent acad-emic position will have been in a high-dependency position with a seniorprofessor will have had limited autonomy to dene a research projectand will be constrained from pursuing a high-risk research project Theyoung German will have an extraordinarily high level of training butwill have faced a number of disincentives to undertake high-risk researchAnd as soon as heshe has completed the habilitation and becomes aprofessor administrative responsibilities constitute another disincentivefrom engaging in high-risk or long-term research projects Moreover theGerman professor is a civil servant with a more stable and secure incomethan hisher American counterpart In a very different institutionalenvironment American professors in large research universities tend tobecome much more entrepreneurial Today many American professorsbehave as though they were operating their own rms within univer-sities and thus devote considerable time to generating funds for researchprojects assistants secretaries supplies etc And the variation in thisentrepreneurial spirit across countries is associated with variation in thetaking of high-risk research strategies and the making of major researchbreakthroughsIn sum societies vary in terms of the institutional environment in

which their organizations are embedded and this variation inuencesthe degree to which they have industrial activities in the early or latestages of the industry life cycle successful new or mature industriesand radical vs incremental innovations And it is this variation whichis encouraging rms in countries with relatively few new industries and with few or no radical innovations to invest in those countries with many new industries and numerous radical innovations ndash as astrategy to be able to transfer new technologies from the latter institu-tional environment to the former

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

629

Table 4 summarizes the way that the norms rules and values of a society are associated with Americarsquos and Germanyrsquos institutionalarrangements their social system of production and their styles of inno-vativeness

CONCLUSION

Several themes deserve to be reemphasized Most importantly institu-tional analysis is emerging as a dominant issue in several social sciencedisciplines The study of various institutional components discussedabove is a complex subject requiring multiple levels of analysis and aknowledge of multiple sectors within societies It requires a theoreticalperspective that evolves from an empirical analysis in different societiesIf we are to advance the study of institutional analysis we need cross-disciplinary collaboration and team research which are not readilyachieved given the structure of most of our research universitiesAt present scholars in each academic discipline have their own distinc-

tive strategies for studying institutional analysis Thus economists tendto use several approaches (North 1990 Williamson 1985 Hodgson 19881998 1999) anthropologists another (Geertz 1995) and sociologistspolitical scientists and historians employ other strategies (Campbell etal 1991 Powell and DiMaggio 1991 Katznelson 1998 Hall and Taylor1996) With practitioners in various disciplines pursuing their distinctiveapproaches to institutional analysis there is low potential for collabo-rative institutional analysis across the social sciences The lessons fromthe biological sciences over the last 40 years are very instructive oncepractitioners from various biological disciplines began to cooperate instudying biological phenomena by using the same concepts at the mole-cular level theoretical advance was very rapid Comparable phenomenatook place in the development of other hybrid elds eg biochemistrybiophysics (Judson 1979 Olby 1979 Kohler 1982 1994) Similarly weresocial scientists in various disciplines to work collaboratively in the areaof institutional analysis there would very likely be an acceleration intheir theoretical knowledgeThis is not a plea that everyone engaged in institutional analysis should

do the same kind of research Indeed just as biologists working at themolecular level study many different kinds of problems similarly socialscientists engaged in institutional analysis would also work on manyseparate problems but in the spirit of a collective enterprise Some wouldconcentrate their attention on the study of rules norms habits conven-tions and values while others would study how these are associatedwith congurations of various institutional arrangements (types ofmarkets hierarchies networks associations communities clans statesetc) Others would work on specic institutional sectors (eg education

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

630

Table 4

Rules norms hab

its an

d values associated w

ith sp

ecic co

ngu

ration

s of coo

rdinating mecha

nism

s soc

ial sy

stem

s of

prod

uction

an

d nationa

l styles of inno

vation

USA

Germany

Norms rules values which provide incentives for

Norms rules values which provide incentives for

short-term horizons

long-term horizons

Con

gu

rative

forms of

Marke

ts c

orpo

rate hierarchies reg

ulatory state

Assoc

iation

s (busine

ss assoc

iation

s la

bor union

s)

coordinating

corp

oratist-type

networks

marke

ts c

orporatist-

mecha

nism

s (listed in

type

state

order of im

portanc

e in

the particu

lar social

system

)

Social sys

tem of

prod

uction

Bus

iness system

Firm

struc

tures

Con

glom

erates b

ut with mov

emen

t toward

Firm

more co

ngru

ent with pa

rticular prod

ucts

elim

ination of produc

tion

in som

e sectors

emph

asis on long

-term strateg

ies rm

s are well

emph

asis on sh

ort-term

strateg

ies high

integrated

differentiation

between divisions

of rms

Firm

s mob

ilized in

Low

High

and in

tegrated

into busine

ss

associations

Indu

strial relations

Low

deg

ree of jo

b secu

rity rigid in

ternal lab

or

High deg

ree of jo

b secu

rity

exible interna

l lab

or

system

marke

t ex

ible externa

l labo

r marke

tsmarke

t rigid externa

l lab

or m

arke

tPatterns of

Dispersed owne

rship frequ

ent turnov

erMore stab

le owne

rship patterns

owne

rship

Table 4

continue

d

USA

Germany

Norms rules values which provide incentives for

Norms rules values which provide incentives for

short-term horizons

long-term horizons

Social system of

Prod

uction

Bus

iness system

Training for labo

r an

dHigh deg

ree of voluntary decision mak

ing by

Collective an

d m

ore co

mpulso

ry d

ecisions

abo

ut

man

agem

ent

individuals as to ho

w m

uch

tim

e an

d ene

rgy

how m

uch ed

ucation

is ne

eded

he

avy em

pha

sis

to inv

est in educ

ation heavy

emph

asis on

on techn

ical training (eg en

gine

ering)

busine

ss sch

ools training

in marke

ting

and

sales accou

nting

Fina

ncial marke

tsEqu

ity an

d ven

ture cap

ital m

arke

ts highly

Equ

ity an

d ven

ture cap

ital m

arke

ts w

eakly

dev

elop

ed na

ncing from

cen

tral gov

ernm

ent

deve

lope

d he

avy em

pha

sis on

corporate loan

sto priva

te rm

s for dev

elop

men

t of new

tech

nologies related

to na

tion

al security

Unive

rsity research

High individual auton

omy high poten

tial for

Low

ind

ividua

l au

tono

my until almost 40

yea

rs of

system

individual creativity u

nive

rsity pr

ofesso

rs

age poten

tial for ind

ividual creativity high

ly

high

ly entreprene

urial

cons

traine

d unive

rsity pr

ofessors not so

entrep

rene

urial

Inno

vative

styles

Highly su

ccessful in m

ajor break

through

s in

High pe

rforman

ce in proc

ess an

d inc

remen

tal

basic an

d app

lied scien

ce go

od in dev

elop

ing

inno

vation

s in existing produ

cts not so su

ccessful

totally

new

lines of pr

oducts an

d in m

ajor break

through

s in basic scien

ce or in

tech

nologies high

deg

ree of entrepr

eneu

rship

deve

loping

totally new

lines of prod

ucts of

amon

g research

ers

tech

nologies low d

egree of entrepr

eneu

rship

amon

g research

ers

Table 4

continue

d

USA

Germany

Norms rules values which provide incentives for

Norms rules values which provide incentives for

short-term horizons

long-term horizons

Prod

ucts in w

hich

Pr

oducts with sh

ort liv

es prod

ucts in w

hich

Older produ

cts with long

history machine

ry pap

er

society excels

tech

nology

is bo

th highly co

mplex

and

pr

inting

textile

s m

otor veh

icles m

achine

too

lsch

ange

s rapidly en

tertainm

ent indus

try

pha

rmaceu

ticals biotechn

olog

y softw

are

aerosp

ace

business systems nancial markets and systems of research) Distinctivewould be the recognition of how these separate studies are linkedtogether how each institutional sector is linked to a societyrsquos normsrules values etc and its conguration of institutional arrangementsSimilarly other institutional analysts might focus on the study of orga-nizations with a major concern for how the institutional environmentof organizations (eg norms rules values congurations of institutionalarrangements institutional sectors) inuenced the structure culture andoutcomes of organizations And nally others might study how all ofthe aforementioned aspects of institutional analysis inuenced a societyrsquosoverall performance Nelson and Sampat (1998) have appropriatelyreminded us that we should engage in institutional analysis in order toaddress specic problems but the view developed in this article is thatwe rst need some mapping of the terrain of institutional analysis inorder for our diverse enterprises to become a meaningful collective enter-prise especially if we are to understand how institutions inuence theinnovative processThe argument developed here is not that all social scientists should engage

in institutional analysis Obviously there are numerous other importantareas of research which lie beyond the eld of institutional analysis Nordoes this article argue that institutional analysis is the most importantsubject for social scientistsSince the structure and culture of our research universities reect an

enormous amount of disciplinary fragmentation what are the prospectsfor promoting the kind of collective and interdisciplinary research agendaproposed above In one respect the prospects are not encouragingDuring the past half century the efforts of American universities to have effective communication across the biological sciences have variedgreatly in their success At the University of California-Berkeley it was not until the 1980s that there was a serious effort to integrate thebiological sciences and at that time the campus abolished seventeendepartments to promote an integrated biological program But at theUniversity of Wisconsin the biological sciences are still very fragmentedinto dozens of different departments in botany plant pathology zoologyand physiology with two departments of biochemistry and two depart-ments of geneticsThere have been other approaches which have worked very well histor-

ically for promoting scientic integration amidst scientic diversity Forexample several summer schools and special institutes were organizedin the United States during the 1940s in an effort to overcome the frag-mentation in the biological sciences These programs made it possiblefor scientists in diverse academic disciplines and from different univer-sities and countries to come together to engage in intense and frequentinteraction Perhaps the most effective program was that held in the

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

634

summers at Cold Spring Harbor New York where a research agendafor the study of molecular biology was developed (Fischer and Lipson1988 Stent and Watson 1966) Scientists went there summer aftersummer and when they returned to their universities many began therestructuring of the biological sciences in their universitiesIn our own day we need summer institutes and other similar programs

so that scholars from diverse backgrounds can come together anddevelop a collective agenda for doing institutional analysis and forstudying the impact of institutions on the performance of societies ndashinnovativeness being one such subject If we can develop such instituteswe will increase the prospects for transforming our universities Thereis reason to be optimistic that once an eclectic group of scholars workwith a common set of concepts we can have genuine theoretical advanceif researchers can bring their diverse expertise to bear on a common setof problems There is considerable evidence that institutional analysisoffers high potential for advancing our knowledge of innovations andtechnological change (Hollingsworth and Hollingsworth 2000) If we aregoing to make substantial theoretical advance it is not likely to occurfrom the vantage point of a multiplicity of disciplines Rather it is likelyto occur as we develop a new intellectual framework and our own acad-emic journals This is a tall agenda which requires us to reassess theway that knowledge is presently produced in our societies

NOTE

This article is part of a much larger agenda involving the study of institutionsorganizations and innovations in which I am involved My debts are many andvaried The rst version was written while I was in residence in the ResearchUnit for Institutional Change and European Integration of the Austrian Academyof Sciences in Vienna in the summer of 1998 I am especially grateful to ProfessorsEgon Matzner and Sonja Puntscher-Riekmann for providing the stimulating envi-ronment in which to think through the ideas developed here Egon Matzner hasengaged in very stimulating discussions with me about the importance of insti-tutions in the general eld of socio-economics The nal version of the articlewas completed during my residence as a fellow at the Neurosciences Institutein La Jolla California I also wish to thank my colleagues in the InnovationTheme Project at the Netherlands Institute for Advanced Study (NIAS) during1998ndash9 and Edgar Grande and Karl Muumlller for stimulating conversations aboutthe interaction of institutions and organizations in facilitating particular typesof innovation I am especially grateful to Jerald Hage for helping me to developa number of issues in the article and for his useful comments on the rst draftFrans van Waarden early on helped me to work through the idea of multiplelevels of institutional analysis Others who have helped to develop the ideas inthe article have been my colleagues in the research project lsquoComparative socialsystems of productionrsquo especially Tom Burns Christel Lane Yoshitaka OkadaWolfgang Streeck and Richard Whitley My colleagues Robert Boyer SteveCasper and David Soskice have taught me much about how incentives providedby the institutional environments of rms inuence individual decision making

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

635

But my most important debt is to Ellen Jane Hollingsworth who has repeatedlyquestioned every single assumption of this article Without her rich and contin-uous tutoring this article would be even less developed than it is at presentFinally without the insightful assistance of David Gear in many phases of thisarticle it would not have been completed

REFERENCES

Aoki Masahiko (1990) lsquoTowards an economic model of the Japanese rmrsquo Journalof Economic Literature March 1ndash27

Arora A Landau Ralph and Rosenberg Nathan (eds) (1998) Chemicals and Long-Term Economic Growth Insights from the Chemical Industry New York JohnWiley amp Sons

Arrow Kenneth (1974) The Limits of Organization New York NortonArthur W B (1988a) lsquoSelf-reinforcing mechanisms in economicsrsquo in P W

Anderson K J Arrow and D Pines (eds) The Economy as an Evolving ComplexSystem Redwood City Calif Addison-Wesley

mdashmdash(1988b) lsquoCompeting technologies an overviewrsquo in G Dosi C Freeman RNelson G Silverberg and L Soete (eds) Technical Change and Economic TheoryLondon Pinter Publishers pp 590ndash607

Audretsch David B (1995) Innovation and Industry Evolution Cambridge MassMIT Press

Audretsch David B and Feldman Maryann P (1995) lsquoInnovative clusters andthe industry life cyclersquo Discussion Paper Wissenschaftszentrum BerlinMarket Processes and Corporate Development Research Unit FS-IV-95-7

Baum Joel A C and Oliver Christine (1992) lsquoInstitutional embeddedness andthe dynamics of organizational populationsrsquo American Sociological Review 57540ndash59

Baum Joel A C and Singh Jitendra V (1994a) lsquoOrganizationndashenvironment coevo-lutionrsquo in Joel A C Baum and Jitendra V Singh (eds) The EvolutionaryDynamics of Organizations New York Oxford University Press pp 379ndash402

mdashmdash(eds) (1994b) The Evolutionary Dynamics of Organizations New York OxfordUniversity Press

Berger Suzanne and Dore Ronald (eds) (1996) National Diversity and GlobalCapitalism Ithaca NY Cornell University Press

Boyer Robert (1997) lsquoThe variety and unequal performance of really existingmarkets farewell to Doctor Panglossrsquo in J Rogers Hollingsworth and RobertBoyer (eds) Contemporary Capitalism the Embeddedness of InstitutionsCambridge and New York Cambridge University Press pp 55ndash93

Burns Tom R and Dietz Thomas (2001 forthcoming) lsquoSocial rule system theorysocial action institutional arrangements and evolutionary processesrsquo in JRogers Hollingsworth and Karl Muumlller (eds) Advances in SocioeconomicsLanham Md Rowman amp Littleeld

Burns Tom R and Flam Helena (1987) The Shaping of Social Organization BeverlyHills Calif Sage

Calvert Randall L (1995) lsquoThe rational choice theory of social institutionsrsquo inJeffrey S Banks and Eric A Hanuskek (eds) Modern Political Economy NewYork Cambridge University Press pp 216ndash66

Camic Charles (1986) lsquoThe matter of habitrsquo American Journal of Sociology 911039ndash87

Campbell John L (1997) lsquoMechanisms of evolutionary change in economicsgovernance interaction interpretation and bricolagersquo in Lars Magnusson

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

636

The vertical axis deals with action motives Toward the upper part ofFigure 2 actors are engaged in individualistically oriented behaviorwhereas toward the lower part actors are more engaged in collectivebehavior and strive to cope with problems of common interest Cell 3 ndashcommunities and clans ndash consists of institutional arrangements based ontrust reciprocity or obligation and thus not derived from the pure selshcomputation of pleasures and pains This is an unconventional form ofcoordination for most neoclassical economists (however see Arrow1974) but not for many anthropologists political scientists and sociolo-gists (Streeck and Schmitter 1985a Polanyi 1944 Gambetta 1988Fukuyama 1995 Sabel 1992 Putnam 1993)In the neoclassical paradigm theorists argue that actors engage in

forms of exchanges that best promote their individual interests If somestructural conditions are fullled (absence of increasing returns to scalethe reversibility of transactions absence of uncertainty and completecontingent markets with no collusion between actors) then the invis-ible hand theorem applies and market-type activity functions quite welland also provides the optimum for society thereby combining efciencyharmony and order However an excess of market activity may welllead to ruinous competition and excessive conict There is variation inthe extent to which ruinous competition occurs depending on the socialcontext in which transactions take place Thus it is important that webe sensitive to the institutional context in which transactions areembedded and that we understand the degree to which social bondsexist at both the micro and macro levels of analysis Micro bonds facil-itate exchanges in a society but at the societal level social bonds existat the level of the collective ndash in the community or region and amongmembers of racial religious and ethnic groups All other things beingequal the more powerful the social bonds among transacting partnersthe more economic competition is likely to be restrained Thus mosttransactions occur not simply in an impersonal calculative system ofautonomous actors unrestrained by social ties ndash as implied by theneoclassical paradigm ndash but in the context of social ties variation in the strength of which leads to variation in levels of trust and transac-tion costs (Etzioni 1988 211 Granovetter 1985 Streeck and Schmitter1985a Hollingsworth and Boyer 1997 Hodgson 1988 1999 Schneibergand Hollingsworth 1990)Another form of multilateral institutional arrangement is various

types of associations (cell 6) Unlike networks clans and communitiesassociations are more formal organizations Whereas markets corporatehierarchies and networks tend to coordinate economic activity amongdifferent types of actors (eg producers with suppliers capital withlabor) associations typically coordinate actors engaged in the same orsimilar kinds of activities Business associations and labor unions are

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

608

some of the most common forms of associations for coordinatingeconomic activity in capitalist economies (Schneiberg and Hollingsworth1990 Streeck and Schmitter 1985a)Finally there is the state (cell 4) which is an institutional arrangement

quite unlike any of the others It is the state that sanctions and regu-lates the various non-state coordinating mechanisms that is the ultimateenforcer of rules of the various mechanisms that denes and enforcesproperty rights and that manipulates scal and monetary policy Overallit is the state that inuences the total incentive system of a society(Lindberg and Campbell 1991 Johnson 1992 40) At the same time thestate may also be an economic actor by engaging directly in productionand exchange relationsThe choices of institutional arrangements in Figure 2 are constrained

by the social context in which they are embedded Depending on thenature of that embeddedness there is variation in the collective formsof governance some of which are specied in the lower part of thegure Some modes of coordination in the bottom and upper levels ofthe typology are often mixed together though one coordinating modeis likely to be more prominent than another But actors are often engagedin complex congurations involving several kinds of institutionalarrangementsEach of these various institutional arrangements has its own logic ndash

its own rules its own procedures for enforcing compliance its ownnorms and ideologies which help to reduce the costs of enforcementThese are summarized in Table 2 which provides further elaborationabout the various coordinating mechanisms one nds in almost everycapitalist society Table 2 lists the organizational arrangements and theirstructures rules of exchange and means of enforcing compliance asso-ciated with each type of coordinating mechanism While each type ofinstitutional arrangement has various positive features each institutionalarrangement also has failures and these are featured in Table 3 It is thecontest between those who support and those who oppose these variousinstitutional arrangements that tends to lead to transformations in insti-tutional arrangements over time (Campbell et al 1991 Campbell 1997)Again it is difcult to conceive of pure institutional arrangements ndash

either exclusively markets or exclusively hierarchies ndash since it has beenwell known since Adam Smithrsquos Wealth of Nations that the division oflabor within the rm cannot be disentangled from the existence andextent of the market Institutional arrangements have their own distinc-tive form of efciency (good static efciency for the market dynamicefciency for rms) and inefciency often leading to considerableinequality Neither networks nor communities are panaceas for economiccoordination without being congured with other types of institutionalarrangements Networks and communities may solve certain issues but

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

609

Table 2

Log

ics of ins

titution

al arran

gemen

ts

Coordination

Organizational

Rules of

Individual means

Collective means

mechanisms

structure

exchange

of compliance

of compliance

Marke

tsEasy en

try an

d exit

Voluntary sp

ot excha

nge

Leg

al enforcemen

t of

Norm of pr

ivate

Bila

teral ex

chan

ge or

control

prope

rty

marke

t site (W

all Street)

Reg

ulations

to

Leg

itim

acy of m

arke

t en

force contracts

men

tality

Com

mun

ities

Inform

al m

embe

rship

Voluntary exch

ange

based

So

cial norms an

d m

oral

Highly institutiona

lized

ev

olve

s ov

er lon

g period

on social solid

arity an

d

princ

iples im

pose

norm

s an

d rules requ

ire

of tim

ehigh

deg

ree of tru

stob

ligations

mem

bers to accept

Kno

wledg

e of others an

d

lsquocorpo

ratersquoo

bligations

recipr

ocity ov

er tim

e

Networks

Semi-form

al m

embe

rship

Voluntary exch

ange

ove

r a

Con

tractual bon

dsPersona

l relation

sBila

teral or m

ultilateral

time pe

riod

Resou

rce de

pend

ence

Tru

st built outside

the

ex

chan

geecon

omic arena

Associatio

nsFo

rmal m

embe

rship

Restricted to mem

bers

Self-interest

Some deg

ree of

Multilateral excha

nge

Emph

asis on insider

Rep

utation

effects

compu

lsion

outsider or we

they

Priva

te interest type

of

men

tality

gove

rnan

ce

Private

Com

plex organ

izations

Restricted to mem

bers

Rew

ards to ind

ividuals

Highly institutiona

lized

hierarch

ies

which

ten

d to becom

e ex

chan

ge based

on

Asymmetric po

wer threat

rules

bureau

cratic hierarchies

asym

metric po

wer

of san

ctions

Mem

bers soc

ializ

ed into

bureau

cratic rules

corp

orate cu

lture us

e of

sanc

tions

State

Pub

lic hierarchy

Unilateral action

Exit vo

ice (vote lob

bying)

Coe

rcion

De jure

and impo

sed

Indirect political an

d

loya

ltyNorms an

d pu

blic rules

mem

bership

econ

omic excha

nge

Table 3

Failu

res of coo

rdinating mecha

nism

s

Coordinating mechanisms

Type of failure

Associations

Private hierarchies

State

Market

Communities

Networks

Enforcem

ent

Usu

ally relies on

Might enh

ance

Needs co

ntrol

Needs an

internal

Needs

tru

st and

Need an external

the state as an

oppo

rtun

istic

extern

al to state

enforcem

ent

loya

lty often

en

forcem

ent

enforcer

beha

vior

bureau

cracy

authority

coming from

au

thority

Resem

bles

The

ideal o

f (ju

dge

s

Facilitates

outside (fam

ily May

facilitate

enforcem

ent

internal m

arke

ts

parliamen

t collu

sion

and

relig

ion

cartelization an

d mecha

nism

might hur

t marke

t) to

impe

rfect

ethn

icity)

mon

opoly

of cartels

incu

mbe

nt

correct state

compe

titio

nCom

patible with

worke

rsab

uses

variou

s types of

Lob

bies can

co

mpe

tition

captur

e pu

blic

interest goa

ls

Public goo

d

Useful for

Gov

erna

nce co

sts

Can

provide

Can

not provide

Can

internalize

Useful for

and

establishing

might exceed

public goo

ds

colle

ctive go

ods

some co

llective

enha

ncing qu

ality

externality

stan

dards an

d

the be

nets of

but ha

s or deal w

ith

good

s (qua

lity

and training bu

t qu

ality for

internal division

difc

ultie

s in

externalities

training

) bu

t no

t no

t ve

ry goo

d in

setting ru

les of

of la

bor

prov

iding them

Inad

equate

othe

rs (welfare

prov

iding for

compe

titio

n in

Slow

to react to

in precise

mon

itoring of

gene

ral p

ublic

societal gen

eral

the indu

stry

chan

ges in the

am

ounts

tech

nical ch

ange

go

ods)

welfare

Useful for

environm

ent

Might fail in

and inno

vation

Mem

bers tightly

Weak in the

prov

iding

induc

ing

integrated

into

prov

ision of

man

y go

ods

tech

nical

community

colle

ctive go

als

colle

ctively

chan

geha

ve limite

d

that in

dividua

lcapa

city for

mem

bers can

not

inno

vatio

nsprov

ide for

them

selves

Table 3

continue

d

Coordinating mechanisms

Type of failure

Associations

Private hierarchies

State

Market

Communities

Networks

Efcien

cyFa

cilitates

De

cien

t in

Can

be high

ly

Some ba

sic social So

me go

ods

Slow

to en

hanc

e co

operation an

d

coop

eration an

d

bureau

cratic

relation

s cann

ot

cann

ot be

efc

ienc

y an

d

X-efcien

cy but

X-efcien

cyan

d can

not

be provided

by

delivered

at

speed of

not allocative

easily deliver

pure m

arke

t su

fcien

tly

adap

tive

ness

efcien

cy

good

s at

mecha

nism

slow costs

except in

low cost

indus

tries whe

re

tech

nology

is

complex an

d rapidly cha

nging

Equity

Narrow

Excessive

Might enh

ance

Facilitates

Might lead to

Whe

n widely

encompa

ssing

multiplication

ineq

uality

ineq

uality in

retard

ed

dev

elop

ed in

to

associationa

l of con

trollers

(pow

er

income an

d dev

elop

men

tindus

trial

stru

ctures lead

(fru

stration

and

an

d privilege

)wealth

districts

to in

come

ineq

uality)

netw

orks m

ay

equa

lity

facilitate greater

equality an

d

inco

me

distribution

whe

n wea

kly

dev

elop

ed

netw

orks ten

d to

increase soc

ial

ineq

uality

Note For allo

cativ

e efc

ienc

y an

d X

-efcien

cy see Leibe

nstein (19

66 19

76)

they raise other no less severe problems It is important to recognizethe imperfection of any single institutional arrangement in order tocomprehend the origin and transformation of any other institutionalarrangement (Hodgson 1999 ch 3)As suggested above each institutional arrangement is congured with

other institutional arrangements Usually one particular institutionalarrangement is more dominant in a particular conguration but becauseeach type of institutional arrangement has its own strengths andweaknesses there is no simple structural logic in the governance or coor-dination of a society Each institutional arrangement constrains the otherbut the inherent tension among the various institutional arrangementswithin a conguration contributes to changes in the governance orcoordination of a society The routines and logics of each of these cong-urations of institutional arrangements provide constraints and incentivesfor actors The way that actors perceive the incentives and constraintsof these governance congurations leads to a particular market logicand it is the specic market logic of a society which inuences its speciccapacities and weaknesses The inherent strain among the different logicsin a conguration of coordination helps to provide the exibility for asociety to adapt to new circumstancesIf there were a society with pure markets or a society coordinated

only by the state there would be too much rigidity and too little diver-sity to cope with the vast uncertainty in the global environment A societywith very little diversity in its coordination mechanisms would havelittle capacity to adapt to new circumstances Congurations with consid-erable diversity of institutional arrangements provide for a certainamount of incoherence in governance but they also provide for thecapacity to adapt to new circumstances While the Soviet Union wasdominated by the state as a coordinating mechanism there were alwaysfunctioning markets in the system Moreover the feudal society of thetwelfth and thirteenth centuries consisted not only of hierarchical rela-tionships between serfs and masters but also of urban guilds clericalhierarchies and markets (Johnson 1992 38) In sum the robustness ofinstitutions often depends on multiple and diverse principles and logicsof action on the inconsistency of principles and procedures on patternedforms of disorder (Lanzara 1998 Orren and Skowronek 1991 320 329)

Third level of analysis

The next level consists of the institutional sectors of a society The rulesnorms and values of a society inuence the array of institutional arrange-ments and levels one and two inuence the nature of and the relation-ships among various institutional sectors and all three levels areintricately linked together to form a social system of production Together

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

613

all of these components inuence the performance of economic sectorswithin a society as well as the performance of the total society A socialsystem of production is the way that a societyrsquos institutions (see rstlevel of analysis above) its institutional arrangements (see second levelof analysis above) and its institutional sectors are integrated into a socialcongurationAn institutional sector includes all organizations in a society which

supply a given service or product along with their associated focal orga-nizations (eg major suppliers funding sources regulators and so forth(Scott et al 1994 108 117 Campbell et al 1991 Hollingsworth et al1994) Institutional sectors include but are not limited to the societyrsquossystem of education system of research business system nancialmarkets legal system and the state The structure of the nancial marketssystem of training and education industrial relations system system ofresearch and state and legal systems are distinct and idiosyncratic ineach society In short they are system specicAll of these institutional sectors tend to cohere with each other

although they vary in the degree to which they are tightly coupled intoa full-edged system While each of the institutional sectors has someautonomy and may have some goals that are contradictory to the goalsof other institutional sectors with which it is linked an institutional logicin each society leads institutions to coalesce into a complex social con-guration (Hollingsworth 1991) This occurs because the institutionalsectors are embedded in a culture in which their logics are symbolicallygrounded organizationally structured technically and materially con-strained and politically defended The conguration of institutionalsectors usually exhibits some degree of adaptability to new challengesbut continues to evolve within an existing style But under new circum-stances or unprecedented disturbances these institutional congurationsare exposed to sharp historical limits as to what they may or may notdo (Schumpeter 1983 David 1988 Arthur 1988a 1988b Haringkanssen andLundgren 1997)Why do all of these different institutions coalesce into a complex

social conguration which is labeled here as a social system of pro-duction The literature suggests two contrasting interpretations Part ofthe answer ndash indeed a controversial one ndash is that these institutionalsectors are functionally determined by the requirements of the practiceof capitalism in each time and place (Habermas 1975) Another expla-nation emphasizes the genesis of the actual conguration via a trial and error process according to which the survival of rms regions orcountries is the outcome of complex evolutionary mechanisms (Maynard-Smith 1982 Nelson and Winter 1982) However the problem is evenmore complex Markets and other mechanisms for coordinating rela-tionships among economic actors place constraints on the means and

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

614

ends of economic activity to be achieved in any society The othercoordinating mechanisms include different kinds of hierarchies varioustypes of networks and associations ndash trade unions employers and busi-ness associations (Hollingsworth and Lindberg 1985 Campbell et al1991) The logic of these various institutional arrangements providesactors with vocabularies and logics for pursuing their goals for deningwhat is valued and for shaping the norms and rules by which theyabide (see Tables 2 and 3) In short in contrast to the logic of theneoclassical paradigm the argument here is that the dominant type ofinstitutional arrangements places severe constraints on the denition of needs preferences and choices of economic actors Whereas the neo-classical paradigm assumes that individuals and rms are sovereign thisarticle is based on the assumption that rms are inuenced by the hold that the institutional congurations making up a social system ofproduction have on individual decision making (Campbell et al 1991Etzioni 1988 Streeck and Schmitter 1985a Hollingsworth et al 1994Hollingsworth and Boyer 1997 Magnusson and Ottosson 1997 North1990 Hodgson 1999)Standard neoclassical economic theory has tended to downplay the

role of production and consequently of rms Even transaction cost theo-rists who are concerned with analyzing the rm as a coordinatingmechanism have been relatively unconcerned with the various compo-nents of a social system of production Indeed as long as there waswidespread optimism about the efcacy of Keynesian economics therewas relatively little concern among neoclassical economists with thesupply side of the economy Even in the opinion of most Keynesians agroup of experts should ideally be able to shape the size of aggregatedemand while the supply side of the economy would be left to the twominimalist institutions of neoclassical economics ndash markets and manage-rial hierarchies For several decades however it has become increasinglyobvious that some of the most competitive and successful patterns ofindustrial output and industrial production in capitalist economies donot derive from the neoclassical prescription of unregulated markets andcorporate hierarchies complemented by a neoliberal democratic stateIndeed it is now well understood that certain highly successful produc-tion patterns require for their emergence and survival institutionalarrangements and environments the very opposite of the prescriptionsfound in the neoclassical paradigm (see especially Streeck 1992 but alsoHollingsworth and Streeck 1994) Thus social scientists have come tounderstand that if they are to comprehend the behavior and performanceof contemporary economies concerns about social systems of produc-tion must be brought into the pictureBecause production involves more than technology a number of social

scientists have an increasing concern with social systems of production

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

615

The same equipment is frequently operated quite differently in the samesectors in different countries even when rms are competing in the same market (Maurice et al 1980 Sorge and Streeck 1988 Sorge 1989Hollingsworth et al 1994) Variations in production and process tech-nologies are inuenced partly by variations in the social environmentsin which they are embedded In other words rms are embedded intocomplex environments which among other things place constraints ontheir behavior Thus a social system of production is of major impor-tance in understanding the behavior and performance of an economyHow the state and other institutional arrangements (eg marketsnetworks private hierarchies associations) coalesce and are related toparticular social systems of production are important determinants ofeconomic performanceThe congurations of institutional arrangements that coordinate or

govern the behavior of actors in one society and its structure of specicinstitutional sectors cannot easily be transferred to another society forthey are embedded into a social system of production that is societallydistinct (Hollingsworth 1997) Societies borrow selected principles offoreign management styles and work practices but the effectiveness ofsuch borrowing is generally limited Economic behavior and performanceare shaped by the entire social system of production in which actors areembedded and not simply by specic principles of particular manage-ment styles and work practices Moreover a societyrsquos social system ofproduction tends to limit the kind of goods that it can produce and withwhich it can compete successfully in international marketsSome scholars (Kenney and Florida 1993 Oliver and Wilkinson 1988)

have assumed that the diffusion of particular forms of management stylesand work practices across societies could lead to system convergenceReferring to the Japanese production system Kenney and Florida haveargued that it consists of organizational practices whose fundamentallsquogenetic codersquo can successfully be inserted into another society and beginsuccessfully to reproduce its behavior in the new environment (1993 8)Their position is in the intellectual tradition of Antonio Gramsci (1971)who contended decades ago that the American system of mass produc-tion would most certainly diffuse to Europe over timeHowever the argument of this article is that even though British

French and American rms may adopt certain aspects of Japanesemanagement styles (eg just-in-time production self-managing teamsquality circles the use of statistical process controls) or some variant ofthe German vocational system their social systems of production willnot converge Moreover the overall conguration of a societyrsquos socialsystem of production inuences its sectoral and overall national economicperformance This helps to explain why societies have different systemsof innovation and why they vary in the clusters of industries in which

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

616

they are highly competitive in the international markets (Nelson 19931999 Hage and Hollingsworth 2000 Berger and Dore 1996)In order to understand how and why a societyrsquos economy performs

as it does it is necessary to understand its entire social system of produc-tion If a society is to modify substantially the performance of itsinstitutional sectors and its entire economy it cannot adapt only someof the management and work practices of its foreign competitors Ratherit must alter its entire social system of production Because a societyrsquosmodes of economic governance and coordination and its institutionalsectors develop according to particular logics and are system specicthere are serious limitations to the extent to which a society may mimicthe institutions (eg the rules and norms) institutional arrangements andinstitutional sectors of other societiesIn the history of modern societies there are logics by which institu-

tions coalesce into a social system of production (Hollingsworth 1991)Though institutions are constantly changing there are sharp limits tothe type and the direction of change that any particular institution canundergo because of its linkages with institutional arrangements and insti-tutional sectors Thus a societyrsquos business rms educational systemnancial markets industrial relations system etc can engage in seriousrestructuring only if most of the other institutional sectors also changeIn social systems of production there are pressures toward consistencyin the norms rules and values across institutional sectors though in anycomplex society social systems are obviously imperfectly integratedIndeed the degree to which the institutional norms and rules makingup a social system of production are loosely or tightly coupled is a vari-able of considerable importance In general the institutional sectorsmaking up a social system of production are interdependent and changesin one generally result in changes in the others Since each institutionallevel is dependent on the others for various types of resources there isinterdependence among the differing institutional spheres Moreovereach society has its norms moral principles rules and laws and recipesfor action as well as its own idiosyncratic customs traditions and prin-ciples of justice (Burns and Flam 1987)There are also other inherent obstacles to convergence among social

systems of production for where a system is at any one point in timeis inuenced by its initial state Systems with quite different initial statesare unlikely to converge with one anotherrsquos institutional practicesExisting institutional arrangements block certain institutional innova-tions and facilitate others (Roland 1990) Thus the institutions makingup a social system of production provide continuity even though insti-tutional arrangements are always changing but with a particular logicWhile Williamson (1975 1985) suggests that actors tend to choose theinstitutional arrangements which are most efcient North (1990) is

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

617

much closer to the mark in his argument that most societal institutionalarrangements exist as a result of custom and habit and are thereforeinefcient At any moment in time the world often appears to its actorsas very complex and uncertain For this reason actors often engage incontradictory forms of behavior pursuing different strategies as hedgesagainst a very uncertain world (Lanzara 1998) And their hedging andcontradictory forms of behavior may lead in somewhat different soci-etal directions all constrained by the institutional fabric within whichthe actors are embedded This kind of contradictory behavior occurs inpart because of the contradictions inherent in the dominant congura-tions of institutional arrangements with their contradictory logics (seelsquoSecond level of analysisrsquo above)Despite the emphasis on the logic of institutional continuity this is

not an argument that systems change along some predetermined pathThere are critical turning points in the history of highly industrializedsocieties but with the choices limited by the existing institutional terrainBeing path-dependent a social system of production continues along aparticular logic until or unless a fundamental societal crisis intervenes(Milgrom et al 1991 Krugman 1991 Durlauf 1991 Hollingsworth 1991Pred 1966 David 1988 Hodgson 1998)At this point it is important to confront the question of how social

systems of production evolve And this question gets to the heart of theproblem of building complex societal structures Certainly social systemsof production did not emerge from some process of social engineeringMoreover the various component parts of each social system of produc-tion have often not been designed to be part of a social system ofproduction The component parts of each social system of productionhave emerged more often than not as unintended by-products of goalswhich various actors had in mind at earlier moments in time It is usuallythe case that actors at time lsquotrsquo fail to comprehend the long-term conse-quences of their actions Wolfgang Streeck the author of several brilliantpapers on social systems of production argues that institutional sectordesigns created for one purpose generally address that goal but overtime those designs have quite unintended consequences (Streeck 1997a1997b Wright 1998) Describing the conguration of the German socialsystem of production of the 1970s and 1980s Streeck points out that itwas the unintended by-product of multiple points in history Someelements were pre-Wilhelmian others were introduced by the Alliedpowers after 1945 and others emerged during the years of the GermanFederal Republic All component parts of the German social system ofproduction

were and continue to be changing for their own reasons as wellas in reaction to each other and certainly there can be no presump-tion of a pre-established t between them even though one might

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

618

want to allow for some reinforcement effects of the modelrsquos histor-ically contingent social and economic success

(Streeck 1997b 54)

In sum the emergence of social systems of production is a long-termevolutionary process with each part interacting with its environment andresulting in a congurative whole but with no previous design by eithera single actor or a collectivity of actors As suggested above actors do not recreate their world anew At any moment in history the totalinstitutional complex of a society necessarily contains the resources and legacies of its past Preexisting institutional complexes are nevercompletely wiped out Societies consist of multiple layers of history withtheir diverse logics of action However the cumulative effect of smallperipheral changes in altering particular institutional sectors and insti-tutional arrangements can be substantial (Murmann 1998 Lanzara 1998)lsquoInstitution building is affected more by the ways in which people havecodied the past than by how they envision the future There are manymore potential resources in pre-existing institutional arrangements thanit is usually assumed or suspectedrsquo (Lanzara 1998 30)

Fourth level of analysis

The next level of institutional analysis organizational structures is some-what more controversial As suggested above Douglass North (1990)draws a sharp distinction between institutions and organizations Morerecently many organizational theorists (Powell 1991 Powell andDiMaggio 1991 DiMaggio and Powell 1983 Baum and Oliver 1992Kondra and Hinings 1998 Townley 1997) argue that institutional rulesnorms and conventions unfold in tandem with organizational structuresand this is my positionThe literature which focuses on how institutions inuence organiza-

tions is quite different from two other theoretical literatures which alsoare concerned with how organizational environments shape the behaviorof organizations There are resource dependency theorists (Pfeffer andSalancik 1978) who emphasize the role of environmental resources inshaping organizations while Hannan and Freeman (1977 1984) and otherpopulation ecologists emphasize the survival of organizations givencertain kinds of organizational conditions (Orru et al 1991)The institutionalist perspective as implied above emphasizes the

normative environment in which organizations are embedded It is a perspective which focuses on the way in which organizations in their behavior tend to conform to the institutional rules and normswhich are dominant in the organizational environment However allthree perspectives ndash the institutionalist the population ecology and

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

619

the resource dependency ndash emphasize how the environment inuencesorganizations and how organizations which are subject to the sameenvironment tend to converge in their behavior to have what DiMaggioand Powell have labeled organizational isomorphism (DiMaggio andPowell 1983)There has emerged a vast literature which demonstrates that within

each society there are scal political judicial and other regulatory normswhich limit and shape the culture and structure of organizationalbehavior The normative institutional environment of organizations limits the options of what organizations do in a particular society and inuences the patterns of ownership relations with suppliers andcustomers In short it is the normative environment of organizationswhich denes within a particular society what is socially acceptablebehavior for organizations (Hamilton and Biggart 1988 Hollingsworthand Hollingsworth 2000 Meyer and Rowan 1991 Zucker 1987 1988Orru et al 1991 Townley 1997)Thus far most of the analyses on isomorphic pressures operating on

organizations have been cross-sectional in nature However the entireconcept of isomorphism implies that there are strong environmental pres-sures exerted on organizations and in order to observe this phenomenonit is important that we have longitudinal studies which assess howchanges in an organizationrsquos institutional environment inuence changesin the structure and culture of organizations Moreover the expectationhere is that when one analyzes the patterns of organizational changeover time the isomorphism will be holistic ndash that is within a particularsociety there will be a tendency for the internal structure of particulartypes of organizations to converge in large part because of the pres-sures to conform to the changes in the external norms rules and valuesA close reading of Aoki (1990) and Slack and Hinings (1994) nds thatchanges in the national institutional environment of organizations haveinuenced changes in the internal structure of organizations within aspecic industry of a particular societyHowever it is inappropriate to view organizations as changing only

in relation to exogenous environmental change There is a kind of co-evolution which occurs between organizations and their institutionalenvironments Co-evolution implies nonlinear feedback between orga-nizations and their institutional environments This is of course quitedifferent from the more common way of studying organizations bymodeling relations between independent and dependent variables Ofcourse the more common type of analysis is quite appropriate whenthere are simple relationships which do not involve complex feedbackprocesses But in the strategy for institutional analysis which is proposedherein it is less useful to separate independent from dependent vari-ables and more useful to understand the interacting and co-evolutionary

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

620

processes among all of the levels of institutional analysis discussed inthe various sections above (Baum and Singh 1994a 379ndash80 Mowery andNelson 1999 Murmann 1998)The logic of the perspective herein suggests that within every

society there is variation in the structure and culture of business rmsuniversities and other complex organizations (Kondra and Hinings1998) However this variation takes place within parameters which aresystem specic For example every German rm and every Germanuniversity is different from every other German university and rm but there are a set of cultural and structural characteristics which distin-guish German rms and universities from those in America Similarlyeach Japanese rm and university is unique yet they have a set ofcultural and structural characteristics which are system specic andwhich differentiate them from business rms and universities in other countriesIn all societies each organization has its own distinctive organizational

rules norms and conventions which are subordinate to the meta norms and rules of the larger society within which they are embeddedHowever the strength of the institutional environment within whichorganizations are embedded varies from society to society Some societies have multiple institutional environments and there is hetero-geneity in terms of what constitutes appropriate behavior oforganizations In such societies there can be conict over whatinstitutional logics should regulate specic organizational functions(Friedland and Alford 1991 Townley 1997 262ndash3) In those societies inwhich the institutional norms habits and rules are most developed and in which the institutional pressures to conform are greatest thereis less variation in the structure and culture of business rms and variouskinds of research organizations In such societies the connectednessbetween research organizations and their institutional environments is sufciently strong for organizations to have low autonomy to pursue independent strategies and goals in these societies there is a great deal of organizational isomorphism Conversely the weaker the institutional environment in which research organizations areembedded the greater the variation in the structure and culture of business rms and research organizations Moreover where the institu-tional environments are more weakly developed organizations havegreater autonomy and exibility to respond to the development of new knowledge and to be highly innovative Hence it is in those societies where the institutional environments are most developed and most rigid and when there is less organizational autonomy and exibility that fewer radical innovations in basic and applied science as well as in totally new products and industrial sectors havebeen made

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

621

Fifth level of analysis

We come next to the outputs and performance of the various institutionalcomponents of a society It is at this level that institutional componentsare more pragmatic and exible For example in the legal sector thereare specic statutes and court rulings in the state sector there are specicpolicies in the business sector there are new products new technolo-gies and market strategies It is at this nal level of analysis thatinstitutional spheres are most open and susceptible to change and cross-national mimicry is easiest and most commonThrough the outputs of the society we can obtain some assessment as

to how well a society is performing Moreover we can assess how inno-vative it is how egalitarian it is in the distribution of its resources howegalitarian it is in terms of levels of health education etc Just as it iscomplicated to measure how egalitarian a society is similarly it is dif-cult to assess how innovative it is For example societies vary in makingincremental and radical breakthroughs in basic and applied science in developing totally new products and new kinds of organizations inincremental and process innovations in existing products and organiza-tions in developing new and different forms of marketing ndash bothdomestically and globally (Hage and Hollingsworth 2000) Of coursesocieties vary in their rates of economic growth in their rates of economicproductivity in their quality of life (eg rates of crime life expectancyetc) All of these performance criteria feed back and inuence each ofthe levels of institutional analysis discussed above ndash rules norms valuesetc institutional arrangements institutional sectors the structure andform of organizations Moreover performance measures may inuenceeach other For example if a society has very low rates of growth itmay not be very innovative in some of the types of innovative activitiesmentioned above Good or poor performances of certain kinds inuenceother performance indicators (Hollingsworth et al 1990)Different institutional arrangements and different social systems of

production result in different types of economic performance Hence solong as societies have different social systems of production there areserious constraints on the degree to which they can converge in theirinnovative styles Different social systems of production tend to maxi-mize in a more or less explicit manner different performance criteriausually mixing considerations about static and dynamic efciency protsecurity social peace and economic andor political power In short incontrast to the implications of neoclassical economic theory in real-worldeconomies there are no universal standards all economically rationalactors attempt to maximize Economic history provides numerous exam-ples of how a great variety of principles of rationality are implementedin different societies

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

622

A critical question in institutional analysis is whether an existing socialsystem of production which supports a set of routines for a particularkind of technology and industry can shift from old practices to newones There are numerous historical accounts of how the social systemof production in a particular society which worked so well for a numberof years could not adapt to new technologies (Lazonick 1990 Veblen1915 Nelson 1994) Schumpeter (1983) and Freeman (1991) have devel-oped the idea of a techno-economic paradigm which suggests thatdifferent technologies require specic forms of organizational arrange-ments ndash or what I called social systems of production Because oflsquoinstitutional inertiarsquo the social system of production of a society maynot be able to adapt to a new techno-economic paradigm (Nelson 199458 Lanzara 1998)Whether a social system of production can sustain its performance

standards depends not only on its intrinsic economic rationality but alsoon where it ts into a larger system If a particular social system ofproduction is immune from the innovativeness and competition of analternative system survival can be long lasting But if different socialsystems of production with diverging criteria of good economic perfor-mance meet in the world arena the arbitrariness of nationally imposedconstructed performance standards may be superseded by alternativeperformance criteria as a result of international competitivenessThe world economy is also socially constructed just as are national

economies Even if different social systems of production are competingin the international arena it is not always possible to determine whichis more competitively effective at any moment in time Hegemonicnation-states in the short run shape the rules of trade to favor their indus-trial sectors and rms But the history of hegemonic powers suggeststhat in the longer run such social systems of production sustainedlargely by military and political power eventually give way to moreinnovative and competitive social systems of production In our ownday as nation-states are increasingly integrated into a world economyeven hegemonic powers lose their innovative and competitive advan-tage and their share of world output decreases Such a country mayattempt to restructure its institutional arrangements and to readjust itsperformance preferences But to restructure its system of innovationgenerally calls for a major redistribution of power within a societyLargely for this reason societies have historically had limited capacityto reconstruct a system of innovation in the image of their major competi-tors Each countryrsquos social system of production is a conguration of ahost of norms rules and values as well as of institutional arrangementsEach system is constantly changing and is open to inuence from othersystems And indeed many technologies and practices diffuse from onesociety to another but a societyrsquos capacity to be innovative is constrained

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

623

by the existing social system of production Thus the same technologymay exist in numerous countries but how it is employed and how itinuences societal outcomes and performances varies from one institu-tional conguration or society to another (Hollingsworth 1998)

PERSPECTIVES ON INSTITUTIONAL CHANGE

A fundamental problem which remains unresolved in institutionalanalysis involves the nature of institutional change There is much confu-sion and miscommunication in the social sciences about institutionalchange in large part because we do not even know how to measure therate of institutional change Of course we know that there are bothexternal and internal focuses for change and we have had a good bitof scholarship on this subject What is not clearly understood is thatinstitutions and institutional arrangements within societies are histori-cally rooted that there is a great deal of path-dependency to the waythat institutions evolve Moreover the more intricately linked each sectoris with each other sector and with a societyrsquos rules and norms the lesschoice actors have to devise new institutions and institutional arrange-ments On the other hand if a sector or actor is too isolated it may betoo weak within a society to bring about any kind of effective changeThese considerations are very critical to the problem of how muchfreedom actors have to create new institutions afresh to what extentinstitutions diffuse from one society to another (Murmann 1998 Lanzara1998) This is a fundamental question for students of institutionalanalysis as we attempt to understand how the different institutionallevels are linked togetherThere are a variety of reasons why there is confusion in studying insti-

tutional change There is disagreement over the extent to which actorshave the freedom to build new institutional arrangements and if theydo to what extent the new arrangements and organizational patternsmay depart from past practices Among some institutional economists(Williamson 1985) there is a basic assumption that actors build efcientinstitutions by pursuing their self-interest and by promoting theirstrategic goals Moreover there is the assumption that a mixing of fullrationality with market competition tends to produce all of the optimalinstitutions that are needed in order to coordinate a complex capitalisteconomy (Hollingsworth and Boyer 1997 ch 14 Hodgson 1999) Incontrast the argument developed here is much more complexSocial institutions are historically rooted and there is a great deal of

path-dependency in the way that various institutional componentsevolve The shape of institutional congurations at any moment limitsthe type of options for change Because there is a great deal of institu-tional inertia radical change in the institutional components of a society

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

624

is uncommon (Lanzara 1998) Recognizing the constraints existinginstitutions exert on actors we note that existing congurations of insti-tutional arrangements institutional sectors and organizational propertiescan limit but also inuence the degree to which norms and rules canchange Many scholars (North 1990 Ostrom 1990) have emphasized therole of norms rules ideologies and values in limiting the ability of actorsto develop new institutional arrangements which radically depart fromexisting ones However there is a two-way process of interaction aprocess of co-evolution (Baum and Singh 1994a Nelson 1994) Thus asFriedland and Alford (1991 244) argue actors develop new institutionalarrangements by recognizing organizational failures and low perfor-mance but these new arrangements do not necessarily change a societyrsquosnorms rules habits and its underlying value systemCampbell (1997) advances our understanding of how existing institu-

tions enable actors to construct radically new institutions He recognizesthat most institutions and institutional arrangements embody a degreeof inertia from which it is difcult for actors to depart Moreover thedifferential power relations among actors make it unlikely that lesspowerful actors can change the existing power structure of a societyMost change in the institutional components of a society evolves througha process of constrained selection which reects to a considerable degreethe existing arrangements and power relations This becomes the inter-pretative frame for social actors with institutionalized scripts and ritualswhich tend to be taken for granted and appear to be quite ordinary andnatural This institutionalized scripting of the social world is an extremelyimportant source of social stability and inertiaCampbell has developed a very fruitful explanation of change in terms

of the various institutional components outlined here He distinguishesbetween incremental and radical institutional change and argues thatradical institutional change occurs when social actors with widelydiffering norms cultural scripts and rituals for action engage in intenseinteraction with each other (see also Knight 1992) The more diverse theinteractions the greater the potential for institutional change He arguesthat as shifts in the composition of interaction among social actors occurchanges in the interpretation of problems and interests will follow Ifthere are only minor changes in the extent to which diverse actorsinteract there will be only modest institutional change But the morefundamental the changes in the interaction of diverse social actors thegreater the change in the way that actors interpret their world And thegreater the change in the way that actors interpret their world the morelikely that radical changes in various institutional components willfollow In other words diverse decision making tends to create a widerrange of interpretative frameworks of the social world than is likely tobe the case if social actors operate in isolation

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

625

This perspective of diverse interactions among social actors facilitatesour ability to move beyond the traditional view that existing institutionsconstrain the range of institutional alternatives which actors face Theolder view has tended to suggest that actors engage in institutionalchange by extending existing institutional principles habits and conven-tions Specifying the conditions which facilitate fundamental institutionalchange permits a much richer perspective for institutional analysisCampbellrsquos perspective about the conditions under which fundamentalor radical change in institutional components occurs is consistent withmy own research about the conditions under which major discoveriesor fundamental new knowledge occur the more scholars with diversebackgrounds interact with intensity and frequency the greater the like-lihood that they will develop new and alternative ways of thinking(Hollingsworth and Hollingsworth 2000)These views of institutional change are also consistent with the recent

work of Nelson and his colleagues (Mowery and Nelson 1999 Murmann1998) They report the research on a variety of industrial sectors in anumber of different countries revealing that networks of actors in partic-ular industries business and professional associations universityresearch departments and institutes are frequently able to use their collec-tive power to modify a societyrsquos property rights research system andgovernmental policies It is the intense and frequent interaction of thesediverse complex sets of actors ndash sometimes constituting a technologysystem ndash that often leads to incremental changes in a societyrsquos total socialsystem of production It is this kind of collective action that brings aboutorganizational change change in institutional sectors institutionalarrangements and even in the rules and norms of a society And if weare to understand how the style of innovativeness of a society changeswe must rst understand the changes in its institutional makeup It ishighly unlikely that a society will develop a new style of innovativenesswithout changing its institutional structure

FIRMS THEIR INSTITUTIONAL ENVIRONMENT ANDNATIONAL SYSTEMS OF INNOVATION

Variation in the innovative style of societies is shaped by their insti-tutional makeup ndash the various levels which are discussed aboveinstitutions institutional arrangements the structure and cohesion of institutional sectors which constitute a societyrsquos social system of pro-duction and the structure and culture of its organizations ndash especiallyits business rms and other research organizations By undertaking aninstitutional analysis of a society one can begin to understand in whatkinds of organizations the production of specic kinds of knowledgetakes place and how this is linked to particular kinds of innovativeness

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

626

With such a perspective one can gain a rich understanding of why somesocieties excel in the production of radical breakthroughs in basic andapplied science and in developing radically new products and why othersocieties excel in more incremental innovationsReecting on these issues one cannot help being impressed with the

contrasts between the innovative styles of German and American orga-nizations For example Germany has continued to be very successful in making incremental innovations in many industries in which it wasvery competitive well before World War II paper printing materialsmachinery electrotechnical products motor vehicles chemicals textileyarns and fabrics But the Germans have been much less successful indeveloping totally new products or related innovations in biotechnologyelectronics telecommunications aircraft as well as other newer indus-trial areas In contrast the Americans have been very innovative in these and other industries with products having very short lives andwith technologies which change very rapidly and are very complex(Hollingsworth 1991 Soskice 1997 1999)The importance of institutions and innovativeness for international

competitiveness has led to the recognition that there are national systemsof innovation (Lundvall 1992 Nelson 1993 Edquist 1997 Hage andHollingsworth 2000) But even before the recent studies of nationalsystems of innovation Landes (1969) was a proponent of these viewswhen he made the argument that Germany had developed a nationalsystem of innovation by building on its education and scientic researchsystems which were lacking in Britain and elsewhere However the concept of a national system of innovation immediately poses theproblem of what kind of innovations should be consideredDavid Soskice an economist at the Wissenschaftszentrum Berlin has

addressed a set of problems somewhat complementary to the perspec-tive raised here His work (Soskice 1990 1997 1999) suggests that weare sensitive to the institutional environment in which the followingactors are embedded the employees of rms the owners and nanciersof rms the competitors of rms and nally their collaborators ndash aperspective complementary to Hollingsworthrsquos work on social systemsof production (Hollingsworth 1997 1998) When one reects on the insti-tutional contextsocial system of production within which these actorsare embedded one becomes aware that those countries which have some-what inexible labor markets external to rms but a business system in which rms have long-term commitments to their employees arestrongly associated with incremental but hardly at all with radical inno-vations In the rms of such countries (eg Germany and Japan) thereis much more consensus decision making than in societies where a lsquohireand rersquo set of practices is quite pervasive And this kind of consensusdecision making ndash while highly conducive to incremental innovativeness

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

627

ndash limits the capacity for radical innovations to occur This kind of lsquolongtermismrsquo set of norms and rules occurs not only in the employmentpractices of incrementally innovative rms but also in their nancingand ownership patterns (Hollingsworth 1997) During the past thirtyyears these were characteristics of many rms in Germany and JapanSignicantly in both of these countries venture capital markets werenot highly developed nancing tended to be based more on long-termbank loans than on the equity markets and the relationships of rmstended to be relatively stable And these stable institutional arrange-ments also contributed to a high degree of innovativeness ndash butincremental rather than radical innovationsMeantime we should also pursue Soskicersquos ideas (1999) about how

rmsrsquo relationships with their competitors and collaborators shape theirstyles of innovativeness In those advanced industrial societies in whichbusiness rms are highly mobilized into business associations theretends to be a high degree of standardized quality controls over technicalnorms and a high degree of reliance on legal sanctions as monitoringinstruments vis-agrave-vis association members As a result of this high degreeof integration of rms into associations member rms are more likelyto be associated with incremental innovations than is likely to be the case with rms in those societies where associations are less welldeveloped In the latter type of society radical innovations are morelikely to occur In those advanced industrial societies where rms havelong-term stable relationships with collaborators in product design andproduction the style of innovation is likely to be incremental And it is in those societies in which there is a weaker tradition of strongcollaboration among rms that radical innovations are more likely tooccurObviously even in a country where there may be a great deal of

radical innovativeness most innovations will be of an incremental typefor there are simply fewer radical innovations But Germany and therest of Europe not only have had relatively few radical innovations incomparison with the United States but they also have lagged behindthe USA in developing relatively totally new industries Thus there hasbeen much less innovative activity in Europe in high-technology andnewly emerging industries than in the USAIn complementary research Audretsch and Feldman (1995) have

demonstrated that the USA has been more successful in spawning rmsengaged in the early stages of the industry life cycle but relative toGermany had an environment unfavorable to innovative activity byestablished rms especially in more mature industries German rmshave tended to be associated with a routinized technological regime ndashwhere established enterprises have had strong incentives to engage inincremental innovations but new rms have tended to have fundamental

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

628

innovative disadvantages relative to the Americans (see Audretsch 1995Winter 1982 Casper 2000)On both sides of the Atlantic universities engage in basic and applied

research But signicantly the institutional environment in which univer-sities are embedded inuences the degree to which the research is likelyto represent major breakthroughs or is likely to have more modestimplications Recent research in biomedical science demonstrates that inthose societies in which university professors andor graduate studentshave the following characteristics there are likely to be relatively majorresearch breakthroughs graduate students have a long training periodhigh dependency on a major professor low autonomy in dening thenature of a research project restricted job options and a university envi-ronment where there is a very modest or nonexistent entrepreneurialculture For example in Germany ndash a country with relatively few radicalbreakthroughs ndash the scholar who nishes a habilitation will have spentmany years on a research project before qualifying for a permanent acad-emic position will have been in a high-dependency position with a seniorprofessor will have had limited autonomy to dene a research projectand will be constrained from pursuing a high-risk research project Theyoung German will have an extraordinarily high level of training butwill have faced a number of disincentives to undertake high-risk researchAnd as soon as heshe has completed the habilitation and becomes aprofessor administrative responsibilities constitute another disincentivefrom engaging in high-risk or long-term research projects Moreover theGerman professor is a civil servant with a more stable and secure incomethan hisher American counterpart In a very different institutionalenvironment American professors in large research universities tend tobecome much more entrepreneurial Today many American professorsbehave as though they were operating their own rms within univer-sities and thus devote considerable time to generating funds for researchprojects assistants secretaries supplies etc And the variation in thisentrepreneurial spirit across countries is associated with variation in thetaking of high-risk research strategies and the making of major researchbreakthroughsIn sum societies vary in terms of the institutional environment in

which their organizations are embedded and this variation inuencesthe degree to which they have industrial activities in the early or latestages of the industry life cycle successful new or mature industriesand radical vs incremental innovations And it is this variation whichis encouraging rms in countries with relatively few new industries and with few or no radical innovations to invest in those countries with many new industries and numerous radical innovations ndash as astrategy to be able to transfer new technologies from the latter institu-tional environment to the former

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

629

Table 4 summarizes the way that the norms rules and values of a society are associated with Americarsquos and Germanyrsquos institutionalarrangements their social system of production and their styles of inno-vativeness

CONCLUSION

Several themes deserve to be reemphasized Most importantly institu-tional analysis is emerging as a dominant issue in several social sciencedisciplines The study of various institutional components discussedabove is a complex subject requiring multiple levels of analysis and aknowledge of multiple sectors within societies It requires a theoreticalperspective that evolves from an empirical analysis in different societiesIf we are to advance the study of institutional analysis we need cross-disciplinary collaboration and team research which are not readilyachieved given the structure of most of our research universitiesAt present scholars in each academic discipline have their own distinc-

tive strategies for studying institutional analysis Thus economists tendto use several approaches (North 1990 Williamson 1985 Hodgson 19881998 1999) anthropologists another (Geertz 1995) and sociologistspolitical scientists and historians employ other strategies (Campbell etal 1991 Powell and DiMaggio 1991 Katznelson 1998 Hall and Taylor1996) With practitioners in various disciplines pursuing their distinctiveapproaches to institutional analysis there is low potential for collabo-rative institutional analysis across the social sciences The lessons fromthe biological sciences over the last 40 years are very instructive oncepractitioners from various biological disciplines began to cooperate instudying biological phenomena by using the same concepts at the mole-cular level theoretical advance was very rapid Comparable phenomenatook place in the development of other hybrid elds eg biochemistrybiophysics (Judson 1979 Olby 1979 Kohler 1982 1994) Similarly weresocial scientists in various disciplines to work collaboratively in the areaof institutional analysis there would very likely be an acceleration intheir theoretical knowledgeThis is not a plea that everyone engaged in institutional analysis should

do the same kind of research Indeed just as biologists working at themolecular level study many different kinds of problems similarly socialscientists engaged in institutional analysis would also work on manyseparate problems but in the spirit of a collective enterprise Some wouldconcentrate their attention on the study of rules norms habits conven-tions and values while others would study how these are associatedwith congurations of various institutional arrangements (types ofmarkets hierarchies networks associations communities clans statesetc) Others would work on specic institutional sectors (eg education

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

630

Table 4

Rules norms hab

its an

d values associated w

ith sp

ecic co

ngu

ration

s of coo

rdinating mecha

nism

s soc

ial sy

stem

s of

prod

uction

an

d nationa

l styles of inno

vation

USA

Germany

Norms rules values which provide incentives for

Norms rules values which provide incentives for

short-term horizons

long-term horizons

Con

gu

rative

forms of

Marke

ts c

orpo

rate hierarchies reg

ulatory state

Assoc

iation

s (busine

ss assoc

iation

s la

bor union

s)

coordinating

corp

oratist-type

networks

marke

ts c

orporatist-

mecha

nism

s (listed in

type

state

order of im

portanc

e in

the particu

lar social

system

)

Social sys

tem of

prod

uction

Bus

iness system

Firm

struc

tures

Con

glom

erates b

ut with mov

emen

t toward

Firm

more co

ngru

ent with pa

rticular prod

ucts

elim

ination of produc

tion

in som

e sectors

emph

asis on long

-term strateg

ies rm

s are well

emph

asis on sh

ort-term

strateg

ies high

integrated

differentiation

between divisions

of rms

Firm

s mob

ilized in

Low

High

and in

tegrated

into busine

ss

associations

Indu

strial relations

Low

deg

ree of jo

b secu

rity rigid in

ternal lab

or

High deg

ree of jo

b secu

rity

exible interna

l lab

or

system

marke

t ex

ible externa

l labo

r marke

tsmarke

t rigid externa

l lab

or m

arke

tPatterns of

Dispersed owne

rship frequ

ent turnov

erMore stab

le owne

rship patterns

owne

rship

Table 4

continue

d

USA

Germany

Norms rules values which provide incentives for

Norms rules values which provide incentives for

short-term horizons

long-term horizons

Social system of

Prod

uction

Bus

iness system

Training for labo

r an

dHigh deg

ree of voluntary decision mak

ing by

Collective an

d m

ore co

mpulso

ry d

ecisions

abo

ut

man

agem

ent

individuals as to ho

w m

uch

tim

e an

d ene

rgy

how m

uch ed

ucation

is ne

eded

he

avy em

pha

sis

to inv

est in educ

ation heavy

emph

asis on

on techn

ical training (eg en

gine

ering)

busine

ss sch

ools training

in marke

ting

and

sales accou

nting

Fina

ncial marke

tsEqu

ity an

d ven

ture cap

ital m

arke

ts highly

Equ

ity an

d ven

ture cap

ital m

arke

ts w

eakly

dev

elop

ed na

ncing from

cen

tral gov

ernm

ent

deve

lope

d he

avy em

pha

sis on

corporate loan

sto priva

te rm

s for dev

elop

men

t of new

tech

nologies related

to na

tion

al security

Unive

rsity research

High individual auton

omy high poten

tial for

Low

ind

ividua

l au

tono

my until almost 40

yea

rs of

system

individual creativity u

nive

rsity pr

ofesso

rs

age poten

tial for ind

ividual creativity high

ly

high

ly entreprene

urial

cons

traine

d unive

rsity pr

ofessors not so

entrep

rene

urial

Inno

vative

styles

Highly su

ccessful in m

ajor break

through

s in

High pe

rforman

ce in proc

ess an

d inc

remen

tal

basic an

d app

lied scien

ce go

od in dev

elop

ing

inno

vation

s in existing produ

cts not so su

ccessful

totally

new

lines of pr

oducts an

d in m

ajor break

through

s in basic scien

ce or in

tech

nologies high

deg

ree of entrepr

eneu

rship

deve

loping

totally new

lines of prod

ucts of

amon

g research

ers

tech

nologies low d

egree of entrepr

eneu

rship

amon

g research

ers

Table 4

continue

d

USA

Germany

Norms rules values which provide incentives for

Norms rules values which provide incentives for

short-term horizons

long-term horizons

Prod

ucts in w

hich

Pr

oducts with sh

ort liv

es prod

ucts in w

hich

Older produ

cts with long

history machine

ry pap

er

society excels

tech

nology

is bo

th highly co

mplex

and

pr

inting

textile

s m

otor veh

icles m

achine

too

lsch

ange

s rapidly en

tertainm

ent indus

try

pha

rmaceu

ticals biotechn

olog

y softw

are

aerosp

ace

business systems nancial markets and systems of research) Distinctivewould be the recognition of how these separate studies are linkedtogether how each institutional sector is linked to a societyrsquos normsrules values etc and its conguration of institutional arrangementsSimilarly other institutional analysts might focus on the study of orga-nizations with a major concern for how the institutional environmentof organizations (eg norms rules values congurations of institutionalarrangements institutional sectors) inuenced the structure culture andoutcomes of organizations And nally others might study how all ofthe aforementioned aspects of institutional analysis inuenced a societyrsquosoverall performance Nelson and Sampat (1998) have appropriatelyreminded us that we should engage in institutional analysis in order toaddress specic problems but the view developed in this article is thatwe rst need some mapping of the terrain of institutional analysis inorder for our diverse enterprises to become a meaningful collective enter-prise especially if we are to understand how institutions inuence theinnovative processThe argument developed here is not that all social scientists should engage

in institutional analysis Obviously there are numerous other importantareas of research which lie beyond the eld of institutional analysis Nordoes this article argue that institutional analysis is the most importantsubject for social scientistsSince the structure and culture of our research universities reect an

enormous amount of disciplinary fragmentation what are the prospectsfor promoting the kind of collective and interdisciplinary research agendaproposed above In one respect the prospects are not encouragingDuring the past half century the efforts of American universities to have effective communication across the biological sciences have variedgreatly in their success At the University of California-Berkeley it was not until the 1980s that there was a serious effort to integrate thebiological sciences and at that time the campus abolished seventeendepartments to promote an integrated biological program But at theUniversity of Wisconsin the biological sciences are still very fragmentedinto dozens of different departments in botany plant pathology zoologyand physiology with two departments of biochemistry and two depart-ments of geneticsThere have been other approaches which have worked very well histor-

ically for promoting scientic integration amidst scientic diversity Forexample several summer schools and special institutes were organizedin the United States during the 1940s in an effort to overcome the frag-mentation in the biological sciences These programs made it possiblefor scientists in diverse academic disciplines and from different univer-sities and countries to come together to engage in intense and frequentinteraction Perhaps the most effective program was that held in the

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

634

summers at Cold Spring Harbor New York where a research agendafor the study of molecular biology was developed (Fischer and Lipson1988 Stent and Watson 1966) Scientists went there summer aftersummer and when they returned to their universities many began therestructuring of the biological sciences in their universitiesIn our own day we need summer institutes and other similar programs

so that scholars from diverse backgrounds can come together anddevelop a collective agenda for doing institutional analysis and forstudying the impact of institutions on the performance of societies ndashinnovativeness being one such subject If we can develop such instituteswe will increase the prospects for transforming our universities Thereis reason to be optimistic that once an eclectic group of scholars workwith a common set of concepts we can have genuine theoretical advanceif researchers can bring their diverse expertise to bear on a common setof problems There is considerable evidence that institutional analysisoffers high potential for advancing our knowledge of innovations andtechnological change (Hollingsworth and Hollingsworth 2000) If we aregoing to make substantial theoretical advance it is not likely to occurfrom the vantage point of a multiplicity of disciplines Rather it is likelyto occur as we develop a new intellectual framework and our own acad-emic journals This is a tall agenda which requires us to reassess theway that knowledge is presently produced in our societies

NOTE

This article is part of a much larger agenda involving the study of institutionsorganizations and innovations in which I am involved My debts are many andvaried The rst version was written while I was in residence in the ResearchUnit for Institutional Change and European Integration of the Austrian Academyof Sciences in Vienna in the summer of 1998 I am especially grateful to ProfessorsEgon Matzner and Sonja Puntscher-Riekmann for providing the stimulating envi-ronment in which to think through the ideas developed here Egon Matzner hasengaged in very stimulating discussions with me about the importance of insti-tutions in the general eld of socio-economics The nal version of the articlewas completed during my residence as a fellow at the Neurosciences Institutein La Jolla California I also wish to thank my colleagues in the InnovationTheme Project at the Netherlands Institute for Advanced Study (NIAS) during1998ndash9 and Edgar Grande and Karl Muumlller for stimulating conversations aboutthe interaction of institutions and organizations in facilitating particular typesof innovation I am especially grateful to Jerald Hage for helping me to developa number of issues in the article and for his useful comments on the rst draftFrans van Waarden early on helped me to work through the idea of multiplelevels of institutional analysis Others who have helped to develop the ideas inthe article have been my colleagues in the research project lsquoComparative socialsystems of productionrsquo especially Tom Burns Christel Lane Yoshitaka OkadaWolfgang Streeck and Richard Whitley My colleagues Robert Boyer SteveCasper and David Soskice have taught me much about how incentives providedby the institutional environments of rms inuence individual decision making

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

635

But my most important debt is to Ellen Jane Hollingsworth who has repeatedlyquestioned every single assumption of this article Without her rich and contin-uous tutoring this article would be even less developed than it is at presentFinally without the insightful assistance of David Gear in many phases of thisarticle it would not have been completed

REFERENCES

Aoki Masahiko (1990) lsquoTowards an economic model of the Japanese rmrsquo Journalof Economic Literature March 1ndash27

Arora A Landau Ralph and Rosenberg Nathan (eds) (1998) Chemicals and Long-Term Economic Growth Insights from the Chemical Industry New York JohnWiley amp Sons

Arrow Kenneth (1974) The Limits of Organization New York NortonArthur W B (1988a) lsquoSelf-reinforcing mechanisms in economicsrsquo in P W

Anderson K J Arrow and D Pines (eds) The Economy as an Evolving ComplexSystem Redwood City Calif Addison-Wesley

mdashmdash(1988b) lsquoCompeting technologies an overviewrsquo in G Dosi C Freeman RNelson G Silverberg and L Soete (eds) Technical Change and Economic TheoryLondon Pinter Publishers pp 590ndash607

Audretsch David B (1995) Innovation and Industry Evolution Cambridge MassMIT Press

Audretsch David B and Feldman Maryann P (1995) lsquoInnovative clusters andthe industry life cyclersquo Discussion Paper Wissenschaftszentrum BerlinMarket Processes and Corporate Development Research Unit FS-IV-95-7

Baum Joel A C and Oliver Christine (1992) lsquoInstitutional embeddedness andthe dynamics of organizational populationsrsquo American Sociological Review 57540ndash59

Baum Joel A C and Singh Jitendra V (1994a) lsquoOrganizationndashenvironment coevo-lutionrsquo in Joel A C Baum and Jitendra V Singh (eds) The EvolutionaryDynamics of Organizations New York Oxford University Press pp 379ndash402

mdashmdash(eds) (1994b) The Evolutionary Dynamics of Organizations New York OxfordUniversity Press

Berger Suzanne and Dore Ronald (eds) (1996) National Diversity and GlobalCapitalism Ithaca NY Cornell University Press

Boyer Robert (1997) lsquoThe variety and unequal performance of really existingmarkets farewell to Doctor Panglossrsquo in J Rogers Hollingsworth and RobertBoyer (eds) Contemporary Capitalism the Embeddedness of InstitutionsCambridge and New York Cambridge University Press pp 55ndash93

Burns Tom R and Dietz Thomas (2001 forthcoming) lsquoSocial rule system theorysocial action institutional arrangements and evolutionary processesrsquo in JRogers Hollingsworth and Karl Muumlller (eds) Advances in SocioeconomicsLanham Md Rowman amp Littleeld

Burns Tom R and Flam Helena (1987) The Shaping of Social Organization BeverlyHills Calif Sage

Calvert Randall L (1995) lsquoThe rational choice theory of social institutionsrsquo inJeffrey S Banks and Eric A Hanuskek (eds) Modern Political Economy NewYork Cambridge University Press pp 216ndash66

Camic Charles (1986) lsquoThe matter of habitrsquo American Journal of Sociology 911039ndash87

Campbell John L (1997) lsquoMechanisms of evolutionary change in economicsgovernance interaction interpretation and bricolagersquo in Lars Magnusson

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

636

some of the most common forms of associations for coordinatingeconomic activity in capitalist economies (Schneiberg and Hollingsworth1990 Streeck and Schmitter 1985a)Finally there is the state (cell 4) which is an institutional arrangement

quite unlike any of the others It is the state that sanctions and regu-lates the various non-state coordinating mechanisms that is the ultimateenforcer of rules of the various mechanisms that denes and enforcesproperty rights and that manipulates scal and monetary policy Overallit is the state that inuences the total incentive system of a society(Lindberg and Campbell 1991 Johnson 1992 40) At the same time thestate may also be an economic actor by engaging directly in productionand exchange relationsThe choices of institutional arrangements in Figure 2 are constrained

by the social context in which they are embedded Depending on thenature of that embeddedness there is variation in the collective formsof governance some of which are specied in the lower part of thegure Some modes of coordination in the bottom and upper levels ofthe typology are often mixed together though one coordinating modeis likely to be more prominent than another But actors are often engagedin complex congurations involving several kinds of institutionalarrangementsEach of these various institutional arrangements has its own logic ndash

its own rules its own procedures for enforcing compliance its ownnorms and ideologies which help to reduce the costs of enforcementThese are summarized in Table 2 which provides further elaborationabout the various coordinating mechanisms one nds in almost everycapitalist society Table 2 lists the organizational arrangements and theirstructures rules of exchange and means of enforcing compliance asso-ciated with each type of coordinating mechanism While each type ofinstitutional arrangement has various positive features each institutionalarrangement also has failures and these are featured in Table 3 It is thecontest between those who support and those who oppose these variousinstitutional arrangements that tends to lead to transformations in insti-tutional arrangements over time (Campbell et al 1991 Campbell 1997)Again it is difcult to conceive of pure institutional arrangements ndash

either exclusively markets or exclusively hierarchies ndash since it has beenwell known since Adam Smithrsquos Wealth of Nations that the division oflabor within the rm cannot be disentangled from the existence andextent of the market Institutional arrangements have their own distinc-tive form of efciency (good static efciency for the market dynamicefciency for rms) and inefciency often leading to considerableinequality Neither networks nor communities are panaceas for economiccoordination without being congured with other types of institutionalarrangements Networks and communities may solve certain issues but

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

609

Table 2

Log

ics of ins

titution

al arran

gemen

ts

Coordination

Organizational

Rules of

Individual means

Collective means

mechanisms

structure

exchange

of compliance

of compliance

Marke

tsEasy en

try an

d exit

Voluntary sp

ot excha

nge

Leg

al enforcemen

t of

Norm of pr

ivate

Bila

teral ex

chan

ge or

control

prope

rty

marke

t site (W

all Street)

Reg

ulations

to

Leg

itim

acy of m

arke

t en

force contracts

men

tality

Com

mun

ities

Inform

al m

embe

rship

Voluntary exch

ange

based

So

cial norms an

d m

oral

Highly institutiona

lized

ev

olve

s ov

er lon

g period

on social solid

arity an

d

princ

iples im

pose

norm

s an

d rules requ

ire

of tim

ehigh

deg

ree of tru

stob

ligations

mem

bers to accept

Kno

wledg

e of others an

d

lsquocorpo

ratersquoo

bligations

recipr

ocity ov

er tim

e

Networks

Semi-form

al m

embe

rship

Voluntary exch

ange

ove

r a

Con

tractual bon

dsPersona

l relation

sBila

teral or m

ultilateral

time pe

riod

Resou

rce de

pend

ence

Tru

st built outside

the

ex

chan

geecon

omic arena

Associatio

nsFo

rmal m

embe

rship

Restricted to mem

bers

Self-interest

Some deg

ree of

Multilateral excha

nge

Emph

asis on insider

Rep

utation

effects

compu

lsion

outsider or we

they

Priva

te interest type

of

men

tality

gove

rnan

ce

Private

Com

plex organ

izations

Restricted to mem

bers

Rew

ards to ind

ividuals

Highly institutiona

lized

hierarch

ies

which

ten

d to becom

e ex

chan

ge based

on

Asymmetric po

wer threat

rules

bureau

cratic hierarchies

asym

metric po

wer

of san

ctions

Mem

bers soc

ializ

ed into

bureau

cratic rules

corp

orate cu

lture us

e of

sanc

tions

State

Pub

lic hierarchy

Unilateral action

Exit vo

ice (vote lob

bying)

Coe

rcion

De jure

and impo

sed

Indirect political an

d

loya

ltyNorms an

d pu

blic rules

mem

bership

econ

omic excha

nge

Table 3

Failu

res of coo

rdinating mecha

nism

s

Coordinating mechanisms

Type of failure

Associations

Private hierarchies

State

Market

Communities

Networks

Enforcem

ent

Usu

ally relies on

Might enh

ance

Needs co

ntrol

Needs an

internal

Needs

tru

st and

Need an external

the state as an

oppo

rtun

istic

extern

al to state

enforcem

ent

loya

lty often

en

forcem

ent

enforcer

beha

vior

bureau

cracy

authority

coming from

au

thority

Resem

bles

The

ideal o

f (ju

dge

s

Facilitates

outside (fam

ily May

facilitate

enforcem

ent

internal m

arke

ts

parliamen

t collu

sion

and

relig

ion

cartelization an

d mecha

nism

might hur

t marke

t) to

impe

rfect

ethn

icity)

mon

opoly

of cartels

incu

mbe

nt

correct state

compe

titio

nCom

patible with

worke

rsab

uses

variou

s types of

Lob

bies can

co

mpe

tition

captur

e pu

blic

interest goa

ls

Public goo

d

Useful for

Gov

erna

nce co

sts

Can

provide

Can

not provide

Can

internalize

Useful for

and

establishing

might exceed

public goo

ds

colle

ctive go

ods

some co

llective

enha

ncing qu

ality

externality

stan

dards an

d

the be

nets of

but ha

s or deal w

ith

good

s (qua

lity

and training bu

t qu

ality for

internal division

difc

ultie

s in

externalities

training

) bu

t no

t no

t ve

ry goo

d in

setting ru

les of

of la

bor

prov

iding them

Inad

equate

othe

rs (welfare

prov

iding for

compe

titio

n in

Slow

to react to

in precise

mon

itoring of

gene

ral p

ublic

societal gen

eral

the indu

stry

chan

ges in the

am

ounts

tech

nical ch

ange

go

ods)

welfare

Useful for

environm

ent

Might fail in

and inno

vation

Mem

bers tightly

Weak in the

prov

iding

induc

ing

integrated

into

prov

ision of

man

y go

ods

tech

nical

community

colle

ctive go

als

colle

ctively

chan

geha

ve limite

d

that in

dividua

lcapa

city for

mem

bers can

not

inno

vatio

nsprov

ide for

them

selves

Table 3

continue

d

Coordinating mechanisms

Type of failure

Associations

Private hierarchies

State

Market

Communities

Networks

Efcien

cyFa

cilitates

De

cien

t in

Can

be high

ly

Some ba

sic social So

me go

ods

Slow

to en

hanc

e co

operation an

d

coop

eration an

d

bureau

cratic

relation

s cann

ot

cann

ot be

efc

ienc

y an

d

X-efcien

cy but

X-efcien

cyan

d can

not

be provided

by

delivered

at

speed of

not allocative

easily deliver

pure m

arke

t su

fcien

tly

adap

tive

ness

efcien

cy

good

s at

mecha

nism

slow costs

except in

low cost

indus

tries whe

re

tech

nology

is

complex an

d rapidly cha

nging

Equity

Narrow

Excessive

Might enh

ance

Facilitates

Might lead to

Whe

n widely

encompa

ssing

multiplication

ineq

uality

ineq

uality in

retard

ed

dev

elop

ed in

to

associationa

l of con

trollers

(pow

er

income an

d dev

elop

men

tindus

trial

stru

ctures lead

(fru

stration

and

an

d privilege

)wealth

districts

to in

come

ineq

uality)

netw

orks m

ay

equa

lity

facilitate greater

equality an

d

inco

me

distribution

whe

n wea

kly

dev

elop

ed

netw

orks ten

d to

increase soc

ial

ineq

uality

Note For allo

cativ

e efc

ienc

y an

d X

-efcien

cy see Leibe

nstein (19

66 19

76)

they raise other no less severe problems It is important to recognizethe imperfection of any single institutional arrangement in order tocomprehend the origin and transformation of any other institutionalarrangement (Hodgson 1999 ch 3)As suggested above each institutional arrangement is congured with

other institutional arrangements Usually one particular institutionalarrangement is more dominant in a particular conguration but becauseeach type of institutional arrangement has its own strengths andweaknesses there is no simple structural logic in the governance or coor-dination of a society Each institutional arrangement constrains the otherbut the inherent tension among the various institutional arrangementswithin a conguration contributes to changes in the governance orcoordination of a society The routines and logics of each of these cong-urations of institutional arrangements provide constraints and incentivesfor actors The way that actors perceive the incentives and constraintsof these governance congurations leads to a particular market logicand it is the specic market logic of a society which inuences its speciccapacities and weaknesses The inherent strain among the different logicsin a conguration of coordination helps to provide the exibility for asociety to adapt to new circumstancesIf there were a society with pure markets or a society coordinated

only by the state there would be too much rigidity and too little diver-sity to cope with the vast uncertainty in the global environment A societywith very little diversity in its coordination mechanisms would havelittle capacity to adapt to new circumstances Congurations with consid-erable diversity of institutional arrangements provide for a certainamount of incoherence in governance but they also provide for thecapacity to adapt to new circumstances While the Soviet Union wasdominated by the state as a coordinating mechanism there were alwaysfunctioning markets in the system Moreover the feudal society of thetwelfth and thirteenth centuries consisted not only of hierarchical rela-tionships between serfs and masters but also of urban guilds clericalhierarchies and markets (Johnson 1992 38) In sum the robustness ofinstitutions often depends on multiple and diverse principles and logicsof action on the inconsistency of principles and procedures on patternedforms of disorder (Lanzara 1998 Orren and Skowronek 1991 320 329)

Third level of analysis

The next level consists of the institutional sectors of a society The rulesnorms and values of a society inuence the array of institutional arrange-ments and levels one and two inuence the nature of and the relation-ships among various institutional sectors and all three levels areintricately linked together to form a social system of production Together

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

613

all of these components inuence the performance of economic sectorswithin a society as well as the performance of the total society A socialsystem of production is the way that a societyrsquos institutions (see rstlevel of analysis above) its institutional arrangements (see second levelof analysis above) and its institutional sectors are integrated into a socialcongurationAn institutional sector includes all organizations in a society which

supply a given service or product along with their associated focal orga-nizations (eg major suppliers funding sources regulators and so forth(Scott et al 1994 108 117 Campbell et al 1991 Hollingsworth et al1994) Institutional sectors include but are not limited to the societyrsquossystem of education system of research business system nancialmarkets legal system and the state The structure of the nancial marketssystem of training and education industrial relations system system ofresearch and state and legal systems are distinct and idiosyncratic ineach society In short they are system specicAll of these institutional sectors tend to cohere with each other

although they vary in the degree to which they are tightly coupled intoa full-edged system While each of the institutional sectors has someautonomy and may have some goals that are contradictory to the goalsof other institutional sectors with which it is linked an institutional logicin each society leads institutions to coalesce into a complex social con-guration (Hollingsworth 1991) This occurs because the institutionalsectors are embedded in a culture in which their logics are symbolicallygrounded organizationally structured technically and materially con-strained and politically defended The conguration of institutionalsectors usually exhibits some degree of adaptability to new challengesbut continues to evolve within an existing style But under new circum-stances or unprecedented disturbances these institutional congurationsare exposed to sharp historical limits as to what they may or may notdo (Schumpeter 1983 David 1988 Arthur 1988a 1988b Haringkanssen andLundgren 1997)Why do all of these different institutions coalesce into a complex

social conguration which is labeled here as a social system of pro-duction The literature suggests two contrasting interpretations Part ofthe answer ndash indeed a controversial one ndash is that these institutionalsectors are functionally determined by the requirements of the practiceof capitalism in each time and place (Habermas 1975) Another expla-nation emphasizes the genesis of the actual conguration via a trial and error process according to which the survival of rms regions orcountries is the outcome of complex evolutionary mechanisms (Maynard-Smith 1982 Nelson and Winter 1982) However the problem is evenmore complex Markets and other mechanisms for coordinating rela-tionships among economic actors place constraints on the means and

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

614

ends of economic activity to be achieved in any society The othercoordinating mechanisms include different kinds of hierarchies varioustypes of networks and associations ndash trade unions employers and busi-ness associations (Hollingsworth and Lindberg 1985 Campbell et al1991) The logic of these various institutional arrangements providesactors with vocabularies and logics for pursuing their goals for deningwhat is valued and for shaping the norms and rules by which theyabide (see Tables 2 and 3) In short in contrast to the logic of theneoclassical paradigm the argument here is that the dominant type ofinstitutional arrangements places severe constraints on the denition of needs preferences and choices of economic actors Whereas the neo-classical paradigm assumes that individuals and rms are sovereign thisarticle is based on the assumption that rms are inuenced by the hold that the institutional congurations making up a social system ofproduction have on individual decision making (Campbell et al 1991Etzioni 1988 Streeck and Schmitter 1985a Hollingsworth et al 1994Hollingsworth and Boyer 1997 Magnusson and Ottosson 1997 North1990 Hodgson 1999)Standard neoclassical economic theory has tended to downplay the

role of production and consequently of rms Even transaction cost theo-rists who are concerned with analyzing the rm as a coordinatingmechanism have been relatively unconcerned with the various compo-nents of a social system of production Indeed as long as there waswidespread optimism about the efcacy of Keynesian economics therewas relatively little concern among neoclassical economists with thesupply side of the economy Even in the opinion of most Keynesians agroup of experts should ideally be able to shape the size of aggregatedemand while the supply side of the economy would be left to the twominimalist institutions of neoclassical economics ndash markets and manage-rial hierarchies For several decades however it has become increasinglyobvious that some of the most competitive and successful patterns ofindustrial output and industrial production in capitalist economies donot derive from the neoclassical prescription of unregulated markets andcorporate hierarchies complemented by a neoliberal democratic stateIndeed it is now well understood that certain highly successful produc-tion patterns require for their emergence and survival institutionalarrangements and environments the very opposite of the prescriptionsfound in the neoclassical paradigm (see especially Streeck 1992 but alsoHollingsworth and Streeck 1994) Thus social scientists have come tounderstand that if they are to comprehend the behavior and performanceof contemporary economies concerns about social systems of produc-tion must be brought into the pictureBecause production involves more than technology a number of social

scientists have an increasing concern with social systems of production

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

615

The same equipment is frequently operated quite differently in the samesectors in different countries even when rms are competing in the same market (Maurice et al 1980 Sorge and Streeck 1988 Sorge 1989Hollingsworth et al 1994) Variations in production and process tech-nologies are inuenced partly by variations in the social environmentsin which they are embedded In other words rms are embedded intocomplex environments which among other things place constraints ontheir behavior Thus a social system of production is of major impor-tance in understanding the behavior and performance of an economyHow the state and other institutional arrangements (eg marketsnetworks private hierarchies associations) coalesce and are related toparticular social systems of production are important determinants ofeconomic performanceThe congurations of institutional arrangements that coordinate or

govern the behavior of actors in one society and its structure of specicinstitutional sectors cannot easily be transferred to another society forthey are embedded into a social system of production that is societallydistinct (Hollingsworth 1997) Societies borrow selected principles offoreign management styles and work practices but the effectiveness ofsuch borrowing is generally limited Economic behavior and performanceare shaped by the entire social system of production in which actors areembedded and not simply by specic principles of particular manage-ment styles and work practices Moreover a societyrsquos social system ofproduction tends to limit the kind of goods that it can produce and withwhich it can compete successfully in international marketsSome scholars (Kenney and Florida 1993 Oliver and Wilkinson 1988)

have assumed that the diffusion of particular forms of management stylesand work practices across societies could lead to system convergenceReferring to the Japanese production system Kenney and Florida haveargued that it consists of organizational practices whose fundamentallsquogenetic codersquo can successfully be inserted into another society and beginsuccessfully to reproduce its behavior in the new environment (1993 8)Their position is in the intellectual tradition of Antonio Gramsci (1971)who contended decades ago that the American system of mass produc-tion would most certainly diffuse to Europe over timeHowever the argument of this article is that even though British

French and American rms may adopt certain aspects of Japanesemanagement styles (eg just-in-time production self-managing teamsquality circles the use of statistical process controls) or some variant ofthe German vocational system their social systems of production willnot converge Moreover the overall conguration of a societyrsquos socialsystem of production inuences its sectoral and overall national economicperformance This helps to explain why societies have different systemsof innovation and why they vary in the clusters of industries in which

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

616

they are highly competitive in the international markets (Nelson 19931999 Hage and Hollingsworth 2000 Berger and Dore 1996)In order to understand how and why a societyrsquos economy performs

as it does it is necessary to understand its entire social system of produc-tion If a society is to modify substantially the performance of itsinstitutional sectors and its entire economy it cannot adapt only someof the management and work practices of its foreign competitors Ratherit must alter its entire social system of production Because a societyrsquosmodes of economic governance and coordination and its institutionalsectors develop according to particular logics and are system specicthere are serious limitations to the extent to which a society may mimicthe institutions (eg the rules and norms) institutional arrangements andinstitutional sectors of other societiesIn the history of modern societies there are logics by which institu-

tions coalesce into a social system of production (Hollingsworth 1991)Though institutions are constantly changing there are sharp limits tothe type and the direction of change that any particular institution canundergo because of its linkages with institutional arrangements and insti-tutional sectors Thus a societyrsquos business rms educational systemnancial markets industrial relations system etc can engage in seriousrestructuring only if most of the other institutional sectors also changeIn social systems of production there are pressures toward consistencyin the norms rules and values across institutional sectors though in anycomplex society social systems are obviously imperfectly integratedIndeed the degree to which the institutional norms and rules makingup a social system of production are loosely or tightly coupled is a vari-able of considerable importance In general the institutional sectorsmaking up a social system of production are interdependent and changesin one generally result in changes in the others Since each institutionallevel is dependent on the others for various types of resources there isinterdependence among the differing institutional spheres Moreovereach society has its norms moral principles rules and laws and recipesfor action as well as its own idiosyncratic customs traditions and prin-ciples of justice (Burns and Flam 1987)There are also other inherent obstacles to convergence among social

systems of production for where a system is at any one point in timeis inuenced by its initial state Systems with quite different initial statesare unlikely to converge with one anotherrsquos institutional practicesExisting institutional arrangements block certain institutional innova-tions and facilitate others (Roland 1990) Thus the institutions makingup a social system of production provide continuity even though insti-tutional arrangements are always changing but with a particular logicWhile Williamson (1975 1985) suggests that actors tend to choose theinstitutional arrangements which are most efcient North (1990) is

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

617

much closer to the mark in his argument that most societal institutionalarrangements exist as a result of custom and habit and are thereforeinefcient At any moment in time the world often appears to its actorsas very complex and uncertain For this reason actors often engage incontradictory forms of behavior pursuing different strategies as hedgesagainst a very uncertain world (Lanzara 1998) And their hedging andcontradictory forms of behavior may lead in somewhat different soci-etal directions all constrained by the institutional fabric within whichthe actors are embedded This kind of contradictory behavior occurs inpart because of the contradictions inherent in the dominant congura-tions of institutional arrangements with their contradictory logics (seelsquoSecond level of analysisrsquo above)Despite the emphasis on the logic of institutional continuity this is

not an argument that systems change along some predetermined pathThere are critical turning points in the history of highly industrializedsocieties but with the choices limited by the existing institutional terrainBeing path-dependent a social system of production continues along aparticular logic until or unless a fundamental societal crisis intervenes(Milgrom et al 1991 Krugman 1991 Durlauf 1991 Hollingsworth 1991Pred 1966 David 1988 Hodgson 1998)At this point it is important to confront the question of how social

systems of production evolve And this question gets to the heart of theproblem of building complex societal structures Certainly social systemsof production did not emerge from some process of social engineeringMoreover the various component parts of each social system of produc-tion have often not been designed to be part of a social system ofproduction The component parts of each social system of productionhave emerged more often than not as unintended by-products of goalswhich various actors had in mind at earlier moments in time It is usuallythe case that actors at time lsquotrsquo fail to comprehend the long-term conse-quences of their actions Wolfgang Streeck the author of several brilliantpapers on social systems of production argues that institutional sectordesigns created for one purpose generally address that goal but overtime those designs have quite unintended consequences (Streeck 1997a1997b Wright 1998) Describing the conguration of the German socialsystem of production of the 1970s and 1980s Streeck points out that itwas the unintended by-product of multiple points in history Someelements were pre-Wilhelmian others were introduced by the Alliedpowers after 1945 and others emerged during the years of the GermanFederal Republic All component parts of the German social system ofproduction

were and continue to be changing for their own reasons as wellas in reaction to each other and certainly there can be no presump-tion of a pre-established t between them even though one might

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

618

want to allow for some reinforcement effects of the modelrsquos histor-ically contingent social and economic success

(Streeck 1997b 54)

In sum the emergence of social systems of production is a long-termevolutionary process with each part interacting with its environment andresulting in a congurative whole but with no previous design by eithera single actor or a collectivity of actors As suggested above actors do not recreate their world anew At any moment in history the totalinstitutional complex of a society necessarily contains the resources and legacies of its past Preexisting institutional complexes are nevercompletely wiped out Societies consist of multiple layers of history withtheir diverse logics of action However the cumulative effect of smallperipheral changes in altering particular institutional sectors and insti-tutional arrangements can be substantial (Murmann 1998 Lanzara 1998)lsquoInstitution building is affected more by the ways in which people havecodied the past than by how they envision the future There are manymore potential resources in pre-existing institutional arrangements thanit is usually assumed or suspectedrsquo (Lanzara 1998 30)

Fourth level of analysis

The next level of institutional analysis organizational structures is some-what more controversial As suggested above Douglass North (1990)draws a sharp distinction between institutions and organizations Morerecently many organizational theorists (Powell 1991 Powell andDiMaggio 1991 DiMaggio and Powell 1983 Baum and Oliver 1992Kondra and Hinings 1998 Townley 1997) argue that institutional rulesnorms and conventions unfold in tandem with organizational structuresand this is my positionThe literature which focuses on how institutions inuence organiza-

tions is quite different from two other theoretical literatures which alsoare concerned with how organizational environments shape the behaviorof organizations There are resource dependency theorists (Pfeffer andSalancik 1978) who emphasize the role of environmental resources inshaping organizations while Hannan and Freeman (1977 1984) and otherpopulation ecologists emphasize the survival of organizations givencertain kinds of organizational conditions (Orru et al 1991)The institutionalist perspective as implied above emphasizes the

normative environment in which organizations are embedded It is a perspective which focuses on the way in which organizations in their behavior tend to conform to the institutional rules and normswhich are dominant in the organizational environment However allthree perspectives ndash the institutionalist the population ecology and

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

619

the resource dependency ndash emphasize how the environment inuencesorganizations and how organizations which are subject to the sameenvironment tend to converge in their behavior to have what DiMaggioand Powell have labeled organizational isomorphism (DiMaggio andPowell 1983)There has emerged a vast literature which demonstrates that within

each society there are scal political judicial and other regulatory normswhich limit and shape the culture and structure of organizationalbehavior The normative institutional environment of organizations limits the options of what organizations do in a particular society and inuences the patterns of ownership relations with suppliers andcustomers In short it is the normative environment of organizationswhich denes within a particular society what is socially acceptablebehavior for organizations (Hamilton and Biggart 1988 Hollingsworthand Hollingsworth 2000 Meyer and Rowan 1991 Zucker 1987 1988Orru et al 1991 Townley 1997)Thus far most of the analyses on isomorphic pressures operating on

organizations have been cross-sectional in nature However the entireconcept of isomorphism implies that there are strong environmental pres-sures exerted on organizations and in order to observe this phenomenonit is important that we have longitudinal studies which assess howchanges in an organizationrsquos institutional environment inuence changesin the structure and culture of organizations Moreover the expectationhere is that when one analyzes the patterns of organizational changeover time the isomorphism will be holistic ndash that is within a particularsociety there will be a tendency for the internal structure of particulartypes of organizations to converge in large part because of the pres-sures to conform to the changes in the external norms rules and valuesA close reading of Aoki (1990) and Slack and Hinings (1994) nds thatchanges in the national institutional environment of organizations haveinuenced changes in the internal structure of organizations within aspecic industry of a particular societyHowever it is inappropriate to view organizations as changing only

in relation to exogenous environmental change There is a kind of co-evolution which occurs between organizations and their institutionalenvironments Co-evolution implies nonlinear feedback between orga-nizations and their institutional environments This is of course quitedifferent from the more common way of studying organizations bymodeling relations between independent and dependent variables Ofcourse the more common type of analysis is quite appropriate whenthere are simple relationships which do not involve complex feedbackprocesses But in the strategy for institutional analysis which is proposedherein it is less useful to separate independent from dependent vari-ables and more useful to understand the interacting and co-evolutionary

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

620

processes among all of the levels of institutional analysis discussed inthe various sections above (Baum and Singh 1994a 379ndash80 Mowery andNelson 1999 Murmann 1998)The logic of the perspective herein suggests that within every

society there is variation in the structure and culture of business rmsuniversities and other complex organizations (Kondra and Hinings1998) However this variation takes place within parameters which aresystem specic For example every German rm and every Germanuniversity is different from every other German university and rm but there are a set of cultural and structural characteristics which distin-guish German rms and universities from those in America Similarlyeach Japanese rm and university is unique yet they have a set ofcultural and structural characteristics which are system specic andwhich differentiate them from business rms and universities in other countriesIn all societies each organization has its own distinctive organizational

rules norms and conventions which are subordinate to the meta norms and rules of the larger society within which they are embeddedHowever the strength of the institutional environment within whichorganizations are embedded varies from society to society Some societies have multiple institutional environments and there is hetero-geneity in terms of what constitutes appropriate behavior oforganizations In such societies there can be conict over whatinstitutional logics should regulate specic organizational functions(Friedland and Alford 1991 Townley 1997 262ndash3) In those societies inwhich the institutional norms habits and rules are most developed and in which the institutional pressures to conform are greatest thereis less variation in the structure and culture of business rms and variouskinds of research organizations In such societies the connectednessbetween research organizations and their institutional environments is sufciently strong for organizations to have low autonomy to pursue independent strategies and goals in these societies there is a great deal of organizational isomorphism Conversely the weaker the institutional environment in which research organizations areembedded the greater the variation in the structure and culture of business rms and research organizations Moreover where the institu-tional environments are more weakly developed organizations havegreater autonomy and exibility to respond to the development of new knowledge and to be highly innovative Hence it is in those societies where the institutional environments are most developed and most rigid and when there is less organizational autonomy and exibility that fewer radical innovations in basic and applied science as well as in totally new products and industrial sectors havebeen made

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

621

Fifth level of analysis

We come next to the outputs and performance of the various institutionalcomponents of a society It is at this level that institutional componentsare more pragmatic and exible For example in the legal sector thereare specic statutes and court rulings in the state sector there are specicpolicies in the business sector there are new products new technolo-gies and market strategies It is at this nal level of analysis thatinstitutional spheres are most open and susceptible to change and cross-national mimicry is easiest and most commonThrough the outputs of the society we can obtain some assessment as

to how well a society is performing Moreover we can assess how inno-vative it is how egalitarian it is in the distribution of its resources howegalitarian it is in terms of levels of health education etc Just as it iscomplicated to measure how egalitarian a society is similarly it is dif-cult to assess how innovative it is For example societies vary in makingincremental and radical breakthroughs in basic and applied science in developing totally new products and new kinds of organizations inincremental and process innovations in existing products and organiza-tions in developing new and different forms of marketing ndash bothdomestically and globally (Hage and Hollingsworth 2000) Of coursesocieties vary in their rates of economic growth in their rates of economicproductivity in their quality of life (eg rates of crime life expectancyetc) All of these performance criteria feed back and inuence each ofthe levels of institutional analysis discussed above ndash rules norms valuesetc institutional arrangements institutional sectors the structure andform of organizations Moreover performance measures may inuenceeach other For example if a society has very low rates of growth itmay not be very innovative in some of the types of innovative activitiesmentioned above Good or poor performances of certain kinds inuenceother performance indicators (Hollingsworth et al 1990)Different institutional arrangements and different social systems of

production result in different types of economic performance Hence solong as societies have different social systems of production there areserious constraints on the degree to which they can converge in theirinnovative styles Different social systems of production tend to maxi-mize in a more or less explicit manner different performance criteriausually mixing considerations about static and dynamic efciency protsecurity social peace and economic andor political power In short incontrast to the implications of neoclassical economic theory in real-worldeconomies there are no universal standards all economically rationalactors attempt to maximize Economic history provides numerous exam-ples of how a great variety of principles of rationality are implementedin different societies

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

622

A critical question in institutional analysis is whether an existing socialsystem of production which supports a set of routines for a particularkind of technology and industry can shift from old practices to newones There are numerous historical accounts of how the social systemof production in a particular society which worked so well for a numberof years could not adapt to new technologies (Lazonick 1990 Veblen1915 Nelson 1994) Schumpeter (1983) and Freeman (1991) have devel-oped the idea of a techno-economic paradigm which suggests thatdifferent technologies require specic forms of organizational arrange-ments ndash or what I called social systems of production Because oflsquoinstitutional inertiarsquo the social system of production of a society maynot be able to adapt to a new techno-economic paradigm (Nelson 199458 Lanzara 1998)Whether a social system of production can sustain its performance

standards depends not only on its intrinsic economic rationality but alsoon where it ts into a larger system If a particular social system ofproduction is immune from the innovativeness and competition of analternative system survival can be long lasting But if different socialsystems of production with diverging criteria of good economic perfor-mance meet in the world arena the arbitrariness of nationally imposedconstructed performance standards may be superseded by alternativeperformance criteria as a result of international competitivenessThe world economy is also socially constructed just as are national

economies Even if different social systems of production are competingin the international arena it is not always possible to determine whichis more competitively effective at any moment in time Hegemonicnation-states in the short run shape the rules of trade to favor their indus-trial sectors and rms But the history of hegemonic powers suggeststhat in the longer run such social systems of production sustainedlargely by military and political power eventually give way to moreinnovative and competitive social systems of production In our ownday as nation-states are increasingly integrated into a world economyeven hegemonic powers lose their innovative and competitive advan-tage and their share of world output decreases Such a country mayattempt to restructure its institutional arrangements and to readjust itsperformance preferences But to restructure its system of innovationgenerally calls for a major redistribution of power within a societyLargely for this reason societies have historically had limited capacityto reconstruct a system of innovation in the image of their major competi-tors Each countryrsquos social system of production is a conguration of ahost of norms rules and values as well as of institutional arrangementsEach system is constantly changing and is open to inuence from othersystems And indeed many technologies and practices diffuse from onesociety to another but a societyrsquos capacity to be innovative is constrained

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

623

by the existing social system of production Thus the same technologymay exist in numerous countries but how it is employed and how itinuences societal outcomes and performances varies from one institu-tional conguration or society to another (Hollingsworth 1998)

PERSPECTIVES ON INSTITUTIONAL CHANGE

A fundamental problem which remains unresolved in institutionalanalysis involves the nature of institutional change There is much confu-sion and miscommunication in the social sciences about institutionalchange in large part because we do not even know how to measure therate of institutional change Of course we know that there are bothexternal and internal focuses for change and we have had a good bitof scholarship on this subject What is not clearly understood is thatinstitutions and institutional arrangements within societies are histori-cally rooted that there is a great deal of path-dependency to the waythat institutions evolve Moreover the more intricately linked each sectoris with each other sector and with a societyrsquos rules and norms the lesschoice actors have to devise new institutions and institutional arrange-ments On the other hand if a sector or actor is too isolated it may betoo weak within a society to bring about any kind of effective changeThese considerations are very critical to the problem of how muchfreedom actors have to create new institutions afresh to what extentinstitutions diffuse from one society to another (Murmann 1998 Lanzara1998) This is a fundamental question for students of institutionalanalysis as we attempt to understand how the different institutionallevels are linked togetherThere are a variety of reasons why there is confusion in studying insti-

tutional change There is disagreement over the extent to which actorshave the freedom to build new institutional arrangements and if theydo to what extent the new arrangements and organizational patternsmay depart from past practices Among some institutional economists(Williamson 1985) there is a basic assumption that actors build efcientinstitutions by pursuing their self-interest and by promoting theirstrategic goals Moreover there is the assumption that a mixing of fullrationality with market competition tends to produce all of the optimalinstitutions that are needed in order to coordinate a complex capitalisteconomy (Hollingsworth and Boyer 1997 ch 14 Hodgson 1999) Incontrast the argument developed here is much more complexSocial institutions are historically rooted and there is a great deal of

path-dependency in the way that various institutional componentsevolve The shape of institutional congurations at any moment limitsthe type of options for change Because there is a great deal of institu-tional inertia radical change in the institutional components of a society

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

624

is uncommon (Lanzara 1998) Recognizing the constraints existinginstitutions exert on actors we note that existing congurations of insti-tutional arrangements institutional sectors and organizational propertiescan limit but also inuence the degree to which norms and rules canchange Many scholars (North 1990 Ostrom 1990) have emphasized therole of norms rules ideologies and values in limiting the ability of actorsto develop new institutional arrangements which radically depart fromexisting ones However there is a two-way process of interaction aprocess of co-evolution (Baum and Singh 1994a Nelson 1994) Thus asFriedland and Alford (1991 244) argue actors develop new institutionalarrangements by recognizing organizational failures and low perfor-mance but these new arrangements do not necessarily change a societyrsquosnorms rules habits and its underlying value systemCampbell (1997) advances our understanding of how existing institu-

tions enable actors to construct radically new institutions He recognizesthat most institutions and institutional arrangements embody a degreeof inertia from which it is difcult for actors to depart Moreover thedifferential power relations among actors make it unlikely that lesspowerful actors can change the existing power structure of a societyMost change in the institutional components of a society evolves througha process of constrained selection which reects to a considerable degreethe existing arrangements and power relations This becomes the inter-pretative frame for social actors with institutionalized scripts and ritualswhich tend to be taken for granted and appear to be quite ordinary andnatural This institutionalized scripting of the social world is an extremelyimportant source of social stability and inertiaCampbell has developed a very fruitful explanation of change in terms

of the various institutional components outlined here He distinguishesbetween incremental and radical institutional change and argues thatradical institutional change occurs when social actors with widelydiffering norms cultural scripts and rituals for action engage in intenseinteraction with each other (see also Knight 1992) The more diverse theinteractions the greater the potential for institutional change He arguesthat as shifts in the composition of interaction among social actors occurchanges in the interpretation of problems and interests will follow Ifthere are only minor changes in the extent to which diverse actorsinteract there will be only modest institutional change But the morefundamental the changes in the interaction of diverse social actors thegreater the change in the way that actors interpret their world And thegreater the change in the way that actors interpret their world the morelikely that radical changes in various institutional components willfollow In other words diverse decision making tends to create a widerrange of interpretative frameworks of the social world than is likely tobe the case if social actors operate in isolation

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

625

This perspective of diverse interactions among social actors facilitatesour ability to move beyond the traditional view that existing institutionsconstrain the range of institutional alternatives which actors face Theolder view has tended to suggest that actors engage in institutionalchange by extending existing institutional principles habits and conven-tions Specifying the conditions which facilitate fundamental institutionalchange permits a much richer perspective for institutional analysisCampbellrsquos perspective about the conditions under which fundamentalor radical change in institutional components occurs is consistent withmy own research about the conditions under which major discoveriesor fundamental new knowledge occur the more scholars with diversebackgrounds interact with intensity and frequency the greater the like-lihood that they will develop new and alternative ways of thinking(Hollingsworth and Hollingsworth 2000)These views of institutional change are also consistent with the recent

work of Nelson and his colleagues (Mowery and Nelson 1999 Murmann1998) They report the research on a variety of industrial sectors in anumber of different countries revealing that networks of actors in partic-ular industries business and professional associations universityresearch departments and institutes are frequently able to use their collec-tive power to modify a societyrsquos property rights research system andgovernmental policies It is the intense and frequent interaction of thesediverse complex sets of actors ndash sometimes constituting a technologysystem ndash that often leads to incremental changes in a societyrsquos total socialsystem of production It is this kind of collective action that brings aboutorganizational change change in institutional sectors institutionalarrangements and even in the rules and norms of a society And if weare to understand how the style of innovativeness of a society changeswe must rst understand the changes in its institutional makeup It ishighly unlikely that a society will develop a new style of innovativenesswithout changing its institutional structure

FIRMS THEIR INSTITUTIONAL ENVIRONMENT ANDNATIONAL SYSTEMS OF INNOVATION

Variation in the innovative style of societies is shaped by their insti-tutional makeup ndash the various levels which are discussed aboveinstitutions institutional arrangements the structure and cohesion of institutional sectors which constitute a societyrsquos social system of pro-duction and the structure and culture of its organizations ndash especiallyits business rms and other research organizations By undertaking aninstitutional analysis of a society one can begin to understand in whatkinds of organizations the production of specic kinds of knowledgetakes place and how this is linked to particular kinds of innovativeness

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

626

With such a perspective one can gain a rich understanding of why somesocieties excel in the production of radical breakthroughs in basic andapplied science and in developing radically new products and why othersocieties excel in more incremental innovationsReecting on these issues one cannot help being impressed with the

contrasts between the innovative styles of German and American orga-nizations For example Germany has continued to be very successful in making incremental innovations in many industries in which it wasvery competitive well before World War II paper printing materialsmachinery electrotechnical products motor vehicles chemicals textileyarns and fabrics But the Germans have been much less successful indeveloping totally new products or related innovations in biotechnologyelectronics telecommunications aircraft as well as other newer indus-trial areas In contrast the Americans have been very innovative in these and other industries with products having very short lives andwith technologies which change very rapidly and are very complex(Hollingsworth 1991 Soskice 1997 1999)The importance of institutions and innovativeness for international

competitiveness has led to the recognition that there are national systemsof innovation (Lundvall 1992 Nelson 1993 Edquist 1997 Hage andHollingsworth 2000) But even before the recent studies of nationalsystems of innovation Landes (1969) was a proponent of these viewswhen he made the argument that Germany had developed a nationalsystem of innovation by building on its education and scientic researchsystems which were lacking in Britain and elsewhere However the concept of a national system of innovation immediately poses theproblem of what kind of innovations should be consideredDavid Soskice an economist at the Wissenschaftszentrum Berlin has

addressed a set of problems somewhat complementary to the perspec-tive raised here His work (Soskice 1990 1997 1999) suggests that weare sensitive to the institutional environment in which the followingactors are embedded the employees of rms the owners and nanciersof rms the competitors of rms and nally their collaborators ndash aperspective complementary to Hollingsworthrsquos work on social systemsof production (Hollingsworth 1997 1998) When one reects on the insti-tutional contextsocial system of production within which these actorsare embedded one becomes aware that those countries which have some-what inexible labor markets external to rms but a business system in which rms have long-term commitments to their employees arestrongly associated with incremental but hardly at all with radical inno-vations In the rms of such countries (eg Germany and Japan) thereis much more consensus decision making than in societies where a lsquohireand rersquo set of practices is quite pervasive And this kind of consensusdecision making ndash while highly conducive to incremental innovativeness

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

627

ndash limits the capacity for radical innovations to occur This kind of lsquolongtermismrsquo set of norms and rules occurs not only in the employmentpractices of incrementally innovative rms but also in their nancingand ownership patterns (Hollingsworth 1997) During the past thirtyyears these were characteristics of many rms in Germany and JapanSignicantly in both of these countries venture capital markets werenot highly developed nancing tended to be based more on long-termbank loans than on the equity markets and the relationships of rmstended to be relatively stable And these stable institutional arrange-ments also contributed to a high degree of innovativeness ndash butincremental rather than radical innovationsMeantime we should also pursue Soskicersquos ideas (1999) about how

rmsrsquo relationships with their competitors and collaborators shape theirstyles of innovativeness In those advanced industrial societies in whichbusiness rms are highly mobilized into business associations theretends to be a high degree of standardized quality controls over technicalnorms and a high degree of reliance on legal sanctions as monitoringinstruments vis-agrave-vis association members As a result of this high degreeof integration of rms into associations member rms are more likelyto be associated with incremental innovations than is likely to be the case with rms in those societies where associations are less welldeveloped In the latter type of society radical innovations are morelikely to occur In those advanced industrial societies where rms havelong-term stable relationships with collaborators in product design andproduction the style of innovation is likely to be incremental And it is in those societies in which there is a weaker tradition of strongcollaboration among rms that radical innovations are more likely tooccurObviously even in a country where there may be a great deal of

radical innovativeness most innovations will be of an incremental typefor there are simply fewer radical innovations But Germany and therest of Europe not only have had relatively few radical innovations incomparison with the United States but they also have lagged behindthe USA in developing relatively totally new industries Thus there hasbeen much less innovative activity in Europe in high-technology andnewly emerging industries than in the USAIn complementary research Audretsch and Feldman (1995) have

demonstrated that the USA has been more successful in spawning rmsengaged in the early stages of the industry life cycle but relative toGermany had an environment unfavorable to innovative activity byestablished rms especially in more mature industries German rmshave tended to be associated with a routinized technological regime ndashwhere established enterprises have had strong incentives to engage inincremental innovations but new rms have tended to have fundamental

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

628

innovative disadvantages relative to the Americans (see Audretsch 1995Winter 1982 Casper 2000)On both sides of the Atlantic universities engage in basic and applied

research But signicantly the institutional environment in which univer-sities are embedded inuences the degree to which the research is likelyto represent major breakthroughs or is likely to have more modestimplications Recent research in biomedical science demonstrates that inthose societies in which university professors andor graduate studentshave the following characteristics there are likely to be relatively majorresearch breakthroughs graduate students have a long training periodhigh dependency on a major professor low autonomy in dening thenature of a research project restricted job options and a university envi-ronment where there is a very modest or nonexistent entrepreneurialculture For example in Germany ndash a country with relatively few radicalbreakthroughs ndash the scholar who nishes a habilitation will have spentmany years on a research project before qualifying for a permanent acad-emic position will have been in a high-dependency position with a seniorprofessor will have had limited autonomy to dene a research projectand will be constrained from pursuing a high-risk research project Theyoung German will have an extraordinarily high level of training butwill have faced a number of disincentives to undertake high-risk researchAnd as soon as heshe has completed the habilitation and becomes aprofessor administrative responsibilities constitute another disincentivefrom engaging in high-risk or long-term research projects Moreover theGerman professor is a civil servant with a more stable and secure incomethan hisher American counterpart In a very different institutionalenvironment American professors in large research universities tend tobecome much more entrepreneurial Today many American professorsbehave as though they were operating their own rms within univer-sities and thus devote considerable time to generating funds for researchprojects assistants secretaries supplies etc And the variation in thisentrepreneurial spirit across countries is associated with variation in thetaking of high-risk research strategies and the making of major researchbreakthroughsIn sum societies vary in terms of the institutional environment in

which their organizations are embedded and this variation inuencesthe degree to which they have industrial activities in the early or latestages of the industry life cycle successful new or mature industriesand radical vs incremental innovations And it is this variation whichis encouraging rms in countries with relatively few new industries and with few or no radical innovations to invest in those countries with many new industries and numerous radical innovations ndash as astrategy to be able to transfer new technologies from the latter institu-tional environment to the former

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

629

Table 4 summarizes the way that the norms rules and values of a society are associated with Americarsquos and Germanyrsquos institutionalarrangements their social system of production and their styles of inno-vativeness

CONCLUSION

Several themes deserve to be reemphasized Most importantly institu-tional analysis is emerging as a dominant issue in several social sciencedisciplines The study of various institutional components discussedabove is a complex subject requiring multiple levels of analysis and aknowledge of multiple sectors within societies It requires a theoreticalperspective that evolves from an empirical analysis in different societiesIf we are to advance the study of institutional analysis we need cross-disciplinary collaboration and team research which are not readilyachieved given the structure of most of our research universitiesAt present scholars in each academic discipline have their own distinc-

tive strategies for studying institutional analysis Thus economists tendto use several approaches (North 1990 Williamson 1985 Hodgson 19881998 1999) anthropologists another (Geertz 1995) and sociologistspolitical scientists and historians employ other strategies (Campbell etal 1991 Powell and DiMaggio 1991 Katznelson 1998 Hall and Taylor1996) With practitioners in various disciplines pursuing their distinctiveapproaches to institutional analysis there is low potential for collabo-rative institutional analysis across the social sciences The lessons fromthe biological sciences over the last 40 years are very instructive oncepractitioners from various biological disciplines began to cooperate instudying biological phenomena by using the same concepts at the mole-cular level theoretical advance was very rapid Comparable phenomenatook place in the development of other hybrid elds eg biochemistrybiophysics (Judson 1979 Olby 1979 Kohler 1982 1994) Similarly weresocial scientists in various disciplines to work collaboratively in the areaof institutional analysis there would very likely be an acceleration intheir theoretical knowledgeThis is not a plea that everyone engaged in institutional analysis should

do the same kind of research Indeed just as biologists working at themolecular level study many different kinds of problems similarly socialscientists engaged in institutional analysis would also work on manyseparate problems but in the spirit of a collective enterprise Some wouldconcentrate their attention on the study of rules norms habits conven-tions and values while others would study how these are associatedwith congurations of various institutional arrangements (types ofmarkets hierarchies networks associations communities clans statesetc) Others would work on specic institutional sectors (eg education

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

630

Table 4

Rules norms hab

its an

d values associated w

ith sp

ecic co

ngu

ration

s of coo

rdinating mecha

nism

s soc

ial sy

stem

s of

prod

uction

an

d nationa

l styles of inno

vation

USA

Germany

Norms rules values which provide incentives for

Norms rules values which provide incentives for

short-term horizons

long-term horizons

Con

gu

rative

forms of

Marke

ts c

orpo

rate hierarchies reg

ulatory state

Assoc

iation

s (busine

ss assoc

iation

s la

bor union

s)

coordinating

corp

oratist-type

networks

marke

ts c

orporatist-

mecha

nism

s (listed in

type

state

order of im

portanc

e in

the particu

lar social

system

)

Social sys

tem of

prod

uction

Bus

iness system

Firm

struc

tures

Con

glom

erates b

ut with mov

emen

t toward

Firm

more co

ngru

ent with pa

rticular prod

ucts

elim

ination of produc

tion

in som

e sectors

emph

asis on long

-term strateg

ies rm

s are well

emph

asis on sh

ort-term

strateg

ies high

integrated

differentiation

between divisions

of rms

Firm

s mob

ilized in

Low

High

and in

tegrated

into busine

ss

associations

Indu

strial relations

Low

deg

ree of jo

b secu

rity rigid in

ternal lab

or

High deg

ree of jo

b secu

rity

exible interna

l lab

or

system

marke

t ex

ible externa

l labo

r marke

tsmarke

t rigid externa

l lab

or m

arke

tPatterns of

Dispersed owne

rship frequ

ent turnov

erMore stab

le owne

rship patterns

owne

rship

Table 4

continue

d

USA

Germany

Norms rules values which provide incentives for

Norms rules values which provide incentives for

short-term horizons

long-term horizons

Social system of

Prod

uction

Bus

iness system

Training for labo

r an

dHigh deg

ree of voluntary decision mak

ing by

Collective an

d m

ore co

mpulso

ry d

ecisions

abo

ut

man

agem

ent

individuals as to ho

w m

uch

tim

e an

d ene

rgy

how m

uch ed

ucation

is ne

eded

he

avy em

pha

sis

to inv

est in educ

ation heavy

emph

asis on

on techn

ical training (eg en

gine

ering)

busine

ss sch

ools training

in marke

ting

and

sales accou

nting

Fina

ncial marke

tsEqu

ity an

d ven

ture cap

ital m

arke

ts highly

Equ

ity an

d ven

ture cap

ital m

arke

ts w

eakly

dev

elop

ed na

ncing from

cen

tral gov

ernm

ent

deve

lope

d he

avy em

pha

sis on

corporate loan

sto priva

te rm

s for dev

elop

men

t of new

tech

nologies related

to na

tion

al security

Unive

rsity research

High individual auton

omy high poten

tial for

Low

ind

ividua

l au

tono

my until almost 40

yea

rs of

system

individual creativity u

nive

rsity pr

ofesso

rs

age poten

tial for ind

ividual creativity high

ly

high

ly entreprene

urial

cons

traine

d unive

rsity pr

ofessors not so

entrep

rene

urial

Inno

vative

styles

Highly su

ccessful in m

ajor break

through

s in

High pe

rforman

ce in proc

ess an

d inc

remen

tal

basic an

d app

lied scien

ce go

od in dev

elop

ing

inno

vation

s in existing produ

cts not so su

ccessful

totally

new

lines of pr

oducts an

d in m

ajor break

through

s in basic scien

ce or in

tech

nologies high

deg

ree of entrepr

eneu

rship

deve

loping

totally new

lines of prod

ucts of

amon

g research

ers

tech

nologies low d

egree of entrepr

eneu

rship

amon

g research

ers

Table 4

continue

d

USA

Germany

Norms rules values which provide incentives for

Norms rules values which provide incentives for

short-term horizons

long-term horizons

Prod

ucts in w

hich

Pr

oducts with sh

ort liv

es prod

ucts in w

hich

Older produ

cts with long

history machine

ry pap

er

society excels

tech

nology

is bo

th highly co

mplex

and

pr

inting

textile

s m

otor veh

icles m

achine

too

lsch

ange

s rapidly en

tertainm

ent indus

try

pha

rmaceu

ticals biotechn

olog

y softw

are

aerosp

ace

business systems nancial markets and systems of research) Distinctivewould be the recognition of how these separate studies are linkedtogether how each institutional sector is linked to a societyrsquos normsrules values etc and its conguration of institutional arrangementsSimilarly other institutional analysts might focus on the study of orga-nizations with a major concern for how the institutional environmentof organizations (eg norms rules values congurations of institutionalarrangements institutional sectors) inuenced the structure culture andoutcomes of organizations And nally others might study how all ofthe aforementioned aspects of institutional analysis inuenced a societyrsquosoverall performance Nelson and Sampat (1998) have appropriatelyreminded us that we should engage in institutional analysis in order toaddress specic problems but the view developed in this article is thatwe rst need some mapping of the terrain of institutional analysis inorder for our diverse enterprises to become a meaningful collective enter-prise especially if we are to understand how institutions inuence theinnovative processThe argument developed here is not that all social scientists should engage

in institutional analysis Obviously there are numerous other importantareas of research which lie beyond the eld of institutional analysis Nordoes this article argue that institutional analysis is the most importantsubject for social scientistsSince the structure and culture of our research universities reect an

enormous amount of disciplinary fragmentation what are the prospectsfor promoting the kind of collective and interdisciplinary research agendaproposed above In one respect the prospects are not encouragingDuring the past half century the efforts of American universities to have effective communication across the biological sciences have variedgreatly in their success At the University of California-Berkeley it was not until the 1980s that there was a serious effort to integrate thebiological sciences and at that time the campus abolished seventeendepartments to promote an integrated biological program But at theUniversity of Wisconsin the biological sciences are still very fragmentedinto dozens of different departments in botany plant pathology zoologyand physiology with two departments of biochemistry and two depart-ments of geneticsThere have been other approaches which have worked very well histor-

ically for promoting scientic integration amidst scientic diversity Forexample several summer schools and special institutes were organizedin the United States during the 1940s in an effort to overcome the frag-mentation in the biological sciences These programs made it possiblefor scientists in diverse academic disciplines and from different univer-sities and countries to come together to engage in intense and frequentinteraction Perhaps the most effective program was that held in the

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

634

summers at Cold Spring Harbor New York where a research agendafor the study of molecular biology was developed (Fischer and Lipson1988 Stent and Watson 1966) Scientists went there summer aftersummer and when they returned to their universities many began therestructuring of the biological sciences in their universitiesIn our own day we need summer institutes and other similar programs

so that scholars from diverse backgrounds can come together anddevelop a collective agenda for doing institutional analysis and forstudying the impact of institutions on the performance of societies ndashinnovativeness being one such subject If we can develop such instituteswe will increase the prospects for transforming our universities Thereis reason to be optimistic that once an eclectic group of scholars workwith a common set of concepts we can have genuine theoretical advanceif researchers can bring their diverse expertise to bear on a common setof problems There is considerable evidence that institutional analysisoffers high potential for advancing our knowledge of innovations andtechnological change (Hollingsworth and Hollingsworth 2000) If we aregoing to make substantial theoretical advance it is not likely to occurfrom the vantage point of a multiplicity of disciplines Rather it is likelyto occur as we develop a new intellectual framework and our own acad-emic journals This is a tall agenda which requires us to reassess theway that knowledge is presently produced in our societies

NOTE

This article is part of a much larger agenda involving the study of institutionsorganizations and innovations in which I am involved My debts are many andvaried The rst version was written while I was in residence in the ResearchUnit for Institutional Change and European Integration of the Austrian Academyof Sciences in Vienna in the summer of 1998 I am especially grateful to ProfessorsEgon Matzner and Sonja Puntscher-Riekmann for providing the stimulating envi-ronment in which to think through the ideas developed here Egon Matzner hasengaged in very stimulating discussions with me about the importance of insti-tutions in the general eld of socio-economics The nal version of the articlewas completed during my residence as a fellow at the Neurosciences Institutein La Jolla California I also wish to thank my colleagues in the InnovationTheme Project at the Netherlands Institute for Advanced Study (NIAS) during1998ndash9 and Edgar Grande and Karl Muumlller for stimulating conversations aboutthe interaction of institutions and organizations in facilitating particular typesof innovation I am especially grateful to Jerald Hage for helping me to developa number of issues in the article and for his useful comments on the rst draftFrans van Waarden early on helped me to work through the idea of multiplelevels of institutional analysis Others who have helped to develop the ideas inthe article have been my colleagues in the research project lsquoComparative socialsystems of productionrsquo especially Tom Burns Christel Lane Yoshitaka OkadaWolfgang Streeck and Richard Whitley My colleagues Robert Boyer SteveCasper and David Soskice have taught me much about how incentives providedby the institutional environments of rms inuence individual decision making

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

635

But my most important debt is to Ellen Jane Hollingsworth who has repeatedlyquestioned every single assumption of this article Without her rich and contin-uous tutoring this article would be even less developed than it is at presentFinally without the insightful assistance of David Gear in many phases of thisarticle it would not have been completed

REFERENCES

Aoki Masahiko (1990) lsquoTowards an economic model of the Japanese rmrsquo Journalof Economic Literature March 1ndash27

Arora A Landau Ralph and Rosenberg Nathan (eds) (1998) Chemicals and Long-Term Economic Growth Insights from the Chemical Industry New York JohnWiley amp Sons

Arrow Kenneth (1974) The Limits of Organization New York NortonArthur W B (1988a) lsquoSelf-reinforcing mechanisms in economicsrsquo in P W

Anderson K J Arrow and D Pines (eds) The Economy as an Evolving ComplexSystem Redwood City Calif Addison-Wesley

mdashmdash(1988b) lsquoCompeting technologies an overviewrsquo in G Dosi C Freeman RNelson G Silverberg and L Soete (eds) Technical Change and Economic TheoryLondon Pinter Publishers pp 590ndash607

Audretsch David B (1995) Innovation and Industry Evolution Cambridge MassMIT Press

Audretsch David B and Feldman Maryann P (1995) lsquoInnovative clusters andthe industry life cyclersquo Discussion Paper Wissenschaftszentrum BerlinMarket Processes and Corporate Development Research Unit FS-IV-95-7

Baum Joel A C and Oliver Christine (1992) lsquoInstitutional embeddedness andthe dynamics of organizational populationsrsquo American Sociological Review 57540ndash59

Baum Joel A C and Singh Jitendra V (1994a) lsquoOrganizationndashenvironment coevo-lutionrsquo in Joel A C Baum and Jitendra V Singh (eds) The EvolutionaryDynamics of Organizations New York Oxford University Press pp 379ndash402

mdashmdash(eds) (1994b) The Evolutionary Dynamics of Organizations New York OxfordUniversity Press

Berger Suzanne and Dore Ronald (eds) (1996) National Diversity and GlobalCapitalism Ithaca NY Cornell University Press

Boyer Robert (1997) lsquoThe variety and unequal performance of really existingmarkets farewell to Doctor Panglossrsquo in J Rogers Hollingsworth and RobertBoyer (eds) Contemporary Capitalism the Embeddedness of InstitutionsCambridge and New York Cambridge University Press pp 55ndash93

Burns Tom R and Dietz Thomas (2001 forthcoming) lsquoSocial rule system theorysocial action institutional arrangements and evolutionary processesrsquo in JRogers Hollingsworth and Karl Muumlller (eds) Advances in SocioeconomicsLanham Md Rowman amp Littleeld

Burns Tom R and Flam Helena (1987) The Shaping of Social Organization BeverlyHills Calif Sage

Calvert Randall L (1995) lsquoThe rational choice theory of social institutionsrsquo inJeffrey S Banks and Eric A Hanuskek (eds) Modern Political Economy NewYork Cambridge University Press pp 216ndash66

Camic Charles (1986) lsquoThe matter of habitrsquo American Journal of Sociology 911039ndash87

Campbell John L (1997) lsquoMechanisms of evolutionary change in economicsgovernance interaction interpretation and bricolagersquo in Lars Magnusson

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

636

Table 2

Log

ics of ins

titution

al arran

gemen

ts

Coordination

Organizational

Rules of

Individual means

Collective means

mechanisms

structure

exchange

of compliance

of compliance

Marke

tsEasy en

try an

d exit

Voluntary sp

ot excha

nge

Leg

al enforcemen

t of

Norm of pr

ivate

Bila

teral ex

chan

ge or

control

prope

rty

marke

t site (W

all Street)

Reg

ulations

to

Leg

itim

acy of m

arke

t en

force contracts

men

tality

Com

mun

ities

Inform

al m

embe

rship

Voluntary exch

ange

based

So

cial norms an

d m

oral

Highly institutiona

lized

ev

olve

s ov

er lon

g period

on social solid

arity an

d

princ

iples im

pose

norm

s an

d rules requ

ire

of tim

ehigh

deg

ree of tru

stob

ligations

mem

bers to accept

Kno

wledg

e of others an

d

lsquocorpo

ratersquoo

bligations

recipr

ocity ov

er tim

e

Networks

Semi-form

al m

embe

rship

Voluntary exch

ange

ove

r a

Con

tractual bon

dsPersona

l relation

sBila

teral or m

ultilateral

time pe

riod

Resou

rce de

pend

ence

Tru

st built outside

the

ex

chan

geecon

omic arena

Associatio

nsFo

rmal m

embe

rship

Restricted to mem

bers

Self-interest

Some deg

ree of

Multilateral excha

nge

Emph

asis on insider

Rep

utation

effects

compu

lsion

outsider or we

they

Priva

te interest type

of

men

tality

gove

rnan

ce

Private

Com

plex organ

izations

Restricted to mem

bers

Rew

ards to ind

ividuals

Highly institutiona

lized

hierarch

ies

which

ten

d to becom

e ex

chan

ge based

on

Asymmetric po

wer threat

rules

bureau

cratic hierarchies

asym

metric po

wer

of san

ctions

Mem

bers soc

ializ

ed into

bureau

cratic rules

corp

orate cu

lture us

e of

sanc

tions

State

Pub

lic hierarchy

Unilateral action

Exit vo

ice (vote lob

bying)

Coe

rcion

De jure

and impo

sed

Indirect political an

d

loya

ltyNorms an

d pu

blic rules

mem

bership

econ

omic excha

nge

Table 3

Failu

res of coo

rdinating mecha

nism

s

Coordinating mechanisms

Type of failure

Associations

Private hierarchies

State

Market

Communities

Networks

Enforcem

ent

Usu

ally relies on

Might enh

ance

Needs co

ntrol

Needs an

internal

Needs

tru

st and

Need an external

the state as an

oppo

rtun

istic

extern

al to state

enforcem

ent

loya

lty often

en

forcem

ent

enforcer

beha

vior

bureau

cracy

authority

coming from

au

thority

Resem

bles

The

ideal o

f (ju

dge

s

Facilitates

outside (fam

ily May

facilitate

enforcem

ent

internal m

arke

ts

parliamen

t collu

sion

and

relig

ion

cartelization an

d mecha

nism

might hur

t marke

t) to

impe

rfect

ethn

icity)

mon

opoly

of cartels

incu

mbe

nt

correct state

compe

titio

nCom

patible with

worke

rsab

uses

variou

s types of

Lob

bies can

co

mpe

tition

captur

e pu

blic

interest goa

ls

Public goo

d

Useful for

Gov

erna

nce co

sts

Can

provide

Can

not provide

Can

internalize

Useful for

and

establishing

might exceed

public goo

ds

colle

ctive go

ods

some co

llective

enha

ncing qu

ality

externality

stan

dards an

d

the be

nets of

but ha

s or deal w

ith

good

s (qua

lity

and training bu

t qu

ality for

internal division

difc

ultie

s in

externalities

training

) bu

t no

t no

t ve

ry goo

d in

setting ru

les of

of la

bor

prov

iding them

Inad

equate

othe

rs (welfare

prov

iding for

compe

titio

n in

Slow

to react to

in precise

mon

itoring of

gene

ral p

ublic

societal gen

eral

the indu

stry

chan

ges in the

am

ounts

tech

nical ch

ange

go

ods)

welfare

Useful for

environm

ent

Might fail in

and inno

vation

Mem

bers tightly

Weak in the

prov

iding

induc

ing

integrated

into

prov

ision of

man

y go

ods

tech

nical

community

colle

ctive go

als

colle

ctively

chan

geha

ve limite

d

that in

dividua

lcapa

city for

mem

bers can

not

inno

vatio

nsprov

ide for

them

selves

Table 3

continue

d

Coordinating mechanisms

Type of failure

Associations

Private hierarchies

State

Market

Communities

Networks

Efcien

cyFa

cilitates

De

cien

t in

Can

be high

ly

Some ba

sic social So

me go

ods

Slow

to en

hanc

e co

operation an

d

coop

eration an

d

bureau

cratic

relation

s cann

ot

cann

ot be

efc

ienc

y an

d

X-efcien

cy but

X-efcien

cyan

d can

not

be provided

by

delivered

at

speed of

not allocative

easily deliver

pure m

arke

t su

fcien

tly

adap

tive

ness

efcien

cy

good

s at

mecha

nism

slow costs

except in

low cost

indus

tries whe

re

tech

nology

is

complex an

d rapidly cha

nging

Equity

Narrow

Excessive

Might enh

ance

Facilitates

Might lead to

Whe

n widely

encompa

ssing

multiplication

ineq

uality

ineq

uality in

retard

ed

dev

elop

ed in

to

associationa

l of con

trollers

(pow

er

income an

d dev

elop

men

tindus

trial

stru

ctures lead

(fru

stration

and

an

d privilege

)wealth

districts

to in

come

ineq

uality)

netw

orks m

ay

equa

lity

facilitate greater

equality an

d

inco

me

distribution

whe

n wea

kly

dev

elop

ed

netw

orks ten

d to

increase soc

ial

ineq

uality

Note For allo

cativ

e efc

ienc

y an

d X

-efcien

cy see Leibe

nstein (19

66 19

76)

they raise other no less severe problems It is important to recognizethe imperfection of any single institutional arrangement in order tocomprehend the origin and transformation of any other institutionalarrangement (Hodgson 1999 ch 3)As suggested above each institutional arrangement is congured with

other institutional arrangements Usually one particular institutionalarrangement is more dominant in a particular conguration but becauseeach type of institutional arrangement has its own strengths andweaknesses there is no simple structural logic in the governance or coor-dination of a society Each institutional arrangement constrains the otherbut the inherent tension among the various institutional arrangementswithin a conguration contributes to changes in the governance orcoordination of a society The routines and logics of each of these cong-urations of institutional arrangements provide constraints and incentivesfor actors The way that actors perceive the incentives and constraintsof these governance congurations leads to a particular market logicand it is the specic market logic of a society which inuences its speciccapacities and weaknesses The inherent strain among the different logicsin a conguration of coordination helps to provide the exibility for asociety to adapt to new circumstancesIf there were a society with pure markets or a society coordinated

only by the state there would be too much rigidity and too little diver-sity to cope with the vast uncertainty in the global environment A societywith very little diversity in its coordination mechanisms would havelittle capacity to adapt to new circumstances Congurations with consid-erable diversity of institutional arrangements provide for a certainamount of incoherence in governance but they also provide for thecapacity to adapt to new circumstances While the Soviet Union wasdominated by the state as a coordinating mechanism there were alwaysfunctioning markets in the system Moreover the feudal society of thetwelfth and thirteenth centuries consisted not only of hierarchical rela-tionships between serfs and masters but also of urban guilds clericalhierarchies and markets (Johnson 1992 38) In sum the robustness ofinstitutions often depends on multiple and diverse principles and logicsof action on the inconsistency of principles and procedures on patternedforms of disorder (Lanzara 1998 Orren and Skowronek 1991 320 329)

Third level of analysis

The next level consists of the institutional sectors of a society The rulesnorms and values of a society inuence the array of institutional arrange-ments and levels one and two inuence the nature of and the relation-ships among various institutional sectors and all three levels areintricately linked together to form a social system of production Together

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

613

all of these components inuence the performance of economic sectorswithin a society as well as the performance of the total society A socialsystem of production is the way that a societyrsquos institutions (see rstlevel of analysis above) its institutional arrangements (see second levelof analysis above) and its institutional sectors are integrated into a socialcongurationAn institutional sector includes all organizations in a society which

supply a given service or product along with their associated focal orga-nizations (eg major suppliers funding sources regulators and so forth(Scott et al 1994 108 117 Campbell et al 1991 Hollingsworth et al1994) Institutional sectors include but are not limited to the societyrsquossystem of education system of research business system nancialmarkets legal system and the state The structure of the nancial marketssystem of training and education industrial relations system system ofresearch and state and legal systems are distinct and idiosyncratic ineach society In short they are system specicAll of these institutional sectors tend to cohere with each other

although they vary in the degree to which they are tightly coupled intoa full-edged system While each of the institutional sectors has someautonomy and may have some goals that are contradictory to the goalsof other institutional sectors with which it is linked an institutional logicin each society leads institutions to coalesce into a complex social con-guration (Hollingsworth 1991) This occurs because the institutionalsectors are embedded in a culture in which their logics are symbolicallygrounded organizationally structured technically and materially con-strained and politically defended The conguration of institutionalsectors usually exhibits some degree of adaptability to new challengesbut continues to evolve within an existing style But under new circum-stances or unprecedented disturbances these institutional congurationsare exposed to sharp historical limits as to what they may or may notdo (Schumpeter 1983 David 1988 Arthur 1988a 1988b Haringkanssen andLundgren 1997)Why do all of these different institutions coalesce into a complex

social conguration which is labeled here as a social system of pro-duction The literature suggests two contrasting interpretations Part ofthe answer ndash indeed a controversial one ndash is that these institutionalsectors are functionally determined by the requirements of the practiceof capitalism in each time and place (Habermas 1975) Another expla-nation emphasizes the genesis of the actual conguration via a trial and error process according to which the survival of rms regions orcountries is the outcome of complex evolutionary mechanisms (Maynard-Smith 1982 Nelson and Winter 1982) However the problem is evenmore complex Markets and other mechanisms for coordinating rela-tionships among economic actors place constraints on the means and

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

614

ends of economic activity to be achieved in any society The othercoordinating mechanisms include different kinds of hierarchies varioustypes of networks and associations ndash trade unions employers and busi-ness associations (Hollingsworth and Lindberg 1985 Campbell et al1991) The logic of these various institutional arrangements providesactors with vocabularies and logics for pursuing their goals for deningwhat is valued and for shaping the norms and rules by which theyabide (see Tables 2 and 3) In short in contrast to the logic of theneoclassical paradigm the argument here is that the dominant type ofinstitutional arrangements places severe constraints on the denition of needs preferences and choices of economic actors Whereas the neo-classical paradigm assumes that individuals and rms are sovereign thisarticle is based on the assumption that rms are inuenced by the hold that the institutional congurations making up a social system ofproduction have on individual decision making (Campbell et al 1991Etzioni 1988 Streeck and Schmitter 1985a Hollingsworth et al 1994Hollingsworth and Boyer 1997 Magnusson and Ottosson 1997 North1990 Hodgson 1999)Standard neoclassical economic theory has tended to downplay the

role of production and consequently of rms Even transaction cost theo-rists who are concerned with analyzing the rm as a coordinatingmechanism have been relatively unconcerned with the various compo-nents of a social system of production Indeed as long as there waswidespread optimism about the efcacy of Keynesian economics therewas relatively little concern among neoclassical economists with thesupply side of the economy Even in the opinion of most Keynesians agroup of experts should ideally be able to shape the size of aggregatedemand while the supply side of the economy would be left to the twominimalist institutions of neoclassical economics ndash markets and manage-rial hierarchies For several decades however it has become increasinglyobvious that some of the most competitive and successful patterns ofindustrial output and industrial production in capitalist economies donot derive from the neoclassical prescription of unregulated markets andcorporate hierarchies complemented by a neoliberal democratic stateIndeed it is now well understood that certain highly successful produc-tion patterns require for their emergence and survival institutionalarrangements and environments the very opposite of the prescriptionsfound in the neoclassical paradigm (see especially Streeck 1992 but alsoHollingsworth and Streeck 1994) Thus social scientists have come tounderstand that if they are to comprehend the behavior and performanceof contemporary economies concerns about social systems of produc-tion must be brought into the pictureBecause production involves more than technology a number of social

scientists have an increasing concern with social systems of production

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

615

The same equipment is frequently operated quite differently in the samesectors in different countries even when rms are competing in the same market (Maurice et al 1980 Sorge and Streeck 1988 Sorge 1989Hollingsworth et al 1994) Variations in production and process tech-nologies are inuenced partly by variations in the social environmentsin which they are embedded In other words rms are embedded intocomplex environments which among other things place constraints ontheir behavior Thus a social system of production is of major impor-tance in understanding the behavior and performance of an economyHow the state and other institutional arrangements (eg marketsnetworks private hierarchies associations) coalesce and are related toparticular social systems of production are important determinants ofeconomic performanceThe congurations of institutional arrangements that coordinate or

govern the behavior of actors in one society and its structure of specicinstitutional sectors cannot easily be transferred to another society forthey are embedded into a social system of production that is societallydistinct (Hollingsworth 1997) Societies borrow selected principles offoreign management styles and work practices but the effectiveness ofsuch borrowing is generally limited Economic behavior and performanceare shaped by the entire social system of production in which actors areembedded and not simply by specic principles of particular manage-ment styles and work practices Moreover a societyrsquos social system ofproduction tends to limit the kind of goods that it can produce and withwhich it can compete successfully in international marketsSome scholars (Kenney and Florida 1993 Oliver and Wilkinson 1988)

have assumed that the diffusion of particular forms of management stylesand work practices across societies could lead to system convergenceReferring to the Japanese production system Kenney and Florida haveargued that it consists of organizational practices whose fundamentallsquogenetic codersquo can successfully be inserted into another society and beginsuccessfully to reproduce its behavior in the new environment (1993 8)Their position is in the intellectual tradition of Antonio Gramsci (1971)who contended decades ago that the American system of mass produc-tion would most certainly diffuse to Europe over timeHowever the argument of this article is that even though British

French and American rms may adopt certain aspects of Japanesemanagement styles (eg just-in-time production self-managing teamsquality circles the use of statistical process controls) or some variant ofthe German vocational system their social systems of production willnot converge Moreover the overall conguration of a societyrsquos socialsystem of production inuences its sectoral and overall national economicperformance This helps to explain why societies have different systemsof innovation and why they vary in the clusters of industries in which

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

616

they are highly competitive in the international markets (Nelson 19931999 Hage and Hollingsworth 2000 Berger and Dore 1996)In order to understand how and why a societyrsquos economy performs

as it does it is necessary to understand its entire social system of produc-tion If a society is to modify substantially the performance of itsinstitutional sectors and its entire economy it cannot adapt only someof the management and work practices of its foreign competitors Ratherit must alter its entire social system of production Because a societyrsquosmodes of economic governance and coordination and its institutionalsectors develop according to particular logics and are system specicthere are serious limitations to the extent to which a society may mimicthe institutions (eg the rules and norms) institutional arrangements andinstitutional sectors of other societiesIn the history of modern societies there are logics by which institu-

tions coalesce into a social system of production (Hollingsworth 1991)Though institutions are constantly changing there are sharp limits tothe type and the direction of change that any particular institution canundergo because of its linkages with institutional arrangements and insti-tutional sectors Thus a societyrsquos business rms educational systemnancial markets industrial relations system etc can engage in seriousrestructuring only if most of the other institutional sectors also changeIn social systems of production there are pressures toward consistencyin the norms rules and values across institutional sectors though in anycomplex society social systems are obviously imperfectly integratedIndeed the degree to which the institutional norms and rules makingup a social system of production are loosely or tightly coupled is a vari-able of considerable importance In general the institutional sectorsmaking up a social system of production are interdependent and changesin one generally result in changes in the others Since each institutionallevel is dependent on the others for various types of resources there isinterdependence among the differing institutional spheres Moreovereach society has its norms moral principles rules and laws and recipesfor action as well as its own idiosyncratic customs traditions and prin-ciples of justice (Burns and Flam 1987)There are also other inherent obstacles to convergence among social

systems of production for where a system is at any one point in timeis inuenced by its initial state Systems with quite different initial statesare unlikely to converge with one anotherrsquos institutional practicesExisting institutional arrangements block certain institutional innova-tions and facilitate others (Roland 1990) Thus the institutions makingup a social system of production provide continuity even though insti-tutional arrangements are always changing but with a particular logicWhile Williamson (1975 1985) suggests that actors tend to choose theinstitutional arrangements which are most efcient North (1990) is

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

617

much closer to the mark in his argument that most societal institutionalarrangements exist as a result of custom and habit and are thereforeinefcient At any moment in time the world often appears to its actorsas very complex and uncertain For this reason actors often engage incontradictory forms of behavior pursuing different strategies as hedgesagainst a very uncertain world (Lanzara 1998) And their hedging andcontradictory forms of behavior may lead in somewhat different soci-etal directions all constrained by the institutional fabric within whichthe actors are embedded This kind of contradictory behavior occurs inpart because of the contradictions inherent in the dominant congura-tions of institutional arrangements with their contradictory logics (seelsquoSecond level of analysisrsquo above)Despite the emphasis on the logic of institutional continuity this is

not an argument that systems change along some predetermined pathThere are critical turning points in the history of highly industrializedsocieties but with the choices limited by the existing institutional terrainBeing path-dependent a social system of production continues along aparticular logic until or unless a fundamental societal crisis intervenes(Milgrom et al 1991 Krugman 1991 Durlauf 1991 Hollingsworth 1991Pred 1966 David 1988 Hodgson 1998)At this point it is important to confront the question of how social

systems of production evolve And this question gets to the heart of theproblem of building complex societal structures Certainly social systemsof production did not emerge from some process of social engineeringMoreover the various component parts of each social system of produc-tion have often not been designed to be part of a social system ofproduction The component parts of each social system of productionhave emerged more often than not as unintended by-products of goalswhich various actors had in mind at earlier moments in time It is usuallythe case that actors at time lsquotrsquo fail to comprehend the long-term conse-quences of their actions Wolfgang Streeck the author of several brilliantpapers on social systems of production argues that institutional sectordesigns created for one purpose generally address that goal but overtime those designs have quite unintended consequences (Streeck 1997a1997b Wright 1998) Describing the conguration of the German socialsystem of production of the 1970s and 1980s Streeck points out that itwas the unintended by-product of multiple points in history Someelements were pre-Wilhelmian others were introduced by the Alliedpowers after 1945 and others emerged during the years of the GermanFederal Republic All component parts of the German social system ofproduction

were and continue to be changing for their own reasons as wellas in reaction to each other and certainly there can be no presump-tion of a pre-established t between them even though one might

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

618

want to allow for some reinforcement effects of the modelrsquos histor-ically contingent social and economic success

(Streeck 1997b 54)

In sum the emergence of social systems of production is a long-termevolutionary process with each part interacting with its environment andresulting in a congurative whole but with no previous design by eithera single actor or a collectivity of actors As suggested above actors do not recreate their world anew At any moment in history the totalinstitutional complex of a society necessarily contains the resources and legacies of its past Preexisting institutional complexes are nevercompletely wiped out Societies consist of multiple layers of history withtheir diverse logics of action However the cumulative effect of smallperipheral changes in altering particular institutional sectors and insti-tutional arrangements can be substantial (Murmann 1998 Lanzara 1998)lsquoInstitution building is affected more by the ways in which people havecodied the past than by how they envision the future There are manymore potential resources in pre-existing institutional arrangements thanit is usually assumed or suspectedrsquo (Lanzara 1998 30)

Fourth level of analysis

The next level of institutional analysis organizational structures is some-what more controversial As suggested above Douglass North (1990)draws a sharp distinction between institutions and organizations Morerecently many organizational theorists (Powell 1991 Powell andDiMaggio 1991 DiMaggio and Powell 1983 Baum and Oliver 1992Kondra and Hinings 1998 Townley 1997) argue that institutional rulesnorms and conventions unfold in tandem with organizational structuresand this is my positionThe literature which focuses on how institutions inuence organiza-

tions is quite different from two other theoretical literatures which alsoare concerned with how organizational environments shape the behaviorof organizations There are resource dependency theorists (Pfeffer andSalancik 1978) who emphasize the role of environmental resources inshaping organizations while Hannan and Freeman (1977 1984) and otherpopulation ecologists emphasize the survival of organizations givencertain kinds of organizational conditions (Orru et al 1991)The institutionalist perspective as implied above emphasizes the

normative environment in which organizations are embedded It is a perspective which focuses on the way in which organizations in their behavior tend to conform to the institutional rules and normswhich are dominant in the organizational environment However allthree perspectives ndash the institutionalist the population ecology and

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

619

the resource dependency ndash emphasize how the environment inuencesorganizations and how organizations which are subject to the sameenvironment tend to converge in their behavior to have what DiMaggioand Powell have labeled organizational isomorphism (DiMaggio andPowell 1983)There has emerged a vast literature which demonstrates that within

each society there are scal political judicial and other regulatory normswhich limit and shape the culture and structure of organizationalbehavior The normative institutional environment of organizations limits the options of what organizations do in a particular society and inuences the patterns of ownership relations with suppliers andcustomers In short it is the normative environment of organizationswhich denes within a particular society what is socially acceptablebehavior for organizations (Hamilton and Biggart 1988 Hollingsworthand Hollingsworth 2000 Meyer and Rowan 1991 Zucker 1987 1988Orru et al 1991 Townley 1997)Thus far most of the analyses on isomorphic pressures operating on

organizations have been cross-sectional in nature However the entireconcept of isomorphism implies that there are strong environmental pres-sures exerted on organizations and in order to observe this phenomenonit is important that we have longitudinal studies which assess howchanges in an organizationrsquos institutional environment inuence changesin the structure and culture of organizations Moreover the expectationhere is that when one analyzes the patterns of organizational changeover time the isomorphism will be holistic ndash that is within a particularsociety there will be a tendency for the internal structure of particulartypes of organizations to converge in large part because of the pres-sures to conform to the changes in the external norms rules and valuesA close reading of Aoki (1990) and Slack and Hinings (1994) nds thatchanges in the national institutional environment of organizations haveinuenced changes in the internal structure of organizations within aspecic industry of a particular societyHowever it is inappropriate to view organizations as changing only

in relation to exogenous environmental change There is a kind of co-evolution which occurs between organizations and their institutionalenvironments Co-evolution implies nonlinear feedback between orga-nizations and their institutional environments This is of course quitedifferent from the more common way of studying organizations bymodeling relations between independent and dependent variables Ofcourse the more common type of analysis is quite appropriate whenthere are simple relationships which do not involve complex feedbackprocesses But in the strategy for institutional analysis which is proposedherein it is less useful to separate independent from dependent vari-ables and more useful to understand the interacting and co-evolutionary

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

620

processes among all of the levels of institutional analysis discussed inthe various sections above (Baum and Singh 1994a 379ndash80 Mowery andNelson 1999 Murmann 1998)The logic of the perspective herein suggests that within every

society there is variation in the structure and culture of business rmsuniversities and other complex organizations (Kondra and Hinings1998) However this variation takes place within parameters which aresystem specic For example every German rm and every Germanuniversity is different from every other German university and rm but there are a set of cultural and structural characteristics which distin-guish German rms and universities from those in America Similarlyeach Japanese rm and university is unique yet they have a set ofcultural and structural characteristics which are system specic andwhich differentiate them from business rms and universities in other countriesIn all societies each organization has its own distinctive organizational

rules norms and conventions which are subordinate to the meta norms and rules of the larger society within which they are embeddedHowever the strength of the institutional environment within whichorganizations are embedded varies from society to society Some societies have multiple institutional environments and there is hetero-geneity in terms of what constitutes appropriate behavior oforganizations In such societies there can be conict over whatinstitutional logics should regulate specic organizational functions(Friedland and Alford 1991 Townley 1997 262ndash3) In those societies inwhich the institutional norms habits and rules are most developed and in which the institutional pressures to conform are greatest thereis less variation in the structure and culture of business rms and variouskinds of research organizations In such societies the connectednessbetween research organizations and their institutional environments is sufciently strong for organizations to have low autonomy to pursue independent strategies and goals in these societies there is a great deal of organizational isomorphism Conversely the weaker the institutional environment in which research organizations areembedded the greater the variation in the structure and culture of business rms and research organizations Moreover where the institu-tional environments are more weakly developed organizations havegreater autonomy and exibility to respond to the development of new knowledge and to be highly innovative Hence it is in those societies where the institutional environments are most developed and most rigid and when there is less organizational autonomy and exibility that fewer radical innovations in basic and applied science as well as in totally new products and industrial sectors havebeen made

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

621

Fifth level of analysis

We come next to the outputs and performance of the various institutionalcomponents of a society It is at this level that institutional componentsare more pragmatic and exible For example in the legal sector thereare specic statutes and court rulings in the state sector there are specicpolicies in the business sector there are new products new technolo-gies and market strategies It is at this nal level of analysis thatinstitutional spheres are most open and susceptible to change and cross-national mimicry is easiest and most commonThrough the outputs of the society we can obtain some assessment as

to how well a society is performing Moreover we can assess how inno-vative it is how egalitarian it is in the distribution of its resources howegalitarian it is in terms of levels of health education etc Just as it iscomplicated to measure how egalitarian a society is similarly it is dif-cult to assess how innovative it is For example societies vary in makingincremental and radical breakthroughs in basic and applied science in developing totally new products and new kinds of organizations inincremental and process innovations in existing products and organiza-tions in developing new and different forms of marketing ndash bothdomestically and globally (Hage and Hollingsworth 2000) Of coursesocieties vary in their rates of economic growth in their rates of economicproductivity in their quality of life (eg rates of crime life expectancyetc) All of these performance criteria feed back and inuence each ofthe levels of institutional analysis discussed above ndash rules norms valuesetc institutional arrangements institutional sectors the structure andform of organizations Moreover performance measures may inuenceeach other For example if a society has very low rates of growth itmay not be very innovative in some of the types of innovative activitiesmentioned above Good or poor performances of certain kinds inuenceother performance indicators (Hollingsworth et al 1990)Different institutional arrangements and different social systems of

production result in different types of economic performance Hence solong as societies have different social systems of production there areserious constraints on the degree to which they can converge in theirinnovative styles Different social systems of production tend to maxi-mize in a more or less explicit manner different performance criteriausually mixing considerations about static and dynamic efciency protsecurity social peace and economic andor political power In short incontrast to the implications of neoclassical economic theory in real-worldeconomies there are no universal standards all economically rationalactors attempt to maximize Economic history provides numerous exam-ples of how a great variety of principles of rationality are implementedin different societies

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

622

A critical question in institutional analysis is whether an existing socialsystem of production which supports a set of routines for a particularkind of technology and industry can shift from old practices to newones There are numerous historical accounts of how the social systemof production in a particular society which worked so well for a numberof years could not adapt to new technologies (Lazonick 1990 Veblen1915 Nelson 1994) Schumpeter (1983) and Freeman (1991) have devel-oped the idea of a techno-economic paradigm which suggests thatdifferent technologies require specic forms of organizational arrange-ments ndash or what I called social systems of production Because oflsquoinstitutional inertiarsquo the social system of production of a society maynot be able to adapt to a new techno-economic paradigm (Nelson 199458 Lanzara 1998)Whether a social system of production can sustain its performance

standards depends not only on its intrinsic economic rationality but alsoon where it ts into a larger system If a particular social system ofproduction is immune from the innovativeness and competition of analternative system survival can be long lasting But if different socialsystems of production with diverging criteria of good economic perfor-mance meet in the world arena the arbitrariness of nationally imposedconstructed performance standards may be superseded by alternativeperformance criteria as a result of international competitivenessThe world economy is also socially constructed just as are national

economies Even if different social systems of production are competingin the international arena it is not always possible to determine whichis more competitively effective at any moment in time Hegemonicnation-states in the short run shape the rules of trade to favor their indus-trial sectors and rms But the history of hegemonic powers suggeststhat in the longer run such social systems of production sustainedlargely by military and political power eventually give way to moreinnovative and competitive social systems of production In our ownday as nation-states are increasingly integrated into a world economyeven hegemonic powers lose their innovative and competitive advan-tage and their share of world output decreases Such a country mayattempt to restructure its institutional arrangements and to readjust itsperformance preferences But to restructure its system of innovationgenerally calls for a major redistribution of power within a societyLargely for this reason societies have historically had limited capacityto reconstruct a system of innovation in the image of their major competi-tors Each countryrsquos social system of production is a conguration of ahost of norms rules and values as well as of institutional arrangementsEach system is constantly changing and is open to inuence from othersystems And indeed many technologies and practices diffuse from onesociety to another but a societyrsquos capacity to be innovative is constrained

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

623

by the existing social system of production Thus the same technologymay exist in numerous countries but how it is employed and how itinuences societal outcomes and performances varies from one institu-tional conguration or society to another (Hollingsworth 1998)

PERSPECTIVES ON INSTITUTIONAL CHANGE

A fundamental problem which remains unresolved in institutionalanalysis involves the nature of institutional change There is much confu-sion and miscommunication in the social sciences about institutionalchange in large part because we do not even know how to measure therate of institutional change Of course we know that there are bothexternal and internal focuses for change and we have had a good bitof scholarship on this subject What is not clearly understood is thatinstitutions and institutional arrangements within societies are histori-cally rooted that there is a great deal of path-dependency to the waythat institutions evolve Moreover the more intricately linked each sectoris with each other sector and with a societyrsquos rules and norms the lesschoice actors have to devise new institutions and institutional arrange-ments On the other hand if a sector or actor is too isolated it may betoo weak within a society to bring about any kind of effective changeThese considerations are very critical to the problem of how muchfreedom actors have to create new institutions afresh to what extentinstitutions diffuse from one society to another (Murmann 1998 Lanzara1998) This is a fundamental question for students of institutionalanalysis as we attempt to understand how the different institutionallevels are linked togetherThere are a variety of reasons why there is confusion in studying insti-

tutional change There is disagreement over the extent to which actorshave the freedom to build new institutional arrangements and if theydo to what extent the new arrangements and organizational patternsmay depart from past practices Among some institutional economists(Williamson 1985) there is a basic assumption that actors build efcientinstitutions by pursuing their self-interest and by promoting theirstrategic goals Moreover there is the assumption that a mixing of fullrationality with market competition tends to produce all of the optimalinstitutions that are needed in order to coordinate a complex capitalisteconomy (Hollingsworth and Boyer 1997 ch 14 Hodgson 1999) Incontrast the argument developed here is much more complexSocial institutions are historically rooted and there is a great deal of

path-dependency in the way that various institutional componentsevolve The shape of institutional congurations at any moment limitsthe type of options for change Because there is a great deal of institu-tional inertia radical change in the institutional components of a society

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

624

is uncommon (Lanzara 1998) Recognizing the constraints existinginstitutions exert on actors we note that existing congurations of insti-tutional arrangements institutional sectors and organizational propertiescan limit but also inuence the degree to which norms and rules canchange Many scholars (North 1990 Ostrom 1990) have emphasized therole of norms rules ideologies and values in limiting the ability of actorsto develop new institutional arrangements which radically depart fromexisting ones However there is a two-way process of interaction aprocess of co-evolution (Baum and Singh 1994a Nelson 1994) Thus asFriedland and Alford (1991 244) argue actors develop new institutionalarrangements by recognizing organizational failures and low perfor-mance but these new arrangements do not necessarily change a societyrsquosnorms rules habits and its underlying value systemCampbell (1997) advances our understanding of how existing institu-

tions enable actors to construct radically new institutions He recognizesthat most institutions and institutional arrangements embody a degreeof inertia from which it is difcult for actors to depart Moreover thedifferential power relations among actors make it unlikely that lesspowerful actors can change the existing power structure of a societyMost change in the institutional components of a society evolves througha process of constrained selection which reects to a considerable degreethe existing arrangements and power relations This becomes the inter-pretative frame for social actors with institutionalized scripts and ritualswhich tend to be taken for granted and appear to be quite ordinary andnatural This institutionalized scripting of the social world is an extremelyimportant source of social stability and inertiaCampbell has developed a very fruitful explanation of change in terms

of the various institutional components outlined here He distinguishesbetween incremental and radical institutional change and argues thatradical institutional change occurs when social actors with widelydiffering norms cultural scripts and rituals for action engage in intenseinteraction with each other (see also Knight 1992) The more diverse theinteractions the greater the potential for institutional change He arguesthat as shifts in the composition of interaction among social actors occurchanges in the interpretation of problems and interests will follow Ifthere are only minor changes in the extent to which diverse actorsinteract there will be only modest institutional change But the morefundamental the changes in the interaction of diverse social actors thegreater the change in the way that actors interpret their world And thegreater the change in the way that actors interpret their world the morelikely that radical changes in various institutional components willfollow In other words diverse decision making tends to create a widerrange of interpretative frameworks of the social world than is likely tobe the case if social actors operate in isolation

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

625

This perspective of diverse interactions among social actors facilitatesour ability to move beyond the traditional view that existing institutionsconstrain the range of institutional alternatives which actors face Theolder view has tended to suggest that actors engage in institutionalchange by extending existing institutional principles habits and conven-tions Specifying the conditions which facilitate fundamental institutionalchange permits a much richer perspective for institutional analysisCampbellrsquos perspective about the conditions under which fundamentalor radical change in institutional components occurs is consistent withmy own research about the conditions under which major discoveriesor fundamental new knowledge occur the more scholars with diversebackgrounds interact with intensity and frequency the greater the like-lihood that they will develop new and alternative ways of thinking(Hollingsworth and Hollingsworth 2000)These views of institutional change are also consistent with the recent

work of Nelson and his colleagues (Mowery and Nelson 1999 Murmann1998) They report the research on a variety of industrial sectors in anumber of different countries revealing that networks of actors in partic-ular industries business and professional associations universityresearch departments and institutes are frequently able to use their collec-tive power to modify a societyrsquos property rights research system andgovernmental policies It is the intense and frequent interaction of thesediverse complex sets of actors ndash sometimes constituting a technologysystem ndash that often leads to incremental changes in a societyrsquos total socialsystem of production It is this kind of collective action that brings aboutorganizational change change in institutional sectors institutionalarrangements and even in the rules and norms of a society And if weare to understand how the style of innovativeness of a society changeswe must rst understand the changes in its institutional makeup It ishighly unlikely that a society will develop a new style of innovativenesswithout changing its institutional structure

FIRMS THEIR INSTITUTIONAL ENVIRONMENT ANDNATIONAL SYSTEMS OF INNOVATION

Variation in the innovative style of societies is shaped by their insti-tutional makeup ndash the various levels which are discussed aboveinstitutions institutional arrangements the structure and cohesion of institutional sectors which constitute a societyrsquos social system of pro-duction and the structure and culture of its organizations ndash especiallyits business rms and other research organizations By undertaking aninstitutional analysis of a society one can begin to understand in whatkinds of organizations the production of specic kinds of knowledgetakes place and how this is linked to particular kinds of innovativeness

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

626

With such a perspective one can gain a rich understanding of why somesocieties excel in the production of radical breakthroughs in basic andapplied science and in developing radically new products and why othersocieties excel in more incremental innovationsReecting on these issues one cannot help being impressed with the

contrasts between the innovative styles of German and American orga-nizations For example Germany has continued to be very successful in making incremental innovations in many industries in which it wasvery competitive well before World War II paper printing materialsmachinery electrotechnical products motor vehicles chemicals textileyarns and fabrics But the Germans have been much less successful indeveloping totally new products or related innovations in biotechnologyelectronics telecommunications aircraft as well as other newer indus-trial areas In contrast the Americans have been very innovative in these and other industries with products having very short lives andwith technologies which change very rapidly and are very complex(Hollingsworth 1991 Soskice 1997 1999)The importance of institutions and innovativeness for international

competitiveness has led to the recognition that there are national systemsof innovation (Lundvall 1992 Nelson 1993 Edquist 1997 Hage andHollingsworth 2000) But even before the recent studies of nationalsystems of innovation Landes (1969) was a proponent of these viewswhen he made the argument that Germany had developed a nationalsystem of innovation by building on its education and scientic researchsystems which were lacking in Britain and elsewhere However the concept of a national system of innovation immediately poses theproblem of what kind of innovations should be consideredDavid Soskice an economist at the Wissenschaftszentrum Berlin has

addressed a set of problems somewhat complementary to the perspec-tive raised here His work (Soskice 1990 1997 1999) suggests that weare sensitive to the institutional environment in which the followingactors are embedded the employees of rms the owners and nanciersof rms the competitors of rms and nally their collaborators ndash aperspective complementary to Hollingsworthrsquos work on social systemsof production (Hollingsworth 1997 1998) When one reects on the insti-tutional contextsocial system of production within which these actorsare embedded one becomes aware that those countries which have some-what inexible labor markets external to rms but a business system in which rms have long-term commitments to their employees arestrongly associated with incremental but hardly at all with radical inno-vations In the rms of such countries (eg Germany and Japan) thereis much more consensus decision making than in societies where a lsquohireand rersquo set of practices is quite pervasive And this kind of consensusdecision making ndash while highly conducive to incremental innovativeness

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

627

ndash limits the capacity for radical innovations to occur This kind of lsquolongtermismrsquo set of norms and rules occurs not only in the employmentpractices of incrementally innovative rms but also in their nancingand ownership patterns (Hollingsworth 1997) During the past thirtyyears these were characteristics of many rms in Germany and JapanSignicantly in both of these countries venture capital markets werenot highly developed nancing tended to be based more on long-termbank loans than on the equity markets and the relationships of rmstended to be relatively stable And these stable institutional arrange-ments also contributed to a high degree of innovativeness ndash butincremental rather than radical innovationsMeantime we should also pursue Soskicersquos ideas (1999) about how

rmsrsquo relationships with their competitors and collaborators shape theirstyles of innovativeness In those advanced industrial societies in whichbusiness rms are highly mobilized into business associations theretends to be a high degree of standardized quality controls over technicalnorms and a high degree of reliance on legal sanctions as monitoringinstruments vis-agrave-vis association members As a result of this high degreeof integration of rms into associations member rms are more likelyto be associated with incremental innovations than is likely to be the case with rms in those societies where associations are less welldeveloped In the latter type of society radical innovations are morelikely to occur In those advanced industrial societies where rms havelong-term stable relationships with collaborators in product design andproduction the style of innovation is likely to be incremental And it is in those societies in which there is a weaker tradition of strongcollaboration among rms that radical innovations are more likely tooccurObviously even in a country where there may be a great deal of

radical innovativeness most innovations will be of an incremental typefor there are simply fewer radical innovations But Germany and therest of Europe not only have had relatively few radical innovations incomparison with the United States but they also have lagged behindthe USA in developing relatively totally new industries Thus there hasbeen much less innovative activity in Europe in high-technology andnewly emerging industries than in the USAIn complementary research Audretsch and Feldman (1995) have

demonstrated that the USA has been more successful in spawning rmsengaged in the early stages of the industry life cycle but relative toGermany had an environment unfavorable to innovative activity byestablished rms especially in more mature industries German rmshave tended to be associated with a routinized technological regime ndashwhere established enterprises have had strong incentives to engage inincremental innovations but new rms have tended to have fundamental

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

628

innovative disadvantages relative to the Americans (see Audretsch 1995Winter 1982 Casper 2000)On both sides of the Atlantic universities engage in basic and applied

research But signicantly the institutional environment in which univer-sities are embedded inuences the degree to which the research is likelyto represent major breakthroughs or is likely to have more modestimplications Recent research in biomedical science demonstrates that inthose societies in which university professors andor graduate studentshave the following characteristics there are likely to be relatively majorresearch breakthroughs graduate students have a long training periodhigh dependency on a major professor low autonomy in dening thenature of a research project restricted job options and a university envi-ronment where there is a very modest or nonexistent entrepreneurialculture For example in Germany ndash a country with relatively few radicalbreakthroughs ndash the scholar who nishes a habilitation will have spentmany years on a research project before qualifying for a permanent acad-emic position will have been in a high-dependency position with a seniorprofessor will have had limited autonomy to dene a research projectand will be constrained from pursuing a high-risk research project Theyoung German will have an extraordinarily high level of training butwill have faced a number of disincentives to undertake high-risk researchAnd as soon as heshe has completed the habilitation and becomes aprofessor administrative responsibilities constitute another disincentivefrom engaging in high-risk or long-term research projects Moreover theGerman professor is a civil servant with a more stable and secure incomethan hisher American counterpart In a very different institutionalenvironment American professors in large research universities tend tobecome much more entrepreneurial Today many American professorsbehave as though they were operating their own rms within univer-sities and thus devote considerable time to generating funds for researchprojects assistants secretaries supplies etc And the variation in thisentrepreneurial spirit across countries is associated with variation in thetaking of high-risk research strategies and the making of major researchbreakthroughsIn sum societies vary in terms of the institutional environment in

which their organizations are embedded and this variation inuencesthe degree to which they have industrial activities in the early or latestages of the industry life cycle successful new or mature industriesand radical vs incremental innovations And it is this variation whichis encouraging rms in countries with relatively few new industries and with few or no radical innovations to invest in those countries with many new industries and numerous radical innovations ndash as astrategy to be able to transfer new technologies from the latter institu-tional environment to the former

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

629

Table 4 summarizes the way that the norms rules and values of a society are associated with Americarsquos and Germanyrsquos institutionalarrangements their social system of production and their styles of inno-vativeness

CONCLUSION

Several themes deserve to be reemphasized Most importantly institu-tional analysis is emerging as a dominant issue in several social sciencedisciplines The study of various institutional components discussedabove is a complex subject requiring multiple levels of analysis and aknowledge of multiple sectors within societies It requires a theoreticalperspective that evolves from an empirical analysis in different societiesIf we are to advance the study of institutional analysis we need cross-disciplinary collaboration and team research which are not readilyachieved given the structure of most of our research universitiesAt present scholars in each academic discipline have their own distinc-

tive strategies for studying institutional analysis Thus economists tendto use several approaches (North 1990 Williamson 1985 Hodgson 19881998 1999) anthropologists another (Geertz 1995) and sociologistspolitical scientists and historians employ other strategies (Campbell etal 1991 Powell and DiMaggio 1991 Katznelson 1998 Hall and Taylor1996) With practitioners in various disciplines pursuing their distinctiveapproaches to institutional analysis there is low potential for collabo-rative institutional analysis across the social sciences The lessons fromthe biological sciences over the last 40 years are very instructive oncepractitioners from various biological disciplines began to cooperate instudying biological phenomena by using the same concepts at the mole-cular level theoretical advance was very rapid Comparable phenomenatook place in the development of other hybrid elds eg biochemistrybiophysics (Judson 1979 Olby 1979 Kohler 1982 1994) Similarly weresocial scientists in various disciplines to work collaboratively in the areaof institutional analysis there would very likely be an acceleration intheir theoretical knowledgeThis is not a plea that everyone engaged in institutional analysis should

do the same kind of research Indeed just as biologists working at themolecular level study many different kinds of problems similarly socialscientists engaged in institutional analysis would also work on manyseparate problems but in the spirit of a collective enterprise Some wouldconcentrate their attention on the study of rules norms habits conven-tions and values while others would study how these are associatedwith congurations of various institutional arrangements (types ofmarkets hierarchies networks associations communities clans statesetc) Others would work on specic institutional sectors (eg education

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

630

Table 4

Rules norms hab

its an

d values associated w

ith sp

ecic co

ngu

ration

s of coo

rdinating mecha

nism

s soc

ial sy

stem

s of

prod

uction

an

d nationa

l styles of inno

vation

USA

Germany

Norms rules values which provide incentives for

Norms rules values which provide incentives for

short-term horizons

long-term horizons

Con

gu

rative

forms of

Marke

ts c

orpo

rate hierarchies reg

ulatory state

Assoc

iation

s (busine

ss assoc

iation

s la

bor union

s)

coordinating

corp

oratist-type

networks

marke

ts c

orporatist-

mecha

nism

s (listed in

type

state

order of im

portanc

e in

the particu

lar social

system

)

Social sys

tem of

prod

uction

Bus

iness system

Firm

struc

tures

Con

glom

erates b

ut with mov

emen

t toward

Firm

more co

ngru

ent with pa

rticular prod

ucts

elim

ination of produc

tion

in som

e sectors

emph

asis on long

-term strateg

ies rm

s are well

emph

asis on sh

ort-term

strateg

ies high

integrated

differentiation

between divisions

of rms

Firm

s mob

ilized in

Low

High

and in

tegrated

into busine

ss

associations

Indu

strial relations

Low

deg

ree of jo

b secu

rity rigid in

ternal lab

or

High deg

ree of jo

b secu

rity

exible interna

l lab

or

system

marke

t ex

ible externa

l labo

r marke

tsmarke

t rigid externa

l lab

or m

arke

tPatterns of

Dispersed owne

rship frequ

ent turnov

erMore stab

le owne

rship patterns

owne

rship

Table 4

continue

d

USA

Germany

Norms rules values which provide incentives for

Norms rules values which provide incentives for

short-term horizons

long-term horizons

Social system of

Prod

uction

Bus

iness system

Training for labo

r an

dHigh deg

ree of voluntary decision mak

ing by

Collective an

d m

ore co

mpulso

ry d

ecisions

abo

ut

man

agem

ent

individuals as to ho

w m

uch

tim

e an

d ene

rgy

how m

uch ed

ucation

is ne

eded

he

avy em

pha

sis

to inv

est in educ

ation heavy

emph

asis on

on techn

ical training (eg en

gine

ering)

busine

ss sch

ools training

in marke

ting

and

sales accou

nting

Fina

ncial marke

tsEqu

ity an

d ven

ture cap

ital m

arke

ts highly

Equ

ity an

d ven

ture cap

ital m

arke

ts w

eakly

dev

elop

ed na

ncing from

cen

tral gov

ernm

ent

deve

lope

d he

avy em

pha

sis on

corporate loan

sto priva

te rm

s for dev

elop

men

t of new

tech

nologies related

to na

tion

al security

Unive

rsity research

High individual auton

omy high poten

tial for

Low

ind

ividua

l au

tono

my until almost 40

yea

rs of

system

individual creativity u

nive

rsity pr

ofesso

rs

age poten

tial for ind

ividual creativity high

ly

high

ly entreprene

urial

cons

traine

d unive

rsity pr

ofessors not so

entrep

rene

urial

Inno

vative

styles

Highly su

ccessful in m

ajor break

through

s in

High pe

rforman

ce in proc

ess an

d inc

remen

tal

basic an

d app

lied scien

ce go

od in dev

elop

ing

inno

vation

s in existing produ

cts not so su

ccessful

totally

new

lines of pr

oducts an

d in m

ajor break

through

s in basic scien

ce or in

tech

nologies high

deg

ree of entrepr

eneu

rship

deve

loping

totally new

lines of prod

ucts of

amon

g research

ers

tech

nologies low d

egree of entrepr

eneu

rship

amon

g research

ers

Table 4

continue

d

USA

Germany

Norms rules values which provide incentives for

Norms rules values which provide incentives for

short-term horizons

long-term horizons

Prod

ucts in w

hich

Pr

oducts with sh

ort liv

es prod

ucts in w

hich

Older produ

cts with long

history machine

ry pap

er

society excels

tech

nology

is bo

th highly co

mplex

and

pr

inting

textile

s m

otor veh

icles m

achine

too

lsch

ange

s rapidly en

tertainm

ent indus

try

pha

rmaceu

ticals biotechn

olog

y softw

are

aerosp

ace

business systems nancial markets and systems of research) Distinctivewould be the recognition of how these separate studies are linkedtogether how each institutional sector is linked to a societyrsquos normsrules values etc and its conguration of institutional arrangementsSimilarly other institutional analysts might focus on the study of orga-nizations with a major concern for how the institutional environmentof organizations (eg norms rules values congurations of institutionalarrangements institutional sectors) inuenced the structure culture andoutcomes of organizations And nally others might study how all ofthe aforementioned aspects of institutional analysis inuenced a societyrsquosoverall performance Nelson and Sampat (1998) have appropriatelyreminded us that we should engage in institutional analysis in order toaddress specic problems but the view developed in this article is thatwe rst need some mapping of the terrain of institutional analysis inorder for our diverse enterprises to become a meaningful collective enter-prise especially if we are to understand how institutions inuence theinnovative processThe argument developed here is not that all social scientists should engage

in institutional analysis Obviously there are numerous other importantareas of research which lie beyond the eld of institutional analysis Nordoes this article argue that institutional analysis is the most importantsubject for social scientistsSince the structure and culture of our research universities reect an

enormous amount of disciplinary fragmentation what are the prospectsfor promoting the kind of collective and interdisciplinary research agendaproposed above In one respect the prospects are not encouragingDuring the past half century the efforts of American universities to have effective communication across the biological sciences have variedgreatly in their success At the University of California-Berkeley it was not until the 1980s that there was a serious effort to integrate thebiological sciences and at that time the campus abolished seventeendepartments to promote an integrated biological program But at theUniversity of Wisconsin the biological sciences are still very fragmentedinto dozens of different departments in botany plant pathology zoologyand physiology with two departments of biochemistry and two depart-ments of geneticsThere have been other approaches which have worked very well histor-

ically for promoting scientic integration amidst scientic diversity Forexample several summer schools and special institutes were organizedin the United States during the 1940s in an effort to overcome the frag-mentation in the biological sciences These programs made it possiblefor scientists in diverse academic disciplines and from different univer-sities and countries to come together to engage in intense and frequentinteraction Perhaps the most effective program was that held in the

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

634

summers at Cold Spring Harbor New York where a research agendafor the study of molecular biology was developed (Fischer and Lipson1988 Stent and Watson 1966) Scientists went there summer aftersummer and when they returned to their universities many began therestructuring of the biological sciences in their universitiesIn our own day we need summer institutes and other similar programs

so that scholars from diverse backgrounds can come together anddevelop a collective agenda for doing institutional analysis and forstudying the impact of institutions on the performance of societies ndashinnovativeness being one such subject If we can develop such instituteswe will increase the prospects for transforming our universities Thereis reason to be optimistic that once an eclectic group of scholars workwith a common set of concepts we can have genuine theoretical advanceif researchers can bring their diverse expertise to bear on a common setof problems There is considerable evidence that institutional analysisoffers high potential for advancing our knowledge of innovations andtechnological change (Hollingsworth and Hollingsworth 2000) If we aregoing to make substantial theoretical advance it is not likely to occurfrom the vantage point of a multiplicity of disciplines Rather it is likelyto occur as we develop a new intellectual framework and our own acad-emic journals This is a tall agenda which requires us to reassess theway that knowledge is presently produced in our societies

NOTE

This article is part of a much larger agenda involving the study of institutionsorganizations and innovations in which I am involved My debts are many andvaried The rst version was written while I was in residence in the ResearchUnit for Institutional Change and European Integration of the Austrian Academyof Sciences in Vienna in the summer of 1998 I am especially grateful to ProfessorsEgon Matzner and Sonja Puntscher-Riekmann for providing the stimulating envi-ronment in which to think through the ideas developed here Egon Matzner hasengaged in very stimulating discussions with me about the importance of insti-tutions in the general eld of socio-economics The nal version of the articlewas completed during my residence as a fellow at the Neurosciences Institutein La Jolla California I also wish to thank my colleagues in the InnovationTheme Project at the Netherlands Institute for Advanced Study (NIAS) during1998ndash9 and Edgar Grande and Karl Muumlller for stimulating conversations aboutthe interaction of institutions and organizations in facilitating particular typesof innovation I am especially grateful to Jerald Hage for helping me to developa number of issues in the article and for his useful comments on the rst draftFrans van Waarden early on helped me to work through the idea of multiplelevels of institutional analysis Others who have helped to develop the ideas inthe article have been my colleagues in the research project lsquoComparative socialsystems of productionrsquo especially Tom Burns Christel Lane Yoshitaka OkadaWolfgang Streeck and Richard Whitley My colleagues Robert Boyer SteveCasper and David Soskice have taught me much about how incentives providedby the institutional environments of rms inuence individual decision making

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

635

But my most important debt is to Ellen Jane Hollingsworth who has repeatedlyquestioned every single assumption of this article Without her rich and contin-uous tutoring this article would be even less developed than it is at presentFinally without the insightful assistance of David Gear in many phases of thisarticle it would not have been completed

REFERENCES

Aoki Masahiko (1990) lsquoTowards an economic model of the Japanese rmrsquo Journalof Economic Literature March 1ndash27

Arora A Landau Ralph and Rosenberg Nathan (eds) (1998) Chemicals and Long-Term Economic Growth Insights from the Chemical Industry New York JohnWiley amp Sons

Arrow Kenneth (1974) The Limits of Organization New York NortonArthur W B (1988a) lsquoSelf-reinforcing mechanisms in economicsrsquo in P W

Anderson K J Arrow and D Pines (eds) The Economy as an Evolving ComplexSystem Redwood City Calif Addison-Wesley

mdashmdash(1988b) lsquoCompeting technologies an overviewrsquo in G Dosi C Freeman RNelson G Silverberg and L Soete (eds) Technical Change and Economic TheoryLondon Pinter Publishers pp 590ndash607

Audretsch David B (1995) Innovation and Industry Evolution Cambridge MassMIT Press

Audretsch David B and Feldman Maryann P (1995) lsquoInnovative clusters andthe industry life cyclersquo Discussion Paper Wissenschaftszentrum BerlinMarket Processes and Corporate Development Research Unit FS-IV-95-7

Baum Joel A C and Oliver Christine (1992) lsquoInstitutional embeddedness andthe dynamics of organizational populationsrsquo American Sociological Review 57540ndash59

Baum Joel A C and Singh Jitendra V (1994a) lsquoOrganizationndashenvironment coevo-lutionrsquo in Joel A C Baum and Jitendra V Singh (eds) The EvolutionaryDynamics of Organizations New York Oxford University Press pp 379ndash402

mdashmdash(eds) (1994b) The Evolutionary Dynamics of Organizations New York OxfordUniversity Press

Berger Suzanne and Dore Ronald (eds) (1996) National Diversity and GlobalCapitalism Ithaca NY Cornell University Press

Boyer Robert (1997) lsquoThe variety and unequal performance of really existingmarkets farewell to Doctor Panglossrsquo in J Rogers Hollingsworth and RobertBoyer (eds) Contemporary Capitalism the Embeddedness of InstitutionsCambridge and New York Cambridge University Press pp 55ndash93

Burns Tom R and Dietz Thomas (2001 forthcoming) lsquoSocial rule system theorysocial action institutional arrangements and evolutionary processesrsquo in JRogers Hollingsworth and Karl Muumlller (eds) Advances in SocioeconomicsLanham Md Rowman amp Littleeld

Burns Tom R and Flam Helena (1987) The Shaping of Social Organization BeverlyHills Calif Sage

Calvert Randall L (1995) lsquoThe rational choice theory of social institutionsrsquo inJeffrey S Banks and Eric A Hanuskek (eds) Modern Political Economy NewYork Cambridge University Press pp 216ndash66

Camic Charles (1986) lsquoThe matter of habitrsquo American Journal of Sociology 911039ndash87

Campbell John L (1997) lsquoMechanisms of evolutionary change in economicsgovernance interaction interpretation and bricolagersquo in Lars Magnusson

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

636

Table 3

Failu

res of coo

rdinating mecha

nism

s

Coordinating mechanisms

Type of failure

Associations

Private hierarchies

State

Market

Communities

Networks

Enforcem

ent

Usu

ally relies on

Might enh

ance

Needs co

ntrol

Needs an

internal

Needs

tru

st and

Need an external

the state as an

oppo

rtun

istic

extern

al to state

enforcem

ent

loya

lty often

en

forcem

ent

enforcer

beha

vior

bureau

cracy

authority

coming from

au

thority

Resem

bles

The

ideal o

f (ju

dge

s

Facilitates

outside (fam

ily May

facilitate

enforcem

ent

internal m

arke

ts

parliamen

t collu

sion

and

relig

ion

cartelization an

d mecha

nism

might hur

t marke

t) to

impe

rfect

ethn

icity)

mon

opoly

of cartels

incu

mbe

nt

correct state

compe

titio

nCom

patible with

worke

rsab

uses

variou

s types of

Lob

bies can

co

mpe

tition

captur

e pu

blic

interest goa

ls

Public goo

d

Useful for

Gov

erna

nce co

sts

Can

provide

Can

not provide

Can

internalize

Useful for

and

establishing

might exceed

public goo

ds

colle

ctive go

ods

some co

llective

enha

ncing qu

ality

externality

stan

dards an

d

the be

nets of

but ha

s or deal w

ith

good

s (qua

lity

and training bu

t qu

ality for

internal division

difc

ultie

s in

externalities

training

) bu

t no

t no

t ve

ry goo

d in

setting ru

les of

of la

bor

prov

iding them

Inad

equate

othe

rs (welfare

prov

iding for

compe

titio

n in

Slow

to react to

in precise

mon

itoring of

gene

ral p

ublic

societal gen

eral

the indu

stry

chan

ges in the

am

ounts

tech

nical ch

ange

go

ods)

welfare

Useful for

environm

ent

Might fail in

and inno

vation

Mem

bers tightly

Weak in the

prov

iding

induc

ing

integrated

into

prov

ision of

man

y go

ods

tech

nical

community

colle

ctive go

als

colle

ctively

chan

geha

ve limite

d

that in

dividua

lcapa

city for

mem

bers can

not

inno

vatio

nsprov

ide for

them

selves

Table 3

continue

d

Coordinating mechanisms

Type of failure

Associations

Private hierarchies

State

Market

Communities

Networks

Efcien

cyFa

cilitates

De

cien

t in

Can

be high

ly

Some ba

sic social So

me go

ods

Slow

to en

hanc

e co

operation an

d

coop

eration an

d

bureau

cratic

relation

s cann

ot

cann

ot be

efc

ienc

y an

d

X-efcien

cy but

X-efcien

cyan

d can

not

be provided

by

delivered

at

speed of

not allocative

easily deliver

pure m

arke

t su

fcien

tly

adap

tive

ness

efcien

cy

good

s at

mecha

nism

slow costs

except in

low cost

indus

tries whe

re

tech

nology

is

complex an

d rapidly cha

nging

Equity

Narrow

Excessive

Might enh

ance

Facilitates

Might lead to

Whe

n widely

encompa

ssing

multiplication

ineq

uality

ineq

uality in

retard

ed

dev

elop

ed in

to

associationa

l of con

trollers

(pow

er

income an

d dev

elop

men

tindus

trial

stru

ctures lead

(fru

stration

and

an

d privilege

)wealth

districts

to in

come

ineq

uality)

netw

orks m

ay

equa

lity

facilitate greater

equality an

d

inco

me

distribution

whe

n wea

kly

dev

elop

ed

netw

orks ten

d to

increase soc

ial

ineq

uality

Note For allo

cativ

e efc

ienc

y an

d X

-efcien

cy see Leibe

nstein (19

66 19

76)

they raise other no less severe problems It is important to recognizethe imperfection of any single institutional arrangement in order tocomprehend the origin and transformation of any other institutionalarrangement (Hodgson 1999 ch 3)As suggested above each institutional arrangement is congured with

other institutional arrangements Usually one particular institutionalarrangement is more dominant in a particular conguration but becauseeach type of institutional arrangement has its own strengths andweaknesses there is no simple structural logic in the governance or coor-dination of a society Each institutional arrangement constrains the otherbut the inherent tension among the various institutional arrangementswithin a conguration contributes to changes in the governance orcoordination of a society The routines and logics of each of these cong-urations of institutional arrangements provide constraints and incentivesfor actors The way that actors perceive the incentives and constraintsof these governance congurations leads to a particular market logicand it is the specic market logic of a society which inuences its speciccapacities and weaknesses The inherent strain among the different logicsin a conguration of coordination helps to provide the exibility for asociety to adapt to new circumstancesIf there were a society with pure markets or a society coordinated

only by the state there would be too much rigidity and too little diver-sity to cope with the vast uncertainty in the global environment A societywith very little diversity in its coordination mechanisms would havelittle capacity to adapt to new circumstances Congurations with consid-erable diversity of institutional arrangements provide for a certainamount of incoherence in governance but they also provide for thecapacity to adapt to new circumstances While the Soviet Union wasdominated by the state as a coordinating mechanism there were alwaysfunctioning markets in the system Moreover the feudal society of thetwelfth and thirteenth centuries consisted not only of hierarchical rela-tionships between serfs and masters but also of urban guilds clericalhierarchies and markets (Johnson 1992 38) In sum the robustness ofinstitutions often depends on multiple and diverse principles and logicsof action on the inconsistency of principles and procedures on patternedforms of disorder (Lanzara 1998 Orren and Skowronek 1991 320 329)

Third level of analysis

The next level consists of the institutional sectors of a society The rulesnorms and values of a society inuence the array of institutional arrange-ments and levels one and two inuence the nature of and the relation-ships among various institutional sectors and all three levels areintricately linked together to form a social system of production Together

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

613

all of these components inuence the performance of economic sectorswithin a society as well as the performance of the total society A socialsystem of production is the way that a societyrsquos institutions (see rstlevel of analysis above) its institutional arrangements (see second levelof analysis above) and its institutional sectors are integrated into a socialcongurationAn institutional sector includes all organizations in a society which

supply a given service or product along with their associated focal orga-nizations (eg major suppliers funding sources regulators and so forth(Scott et al 1994 108 117 Campbell et al 1991 Hollingsworth et al1994) Institutional sectors include but are not limited to the societyrsquossystem of education system of research business system nancialmarkets legal system and the state The structure of the nancial marketssystem of training and education industrial relations system system ofresearch and state and legal systems are distinct and idiosyncratic ineach society In short they are system specicAll of these institutional sectors tend to cohere with each other

although they vary in the degree to which they are tightly coupled intoa full-edged system While each of the institutional sectors has someautonomy and may have some goals that are contradictory to the goalsof other institutional sectors with which it is linked an institutional logicin each society leads institutions to coalesce into a complex social con-guration (Hollingsworth 1991) This occurs because the institutionalsectors are embedded in a culture in which their logics are symbolicallygrounded organizationally structured technically and materially con-strained and politically defended The conguration of institutionalsectors usually exhibits some degree of adaptability to new challengesbut continues to evolve within an existing style But under new circum-stances or unprecedented disturbances these institutional congurationsare exposed to sharp historical limits as to what they may or may notdo (Schumpeter 1983 David 1988 Arthur 1988a 1988b Haringkanssen andLundgren 1997)Why do all of these different institutions coalesce into a complex

social conguration which is labeled here as a social system of pro-duction The literature suggests two contrasting interpretations Part ofthe answer ndash indeed a controversial one ndash is that these institutionalsectors are functionally determined by the requirements of the practiceof capitalism in each time and place (Habermas 1975) Another expla-nation emphasizes the genesis of the actual conguration via a trial and error process according to which the survival of rms regions orcountries is the outcome of complex evolutionary mechanisms (Maynard-Smith 1982 Nelson and Winter 1982) However the problem is evenmore complex Markets and other mechanisms for coordinating rela-tionships among economic actors place constraints on the means and

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

614

ends of economic activity to be achieved in any society The othercoordinating mechanisms include different kinds of hierarchies varioustypes of networks and associations ndash trade unions employers and busi-ness associations (Hollingsworth and Lindberg 1985 Campbell et al1991) The logic of these various institutional arrangements providesactors with vocabularies and logics for pursuing their goals for deningwhat is valued and for shaping the norms and rules by which theyabide (see Tables 2 and 3) In short in contrast to the logic of theneoclassical paradigm the argument here is that the dominant type ofinstitutional arrangements places severe constraints on the denition of needs preferences and choices of economic actors Whereas the neo-classical paradigm assumes that individuals and rms are sovereign thisarticle is based on the assumption that rms are inuenced by the hold that the institutional congurations making up a social system ofproduction have on individual decision making (Campbell et al 1991Etzioni 1988 Streeck and Schmitter 1985a Hollingsworth et al 1994Hollingsworth and Boyer 1997 Magnusson and Ottosson 1997 North1990 Hodgson 1999)Standard neoclassical economic theory has tended to downplay the

role of production and consequently of rms Even transaction cost theo-rists who are concerned with analyzing the rm as a coordinatingmechanism have been relatively unconcerned with the various compo-nents of a social system of production Indeed as long as there waswidespread optimism about the efcacy of Keynesian economics therewas relatively little concern among neoclassical economists with thesupply side of the economy Even in the opinion of most Keynesians agroup of experts should ideally be able to shape the size of aggregatedemand while the supply side of the economy would be left to the twominimalist institutions of neoclassical economics ndash markets and manage-rial hierarchies For several decades however it has become increasinglyobvious that some of the most competitive and successful patterns ofindustrial output and industrial production in capitalist economies donot derive from the neoclassical prescription of unregulated markets andcorporate hierarchies complemented by a neoliberal democratic stateIndeed it is now well understood that certain highly successful produc-tion patterns require for their emergence and survival institutionalarrangements and environments the very opposite of the prescriptionsfound in the neoclassical paradigm (see especially Streeck 1992 but alsoHollingsworth and Streeck 1994) Thus social scientists have come tounderstand that if they are to comprehend the behavior and performanceof contemporary economies concerns about social systems of produc-tion must be brought into the pictureBecause production involves more than technology a number of social

scientists have an increasing concern with social systems of production

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

615

The same equipment is frequently operated quite differently in the samesectors in different countries even when rms are competing in the same market (Maurice et al 1980 Sorge and Streeck 1988 Sorge 1989Hollingsworth et al 1994) Variations in production and process tech-nologies are inuenced partly by variations in the social environmentsin which they are embedded In other words rms are embedded intocomplex environments which among other things place constraints ontheir behavior Thus a social system of production is of major impor-tance in understanding the behavior and performance of an economyHow the state and other institutional arrangements (eg marketsnetworks private hierarchies associations) coalesce and are related toparticular social systems of production are important determinants ofeconomic performanceThe congurations of institutional arrangements that coordinate or

govern the behavior of actors in one society and its structure of specicinstitutional sectors cannot easily be transferred to another society forthey are embedded into a social system of production that is societallydistinct (Hollingsworth 1997) Societies borrow selected principles offoreign management styles and work practices but the effectiveness ofsuch borrowing is generally limited Economic behavior and performanceare shaped by the entire social system of production in which actors areembedded and not simply by specic principles of particular manage-ment styles and work practices Moreover a societyrsquos social system ofproduction tends to limit the kind of goods that it can produce and withwhich it can compete successfully in international marketsSome scholars (Kenney and Florida 1993 Oliver and Wilkinson 1988)

have assumed that the diffusion of particular forms of management stylesand work practices across societies could lead to system convergenceReferring to the Japanese production system Kenney and Florida haveargued that it consists of organizational practices whose fundamentallsquogenetic codersquo can successfully be inserted into another society and beginsuccessfully to reproduce its behavior in the new environment (1993 8)Their position is in the intellectual tradition of Antonio Gramsci (1971)who contended decades ago that the American system of mass produc-tion would most certainly diffuse to Europe over timeHowever the argument of this article is that even though British

French and American rms may adopt certain aspects of Japanesemanagement styles (eg just-in-time production self-managing teamsquality circles the use of statistical process controls) or some variant ofthe German vocational system their social systems of production willnot converge Moreover the overall conguration of a societyrsquos socialsystem of production inuences its sectoral and overall national economicperformance This helps to explain why societies have different systemsof innovation and why they vary in the clusters of industries in which

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

616

they are highly competitive in the international markets (Nelson 19931999 Hage and Hollingsworth 2000 Berger and Dore 1996)In order to understand how and why a societyrsquos economy performs

as it does it is necessary to understand its entire social system of produc-tion If a society is to modify substantially the performance of itsinstitutional sectors and its entire economy it cannot adapt only someof the management and work practices of its foreign competitors Ratherit must alter its entire social system of production Because a societyrsquosmodes of economic governance and coordination and its institutionalsectors develop according to particular logics and are system specicthere are serious limitations to the extent to which a society may mimicthe institutions (eg the rules and norms) institutional arrangements andinstitutional sectors of other societiesIn the history of modern societies there are logics by which institu-

tions coalesce into a social system of production (Hollingsworth 1991)Though institutions are constantly changing there are sharp limits tothe type and the direction of change that any particular institution canundergo because of its linkages with institutional arrangements and insti-tutional sectors Thus a societyrsquos business rms educational systemnancial markets industrial relations system etc can engage in seriousrestructuring only if most of the other institutional sectors also changeIn social systems of production there are pressures toward consistencyin the norms rules and values across institutional sectors though in anycomplex society social systems are obviously imperfectly integratedIndeed the degree to which the institutional norms and rules makingup a social system of production are loosely or tightly coupled is a vari-able of considerable importance In general the institutional sectorsmaking up a social system of production are interdependent and changesin one generally result in changes in the others Since each institutionallevel is dependent on the others for various types of resources there isinterdependence among the differing institutional spheres Moreovereach society has its norms moral principles rules and laws and recipesfor action as well as its own idiosyncratic customs traditions and prin-ciples of justice (Burns and Flam 1987)There are also other inherent obstacles to convergence among social

systems of production for where a system is at any one point in timeis inuenced by its initial state Systems with quite different initial statesare unlikely to converge with one anotherrsquos institutional practicesExisting institutional arrangements block certain institutional innova-tions and facilitate others (Roland 1990) Thus the institutions makingup a social system of production provide continuity even though insti-tutional arrangements are always changing but with a particular logicWhile Williamson (1975 1985) suggests that actors tend to choose theinstitutional arrangements which are most efcient North (1990) is

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

617

much closer to the mark in his argument that most societal institutionalarrangements exist as a result of custom and habit and are thereforeinefcient At any moment in time the world often appears to its actorsas very complex and uncertain For this reason actors often engage incontradictory forms of behavior pursuing different strategies as hedgesagainst a very uncertain world (Lanzara 1998) And their hedging andcontradictory forms of behavior may lead in somewhat different soci-etal directions all constrained by the institutional fabric within whichthe actors are embedded This kind of contradictory behavior occurs inpart because of the contradictions inherent in the dominant congura-tions of institutional arrangements with their contradictory logics (seelsquoSecond level of analysisrsquo above)Despite the emphasis on the logic of institutional continuity this is

not an argument that systems change along some predetermined pathThere are critical turning points in the history of highly industrializedsocieties but with the choices limited by the existing institutional terrainBeing path-dependent a social system of production continues along aparticular logic until or unless a fundamental societal crisis intervenes(Milgrom et al 1991 Krugman 1991 Durlauf 1991 Hollingsworth 1991Pred 1966 David 1988 Hodgson 1998)At this point it is important to confront the question of how social

systems of production evolve And this question gets to the heart of theproblem of building complex societal structures Certainly social systemsof production did not emerge from some process of social engineeringMoreover the various component parts of each social system of produc-tion have often not been designed to be part of a social system ofproduction The component parts of each social system of productionhave emerged more often than not as unintended by-products of goalswhich various actors had in mind at earlier moments in time It is usuallythe case that actors at time lsquotrsquo fail to comprehend the long-term conse-quences of their actions Wolfgang Streeck the author of several brilliantpapers on social systems of production argues that institutional sectordesigns created for one purpose generally address that goal but overtime those designs have quite unintended consequences (Streeck 1997a1997b Wright 1998) Describing the conguration of the German socialsystem of production of the 1970s and 1980s Streeck points out that itwas the unintended by-product of multiple points in history Someelements were pre-Wilhelmian others were introduced by the Alliedpowers after 1945 and others emerged during the years of the GermanFederal Republic All component parts of the German social system ofproduction

were and continue to be changing for their own reasons as wellas in reaction to each other and certainly there can be no presump-tion of a pre-established t between them even though one might

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

618

want to allow for some reinforcement effects of the modelrsquos histor-ically contingent social and economic success

(Streeck 1997b 54)

In sum the emergence of social systems of production is a long-termevolutionary process with each part interacting with its environment andresulting in a congurative whole but with no previous design by eithera single actor or a collectivity of actors As suggested above actors do not recreate their world anew At any moment in history the totalinstitutional complex of a society necessarily contains the resources and legacies of its past Preexisting institutional complexes are nevercompletely wiped out Societies consist of multiple layers of history withtheir diverse logics of action However the cumulative effect of smallperipheral changes in altering particular institutional sectors and insti-tutional arrangements can be substantial (Murmann 1998 Lanzara 1998)lsquoInstitution building is affected more by the ways in which people havecodied the past than by how they envision the future There are manymore potential resources in pre-existing institutional arrangements thanit is usually assumed or suspectedrsquo (Lanzara 1998 30)

Fourth level of analysis

The next level of institutional analysis organizational structures is some-what more controversial As suggested above Douglass North (1990)draws a sharp distinction between institutions and organizations Morerecently many organizational theorists (Powell 1991 Powell andDiMaggio 1991 DiMaggio and Powell 1983 Baum and Oliver 1992Kondra and Hinings 1998 Townley 1997) argue that institutional rulesnorms and conventions unfold in tandem with organizational structuresand this is my positionThe literature which focuses on how institutions inuence organiza-

tions is quite different from two other theoretical literatures which alsoare concerned with how organizational environments shape the behaviorof organizations There are resource dependency theorists (Pfeffer andSalancik 1978) who emphasize the role of environmental resources inshaping organizations while Hannan and Freeman (1977 1984) and otherpopulation ecologists emphasize the survival of organizations givencertain kinds of organizational conditions (Orru et al 1991)The institutionalist perspective as implied above emphasizes the

normative environment in which organizations are embedded It is a perspective which focuses on the way in which organizations in their behavior tend to conform to the institutional rules and normswhich are dominant in the organizational environment However allthree perspectives ndash the institutionalist the population ecology and

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

619

the resource dependency ndash emphasize how the environment inuencesorganizations and how organizations which are subject to the sameenvironment tend to converge in their behavior to have what DiMaggioand Powell have labeled organizational isomorphism (DiMaggio andPowell 1983)There has emerged a vast literature which demonstrates that within

each society there are scal political judicial and other regulatory normswhich limit and shape the culture and structure of organizationalbehavior The normative institutional environment of organizations limits the options of what organizations do in a particular society and inuences the patterns of ownership relations with suppliers andcustomers In short it is the normative environment of organizationswhich denes within a particular society what is socially acceptablebehavior for organizations (Hamilton and Biggart 1988 Hollingsworthand Hollingsworth 2000 Meyer and Rowan 1991 Zucker 1987 1988Orru et al 1991 Townley 1997)Thus far most of the analyses on isomorphic pressures operating on

organizations have been cross-sectional in nature However the entireconcept of isomorphism implies that there are strong environmental pres-sures exerted on organizations and in order to observe this phenomenonit is important that we have longitudinal studies which assess howchanges in an organizationrsquos institutional environment inuence changesin the structure and culture of organizations Moreover the expectationhere is that when one analyzes the patterns of organizational changeover time the isomorphism will be holistic ndash that is within a particularsociety there will be a tendency for the internal structure of particulartypes of organizations to converge in large part because of the pres-sures to conform to the changes in the external norms rules and valuesA close reading of Aoki (1990) and Slack and Hinings (1994) nds thatchanges in the national institutional environment of organizations haveinuenced changes in the internal structure of organizations within aspecic industry of a particular societyHowever it is inappropriate to view organizations as changing only

in relation to exogenous environmental change There is a kind of co-evolution which occurs between organizations and their institutionalenvironments Co-evolution implies nonlinear feedback between orga-nizations and their institutional environments This is of course quitedifferent from the more common way of studying organizations bymodeling relations between independent and dependent variables Ofcourse the more common type of analysis is quite appropriate whenthere are simple relationships which do not involve complex feedbackprocesses But in the strategy for institutional analysis which is proposedherein it is less useful to separate independent from dependent vari-ables and more useful to understand the interacting and co-evolutionary

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

620

processes among all of the levels of institutional analysis discussed inthe various sections above (Baum and Singh 1994a 379ndash80 Mowery andNelson 1999 Murmann 1998)The logic of the perspective herein suggests that within every

society there is variation in the structure and culture of business rmsuniversities and other complex organizations (Kondra and Hinings1998) However this variation takes place within parameters which aresystem specic For example every German rm and every Germanuniversity is different from every other German university and rm but there are a set of cultural and structural characteristics which distin-guish German rms and universities from those in America Similarlyeach Japanese rm and university is unique yet they have a set ofcultural and structural characteristics which are system specic andwhich differentiate them from business rms and universities in other countriesIn all societies each organization has its own distinctive organizational

rules norms and conventions which are subordinate to the meta norms and rules of the larger society within which they are embeddedHowever the strength of the institutional environment within whichorganizations are embedded varies from society to society Some societies have multiple institutional environments and there is hetero-geneity in terms of what constitutes appropriate behavior oforganizations In such societies there can be conict over whatinstitutional logics should regulate specic organizational functions(Friedland and Alford 1991 Townley 1997 262ndash3) In those societies inwhich the institutional norms habits and rules are most developed and in which the institutional pressures to conform are greatest thereis less variation in the structure and culture of business rms and variouskinds of research organizations In such societies the connectednessbetween research organizations and their institutional environments is sufciently strong for organizations to have low autonomy to pursue independent strategies and goals in these societies there is a great deal of organizational isomorphism Conversely the weaker the institutional environment in which research organizations areembedded the greater the variation in the structure and culture of business rms and research organizations Moreover where the institu-tional environments are more weakly developed organizations havegreater autonomy and exibility to respond to the development of new knowledge and to be highly innovative Hence it is in those societies where the institutional environments are most developed and most rigid and when there is less organizational autonomy and exibility that fewer radical innovations in basic and applied science as well as in totally new products and industrial sectors havebeen made

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

621

Fifth level of analysis

We come next to the outputs and performance of the various institutionalcomponents of a society It is at this level that institutional componentsare more pragmatic and exible For example in the legal sector thereare specic statutes and court rulings in the state sector there are specicpolicies in the business sector there are new products new technolo-gies and market strategies It is at this nal level of analysis thatinstitutional spheres are most open and susceptible to change and cross-national mimicry is easiest and most commonThrough the outputs of the society we can obtain some assessment as

to how well a society is performing Moreover we can assess how inno-vative it is how egalitarian it is in the distribution of its resources howegalitarian it is in terms of levels of health education etc Just as it iscomplicated to measure how egalitarian a society is similarly it is dif-cult to assess how innovative it is For example societies vary in makingincremental and radical breakthroughs in basic and applied science in developing totally new products and new kinds of organizations inincremental and process innovations in existing products and organiza-tions in developing new and different forms of marketing ndash bothdomestically and globally (Hage and Hollingsworth 2000) Of coursesocieties vary in their rates of economic growth in their rates of economicproductivity in their quality of life (eg rates of crime life expectancyetc) All of these performance criteria feed back and inuence each ofthe levels of institutional analysis discussed above ndash rules norms valuesetc institutional arrangements institutional sectors the structure andform of organizations Moreover performance measures may inuenceeach other For example if a society has very low rates of growth itmay not be very innovative in some of the types of innovative activitiesmentioned above Good or poor performances of certain kinds inuenceother performance indicators (Hollingsworth et al 1990)Different institutional arrangements and different social systems of

production result in different types of economic performance Hence solong as societies have different social systems of production there areserious constraints on the degree to which they can converge in theirinnovative styles Different social systems of production tend to maxi-mize in a more or less explicit manner different performance criteriausually mixing considerations about static and dynamic efciency protsecurity social peace and economic andor political power In short incontrast to the implications of neoclassical economic theory in real-worldeconomies there are no universal standards all economically rationalactors attempt to maximize Economic history provides numerous exam-ples of how a great variety of principles of rationality are implementedin different societies

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

622

A critical question in institutional analysis is whether an existing socialsystem of production which supports a set of routines for a particularkind of technology and industry can shift from old practices to newones There are numerous historical accounts of how the social systemof production in a particular society which worked so well for a numberof years could not adapt to new technologies (Lazonick 1990 Veblen1915 Nelson 1994) Schumpeter (1983) and Freeman (1991) have devel-oped the idea of a techno-economic paradigm which suggests thatdifferent technologies require specic forms of organizational arrange-ments ndash or what I called social systems of production Because oflsquoinstitutional inertiarsquo the social system of production of a society maynot be able to adapt to a new techno-economic paradigm (Nelson 199458 Lanzara 1998)Whether a social system of production can sustain its performance

standards depends not only on its intrinsic economic rationality but alsoon where it ts into a larger system If a particular social system ofproduction is immune from the innovativeness and competition of analternative system survival can be long lasting But if different socialsystems of production with diverging criteria of good economic perfor-mance meet in the world arena the arbitrariness of nationally imposedconstructed performance standards may be superseded by alternativeperformance criteria as a result of international competitivenessThe world economy is also socially constructed just as are national

economies Even if different social systems of production are competingin the international arena it is not always possible to determine whichis more competitively effective at any moment in time Hegemonicnation-states in the short run shape the rules of trade to favor their indus-trial sectors and rms But the history of hegemonic powers suggeststhat in the longer run such social systems of production sustainedlargely by military and political power eventually give way to moreinnovative and competitive social systems of production In our ownday as nation-states are increasingly integrated into a world economyeven hegemonic powers lose their innovative and competitive advan-tage and their share of world output decreases Such a country mayattempt to restructure its institutional arrangements and to readjust itsperformance preferences But to restructure its system of innovationgenerally calls for a major redistribution of power within a societyLargely for this reason societies have historically had limited capacityto reconstruct a system of innovation in the image of their major competi-tors Each countryrsquos social system of production is a conguration of ahost of norms rules and values as well as of institutional arrangementsEach system is constantly changing and is open to inuence from othersystems And indeed many technologies and practices diffuse from onesociety to another but a societyrsquos capacity to be innovative is constrained

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

623

by the existing social system of production Thus the same technologymay exist in numerous countries but how it is employed and how itinuences societal outcomes and performances varies from one institu-tional conguration or society to another (Hollingsworth 1998)

PERSPECTIVES ON INSTITUTIONAL CHANGE

A fundamental problem which remains unresolved in institutionalanalysis involves the nature of institutional change There is much confu-sion and miscommunication in the social sciences about institutionalchange in large part because we do not even know how to measure therate of institutional change Of course we know that there are bothexternal and internal focuses for change and we have had a good bitof scholarship on this subject What is not clearly understood is thatinstitutions and institutional arrangements within societies are histori-cally rooted that there is a great deal of path-dependency to the waythat institutions evolve Moreover the more intricately linked each sectoris with each other sector and with a societyrsquos rules and norms the lesschoice actors have to devise new institutions and institutional arrange-ments On the other hand if a sector or actor is too isolated it may betoo weak within a society to bring about any kind of effective changeThese considerations are very critical to the problem of how muchfreedom actors have to create new institutions afresh to what extentinstitutions diffuse from one society to another (Murmann 1998 Lanzara1998) This is a fundamental question for students of institutionalanalysis as we attempt to understand how the different institutionallevels are linked togetherThere are a variety of reasons why there is confusion in studying insti-

tutional change There is disagreement over the extent to which actorshave the freedom to build new institutional arrangements and if theydo to what extent the new arrangements and organizational patternsmay depart from past practices Among some institutional economists(Williamson 1985) there is a basic assumption that actors build efcientinstitutions by pursuing their self-interest and by promoting theirstrategic goals Moreover there is the assumption that a mixing of fullrationality with market competition tends to produce all of the optimalinstitutions that are needed in order to coordinate a complex capitalisteconomy (Hollingsworth and Boyer 1997 ch 14 Hodgson 1999) Incontrast the argument developed here is much more complexSocial institutions are historically rooted and there is a great deal of

path-dependency in the way that various institutional componentsevolve The shape of institutional congurations at any moment limitsthe type of options for change Because there is a great deal of institu-tional inertia radical change in the institutional components of a society

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

624

is uncommon (Lanzara 1998) Recognizing the constraints existinginstitutions exert on actors we note that existing congurations of insti-tutional arrangements institutional sectors and organizational propertiescan limit but also inuence the degree to which norms and rules canchange Many scholars (North 1990 Ostrom 1990) have emphasized therole of norms rules ideologies and values in limiting the ability of actorsto develop new institutional arrangements which radically depart fromexisting ones However there is a two-way process of interaction aprocess of co-evolution (Baum and Singh 1994a Nelson 1994) Thus asFriedland and Alford (1991 244) argue actors develop new institutionalarrangements by recognizing organizational failures and low perfor-mance but these new arrangements do not necessarily change a societyrsquosnorms rules habits and its underlying value systemCampbell (1997) advances our understanding of how existing institu-

tions enable actors to construct radically new institutions He recognizesthat most institutions and institutional arrangements embody a degreeof inertia from which it is difcult for actors to depart Moreover thedifferential power relations among actors make it unlikely that lesspowerful actors can change the existing power structure of a societyMost change in the institutional components of a society evolves througha process of constrained selection which reects to a considerable degreethe existing arrangements and power relations This becomes the inter-pretative frame for social actors with institutionalized scripts and ritualswhich tend to be taken for granted and appear to be quite ordinary andnatural This institutionalized scripting of the social world is an extremelyimportant source of social stability and inertiaCampbell has developed a very fruitful explanation of change in terms

of the various institutional components outlined here He distinguishesbetween incremental and radical institutional change and argues thatradical institutional change occurs when social actors with widelydiffering norms cultural scripts and rituals for action engage in intenseinteraction with each other (see also Knight 1992) The more diverse theinteractions the greater the potential for institutional change He arguesthat as shifts in the composition of interaction among social actors occurchanges in the interpretation of problems and interests will follow Ifthere are only minor changes in the extent to which diverse actorsinteract there will be only modest institutional change But the morefundamental the changes in the interaction of diverse social actors thegreater the change in the way that actors interpret their world And thegreater the change in the way that actors interpret their world the morelikely that radical changes in various institutional components willfollow In other words diverse decision making tends to create a widerrange of interpretative frameworks of the social world than is likely tobe the case if social actors operate in isolation

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

625

This perspective of diverse interactions among social actors facilitatesour ability to move beyond the traditional view that existing institutionsconstrain the range of institutional alternatives which actors face Theolder view has tended to suggest that actors engage in institutionalchange by extending existing institutional principles habits and conven-tions Specifying the conditions which facilitate fundamental institutionalchange permits a much richer perspective for institutional analysisCampbellrsquos perspective about the conditions under which fundamentalor radical change in institutional components occurs is consistent withmy own research about the conditions under which major discoveriesor fundamental new knowledge occur the more scholars with diversebackgrounds interact with intensity and frequency the greater the like-lihood that they will develop new and alternative ways of thinking(Hollingsworth and Hollingsworth 2000)These views of institutional change are also consistent with the recent

work of Nelson and his colleagues (Mowery and Nelson 1999 Murmann1998) They report the research on a variety of industrial sectors in anumber of different countries revealing that networks of actors in partic-ular industries business and professional associations universityresearch departments and institutes are frequently able to use their collec-tive power to modify a societyrsquos property rights research system andgovernmental policies It is the intense and frequent interaction of thesediverse complex sets of actors ndash sometimes constituting a technologysystem ndash that often leads to incremental changes in a societyrsquos total socialsystem of production It is this kind of collective action that brings aboutorganizational change change in institutional sectors institutionalarrangements and even in the rules and norms of a society And if weare to understand how the style of innovativeness of a society changeswe must rst understand the changes in its institutional makeup It ishighly unlikely that a society will develop a new style of innovativenesswithout changing its institutional structure

FIRMS THEIR INSTITUTIONAL ENVIRONMENT ANDNATIONAL SYSTEMS OF INNOVATION

Variation in the innovative style of societies is shaped by their insti-tutional makeup ndash the various levels which are discussed aboveinstitutions institutional arrangements the structure and cohesion of institutional sectors which constitute a societyrsquos social system of pro-duction and the structure and culture of its organizations ndash especiallyits business rms and other research organizations By undertaking aninstitutional analysis of a society one can begin to understand in whatkinds of organizations the production of specic kinds of knowledgetakes place and how this is linked to particular kinds of innovativeness

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

626

With such a perspective one can gain a rich understanding of why somesocieties excel in the production of radical breakthroughs in basic andapplied science and in developing radically new products and why othersocieties excel in more incremental innovationsReecting on these issues one cannot help being impressed with the

contrasts between the innovative styles of German and American orga-nizations For example Germany has continued to be very successful in making incremental innovations in many industries in which it wasvery competitive well before World War II paper printing materialsmachinery electrotechnical products motor vehicles chemicals textileyarns and fabrics But the Germans have been much less successful indeveloping totally new products or related innovations in biotechnologyelectronics telecommunications aircraft as well as other newer indus-trial areas In contrast the Americans have been very innovative in these and other industries with products having very short lives andwith technologies which change very rapidly and are very complex(Hollingsworth 1991 Soskice 1997 1999)The importance of institutions and innovativeness for international

competitiveness has led to the recognition that there are national systemsof innovation (Lundvall 1992 Nelson 1993 Edquist 1997 Hage andHollingsworth 2000) But even before the recent studies of nationalsystems of innovation Landes (1969) was a proponent of these viewswhen he made the argument that Germany had developed a nationalsystem of innovation by building on its education and scientic researchsystems which were lacking in Britain and elsewhere However the concept of a national system of innovation immediately poses theproblem of what kind of innovations should be consideredDavid Soskice an economist at the Wissenschaftszentrum Berlin has

addressed a set of problems somewhat complementary to the perspec-tive raised here His work (Soskice 1990 1997 1999) suggests that weare sensitive to the institutional environment in which the followingactors are embedded the employees of rms the owners and nanciersof rms the competitors of rms and nally their collaborators ndash aperspective complementary to Hollingsworthrsquos work on social systemsof production (Hollingsworth 1997 1998) When one reects on the insti-tutional contextsocial system of production within which these actorsare embedded one becomes aware that those countries which have some-what inexible labor markets external to rms but a business system in which rms have long-term commitments to their employees arestrongly associated with incremental but hardly at all with radical inno-vations In the rms of such countries (eg Germany and Japan) thereis much more consensus decision making than in societies where a lsquohireand rersquo set of practices is quite pervasive And this kind of consensusdecision making ndash while highly conducive to incremental innovativeness

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

627

ndash limits the capacity for radical innovations to occur This kind of lsquolongtermismrsquo set of norms and rules occurs not only in the employmentpractices of incrementally innovative rms but also in their nancingand ownership patterns (Hollingsworth 1997) During the past thirtyyears these were characteristics of many rms in Germany and JapanSignicantly in both of these countries venture capital markets werenot highly developed nancing tended to be based more on long-termbank loans than on the equity markets and the relationships of rmstended to be relatively stable And these stable institutional arrange-ments also contributed to a high degree of innovativeness ndash butincremental rather than radical innovationsMeantime we should also pursue Soskicersquos ideas (1999) about how

rmsrsquo relationships with their competitors and collaborators shape theirstyles of innovativeness In those advanced industrial societies in whichbusiness rms are highly mobilized into business associations theretends to be a high degree of standardized quality controls over technicalnorms and a high degree of reliance on legal sanctions as monitoringinstruments vis-agrave-vis association members As a result of this high degreeof integration of rms into associations member rms are more likelyto be associated with incremental innovations than is likely to be the case with rms in those societies where associations are less welldeveloped In the latter type of society radical innovations are morelikely to occur In those advanced industrial societies where rms havelong-term stable relationships with collaborators in product design andproduction the style of innovation is likely to be incremental And it is in those societies in which there is a weaker tradition of strongcollaboration among rms that radical innovations are more likely tooccurObviously even in a country where there may be a great deal of

radical innovativeness most innovations will be of an incremental typefor there are simply fewer radical innovations But Germany and therest of Europe not only have had relatively few radical innovations incomparison with the United States but they also have lagged behindthe USA in developing relatively totally new industries Thus there hasbeen much less innovative activity in Europe in high-technology andnewly emerging industries than in the USAIn complementary research Audretsch and Feldman (1995) have

demonstrated that the USA has been more successful in spawning rmsengaged in the early stages of the industry life cycle but relative toGermany had an environment unfavorable to innovative activity byestablished rms especially in more mature industries German rmshave tended to be associated with a routinized technological regime ndashwhere established enterprises have had strong incentives to engage inincremental innovations but new rms have tended to have fundamental

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

628

innovative disadvantages relative to the Americans (see Audretsch 1995Winter 1982 Casper 2000)On both sides of the Atlantic universities engage in basic and applied

research But signicantly the institutional environment in which univer-sities are embedded inuences the degree to which the research is likelyto represent major breakthroughs or is likely to have more modestimplications Recent research in biomedical science demonstrates that inthose societies in which university professors andor graduate studentshave the following characteristics there are likely to be relatively majorresearch breakthroughs graduate students have a long training periodhigh dependency on a major professor low autonomy in dening thenature of a research project restricted job options and a university envi-ronment where there is a very modest or nonexistent entrepreneurialculture For example in Germany ndash a country with relatively few radicalbreakthroughs ndash the scholar who nishes a habilitation will have spentmany years on a research project before qualifying for a permanent acad-emic position will have been in a high-dependency position with a seniorprofessor will have had limited autonomy to dene a research projectand will be constrained from pursuing a high-risk research project Theyoung German will have an extraordinarily high level of training butwill have faced a number of disincentives to undertake high-risk researchAnd as soon as heshe has completed the habilitation and becomes aprofessor administrative responsibilities constitute another disincentivefrom engaging in high-risk or long-term research projects Moreover theGerman professor is a civil servant with a more stable and secure incomethan hisher American counterpart In a very different institutionalenvironment American professors in large research universities tend tobecome much more entrepreneurial Today many American professorsbehave as though they were operating their own rms within univer-sities and thus devote considerable time to generating funds for researchprojects assistants secretaries supplies etc And the variation in thisentrepreneurial spirit across countries is associated with variation in thetaking of high-risk research strategies and the making of major researchbreakthroughsIn sum societies vary in terms of the institutional environment in

which their organizations are embedded and this variation inuencesthe degree to which they have industrial activities in the early or latestages of the industry life cycle successful new or mature industriesand radical vs incremental innovations And it is this variation whichis encouraging rms in countries with relatively few new industries and with few or no radical innovations to invest in those countries with many new industries and numerous radical innovations ndash as astrategy to be able to transfer new technologies from the latter institu-tional environment to the former

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

629

Table 4 summarizes the way that the norms rules and values of a society are associated with Americarsquos and Germanyrsquos institutionalarrangements their social system of production and their styles of inno-vativeness

CONCLUSION

Several themes deserve to be reemphasized Most importantly institu-tional analysis is emerging as a dominant issue in several social sciencedisciplines The study of various institutional components discussedabove is a complex subject requiring multiple levels of analysis and aknowledge of multiple sectors within societies It requires a theoreticalperspective that evolves from an empirical analysis in different societiesIf we are to advance the study of institutional analysis we need cross-disciplinary collaboration and team research which are not readilyachieved given the structure of most of our research universitiesAt present scholars in each academic discipline have their own distinc-

tive strategies for studying institutional analysis Thus economists tendto use several approaches (North 1990 Williamson 1985 Hodgson 19881998 1999) anthropologists another (Geertz 1995) and sociologistspolitical scientists and historians employ other strategies (Campbell etal 1991 Powell and DiMaggio 1991 Katznelson 1998 Hall and Taylor1996) With practitioners in various disciplines pursuing their distinctiveapproaches to institutional analysis there is low potential for collabo-rative institutional analysis across the social sciences The lessons fromthe biological sciences over the last 40 years are very instructive oncepractitioners from various biological disciplines began to cooperate instudying biological phenomena by using the same concepts at the mole-cular level theoretical advance was very rapid Comparable phenomenatook place in the development of other hybrid elds eg biochemistrybiophysics (Judson 1979 Olby 1979 Kohler 1982 1994) Similarly weresocial scientists in various disciplines to work collaboratively in the areaof institutional analysis there would very likely be an acceleration intheir theoretical knowledgeThis is not a plea that everyone engaged in institutional analysis should

do the same kind of research Indeed just as biologists working at themolecular level study many different kinds of problems similarly socialscientists engaged in institutional analysis would also work on manyseparate problems but in the spirit of a collective enterprise Some wouldconcentrate their attention on the study of rules norms habits conven-tions and values while others would study how these are associatedwith congurations of various institutional arrangements (types ofmarkets hierarchies networks associations communities clans statesetc) Others would work on specic institutional sectors (eg education

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

630

Table 4

Rules norms hab

its an

d values associated w

ith sp

ecic co

ngu

ration

s of coo

rdinating mecha

nism

s soc

ial sy

stem

s of

prod

uction

an

d nationa

l styles of inno

vation

USA

Germany

Norms rules values which provide incentives for

Norms rules values which provide incentives for

short-term horizons

long-term horizons

Con

gu

rative

forms of

Marke

ts c

orpo

rate hierarchies reg

ulatory state

Assoc

iation

s (busine

ss assoc

iation

s la

bor union

s)

coordinating

corp

oratist-type

networks

marke

ts c

orporatist-

mecha

nism

s (listed in

type

state

order of im

portanc

e in

the particu

lar social

system

)

Social sys

tem of

prod

uction

Bus

iness system

Firm

struc

tures

Con

glom

erates b

ut with mov

emen

t toward

Firm

more co

ngru

ent with pa

rticular prod

ucts

elim

ination of produc

tion

in som

e sectors

emph

asis on long

-term strateg

ies rm

s are well

emph

asis on sh

ort-term

strateg

ies high

integrated

differentiation

between divisions

of rms

Firm

s mob

ilized in

Low

High

and in

tegrated

into busine

ss

associations

Indu

strial relations

Low

deg

ree of jo

b secu

rity rigid in

ternal lab

or

High deg

ree of jo

b secu

rity

exible interna

l lab

or

system

marke

t ex

ible externa

l labo

r marke

tsmarke

t rigid externa

l lab

or m

arke

tPatterns of

Dispersed owne

rship frequ

ent turnov

erMore stab

le owne

rship patterns

owne

rship

Table 4

continue

d

USA

Germany

Norms rules values which provide incentives for

Norms rules values which provide incentives for

short-term horizons

long-term horizons

Social system of

Prod

uction

Bus

iness system

Training for labo

r an

dHigh deg

ree of voluntary decision mak

ing by

Collective an

d m

ore co

mpulso

ry d

ecisions

abo

ut

man

agem

ent

individuals as to ho

w m

uch

tim

e an

d ene

rgy

how m

uch ed

ucation

is ne

eded

he

avy em

pha

sis

to inv

est in educ

ation heavy

emph

asis on

on techn

ical training (eg en

gine

ering)

busine

ss sch

ools training

in marke

ting

and

sales accou

nting

Fina

ncial marke

tsEqu

ity an

d ven

ture cap

ital m

arke

ts highly

Equ

ity an

d ven

ture cap

ital m

arke

ts w

eakly

dev

elop

ed na

ncing from

cen

tral gov

ernm

ent

deve

lope

d he

avy em

pha

sis on

corporate loan

sto priva

te rm

s for dev

elop

men

t of new

tech

nologies related

to na

tion

al security

Unive

rsity research

High individual auton

omy high poten

tial for

Low

ind

ividua

l au

tono

my until almost 40

yea

rs of

system

individual creativity u

nive

rsity pr

ofesso

rs

age poten

tial for ind

ividual creativity high

ly

high

ly entreprene

urial

cons

traine

d unive

rsity pr

ofessors not so

entrep

rene

urial

Inno

vative

styles

Highly su

ccessful in m

ajor break

through

s in

High pe

rforman

ce in proc

ess an

d inc

remen

tal

basic an

d app

lied scien

ce go

od in dev

elop

ing

inno

vation

s in existing produ

cts not so su

ccessful

totally

new

lines of pr

oducts an

d in m

ajor break

through

s in basic scien

ce or in

tech

nologies high

deg

ree of entrepr

eneu

rship

deve

loping

totally new

lines of prod

ucts of

amon

g research

ers

tech

nologies low d

egree of entrepr

eneu

rship

amon

g research

ers

Table 4

continue

d

USA

Germany

Norms rules values which provide incentives for

Norms rules values which provide incentives for

short-term horizons

long-term horizons

Prod

ucts in w

hich

Pr

oducts with sh

ort liv

es prod

ucts in w

hich

Older produ

cts with long

history machine

ry pap

er

society excels

tech

nology

is bo

th highly co

mplex

and

pr

inting

textile

s m

otor veh

icles m

achine

too

lsch

ange

s rapidly en

tertainm

ent indus

try

pha

rmaceu

ticals biotechn

olog

y softw

are

aerosp

ace

business systems nancial markets and systems of research) Distinctivewould be the recognition of how these separate studies are linkedtogether how each institutional sector is linked to a societyrsquos normsrules values etc and its conguration of institutional arrangementsSimilarly other institutional analysts might focus on the study of orga-nizations with a major concern for how the institutional environmentof organizations (eg norms rules values congurations of institutionalarrangements institutional sectors) inuenced the structure culture andoutcomes of organizations And nally others might study how all ofthe aforementioned aspects of institutional analysis inuenced a societyrsquosoverall performance Nelson and Sampat (1998) have appropriatelyreminded us that we should engage in institutional analysis in order toaddress specic problems but the view developed in this article is thatwe rst need some mapping of the terrain of institutional analysis inorder for our diverse enterprises to become a meaningful collective enter-prise especially if we are to understand how institutions inuence theinnovative processThe argument developed here is not that all social scientists should engage

in institutional analysis Obviously there are numerous other importantareas of research which lie beyond the eld of institutional analysis Nordoes this article argue that institutional analysis is the most importantsubject for social scientistsSince the structure and culture of our research universities reect an

enormous amount of disciplinary fragmentation what are the prospectsfor promoting the kind of collective and interdisciplinary research agendaproposed above In one respect the prospects are not encouragingDuring the past half century the efforts of American universities to have effective communication across the biological sciences have variedgreatly in their success At the University of California-Berkeley it was not until the 1980s that there was a serious effort to integrate thebiological sciences and at that time the campus abolished seventeendepartments to promote an integrated biological program But at theUniversity of Wisconsin the biological sciences are still very fragmentedinto dozens of different departments in botany plant pathology zoologyand physiology with two departments of biochemistry and two depart-ments of geneticsThere have been other approaches which have worked very well histor-

ically for promoting scientic integration amidst scientic diversity Forexample several summer schools and special institutes were organizedin the United States during the 1940s in an effort to overcome the frag-mentation in the biological sciences These programs made it possiblefor scientists in diverse academic disciplines and from different univer-sities and countries to come together to engage in intense and frequentinteraction Perhaps the most effective program was that held in the

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

634

summers at Cold Spring Harbor New York where a research agendafor the study of molecular biology was developed (Fischer and Lipson1988 Stent and Watson 1966) Scientists went there summer aftersummer and when they returned to their universities many began therestructuring of the biological sciences in their universitiesIn our own day we need summer institutes and other similar programs

so that scholars from diverse backgrounds can come together anddevelop a collective agenda for doing institutional analysis and forstudying the impact of institutions on the performance of societies ndashinnovativeness being one such subject If we can develop such instituteswe will increase the prospects for transforming our universities Thereis reason to be optimistic that once an eclectic group of scholars workwith a common set of concepts we can have genuine theoretical advanceif researchers can bring their diverse expertise to bear on a common setof problems There is considerable evidence that institutional analysisoffers high potential for advancing our knowledge of innovations andtechnological change (Hollingsworth and Hollingsworth 2000) If we aregoing to make substantial theoretical advance it is not likely to occurfrom the vantage point of a multiplicity of disciplines Rather it is likelyto occur as we develop a new intellectual framework and our own acad-emic journals This is a tall agenda which requires us to reassess theway that knowledge is presently produced in our societies

NOTE

This article is part of a much larger agenda involving the study of institutionsorganizations and innovations in which I am involved My debts are many andvaried The rst version was written while I was in residence in the ResearchUnit for Institutional Change and European Integration of the Austrian Academyof Sciences in Vienna in the summer of 1998 I am especially grateful to ProfessorsEgon Matzner and Sonja Puntscher-Riekmann for providing the stimulating envi-ronment in which to think through the ideas developed here Egon Matzner hasengaged in very stimulating discussions with me about the importance of insti-tutions in the general eld of socio-economics The nal version of the articlewas completed during my residence as a fellow at the Neurosciences Institutein La Jolla California I also wish to thank my colleagues in the InnovationTheme Project at the Netherlands Institute for Advanced Study (NIAS) during1998ndash9 and Edgar Grande and Karl Muumlller for stimulating conversations aboutthe interaction of institutions and organizations in facilitating particular typesof innovation I am especially grateful to Jerald Hage for helping me to developa number of issues in the article and for his useful comments on the rst draftFrans van Waarden early on helped me to work through the idea of multiplelevels of institutional analysis Others who have helped to develop the ideas inthe article have been my colleagues in the research project lsquoComparative socialsystems of productionrsquo especially Tom Burns Christel Lane Yoshitaka OkadaWolfgang Streeck and Richard Whitley My colleagues Robert Boyer SteveCasper and David Soskice have taught me much about how incentives providedby the institutional environments of rms inuence individual decision making

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

635

But my most important debt is to Ellen Jane Hollingsworth who has repeatedlyquestioned every single assumption of this article Without her rich and contin-uous tutoring this article would be even less developed than it is at presentFinally without the insightful assistance of David Gear in many phases of thisarticle it would not have been completed

REFERENCES

Aoki Masahiko (1990) lsquoTowards an economic model of the Japanese rmrsquo Journalof Economic Literature March 1ndash27

Arora A Landau Ralph and Rosenberg Nathan (eds) (1998) Chemicals and Long-Term Economic Growth Insights from the Chemical Industry New York JohnWiley amp Sons

Arrow Kenneth (1974) The Limits of Organization New York NortonArthur W B (1988a) lsquoSelf-reinforcing mechanisms in economicsrsquo in P W

Anderson K J Arrow and D Pines (eds) The Economy as an Evolving ComplexSystem Redwood City Calif Addison-Wesley

mdashmdash(1988b) lsquoCompeting technologies an overviewrsquo in G Dosi C Freeman RNelson G Silverberg and L Soete (eds) Technical Change and Economic TheoryLondon Pinter Publishers pp 590ndash607

Audretsch David B (1995) Innovation and Industry Evolution Cambridge MassMIT Press

Audretsch David B and Feldman Maryann P (1995) lsquoInnovative clusters andthe industry life cyclersquo Discussion Paper Wissenschaftszentrum BerlinMarket Processes and Corporate Development Research Unit FS-IV-95-7

Baum Joel A C and Oliver Christine (1992) lsquoInstitutional embeddedness andthe dynamics of organizational populationsrsquo American Sociological Review 57540ndash59

Baum Joel A C and Singh Jitendra V (1994a) lsquoOrganizationndashenvironment coevo-lutionrsquo in Joel A C Baum and Jitendra V Singh (eds) The EvolutionaryDynamics of Organizations New York Oxford University Press pp 379ndash402

mdashmdash(eds) (1994b) The Evolutionary Dynamics of Organizations New York OxfordUniversity Press

Berger Suzanne and Dore Ronald (eds) (1996) National Diversity and GlobalCapitalism Ithaca NY Cornell University Press

Boyer Robert (1997) lsquoThe variety and unequal performance of really existingmarkets farewell to Doctor Panglossrsquo in J Rogers Hollingsworth and RobertBoyer (eds) Contemporary Capitalism the Embeddedness of InstitutionsCambridge and New York Cambridge University Press pp 55ndash93

Burns Tom R and Dietz Thomas (2001 forthcoming) lsquoSocial rule system theorysocial action institutional arrangements and evolutionary processesrsquo in JRogers Hollingsworth and Karl Muumlller (eds) Advances in SocioeconomicsLanham Md Rowman amp Littleeld

Burns Tom R and Flam Helena (1987) The Shaping of Social Organization BeverlyHills Calif Sage

Calvert Randall L (1995) lsquoThe rational choice theory of social institutionsrsquo inJeffrey S Banks and Eric A Hanuskek (eds) Modern Political Economy NewYork Cambridge University Press pp 216ndash66

Camic Charles (1986) lsquoThe matter of habitrsquo American Journal of Sociology 911039ndash87

Campbell John L (1997) lsquoMechanisms of evolutionary change in economicsgovernance interaction interpretation and bricolagersquo in Lars Magnusson

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

636

Table 3

continue

d

Coordinating mechanisms

Type of failure

Associations

Private hierarchies

State

Market

Communities

Networks

Efcien

cyFa

cilitates

De

cien

t in

Can

be high

ly

Some ba

sic social So

me go

ods

Slow

to en

hanc

e co

operation an

d

coop

eration an

d

bureau

cratic

relation

s cann

ot

cann

ot be

efc

ienc

y an

d

X-efcien

cy but

X-efcien

cyan

d can

not

be provided

by

delivered

at

speed of

not allocative

easily deliver

pure m

arke

t su

fcien

tly

adap

tive

ness

efcien

cy

good

s at

mecha

nism

slow costs

except in

low cost

indus

tries whe

re

tech

nology

is

complex an

d rapidly cha

nging

Equity

Narrow

Excessive

Might enh

ance

Facilitates

Might lead to

Whe

n widely

encompa

ssing

multiplication

ineq

uality

ineq

uality in

retard

ed

dev

elop

ed in

to

associationa

l of con

trollers

(pow

er

income an

d dev

elop

men

tindus

trial

stru

ctures lead

(fru

stration

and

an

d privilege

)wealth

districts

to in

come

ineq

uality)

netw

orks m

ay

equa

lity

facilitate greater

equality an

d

inco

me

distribution

whe

n wea

kly

dev

elop

ed

netw

orks ten

d to

increase soc

ial

ineq

uality

Note For allo

cativ

e efc

ienc

y an

d X

-efcien

cy see Leibe

nstein (19

66 19

76)

they raise other no less severe problems It is important to recognizethe imperfection of any single institutional arrangement in order tocomprehend the origin and transformation of any other institutionalarrangement (Hodgson 1999 ch 3)As suggested above each institutional arrangement is congured with

other institutional arrangements Usually one particular institutionalarrangement is more dominant in a particular conguration but becauseeach type of institutional arrangement has its own strengths andweaknesses there is no simple structural logic in the governance or coor-dination of a society Each institutional arrangement constrains the otherbut the inherent tension among the various institutional arrangementswithin a conguration contributes to changes in the governance orcoordination of a society The routines and logics of each of these cong-urations of institutional arrangements provide constraints and incentivesfor actors The way that actors perceive the incentives and constraintsof these governance congurations leads to a particular market logicand it is the specic market logic of a society which inuences its speciccapacities and weaknesses The inherent strain among the different logicsin a conguration of coordination helps to provide the exibility for asociety to adapt to new circumstancesIf there were a society with pure markets or a society coordinated

only by the state there would be too much rigidity and too little diver-sity to cope with the vast uncertainty in the global environment A societywith very little diversity in its coordination mechanisms would havelittle capacity to adapt to new circumstances Congurations with consid-erable diversity of institutional arrangements provide for a certainamount of incoherence in governance but they also provide for thecapacity to adapt to new circumstances While the Soviet Union wasdominated by the state as a coordinating mechanism there were alwaysfunctioning markets in the system Moreover the feudal society of thetwelfth and thirteenth centuries consisted not only of hierarchical rela-tionships between serfs and masters but also of urban guilds clericalhierarchies and markets (Johnson 1992 38) In sum the robustness ofinstitutions often depends on multiple and diverse principles and logicsof action on the inconsistency of principles and procedures on patternedforms of disorder (Lanzara 1998 Orren and Skowronek 1991 320 329)

Third level of analysis

The next level consists of the institutional sectors of a society The rulesnorms and values of a society inuence the array of institutional arrange-ments and levels one and two inuence the nature of and the relation-ships among various institutional sectors and all three levels areintricately linked together to form a social system of production Together

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

613

all of these components inuence the performance of economic sectorswithin a society as well as the performance of the total society A socialsystem of production is the way that a societyrsquos institutions (see rstlevel of analysis above) its institutional arrangements (see second levelof analysis above) and its institutional sectors are integrated into a socialcongurationAn institutional sector includes all organizations in a society which

supply a given service or product along with their associated focal orga-nizations (eg major suppliers funding sources regulators and so forth(Scott et al 1994 108 117 Campbell et al 1991 Hollingsworth et al1994) Institutional sectors include but are not limited to the societyrsquossystem of education system of research business system nancialmarkets legal system and the state The structure of the nancial marketssystem of training and education industrial relations system system ofresearch and state and legal systems are distinct and idiosyncratic ineach society In short they are system specicAll of these institutional sectors tend to cohere with each other

although they vary in the degree to which they are tightly coupled intoa full-edged system While each of the institutional sectors has someautonomy and may have some goals that are contradictory to the goalsof other institutional sectors with which it is linked an institutional logicin each society leads institutions to coalesce into a complex social con-guration (Hollingsworth 1991) This occurs because the institutionalsectors are embedded in a culture in which their logics are symbolicallygrounded organizationally structured technically and materially con-strained and politically defended The conguration of institutionalsectors usually exhibits some degree of adaptability to new challengesbut continues to evolve within an existing style But under new circum-stances or unprecedented disturbances these institutional congurationsare exposed to sharp historical limits as to what they may or may notdo (Schumpeter 1983 David 1988 Arthur 1988a 1988b Haringkanssen andLundgren 1997)Why do all of these different institutions coalesce into a complex

social conguration which is labeled here as a social system of pro-duction The literature suggests two contrasting interpretations Part ofthe answer ndash indeed a controversial one ndash is that these institutionalsectors are functionally determined by the requirements of the practiceof capitalism in each time and place (Habermas 1975) Another expla-nation emphasizes the genesis of the actual conguration via a trial and error process according to which the survival of rms regions orcountries is the outcome of complex evolutionary mechanisms (Maynard-Smith 1982 Nelson and Winter 1982) However the problem is evenmore complex Markets and other mechanisms for coordinating rela-tionships among economic actors place constraints on the means and

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

614

ends of economic activity to be achieved in any society The othercoordinating mechanisms include different kinds of hierarchies varioustypes of networks and associations ndash trade unions employers and busi-ness associations (Hollingsworth and Lindberg 1985 Campbell et al1991) The logic of these various institutional arrangements providesactors with vocabularies and logics for pursuing their goals for deningwhat is valued and for shaping the norms and rules by which theyabide (see Tables 2 and 3) In short in contrast to the logic of theneoclassical paradigm the argument here is that the dominant type ofinstitutional arrangements places severe constraints on the denition of needs preferences and choices of economic actors Whereas the neo-classical paradigm assumes that individuals and rms are sovereign thisarticle is based on the assumption that rms are inuenced by the hold that the institutional congurations making up a social system ofproduction have on individual decision making (Campbell et al 1991Etzioni 1988 Streeck and Schmitter 1985a Hollingsworth et al 1994Hollingsworth and Boyer 1997 Magnusson and Ottosson 1997 North1990 Hodgson 1999)Standard neoclassical economic theory has tended to downplay the

role of production and consequently of rms Even transaction cost theo-rists who are concerned with analyzing the rm as a coordinatingmechanism have been relatively unconcerned with the various compo-nents of a social system of production Indeed as long as there waswidespread optimism about the efcacy of Keynesian economics therewas relatively little concern among neoclassical economists with thesupply side of the economy Even in the opinion of most Keynesians agroup of experts should ideally be able to shape the size of aggregatedemand while the supply side of the economy would be left to the twominimalist institutions of neoclassical economics ndash markets and manage-rial hierarchies For several decades however it has become increasinglyobvious that some of the most competitive and successful patterns ofindustrial output and industrial production in capitalist economies donot derive from the neoclassical prescription of unregulated markets andcorporate hierarchies complemented by a neoliberal democratic stateIndeed it is now well understood that certain highly successful produc-tion patterns require for their emergence and survival institutionalarrangements and environments the very opposite of the prescriptionsfound in the neoclassical paradigm (see especially Streeck 1992 but alsoHollingsworth and Streeck 1994) Thus social scientists have come tounderstand that if they are to comprehend the behavior and performanceof contemporary economies concerns about social systems of produc-tion must be brought into the pictureBecause production involves more than technology a number of social

scientists have an increasing concern with social systems of production

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

615

The same equipment is frequently operated quite differently in the samesectors in different countries even when rms are competing in the same market (Maurice et al 1980 Sorge and Streeck 1988 Sorge 1989Hollingsworth et al 1994) Variations in production and process tech-nologies are inuenced partly by variations in the social environmentsin which they are embedded In other words rms are embedded intocomplex environments which among other things place constraints ontheir behavior Thus a social system of production is of major impor-tance in understanding the behavior and performance of an economyHow the state and other institutional arrangements (eg marketsnetworks private hierarchies associations) coalesce and are related toparticular social systems of production are important determinants ofeconomic performanceThe congurations of institutional arrangements that coordinate or

govern the behavior of actors in one society and its structure of specicinstitutional sectors cannot easily be transferred to another society forthey are embedded into a social system of production that is societallydistinct (Hollingsworth 1997) Societies borrow selected principles offoreign management styles and work practices but the effectiveness ofsuch borrowing is generally limited Economic behavior and performanceare shaped by the entire social system of production in which actors areembedded and not simply by specic principles of particular manage-ment styles and work practices Moreover a societyrsquos social system ofproduction tends to limit the kind of goods that it can produce and withwhich it can compete successfully in international marketsSome scholars (Kenney and Florida 1993 Oliver and Wilkinson 1988)

have assumed that the diffusion of particular forms of management stylesand work practices across societies could lead to system convergenceReferring to the Japanese production system Kenney and Florida haveargued that it consists of organizational practices whose fundamentallsquogenetic codersquo can successfully be inserted into another society and beginsuccessfully to reproduce its behavior in the new environment (1993 8)Their position is in the intellectual tradition of Antonio Gramsci (1971)who contended decades ago that the American system of mass produc-tion would most certainly diffuse to Europe over timeHowever the argument of this article is that even though British

French and American rms may adopt certain aspects of Japanesemanagement styles (eg just-in-time production self-managing teamsquality circles the use of statistical process controls) or some variant ofthe German vocational system their social systems of production willnot converge Moreover the overall conguration of a societyrsquos socialsystem of production inuences its sectoral and overall national economicperformance This helps to explain why societies have different systemsof innovation and why they vary in the clusters of industries in which

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

616

they are highly competitive in the international markets (Nelson 19931999 Hage and Hollingsworth 2000 Berger and Dore 1996)In order to understand how and why a societyrsquos economy performs

as it does it is necessary to understand its entire social system of produc-tion If a society is to modify substantially the performance of itsinstitutional sectors and its entire economy it cannot adapt only someof the management and work practices of its foreign competitors Ratherit must alter its entire social system of production Because a societyrsquosmodes of economic governance and coordination and its institutionalsectors develop according to particular logics and are system specicthere are serious limitations to the extent to which a society may mimicthe institutions (eg the rules and norms) institutional arrangements andinstitutional sectors of other societiesIn the history of modern societies there are logics by which institu-

tions coalesce into a social system of production (Hollingsworth 1991)Though institutions are constantly changing there are sharp limits tothe type and the direction of change that any particular institution canundergo because of its linkages with institutional arrangements and insti-tutional sectors Thus a societyrsquos business rms educational systemnancial markets industrial relations system etc can engage in seriousrestructuring only if most of the other institutional sectors also changeIn social systems of production there are pressures toward consistencyin the norms rules and values across institutional sectors though in anycomplex society social systems are obviously imperfectly integratedIndeed the degree to which the institutional norms and rules makingup a social system of production are loosely or tightly coupled is a vari-able of considerable importance In general the institutional sectorsmaking up a social system of production are interdependent and changesin one generally result in changes in the others Since each institutionallevel is dependent on the others for various types of resources there isinterdependence among the differing institutional spheres Moreovereach society has its norms moral principles rules and laws and recipesfor action as well as its own idiosyncratic customs traditions and prin-ciples of justice (Burns and Flam 1987)There are also other inherent obstacles to convergence among social

systems of production for where a system is at any one point in timeis inuenced by its initial state Systems with quite different initial statesare unlikely to converge with one anotherrsquos institutional practicesExisting institutional arrangements block certain institutional innova-tions and facilitate others (Roland 1990) Thus the institutions makingup a social system of production provide continuity even though insti-tutional arrangements are always changing but with a particular logicWhile Williamson (1975 1985) suggests that actors tend to choose theinstitutional arrangements which are most efcient North (1990) is

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

617

much closer to the mark in his argument that most societal institutionalarrangements exist as a result of custom and habit and are thereforeinefcient At any moment in time the world often appears to its actorsas very complex and uncertain For this reason actors often engage incontradictory forms of behavior pursuing different strategies as hedgesagainst a very uncertain world (Lanzara 1998) And their hedging andcontradictory forms of behavior may lead in somewhat different soci-etal directions all constrained by the institutional fabric within whichthe actors are embedded This kind of contradictory behavior occurs inpart because of the contradictions inherent in the dominant congura-tions of institutional arrangements with their contradictory logics (seelsquoSecond level of analysisrsquo above)Despite the emphasis on the logic of institutional continuity this is

not an argument that systems change along some predetermined pathThere are critical turning points in the history of highly industrializedsocieties but with the choices limited by the existing institutional terrainBeing path-dependent a social system of production continues along aparticular logic until or unless a fundamental societal crisis intervenes(Milgrom et al 1991 Krugman 1991 Durlauf 1991 Hollingsworth 1991Pred 1966 David 1988 Hodgson 1998)At this point it is important to confront the question of how social

systems of production evolve And this question gets to the heart of theproblem of building complex societal structures Certainly social systemsof production did not emerge from some process of social engineeringMoreover the various component parts of each social system of produc-tion have often not been designed to be part of a social system ofproduction The component parts of each social system of productionhave emerged more often than not as unintended by-products of goalswhich various actors had in mind at earlier moments in time It is usuallythe case that actors at time lsquotrsquo fail to comprehend the long-term conse-quences of their actions Wolfgang Streeck the author of several brilliantpapers on social systems of production argues that institutional sectordesigns created for one purpose generally address that goal but overtime those designs have quite unintended consequences (Streeck 1997a1997b Wright 1998) Describing the conguration of the German socialsystem of production of the 1970s and 1980s Streeck points out that itwas the unintended by-product of multiple points in history Someelements were pre-Wilhelmian others were introduced by the Alliedpowers after 1945 and others emerged during the years of the GermanFederal Republic All component parts of the German social system ofproduction

were and continue to be changing for their own reasons as wellas in reaction to each other and certainly there can be no presump-tion of a pre-established t between them even though one might

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

618

want to allow for some reinforcement effects of the modelrsquos histor-ically contingent social and economic success

(Streeck 1997b 54)

In sum the emergence of social systems of production is a long-termevolutionary process with each part interacting with its environment andresulting in a congurative whole but with no previous design by eithera single actor or a collectivity of actors As suggested above actors do not recreate their world anew At any moment in history the totalinstitutional complex of a society necessarily contains the resources and legacies of its past Preexisting institutional complexes are nevercompletely wiped out Societies consist of multiple layers of history withtheir diverse logics of action However the cumulative effect of smallperipheral changes in altering particular institutional sectors and insti-tutional arrangements can be substantial (Murmann 1998 Lanzara 1998)lsquoInstitution building is affected more by the ways in which people havecodied the past than by how they envision the future There are manymore potential resources in pre-existing institutional arrangements thanit is usually assumed or suspectedrsquo (Lanzara 1998 30)

Fourth level of analysis

The next level of institutional analysis organizational structures is some-what more controversial As suggested above Douglass North (1990)draws a sharp distinction between institutions and organizations Morerecently many organizational theorists (Powell 1991 Powell andDiMaggio 1991 DiMaggio and Powell 1983 Baum and Oliver 1992Kondra and Hinings 1998 Townley 1997) argue that institutional rulesnorms and conventions unfold in tandem with organizational structuresand this is my positionThe literature which focuses on how institutions inuence organiza-

tions is quite different from two other theoretical literatures which alsoare concerned with how organizational environments shape the behaviorof organizations There are resource dependency theorists (Pfeffer andSalancik 1978) who emphasize the role of environmental resources inshaping organizations while Hannan and Freeman (1977 1984) and otherpopulation ecologists emphasize the survival of organizations givencertain kinds of organizational conditions (Orru et al 1991)The institutionalist perspective as implied above emphasizes the

normative environment in which organizations are embedded It is a perspective which focuses on the way in which organizations in their behavior tend to conform to the institutional rules and normswhich are dominant in the organizational environment However allthree perspectives ndash the institutionalist the population ecology and

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

619

the resource dependency ndash emphasize how the environment inuencesorganizations and how organizations which are subject to the sameenvironment tend to converge in their behavior to have what DiMaggioand Powell have labeled organizational isomorphism (DiMaggio andPowell 1983)There has emerged a vast literature which demonstrates that within

each society there are scal political judicial and other regulatory normswhich limit and shape the culture and structure of organizationalbehavior The normative institutional environment of organizations limits the options of what organizations do in a particular society and inuences the patterns of ownership relations with suppliers andcustomers In short it is the normative environment of organizationswhich denes within a particular society what is socially acceptablebehavior for organizations (Hamilton and Biggart 1988 Hollingsworthand Hollingsworth 2000 Meyer and Rowan 1991 Zucker 1987 1988Orru et al 1991 Townley 1997)Thus far most of the analyses on isomorphic pressures operating on

organizations have been cross-sectional in nature However the entireconcept of isomorphism implies that there are strong environmental pres-sures exerted on organizations and in order to observe this phenomenonit is important that we have longitudinal studies which assess howchanges in an organizationrsquos institutional environment inuence changesin the structure and culture of organizations Moreover the expectationhere is that when one analyzes the patterns of organizational changeover time the isomorphism will be holistic ndash that is within a particularsociety there will be a tendency for the internal structure of particulartypes of organizations to converge in large part because of the pres-sures to conform to the changes in the external norms rules and valuesA close reading of Aoki (1990) and Slack and Hinings (1994) nds thatchanges in the national institutional environment of organizations haveinuenced changes in the internal structure of organizations within aspecic industry of a particular societyHowever it is inappropriate to view organizations as changing only

in relation to exogenous environmental change There is a kind of co-evolution which occurs between organizations and their institutionalenvironments Co-evolution implies nonlinear feedback between orga-nizations and their institutional environments This is of course quitedifferent from the more common way of studying organizations bymodeling relations between independent and dependent variables Ofcourse the more common type of analysis is quite appropriate whenthere are simple relationships which do not involve complex feedbackprocesses But in the strategy for institutional analysis which is proposedherein it is less useful to separate independent from dependent vari-ables and more useful to understand the interacting and co-evolutionary

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

620

processes among all of the levels of institutional analysis discussed inthe various sections above (Baum and Singh 1994a 379ndash80 Mowery andNelson 1999 Murmann 1998)The logic of the perspective herein suggests that within every

society there is variation in the structure and culture of business rmsuniversities and other complex organizations (Kondra and Hinings1998) However this variation takes place within parameters which aresystem specic For example every German rm and every Germanuniversity is different from every other German university and rm but there are a set of cultural and structural characteristics which distin-guish German rms and universities from those in America Similarlyeach Japanese rm and university is unique yet they have a set ofcultural and structural characteristics which are system specic andwhich differentiate them from business rms and universities in other countriesIn all societies each organization has its own distinctive organizational

rules norms and conventions which are subordinate to the meta norms and rules of the larger society within which they are embeddedHowever the strength of the institutional environment within whichorganizations are embedded varies from society to society Some societies have multiple institutional environments and there is hetero-geneity in terms of what constitutes appropriate behavior oforganizations In such societies there can be conict over whatinstitutional logics should regulate specic organizational functions(Friedland and Alford 1991 Townley 1997 262ndash3) In those societies inwhich the institutional norms habits and rules are most developed and in which the institutional pressures to conform are greatest thereis less variation in the structure and culture of business rms and variouskinds of research organizations In such societies the connectednessbetween research organizations and their institutional environments is sufciently strong for organizations to have low autonomy to pursue independent strategies and goals in these societies there is a great deal of organizational isomorphism Conversely the weaker the institutional environment in which research organizations areembedded the greater the variation in the structure and culture of business rms and research organizations Moreover where the institu-tional environments are more weakly developed organizations havegreater autonomy and exibility to respond to the development of new knowledge and to be highly innovative Hence it is in those societies where the institutional environments are most developed and most rigid and when there is less organizational autonomy and exibility that fewer radical innovations in basic and applied science as well as in totally new products and industrial sectors havebeen made

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

621

Fifth level of analysis

We come next to the outputs and performance of the various institutionalcomponents of a society It is at this level that institutional componentsare more pragmatic and exible For example in the legal sector thereare specic statutes and court rulings in the state sector there are specicpolicies in the business sector there are new products new technolo-gies and market strategies It is at this nal level of analysis thatinstitutional spheres are most open and susceptible to change and cross-national mimicry is easiest and most commonThrough the outputs of the society we can obtain some assessment as

to how well a society is performing Moreover we can assess how inno-vative it is how egalitarian it is in the distribution of its resources howegalitarian it is in terms of levels of health education etc Just as it iscomplicated to measure how egalitarian a society is similarly it is dif-cult to assess how innovative it is For example societies vary in makingincremental and radical breakthroughs in basic and applied science in developing totally new products and new kinds of organizations inincremental and process innovations in existing products and organiza-tions in developing new and different forms of marketing ndash bothdomestically and globally (Hage and Hollingsworth 2000) Of coursesocieties vary in their rates of economic growth in their rates of economicproductivity in their quality of life (eg rates of crime life expectancyetc) All of these performance criteria feed back and inuence each ofthe levels of institutional analysis discussed above ndash rules norms valuesetc institutional arrangements institutional sectors the structure andform of organizations Moreover performance measures may inuenceeach other For example if a society has very low rates of growth itmay not be very innovative in some of the types of innovative activitiesmentioned above Good or poor performances of certain kinds inuenceother performance indicators (Hollingsworth et al 1990)Different institutional arrangements and different social systems of

production result in different types of economic performance Hence solong as societies have different social systems of production there areserious constraints on the degree to which they can converge in theirinnovative styles Different social systems of production tend to maxi-mize in a more or less explicit manner different performance criteriausually mixing considerations about static and dynamic efciency protsecurity social peace and economic andor political power In short incontrast to the implications of neoclassical economic theory in real-worldeconomies there are no universal standards all economically rationalactors attempt to maximize Economic history provides numerous exam-ples of how a great variety of principles of rationality are implementedin different societies

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

622

A critical question in institutional analysis is whether an existing socialsystem of production which supports a set of routines for a particularkind of technology and industry can shift from old practices to newones There are numerous historical accounts of how the social systemof production in a particular society which worked so well for a numberof years could not adapt to new technologies (Lazonick 1990 Veblen1915 Nelson 1994) Schumpeter (1983) and Freeman (1991) have devel-oped the idea of a techno-economic paradigm which suggests thatdifferent technologies require specic forms of organizational arrange-ments ndash or what I called social systems of production Because oflsquoinstitutional inertiarsquo the social system of production of a society maynot be able to adapt to a new techno-economic paradigm (Nelson 199458 Lanzara 1998)Whether a social system of production can sustain its performance

standards depends not only on its intrinsic economic rationality but alsoon where it ts into a larger system If a particular social system ofproduction is immune from the innovativeness and competition of analternative system survival can be long lasting But if different socialsystems of production with diverging criteria of good economic perfor-mance meet in the world arena the arbitrariness of nationally imposedconstructed performance standards may be superseded by alternativeperformance criteria as a result of international competitivenessThe world economy is also socially constructed just as are national

economies Even if different social systems of production are competingin the international arena it is not always possible to determine whichis more competitively effective at any moment in time Hegemonicnation-states in the short run shape the rules of trade to favor their indus-trial sectors and rms But the history of hegemonic powers suggeststhat in the longer run such social systems of production sustainedlargely by military and political power eventually give way to moreinnovative and competitive social systems of production In our ownday as nation-states are increasingly integrated into a world economyeven hegemonic powers lose their innovative and competitive advan-tage and their share of world output decreases Such a country mayattempt to restructure its institutional arrangements and to readjust itsperformance preferences But to restructure its system of innovationgenerally calls for a major redistribution of power within a societyLargely for this reason societies have historically had limited capacityto reconstruct a system of innovation in the image of their major competi-tors Each countryrsquos social system of production is a conguration of ahost of norms rules and values as well as of institutional arrangementsEach system is constantly changing and is open to inuence from othersystems And indeed many technologies and practices diffuse from onesociety to another but a societyrsquos capacity to be innovative is constrained

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

623

by the existing social system of production Thus the same technologymay exist in numerous countries but how it is employed and how itinuences societal outcomes and performances varies from one institu-tional conguration or society to another (Hollingsworth 1998)

PERSPECTIVES ON INSTITUTIONAL CHANGE

A fundamental problem which remains unresolved in institutionalanalysis involves the nature of institutional change There is much confu-sion and miscommunication in the social sciences about institutionalchange in large part because we do not even know how to measure therate of institutional change Of course we know that there are bothexternal and internal focuses for change and we have had a good bitof scholarship on this subject What is not clearly understood is thatinstitutions and institutional arrangements within societies are histori-cally rooted that there is a great deal of path-dependency to the waythat institutions evolve Moreover the more intricately linked each sectoris with each other sector and with a societyrsquos rules and norms the lesschoice actors have to devise new institutions and institutional arrange-ments On the other hand if a sector or actor is too isolated it may betoo weak within a society to bring about any kind of effective changeThese considerations are very critical to the problem of how muchfreedom actors have to create new institutions afresh to what extentinstitutions diffuse from one society to another (Murmann 1998 Lanzara1998) This is a fundamental question for students of institutionalanalysis as we attempt to understand how the different institutionallevels are linked togetherThere are a variety of reasons why there is confusion in studying insti-

tutional change There is disagreement over the extent to which actorshave the freedom to build new institutional arrangements and if theydo to what extent the new arrangements and organizational patternsmay depart from past practices Among some institutional economists(Williamson 1985) there is a basic assumption that actors build efcientinstitutions by pursuing their self-interest and by promoting theirstrategic goals Moreover there is the assumption that a mixing of fullrationality with market competition tends to produce all of the optimalinstitutions that are needed in order to coordinate a complex capitalisteconomy (Hollingsworth and Boyer 1997 ch 14 Hodgson 1999) Incontrast the argument developed here is much more complexSocial institutions are historically rooted and there is a great deal of

path-dependency in the way that various institutional componentsevolve The shape of institutional congurations at any moment limitsthe type of options for change Because there is a great deal of institu-tional inertia radical change in the institutional components of a society

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

624

is uncommon (Lanzara 1998) Recognizing the constraints existinginstitutions exert on actors we note that existing congurations of insti-tutional arrangements institutional sectors and organizational propertiescan limit but also inuence the degree to which norms and rules canchange Many scholars (North 1990 Ostrom 1990) have emphasized therole of norms rules ideologies and values in limiting the ability of actorsto develop new institutional arrangements which radically depart fromexisting ones However there is a two-way process of interaction aprocess of co-evolution (Baum and Singh 1994a Nelson 1994) Thus asFriedland and Alford (1991 244) argue actors develop new institutionalarrangements by recognizing organizational failures and low perfor-mance but these new arrangements do not necessarily change a societyrsquosnorms rules habits and its underlying value systemCampbell (1997) advances our understanding of how existing institu-

tions enable actors to construct radically new institutions He recognizesthat most institutions and institutional arrangements embody a degreeof inertia from which it is difcult for actors to depart Moreover thedifferential power relations among actors make it unlikely that lesspowerful actors can change the existing power structure of a societyMost change in the institutional components of a society evolves througha process of constrained selection which reects to a considerable degreethe existing arrangements and power relations This becomes the inter-pretative frame for social actors with institutionalized scripts and ritualswhich tend to be taken for granted and appear to be quite ordinary andnatural This institutionalized scripting of the social world is an extremelyimportant source of social stability and inertiaCampbell has developed a very fruitful explanation of change in terms

of the various institutional components outlined here He distinguishesbetween incremental and radical institutional change and argues thatradical institutional change occurs when social actors with widelydiffering norms cultural scripts and rituals for action engage in intenseinteraction with each other (see also Knight 1992) The more diverse theinteractions the greater the potential for institutional change He arguesthat as shifts in the composition of interaction among social actors occurchanges in the interpretation of problems and interests will follow Ifthere are only minor changes in the extent to which diverse actorsinteract there will be only modest institutional change But the morefundamental the changes in the interaction of diverse social actors thegreater the change in the way that actors interpret their world And thegreater the change in the way that actors interpret their world the morelikely that radical changes in various institutional components willfollow In other words diverse decision making tends to create a widerrange of interpretative frameworks of the social world than is likely tobe the case if social actors operate in isolation

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

625

This perspective of diverse interactions among social actors facilitatesour ability to move beyond the traditional view that existing institutionsconstrain the range of institutional alternatives which actors face Theolder view has tended to suggest that actors engage in institutionalchange by extending existing institutional principles habits and conven-tions Specifying the conditions which facilitate fundamental institutionalchange permits a much richer perspective for institutional analysisCampbellrsquos perspective about the conditions under which fundamentalor radical change in institutional components occurs is consistent withmy own research about the conditions under which major discoveriesor fundamental new knowledge occur the more scholars with diversebackgrounds interact with intensity and frequency the greater the like-lihood that they will develop new and alternative ways of thinking(Hollingsworth and Hollingsworth 2000)These views of institutional change are also consistent with the recent

work of Nelson and his colleagues (Mowery and Nelson 1999 Murmann1998) They report the research on a variety of industrial sectors in anumber of different countries revealing that networks of actors in partic-ular industries business and professional associations universityresearch departments and institutes are frequently able to use their collec-tive power to modify a societyrsquos property rights research system andgovernmental policies It is the intense and frequent interaction of thesediverse complex sets of actors ndash sometimes constituting a technologysystem ndash that often leads to incremental changes in a societyrsquos total socialsystem of production It is this kind of collective action that brings aboutorganizational change change in institutional sectors institutionalarrangements and even in the rules and norms of a society And if weare to understand how the style of innovativeness of a society changeswe must rst understand the changes in its institutional makeup It ishighly unlikely that a society will develop a new style of innovativenesswithout changing its institutional structure

FIRMS THEIR INSTITUTIONAL ENVIRONMENT ANDNATIONAL SYSTEMS OF INNOVATION

Variation in the innovative style of societies is shaped by their insti-tutional makeup ndash the various levels which are discussed aboveinstitutions institutional arrangements the structure and cohesion of institutional sectors which constitute a societyrsquos social system of pro-duction and the structure and culture of its organizations ndash especiallyits business rms and other research organizations By undertaking aninstitutional analysis of a society one can begin to understand in whatkinds of organizations the production of specic kinds of knowledgetakes place and how this is linked to particular kinds of innovativeness

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

626

With such a perspective one can gain a rich understanding of why somesocieties excel in the production of radical breakthroughs in basic andapplied science and in developing radically new products and why othersocieties excel in more incremental innovationsReecting on these issues one cannot help being impressed with the

contrasts between the innovative styles of German and American orga-nizations For example Germany has continued to be very successful in making incremental innovations in many industries in which it wasvery competitive well before World War II paper printing materialsmachinery electrotechnical products motor vehicles chemicals textileyarns and fabrics But the Germans have been much less successful indeveloping totally new products or related innovations in biotechnologyelectronics telecommunications aircraft as well as other newer indus-trial areas In contrast the Americans have been very innovative in these and other industries with products having very short lives andwith technologies which change very rapidly and are very complex(Hollingsworth 1991 Soskice 1997 1999)The importance of institutions and innovativeness for international

competitiveness has led to the recognition that there are national systemsof innovation (Lundvall 1992 Nelson 1993 Edquist 1997 Hage andHollingsworth 2000) But even before the recent studies of nationalsystems of innovation Landes (1969) was a proponent of these viewswhen he made the argument that Germany had developed a nationalsystem of innovation by building on its education and scientic researchsystems which were lacking in Britain and elsewhere However the concept of a national system of innovation immediately poses theproblem of what kind of innovations should be consideredDavid Soskice an economist at the Wissenschaftszentrum Berlin has

addressed a set of problems somewhat complementary to the perspec-tive raised here His work (Soskice 1990 1997 1999) suggests that weare sensitive to the institutional environment in which the followingactors are embedded the employees of rms the owners and nanciersof rms the competitors of rms and nally their collaborators ndash aperspective complementary to Hollingsworthrsquos work on social systemsof production (Hollingsworth 1997 1998) When one reects on the insti-tutional contextsocial system of production within which these actorsare embedded one becomes aware that those countries which have some-what inexible labor markets external to rms but a business system in which rms have long-term commitments to their employees arestrongly associated with incremental but hardly at all with radical inno-vations In the rms of such countries (eg Germany and Japan) thereis much more consensus decision making than in societies where a lsquohireand rersquo set of practices is quite pervasive And this kind of consensusdecision making ndash while highly conducive to incremental innovativeness

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

627

ndash limits the capacity for radical innovations to occur This kind of lsquolongtermismrsquo set of norms and rules occurs not only in the employmentpractices of incrementally innovative rms but also in their nancingand ownership patterns (Hollingsworth 1997) During the past thirtyyears these were characteristics of many rms in Germany and JapanSignicantly in both of these countries venture capital markets werenot highly developed nancing tended to be based more on long-termbank loans than on the equity markets and the relationships of rmstended to be relatively stable And these stable institutional arrange-ments also contributed to a high degree of innovativeness ndash butincremental rather than radical innovationsMeantime we should also pursue Soskicersquos ideas (1999) about how

rmsrsquo relationships with their competitors and collaborators shape theirstyles of innovativeness In those advanced industrial societies in whichbusiness rms are highly mobilized into business associations theretends to be a high degree of standardized quality controls over technicalnorms and a high degree of reliance on legal sanctions as monitoringinstruments vis-agrave-vis association members As a result of this high degreeof integration of rms into associations member rms are more likelyto be associated with incremental innovations than is likely to be the case with rms in those societies where associations are less welldeveloped In the latter type of society radical innovations are morelikely to occur In those advanced industrial societies where rms havelong-term stable relationships with collaborators in product design andproduction the style of innovation is likely to be incremental And it is in those societies in which there is a weaker tradition of strongcollaboration among rms that radical innovations are more likely tooccurObviously even in a country where there may be a great deal of

radical innovativeness most innovations will be of an incremental typefor there are simply fewer radical innovations But Germany and therest of Europe not only have had relatively few radical innovations incomparison with the United States but they also have lagged behindthe USA in developing relatively totally new industries Thus there hasbeen much less innovative activity in Europe in high-technology andnewly emerging industries than in the USAIn complementary research Audretsch and Feldman (1995) have

demonstrated that the USA has been more successful in spawning rmsengaged in the early stages of the industry life cycle but relative toGermany had an environment unfavorable to innovative activity byestablished rms especially in more mature industries German rmshave tended to be associated with a routinized technological regime ndashwhere established enterprises have had strong incentives to engage inincremental innovations but new rms have tended to have fundamental

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

628

innovative disadvantages relative to the Americans (see Audretsch 1995Winter 1982 Casper 2000)On both sides of the Atlantic universities engage in basic and applied

research But signicantly the institutional environment in which univer-sities are embedded inuences the degree to which the research is likelyto represent major breakthroughs or is likely to have more modestimplications Recent research in biomedical science demonstrates that inthose societies in which university professors andor graduate studentshave the following characteristics there are likely to be relatively majorresearch breakthroughs graduate students have a long training periodhigh dependency on a major professor low autonomy in dening thenature of a research project restricted job options and a university envi-ronment where there is a very modest or nonexistent entrepreneurialculture For example in Germany ndash a country with relatively few radicalbreakthroughs ndash the scholar who nishes a habilitation will have spentmany years on a research project before qualifying for a permanent acad-emic position will have been in a high-dependency position with a seniorprofessor will have had limited autonomy to dene a research projectand will be constrained from pursuing a high-risk research project Theyoung German will have an extraordinarily high level of training butwill have faced a number of disincentives to undertake high-risk researchAnd as soon as heshe has completed the habilitation and becomes aprofessor administrative responsibilities constitute another disincentivefrom engaging in high-risk or long-term research projects Moreover theGerman professor is a civil servant with a more stable and secure incomethan hisher American counterpart In a very different institutionalenvironment American professors in large research universities tend tobecome much more entrepreneurial Today many American professorsbehave as though they were operating their own rms within univer-sities and thus devote considerable time to generating funds for researchprojects assistants secretaries supplies etc And the variation in thisentrepreneurial spirit across countries is associated with variation in thetaking of high-risk research strategies and the making of major researchbreakthroughsIn sum societies vary in terms of the institutional environment in

which their organizations are embedded and this variation inuencesthe degree to which they have industrial activities in the early or latestages of the industry life cycle successful new or mature industriesand radical vs incremental innovations And it is this variation whichis encouraging rms in countries with relatively few new industries and with few or no radical innovations to invest in those countries with many new industries and numerous radical innovations ndash as astrategy to be able to transfer new technologies from the latter institu-tional environment to the former

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

629

Table 4 summarizes the way that the norms rules and values of a society are associated with Americarsquos and Germanyrsquos institutionalarrangements their social system of production and their styles of inno-vativeness

CONCLUSION

Several themes deserve to be reemphasized Most importantly institu-tional analysis is emerging as a dominant issue in several social sciencedisciplines The study of various institutional components discussedabove is a complex subject requiring multiple levels of analysis and aknowledge of multiple sectors within societies It requires a theoreticalperspective that evolves from an empirical analysis in different societiesIf we are to advance the study of institutional analysis we need cross-disciplinary collaboration and team research which are not readilyachieved given the structure of most of our research universitiesAt present scholars in each academic discipline have their own distinc-

tive strategies for studying institutional analysis Thus economists tendto use several approaches (North 1990 Williamson 1985 Hodgson 19881998 1999) anthropologists another (Geertz 1995) and sociologistspolitical scientists and historians employ other strategies (Campbell etal 1991 Powell and DiMaggio 1991 Katznelson 1998 Hall and Taylor1996) With practitioners in various disciplines pursuing their distinctiveapproaches to institutional analysis there is low potential for collabo-rative institutional analysis across the social sciences The lessons fromthe biological sciences over the last 40 years are very instructive oncepractitioners from various biological disciplines began to cooperate instudying biological phenomena by using the same concepts at the mole-cular level theoretical advance was very rapid Comparable phenomenatook place in the development of other hybrid elds eg biochemistrybiophysics (Judson 1979 Olby 1979 Kohler 1982 1994) Similarly weresocial scientists in various disciplines to work collaboratively in the areaof institutional analysis there would very likely be an acceleration intheir theoretical knowledgeThis is not a plea that everyone engaged in institutional analysis should

do the same kind of research Indeed just as biologists working at themolecular level study many different kinds of problems similarly socialscientists engaged in institutional analysis would also work on manyseparate problems but in the spirit of a collective enterprise Some wouldconcentrate their attention on the study of rules norms habits conven-tions and values while others would study how these are associatedwith congurations of various institutional arrangements (types ofmarkets hierarchies networks associations communities clans statesetc) Others would work on specic institutional sectors (eg education

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

630

Table 4

Rules norms hab

its an

d values associated w

ith sp

ecic co

ngu

ration

s of coo

rdinating mecha

nism

s soc

ial sy

stem

s of

prod

uction

an

d nationa

l styles of inno

vation

USA

Germany

Norms rules values which provide incentives for

Norms rules values which provide incentives for

short-term horizons

long-term horizons

Con

gu

rative

forms of

Marke

ts c

orpo

rate hierarchies reg

ulatory state

Assoc

iation

s (busine

ss assoc

iation

s la

bor union

s)

coordinating

corp

oratist-type

networks

marke

ts c

orporatist-

mecha

nism

s (listed in

type

state

order of im

portanc

e in

the particu

lar social

system

)

Social sys

tem of

prod

uction

Bus

iness system

Firm

struc

tures

Con

glom

erates b

ut with mov

emen

t toward

Firm

more co

ngru

ent with pa

rticular prod

ucts

elim

ination of produc

tion

in som

e sectors

emph

asis on long

-term strateg

ies rm

s are well

emph

asis on sh

ort-term

strateg

ies high

integrated

differentiation

between divisions

of rms

Firm

s mob

ilized in

Low

High

and in

tegrated

into busine

ss

associations

Indu

strial relations

Low

deg

ree of jo

b secu

rity rigid in

ternal lab

or

High deg

ree of jo

b secu

rity

exible interna

l lab

or

system

marke

t ex

ible externa

l labo

r marke

tsmarke

t rigid externa

l lab

or m

arke

tPatterns of

Dispersed owne

rship frequ

ent turnov

erMore stab

le owne

rship patterns

owne

rship

Table 4

continue

d

USA

Germany

Norms rules values which provide incentives for

Norms rules values which provide incentives for

short-term horizons

long-term horizons

Social system of

Prod

uction

Bus

iness system

Training for labo

r an

dHigh deg

ree of voluntary decision mak

ing by

Collective an

d m

ore co

mpulso

ry d

ecisions

abo

ut

man

agem

ent

individuals as to ho

w m

uch

tim

e an

d ene

rgy

how m

uch ed

ucation

is ne

eded

he

avy em

pha

sis

to inv

est in educ

ation heavy

emph

asis on

on techn

ical training (eg en

gine

ering)

busine

ss sch

ools training

in marke

ting

and

sales accou

nting

Fina

ncial marke

tsEqu

ity an

d ven

ture cap

ital m

arke

ts highly

Equ

ity an

d ven

ture cap

ital m

arke

ts w

eakly

dev

elop

ed na

ncing from

cen

tral gov

ernm

ent

deve

lope

d he

avy em

pha

sis on

corporate loan

sto priva

te rm

s for dev

elop

men

t of new

tech

nologies related

to na

tion

al security

Unive

rsity research

High individual auton

omy high poten

tial for

Low

ind

ividua

l au

tono

my until almost 40

yea

rs of

system

individual creativity u

nive

rsity pr

ofesso

rs

age poten

tial for ind

ividual creativity high

ly

high

ly entreprene

urial

cons

traine

d unive

rsity pr

ofessors not so

entrep

rene

urial

Inno

vative

styles

Highly su

ccessful in m

ajor break

through

s in

High pe

rforman

ce in proc

ess an

d inc

remen

tal

basic an

d app

lied scien

ce go

od in dev

elop

ing

inno

vation

s in existing produ

cts not so su

ccessful

totally

new

lines of pr

oducts an

d in m

ajor break

through

s in basic scien

ce or in

tech

nologies high

deg

ree of entrepr

eneu

rship

deve

loping

totally new

lines of prod

ucts of

amon

g research

ers

tech

nologies low d

egree of entrepr

eneu

rship

amon

g research

ers

Table 4

continue

d

USA

Germany

Norms rules values which provide incentives for

Norms rules values which provide incentives for

short-term horizons

long-term horizons

Prod

ucts in w

hich

Pr

oducts with sh

ort liv

es prod

ucts in w

hich

Older produ

cts with long

history machine

ry pap

er

society excels

tech

nology

is bo

th highly co

mplex

and

pr

inting

textile

s m

otor veh

icles m

achine

too

lsch

ange

s rapidly en

tertainm

ent indus

try

pha

rmaceu

ticals biotechn

olog

y softw

are

aerosp

ace

business systems nancial markets and systems of research) Distinctivewould be the recognition of how these separate studies are linkedtogether how each institutional sector is linked to a societyrsquos normsrules values etc and its conguration of institutional arrangementsSimilarly other institutional analysts might focus on the study of orga-nizations with a major concern for how the institutional environmentof organizations (eg norms rules values congurations of institutionalarrangements institutional sectors) inuenced the structure culture andoutcomes of organizations And nally others might study how all ofthe aforementioned aspects of institutional analysis inuenced a societyrsquosoverall performance Nelson and Sampat (1998) have appropriatelyreminded us that we should engage in institutional analysis in order toaddress specic problems but the view developed in this article is thatwe rst need some mapping of the terrain of institutional analysis inorder for our diverse enterprises to become a meaningful collective enter-prise especially if we are to understand how institutions inuence theinnovative processThe argument developed here is not that all social scientists should engage

in institutional analysis Obviously there are numerous other importantareas of research which lie beyond the eld of institutional analysis Nordoes this article argue that institutional analysis is the most importantsubject for social scientistsSince the structure and culture of our research universities reect an

enormous amount of disciplinary fragmentation what are the prospectsfor promoting the kind of collective and interdisciplinary research agendaproposed above In one respect the prospects are not encouragingDuring the past half century the efforts of American universities to have effective communication across the biological sciences have variedgreatly in their success At the University of California-Berkeley it was not until the 1980s that there was a serious effort to integrate thebiological sciences and at that time the campus abolished seventeendepartments to promote an integrated biological program But at theUniversity of Wisconsin the biological sciences are still very fragmentedinto dozens of different departments in botany plant pathology zoologyand physiology with two departments of biochemistry and two depart-ments of geneticsThere have been other approaches which have worked very well histor-

ically for promoting scientic integration amidst scientic diversity Forexample several summer schools and special institutes were organizedin the United States during the 1940s in an effort to overcome the frag-mentation in the biological sciences These programs made it possiblefor scientists in diverse academic disciplines and from different univer-sities and countries to come together to engage in intense and frequentinteraction Perhaps the most effective program was that held in the

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

634

summers at Cold Spring Harbor New York where a research agendafor the study of molecular biology was developed (Fischer and Lipson1988 Stent and Watson 1966) Scientists went there summer aftersummer and when they returned to their universities many began therestructuring of the biological sciences in their universitiesIn our own day we need summer institutes and other similar programs

so that scholars from diverse backgrounds can come together anddevelop a collective agenda for doing institutional analysis and forstudying the impact of institutions on the performance of societies ndashinnovativeness being one such subject If we can develop such instituteswe will increase the prospects for transforming our universities Thereis reason to be optimistic that once an eclectic group of scholars workwith a common set of concepts we can have genuine theoretical advanceif researchers can bring their diverse expertise to bear on a common setof problems There is considerable evidence that institutional analysisoffers high potential for advancing our knowledge of innovations andtechnological change (Hollingsworth and Hollingsworth 2000) If we aregoing to make substantial theoretical advance it is not likely to occurfrom the vantage point of a multiplicity of disciplines Rather it is likelyto occur as we develop a new intellectual framework and our own acad-emic journals This is a tall agenda which requires us to reassess theway that knowledge is presently produced in our societies

NOTE

This article is part of a much larger agenda involving the study of institutionsorganizations and innovations in which I am involved My debts are many andvaried The rst version was written while I was in residence in the ResearchUnit for Institutional Change and European Integration of the Austrian Academyof Sciences in Vienna in the summer of 1998 I am especially grateful to ProfessorsEgon Matzner and Sonja Puntscher-Riekmann for providing the stimulating envi-ronment in which to think through the ideas developed here Egon Matzner hasengaged in very stimulating discussions with me about the importance of insti-tutions in the general eld of socio-economics The nal version of the articlewas completed during my residence as a fellow at the Neurosciences Institutein La Jolla California I also wish to thank my colleagues in the InnovationTheme Project at the Netherlands Institute for Advanced Study (NIAS) during1998ndash9 and Edgar Grande and Karl Muumlller for stimulating conversations aboutthe interaction of institutions and organizations in facilitating particular typesof innovation I am especially grateful to Jerald Hage for helping me to developa number of issues in the article and for his useful comments on the rst draftFrans van Waarden early on helped me to work through the idea of multiplelevels of institutional analysis Others who have helped to develop the ideas inthe article have been my colleagues in the research project lsquoComparative socialsystems of productionrsquo especially Tom Burns Christel Lane Yoshitaka OkadaWolfgang Streeck and Richard Whitley My colleagues Robert Boyer SteveCasper and David Soskice have taught me much about how incentives providedby the institutional environments of rms inuence individual decision making

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

635

But my most important debt is to Ellen Jane Hollingsworth who has repeatedlyquestioned every single assumption of this article Without her rich and contin-uous tutoring this article would be even less developed than it is at presentFinally without the insightful assistance of David Gear in many phases of thisarticle it would not have been completed

REFERENCES

Aoki Masahiko (1990) lsquoTowards an economic model of the Japanese rmrsquo Journalof Economic Literature March 1ndash27

Arora A Landau Ralph and Rosenberg Nathan (eds) (1998) Chemicals and Long-Term Economic Growth Insights from the Chemical Industry New York JohnWiley amp Sons

Arrow Kenneth (1974) The Limits of Organization New York NortonArthur W B (1988a) lsquoSelf-reinforcing mechanisms in economicsrsquo in P W

Anderson K J Arrow and D Pines (eds) The Economy as an Evolving ComplexSystem Redwood City Calif Addison-Wesley

mdashmdash(1988b) lsquoCompeting technologies an overviewrsquo in G Dosi C Freeman RNelson G Silverberg and L Soete (eds) Technical Change and Economic TheoryLondon Pinter Publishers pp 590ndash607

Audretsch David B (1995) Innovation and Industry Evolution Cambridge MassMIT Press

Audretsch David B and Feldman Maryann P (1995) lsquoInnovative clusters andthe industry life cyclersquo Discussion Paper Wissenschaftszentrum BerlinMarket Processes and Corporate Development Research Unit FS-IV-95-7

Baum Joel A C and Oliver Christine (1992) lsquoInstitutional embeddedness andthe dynamics of organizational populationsrsquo American Sociological Review 57540ndash59

Baum Joel A C and Singh Jitendra V (1994a) lsquoOrganizationndashenvironment coevo-lutionrsquo in Joel A C Baum and Jitendra V Singh (eds) The EvolutionaryDynamics of Organizations New York Oxford University Press pp 379ndash402

mdashmdash(eds) (1994b) The Evolutionary Dynamics of Organizations New York OxfordUniversity Press

Berger Suzanne and Dore Ronald (eds) (1996) National Diversity and GlobalCapitalism Ithaca NY Cornell University Press

Boyer Robert (1997) lsquoThe variety and unequal performance of really existingmarkets farewell to Doctor Panglossrsquo in J Rogers Hollingsworth and RobertBoyer (eds) Contemporary Capitalism the Embeddedness of InstitutionsCambridge and New York Cambridge University Press pp 55ndash93

Burns Tom R and Dietz Thomas (2001 forthcoming) lsquoSocial rule system theorysocial action institutional arrangements and evolutionary processesrsquo in JRogers Hollingsworth and Karl Muumlller (eds) Advances in SocioeconomicsLanham Md Rowman amp Littleeld

Burns Tom R and Flam Helena (1987) The Shaping of Social Organization BeverlyHills Calif Sage

Calvert Randall L (1995) lsquoThe rational choice theory of social institutionsrsquo inJeffrey S Banks and Eric A Hanuskek (eds) Modern Political Economy NewYork Cambridge University Press pp 216ndash66

Camic Charles (1986) lsquoThe matter of habitrsquo American Journal of Sociology 911039ndash87

Campbell John L (1997) lsquoMechanisms of evolutionary change in economicsgovernance interaction interpretation and bricolagersquo in Lars Magnusson

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

636

they raise other no less severe problems It is important to recognizethe imperfection of any single institutional arrangement in order tocomprehend the origin and transformation of any other institutionalarrangement (Hodgson 1999 ch 3)As suggested above each institutional arrangement is congured with

other institutional arrangements Usually one particular institutionalarrangement is more dominant in a particular conguration but becauseeach type of institutional arrangement has its own strengths andweaknesses there is no simple structural logic in the governance or coor-dination of a society Each institutional arrangement constrains the otherbut the inherent tension among the various institutional arrangementswithin a conguration contributes to changes in the governance orcoordination of a society The routines and logics of each of these cong-urations of institutional arrangements provide constraints and incentivesfor actors The way that actors perceive the incentives and constraintsof these governance congurations leads to a particular market logicand it is the specic market logic of a society which inuences its speciccapacities and weaknesses The inherent strain among the different logicsin a conguration of coordination helps to provide the exibility for asociety to adapt to new circumstancesIf there were a society with pure markets or a society coordinated

only by the state there would be too much rigidity and too little diver-sity to cope with the vast uncertainty in the global environment A societywith very little diversity in its coordination mechanisms would havelittle capacity to adapt to new circumstances Congurations with consid-erable diversity of institutional arrangements provide for a certainamount of incoherence in governance but they also provide for thecapacity to adapt to new circumstances While the Soviet Union wasdominated by the state as a coordinating mechanism there were alwaysfunctioning markets in the system Moreover the feudal society of thetwelfth and thirteenth centuries consisted not only of hierarchical rela-tionships between serfs and masters but also of urban guilds clericalhierarchies and markets (Johnson 1992 38) In sum the robustness ofinstitutions often depends on multiple and diverse principles and logicsof action on the inconsistency of principles and procedures on patternedforms of disorder (Lanzara 1998 Orren and Skowronek 1991 320 329)

Third level of analysis

The next level consists of the institutional sectors of a society The rulesnorms and values of a society inuence the array of institutional arrange-ments and levels one and two inuence the nature of and the relation-ships among various institutional sectors and all three levels areintricately linked together to form a social system of production Together

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

613

all of these components inuence the performance of economic sectorswithin a society as well as the performance of the total society A socialsystem of production is the way that a societyrsquos institutions (see rstlevel of analysis above) its institutional arrangements (see second levelof analysis above) and its institutional sectors are integrated into a socialcongurationAn institutional sector includes all organizations in a society which

supply a given service or product along with their associated focal orga-nizations (eg major suppliers funding sources regulators and so forth(Scott et al 1994 108 117 Campbell et al 1991 Hollingsworth et al1994) Institutional sectors include but are not limited to the societyrsquossystem of education system of research business system nancialmarkets legal system and the state The structure of the nancial marketssystem of training and education industrial relations system system ofresearch and state and legal systems are distinct and idiosyncratic ineach society In short they are system specicAll of these institutional sectors tend to cohere with each other

although they vary in the degree to which they are tightly coupled intoa full-edged system While each of the institutional sectors has someautonomy and may have some goals that are contradictory to the goalsof other institutional sectors with which it is linked an institutional logicin each society leads institutions to coalesce into a complex social con-guration (Hollingsworth 1991) This occurs because the institutionalsectors are embedded in a culture in which their logics are symbolicallygrounded organizationally structured technically and materially con-strained and politically defended The conguration of institutionalsectors usually exhibits some degree of adaptability to new challengesbut continues to evolve within an existing style But under new circum-stances or unprecedented disturbances these institutional congurationsare exposed to sharp historical limits as to what they may or may notdo (Schumpeter 1983 David 1988 Arthur 1988a 1988b Haringkanssen andLundgren 1997)Why do all of these different institutions coalesce into a complex

social conguration which is labeled here as a social system of pro-duction The literature suggests two contrasting interpretations Part ofthe answer ndash indeed a controversial one ndash is that these institutionalsectors are functionally determined by the requirements of the practiceof capitalism in each time and place (Habermas 1975) Another expla-nation emphasizes the genesis of the actual conguration via a trial and error process according to which the survival of rms regions orcountries is the outcome of complex evolutionary mechanisms (Maynard-Smith 1982 Nelson and Winter 1982) However the problem is evenmore complex Markets and other mechanisms for coordinating rela-tionships among economic actors place constraints on the means and

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

614

ends of economic activity to be achieved in any society The othercoordinating mechanisms include different kinds of hierarchies varioustypes of networks and associations ndash trade unions employers and busi-ness associations (Hollingsworth and Lindberg 1985 Campbell et al1991) The logic of these various institutional arrangements providesactors with vocabularies and logics for pursuing their goals for deningwhat is valued and for shaping the norms and rules by which theyabide (see Tables 2 and 3) In short in contrast to the logic of theneoclassical paradigm the argument here is that the dominant type ofinstitutional arrangements places severe constraints on the denition of needs preferences and choices of economic actors Whereas the neo-classical paradigm assumes that individuals and rms are sovereign thisarticle is based on the assumption that rms are inuenced by the hold that the institutional congurations making up a social system ofproduction have on individual decision making (Campbell et al 1991Etzioni 1988 Streeck and Schmitter 1985a Hollingsworth et al 1994Hollingsworth and Boyer 1997 Magnusson and Ottosson 1997 North1990 Hodgson 1999)Standard neoclassical economic theory has tended to downplay the

role of production and consequently of rms Even transaction cost theo-rists who are concerned with analyzing the rm as a coordinatingmechanism have been relatively unconcerned with the various compo-nents of a social system of production Indeed as long as there waswidespread optimism about the efcacy of Keynesian economics therewas relatively little concern among neoclassical economists with thesupply side of the economy Even in the opinion of most Keynesians agroup of experts should ideally be able to shape the size of aggregatedemand while the supply side of the economy would be left to the twominimalist institutions of neoclassical economics ndash markets and manage-rial hierarchies For several decades however it has become increasinglyobvious that some of the most competitive and successful patterns ofindustrial output and industrial production in capitalist economies donot derive from the neoclassical prescription of unregulated markets andcorporate hierarchies complemented by a neoliberal democratic stateIndeed it is now well understood that certain highly successful produc-tion patterns require for their emergence and survival institutionalarrangements and environments the very opposite of the prescriptionsfound in the neoclassical paradigm (see especially Streeck 1992 but alsoHollingsworth and Streeck 1994) Thus social scientists have come tounderstand that if they are to comprehend the behavior and performanceof contemporary economies concerns about social systems of produc-tion must be brought into the pictureBecause production involves more than technology a number of social

scientists have an increasing concern with social systems of production

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

615

The same equipment is frequently operated quite differently in the samesectors in different countries even when rms are competing in the same market (Maurice et al 1980 Sorge and Streeck 1988 Sorge 1989Hollingsworth et al 1994) Variations in production and process tech-nologies are inuenced partly by variations in the social environmentsin which they are embedded In other words rms are embedded intocomplex environments which among other things place constraints ontheir behavior Thus a social system of production is of major impor-tance in understanding the behavior and performance of an economyHow the state and other institutional arrangements (eg marketsnetworks private hierarchies associations) coalesce and are related toparticular social systems of production are important determinants ofeconomic performanceThe congurations of institutional arrangements that coordinate or

govern the behavior of actors in one society and its structure of specicinstitutional sectors cannot easily be transferred to another society forthey are embedded into a social system of production that is societallydistinct (Hollingsworth 1997) Societies borrow selected principles offoreign management styles and work practices but the effectiveness ofsuch borrowing is generally limited Economic behavior and performanceare shaped by the entire social system of production in which actors areembedded and not simply by specic principles of particular manage-ment styles and work practices Moreover a societyrsquos social system ofproduction tends to limit the kind of goods that it can produce and withwhich it can compete successfully in international marketsSome scholars (Kenney and Florida 1993 Oliver and Wilkinson 1988)

have assumed that the diffusion of particular forms of management stylesand work practices across societies could lead to system convergenceReferring to the Japanese production system Kenney and Florida haveargued that it consists of organizational practices whose fundamentallsquogenetic codersquo can successfully be inserted into another society and beginsuccessfully to reproduce its behavior in the new environment (1993 8)Their position is in the intellectual tradition of Antonio Gramsci (1971)who contended decades ago that the American system of mass produc-tion would most certainly diffuse to Europe over timeHowever the argument of this article is that even though British

French and American rms may adopt certain aspects of Japanesemanagement styles (eg just-in-time production self-managing teamsquality circles the use of statistical process controls) or some variant ofthe German vocational system their social systems of production willnot converge Moreover the overall conguration of a societyrsquos socialsystem of production inuences its sectoral and overall national economicperformance This helps to explain why societies have different systemsof innovation and why they vary in the clusters of industries in which

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

616

they are highly competitive in the international markets (Nelson 19931999 Hage and Hollingsworth 2000 Berger and Dore 1996)In order to understand how and why a societyrsquos economy performs

as it does it is necessary to understand its entire social system of produc-tion If a society is to modify substantially the performance of itsinstitutional sectors and its entire economy it cannot adapt only someof the management and work practices of its foreign competitors Ratherit must alter its entire social system of production Because a societyrsquosmodes of economic governance and coordination and its institutionalsectors develop according to particular logics and are system specicthere are serious limitations to the extent to which a society may mimicthe institutions (eg the rules and norms) institutional arrangements andinstitutional sectors of other societiesIn the history of modern societies there are logics by which institu-

tions coalesce into a social system of production (Hollingsworth 1991)Though institutions are constantly changing there are sharp limits tothe type and the direction of change that any particular institution canundergo because of its linkages with institutional arrangements and insti-tutional sectors Thus a societyrsquos business rms educational systemnancial markets industrial relations system etc can engage in seriousrestructuring only if most of the other institutional sectors also changeIn social systems of production there are pressures toward consistencyin the norms rules and values across institutional sectors though in anycomplex society social systems are obviously imperfectly integratedIndeed the degree to which the institutional norms and rules makingup a social system of production are loosely or tightly coupled is a vari-able of considerable importance In general the institutional sectorsmaking up a social system of production are interdependent and changesin one generally result in changes in the others Since each institutionallevel is dependent on the others for various types of resources there isinterdependence among the differing institutional spheres Moreovereach society has its norms moral principles rules and laws and recipesfor action as well as its own idiosyncratic customs traditions and prin-ciples of justice (Burns and Flam 1987)There are also other inherent obstacles to convergence among social

systems of production for where a system is at any one point in timeis inuenced by its initial state Systems with quite different initial statesare unlikely to converge with one anotherrsquos institutional practicesExisting institutional arrangements block certain institutional innova-tions and facilitate others (Roland 1990) Thus the institutions makingup a social system of production provide continuity even though insti-tutional arrangements are always changing but with a particular logicWhile Williamson (1975 1985) suggests that actors tend to choose theinstitutional arrangements which are most efcient North (1990) is

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

617

much closer to the mark in his argument that most societal institutionalarrangements exist as a result of custom and habit and are thereforeinefcient At any moment in time the world often appears to its actorsas very complex and uncertain For this reason actors often engage incontradictory forms of behavior pursuing different strategies as hedgesagainst a very uncertain world (Lanzara 1998) And their hedging andcontradictory forms of behavior may lead in somewhat different soci-etal directions all constrained by the institutional fabric within whichthe actors are embedded This kind of contradictory behavior occurs inpart because of the contradictions inherent in the dominant congura-tions of institutional arrangements with their contradictory logics (seelsquoSecond level of analysisrsquo above)Despite the emphasis on the logic of institutional continuity this is

not an argument that systems change along some predetermined pathThere are critical turning points in the history of highly industrializedsocieties but with the choices limited by the existing institutional terrainBeing path-dependent a social system of production continues along aparticular logic until or unless a fundamental societal crisis intervenes(Milgrom et al 1991 Krugman 1991 Durlauf 1991 Hollingsworth 1991Pred 1966 David 1988 Hodgson 1998)At this point it is important to confront the question of how social

systems of production evolve And this question gets to the heart of theproblem of building complex societal structures Certainly social systemsof production did not emerge from some process of social engineeringMoreover the various component parts of each social system of produc-tion have often not been designed to be part of a social system ofproduction The component parts of each social system of productionhave emerged more often than not as unintended by-products of goalswhich various actors had in mind at earlier moments in time It is usuallythe case that actors at time lsquotrsquo fail to comprehend the long-term conse-quences of their actions Wolfgang Streeck the author of several brilliantpapers on social systems of production argues that institutional sectordesigns created for one purpose generally address that goal but overtime those designs have quite unintended consequences (Streeck 1997a1997b Wright 1998) Describing the conguration of the German socialsystem of production of the 1970s and 1980s Streeck points out that itwas the unintended by-product of multiple points in history Someelements were pre-Wilhelmian others were introduced by the Alliedpowers after 1945 and others emerged during the years of the GermanFederal Republic All component parts of the German social system ofproduction

were and continue to be changing for their own reasons as wellas in reaction to each other and certainly there can be no presump-tion of a pre-established t between them even though one might

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

618

want to allow for some reinforcement effects of the modelrsquos histor-ically contingent social and economic success

(Streeck 1997b 54)

In sum the emergence of social systems of production is a long-termevolutionary process with each part interacting with its environment andresulting in a congurative whole but with no previous design by eithera single actor or a collectivity of actors As suggested above actors do not recreate their world anew At any moment in history the totalinstitutional complex of a society necessarily contains the resources and legacies of its past Preexisting institutional complexes are nevercompletely wiped out Societies consist of multiple layers of history withtheir diverse logics of action However the cumulative effect of smallperipheral changes in altering particular institutional sectors and insti-tutional arrangements can be substantial (Murmann 1998 Lanzara 1998)lsquoInstitution building is affected more by the ways in which people havecodied the past than by how they envision the future There are manymore potential resources in pre-existing institutional arrangements thanit is usually assumed or suspectedrsquo (Lanzara 1998 30)

Fourth level of analysis

The next level of institutional analysis organizational structures is some-what more controversial As suggested above Douglass North (1990)draws a sharp distinction between institutions and organizations Morerecently many organizational theorists (Powell 1991 Powell andDiMaggio 1991 DiMaggio and Powell 1983 Baum and Oliver 1992Kondra and Hinings 1998 Townley 1997) argue that institutional rulesnorms and conventions unfold in tandem with organizational structuresand this is my positionThe literature which focuses on how institutions inuence organiza-

tions is quite different from two other theoretical literatures which alsoare concerned with how organizational environments shape the behaviorof organizations There are resource dependency theorists (Pfeffer andSalancik 1978) who emphasize the role of environmental resources inshaping organizations while Hannan and Freeman (1977 1984) and otherpopulation ecologists emphasize the survival of organizations givencertain kinds of organizational conditions (Orru et al 1991)The institutionalist perspective as implied above emphasizes the

normative environment in which organizations are embedded It is a perspective which focuses on the way in which organizations in their behavior tend to conform to the institutional rules and normswhich are dominant in the organizational environment However allthree perspectives ndash the institutionalist the population ecology and

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

619

the resource dependency ndash emphasize how the environment inuencesorganizations and how organizations which are subject to the sameenvironment tend to converge in their behavior to have what DiMaggioand Powell have labeled organizational isomorphism (DiMaggio andPowell 1983)There has emerged a vast literature which demonstrates that within

each society there are scal political judicial and other regulatory normswhich limit and shape the culture and structure of organizationalbehavior The normative institutional environment of organizations limits the options of what organizations do in a particular society and inuences the patterns of ownership relations with suppliers andcustomers In short it is the normative environment of organizationswhich denes within a particular society what is socially acceptablebehavior for organizations (Hamilton and Biggart 1988 Hollingsworthand Hollingsworth 2000 Meyer and Rowan 1991 Zucker 1987 1988Orru et al 1991 Townley 1997)Thus far most of the analyses on isomorphic pressures operating on

organizations have been cross-sectional in nature However the entireconcept of isomorphism implies that there are strong environmental pres-sures exerted on organizations and in order to observe this phenomenonit is important that we have longitudinal studies which assess howchanges in an organizationrsquos institutional environment inuence changesin the structure and culture of organizations Moreover the expectationhere is that when one analyzes the patterns of organizational changeover time the isomorphism will be holistic ndash that is within a particularsociety there will be a tendency for the internal structure of particulartypes of organizations to converge in large part because of the pres-sures to conform to the changes in the external norms rules and valuesA close reading of Aoki (1990) and Slack and Hinings (1994) nds thatchanges in the national institutional environment of organizations haveinuenced changes in the internal structure of organizations within aspecic industry of a particular societyHowever it is inappropriate to view organizations as changing only

in relation to exogenous environmental change There is a kind of co-evolution which occurs between organizations and their institutionalenvironments Co-evolution implies nonlinear feedback between orga-nizations and their institutional environments This is of course quitedifferent from the more common way of studying organizations bymodeling relations between independent and dependent variables Ofcourse the more common type of analysis is quite appropriate whenthere are simple relationships which do not involve complex feedbackprocesses But in the strategy for institutional analysis which is proposedherein it is less useful to separate independent from dependent vari-ables and more useful to understand the interacting and co-evolutionary

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

620

processes among all of the levels of institutional analysis discussed inthe various sections above (Baum and Singh 1994a 379ndash80 Mowery andNelson 1999 Murmann 1998)The logic of the perspective herein suggests that within every

society there is variation in the structure and culture of business rmsuniversities and other complex organizations (Kondra and Hinings1998) However this variation takes place within parameters which aresystem specic For example every German rm and every Germanuniversity is different from every other German university and rm but there are a set of cultural and structural characteristics which distin-guish German rms and universities from those in America Similarlyeach Japanese rm and university is unique yet they have a set ofcultural and structural characteristics which are system specic andwhich differentiate them from business rms and universities in other countriesIn all societies each organization has its own distinctive organizational

rules norms and conventions which are subordinate to the meta norms and rules of the larger society within which they are embeddedHowever the strength of the institutional environment within whichorganizations are embedded varies from society to society Some societies have multiple institutional environments and there is hetero-geneity in terms of what constitutes appropriate behavior oforganizations In such societies there can be conict over whatinstitutional logics should regulate specic organizational functions(Friedland and Alford 1991 Townley 1997 262ndash3) In those societies inwhich the institutional norms habits and rules are most developed and in which the institutional pressures to conform are greatest thereis less variation in the structure and culture of business rms and variouskinds of research organizations In such societies the connectednessbetween research organizations and their institutional environments is sufciently strong for organizations to have low autonomy to pursue independent strategies and goals in these societies there is a great deal of organizational isomorphism Conversely the weaker the institutional environment in which research organizations areembedded the greater the variation in the structure and culture of business rms and research organizations Moreover where the institu-tional environments are more weakly developed organizations havegreater autonomy and exibility to respond to the development of new knowledge and to be highly innovative Hence it is in those societies where the institutional environments are most developed and most rigid and when there is less organizational autonomy and exibility that fewer radical innovations in basic and applied science as well as in totally new products and industrial sectors havebeen made

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

621

Fifth level of analysis

We come next to the outputs and performance of the various institutionalcomponents of a society It is at this level that institutional componentsare more pragmatic and exible For example in the legal sector thereare specic statutes and court rulings in the state sector there are specicpolicies in the business sector there are new products new technolo-gies and market strategies It is at this nal level of analysis thatinstitutional spheres are most open and susceptible to change and cross-national mimicry is easiest and most commonThrough the outputs of the society we can obtain some assessment as

to how well a society is performing Moreover we can assess how inno-vative it is how egalitarian it is in the distribution of its resources howegalitarian it is in terms of levels of health education etc Just as it iscomplicated to measure how egalitarian a society is similarly it is dif-cult to assess how innovative it is For example societies vary in makingincremental and radical breakthroughs in basic and applied science in developing totally new products and new kinds of organizations inincremental and process innovations in existing products and organiza-tions in developing new and different forms of marketing ndash bothdomestically and globally (Hage and Hollingsworth 2000) Of coursesocieties vary in their rates of economic growth in their rates of economicproductivity in their quality of life (eg rates of crime life expectancyetc) All of these performance criteria feed back and inuence each ofthe levels of institutional analysis discussed above ndash rules norms valuesetc institutional arrangements institutional sectors the structure andform of organizations Moreover performance measures may inuenceeach other For example if a society has very low rates of growth itmay not be very innovative in some of the types of innovative activitiesmentioned above Good or poor performances of certain kinds inuenceother performance indicators (Hollingsworth et al 1990)Different institutional arrangements and different social systems of

production result in different types of economic performance Hence solong as societies have different social systems of production there areserious constraints on the degree to which they can converge in theirinnovative styles Different social systems of production tend to maxi-mize in a more or less explicit manner different performance criteriausually mixing considerations about static and dynamic efciency protsecurity social peace and economic andor political power In short incontrast to the implications of neoclassical economic theory in real-worldeconomies there are no universal standards all economically rationalactors attempt to maximize Economic history provides numerous exam-ples of how a great variety of principles of rationality are implementedin different societies

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

622

A critical question in institutional analysis is whether an existing socialsystem of production which supports a set of routines for a particularkind of technology and industry can shift from old practices to newones There are numerous historical accounts of how the social systemof production in a particular society which worked so well for a numberof years could not adapt to new technologies (Lazonick 1990 Veblen1915 Nelson 1994) Schumpeter (1983) and Freeman (1991) have devel-oped the idea of a techno-economic paradigm which suggests thatdifferent technologies require specic forms of organizational arrange-ments ndash or what I called social systems of production Because oflsquoinstitutional inertiarsquo the social system of production of a society maynot be able to adapt to a new techno-economic paradigm (Nelson 199458 Lanzara 1998)Whether a social system of production can sustain its performance

standards depends not only on its intrinsic economic rationality but alsoon where it ts into a larger system If a particular social system ofproduction is immune from the innovativeness and competition of analternative system survival can be long lasting But if different socialsystems of production with diverging criteria of good economic perfor-mance meet in the world arena the arbitrariness of nationally imposedconstructed performance standards may be superseded by alternativeperformance criteria as a result of international competitivenessThe world economy is also socially constructed just as are national

economies Even if different social systems of production are competingin the international arena it is not always possible to determine whichis more competitively effective at any moment in time Hegemonicnation-states in the short run shape the rules of trade to favor their indus-trial sectors and rms But the history of hegemonic powers suggeststhat in the longer run such social systems of production sustainedlargely by military and political power eventually give way to moreinnovative and competitive social systems of production In our ownday as nation-states are increasingly integrated into a world economyeven hegemonic powers lose their innovative and competitive advan-tage and their share of world output decreases Such a country mayattempt to restructure its institutional arrangements and to readjust itsperformance preferences But to restructure its system of innovationgenerally calls for a major redistribution of power within a societyLargely for this reason societies have historically had limited capacityto reconstruct a system of innovation in the image of their major competi-tors Each countryrsquos social system of production is a conguration of ahost of norms rules and values as well as of institutional arrangementsEach system is constantly changing and is open to inuence from othersystems And indeed many technologies and practices diffuse from onesociety to another but a societyrsquos capacity to be innovative is constrained

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

623

by the existing social system of production Thus the same technologymay exist in numerous countries but how it is employed and how itinuences societal outcomes and performances varies from one institu-tional conguration or society to another (Hollingsworth 1998)

PERSPECTIVES ON INSTITUTIONAL CHANGE

A fundamental problem which remains unresolved in institutionalanalysis involves the nature of institutional change There is much confu-sion and miscommunication in the social sciences about institutionalchange in large part because we do not even know how to measure therate of institutional change Of course we know that there are bothexternal and internal focuses for change and we have had a good bitof scholarship on this subject What is not clearly understood is thatinstitutions and institutional arrangements within societies are histori-cally rooted that there is a great deal of path-dependency to the waythat institutions evolve Moreover the more intricately linked each sectoris with each other sector and with a societyrsquos rules and norms the lesschoice actors have to devise new institutions and institutional arrange-ments On the other hand if a sector or actor is too isolated it may betoo weak within a society to bring about any kind of effective changeThese considerations are very critical to the problem of how muchfreedom actors have to create new institutions afresh to what extentinstitutions diffuse from one society to another (Murmann 1998 Lanzara1998) This is a fundamental question for students of institutionalanalysis as we attempt to understand how the different institutionallevels are linked togetherThere are a variety of reasons why there is confusion in studying insti-

tutional change There is disagreement over the extent to which actorshave the freedom to build new institutional arrangements and if theydo to what extent the new arrangements and organizational patternsmay depart from past practices Among some institutional economists(Williamson 1985) there is a basic assumption that actors build efcientinstitutions by pursuing their self-interest and by promoting theirstrategic goals Moreover there is the assumption that a mixing of fullrationality with market competition tends to produce all of the optimalinstitutions that are needed in order to coordinate a complex capitalisteconomy (Hollingsworth and Boyer 1997 ch 14 Hodgson 1999) Incontrast the argument developed here is much more complexSocial institutions are historically rooted and there is a great deal of

path-dependency in the way that various institutional componentsevolve The shape of institutional congurations at any moment limitsthe type of options for change Because there is a great deal of institu-tional inertia radical change in the institutional components of a society

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

624

is uncommon (Lanzara 1998) Recognizing the constraints existinginstitutions exert on actors we note that existing congurations of insti-tutional arrangements institutional sectors and organizational propertiescan limit but also inuence the degree to which norms and rules canchange Many scholars (North 1990 Ostrom 1990) have emphasized therole of norms rules ideologies and values in limiting the ability of actorsto develop new institutional arrangements which radically depart fromexisting ones However there is a two-way process of interaction aprocess of co-evolution (Baum and Singh 1994a Nelson 1994) Thus asFriedland and Alford (1991 244) argue actors develop new institutionalarrangements by recognizing organizational failures and low perfor-mance but these new arrangements do not necessarily change a societyrsquosnorms rules habits and its underlying value systemCampbell (1997) advances our understanding of how existing institu-

tions enable actors to construct radically new institutions He recognizesthat most institutions and institutional arrangements embody a degreeof inertia from which it is difcult for actors to depart Moreover thedifferential power relations among actors make it unlikely that lesspowerful actors can change the existing power structure of a societyMost change in the institutional components of a society evolves througha process of constrained selection which reects to a considerable degreethe existing arrangements and power relations This becomes the inter-pretative frame for social actors with institutionalized scripts and ritualswhich tend to be taken for granted and appear to be quite ordinary andnatural This institutionalized scripting of the social world is an extremelyimportant source of social stability and inertiaCampbell has developed a very fruitful explanation of change in terms

of the various institutional components outlined here He distinguishesbetween incremental and radical institutional change and argues thatradical institutional change occurs when social actors with widelydiffering norms cultural scripts and rituals for action engage in intenseinteraction with each other (see also Knight 1992) The more diverse theinteractions the greater the potential for institutional change He arguesthat as shifts in the composition of interaction among social actors occurchanges in the interpretation of problems and interests will follow Ifthere are only minor changes in the extent to which diverse actorsinteract there will be only modest institutional change But the morefundamental the changes in the interaction of diverse social actors thegreater the change in the way that actors interpret their world And thegreater the change in the way that actors interpret their world the morelikely that radical changes in various institutional components willfollow In other words diverse decision making tends to create a widerrange of interpretative frameworks of the social world than is likely tobe the case if social actors operate in isolation

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

625

This perspective of diverse interactions among social actors facilitatesour ability to move beyond the traditional view that existing institutionsconstrain the range of institutional alternatives which actors face Theolder view has tended to suggest that actors engage in institutionalchange by extending existing institutional principles habits and conven-tions Specifying the conditions which facilitate fundamental institutionalchange permits a much richer perspective for institutional analysisCampbellrsquos perspective about the conditions under which fundamentalor radical change in institutional components occurs is consistent withmy own research about the conditions under which major discoveriesor fundamental new knowledge occur the more scholars with diversebackgrounds interact with intensity and frequency the greater the like-lihood that they will develop new and alternative ways of thinking(Hollingsworth and Hollingsworth 2000)These views of institutional change are also consistent with the recent

work of Nelson and his colleagues (Mowery and Nelson 1999 Murmann1998) They report the research on a variety of industrial sectors in anumber of different countries revealing that networks of actors in partic-ular industries business and professional associations universityresearch departments and institutes are frequently able to use their collec-tive power to modify a societyrsquos property rights research system andgovernmental policies It is the intense and frequent interaction of thesediverse complex sets of actors ndash sometimes constituting a technologysystem ndash that often leads to incremental changes in a societyrsquos total socialsystem of production It is this kind of collective action that brings aboutorganizational change change in institutional sectors institutionalarrangements and even in the rules and norms of a society And if weare to understand how the style of innovativeness of a society changeswe must rst understand the changes in its institutional makeup It ishighly unlikely that a society will develop a new style of innovativenesswithout changing its institutional structure

FIRMS THEIR INSTITUTIONAL ENVIRONMENT ANDNATIONAL SYSTEMS OF INNOVATION

Variation in the innovative style of societies is shaped by their insti-tutional makeup ndash the various levels which are discussed aboveinstitutions institutional arrangements the structure and cohesion of institutional sectors which constitute a societyrsquos social system of pro-duction and the structure and culture of its organizations ndash especiallyits business rms and other research organizations By undertaking aninstitutional analysis of a society one can begin to understand in whatkinds of organizations the production of specic kinds of knowledgetakes place and how this is linked to particular kinds of innovativeness

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

626

With such a perspective one can gain a rich understanding of why somesocieties excel in the production of radical breakthroughs in basic andapplied science and in developing radically new products and why othersocieties excel in more incremental innovationsReecting on these issues one cannot help being impressed with the

contrasts between the innovative styles of German and American orga-nizations For example Germany has continued to be very successful in making incremental innovations in many industries in which it wasvery competitive well before World War II paper printing materialsmachinery electrotechnical products motor vehicles chemicals textileyarns and fabrics But the Germans have been much less successful indeveloping totally new products or related innovations in biotechnologyelectronics telecommunications aircraft as well as other newer indus-trial areas In contrast the Americans have been very innovative in these and other industries with products having very short lives andwith technologies which change very rapidly and are very complex(Hollingsworth 1991 Soskice 1997 1999)The importance of institutions and innovativeness for international

competitiveness has led to the recognition that there are national systemsof innovation (Lundvall 1992 Nelson 1993 Edquist 1997 Hage andHollingsworth 2000) But even before the recent studies of nationalsystems of innovation Landes (1969) was a proponent of these viewswhen he made the argument that Germany had developed a nationalsystem of innovation by building on its education and scientic researchsystems which were lacking in Britain and elsewhere However the concept of a national system of innovation immediately poses theproblem of what kind of innovations should be consideredDavid Soskice an economist at the Wissenschaftszentrum Berlin has

addressed a set of problems somewhat complementary to the perspec-tive raised here His work (Soskice 1990 1997 1999) suggests that weare sensitive to the institutional environment in which the followingactors are embedded the employees of rms the owners and nanciersof rms the competitors of rms and nally their collaborators ndash aperspective complementary to Hollingsworthrsquos work on social systemsof production (Hollingsworth 1997 1998) When one reects on the insti-tutional contextsocial system of production within which these actorsare embedded one becomes aware that those countries which have some-what inexible labor markets external to rms but a business system in which rms have long-term commitments to their employees arestrongly associated with incremental but hardly at all with radical inno-vations In the rms of such countries (eg Germany and Japan) thereis much more consensus decision making than in societies where a lsquohireand rersquo set of practices is quite pervasive And this kind of consensusdecision making ndash while highly conducive to incremental innovativeness

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

627

ndash limits the capacity for radical innovations to occur This kind of lsquolongtermismrsquo set of norms and rules occurs not only in the employmentpractices of incrementally innovative rms but also in their nancingand ownership patterns (Hollingsworth 1997) During the past thirtyyears these were characteristics of many rms in Germany and JapanSignicantly in both of these countries venture capital markets werenot highly developed nancing tended to be based more on long-termbank loans than on the equity markets and the relationships of rmstended to be relatively stable And these stable institutional arrange-ments also contributed to a high degree of innovativeness ndash butincremental rather than radical innovationsMeantime we should also pursue Soskicersquos ideas (1999) about how

rmsrsquo relationships with their competitors and collaborators shape theirstyles of innovativeness In those advanced industrial societies in whichbusiness rms are highly mobilized into business associations theretends to be a high degree of standardized quality controls over technicalnorms and a high degree of reliance on legal sanctions as monitoringinstruments vis-agrave-vis association members As a result of this high degreeof integration of rms into associations member rms are more likelyto be associated with incremental innovations than is likely to be the case with rms in those societies where associations are less welldeveloped In the latter type of society radical innovations are morelikely to occur In those advanced industrial societies where rms havelong-term stable relationships with collaborators in product design andproduction the style of innovation is likely to be incremental And it is in those societies in which there is a weaker tradition of strongcollaboration among rms that radical innovations are more likely tooccurObviously even in a country where there may be a great deal of

radical innovativeness most innovations will be of an incremental typefor there are simply fewer radical innovations But Germany and therest of Europe not only have had relatively few radical innovations incomparison with the United States but they also have lagged behindthe USA in developing relatively totally new industries Thus there hasbeen much less innovative activity in Europe in high-technology andnewly emerging industries than in the USAIn complementary research Audretsch and Feldman (1995) have

demonstrated that the USA has been more successful in spawning rmsengaged in the early stages of the industry life cycle but relative toGermany had an environment unfavorable to innovative activity byestablished rms especially in more mature industries German rmshave tended to be associated with a routinized technological regime ndashwhere established enterprises have had strong incentives to engage inincremental innovations but new rms have tended to have fundamental

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

628

innovative disadvantages relative to the Americans (see Audretsch 1995Winter 1982 Casper 2000)On both sides of the Atlantic universities engage in basic and applied

research But signicantly the institutional environment in which univer-sities are embedded inuences the degree to which the research is likelyto represent major breakthroughs or is likely to have more modestimplications Recent research in biomedical science demonstrates that inthose societies in which university professors andor graduate studentshave the following characteristics there are likely to be relatively majorresearch breakthroughs graduate students have a long training periodhigh dependency on a major professor low autonomy in dening thenature of a research project restricted job options and a university envi-ronment where there is a very modest or nonexistent entrepreneurialculture For example in Germany ndash a country with relatively few radicalbreakthroughs ndash the scholar who nishes a habilitation will have spentmany years on a research project before qualifying for a permanent acad-emic position will have been in a high-dependency position with a seniorprofessor will have had limited autonomy to dene a research projectand will be constrained from pursuing a high-risk research project Theyoung German will have an extraordinarily high level of training butwill have faced a number of disincentives to undertake high-risk researchAnd as soon as heshe has completed the habilitation and becomes aprofessor administrative responsibilities constitute another disincentivefrom engaging in high-risk or long-term research projects Moreover theGerman professor is a civil servant with a more stable and secure incomethan hisher American counterpart In a very different institutionalenvironment American professors in large research universities tend tobecome much more entrepreneurial Today many American professorsbehave as though they were operating their own rms within univer-sities and thus devote considerable time to generating funds for researchprojects assistants secretaries supplies etc And the variation in thisentrepreneurial spirit across countries is associated with variation in thetaking of high-risk research strategies and the making of major researchbreakthroughsIn sum societies vary in terms of the institutional environment in

which their organizations are embedded and this variation inuencesthe degree to which they have industrial activities in the early or latestages of the industry life cycle successful new or mature industriesand radical vs incremental innovations And it is this variation whichis encouraging rms in countries with relatively few new industries and with few or no radical innovations to invest in those countries with many new industries and numerous radical innovations ndash as astrategy to be able to transfer new technologies from the latter institu-tional environment to the former

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

629

Table 4 summarizes the way that the norms rules and values of a society are associated with Americarsquos and Germanyrsquos institutionalarrangements their social system of production and their styles of inno-vativeness

CONCLUSION

Several themes deserve to be reemphasized Most importantly institu-tional analysis is emerging as a dominant issue in several social sciencedisciplines The study of various institutional components discussedabove is a complex subject requiring multiple levels of analysis and aknowledge of multiple sectors within societies It requires a theoreticalperspective that evolves from an empirical analysis in different societiesIf we are to advance the study of institutional analysis we need cross-disciplinary collaboration and team research which are not readilyachieved given the structure of most of our research universitiesAt present scholars in each academic discipline have their own distinc-

tive strategies for studying institutional analysis Thus economists tendto use several approaches (North 1990 Williamson 1985 Hodgson 19881998 1999) anthropologists another (Geertz 1995) and sociologistspolitical scientists and historians employ other strategies (Campbell etal 1991 Powell and DiMaggio 1991 Katznelson 1998 Hall and Taylor1996) With practitioners in various disciplines pursuing their distinctiveapproaches to institutional analysis there is low potential for collabo-rative institutional analysis across the social sciences The lessons fromthe biological sciences over the last 40 years are very instructive oncepractitioners from various biological disciplines began to cooperate instudying biological phenomena by using the same concepts at the mole-cular level theoretical advance was very rapid Comparable phenomenatook place in the development of other hybrid elds eg biochemistrybiophysics (Judson 1979 Olby 1979 Kohler 1982 1994) Similarly weresocial scientists in various disciplines to work collaboratively in the areaof institutional analysis there would very likely be an acceleration intheir theoretical knowledgeThis is not a plea that everyone engaged in institutional analysis should

do the same kind of research Indeed just as biologists working at themolecular level study many different kinds of problems similarly socialscientists engaged in institutional analysis would also work on manyseparate problems but in the spirit of a collective enterprise Some wouldconcentrate their attention on the study of rules norms habits conven-tions and values while others would study how these are associatedwith congurations of various institutional arrangements (types ofmarkets hierarchies networks associations communities clans statesetc) Others would work on specic institutional sectors (eg education

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

630

Table 4

Rules norms hab

its an

d values associated w

ith sp

ecic co

ngu

ration

s of coo

rdinating mecha

nism

s soc

ial sy

stem

s of

prod

uction

an

d nationa

l styles of inno

vation

USA

Germany

Norms rules values which provide incentives for

Norms rules values which provide incentives for

short-term horizons

long-term horizons

Con

gu

rative

forms of

Marke

ts c

orpo

rate hierarchies reg

ulatory state

Assoc

iation

s (busine

ss assoc

iation

s la

bor union

s)

coordinating

corp

oratist-type

networks

marke

ts c

orporatist-

mecha

nism

s (listed in

type

state

order of im

portanc

e in

the particu

lar social

system

)

Social sys

tem of

prod

uction

Bus

iness system

Firm

struc

tures

Con

glom

erates b

ut with mov

emen

t toward

Firm

more co

ngru

ent with pa

rticular prod

ucts

elim

ination of produc

tion

in som

e sectors

emph

asis on long

-term strateg

ies rm

s are well

emph

asis on sh

ort-term

strateg

ies high

integrated

differentiation

between divisions

of rms

Firm

s mob

ilized in

Low

High

and in

tegrated

into busine

ss

associations

Indu

strial relations

Low

deg

ree of jo

b secu

rity rigid in

ternal lab

or

High deg

ree of jo

b secu

rity

exible interna

l lab

or

system

marke

t ex

ible externa

l labo

r marke

tsmarke

t rigid externa

l lab

or m

arke

tPatterns of

Dispersed owne

rship frequ

ent turnov

erMore stab

le owne

rship patterns

owne

rship

Table 4

continue

d

USA

Germany

Norms rules values which provide incentives for

Norms rules values which provide incentives for

short-term horizons

long-term horizons

Social system of

Prod

uction

Bus

iness system

Training for labo

r an

dHigh deg

ree of voluntary decision mak

ing by

Collective an

d m

ore co

mpulso

ry d

ecisions

abo

ut

man

agem

ent

individuals as to ho

w m

uch

tim

e an

d ene

rgy

how m

uch ed

ucation

is ne

eded

he

avy em

pha

sis

to inv

est in educ

ation heavy

emph

asis on

on techn

ical training (eg en

gine

ering)

busine

ss sch

ools training

in marke

ting

and

sales accou

nting

Fina

ncial marke

tsEqu

ity an

d ven

ture cap

ital m

arke

ts highly

Equ

ity an

d ven

ture cap

ital m

arke

ts w

eakly

dev

elop

ed na

ncing from

cen

tral gov

ernm

ent

deve

lope

d he

avy em

pha

sis on

corporate loan

sto priva

te rm

s for dev

elop

men

t of new

tech

nologies related

to na

tion

al security

Unive

rsity research

High individual auton

omy high poten

tial for

Low

ind

ividua

l au

tono

my until almost 40

yea

rs of

system

individual creativity u

nive

rsity pr

ofesso

rs

age poten

tial for ind

ividual creativity high

ly

high

ly entreprene

urial

cons

traine

d unive

rsity pr

ofessors not so

entrep

rene

urial

Inno

vative

styles

Highly su

ccessful in m

ajor break

through

s in

High pe

rforman

ce in proc

ess an

d inc

remen

tal

basic an

d app

lied scien

ce go

od in dev

elop

ing

inno

vation

s in existing produ

cts not so su

ccessful

totally

new

lines of pr

oducts an

d in m

ajor break

through

s in basic scien

ce or in

tech

nologies high

deg

ree of entrepr

eneu

rship

deve

loping

totally new

lines of prod

ucts of

amon

g research

ers

tech

nologies low d

egree of entrepr

eneu

rship

amon

g research

ers

Table 4

continue

d

USA

Germany

Norms rules values which provide incentives for

Norms rules values which provide incentives for

short-term horizons

long-term horizons

Prod

ucts in w

hich

Pr

oducts with sh

ort liv

es prod

ucts in w

hich

Older produ

cts with long

history machine

ry pap

er

society excels

tech

nology

is bo

th highly co

mplex

and

pr

inting

textile

s m

otor veh

icles m

achine

too

lsch

ange

s rapidly en

tertainm

ent indus

try

pha

rmaceu

ticals biotechn

olog

y softw

are

aerosp

ace

business systems nancial markets and systems of research) Distinctivewould be the recognition of how these separate studies are linkedtogether how each institutional sector is linked to a societyrsquos normsrules values etc and its conguration of institutional arrangementsSimilarly other institutional analysts might focus on the study of orga-nizations with a major concern for how the institutional environmentof organizations (eg norms rules values congurations of institutionalarrangements institutional sectors) inuenced the structure culture andoutcomes of organizations And nally others might study how all ofthe aforementioned aspects of institutional analysis inuenced a societyrsquosoverall performance Nelson and Sampat (1998) have appropriatelyreminded us that we should engage in institutional analysis in order toaddress specic problems but the view developed in this article is thatwe rst need some mapping of the terrain of institutional analysis inorder for our diverse enterprises to become a meaningful collective enter-prise especially if we are to understand how institutions inuence theinnovative processThe argument developed here is not that all social scientists should engage

in institutional analysis Obviously there are numerous other importantareas of research which lie beyond the eld of institutional analysis Nordoes this article argue that institutional analysis is the most importantsubject for social scientistsSince the structure and culture of our research universities reect an

enormous amount of disciplinary fragmentation what are the prospectsfor promoting the kind of collective and interdisciplinary research agendaproposed above In one respect the prospects are not encouragingDuring the past half century the efforts of American universities to have effective communication across the biological sciences have variedgreatly in their success At the University of California-Berkeley it was not until the 1980s that there was a serious effort to integrate thebiological sciences and at that time the campus abolished seventeendepartments to promote an integrated biological program But at theUniversity of Wisconsin the biological sciences are still very fragmentedinto dozens of different departments in botany plant pathology zoologyand physiology with two departments of biochemistry and two depart-ments of geneticsThere have been other approaches which have worked very well histor-

ically for promoting scientic integration amidst scientic diversity Forexample several summer schools and special institutes were organizedin the United States during the 1940s in an effort to overcome the frag-mentation in the biological sciences These programs made it possiblefor scientists in diverse academic disciplines and from different univer-sities and countries to come together to engage in intense and frequentinteraction Perhaps the most effective program was that held in the

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

634

summers at Cold Spring Harbor New York where a research agendafor the study of molecular biology was developed (Fischer and Lipson1988 Stent and Watson 1966) Scientists went there summer aftersummer and when they returned to their universities many began therestructuring of the biological sciences in their universitiesIn our own day we need summer institutes and other similar programs

so that scholars from diverse backgrounds can come together anddevelop a collective agenda for doing institutional analysis and forstudying the impact of institutions on the performance of societies ndashinnovativeness being one such subject If we can develop such instituteswe will increase the prospects for transforming our universities Thereis reason to be optimistic that once an eclectic group of scholars workwith a common set of concepts we can have genuine theoretical advanceif researchers can bring their diverse expertise to bear on a common setof problems There is considerable evidence that institutional analysisoffers high potential for advancing our knowledge of innovations andtechnological change (Hollingsworth and Hollingsworth 2000) If we aregoing to make substantial theoretical advance it is not likely to occurfrom the vantage point of a multiplicity of disciplines Rather it is likelyto occur as we develop a new intellectual framework and our own acad-emic journals This is a tall agenda which requires us to reassess theway that knowledge is presently produced in our societies

NOTE

This article is part of a much larger agenda involving the study of institutionsorganizations and innovations in which I am involved My debts are many andvaried The rst version was written while I was in residence in the ResearchUnit for Institutional Change and European Integration of the Austrian Academyof Sciences in Vienna in the summer of 1998 I am especially grateful to ProfessorsEgon Matzner and Sonja Puntscher-Riekmann for providing the stimulating envi-ronment in which to think through the ideas developed here Egon Matzner hasengaged in very stimulating discussions with me about the importance of insti-tutions in the general eld of socio-economics The nal version of the articlewas completed during my residence as a fellow at the Neurosciences Institutein La Jolla California I also wish to thank my colleagues in the InnovationTheme Project at the Netherlands Institute for Advanced Study (NIAS) during1998ndash9 and Edgar Grande and Karl Muumlller for stimulating conversations aboutthe interaction of institutions and organizations in facilitating particular typesof innovation I am especially grateful to Jerald Hage for helping me to developa number of issues in the article and for his useful comments on the rst draftFrans van Waarden early on helped me to work through the idea of multiplelevels of institutional analysis Others who have helped to develop the ideas inthe article have been my colleagues in the research project lsquoComparative socialsystems of productionrsquo especially Tom Burns Christel Lane Yoshitaka OkadaWolfgang Streeck and Richard Whitley My colleagues Robert Boyer SteveCasper and David Soskice have taught me much about how incentives providedby the institutional environments of rms inuence individual decision making

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

635

But my most important debt is to Ellen Jane Hollingsworth who has repeatedlyquestioned every single assumption of this article Without her rich and contin-uous tutoring this article would be even less developed than it is at presentFinally without the insightful assistance of David Gear in many phases of thisarticle it would not have been completed

REFERENCES

Aoki Masahiko (1990) lsquoTowards an economic model of the Japanese rmrsquo Journalof Economic Literature March 1ndash27

Arora A Landau Ralph and Rosenberg Nathan (eds) (1998) Chemicals and Long-Term Economic Growth Insights from the Chemical Industry New York JohnWiley amp Sons

Arrow Kenneth (1974) The Limits of Organization New York NortonArthur W B (1988a) lsquoSelf-reinforcing mechanisms in economicsrsquo in P W

Anderson K J Arrow and D Pines (eds) The Economy as an Evolving ComplexSystem Redwood City Calif Addison-Wesley

mdashmdash(1988b) lsquoCompeting technologies an overviewrsquo in G Dosi C Freeman RNelson G Silverberg and L Soete (eds) Technical Change and Economic TheoryLondon Pinter Publishers pp 590ndash607

Audretsch David B (1995) Innovation and Industry Evolution Cambridge MassMIT Press

Audretsch David B and Feldman Maryann P (1995) lsquoInnovative clusters andthe industry life cyclersquo Discussion Paper Wissenschaftszentrum BerlinMarket Processes and Corporate Development Research Unit FS-IV-95-7

Baum Joel A C and Oliver Christine (1992) lsquoInstitutional embeddedness andthe dynamics of organizational populationsrsquo American Sociological Review 57540ndash59

Baum Joel A C and Singh Jitendra V (1994a) lsquoOrganizationndashenvironment coevo-lutionrsquo in Joel A C Baum and Jitendra V Singh (eds) The EvolutionaryDynamics of Organizations New York Oxford University Press pp 379ndash402

mdashmdash(eds) (1994b) The Evolutionary Dynamics of Organizations New York OxfordUniversity Press

Berger Suzanne and Dore Ronald (eds) (1996) National Diversity and GlobalCapitalism Ithaca NY Cornell University Press

Boyer Robert (1997) lsquoThe variety and unequal performance of really existingmarkets farewell to Doctor Panglossrsquo in J Rogers Hollingsworth and RobertBoyer (eds) Contemporary Capitalism the Embeddedness of InstitutionsCambridge and New York Cambridge University Press pp 55ndash93

Burns Tom R and Dietz Thomas (2001 forthcoming) lsquoSocial rule system theorysocial action institutional arrangements and evolutionary processesrsquo in JRogers Hollingsworth and Karl Muumlller (eds) Advances in SocioeconomicsLanham Md Rowman amp Littleeld

Burns Tom R and Flam Helena (1987) The Shaping of Social Organization BeverlyHills Calif Sage

Calvert Randall L (1995) lsquoThe rational choice theory of social institutionsrsquo inJeffrey S Banks and Eric A Hanuskek (eds) Modern Political Economy NewYork Cambridge University Press pp 216ndash66

Camic Charles (1986) lsquoThe matter of habitrsquo American Journal of Sociology 911039ndash87

Campbell John L (1997) lsquoMechanisms of evolutionary change in economicsgovernance interaction interpretation and bricolagersquo in Lars Magnusson

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

636

all of these components inuence the performance of economic sectorswithin a society as well as the performance of the total society A socialsystem of production is the way that a societyrsquos institutions (see rstlevel of analysis above) its institutional arrangements (see second levelof analysis above) and its institutional sectors are integrated into a socialcongurationAn institutional sector includes all organizations in a society which

supply a given service or product along with their associated focal orga-nizations (eg major suppliers funding sources regulators and so forth(Scott et al 1994 108 117 Campbell et al 1991 Hollingsworth et al1994) Institutional sectors include but are not limited to the societyrsquossystem of education system of research business system nancialmarkets legal system and the state The structure of the nancial marketssystem of training and education industrial relations system system ofresearch and state and legal systems are distinct and idiosyncratic ineach society In short they are system specicAll of these institutional sectors tend to cohere with each other

although they vary in the degree to which they are tightly coupled intoa full-edged system While each of the institutional sectors has someautonomy and may have some goals that are contradictory to the goalsof other institutional sectors with which it is linked an institutional logicin each society leads institutions to coalesce into a complex social con-guration (Hollingsworth 1991) This occurs because the institutionalsectors are embedded in a culture in which their logics are symbolicallygrounded organizationally structured technically and materially con-strained and politically defended The conguration of institutionalsectors usually exhibits some degree of adaptability to new challengesbut continues to evolve within an existing style But under new circum-stances or unprecedented disturbances these institutional congurationsare exposed to sharp historical limits as to what they may or may notdo (Schumpeter 1983 David 1988 Arthur 1988a 1988b Haringkanssen andLundgren 1997)Why do all of these different institutions coalesce into a complex

social conguration which is labeled here as a social system of pro-duction The literature suggests two contrasting interpretations Part ofthe answer ndash indeed a controversial one ndash is that these institutionalsectors are functionally determined by the requirements of the practiceof capitalism in each time and place (Habermas 1975) Another expla-nation emphasizes the genesis of the actual conguration via a trial and error process according to which the survival of rms regions orcountries is the outcome of complex evolutionary mechanisms (Maynard-Smith 1982 Nelson and Winter 1982) However the problem is evenmore complex Markets and other mechanisms for coordinating rela-tionships among economic actors place constraints on the means and

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

614

ends of economic activity to be achieved in any society The othercoordinating mechanisms include different kinds of hierarchies varioustypes of networks and associations ndash trade unions employers and busi-ness associations (Hollingsworth and Lindberg 1985 Campbell et al1991) The logic of these various institutional arrangements providesactors with vocabularies and logics for pursuing their goals for deningwhat is valued and for shaping the norms and rules by which theyabide (see Tables 2 and 3) In short in contrast to the logic of theneoclassical paradigm the argument here is that the dominant type ofinstitutional arrangements places severe constraints on the denition of needs preferences and choices of economic actors Whereas the neo-classical paradigm assumes that individuals and rms are sovereign thisarticle is based on the assumption that rms are inuenced by the hold that the institutional congurations making up a social system ofproduction have on individual decision making (Campbell et al 1991Etzioni 1988 Streeck and Schmitter 1985a Hollingsworth et al 1994Hollingsworth and Boyer 1997 Magnusson and Ottosson 1997 North1990 Hodgson 1999)Standard neoclassical economic theory has tended to downplay the

role of production and consequently of rms Even transaction cost theo-rists who are concerned with analyzing the rm as a coordinatingmechanism have been relatively unconcerned with the various compo-nents of a social system of production Indeed as long as there waswidespread optimism about the efcacy of Keynesian economics therewas relatively little concern among neoclassical economists with thesupply side of the economy Even in the opinion of most Keynesians agroup of experts should ideally be able to shape the size of aggregatedemand while the supply side of the economy would be left to the twominimalist institutions of neoclassical economics ndash markets and manage-rial hierarchies For several decades however it has become increasinglyobvious that some of the most competitive and successful patterns ofindustrial output and industrial production in capitalist economies donot derive from the neoclassical prescription of unregulated markets andcorporate hierarchies complemented by a neoliberal democratic stateIndeed it is now well understood that certain highly successful produc-tion patterns require for their emergence and survival institutionalarrangements and environments the very opposite of the prescriptionsfound in the neoclassical paradigm (see especially Streeck 1992 but alsoHollingsworth and Streeck 1994) Thus social scientists have come tounderstand that if they are to comprehend the behavior and performanceof contemporary economies concerns about social systems of produc-tion must be brought into the pictureBecause production involves more than technology a number of social

scientists have an increasing concern with social systems of production

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

615

The same equipment is frequently operated quite differently in the samesectors in different countries even when rms are competing in the same market (Maurice et al 1980 Sorge and Streeck 1988 Sorge 1989Hollingsworth et al 1994) Variations in production and process tech-nologies are inuenced partly by variations in the social environmentsin which they are embedded In other words rms are embedded intocomplex environments which among other things place constraints ontheir behavior Thus a social system of production is of major impor-tance in understanding the behavior and performance of an economyHow the state and other institutional arrangements (eg marketsnetworks private hierarchies associations) coalesce and are related toparticular social systems of production are important determinants ofeconomic performanceThe congurations of institutional arrangements that coordinate or

govern the behavior of actors in one society and its structure of specicinstitutional sectors cannot easily be transferred to another society forthey are embedded into a social system of production that is societallydistinct (Hollingsworth 1997) Societies borrow selected principles offoreign management styles and work practices but the effectiveness ofsuch borrowing is generally limited Economic behavior and performanceare shaped by the entire social system of production in which actors areembedded and not simply by specic principles of particular manage-ment styles and work practices Moreover a societyrsquos social system ofproduction tends to limit the kind of goods that it can produce and withwhich it can compete successfully in international marketsSome scholars (Kenney and Florida 1993 Oliver and Wilkinson 1988)

have assumed that the diffusion of particular forms of management stylesand work practices across societies could lead to system convergenceReferring to the Japanese production system Kenney and Florida haveargued that it consists of organizational practices whose fundamentallsquogenetic codersquo can successfully be inserted into another society and beginsuccessfully to reproduce its behavior in the new environment (1993 8)Their position is in the intellectual tradition of Antonio Gramsci (1971)who contended decades ago that the American system of mass produc-tion would most certainly diffuse to Europe over timeHowever the argument of this article is that even though British

French and American rms may adopt certain aspects of Japanesemanagement styles (eg just-in-time production self-managing teamsquality circles the use of statistical process controls) or some variant ofthe German vocational system their social systems of production willnot converge Moreover the overall conguration of a societyrsquos socialsystem of production inuences its sectoral and overall national economicperformance This helps to explain why societies have different systemsof innovation and why they vary in the clusters of industries in which

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

616

they are highly competitive in the international markets (Nelson 19931999 Hage and Hollingsworth 2000 Berger and Dore 1996)In order to understand how and why a societyrsquos economy performs

as it does it is necessary to understand its entire social system of produc-tion If a society is to modify substantially the performance of itsinstitutional sectors and its entire economy it cannot adapt only someof the management and work practices of its foreign competitors Ratherit must alter its entire social system of production Because a societyrsquosmodes of economic governance and coordination and its institutionalsectors develop according to particular logics and are system specicthere are serious limitations to the extent to which a society may mimicthe institutions (eg the rules and norms) institutional arrangements andinstitutional sectors of other societiesIn the history of modern societies there are logics by which institu-

tions coalesce into a social system of production (Hollingsworth 1991)Though institutions are constantly changing there are sharp limits tothe type and the direction of change that any particular institution canundergo because of its linkages with institutional arrangements and insti-tutional sectors Thus a societyrsquos business rms educational systemnancial markets industrial relations system etc can engage in seriousrestructuring only if most of the other institutional sectors also changeIn social systems of production there are pressures toward consistencyin the norms rules and values across institutional sectors though in anycomplex society social systems are obviously imperfectly integratedIndeed the degree to which the institutional norms and rules makingup a social system of production are loosely or tightly coupled is a vari-able of considerable importance In general the institutional sectorsmaking up a social system of production are interdependent and changesin one generally result in changes in the others Since each institutionallevel is dependent on the others for various types of resources there isinterdependence among the differing institutional spheres Moreovereach society has its norms moral principles rules and laws and recipesfor action as well as its own idiosyncratic customs traditions and prin-ciples of justice (Burns and Flam 1987)There are also other inherent obstacles to convergence among social

systems of production for where a system is at any one point in timeis inuenced by its initial state Systems with quite different initial statesare unlikely to converge with one anotherrsquos institutional practicesExisting institutional arrangements block certain institutional innova-tions and facilitate others (Roland 1990) Thus the institutions makingup a social system of production provide continuity even though insti-tutional arrangements are always changing but with a particular logicWhile Williamson (1975 1985) suggests that actors tend to choose theinstitutional arrangements which are most efcient North (1990) is

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

617

much closer to the mark in his argument that most societal institutionalarrangements exist as a result of custom and habit and are thereforeinefcient At any moment in time the world often appears to its actorsas very complex and uncertain For this reason actors often engage incontradictory forms of behavior pursuing different strategies as hedgesagainst a very uncertain world (Lanzara 1998) And their hedging andcontradictory forms of behavior may lead in somewhat different soci-etal directions all constrained by the institutional fabric within whichthe actors are embedded This kind of contradictory behavior occurs inpart because of the contradictions inherent in the dominant congura-tions of institutional arrangements with their contradictory logics (seelsquoSecond level of analysisrsquo above)Despite the emphasis on the logic of institutional continuity this is

not an argument that systems change along some predetermined pathThere are critical turning points in the history of highly industrializedsocieties but with the choices limited by the existing institutional terrainBeing path-dependent a social system of production continues along aparticular logic until or unless a fundamental societal crisis intervenes(Milgrom et al 1991 Krugman 1991 Durlauf 1991 Hollingsworth 1991Pred 1966 David 1988 Hodgson 1998)At this point it is important to confront the question of how social

systems of production evolve And this question gets to the heart of theproblem of building complex societal structures Certainly social systemsof production did not emerge from some process of social engineeringMoreover the various component parts of each social system of produc-tion have often not been designed to be part of a social system ofproduction The component parts of each social system of productionhave emerged more often than not as unintended by-products of goalswhich various actors had in mind at earlier moments in time It is usuallythe case that actors at time lsquotrsquo fail to comprehend the long-term conse-quences of their actions Wolfgang Streeck the author of several brilliantpapers on social systems of production argues that institutional sectordesigns created for one purpose generally address that goal but overtime those designs have quite unintended consequences (Streeck 1997a1997b Wright 1998) Describing the conguration of the German socialsystem of production of the 1970s and 1980s Streeck points out that itwas the unintended by-product of multiple points in history Someelements were pre-Wilhelmian others were introduced by the Alliedpowers after 1945 and others emerged during the years of the GermanFederal Republic All component parts of the German social system ofproduction

were and continue to be changing for their own reasons as wellas in reaction to each other and certainly there can be no presump-tion of a pre-established t between them even though one might

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

618

want to allow for some reinforcement effects of the modelrsquos histor-ically contingent social and economic success

(Streeck 1997b 54)

In sum the emergence of social systems of production is a long-termevolutionary process with each part interacting with its environment andresulting in a congurative whole but with no previous design by eithera single actor or a collectivity of actors As suggested above actors do not recreate their world anew At any moment in history the totalinstitutional complex of a society necessarily contains the resources and legacies of its past Preexisting institutional complexes are nevercompletely wiped out Societies consist of multiple layers of history withtheir diverse logics of action However the cumulative effect of smallperipheral changes in altering particular institutional sectors and insti-tutional arrangements can be substantial (Murmann 1998 Lanzara 1998)lsquoInstitution building is affected more by the ways in which people havecodied the past than by how they envision the future There are manymore potential resources in pre-existing institutional arrangements thanit is usually assumed or suspectedrsquo (Lanzara 1998 30)

Fourth level of analysis

The next level of institutional analysis organizational structures is some-what more controversial As suggested above Douglass North (1990)draws a sharp distinction between institutions and organizations Morerecently many organizational theorists (Powell 1991 Powell andDiMaggio 1991 DiMaggio and Powell 1983 Baum and Oliver 1992Kondra and Hinings 1998 Townley 1997) argue that institutional rulesnorms and conventions unfold in tandem with organizational structuresand this is my positionThe literature which focuses on how institutions inuence organiza-

tions is quite different from two other theoretical literatures which alsoare concerned with how organizational environments shape the behaviorof organizations There are resource dependency theorists (Pfeffer andSalancik 1978) who emphasize the role of environmental resources inshaping organizations while Hannan and Freeman (1977 1984) and otherpopulation ecologists emphasize the survival of organizations givencertain kinds of organizational conditions (Orru et al 1991)The institutionalist perspective as implied above emphasizes the

normative environment in which organizations are embedded It is a perspective which focuses on the way in which organizations in their behavior tend to conform to the institutional rules and normswhich are dominant in the organizational environment However allthree perspectives ndash the institutionalist the population ecology and

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

619

the resource dependency ndash emphasize how the environment inuencesorganizations and how organizations which are subject to the sameenvironment tend to converge in their behavior to have what DiMaggioand Powell have labeled organizational isomorphism (DiMaggio andPowell 1983)There has emerged a vast literature which demonstrates that within

each society there are scal political judicial and other regulatory normswhich limit and shape the culture and structure of organizationalbehavior The normative institutional environment of organizations limits the options of what organizations do in a particular society and inuences the patterns of ownership relations with suppliers andcustomers In short it is the normative environment of organizationswhich denes within a particular society what is socially acceptablebehavior for organizations (Hamilton and Biggart 1988 Hollingsworthand Hollingsworth 2000 Meyer and Rowan 1991 Zucker 1987 1988Orru et al 1991 Townley 1997)Thus far most of the analyses on isomorphic pressures operating on

organizations have been cross-sectional in nature However the entireconcept of isomorphism implies that there are strong environmental pres-sures exerted on organizations and in order to observe this phenomenonit is important that we have longitudinal studies which assess howchanges in an organizationrsquos institutional environment inuence changesin the structure and culture of organizations Moreover the expectationhere is that when one analyzes the patterns of organizational changeover time the isomorphism will be holistic ndash that is within a particularsociety there will be a tendency for the internal structure of particulartypes of organizations to converge in large part because of the pres-sures to conform to the changes in the external norms rules and valuesA close reading of Aoki (1990) and Slack and Hinings (1994) nds thatchanges in the national institutional environment of organizations haveinuenced changes in the internal structure of organizations within aspecic industry of a particular societyHowever it is inappropriate to view organizations as changing only

in relation to exogenous environmental change There is a kind of co-evolution which occurs between organizations and their institutionalenvironments Co-evolution implies nonlinear feedback between orga-nizations and their institutional environments This is of course quitedifferent from the more common way of studying organizations bymodeling relations between independent and dependent variables Ofcourse the more common type of analysis is quite appropriate whenthere are simple relationships which do not involve complex feedbackprocesses But in the strategy for institutional analysis which is proposedherein it is less useful to separate independent from dependent vari-ables and more useful to understand the interacting and co-evolutionary

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

620

processes among all of the levels of institutional analysis discussed inthe various sections above (Baum and Singh 1994a 379ndash80 Mowery andNelson 1999 Murmann 1998)The logic of the perspective herein suggests that within every

society there is variation in the structure and culture of business rmsuniversities and other complex organizations (Kondra and Hinings1998) However this variation takes place within parameters which aresystem specic For example every German rm and every Germanuniversity is different from every other German university and rm but there are a set of cultural and structural characteristics which distin-guish German rms and universities from those in America Similarlyeach Japanese rm and university is unique yet they have a set ofcultural and structural characteristics which are system specic andwhich differentiate them from business rms and universities in other countriesIn all societies each organization has its own distinctive organizational

rules norms and conventions which are subordinate to the meta norms and rules of the larger society within which they are embeddedHowever the strength of the institutional environment within whichorganizations are embedded varies from society to society Some societies have multiple institutional environments and there is hetero-geneity in terms of what constitutes appropriate behavior oforganizations In such societies there can be conict over whatinstitutional logics should regulate specic organizational functions(Friedland and Alford 1991 Townley 1997 262ndash3) In those societies inwhich the institutional norms habits and rules are most developed and in which the institutional pressures to conform are greatest thereis less variation in the structure and culture of business rms and variouskinds of research organizations In such societies the connectednessbetween research organizations and their institutional environments is sufciently strong for organizations to have low autonomy to pursue independent strategies and goals in these societies there is a great deal of organizational isomorphism Conversely the weaker the institutional environment in which research organizations areembedded the greater the variation in the structure and culture of business rms and research organizations Moreover where the institu-tional environments are more weakly developed organizations havegreater autonomy and exibility to respond to the development of new knowledge and to be highly innovative Hence it is in those societies where the institutional environments are most developed and most rigid and when there is less organizational autonomy and exibility that fewer radical innovations in basic and applied science as well as in totally new products and industrial sectors havebeen made

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

621

Fifth level of analysis

We come next to the outputs and performance of the various institutionalcomponents of a society It is at this level that institutional componentsare more pragmatic and exible For example in the legal sector thereare specic statutes and court rulings in the state sector there are specicpolicies in the business sector there are new products new technolo-gies and market strategies It is at this nal level of analysis thatinstitutional spheres are most open and susceptible to change and cross-national mimicry is easiest and most commonThrough the outputs of the society we can obtain some assessment as

to how well a society is performing Moreover we can assess how inno-vative it is how egalitarian it is in the distribution of its resources howegalitarian it is in terms of levels of health education etc Just as it iscomplicated to measure how egalitarian a society is similarly it is dif-cult to assess how innovative it is For example societies vary in makingincremental and radical breakthroughs in basic and applied science in developing totally new products and new kinds of organizations inincremental and process innovations in existing products and organiza-tions in developing new and different forms of marketing ndash bothdomestically and globally (Hage and Hollingsworth 2000) Of coursesocieties vary in their rates of economic growth in their rates of economicproductivity in their quality of life (eg rates of crime life expectancyetc) All of these performance criteria feed back and inuence each ofthe levels of institutional analysis discussed above ndash rules norms valuesetc institutional arrangements institutional sectors the structure andform of organizations Moreover performance measures may inuenceeach other For example if a society has very low rates of growth itmay not be very innovative in some of the types of innovative activitiesmentioned above Good or poor performances of certain kinds inuenceother performance indicators (Hollingsworth et al 1990)Different institutional arrangements and different social systems of

production result in different types of economic performance Hence solong as societies have different social systems of production there areserious constraints on the degree to which they can converge in theirinnovative styles Different social systems of production tend to maxi-mize in a more or less explicit manner different performance criteriausually mixing considerations about static and dynamic efciency protsecurity social peace and economic andor political power In short incontrast to the implications of neoclassical economic theory in real-worldeconomies there are no universal standards all economically rationalactors attempt to maximize Economic history provides numerous exam-ples of how a great variety of principles of rationality are implementedin different societies

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

622

A critical question in institutional analysis is whether an existing socialsystem of production which supports a set of routines for a particularkind of technology and industry can shift from old practices to newones There are numerous historical accounts of how the social systemof production in a particular society which worked so well for a numberof years could not adapt to new technologies (Lazonick 1990 Veblen1915 Nelson 1994) Schumpeter (1983) and Freeman (1991) have devel-oped the idea of a techno-economic paradigm which suggests thatdifferent technologies require specic forms of organizational arrange-ments ndash or what I called social systems of production Because oflsquoinstitutional inertiarsquo the social system of production of a society maynot be able to adapt to a new techno-economic paradigm (Nelson 199458 Lanzara 1998)Whether a social system of production can sustain its performance

standards depends not only on its intrinsic economic rationality but alsoon where it ts into a larger system If a particular social system ofproduction is immune from the innovativeness and competition of analternative system survival can be long lasting But if different socialsystems of production with diverging criteria of good economic perfor-mance meet in the world arena the arbitrariness of nationally imposedconstructed performance standards may be superseded by alternativeperformance criteria as a result of international competitivenessThe world economy is also socially constructed just as are national

economies Even if different social systems of production are competingin the international arena it is not always possible to determine whichis more competitively effective at any moment in time Hegemonicnation-states in the short run shape the rules of trade to favor their indus-trial sectors and rms But the history of hegemonic powers suggeststhat in the longer run such social systems of production sustainedlargely by military and political power eventually give way to moreinnovative and competitive social systems of production In our ownday as nation-states are increasingly integrated into a world economyeven hegemonic powers lose their innovative and competitive advan-tage and their share of world output decreases Such a country mayattempt to restructure its institutional arrangements and to readjust itsperformance preferences But to restructure its system of innovationgenerally calls for a major redistribution of power within a societyLargely for this reason societies have historically had limited capacityto reconstruct a system of innovation in the image of their major competi-tors Each countryrsquos social system of production is a conguration of ahost of norms rules and values as well as of institutional arrangementsEach system is constantly changing and is open to inuence from othersystems And indeed many technologies and practices diffuse from onesociety to another but a societyrsquos capacity to be innovative is constrained

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

623

by the existing social system of production Thus the same technologymay exist in numerous countries but how it is employed and how itinuences societal outcomes and performances varies from one institu-tional conguration or society to another (Hollingsworth 1998)

PERSPECTIVES ON INSTITUTIONAL CHANGE

A fundamental problem which remains unresolved in institutionalanalysis involves the nature of institutional change There is much confu-sion and miscommunication in the social sciences about institutionalchange in large part because we do not even know how to measure therate of institutional change Of course we know that there are bothexternal and internal focuses for change and we have had a good bitof scholarship on this subject What is not clearly understood is thatinstitutions and institutional arrangements within societies are histori-cally rooted that there is a great deal of path-dependency to the waythat institutions evolve Moreover the more intricately linked each sectoris with each other sector and with a societyrsquos rules and norms the lesschoice actors have to devise new institutions and institutional arrange-ments On the other hand if a sector or actor is too isolated it may betoo weak within a society to bring about any kind of effective changeThese considerations are very critical to the problem of how muchfreedom actors have to create new institutions afresh to what extentinstitutions diffuse from one society to another (Murmann 1998 Lanzara1998) This is a fundamental question for students of institutionalanalysis as we attempt to understand how the different institutionallevels are linked togetherThere are a variety of reasons why there is confusion in studying insti-

tutional change There is disagreement over the extent to which actorshave the freedom to build new institutional arrangements and if theydo to what extent the new arrangements and organizational patternsmay depart from past practices Among some institutional economists(Williamson 1985) there is a basic assumption that actors build efcientinstitutions by pursuing their self-interest and by promoting theirstrategic goals Moreover there is the assumption that a mixing of fullrationality with market competition tends to produce all of the optimalinstitutions that are needed in order to coordinate a complex capitalisteconomy (Hollingsworth and Boyer 1997 ch 14 Hodgson 1999) Incontrast the argument developed here is much more complexSocial institutions are historically rooted and there is a great deal of

path-dependency in the way that various institutional componentsevolve The shape of institutional congurations at any moment limitsthe type of options for change Because there is a great deal of institu-tional inertia radical change in the institutional components of a society

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

624

is uncommon (Lanzara 1998) Recognizing the constraints existinginstitutions exert on actors we note that existing congurations of insti-tutional arrangements institutional sectors and organizational propertiescan limit but also inuence the degree to which norms and rules canchange Many scholars (North 1990 Ostrom 1990) have emphasized therole of norms rules ideologies and values in limiting the ability of actorsto develop new institutional arrangements which radically depart fromexisting ones However there is a two-way process of interaction aprocess of co-evolution (Baum and Singh 1994a Nelson 1994) Thus asFriedland and Alford (1991 244) argue actors develop new institutionalarrangements by recognizing organizational failures and low perfor-mance but these new arrangements do not necessarily change a societyrsquosnorms rules habits and its underlying value systemCampbell (1997) advances our understanding of how existing institu-

tions enable actors to construct radically new institutions He recognizesthat most institutions and institutional arrangements embody a degreeof inertia from which it is difcult for actors to depart Moreover thedifferential power relations among actors make it unlikely that lesspowerful actors can change the existing power structure of a societyMost change in the institutional components of a society evolves througha process of constrained selection which reects to a considerable degreethe existing arrangements and power relations This becomes the inter-pretative frame for social actors with institutionalized scripts and ritualswhich tend to be taken for granted and appear to be quite ordinary andnatural This institutionalized scripting of the social world is an extremelyimportant source of social stability and inertiaCampbell has developed a very fruitful explanation of change in terms

of the various institutional components outlined here He distinguishesbetween incremental and radical institutional change and argues thatradical institutional change occurs when social actors with widelydiffering norms cultural scripts and rituals for action engage in intenseinteraction with each other (see also Knight 1992) The more diverse theinteractions the greater the potential for institutional change He arguesthat as shifts in the composition of interaction among social actors occurchanges in the interpretation of problems and interests will follow Ifthere are only minor changes in the extent to which diverse actorsinteract there will be only modest institutional change But the morefundamental the changes in the interaction of diverse social actors thegreater the change in the way that actors interpret their world And thegreater the change in the way that actors interpret their world the morelikely that radical changes in various institutional components willfollow In other words diverse decision making tends to create a widerrange of interpretative frameworks of the social world than is likely tobe the case if social actors operate in isolation

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

625

This perspective of diverse interactions among social actors facilitatesour ability to move beyond the traditional view that existing institutionsconstrain the range of institutional alternatives which actors face Theolder view has tended to suggest that actors engage in institutionalchange by extending existing institutional principles habits and conven-tions Specifying the conditions which facilitate fundamental institutionalchange permits a much richer perspective for institutional analysisCampbellrsquos perspective about the conditions under which fundamentalor radical change in institutional components occurs is consistent withmy own research about the conditions under which major discoveriesor fundamental new knowledge occur the more scholars with diversebackgrounds interact with intensity and frequency the greater the like-lihood that they will develop new and alternative ways of thinking(Hollingsworth and Hollingsworth 2000)These views of institutional change are also consistent with the recent

work of Nelson and his colleagues (Mowery and Nelson 1999 Murmann1998) They report the research on a variety of industrial sectors in anumber of different countries revealing that networks of actors in partic-ular industries business and professional associations universityresearch departments and institutes are frequently able to use their collec-tive power to modify a societyrsquos property rights research system andgovernmental policies It is the intense and frequent interaction of thesediverse complex sets of actors ndash sometimes constituting a technologysystem ndash that often leads to incremental changes in a societyrsquos total socialsystem of production It is this kind of collective action that brings aboutorganizational change change in institutional sectors institutionalarrangements and even in the rules and norms of a society And if weare to understand how the style of innovativeness of a society changeswe must rst understand the changes in its institutional makeup It ishighly unlikely that a society will develop a new style of innovativenesswithout changing its institutional structure

FIRMS THEIR INSTITUTIONAL ENVIRONMENT ANDNATIONAL SYSTEMS OF INNOVATION

Variation in the innovative style of societies is shaped by their insti-tutional makeup ndash the various levels which are discussed aboveinstitutions institutional arrangements the structure and cohesion of institutional sectors which constitute a societyrsquos social system of pro-duction and the structure and culture of its organizations ndash especiallyits business rms and other research organizations By undertaking aninstitutional analysis of a society one can begin to understand in whatkinds of organizations the production of specic kinds of knowledgetakes place and how this is linked to particular kinds of innovativeness

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

626

With such a perspective one can gain a rich understanding of why somesocieties excel in the production of radical breakthroughs in basic andapplied science and in developing radically new products and why othersocieties excel in more incremental innovationsReecting on these issues one cannot help being impressed with the

contrasts between the innovative styles of German and American orga-nizations For example Germany has continued to be very successful in making incremental innovations in many industries in which it wasvery competitive well before World War II paper printing materialsmachinery electrotechnical products motor vehicles chemicals textileyarns and fabrics But the Germans have been much less successful indeveloping totally new products or related innovations in biotechnologyelectronics telecommunications aircraft as well as other newer indus-trial areas In contrast the Americans have been very innovative in these and other industries with products having very short lives andwith technologies which change very rapidly and are very complex(Hollingsworth 1991 Soskice 1997 1999)The importance of institutions and innovativeness for international

competitiveness has led to the recognition that there are national systemsof innovation (Lundvall 1992 Nelson 1993 Edquist 1997 Hage andHollingsworth 2000) But even before the recent studies of nationalsystems of innovation Landes (1969) was a proponent of these viewswhen he made the argument that Germany had developed a nationalsystem of innovation by building on its education and scientic researchsystems which were lacking in Britain and elsewhere However the concept of a national system of innovation immediately poses theproblem of what kind of innovations should be consideredDavid Soskice an economist at the Wissenschaftszentrum Berlin has

addressed a set of problems somewhat complementary to the perspec-tive raised here His work (Soskice 1990 1997 1999) suggests that weare sensitive to the institutional environment in which the followingactors are embedded the employees of rms the owners and nanciersof rms the competitors of rms and nally their collaborators ndash aperspective complementary to Hollingsworthrsquos work on social systemsof production (Hollingsworth 1997 1998) When one reects on the insti-tutional contextsocial system of production within which these actorsare embedded one becomes aware that those countries which have some-what inexible labor markets external to rms but a business system in which rms have long-term commitments to their employees arestrongly associated with incremental but hardly at all with radical inno-vations In the rms of such countries (eg Germany and Japan) thereis much more consensus decision making than in societies where a lsquohireand rersquo set of practices is quite pervasive And this kind of consensusdecision making ndash while highly conducive to incremental innovativeness

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

627

ndash limits the capacity for radical innovations to occur This kind of lsquolongtermismrsquo set of norms and rules occurs not only in the employmentpractices of incrementally innovative rms but also in their nancingand ownership patterns (Hollingsworth 1997) During the past thirtyyears these were characteristics of many rms in Germany and JapanSignicantly in both of these countries venture capital markets werenot highly developed nancing tended to be based more on long-termbank loans than on the equity markets and the relationships of rmstended to be relatively stable And these stable institutional arrange-ments also contributed to a high degree of innovativeness ndash butincremental rather than radical innovationsMeantime we should also pursue Soskicersquos ideas (1999) about how

rmsrsquo relationships with their competitors and collaborators shape theirstyles of innovativeness In those advanced industrial societies in whichbusiness rms are highly mobilized into business associations theretends to be a high degree of standardized quality controls over technicalnorms and a high degree of reliance on legal sanctions as monitoringinstruments vis-agrave-vis association members As a result of this high degreeof integration of rms into associations member rms are more likelyto be associated with incremental innovations than is likely to be the case with rms in those societies where associations are less welldeveloped In the latter type of society radical innovations are morelikely to occur In those advanced industrial societies where rms havelong-term stable relationships with collaborators in product design andproduction the style of innovation is likely to be incremental And it is in those societies in which there is a weaker tradition of strongcollaboration among rms that radical innovations are more likely tooccurObviously even in a country where there may be a great deal of

radical innovativeness most innovations will be of an incremental typefor there are simply fewer radical innovations But Germany and therest of Europe not only have had relatively few radical innovations incomparison with the United States but they also have lagged behindthe USA in developing relatively totally new industries Thus there hasbeen much less innovative activity in Europe in high-technology andnewly emerging industries than in the USAIn complementary research Audretsch and Feldman (1995) have

demonstrated that the USA has been more successful in spawning rmsengaged in the early stages of the industry life cycle but relative toGermany had an environment unfavorable to innovative activity byestablished rms especially in more mature industries German rmshave tended to be associated with a routinized technological regime ndashwhere established enterprises have had strong incentives to engage inincremental innovations but new rms have tended to have fundamental

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

628

innovative disadvantages relative to the Americans (see Audretsch 1995Winter 1982 Casper 2000)On both sides of the Atlantic universities engage in basic and applied

research But signicantly the institutional environment in which univer-sities are embedded inuences the degree to which the research is likelyto represent major breakthroughs or is likely to have more modestimplications Recent research in biomedical science demonstrates that inthose societies in which university professors andor graduate studentshave the following characteristics there are likely to be relatively majorresearch breakthroughs graduate students have a long training periodhigh dependency on a major professor low autonomy in dening thenature of a research project restricted job options and a university envi-ronment where there is a very modest or nonexistent entrepreneurialculture For example in Germany ndash a country with relatively few radicalbreakthroughs ndash the scholar who nishes a habilitation will have spentmany years on a research project before qualifying for a permanent acad-emic position will have been in a high-dependency position with a seniorprofessor will have had limited autonomy to dene a research projectand will be constrained from pursuing a high-risk research project Theyoung German will have an extraordinarily high level of training butwill have faced a number of disincentives to undertake high-risk researchAnd as soon as heshe has completed the habilitation and becomes aprofessor administrative responsibilities constitute another disincentivefrom engaging in high-risk or long-term research projects Moreover theGerman professor is a civil servant with a more stable and secure incomethan hisher American counterpart In a very different institutionalenvironment American professors in large research universities tend tobecome much more entrepreneurial Today many American professorsbehave as though they were operating their own rms within univer-sities and thus devote considerable time to generating funds for researchprojects assistants secretaries supplies etc And the variation in thisentrepreneurial spirit across countries is associated with variation in thetaking of high-risk research strategies and the making of major researchbreakthroughsIn sum societies vary in terms of the institutional environment in

which their organizations are embedded and this variation inuencesthe degree to which they have industrial activities in the early or latestages of the industry life cycle successful new or mature industriesand radical vs incremental innovations And it is this variation whichis encouraging rms in countries with relatively few new industries and with few or no radical innovations to invest in those countries with many new industries and numerous radical innovations ndash as astrategy to be able to transfer new technologies from the latter institu-tional environment to the former

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

629

Table 4 summarizes the way that the norms rules and values of a society are associated with Americarsquos and Germanyrsquos institutionalarrangements their social system of production and their styles of inno-vativeness

CONCLUSION

Several themes deserve to be reemphasized Most importantly institu-tional analysis is emerging as a dominant issue in several social sciencedisciplines The study of various institutional components discussedabove is a complex subject requiring multiple levels of analysis and aknowledge of multiple sectors within societies It requires a theoreticalperspective that evolves from an empirical analysis in different societiesIf we are to advance the study of institutional analysis we need cross-disciplinary collaboration and team research which are not readilyachieved given the structure of most of our research universitiesAt present scholars in each academic discipline have their own distinc-

tive strategies for studying institutional analysis Thus economists tendto use several approaches (North 1990 Williamson 1985 Hodgson 19881998 1999) anthropologists another (Geertz 1995) and sociologistspolitical scientists and historians employ other strategies (Campbell etal 1991 Powell and DiMaggio 1991 Katznelson 1998 Hall and Taylor1996) With practitioners in various disciplines pursuing their distinctiveapproaches to institutional analysis there is low potential for collabo-rative institutional analysis across the social sciences The lessons fromthe biological sciences over the last 40 years are very instructive oncepractitioners from various biological disciplines began to cooperate instudying biological phenomena by using the same concepts at the mole-cular level theoretical advance was very rapid Comparable phenomenatook place in the development of other hybrid elds eg biochemistrybiophysics (Judson 1979 Olby 1979 Kohler 1982 1994) Similarly weresocial scientists in various disciplines to work collaboratively in the areaof institutional analysis there would very likely be an acceleration intheir theoretical knowledgeThis is not a plea that everyone engaged in institutional analysis should

do the same kind of research Indeed just as biologists working at themolecular level study many different kinds of problems similarly socialscientists engaged in institutional analysis would also work on manyseparate problems but in the spirit of a collective enterprise Some wouldconcentrate their attention on the study of rules norms habits conven-tions and values while others would study how these are associatedwith congurations of various institutional arrangements (types ofmarkets hierarchies networks associations communities clans statesetc) Others would work on specic institutional sectors (eg education

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

630

Table 4

Rules norms hab

its an

d values associated w

ith sp

ecic co

ngu

ration

s of coo

rdinating mecha

nism

s soc

ial sy

stem

s of

prod

uction

an

d nationa

l styles of inno

vation

USA

Germany

Norms rules values which provide incentives for

Norms rules values which provide incentives for

short-term horizons

long-term horizons

Con

gu

rative

forms of

Marke

ts c

orpo

rate hierarchies reg

ulatory state

Assoc

iation

s (busine

ss assoc

iation

s la

bor union

s)

coordinating

corp

oratist-type

networks

marke

ts c

orporatist-

mecha

nism

s (listed in

type

state

order of im

portanc

e in

the particu

lar social

system

)

Social sys

tem of

prod

uction

Bus

iness system

Firm

struc

tures

Con

glom

erates b

ut with mov

emen

t toward

Firm

more co

ngru

ent with pa

rticular prod

ucts

elim

ination of produc

tion

in som

e sectors

emph

asis on long

-term strateg

ies rm

s are well

emph

asis on sh

ort-term

strateg

ies high

integrated

differentiation

between divisions

of rms

Firm

s mob

ilized in

Low

High

and in

tegrated

into busine

ss

associations

Indu

strial relations

Low

deg

ree of jo

b secu

rity rigid in

ternal lab

or

High deg

ree of jo

b secu

rity

exible interna

l lab

or

system

marke

t ex

ible externa

l labo

r marke

tsmarke

t rigid externa

l lab

or m

arke

tPatterns of

Dispersed owne

rship frequ

ent turnov

erMore stab

le owne

rship patterns

owne

rship

Table 4

continue

d

USA

Germany

Norms rules values which provide incentives for

Norms rules values which provide incentives for

short-term horizons

long-term horizons

Social system of

Prod

uction

Bus

iness system

Training for labo

r an

dHigh deg

ree of voluntary decision mak

ing by

Collective an

d m

ore co

mpulso

ry d

ecisions

abo

ut

man

agem

ent

individuals as to ho

w m

uch

tim

e an

d ene

rgy

how m

uch ed

ucation

is ne

eded

he

avy em

pha

sis

to inv

est in educ

ation heavy

emph

asis on

on techn

ical training (eg en

gine

ering)

busine

ss sch

ools training

in marke

ting

and

sales accou

nting

Fina

ncial marke

tsEqu

ity an

d ven

ture cap

ital m

arke

ts highly

Equ

ity an

d ven

ture cap

ital m

arke

ts w

eakly

dev

elop

ed na

ncing from

cen

tral gov

ernm

ent

deve

lope

d he

avy em

pha

sis on

corporate loan

sto priva

te rm

s for dev

elop

men

t of new

tech

nologies related

to na

tion

al security

Unive

rsity research

High individual auton

omy high poten

tial for

Low

ind

ividua

l au

tono

my until almost 40

yea

rs of

system

individual creativity u

nive

rsity pr

ofesso

rs

age poten

tial for ind

ividual creativity high

ly

high

ly entreprene

urial

cons

traine

d unive

rsity pr

ofessors not so

entrep

rene

urial

Inno

vative

styles

Highly su

ccessful in m

ajor break

through

s in

High pe

rforman

ce in proc

ess an

d inc

remen

tal

basic an

d app

lied scien

ce go

od in dev

elop

ing

inno

vation

s in existing produ

cts not so su

ccessful

totally

new

lines of pr

oducts an

d in m

ajor break

through

s in basic scien

ce or in

tech

nologies high

deg

ree of entrepr

eneu

rship

deve

loping

totally new

lines of prod

ucts of

amon

g research

ers

tech

nologies low d

egree of entrepr

eneu

rship

amon

g research

ers

Table 4

continue

d

USA

Germany

Norms rules values which provide incentives for

Norms rules values which provide incentives for

short-term horizons

long-term horizons

Prod

ucts in w

hich

Pr

oducts with sh

ort liv

es prod

ucts in w

hich

Older produ

cts with long

history machine

ry pap

er

society excels

tech

nology

is bo

th highly co

mplex

and

pr

inting

textile

s m

otor veh

icles m

achine

too

lsch

ange

s rapidly en

tertainm

ent indus

try

pha

rmaceu

ticals biotechn

olog

y softw

are

aerosp

ace

business systems nancial markets and systems of research) Distinctivewould be the recognition of how these separate studies are linkedtogether how each institutional sector is linked to a societyrsquos normsrules values etc and its conguration of institutional arrangementsSimilarly other institutional analysts might focus on the study of orga-nizations with a major concern for how the institutional environmentof organizations (eg norms rules values congurations of institutionalarrangements institutional sectors) inuenced the structure culture andoutcomes of organizations And nally others might study how all ofthe aforementioned aspects of institutional analysis inuenced a societyrsquosoverall performance Nelson and Sampat (1998) have appropriatelyreminded us that we should engage in institutional analysis in order toaddress specic problems but the view developed in this article is thatwe rst need some mapping of the terrain of institutional analysis inorder for our diverse enterprises to become a meaningful collective enter-prise especially if we are to understand how institutions inuence theinnovative processThe argument developed here is not that all social scientists should engage

in institutional analysis Obviously there are numerous other importantareas of research which lie beyond the eld of institutional analysis Nordoes this article argue that institutional analysis is the most importantsubject for social scientistsSince the structure and culture of our research universities reect an

enormous amount of disciplinary fragmentation what are the prospectsfor promoting the kind of collective and interdisciplinary research agendaproposed above In one respect the prospects are not encouragingDuring the past half century the efforts of American universities to have effective communication across the biological sciences have variedgreatly in their success At the University of California-Berkeley it was not until the 1980s that there was a serious effort to integrate thebiological sciences and at that time the campus abolished seventeendepartments to promote an integrated biological program But at theUniversity of Wisconsin the biological sciences are still very fragmentedinto dozens of different departments in botany plant pathology zoologyand physiology with two departments of biochemistry and two depart-ments of geneticsThere have been other approaches which have worked very well histor-

ically for promoting scientic integration amidst scientic diversity Forexample several summer schools and special institutes were organizedin the United States during the 1940s in an effort to overcome the frag-mentation in the biological sciences These programs made it possiblefor scientists in diverse academic disciplines and from different univer-sities and countries to come together to engage in intense and frequentinteraction Perhaps the most effective program was that held in the

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

634

summers at Cold Spring Harbor New York where a research agendafor the study of molecular biology was developed (Fischer and Lipson1988 Stent and Watson 1966) Scientists went there summer aftersummer and when they returned to their universities many began therestructuring of the biological sciences in their universitiesIn our own day we need summer institutes and other similar programs

so that scholars from diverse backgrounds can come together anddevelop a collective agenda for doing institutional analysis and forstudying the impact of institutions on the performance of societies ndashinnovativeness being one such subject If we can develop such instituteswe will increase the prospects for transforming our universities Thereis reason to be optimistic that once an eclectic group of scholars workwith a common set of concepts we can have genuine theoretical advanceif researchers can bring their diverse expertise to bear on a common setof problems There is considerable evidence that institutional analysisoffers high potential for advancing our knowledge of innovations andtechnological change (Hollingsworth and Hollingsworth 2000) If we aregoing to make substantial theoretical advance it is not likely to occurfrom the vantage point of a multiplicity of disciplines Rather it is likelyto occur as we develop a new intellectual framework and our own acad-emic journals This is a tall agenda which requires us to reassess theway that knowledge is presently produced in our societies

NOTE

This article is part of a much larger agenda involving the study of institutionsorganizations and innovations in which I am involved My debts are many andvaried The rst version was written while I was in residence in the ResearchUnit for Institutional Change and European Integration of the Austrian Academyof Sciences in Vienna in the summer of 1998 I am especially grateful to ProfessorsEgon Matzner and Sonja Puntscher-Riekmann for providing the stimulating envi-ronment in which to think through the ideas developed here Egon Matzner hasengaged in very stimulating discussions with me about the importance of insti-tutions in the general eld of socio-economics The nal version of the articlewas completed during my residence as a fellow at the Neurosciences Institutein La Jolla California I also wish to thank my colleagues in the InnovationTheme Project at the Netherlands Institute for Advanced Study (NIAS) during1998ndash9 and Edgar Grande and Karl Muumlller for stimulating conversations aboutthe interaction of institutions and organizations in facilitating particular typesof innovation I am especially grateful to Jerald Hage for helping me to developa number of issues in the article and for his useful comments on the rst draftFrans van Waarden early on helped me to work through the idea of multiplelevels of institutional analysis Others who have helped to develop the ideas inthe article have been my colleagues in the research project lsquoComparative socialsystems of productionrsquo especially Tom Burns Christel Lane Yoshitaka OkadaWolfgang Streeck and Richard Whitley My colleagues Robert Boyer SteveCasper and David Soskice have taught me much about how incentives providedby the institutional environments of rms inuence individual decision making

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

635

But my most important debt is to Ellen Jane Hollingsworth who has repeatedlyquestioned every single assumption of this article Without her rich and contin-uous tutoring this article would be even less developed than it is at presentFinally without the insightful assistance of David Gear in many phases of thisarticle it would not have been completed

REFERENCES

Aoki Masahiko (1990) lsquoTowards an economic model of the Japanese rmrsquo Journalof Economic Literature March 1ndash27

Arora A Landau Ralph and Rosenberg Nathan (eds) (1998) Chemicals and Long-Term Economic Growth Insights from the Chemical Industry New York JohnWiley amp Sons

Arrow Kenneth (1974) The Limits of Organization New York NortonArthur W B (1988a) lsquoSelf-reinforcing mechanisms in economicsrsquo in P W

Anderson K J Arrow and D Pines (eds) The Economy as an Evolving ComplexSystem Redwood City Calif Addison-Wesley

mdashmdash(1988b) lsquoCompeting technologies an overviewrsquo in G Dosi C Freeman RNelson G Silverberg and L Soete (eds) Technical Change and Economic TheoryLondon Pinter Publishers pp 590ndash607

Audretsch David B (1995) Innovation and Industry Evolution Cambridge MassMIT Press

Audretsch David B and Feldman Maryann P (1995) lsquoInnovative clusters andthe industry life cyclersquo Discussion Paper Wissenschaftszentrum BerlinMarket Processes and Corporate Development Research Unit FS-IV-95-7

Baum Joel A C and Oliver Christine (1992) lsquoInstitutional embeddedness andthe dynamics of organizational populationsrsquo American Sociological Review 57540ndash59

Baum Joel A C and Singh Jitendra V (1994a) lsquoOrganizationndashenvironment coevo-lutionrsquo in Joel A C Baum and Jitendra V Singh (eds) The EvolutionaryDynamics of Organizations New York Oxford University Press pp 379ndash402

mdashmdash(eds) (1994b) The Evolutionary Dynamics of Organizations New York OxfordUniversity Press

Berger Suzanne and Dore Ronald (eds) (1996) National Diversity and GlobalCapitalism Ithaca NY Cornell University Press

Boyer Robert (1997) lsquoThe variety and unequal performance of really existingmarkets farewell to Doctor Panglossrsquo in J Rogers Hollingsworth and RobertBoyer (eds) Contemporary Capitalism the Embeddedness of InstitutionsCambridge and New York Cambridge University Press pp 55ndash93

Burns Tom R and Dietz Thomas (2001 forthcoming) lsquoSocial rule system theorysocial action institutional arrangements and evolutionary processesrsquo in JRogers Hollingsworth and Karl Muumlller (eds) Advances in SocioeconomicsLanham Md Rowman amp Littleeld

Burns Tom R and Flam Helena (1987) The Shaping of Social Organization BeverlyHills Calif Sage

Calvert Randall L (1995) lsquoThe rational choice theory of social institutionsrsquo inJeffrey S Banks and Eric A Hanuskek (eds) Modern Political Economy NewYork Cambridge University Press pp 216ndash66

Camic Charles (1986) lsquoThe matter of habitrsquo American Journal of Sociology 911039ndash87

Campbell John L (1997) lsquoMechanisms of evolutionary change in economicsgovernance interaction interpretation and bricolagersquo in Lars Magnusson

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

636

ends of economic activity to be achieved in any society The othercoordinating mechanisms include different kinds of hierarchies varioustypes of networks and associations ndash trade unions employers and busi-ness associations (Hollingsworth and Lindberg 1985 Campbell et al1991) The logic of these various institutional arrangements providesactors with vocabularies and logics for pursuing their goals for deningwhat is valued and for shaping the norms and rules by which theyabide (see Tables 2 and 3) In short in contrast to the logic of theneoclassical paradigm the argument here is that the dominant type ofinstitutional arrangements places severe constraints on the denition of needs preferences and choices of economic actors Whereas the neo-classical paradigm assumes that individuals and rms are sovereign thisarticle is based on the assumption that rms are inuenced by the hold that the institutional congurations making up a social system ofproduction have on individual decision making (Campbell et al 1991Etzioni 1988 Streeck and Schmitter 1985a Hollingsworth et al 1994Hollingsworth and Boyer 1997 Magnusson and Ottosson 1997 North1990 Hodgson 1999)Standard neoclassical economic theory has tended to downplay the

role of production and consequently of rms Even transaction cost theo-rists who are concerned with analyzing the rm as a coordinatingmechanism have been relatively unconcerned with the various compo-nents of a social system of production Indeed as long as there waswidespread optimism about the efcacy of Keynesian economics therewas relatively little concern among neoclassical economists with thesupply side of the economy Even in the opinion of most Keynesians agroup of experts should ideally be able to shape the size of aggregatedemand while the supply side of the economy would be left to the twominimalist institutions of neoclassical economics ndash markets and manage-rial hierarchies For several decades however it has become increasinglyobvious that some of the most competitive and successful patterns ofindustrial output and industrial production in capitalist economies donot derive from the neoclassical prescription of unregulated markets andcorporate hierarchies complemented by a neoliberal democratic stateIndeed it is now well understood that certain highly successful produc-tion patterns require for their emergence and survival institutionalarrangements and environments the very opposite of the prescriptionsfound in the neoclassical paradigm (see especially Streeck 1992 but alsoHollingsworth and Streeck 1994) Thus social scientists have come tounderstand that if they are to comprehend the behavior and performanceof contemporary economies concerns about social systems of produc-tion must be brought into the pictureBecause production involves more than technology a number of social

scientists have an increasing concern with social systems of production

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

615

The same equipment is frequently operated quite differently in the samesectors in different countries even when rms are competing in the same market (Maurice et al 1980 Sorge and Streeck 1988 Sorge 1989Hollingsworth et al 1994) Variations in production and process tech-nologies are inuenced partly by variations in the social environmentsin which they are embedded In other words rms are embedded intocomplex environments which among other things place constraints ontheir behavior Thus a social system of production is of major impor-tance in understanding the behavior and performance of an economyHow the state and other institutional arrangements (eg marketsnetworks private hierarchies associations) coalesce and are related toparticular social systems of production are important determinants ofeconomic performanceThe congurations of institutional arrangements that coordinate or

govern the behavior of actors in one society and its structure of specicinstitutional sectors cannot easily be transferred to another society forthey are embedded into a social system of production that is societallydistinct (Hollingsworth 1997) Societies borrow selected principles offoreign management styles and work practices but the effectiveness ofsuch borrowing is generally limited Economic behavior and performanceare shaped by the entire social system of production in which actors areembedded and not simply by specic principles of particular manage-ment styles and work practices Moreover a societyrsquos social system ofproduction tends to limit the kind of goods that it can produce and withwhich it can compete successfully in international marketsSome scholars (Kenney and Florida 1993 Oliver and Wilkinson 1988)

have assumed that the diffusion of particular forms of management stylesand work practices across societies could lead to system convergenceReferring to the Japanese production system Kenney and Florida haveargued that it consists of organizational practices whose fundamentallsquogenetic codersquo can successfully be inserted into another society and beginsuccessfully to reproduce its behavior in the new environment (1993 8)Their position is in the intellectual tradition of Antonio Gramsci (1971)who contended decades ago that the American system of mass produc-tion would most certainly diffuse to Europe over timeHowever the argument of this article is that even though British

French and American rms may adopt certain aspects of Japanesemanagement styles (eg just-in-time production self-managing teamsquality circles the use of statistical process controls) or some variant ofthe German vocational system their social systems of production willnot converge Moreover the overall conguration of a societyrsquos socialsystem of production inuences its sectoral and overall national economicperformance This helps to explain why societies have different systemsof innovation and why they vary in the clusters of industries in which

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

616

they are highly competitive in the international markets (Nelson 19931999 Hage and Hollingsworth 2000 Berger and Dore 1996)In order to understand how and why a societyrsquos economy performs

as it does it is necessary to understand its entire social system of produc-tion If a society is to modify substantially the performance of itsinstitutional sectors and its entire economy it cannot adapt only someof the management and work practices of its foreign competitors Ratherit must alter its entire social system of production Because a societyrsquosmodes of economic governance and coordination and its institutionalsectors develop according to particular logics and are system specicthere are serious limitations to the extent to which a society may mimicthe institutions (eg the rules and norms) institutional arrangements andinstitutional sectors of other societiesIn the history of modern societies there are logics by which institu-

tions coalesce into a social system of production (Hollingsworth 1991)Though institutions are constantly changing there are sharp limits tothe type and the direction of change that any particular institution canundergo because of its linkages with institutional arrangements and insti-tutional sectors Thus a societyrsquos business rms educational systemnancial markets industrial relations system etc can engage in seriousrestructuring only if most of the other institutional sectors also changeIn social systems of production there are pressures toward consistencyin the norms rules and values across institutional sectors though in anycomplex society social systems are obviously imperfectly integratedIndeed the degree to which the institutional norms and rules makingup a social system of production are loosely or tightly coupled is a vari-able of considerable importance In general the institutional sectorsmaking up a social system of production are interdependent and changesin one generally result in changes in the others Since each institutionallevel is dependent on the others for various types of resources there isinterdependence among the differing institutional spheres Moreovereach society has its norms moral principles rules and laws and recipesfor action as well as its own idiosyncratic customs traditions and prin-ciples of justice (Burns and Flam 1987)There are also other inherent obstacles to convergence among social

systems of production for where a system is at any one point in timeis inuenced by its initial state Systems with quite different initial statesare unlikely to converge with one anotherrsquos institutional practicesExisting institutional arrangements block certain institutional innova-tions and facilitate others (Roland 1990) Thus the institutions makingup a social system of production provide continuity even though insti-tutional arrangements are always changing but with a particular logicWhile Williamson (1975 1985) suggests that actors tend to choose theinstitutional arrangements which are most efcient North (1990) is

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

617

much closer to the mark in his argument that most societal institutionalarrangements exist as a result of custom and habit and are thereforeinefcient At any moment in time the world often appears to its actorsas very complex and uncertain For this reason actors often engage incontradictory forms of behavior pursuing different strategies as hedgesagainst a very uncertain world (Lanzara 1998) And their hedging andcontradictory forms of behavior may lead in somewhat different soci-etal directions all constrained by the institutional fabric within whichthe actors are embedded This kind of contradictory behavior occurs inpart because of the contradictions inherent in the dominant congura-tions of institutional arrangements with their contradictory logics (seelsquoSecond level of analysisrsquo above)Despite the emphasis on the logic of institutional continuity this is

not an argument that systems change along some predetermined pathThere are critical turning points in the history of highly industrializedsocieties but with the choices limited by the existing institutional terrainBeing path-dependent a social system of production continues along aparticular logic until or unless a fundamental societal crisis intervenes(Milgrom et al 1991 Krugman 1991 Durlauf 1991 Hollingsworth 1991Pred 1966 David 1988 Hodgson 1998)At this point it is important to confront the question of how social

systems of production evolve And this question gets to the heart of theproblem of building complex societal structures Certainly social systemsof production did not emerge from some process of social engineeringMoreover the various component parts of each social system of produc-tion have often not been designed to be part of a social system ofproduction The component parts of each social system of productionhave emerged more often than not as unintended by-products of goalswhich various actors had in mind at earlier moments in time It is usuallythe case that actors at time lsquotrsquo fail to comprehend the long-term conse-quences of their actions Wolfgang Streeck the author of several brilliantpapers on social systems of production argues that institutional sectordesigns created for one purpose generally address that goal but overtime those designs have quite unintended consequences (Streeck 1997a1997b Wright 1998) Describing the conguration of the German socialsystem of production of the 1970s and 1980s Streeck points out that itwas the unintended by-product of multiple points in history Someelements were pre-Wilhelmian others were introduced by the Alliedpowers after 1945 and others emerged during the years of the GermanFederal Republic All component parts of the German social system ofproduction

were and continue to be changing for their own reasons as wellas in reaction to each other and certainly there can be no presump-tion of a pre-established t between them even though one might

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

618

want to allow for some reinforcement effects of the modelrsquos histor-ically contingent social and economic success

(Streeck 1997b 54)

In sum the emergence of social systems of production is a long-termevolutionary process with each part interacting with its environment andresulting in a congurative whole but with no previous design by eithera single actor or a collectivity of actors As suggested above actors do not recreate their world anew At any moment in history the totalinstitutional complex of a society necessarily contains the resources and legacies of its past Preexisting institutional complexes are nevercompletely wiped out Societies consist of multiple layers of history withtheir diverse logics of action However the cumulative effect of smallperipheral changes in altering particular institutional sectors and insti-tutional arrangements can be substantial (Murmann 1998 Lanzara 1998)lsquoInstitution building is affected more by the ways in which people havecodied the past than by how they envision the future There are manymore potential resources in pre-existing institutional arrangements thanit is usually assumed or suspectedrsquo (Lanzara 1998 30)

Fourth level of analysis

The next level of institutional analysis organizational structures is some-what more controversial As suggested above Douglass North (1990)draws a sharp distinction between institutions and organizations Morerecently many organizational theorists (Powell 1991 Powell andDiMaggio 1991 DiMaggio and Powell 1983 Baum and Oliver 1992Kondra and Hinings 1998 Townley 1997) argue that institutional rulesnorms and conventions unfold in tandem with organizational structuresand this is my positionThe literature which focuses on how institutions inuence organiza-

tions is quite different from two other theoretical literatures which alsoare concerned with how organizational environments shape the behaviorof organizations There are resource dependency theorists (Pfeffer andSalancik 1978) who emphasize the role of environmental resources inshaping organizations while Hannan and Freeman (1977 1984) and otherpopulation ecologists emphasize the survival of organizations givencertain kinds of organizational conditions (Orru et al 1991)The institutionalist perspective as implied above emphasizes the

normative environment in which organizations are embedded It is a perspective which focuses on the way in which organizations in their behavior tend to conform to the institutional rules and normswhich are dominant in the organizational environment However allthree perspectives ndash the institutionalist the population ecology and

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

619

the resource dependency ndash emphasize how the environment inuencesorganizations and how organizations which are subject to the sameenvironment tend to converge in their behavior to have what DiMaggioand Powell have labeled organizational isomorphism (DiMaggio andPowell 1983)There has emerged a vast literature which demonstrates that within

each society there are scal political judicial and other regulatory normswhich limit and shape the culture and structure of organizationalbehavior The normative institutional environment of organizations limits the options of what organizations do in a particular society and inuences the patterns of ownership relations with suppliers andcustomers In short it is the normative environment of organizationswhich denes within a particular society what is socially acceptablebehavior for organizations (Hamilton and Biggart 1988 Hollingsworthand Hollingsworth 2000 Meyer and Rowan 1991 Zucker 1987 1988Orru et al 1991 Townley 1997)Thus far most of the analyses on isomorphic pressures operating on

organizations have been cross-sectional in nature However the entireconcept of isomorphism implies that there are strong environmental pres-sures exerted on organizations and in order to observe this phenomenonit is important that we have longitudinal studies which assess howchanges in an organizationrsquos institutional environment inuence changesin the structure and culture of organizations Moreover the expectationhere is that when one analyzes the patterns of organizational changeover time the isomorphism will be holistic ndash that is within a particularsociety there will be a tendency for the internal structure of particulartypes of organizations to converge in large part because of the pres-sures to conform to the changes in the external norms rules and valuesA close reading of Aoki (1990) and Slack and Hinings (1994) nds thatchanges in the national institutional environment of organizations haveinuenced changes in the internal structure of organizations within aspecic industry of a particular societyHowever it is inappropriate to view organizations as changing only

in relation to exogenous environmental change There is a kind of co-evolution which occurs between organizations and their institutionalenvironments Co-evolution implies nonlinear feedback between orga-nizations and their institutional environments This is of course quitedifferent from the more common way of studying organizations bymodeling relations between independent and dependent variables Ofcourse the more common type of analysis is quite appropriate whenthere are simple relationships which do not involve complex feedbackprocesses But in the strategy for institutional analysis which is proposedherein it is less useful to separate independent from dependent vari-ables and more useful to understand the interacting and co-evolutionary

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

620

processes among all of the levels of institutional analysis discussed inthe various sections above (Baum and Singh 1994a 379ndash80 Mowery andNelson 1999 Murmann 1998)The logic of the perspective herein suggests that within every

society there is variation in the structure and culture of business rmsuniversities and other complex organizations (Kondra and Hinings1998) However this variation takes place within parameters which aresystem specic For example every German rm and every Germanuniversity is different from every other German university and rm but there are a set of cultural and structural characteristics which distin-guish German rms and universities from those in America Similarlyeach Japanese rm and university is unique yet they have a set ofcultural and structural characteristics which are system specic andwhich differentiate them from business rms and universities in other countriesIn all societies each organization has its own distinctive organizational

rules norms and conventions which are subordinate to the meta norms and rules of the larger society within which they are embeddedHowever the strength of the institutional environment within whichorganizations are embedded varies from society to society Some societies have multiple institutional environments and there is hetero-geneity in terms of what constitutes appropriate behavior oforganizations In such societies there can be conict over whatinstitutional logics should regulate specic organizational functions(Friedland and Alford 1991 Townley 1997 262ndash3) In those societies inwhich the institutional norms habits and rules are most developed and in which the institutional pressures to conform are greatest thereis less variation in the structure and culture of business rms and variouskinds of research organizations In such societies the connectednessbetween research organizations and their institutional environments is sufciently strong for organizations to have low autonomy to pursue independent strategies and goals in these societies there is a great deal of organizational isomorphism Conversely the weaker the institutional environment in which research organizations areembedded the greater the variation in the structure and culture of business rms and research organizations Moreover where the institu-tional environments are more weakly developed organizations havegreater autonomy and exibility to respond to the development of new knowledge and to be highly innovative Hence it is in those societies where the institutional environments are most developed and most rigid and when there is less organizational autonomy and exibility that fewer radical innovations in basic and applied science as well as in totally new products and industrial sectors havebeen made

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

621

Fifth level of analysis

We come next to the outputs and performance of the various institutionalcomponents of a society It is at this level that institutional componentsare more pragmatic and exible For example in the legal sector thereare specic statutes and court rulings in the state sector there are specicpolicies in the business sector there are new products new technolo-gies and market strategies It is at this nal level of analysis thatinstitutional spheres are most open and susceptible to change and cross-national mimicry is easiest and most commonThrough the outputs of the society we can obtain some assessment as

to how well a society is performing Moreover we can assess how inno-vative it is how egalitarian it is in the distribution of its resources howegalitarian it is in terms of levels of health education etc Just as it iscomplicated to measure how egalitarian a society is similarly it is dif-cult to assess how innovative it is For example societies vary in makingincremental and radical breakthroughs in basic and applied science in developing totally new products and new kinds of organizations inincremental and process innovations in existing products and organiza-tions in developing new and different forms of marketing ndash bothdomestically and globally (Hage and Hollingsworth 2000) Of coursesocieties vary in their rates of economic growth in their rates of economicproductivity in their quality of life (eg rates of crime life expectancyetc) All of these performance criteria feed back and inuence each ofthe levels of institutional analysis discussed above ndash rules norms valuesetc institutional arrangements institutional sectors the structure andform of organizations Moreover performance measures may inuenceeach other For example if a society has very low rates of growth itmay not be very innovative in some of the types of innovative activitiesmentioned above Good or poor performances of certain kinds inuenceother performance indicators (Hollingsworth et al 1990)Different institutional arrangements and different social systems of

production result in different types of economic performance Hence solong as societies have different social systems of production there areserious constraints on the degree to which they can converge in theirinnovative styles Different social systems of production tend to maxi-mize in a more or less explicit manner different performance criteriausually mixing considerations about static and dynamic efciency protsecurity social peace and economic andor political power In short incontrast to the implications of neoclassical economic theory in real-worldeconomies there are no universal standards all economically rationalactors attempt to maximize Economic history provides numerous exam-ples of how a great variety of principles of rationality are implementedin different societies

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

622

A critical question in institutional analysis is whether an existing socialsystem of production which supports a set of routines for a particularkind of technology and industry can shift from old practices to newones There are numerous historical accounts of how the social systemof production in a particular society which worked so well for a numberof years could not adapt to new technologies (Lazonick 1990 Veblen1915 Nelson 1994) Schumpeter (1983) and Freeman (1991) have devel-oped the idea of a techno-economic paradigm which suggests thatdifferent technologies require specic forms of organizational arrange-ments ndash or what I called social systems of production Because oflsquoinstitutional inertiarsquo the social system of production of a society maynot be able to adapt to a new techno-economic paradigm (Nelson 199458 Lanzara 1998)Whether a social system of production can sustain its performance

standards depends not only on its intrinsic economic rationality but alsoon where it ts into a larger system If a particular social system ofproduction is immune from the innovativeness and competition of analternative system survival can be long lasting But if different socialsystems of production with diverging criteria of good economic perfor-mance meet in the world arena the arbitrariness of nationally imposedconstructed performance standards may be superseded by alternativeperformance criteria as a result of international competitivenessThe world economy is also socially constructed just as are national

economies Even if different social systems of production are competingin the international arena it is not always possible to determine whichis more competitively effective at any moment in time Hegemonicnation-states in the short run shape the rules of trade to favor their indus-trial sectors and rms But the history of hegemonic powers suggeststhat in the longer run such social systems of production sustainedlargely by military and political power eventually give way to moreinnovative and competitive social systems of production In our ownday as nation-states are increasingly integrated into a world economyeven hegemonic powers lose their innovative and competitive advan-tage and their share of world output decreases Such a country mayattempt to restructure its institutional arrangements and to readjust itsperformance preferences But to restructure its system of innovationgenerally calls for a major redistribution of power within a societyLargely for this reason societies have historically had limited capacityto reconstruct a system of innovation in the image of their major competi-tors Each countryrsquos social system of production is a conguration of ahost of norms rules and values as well as of institutional arrangementsEach system is constantly changing and is open to inuence from othersystems And indeed many technologies and practices diffuse from onesociety to another but a societyrsquos capacity to be innovative is constrained

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

623

by the existing social system of production Thus the same technologymay exist in numerous countries but how it is employed and how itinuences societal outcomes and performances varies from one institu-tional conguration or society to another (Hollingsworth 1998)

PERSPECTIVES ON INSTITUTIONAL CHANGE

A fundamental problem which remains unresolved in institutionalanalysis involves the nature of institutional change There is much confu-sion and miscommunication in the social sciences about institutionalchange in large part because we do not even know how to measure therate of institutional change Of course we know that there are bothexternal and internal focuses for change and we have had a good bitof scholarship on this subject What is not clearly understood is thatinstitutions and institutional arrangements within societies are histori-cally rooted that there is a great deal of path-dependency to the waythat institutions evolve Moreover the more intricately linked each sectoris with each other sector and with a societyrsquos rules and norms the lesschoice actors have to devise new institutions and institutional arrange-ments On the other hand if a sector or actor is too isolated it may betoo weak within a society to bring about any kind of effective changeThese considerations are very critical to the problem of how muchfreedom actors have to create new institutions afresh to what extentinstitutions diffuse from one society to another (Murmann 1998 Lanzara1998) This is a fundamental question for students of institutionalanalysis as we attempt to understand how the different institutionallevels are linked togetherThere are a variety of reasons why there is confusion in studying insti-

tutional change There is disagreement over the extent to which actorshave the freedom to build new institutional arrangements and if theydo to what extent the new arrangements and organizational patternsmay depart from past practices Among some institutional economists(Williamson 1985) there is a basic assumption that actors build efcientinstitutions by pursuing their self-interest and by promoting theirstrategic goals Moreover there is the assumption that a mixing of fullrationality with market competition tends to produce all of the optimalinstitutions that are needed in order to coordinate a complex capitalisteconomy (Hollingsworth and Boyer 1997 ch 14 Hodgson 1999) Incontrast the argument developed here is much more complexSocial institutions are historically rooted and there is a great deal of

path-dependency in the way that various institutional componentsevolve The shape of institutional congurations at any moment limitsthe type of options for change Because there is a great deal of institu-tional inertia radical change in the institutional components of a society

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

624

is uncommon (Lanzara 1998) Recognizing the constraints existinginstitutions exert on actors we note that existing congurations of insti-tutional arrangements institutional sectors and organizational propertiescan limit but also inuence the degree to which norms and rules canchange Many scholars (North 1990 Ostrom 1990) have emphasized therole of norms rules ideologies and values in limiting the ability of actorsto develop new institutional arrangements which radically depart fromexisting ones However there is a two-way process of interaction aprocess of co-evolution (Baum and Singh 1994a Nelson 1994) Thus asFriedland and Alford (1991 244) argue actors develop new institutionalarrangements by recognizing organizational failures and low perfor-mance but these new arrangements do not necessarily change a societyrsquosnorms rules habits and its underlying value systemCampbell (1997) advances our understanding of how existing institu-

tions enable actors to construct radically new institutions He recognizesthat most institutions and institutional arrangements embody a degreeof inertia from which it is difcult for actors to depart Moreover thedifferential power relations among actors make it unlikely that lesspowerful actors can change the existing power structure of a societyMost change in the institutional components of a society evolves througha process of constrained selection which reects to a considerable degreethe existing arrangements and power relations This becomes the inter-pretative frame for social actors with institutionalized scripts and ritualswhich tend to be taken for granted and appear to be quite ordinary andnatural This institutionalized scripting of the social world is an extremelyimportant source of social stability and inertiaCampbell has developed a very fruitful explanation of change in terms

of the various institutional components outlined here He distinguishesbetween incremental and radical institutional change and argues thatradical institutional change occurs when social actors with widelydiffering norms cultural scripts and rituals for action engage in intenseinteraction with each other (see also Knight 1992) The more diverse theinteractions the greater the potential for institutional change He arguesthat as shifts in the composition of interaction among social actors occurchanges in the interpretation of problems and interests will follow Ifthere are only minor changes in the extent to which diverse actorsinteract there will be only modest institutional change But the morefundamental the changes in the interaction of diverse social actors thegreater the change in the way that actors interpret their world And thegreater the change in the way that actors interpret their world the morelikely that radical changes in various institutional components willfollow In other words diverse decision making tends to create a widerrange of interpretative frameworks of the social world than is likely tobe the case if social actors operate in isolation

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

625

This perspective of diverse interactions among social actors facilitatesour ability to move beyond the traditional view that existing institutionsconstrain the range of institutional alternatives which actors face Theolder view has tended to suggest that actors engage in institutionalchange by extending existing institutional principles habits and conven-tions Specifying the conditions which facilitate fundamental institutionalchange permits a much richer perspective for institutional analysisCampbellrsquos perspective about the conditions under which fundamentalor radical change in institutional components occurs is consistent withmy own research about the conditions under which major discoveriesor fundamental new knowledge occur the more scholars with diversebackgrounds interact with intensity and frequency the greater the like-lihood that they will develop new and alternative ways of thinking(Hollingsworth and Hollingsworth 2000)These views of institutional change are also consistent with the recent

work of Nelson and his colleagues (Mowery and Nelson 1999 Murmann1998) They report the research on a variety of industrial sectors in anumber of different countries revealing that networks of actors in partic-ular industries business and professional associations universityresearch departments and institutes are frequently able to use their collec-tive power to modify a societyrsquos property rights research system andgovernmental policies It is the intense and frequent interaction of thesediverse complex sets of actors ndash sometimes constituting a technologysystem ndash that often leads to incremental changes in a societyrsquos total socialsystem of production It is this kind of collective action that brings aboutorganizational change change in institutional sectors institutionalarrangements and even in the rules and norms of a society And if weare to understand how the style of innovativeness of a society changeswe must rst understand the changes in its institutional makeup It ishighly unlikely that a society will develop a new style of innovativenesswithout changing its institutional structure

FIRMS THEIR INSTITUTIONAL ENVIRONMENT ANDNATIONAL SYSTEMS OF INNOVATION

Variation in the innovative style of societies is shaped by their insti-tutional makeup ndash the various levels which are discussed aboveinstitutions institutional arrangements the structure and cohesion of institutional sectors which constitute a societyrsquos social system of pro-duction and the structure and culture of its organizations ndash especiallyits business rms and other research organizations By undertaking aninstitutional analysis of a society one can begin to understand in whatkinds of organizations the production of specic kinds of knowledgetakes place and how this is linked to particular kinds of innovativeness

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

626

With such a perspective one can gain a rich understanding of why somesocieties excel in the production of radical breakthroughs in basic andapplied science and in developing radically new products and why othersocieties excel in more incremental innovationsReecting on these issues one cannot help being impressed with the

contrasts between the innovative styles of German and American orga-nizations For example Germany has continued to be very successful in making incremental innovations in many industries in which it wasvery competitive well before World War II paper printing materialsmachinery electrotechnical products motor vehicles chemicals textileyarns and fabrics But the Germans have been much less successful indeveloping totally new products or related innovations in biotechnologyelectronics telecommunications aircraft as well as other newer indus-trial areas In contrast the Americans have been very innovative in these and other industries with products having very short lives andwith technologies which change very rapidly and are very complex(Hollingsworth 1991 Soskice 1997 1999)The importance of institutions and innovativeness for international

competitiveness has led to the recognition that there are national systemsof innovation (Lundvall 1992 Nelson 1993 Edquist 1997 Hage andHollingsworth 2000) But even before the recent studies of nationalsystems of innovation Landes (1969) was a proponent of these viewswhen he made the argument that Germany had developed a nationalsystem of innovation by building on its education and scientic researchsystems which were lacking in Britain and elsewhere However the concept of a national system of innovation immediately poses theproblem of what kind of innovations should be consideredDavid Soskice an economist at the Wissenschaftszentrum Berlin has

addressed a set of problems somewhat complementary to the perspec-tive raised here His work (Soskice 1990 1997 1999) suggests that weare sensitive to the institutional environment in which the followingactors are embedded the employees of rms the owners and nanciersof rms the competitors of rms and nally their collaborators ndash aperspective complementary to Hollingsworthrsquos work on social systemsof production (Hollingsworth 1997 1998) When one reects on the insti-tutional contextsocial system of production within which these actorsare embedded one becomes aware that those countries which have some-what inexible labor markets external to rms but a business system in which rms have long-term commitments to their employees arestrongly associated with incremental but hardly at all with radical inno-vations In the rms of such countries (eg Germany and Japan) thereis much more consensus decision making than in societies where a lsquohireand rersquo set of practices is quite pervasive And this kind of consensusdecision making ndash while highly conducive to incremental innovativeness

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

627

ndash limits the capacity for radical innovations to occur This kind of lsquolongtermismrsquo set of norms and rules occurs not only in the employmentpractices of incrementally innovative rms but also in their nancingand ownership patterns (Hollingsworth 1997) During the past thirtyyears these were characteristics of many rms in Germany and JapanSignicantly in both of these countries venture capital markets werenot highly developed nancing tended to be based more on long-termbank loans than on the equity markets and the relationships of rmstended to be relatively stable And these stable institutional arrange-ments also contributed to a high degree of innovativeness ndash butincremental rather than radical innovationsMeantime we should also pursue Soskicersquos ideas (1999) about how

rmsrsquo relationships with their competitors and collaborators shape theirstyles of innovativeness In those advanced industrial societies in whichbusiness rms are highly mobilized into business associations theretends to be a high degree of standardized quality controls over technicalnorms and a high degree of reliance on legal sanctions as monitoringinstruments vis-agrave-vis association members As a result of this high degreeof integration of rms into associations member rms are more likelyto be associated with incremental innovations than is likely to be the case with rms in those societies where associations are less welldeveloped In the latter type of society radical innovations are morelikely to occur In those advanced industrial societies where rms havelong-term stable relationships with collaborators in product design andproduction the style of innovation is likely to be incremental And it is in those societies in which there is a weaker tradition of strongcollaboration among rms that radical innovations are more likely tooccurObviously even in a country where there may be a great deal of

radical innovativeness most innovations will be of an incremental typefor there are simply fewer radical innovations But Germany and therest of Europe not only have had relatively few radical innovations incomparison with the United States but they also have lagged behindthe USA in developing relatively totally new industries Thus there hasbeen much less innovative activity in Europe in high-technology andnewly emerging industries than in the USAIn complementary research Audretsch and Feldman (1995) have

demonstrated that the USA has been more successful in spawning rmsengaged in the early stages of the industry life cycle but relative toGermany had an environment unfavorable to innovative activity byestablished rms especially in more mature industries German rmshave tended to be associated with a routinized technological regime ndashwhere established enterprises have had strong incentives to engage inincremental innovations but new rms have tended to have fundamental

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

628

innovative disadvantages relative to the Americans (see Audretsch 1995Winter 1982 Casper 2000)On both sides of the Atlantic universities engage in basic and applied

research But signicantly the institutional environment in which univer-sities are embedded inuences the degree to which the research is likelyto represent major breakthroughs or is likely to have more modestimplications Recent research in biomedical science demonstrates that inthose societies in which university professors andor graduate studentshave the following characteristics there are likely to be relatively majorresearch breakthroughs graduate students have a long training periodhigh dependency on a major professor low autonomy in dening thenature of a research project restricted job options and a university envi-ronment where there is a very modest or nonexistent entrepreneurialculture For example in Germany ndash a country with relatively few radicalbreakthroughs ndash the scholar who nishes a habilitation will have spentmany years on a research project before qualifying for a permanent acad-emic position will have been in a high-dependency position with a seniorprofessor will have had limited autonomy to dene a research projectand will be constrained from pursuing a high-risk research project Theyoung German will have an extraordinarily high level of training butwill have faced a number of disincentives to undertake high-risk researchAnd as soon as heshe has completed the habilitation and becomes aprofessor administrative responsibilities constitute another disincentivefrom engaging in high-risk or long-term research projects Moreover theGerman professor is a civil servant with a more stable and secure incomethan hisher American counterpart In a very different institutionalenvironment American professors in large research universities tend tobecome much more entrepreneurial Today many American professorsbehave as though they were operating their own rms within univer-sities and thus devote considerable time to generating funds for researchprojects assistants secretaries supplies etc And the variation in thisentrepreneurial spirit across countries is associated with variation in thetaking of high-risk research strategies and the making of major researchbreakthroughsIn sum societies vary in terms of the institutional environment in

which their organizations are embedded and this variation inuencesthe degree to which they have industrial activities in the early or latestages of the industry life cycle successful new or mature industriesand radical vs incremental innovations And it is this variation whichis encouraging rms in countries with relatively few new industries and with few or no radical innovations to invest in those countries with many new industries and numerous radical innovations ndash as astrategy to be able to transfer new technologies from the latter institu-tional environment to the former

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

629

Table 4 summarizes the way that the norms rules and values of a society are associated with Americarsquos and Germanyrsquos institutionalarrangements their social system of production and their styles of inno-vativeness

CONCLUSION

Several themes deserve to be reemphasized Most importantly institu-tional analysis is emerging as a dominant issue in several social sciencedisciplines The study of various institutional components discussedabove is a complex subject requiring multiple levels of analysis and aknowledge of multiple sectors within societies It requires a theoreticalperspective that evolves from an empirical analysis in different societiesIf we are to advance the study of institutional analysis we need cross-disciplinary collaboration and team research which are not readilyachieved given the structure of most of our research universitiesAt present scholars in each academic discipline have their own distinc-

tive strategies for studying institutional analysis Thus economists tendto use several approaches (North 1990 Williamson 1985 Hodgson 19881998 1999) anthropologists another (Geertz 1995) and sociologistspolitical scientists and historians employ other strategies (Campbell etal 1991 Powell and DiMaggio 1991 Katznelson 1998 Hall and Taylor1996) With practitioners in various disciplines pursuing their distinctiveapproaches to institutional analysis there is low potential for collabo-rative institutional analysis across the social sciences The lessons fromthe biological sciences over the last 40 years are very instructive oncepractitioners from various biological disciplines began to cooperate instudying biological phenomena by using the same concepts at the mole-cular level theoretical advance was very rapid Comparable phenomenatook place in the development of other hybrid elds eg biochemistrybiophysics (Judson 1979 Olby 1979 Kohler 1982 1994) Similarly weresocial scientists in various disciplines to work collaboratively in the areaof institutional analysis there would very likely be an acceleration intheir theoretical knowledgeThis is not a plea that everyone engaged in institutional analysis should

do the same kind of research Indeed just as biologists working at themolecular level study many different kinds of problems similarly socialscientists engaged in institutional analysis would also work on manyseparate problems but in the spirit of a collective enterprise Some wouldconcentrate their attention on the study of rules norms habits conven-tions and values while others would study how these are associatedwith congurations of various institutional arrangements (types ofmarkets hierarchies networks associations communities clans statesetc) Others would work on specic institutional sectors (eg education

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

630

Table 4

Rules norms hab

its an

d values associated w

ith sp

ecic co

ngu

ration

s of coo

rdinating mecha

nism

s soc

ial sy

stem

s of

prod

uction

an

d nationa

l styles of inno

vation

USA

Germany

Norms rules values which provide incentives for

Norms rules values which provide incentives for

short-term horizons

long-term horizons

Con

gu

rative

forms of

Marke

ts c

orpo

rate hierarchies reg

ulatory state

Assoc

iation

s (busine

ss assoc

iation

s la

bor union

s)

coordinating

corp

oratist-type

networks

marke

ts c

orporatist-

mecha

nism

s (listed in

type

state

order of im

portanc

e in

the particu

lar social

system

)

Social sys

tem of

prod

uction

Bus

iness system

Firm

struc

tures

Con

glom

erates b

ut with mov

emen

t toward

Firm

more co

ngru

ent with pa

rticular prod

ucts

elim

ination of produc

tion

in som

e sectors

emph

asis on long

-term strateg

ies rm

s are well

emph

asis on sh

ort-term

strateg

ies high

integrated

differentiation

between divisions

of rms

Firm

s mob

ilized in

Low

High

and in

tegrated

into busine

ss

associations

Indu

strial relations

Low

deg

ree of jo

b secu

rity rigid in

ternal lab

or

High deg

ree of jo

b secu

rity

exible interna

l lab

or

system

marke

t ex

ible externa

l labo

r marke

tsmarke

t rigid externa

l lab

or m

arke

tPatterns of

Dispersed owne

rship frequ

ent turnov

erMore stab

le owne

rship patterns

owne

rship

Table 4

continue

d

USA

Germany

Norms rules values which provide incentives for

Norms rules values which provide incentives for

short-term horizons

long-term horizons

Social system of

Prod

uction

Bus

iness system

Training for labo

r an

dHigh deg

ree of voluntary decision mak

ing by

Collective an

d m

ore co

mpulso

ry d

ecisions

abo

ut

man

agem

ent

individuals as to ho

w m

uch

tim

e an

d ene

rgy

how m

uch ed

ucation

is ne

eded

he

avy em

pha

sis

to inv

est in educ

ation heavy

emph

asis on

on techn

ical training (eg en

gine

ering)

busine

ss sch

ools training

in marke

ting

and

sales accou

nting

Fina

ncial marke

tsEqu

ity an

d ven

ture cap

ital m

arke

ts highly

Equ

ity an

d ven

ture cap

ital m

arke

ts w

eakly

dev

elop

ed na

ncing from

cen

tral gov

ernm

ent

deve

lope

d he

avy em

pha

sis on

corporate loan

sto priva

te rm

s for dev

elop

men

t of new

tech

nologies related

to na

tion

al security

Unive

rsity research

High individual auton

omy high poten

tial for

Low

ind

ividua

l au

tono

my until almost 40

yea

rs of

system

individual creativity u

nive

rsity pr

ofesso

rs

age poten

tial for ind

ividual creativity high

ly

high

ly entreprene

urial

cons

traine

d unive

rsity pr

ofessors not so

entrep

rene

urial

Inno

vative

styles

Highly su

ccessful in m

ajor break

through

s in

High pe

rforman

ce in proc

ess an

d inc

remen

tal

basic an

d app

lied scien

ce go

od in dev

elop

ing

inno

vation

s in existing produ

cts not so su

ccessful

totally

new

lines of pr

oducts an

d in m

ajor break

through

s in basic scien

ce or in

tech

nologies high

deg

ree of entrepr

eneu

rship

deve

loping

totally new

lines of prod

ucts of

amon

g research

ers

tech

nologies low d

egree of entrepr

eneu

rship

amon

g research

ers

Table 4

continue

d

USA

Germany

Norms rules values which provide incentives for

Norms rules values which provide incentives for

short-term horizons

long-term horizons

Prod

ucts in w

hich

Pr

oducts with sh

ort liv

es prod

ucts in w

hich

Older produ

cts with long

history machine

ry pap

er

society excels

tech

nology

is bo

th highly co

mplex

and

pr

inting

textile

s m

otor veh

icles m

achine

too

lsch

ange

s rapidly en

tertainm

ent indus

try

pha

rmaceu

ticals biotechn

olog

y softw

are

aerosp

ace

business systems nancial markets and systems of research) Distinctivewould be the recognition of how these separate studies are linkedtogether how each institutional sector is linked to a societyrsquos normsrules values etc and its conguration of institutional arrangementsSimilarly other institutional analysts might focus on the study of orga-nizations with a major concern for how the institutional environmentof organizations (eg norms rules values congurations of institutionalarrangements institutional sectors) inuenced the structure culture andoutcomes of organizations And nally others might study how all ofthe aforementioned aspects of institutional analysis inuenced a societyrsquosoverall performance Nelson and Sampat (1998) have appropriatelyreminded us that we should engage in institutional analysis in order toaddress specic problems but the view developed in this article is thatwe rst need some mapping of the terrain of institutional analysis inorder for our diverse enterprises to become a meaningful collective enter-prise especially if we are to understand how institutions inuence theinnovative processThe argument developed here is not that all social scientists should engage

in institutional analysis Obviously there are numerous other importantareas of research which lie beyond the eld of institutional analysis Nordoes this article argue that institutional analysis is the most importantsubject for social scientistsSince the structure and culture of our research universities reect an

enormous amount of disciplinary fragmentation what are the prospectsfor promoting the kind of collective and interdisciplinary research agendaproposed above In one respect the prospects are not encouragingDuring the past half century the efforts of American universities to have effective communication across the biological sciences have variedgreatly in their success At the University of California-Berkeley it was not until the 1980s that there was a serious effort to integrate thebiological sciences and at that time the campus abolished seventeendepartments to promote an integrated biological program But at theUniversity of Wisconsin the biological sciences are still very fragmentedinto dozens of different departments in botany plant pathology zoologyand physiology with two departments of biochemistry and two depart-ments of geneticsThere have been other approaches which have worked very well histor-

ically for promoting scientic integration amidst scientic diversity Forexample several summer schools and special institutes were organizedin the United States during the 1940s in an effort to overcome the frag-mentation in the biological sciences These programs made it possiblefor scientists in diverse academic disciplines and from different univer-sities and countries to come together to engage in intense and frequentinteraction Perhaps the most effective program was that held in the

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

634

summers at Cold Spring Harbor New York where a research agendafor the study of molecular biology was developed (Fischer and Lipson1988 Stent and Watson 1966) Scientists went there summer aftersummer and when they returned to their universities many began therestructuring of the biological sciences in their universitiesIn our own day we need summer institutes and other similar programs

so that scholars from diverse backgrounds can come together anddevelop a collective agenda for doing institutional analysis and forstudying the impact of institutions on the performance of societies ndashinnovativeness being one such subject If we can develop such instituteswe will increase the prospects for transforming our universities Thereis reason to be optimistic that once an eclectic group of scholars workwith a common set of concepts we can have genuine theoretical advanceif researchers can bring their diverse expertise to bear on a common setof problems There is considerable evidence that institutional analysisoffers high potential for advancing our knowledge of innovations andtechnological change (Hollingsworth and Hollingsworth 2000) If we aregoing to make substantial theoretical advance it is not likely to occurfrom the vantage point of a multiplicity of disciplines Rather it is likelyto occur as we develop a new intellectual framework and our own acad-emic journals This is a tall agenda which requires us to reassess theway that knowledge is presently produced in our societies

NOTE

This article is part of a much larger agenda involving the study of institutionsorganizations and innovations in which I am involved My debts are many andvaried The rst version was written while I was in residence in the ResearchUnit for Institutional Change and European Integration of the Austrian Academyof Sciences in Vienna in the summer of 1998 I am especially grateful to ProfessorsEgon Matzner and Sonja Puntscher-Riekmann for providing the stimulating envi-ronment in which to think through the ideas developed here Egon Matzner hasengaged in very stimulating discussions with me about the importance of insti-tutions in the general eld of socio-economics The nal version of the articlewas completed during my residence as a fellow at the Neurosciences Institutein La Jolla California I also wish to thank my colleagues in the InnovationTheme Project at the Netherlands Institute for Advanced Study (NIAS) during1998ndash9 and Edgar Grande and Karl Muumlller for stimulating conversations aboutthe interaction of institutions and organizations in facilitating particular typesof innovation I am especially grateful to Jerald Hage for helping me to developa number of issues in the article and for his useful comments on the rst draftFrans van Waarden early on helped me to work through the idea of multiplelevels of institutional analysis Others who have helped to develop the ideas inthe article have been my colleagues in the research project lsquoComparative socialsystems of productionrsquo especially Tom Burns Christel Lane Yoshitaka OkadaWolfgang Streeck and Richard Whitley My colleagues Robert Boyer SteveCasper and David Soskice have taught me much about how incentives providedby the institutional environments of rms inuence individual decision making

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

635

But my most important debt is to Ellen Jane Hollingsworth who has repeatedlyquestioned every single assumption of this article Without her rich and contin-uous tutoring this article would be even less developed than it is at presentFinally without the insightful assistance of David Gear in many phases of thisarticle it would not have been completed

REFERENCES

Aoki Masahiko (1990) lsquoTowards an economic model of the Japanese rmrsquo Journalof Economic Literature March 1ndash27

Arora A Landau Ralph and Rosenberg Nathan (eds) (1998) Chemicals and Long-Term Economic Growth Insights from the Chemical Industry New York JohnWiley amp Sons

Arrow Kenneth (1974) The Limits of Organization New York NortonArthur W B (1988a) lsquoSelf-reinforcing mechanisms in economicsrsquo in P W

Anderson K J Arrow and D Pines (eds) The Economy as an Evolving ComplexSystem Redwood City Calif Addison-Wesley

mdashmdash(1988b) lsquoCompeting technologies an overviewrsquo in G Dosi C Freeman RNelson G Silverberg and L Soete (eds) Technical Change and Economic TheoryLondon Pinter Publishers pp 590ndash607

Audretsch David B (1995) Innovation and Industry Evolution Cambridge MassMIT Press

Audretsch David B and Feldman Maryann P (1995) lsquoInnovative clusters andthe industry life cyclersquo Discussion Paper Wissenschaftszentrum BerlinMarket Processes and Corporate Development Research Unit FS-IV-95-7

Baum Joel A C and Oliver Christine (1992) lsquoInstitutional embeddedness andthe dynamics of organizational populationsrsquo American Sociological Review 57540ndash59

Baum Joel A C and Singh Jitendra V (1994a) lsquoOrganizationndashenvironment coevo-lutionrsquo in Joel A C Baum and Jitendra V Singh (eds) The EvolutionaryDynamics of Organizations New York Oxford University Press pp 379ndash402

mdashmdash(eds) (1994b) The Evolutionary Dynamics of Organizations New York OxfordUniversity Press

Berger Suzanne and Dore Ronald (eds) (1996) National Diversity and GlobalCapitalism Ithaca NY Cornell University Press

Boyer Robert (1997) lsquoThe variety and unequal performance of really existingmarkets farewell to Doctor Panglossrsquo in J Rogers Hollingsworth and RobertBoyer (eds) Contemporary Capitalism the Embeddedness of InstitutionsCambridge and New York Cambridge University Press pp 55ndash93

Burns Tom R and Dietz Thomas (2001 forthcoming) lsquoSocial rule system theorysocial action institutional arrangements and evolutionary processesrsquo in JRogers Hollingsworth and Karl Muumlller (eds) Advances in SocioeconomicsLanham Md Rowman amp Littleeld

Burns Tom R and Flam Helena (1987) The Shaping of Social Organization BeverlyHills Calif Sage

Calvert Randall L (1995) lsquoThe rational choice theory of social institutionsrsquo inJeffrey S Banks and Eric A Hanuskek (eds) Modern Political Economy NewYork Cambridge University Press pp 216ndash66

Camic Charles (1986) lsquoThe matter of habitrsquo American Journal of Sociology 911039ndash87

Campbell John L (1997) lsquoMechanisms of evolutionary change in economicsgovernance interaction interpretation and bricolagersquo in Lars Magnusson

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

636

The same equipment is frequently operated quite differently in the samesectors in different countries even when rms are competing in the same market (Maurice et al 1980 Sorge and Streeck 1988 Sorge 1989Hollingsworth et al 1994) Variations in production and process tech-nologies are inuenced partly by variations in the social environmentsin which they are embedded In other words rms are embedded intocomplex environments which among other things place constraints ontheir behavior Thus a social system of production is of major impor-tance in understanding the behavior and performance of an economyHow the state and other institutional arrangements (eg marketsnetworks private hierarchies associations) coalesce and are related toparticular social systems of production are important determinants ofeconomic performanceThe congurations of institutional arrangements that coordinate or

govern the behavior of actors in one society and its structure of specicinstitutional sectors cannot easily be transferred to another society forthey are embedded into a social system of production that is societallydistinct (Hollingsworth 1997) Societies borrow selected principles offoreign management styles and work practices but the effectiveness ofsuch borrowing is generally limited Economic behavior and performanceare shaped by the entire social system of production in which actors areembedded and not simply by specic principles of particular manage-ment styles and work practices Moreover a societyrsquos social system ofproduction tends to limit the kind of goods that it can produce and withwhich it can compete successfully in international marketsSome scholars (Kenney and Florida 1993 Oliver and Wilkinson 1988)

have assumed that the diffusion of particular forms of management stylesand work practices across societies could lead to system convergenceReferring to the Japanese production system Kenney and Florida haveargued that it consists of organizational practices whose fundamentallsquogenetic codersquo can successfully be inserted into another society and beginsuccessfully to reproduce its behavior in the new environment (1993 8)Their position is in the intellectual tradition of Antonio Gramsci (1971)who contended decades ago that the American system of mass produc-tion would most certainly diffuse to Europe over timeHowever the argument of this article is that even though British

French and American rms may adopt certain aspects of Japanesemanagement styles (eg just-in-time production self-managing teamsquality circles the use of statistical process controls) or some variant ofthe German vocational system their social systems of production willnot converge Moreover the overall conguration of a societyrsquos socialsystem of production inuences its sectoral and overall national economicperformance This helps to explain why societies have different systemsof innovation and why they vary in the clusters of industries in which

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

616

they are highly competitive in the international markets (Nelson 19931999 Hage and Hollingsworth 2000 Berger and Dore 1996)In order to understand how and why a societyrsquos economy performs

as it does it is necessary to understand its entire social system of produc-tion If a society is to modify substantially the performance of itsinstitutional sectors and its entire economy it cannot adapt only someof the management and work practices of its foreign competitors Ratherit must alter its entire social system of production Because a societyrsquosmodes of economic governance and coordination and its institutionalsectors develop according to particular logics and are system specicthere are serious limitations to the extent to which a society may mimicthe institutions (eg the rules and norms) institutional arrangements andinstitutional sectors of other societiesIn the history of modern societies there are logics by which institu-

tions coalesce into a social system of production (Hollingsworth 1991)Though institutions are constantly changing there are sharp limits tothe type and the direction of change that any particular institution canundergo because of its linkages with institutional arrangements and insti-tutional sectors Thus a societyrsquos business rms educational systemnancial markets industrial relations system etc can engage in seriousrestructuring only if most of the other institutional sectors also changeIn social systems of production there are pressures toward consistencyin the norms rules and values across institutional sectors though in anycomplex society social systems are obviously imperfectly integratedIndeed the degree to which the institutional norms and rules makingup a social system of production are loosely or tightly coupled is a vari-able of considerable importance In general the institutional sectorsmaking up a social system of production are interdependent and changesin one generally result in changes in the others Since each institutionallevel is dependent on the others for various types of resources there isinterdependence among the differing institutional spheres Moreovereach society has its norms moral principles rules and laws and recipesfor action as well as its own idiosyncratic customs traditions and prin-ciples of justice (Burns and Flam 1987)There are also other inherent obstacles to convergence among social

systems of production for where a system is at any one point in timeis inuenced by its initial state Systems with quite different initial statesare unlikely to converge with one anotherrsquos institutional practicesExisting institutional arrangements block certain institutional innova-tions and facilitate others (Roland 1990) Thus the institutions makingup a social system of production provide continuity even though insti-tutional arrangements are always changing but with a particular logicWhile Williamson (1975 1985) suggests that actors tend to choose theinstitutional arrangements which are most efcient North (1990) is

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

617

much closer to the mark in his argument that most societal institutionalarrangements exist as a result of custom and habit and are thereforeinefcient At any moment in time the world often appears to its actorsas very complex and uncertain For this reason actors often engage incontradictory forms of behavior pursuing different strategies as hedgesagainst a very uncertain world (Lanzara 1998) And their hedging andcontradictory forms of behavior may lead in somewhat different soci-etal directions all constrained by the institutional fabric within whichthe actors are embedded This kind of contradictory behavior occurs inpart because of the contradictions inherent in the dominant congura-tions of institutional arrangements with their contradictory logics (seelsquoSecond level of analysisrsquo above)Despite the emphasis on the logic of institutional continuity this is

not an argument that systems change along some predetermined pathThere are critical turning points in the history of highly industrializedsocieties but with the choices limited by the existing institutional terrainBeing path-dependent a social system of production continues along aparticular logic until or unless a fundamental societal crisis intervenes(Milgrom et al 1991 Krugman 1991 Durlauf 1991 Hollingsworth 1991Pred 1966 David 1988 Hodgson 1998)At this point it is important to confront the question of how social

systems of production evolve And this question gets to the heart of theproblem of building complex societal structures Certainly social systemsof production did not emerge from some process of social engineeringMoreover the various component parts of each social system of produc-tion have often not been designed to be part of a social system ofproduction The component parts of each social system of productionhave emerged more often than not as unintended by-products of goalswhich various actors had in mind at earlier moments in time It is usuallythe case that actors at time lsquotrsquo fail to comprehend the long-term conse-quences of their actions Wolfgang Streeck the author of several brilliantpapers on social systems of production argues that institutional sectordesigns created for one purpose generally address that goal but overtime those designs have quite unintended consequences (Streeck 1997a1997b Wright 1998) Describing the conguration of the German socialsystem of production of the 1970s and 1980s Streeck points out that itwas the unintended by-product of multiple points in history Someelements were pre-Wilhelmian others were introduced by the Alliedpowers after 1945 and others emerged during the years of the GermanFederal Republic All component parts of the German social system ofproduction

were and continue to be changing for their own reasons as wellas in reaction to each other and certainly there can be no presump-tion of a pre-established t between them even though one might

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

618

want to allow for some reinforcement effects of the modelrsquos histor-ically contingent social and economic success

(Streeck 1997b 54)

In sum the emergence of social systems of production is a long-termevolutionary process with each part interacting with its environment andresulting in a congurative whole but with no previous design by eithera single actor or a collectivity of actors As suggested above actors do not recreate their world anew At any moment in history the totalinstitutional complex of a society necessarily contains the resources and legacies of its past Preexisting institutional complexes are nevercompletely wiped out Societies consist of multiple layers of history withtheir diverse logics of action However the cumulative effect of smallperipheral changes in altering particular institutional sectors and insti-tutional arrangements can be substantial (Murmann 1998 Lanzara 1998)lsquoInstitution building is affected more by the ways in which people havecodied the past than by how they envision the future There are manymore potential resources in pre-existing institutional arrangements thanit is usually assumed or suspectedrsquo (Lanzara 1998 30)

Fourth level of analysis

The next level of institutional analysis organizational structures is some-what more controversial As suggested above Douglass North (1990)draws a sharp distinction between institutions and organizations Morerecently many organizational theorists (Powell 1991 Powell andDiMaggio 1991 DiMaggio and Powell 1983 Baum and Oliver 1992Kondra and Hinings 1998 Townley 1997) argue that institutional rulesnorms and conventions unfold in tandem with organizational structuresand this is my positionThe literature which focuses on how institutions inuence organiza-

tions is quite different from two other theoretical literatures which alsoare concerned with how organizational environments shape the behaviorof organizations There are resource dependency theorists (Pfeffer andSalancik 1978) who emphasize the role of environmental resources inshaping organizations while Hannan and Freeman (1977 1984) and otherpopulation ecologists emphasize the survival of organizations givencertain kinds of organizational conditions (Orru et al 1991)The institutionalist perspective as implied above emphasizes the

normative environment in which organizations are embedded It is a perspective which focuses on the way in which organizations in their behavior tend to conform to the institutional rules and normswhich are dominant in the organizational environment However allthree perspectives ndash the institutionalist the population ecology and

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

619

the resource dependency ndash emphasize how the environment inuencesorganizations and how organizations which are subject to the sameenvironment tend to converge in their behavior to have what DiMaggioand Powell have labeled organizational isomorphism (DiMaggio andPowell 1983)There has emerged a vast literature which demonstrates that within

each society there are scal political judicial and other regulatory normswhich limit and shape the culture and structure of organizationalbehavior The normative institutional environment of organizations limits the options of what organizations do in a particular society and inuences the patterns of ownership relations with suppliers andcustomers In short it is the normative environment of organizationswhich denes within a particular society what is socially acceptablebehavior for organizations (Hamilton and Biggart 1988 Hollingsworthand Hollingsworth 2000 Meyer and Rowan 1991 Zucker 1987 1988Orru et al 1991 Townley 1997)Thus far most of the analyses on isomorphic pressures operating on

organizations have been cross-sectional in nature However the entireconcept of isomorphism implies that there are strong environmental pres-sures exerted on organizations and in order to observe this phenomenonit is important that we have longitudinal studies which assess howchanges in an organizationrsquos institutional environment inuence changesin the structure and culture of organizations Moreover the expectationhere is that when one analyzes the patterns of organizational changeover time the isomorphism will be holistic ndash that is within a particularsociety there will be a tendency for the internal structure of particulartypes of organizations to converge in large part because of the pres-sures to conform to the changes in the external norms rules and valuesA close reading of Aoki (1990) and Slack and Hinings (1994) nds thatchanges in the national institutional environment of organizations haveinuenced changes in the internal structure of organizations within aspecic industry of a particular societyHowever it is inappropriate to view organizations as changing only

in relation to exogenous environmental change There is a kind of co-evolution which occurs between organizations and their institutionalenvironments Co-evolution implies nonlinear feedback between orga-nizations and their institutional environments This is of course quitedifferent from the more common way of studying organizations bymodeling relations between independent and dependent variables Ofcourse the more common type of analysis is quite appropriate whenthere are simple relationships which do not involve complex feedbackprocesses But in the strategy for institutional analysis which is proposedherein it is less useful to separate independent from dependent vari-ables and more useful to understand the interacting and co-evolutionary

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

620

processes among all of the levels of institutional analysis discussed inthe various sections above (Baum and Singh 1994a 379ndash80 Mowery andNelson 1999 Murmann 1998)The logic of the perspective herein suggests that within every

society there is variation in the structure and culture of business rmsuniversities and other complex organizations (Kondra and Hinings1998) However this variation takes place within parameters which aresystem specic For example every German rm and every Germanuniversity is different from every other German university and rm but there are a set of cultural and structural characteristics which distin-guish German rms and universities from those in America Similarlyeach Japanese rm and university is unique yet they have a set ofcultural and structural characteristics which are system specic andwhich differentiate them from business rms and universities in other countriesIn all societies each organization has its own distinctive organizational

rules norms and conventions which are subordinate to the meta norms and rules of the larger society within which they are embeddedHowever the strength of the institutional environment within whichorganizations are embedded varies from society to society Some societies have multiple institutional environments and there is hetero-geneity in terms of what constitutes appropriate behavior oforganizations In such societies there can be conict over whatinstitutional logics should regulate specic organizational functions(Friedland and Alford 1991 Townley 1997 262ndash3) In those societies inwhich the institutional norms habits and rules are most developed and in which the institutional pressures to conform are greatest thereis less variation in the structure and culture of business rms and variouskinds of research organizations In such societies the connectednessbetween research organizations and their institutional environments is sufciently strong for organizations to have low autonomy to pursue independent strategies and goals in these societies there is a great deal of organizational isomorphism Conversely the weaker the institutional environment in which research organizations areembedded the greater the variation in the structure and culture of business rms and research organizations Moreover where the institu-tional environments are more weakly developed organizations havegreater autonomy and exibility to respond to the development of new knowledge and to be highly innovative Hence it is in those societies where the institutional environments are most developed and most rigid and when there is less organizational autonomy and exibility that fewer radical innovations in basic and applied science as well as in totally new products and industrial sectors havebeen made

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

621

Fifth level of analysis

We come next to the outputs and performance of the various institutionalcomponents of a society It is at this level that institutional componentsare more pragmatic and exible For example in the legal sector thereare specic statutes and court rulings in the state sector there are specicpolicies in the business sector there are new products new technolo-gies and market strategies It is at this nal level of analysis thatinstitutional spheres are most open and susceptible to change and cross-national mimicry is easiest and most commonThrough the outputs of the society we can obtain some assessment as

to how well a society is performing Moreover we can assess how inno-vative it is how egalitarian it is in the distribution of its resources howegalitarian it is in terms of levels of health education etc Just as it iscomplicated to measure how egalitarian a society is similarly it is dif-cult to assess how innovative it is For example societies vary in makingincremental and radical breakthroughs in basic and applied science in developing totally new products and new kinds of organizations inincremental and process innovations in existing products and organiza-tions in developing new and different forms of marketing ndash bothdomestically and globally (Hage and Hollingsworth 2000) Of coursesocieties vary in their rates of economic growth in their rates of economicproductivity in their quality of life (eg rates of crime life expectancyetc) All of these performance criteria feed back and inuence each ofthe levels of institutional analysis discussed above ndash rules norms valuesetc institutional arrangements institutional sectors the structure andform of organizations Moreover performance measures may inuenceeach other For example if a society has very low rates of growth itmay not be very innovative in some of the types of innovative activitiesmentioned above Good or poor performances of certain kinds inuenceother performance indicators (Hollingsworth et al 1990)Different institutional arrangements and different social systems of

production result in different types of economic performance Hence solong as societies have different social systems of production there areserious constraints on the degree to which they can converge in theirinnovative styles Different social systems of production tend to maxi-mize in a more or less explicit manner different performance criteriausually mixing considerations about static and dynamic efciency protsecurity social peace and economic andor political power In short incontrast to the implications of neoclassical economic theory in real-worldeconomies there are no universal standards all economically rationalactors attempt to maximize Economic history provides numerous exam-ples of how a great variety of principles of rationality are implementedin different societies

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

622

A critical question in institutional analysis is whether an existing socialsystem of production which supports a set of routines for a particularkind of technology and industry can shift from old practices to newones There are numerous historical accounts of how the social systemof production in a particular society which worked so well for a numberof years could not adapt to new technologies (Lazonick 1990 Veblen1915 Nelson 1994) Schumpeter (1983) and Freeman (1991) have devel-oped the idea of a techno-economic paradigm which suggests thatdifferent technologies require specic forms of organizational arrange-ments ndash or what I called social systems of production Because oflsquoinstitutional inertiarsquo the social system of production of a society maynot be able to adapt to a new techno-economic paradigm (Nelson 199458 Lanzara 1998)Whether a social system of production can sustain its performance

standards depends not only on its intrinsic economic rationality but alsoon where it ts into a larger system If a particular social system ofproduction is immune from the innovativeness and competition of analternative system survival can be long lasting But if different socialsystems of production with diverging criteria of good economic perfor-mance meet in the world arena the arbitrariness of nationally imposedconstructed performance standards may be superseded by alternativeperformance criteria as a result of international competitivenessThe world economy is also socially constructed just as are national

economies Even if different social systems of production are competingin the international arena it is not always possible to determine whichis more competitively effective at any moment in time Hegemonicnation-states in the short run shape the rules of trade to favor their indus-trial sectors and rms But the history of hegemonic powers suggeststhat in the longer run such social systems of production sustainedlargely by military and political power eventually give way to moreinnovative and competitive social systems of production In our ownday as nation-states are increasingly integrated into a world economyeven hegemonic powers lose their innovative and competitive advan-tage and their share of world output decreases Such a country mayattempt to restructure its institutional arrangements and to readjust itsperformance preferences But to restructure its system of innovationgenerally calls for a major redistribution of power within a societyLargely for this reason societies have historically had limited capacityto reconstruct a system of innovation in the image of their major competi-tors Each countryrsquos social system of production is a conguration of ahost of norms rules and values as well as of institutional arrangementsEach system is constantly changing and is open to inuence from othersystems And indeed many technologies and practices diffuse from onesociety to another but a societyrsquos capacity to be innovative is constrained

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

623

by the existing social system of production Thus the same technologymay exist in numerous countries but how it is employed and how itinuences societal outcomes and performances varies from one institu-tional conguration or society to another (Hollingsworth 1998)

PERSPECTIVES ON INSTITUTIONAL CHANGE

A fundamental problem which remains unresolved in institutionalanalysis involves the nature of institutional change There is much confu-sion and miscommunication in the social sciences about institutionalchange in large part because we do not even know how to measure therate of institutional change Of course we know that there are bothexternal and internal focuses for change and we have had a good bitof scholarship on this subject What is not clearly understood is thatinstitutions and institutional arrangements within societies are histori-cally rooted that there is a great deal of path-dependency to the waythat institutions evolve Moreover the more intricately linked each sectoris with each other sector and with a societyrsquos rules and norms the lesschoice actors have to devise new institutions and institutional arrange-ments On the other hand if a sector or actor is too isolated it may betoo weak within a society to bring about any kind of effective changeThese considerations are very critical to the problem of how muchfreedom actors have to create new institutions afresh to what extentinstitutions diffuse from one society to another (Murmann 1998 Lanzara1998) This is a fundamental question for students of institutionalanalysis as we attempt to understand how the different institutionallevels are linked togetherThere are a variety of reasons why there is confusion in studying insti-

tutional change There is disagreement over the extent to which actorshave the freedom to build new institutional arrangements and if theydo to what extent the new arrangements and organizational patternsmay depart from past practices Among some institutional economists(Williamson 1985) there is a basic assumption that actors build efcientinstitutions by pursuing their self-interest and by promoting theirstrategic goals Moreover there is the assumption that a mixing of fullrationality with market competition tends to produce all of the optimalinstitutions that are needed in order to coordinate a complex capitalisteconomy (Hollingsworth and Boyer 1997 ch 14 Hodgson 1999) Incontrast the argument developed here is much more complexSocial institutions are historically rooted and there is a great deal of

path-dependency in the way that various institutional componentsevolve The shape of institutional congurations at any moment limitsthe type of options for change Because there is a great deal of institu-tional inertia radical change in the institutional components of a society

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

624

is uncommon (Lanzara 1998) Recognizing the constraints existinginstitutions exert on actors we note that existing congurations of insti-tutional arrangements institutional sectors and organizational propertiescan limit but also inuence the degree to which norms and rules canchange Many scholars (North 1990 Ostrom 1990) have emphasized therole of norms rules ideologies and values in limiting the ability of actorsto develop new institutional arrangements which radically depart fromexisting ones However there is a two-way process of interaction aprocess of co-evolution (Baum and Singh 1994a Nelson 1994) Thus asFriedland and Alford (1991 244) argue actors develop new institutionalarrangements by recognizing organizational failures and low perfor-mance but these new arrangements do not necessarily change a societyrsquosnorms rules habits and its underlying value systemCampbell (1997) advances our understanding of how existing institu-

tions enable actors to construct radically new institutions He recognizesthat most institutions and institutional arrangements embody a degreeof inertia from which it is difcult for actors to depart Moreover thedifferential power relations among actors make it unlikely that lesspowerful actors can change the existing power structure of a societyMost change in the institutional components of a society evolves througha process of constrained selection which reects to a considerable degreethe existing arrangements and power relations This becomes the inter-pretative frame for social actors with institutionalized scripts and ritualswhich tend to be taken for granted and appear to be quite ordinary andnatural This institutionalized scripting of the social world is an extremelyimportant source of social stability and inertiaCampbell has developed a very fruitful explanation of change in terms

of the various institutional components outlined here He distinguishesbetween incremental and radical institutional change and argues thatradical institutional change occurs when social actors with widelydiffering norms cultural scripts and rituals for action engage in intenseinteraction with each other (see also Knight 1992) The more diverse theinteractions the greater the potential for institutional change He arguesthat as shifts in the composition of interaction among social actors occurchanges in the interpretation of problems and interests will follow Ifthere are only minor changes in the extent to which diverse actorsinteract there will be only modest institutional change But the morefundamental the changes in the interaction of diverse social actors thegreater the change in the way that actors interpret their world And thegreater the change in the way that actors interpret their world the morelikely that radical changes in various institutional components willfollow In other words diverse decision making tends to create a widerrange of interpretative frameworks of the social world than is likely tobe the case if social actors operate in isolation

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

625

This perspective of diverse interactions among social actors facilitatesour ability to move beyond the traditional view that existing institutionsconstrain the range of institutional alternatives which actors face Theolder view has tended to suggest that actors engage in institutionalchange by extending existing institutional principles habits and conven-tions Specifying the conditions which facilitate fundamental institutionalchange permits a much richer perspective for institutional analysisCampbellrsquos perspective about the conditions under which fundamentalor radical change in institutional components occurs is consistent withmy own research about the conditions under which major discoveriesor fundamental new knowledge occur the more scholars with diversebackgrounds interact with intensity and frequency the greater the like-lihood that they will develop new and alternative ways of thinking(Hollingsworth and Hollingsworth 2000)These views of institutional change are also consistent with the recent

work of Nelson and his colleagues (Mowery and Nelson 1999 Murmann1998) They report the research on a variety of industrial sectors in anumber of different countries revealing that networks of actors in partic-ular industries business and professional associations universityresearch departments and institutes are frequently able to use their collec-tive power to modify a societyrsquos property rights research system andgovernmental policies It is the intense and frequent interaction of thesediverse complex sets of actors ndash sometimes constituting a technologysystem ndash that often leads to incremental changes in a societyrsquos total socialsystem of production It is this kind of collective action that brings aboutorganizational change change in institutional sectors institutionalarrangements and even in the rules and norms of a society And if weare to understand how the style of innovativeness of a society changeswe must rst understand the changes in its institutional makeup It ishighly unlikely that a society will develop a new style of innovativenesswithout changing its institutional structure

FIRMS THEIR INSTITUTIONAL ENVIRONMENT ANDNATIONAL SYSTEMS OF INNOVATION

Variation in the innovative style of societies is shaped by their insti-tutional makeup ndash the various levels which are discussed aboveinstitutions institutional arrangements the structure and cohesion of institutional sectors which constitute a societyrsquos social system of pro-duction and the structure and culture of its organizations ndash especiallyits business rms and other research organizations By undertaking aninstitutional analysis of a society one can begin to understand in whatkinds of organizations the production of specic kinds of knowledgetakes place and how this is linked to particular kinds of innovativeness

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

626

With such a perspective one can gain a rich understanding of why somesocieties excel in the production of radical breakthroughs in basic andapplied science and in developing radically new products and why othersocieties excel in more incremental innovationsReecting on these issues one cannot help being impressed with the

contrasts between the innovative styles of German and American orga-nizations For example Germany has continued to be very successful in making incremental innovations in many industries in which it wasvery competitive well before World War II paper printing materialsmachinery electrotechnical products motor vehicles chemicals textileyarns and fabrics But the Germans have been much less successful indeveloping totally new products or related innovations in biotechnologyelectronics telecommunications aircraft as well as other newer indus-trial areas In contrast the Americans have been very innovative in these and other industries with products having very short lives andwith technologies which change very rapidly and are very complex(Hollingsworth 1991 Soskice 1997 1999)The importance of institutions and innovativeness for international

competitiveness has led to the recognition that there are national systemsof innovation (Lundvall 1992 Nelson 1993 Edquist 1997 Hage andHollingsworth 2000) But even before the recent studies of nationalsystems of innovation Landes (1969) was a proponent of these viewswhen he made the argument that Germany had developed a nationalsystem of innovation by building on its education and scientic researchsystems which were lacking in Britain and elsewhere However the concept of a national system of innovation immediately poses theproblem of what kind of innovations should be consideredDavid Soskice an economist at the Wissenschaftszentrum Berlin has

addressed a set of problems somewhat complementary to the perspec-tive raised here His work (Soskice 1990 1997 1999) suggests that weare sensitive to the institutional environment in which the followingactors are embedded the employees of rms the owners and nanciersof rms the competitors of rms and nally their collaborators ndash aperspective complementary to Hollingsworthrsquos work on social systemsof production (Hollingsworth 1997 1998) When one reects on the insti-tutional contextsocial system of production within which these actorsare embedded one becomes aware that those countries which have some-what inexible labor markets external to rms but a business system in which rms have long-term commitments to their employees arestrongly associated with incremental but hardly at all with radical inno-vations In the rms of such countries (eg Germany and Japan) thereis much more consensus decision making than in societies where a lsquohireand rersquo set of practices is quite pervasive And this kind of consensusdecision making ndash while highly conducive to incremental innovativeness

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

627

ndash limits the capacity for radical innovations to occur This kind of lsquolongtermismrsquo set of norms and rules occurs not only in the employmentpractices of incrementally innovative rms but also in their nancingand ownership patterns (Hollingsworth 1997) During the past thirtyyears these were characteristics of many rms in Germany and JapanSignicantly in both of these countries venture capital markets werenot highly developed nancing tended to be based more on long-termbank loans than on the equity markets and the relationships of rmstended to be relatively stable And these stable institutional arrange-ments also contributed to a high degree of innovativeness ndash butincremental rather than radical innovationsMeantime we should also pursue Soskicersquos ideas (1999) about how

rmsrsquo relationships with their competitors and collaborators shape theirstyles of innovativeness In those advanced industrial societies in whichbusiness rms are highly mobilized into business associations theretends to be a high degree of standardized quality controls over technicalnorms and a high degree of reliance on legal sanctions as monitoringinstruments vis-agrave-vis association members As a result of this high degreeof integration of rms into associations member rms are more likelyto be associated with incremental innovations than is likely to be the case with rms in those societies where associations are less welldeveloped In the latter type of society radical innovations are morelikely to occur In those advanced industrial societies where rms havelong-term stable relationships with collaborators in product design andproduction the style of innovation is likely to be incremental And it is in those societies in which there is a weaker tradition of strongcollaboration among rms that radical innovations are more likely tooccurObviously even in a country where there may be a great deal of

radical innovativeness most innovations will be of an incremental typefor there are simply fewer radical innovations But Germany and therest of Europe not only have had relatively few radical innovations incomparison with the United States but they also have lagged behindthe USA in developing relatively totally new industries Thus there hasbeen much less innovative activity in Europe in high-technology andnewly emerging industries than in the USAIn complementary research Audretsch and Feldman (1995) have

demonstrated that the USA has been more successful in spawning rmsengaged in the early stages of the industry life cycle but relative toGermany had an environment unfavorable to innovative activity byestablished rms especially in more mature industries German rmshave tended to be associated with a routinized technological regime ndashwhere established enterprises have had strong incentives to engage inincremental innovations but new rms have tended to have fundamental

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

628

innovative disadvantages relative to the Americans (see Audretsch 1995Winter 1982 Casper 2000)On both sides of the Atlantic universities engage in basic and applied

research But signicantly the institutional environment in which univer-sities are embedded inuences the degree to which the research is likelyto represent major breakthroughs or is likely to have more modestimplications Recent research in biomedical science demonstrates that inthose societies in which university professors andor graduate studentshave the following characteristics there are likely to be relatively majorresearch breakthroughs graduate students have a long training periodhigh dependency on a major professor low autonomy in dening thenature of a research project restricted job options and a university envi-ronment where there is a very modest or nonexistent entrepreneurialculture For example in Germany ndash a country with relatively few radicalbreakthroughs ndash the scholar who nishes a habilitation will have spentmany years on a research project before qualifying for a permanent acad-emic position will have been in a high-dependency position with a seniorprofessor will have had limited autonomy to dene a research projectand will be constrained from pursuing a high-risk research project Theyoung German will have an extraordinarily high level of training butwill have faced a number of disincentives to undertake high-risk researchAnd as soon as heshe has completed the habilitation and becomes aprofessor administrative responsibilities constitute another disincentivefrom engaging in high-risk or long-term research projects Moreover theGerman professor is a civil servant with a more stable and secure incomethan hisher American counterpart In a very different institutionalenvironment American professors in large research universities tend tobecome much more entrepreneurial Today many American professorsbehave as though they were operating their own rms within univer-sities and thus devote considerable time to generating funds for researchprojects assistants secretaries supplies etc And the variation in thisentrepreneurial spirit across countries is associated with variation in thetaking of high-risk research strategies and the making of major researchbreakthroughsIn sum societies vary in terms of the institutional environment in

which their organizations are embedded and this variation inuencesthe degree to which they have industrial activities in the early or latestages of the industry life cycle successful new or mature industriesand radical vs incremental innovations And it is this variation whichis encouraging rms in countries with relatively few new industries and with few or no radical innovations to invest in those countries with many new industries and numerous radical innovations ndash as astrategy to be able to transfer new technologies from the latter institu-tional environment to the former

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

629

Table 4 summarizes the way that the norms rules and values of a society are associated with Americarsquos and Germanyrsquos institutionalarrangements their social system of production and their styles of inno-vativeness

CONCLUSION

Several themes deserve to be reemphasized Most importantly institu-tional analysis is emerging as a dominant issue in several social sciencedisciplines The study of various institutional components discussedabove is a complex subject requiring multiple levels of analysis and aknowledge of multiple sectors within societies It requires a theoreticalperspective that evolves from an empirical analysis in different societiesIf we are to advance the study of institutional analysis we need cross-disciplinary collaboration and team research which are not readilyachieved given the structure of most of our research universitiesAt present scholars in each academic discipline have their own distinc-

tive strategies for studying institutional analysis Thus economists tendto use several approaches (North 1990 Williamson 1985 Hodgson 19881998 1999) anthropologists another (Geertz 1995) and sociologistspolitical scientists and historians employ other strategies (Campbell etal 1991 Powell and DiMaggio 1991 Katznelson 1998 Hall and Taylor1996) With practitioners in various disciplines pursuing their distinctiveapproaches to institutional analysis there is low potential for collabo-rative institutional analysis across the social sciences The lessons fromthe biological sciences over the last 40 years are very instructive oncepractitioners from various biological disciplines began to cooperate instudying biological phenomena by using the same concepts at the mole-cular level theoretical advance was very rapid Comparable phenomenatook place in the development of other hybrid elds eg biochemistrybiophysics (Judson 1979 Olby 1979 Kohler 1982 1994) Similarly weresocial scientists in various disciplines to work collaboratively in the areaof institutional analysis there would very likely be an acceleration intheir theoretical knowledgeThis is not a plea that everyone engaged in institutional analysis should

do the same kind of research Indeed just as biologists working at themolecular level study many different kinds of problems similarly socialscientists engaged in institutional analysis would also work on manyseparate problems but in the spirit of a collective enterprise Some wouldconcentrate their attention on the study of rules norms habits conven-tions and values while others would study how these are associatedwith congurations of various institutional arrangements (types ofmarkets hierarchies networks associations communities clans statesetc) Others would work on specic institutional sectors (eg education

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

630

Table 4

Rules norms hab

its an

d values associated w

ith sp

ecic co

ngu

ration

s of coo

rdinating mecha

nism

s soc

ial sy

stem

s of

prod

uction

an

d nationa

l styles of inno

vation

USA

Germany

Norms rules values which provide incentives for

Norms rules values which provide incentives for

short-term horizons

long-term horizons

Con

gu

rative

forms of

Marke

ts c

orpo

rate hierarchies reg

ulatory state

Assoc

iation

s (busine

ss assoc

iation

s la

bor union

s)

coordinating

corp

oratist-type

networks

marke

ts c

orporatist-

mecha

nism

s (listed in

type

state

order of im

portanc

e in

the particu

lar social

system

)

Social sys

tem of

prod

uction

Bus

iness system

Firm

struc

tures

Con

glom

erates b

ut with mov

emen

t toward

Firm

more co

ngru

ent with pa

rticular prod

ucts

elim

ination of produc

tion

in som

e sectors

emph

asis on long

-term strateg

ies rm

s are well

emph

asis on sh

ort-term

strateg

ies high

integrated

differentiation

between divisions

of rms

Firm

s mob

ilized in

Low

High

and in

tegrated

into busine

ss

associations

Indu

strial relations

Low

deg

ree of jo

b secu

rity rigid in

ternal lab

or

High deg

ree of jo

b secu

rity

exible interna

l lab

or

system

marke

t ex

ible externa

l labo

r marke

tsmarke

t rigid externa

l lab

or m

arke

tPatterns of

Dispersed owne

rship frequ

ent turnov

erMore stab

le owne

rship patterns

owne

rship

Table 4

continue

d

USA

Germany

Norms rules values which provide incentives for

Norms rules values which provide incentives for

short-term horizons

long-term horizons

Social system of

Prod

uction

Bus

iness system

Training for labo

r an

dHigh deg

ree of voluntary decision mak

ing by

Collective an

d m

ore co

mpulso

ry d

ecisions

abo

ut

man

agem

ent

individuals as to ho

w m

uch

tim

e an

d ene

rgy

how m

uch ed

ucation

is ne

eded

he

avy em

pha

sis

to inv

est in educ

ation heavy

emph

asis on

on techn

ical training (eg en

gine

ering)

busine

ss sch

ools training

in marke

ting

and

sales accou

nting

Fina

ncial marke

tsEqu

ity an

d ven

ture cap

ital m

arke

ts highly

Equ

ity an

d ven

ture cap

ital m

arke

ts w

eakly

dev

elop

ed na

ncing from

cen

tral gov

ernm

ent

deve

lope

d he

avy em

pha

sis on

corporate loan

sto priva

te rm

s for dev

elop

men

t of new

tech

nologies related

to na

tion

al security

Unive

rsity research

High individual auton

omy high poten

tial for

Low

ind

ividua

l au

tono

my until almost 40

yea

rs of

system

individual creativity u

nive

rsity pr

ofesso

rs

age poten

tial for ind

ividual creativity high

ly

high

ly entreprene

urial

cons

traine

d unive

rsity pr

ofessors not so

entrep

rene

urial

Inno

vative

styles

Highly su

ccessful in m

ajor break

through

s in

High pe

rforman

ce in proc

ess an

d inc

remen

tal

basic an

d app

lied scien

ce go

od in dev

elop

ing

inno

vation

s in existing produ

cts not so su

ccessful

totally

new

lines of pr

oducts an

d in m

ajor break

through

s in basic scien

ce or in

tech

nologies high

deg

ree of entrepr

eneu

rship

deve

loping

totally new

lines of prod

ucts of

amon

g research

ers

tech

nologies low d

egree of entrepr

eneu

rship

amon

g research

ers

Table 4

continue

d

USA

Germany

Norms rules values which provide incentives for

Norms rules values which provide incentives for

short-term horizons

long-term horizons

Prod

ucts in w

hich

Pr

oducts with sh

ort liv

es prod

ucts in w

hich

Older produ

cts with long

history machine

ry pap

er

society excels

tech

nology

is bo

th highly co

mplex

and

pr

inting

textile

s m

otor veh

icles m

achine

too

lsch

ange

s rapidly en

tertainm

ent indus

try

pha

rmaceu

ticals biotechn

olog

y softw

are

aerosp

ace

business systems nancial markets and systems of research) Distinctivewould be the recognition of how these separate studies are linkedtogether how each institutional sector is linked to a societyrsquos normsrules values etc and its conguration of institutional arrangementsSimilarly other institutional analysts might focus on the study of orga-nizations with a major concern for how the institutional environmentof organizations (eg norms rules values congurations of institutionalarrangements institutional sectors) inuenced the structure culture andoutcomes of organizations And nally others might study how all ofthe aforementioned aspects of institutional analysis inuenced a societyrsquosoverall performance Nelson and Sampat (1998) have appropriatelyreminded us that we should engage in institutional analysis in order toaddress specic problems but the view developed in this article is thatwe rst need some mapping of the terrain of institutional analysis inorder for our diverse enterprises to become a meaningful collective enter-prise especially if we are to understand how institutions inuence theinnovative processThe argument developed here is not that all social scientists should engage

in institutional analysis Obviously there are numerous other importantareas of research which lie beyond the eld of institutional analysis Nordoes this article argue that institutional analysis is the most importantsubject for social scientistsSince the structure and culture of our research universities reect an

enormous amount of disciplinary fragmentation what are the prospectsfor promoting the kind of collective and interdisciplinary research agendaproposed above In one respect the prospects are not encouragingDuring the past half century the efforts of American universities to have effective communication across the biological sciences have variedgreatly in their success At the University of California-Berkeley it was not until the 1980s that there was a serious effort to integrate thebiological sciences and at that time the campus abolished seventeendepartments to promote an integrated biological program But at theUniversity of Wisconsin the biological sciences are still very fragmentedinto dozens of different departments in botany plant pathology zoologyand physiology with two departments of biochemistry and two depart-ments of geneticsThere have been other approaches which have worked very well histor-

ically for promoting scientic integration amidst scientic diversity Forexample several summer schools and special institutes were organizedin the United States during the 1940s in an effort to overcome the frag-mentation in the biological sciences These programs made it possiblefor scientists in diverse academic disciplines and from different univer-sities and countries to come together to engage in intense and frequentinteraction Perhaps the most effective program was that held in the

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

634

summers at Cold Spring Harbor New York where a research agendafor the study of molecular biology was developed (Fischer and Lipson1988 Stent and Watson 1966) Scientists went there summer aftersummer and when they returned to their universities many began therestructuring of the biological sciences in their universitiesIn our own day we need summer institutes and other similar programs

so that scholars from diverse backgrounds can come together anddevelop a collective agenda for doing institutional analysis and forstudying the impact of institutions on the performance of societies ndashinnovativeness being one such subject If we can develop such instituteswe will increase the prospects for transforming our universities Thereis reason to be optimistic that once an eclectic group of scholars workwith a common set of concepts we can have genuine theoretical advanceif researchers can bring their diverse expertise to bear on a common setof problems There is considerable evidence that institutional analysisoffers high potential for advancing our knowledge of innovations andtechnological change (Hollingsworth and Hollingsworth 2000) If we aregoing to make substantial theoretical advance it is not likely to occurfrom the vantage point of a multiplicity of disciplines Rather it is likelyto occur as we develop a new intellectual framework and our own acad-emic journals This is a tall agenda which requires us to reassess theway that knowledge is presently produced in our societies

NOTE

This article is part of a much larger agenda involving the study of institutionsorganizations and innovations in which I am involved My debts are many andvaried The rst version was written while I was in residence in the ResearchUnit for Institutional Change and European Integration of the Austrian Academyof Sciences in Vienna in the summer of 1998 I am especially grateful to ProfessorsEgon Matzner and Sonja Puntscher-Riekmann for providing the stimulating envi-ronment in which to think through the ideas developed here Egon Matzner hasengaged in very stimulating discussions with me about the importance of insti-tutions in the general eld of socio-economics The nal version of the articlewas completed during my residence as a fellow at the Neurosciences Institutein La Jolla California I also wish to thank my colleagues in the InnovationTheme Project at the Netherlands Institute for Advanced Study (NIAS) during1998ndash9 and Edgar Grande and Karl Muumlller for stimulating conversations aboutthe interaction of institutions and organizations in facilitating particular typesof innovation I am especially grateful to Jerald Hage for helping me to developa number of issues in the article and for his useful comments on the rst draftFrans van Waarden early on helped me to work through the idea of multiplelevels of institutional analysis Others who have helped to develop the ideas inthe article have been my colleagues in the research project lsquoComparative socialsystems of productionrsquo especially Tom Burns Christel Lane Yoshitaka OkadaWolfgang Streeck and Richard Whitley My colleagues Robert Boyer SteveCasper and David Soskice have taught me much about how incentives providedby the institutional environments of rms inuence individual decision making

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

635

But my most important debt is to Ellen Jane Hollingsworth who has repeatedlyquestioned every single assumption of this article Without her rich and contin-uous tutoring this article would be even less developed than it is at presentFinally without the insightful assistance of David Gear in many phases of thisarticle it would not have been completed

REFERENCES

Aoki Masahiko (1990) lsquoTowards an economic model of the Japanese rmrsquo Journalof Economic Literature March 1ndash27

Arora A Landau Ralph and Rosenberg Nathan (eds) (1998) Chemicals and Long-Term Economic Growth Insights from the Chemical Industry New York JohnWiley amp Sons

Arrow Kenneth (1974) The Limits of Organization New York NortonArthur W B (1988a) lsquoSelf-reinforcing mechanisms in economicsrsquo in P W

Anderson K J Arrow and D Pines (eds) The Economy as an Evolving ComplexSystem Redwood City Calif Addison-Wesley

mdashmdash(1988b) lsquoCompeting technologies an overviewrsquo in G Dosi C Freeman RNelson G Silverberg and L Soete (eds) Technical Change and Economic TheoryLondon Pinter Publishers pp 590ndash607

Audretsch David B (1995) Innovation and Industry Evolution Cambridge MassMIT Press

Audretsch David B and Feldman Maryann P (1995) lsquoInnovative clusters andthe industry life cyclersquo Discussion Paper Wissenschaftszentrum BerlinMarket Processes and Corporate Development Research Unit FS-IV-95-7

Baum Joel A C and Oliver Christine (1992) lsquoInstitutional embeddedness andthe dynamics of organizational populationsrsquo American Sociological Review 57540ndash59

Baum Joel A C and Singh Jitendra V (1994a) lsquoOrganizationndashenvironment coevo-lutionrsquo in Joel A C Baum and Jitendra V Singh (eds) The EvolutionaryDynamics of Organizations New York Oxford University Press pp 379ndash402

mdashmdash(eds) (1994b) The Evolutionary Dynamics of Organizations New York OxfordUniversity Press

Berger Suzanne and Dore Ronald (eds) (1996) National Diversity and GlobalCapitalism Ithaca NY Cornell University Press

Boyer Robert (1997) lsquoThe variety and unequal performance of really existingmarkets farewell to Doctor Panglossrsquo in J Rogers Hollingsworth and RobertBoyer (eds) Contemporary Capitalism the Embeddedness of InstitutionsCambridge and New York Cambridge University Press pp 55ndash93

Burns Tom R and Dietz Thomas (2001 forthcoming) lsquoSocial rule system theorysocial action institutional arrangements and evolutionary processesrsquo in JRogers Hollingsworth and Karl Muumlller (eds) Advances in SocioeconomicsLanham Md Rowman amp Littleeld

Burns Tom R and Flam Helena (1987) The Shaping of Social Organization BeverlyHills Calif Sage

Calvert Randall L (1995) lsquoThe rational choice theory of social institutionsrsquo inJeffrey S Banks and Eric A Hanuskek (eds) Modern Political Economy NewYork Cambridge University Press pp 216ndash66

Camic Charles (1986) lsquoThe matter of habitrsquo American Journal of Sociology 911039ndash87

Campbell John L (1997) lsquoMechanisms of evolutionary change in economicsgovernance interaction interpretation and bricolagersquo in Lars Magnusson

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

636

they are highly competitive in the international markets (Nelson 19931999 Hage and Hollingsworth 2000 Berger and Dore 1996)In order to understand how and why a societyrsquos economy performs

as it does it is necessary to understand its entire social system of produc-tion If a society is to modify substantially the performance of itsinstitutional sectors and its entire economy it cannot adapt only someof the management and work practices of its foreign competitors Ratherit must alter its entire social system of production Because a societyrsquosmodes of economic governance and coordination and its institutionalsectors develop according to particular logics and are system specicthere are serious limitations to the extent to which a society may mimicthe institutions (eg the rules and norms) institutional arrangements andinstitutional sectors of other societiesIn the history of modern societies there are logics by which institu-

tions coalesce into a social system of production (Hollingsworth 1991)Though institutions are constantly changing there are sharp limits tothe type and the direction of change that any particular institution canundergo because of its linkages with institutional arrangements and insti-tutional sectors Thus a societyrsquos business rms educational systemnancial markets industrial relations system etc can engage in seriousrestructuring only if most of the other institutional sectors also changeIn social systems of production there are pressures toward consistencyin the norms rules and values across institutional sectors though in anycomplex society social systems are obviously imperfectly integratedIndeed the degree to which the institutional norms and rules makingup a social system of production are loosely or tightly coupled is a vari-able of considerable importance In general the institutional sectorsmaking up a social system of production are interdependent and changesin one generally result in changes in the others Since each institutionallevel is dependent on the others for various types of resources there isinterdependence among the differing institutional spheres Moreovereach society has its norms moral principles rules and laws and recipesfor action as well as its own idiosyncratic customs traditions and prin-ciples of justice (Burns and Flam 1987)There are also other inherent obstacles to convergence among social

systems of production for where a system is at any one point in timeis inuenced by its initial state Systems with quite different initial statesare unlikely to converge with one anotherrsquos institutional practicesExisting institutional arrangements block certain institutional innova-tions and facilitate others (Roland 1990) Thus the institutions makingup a social system of production provide continuity even though insti-tutional arrangements are always changing but with a particular logicWhile Williamson (1975 1985) suggests that actors tend to choose theinstitutional arrangements which are most efcient North (1990) is

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

617

much closer to the mark in his argument that most societal institutionalarrangements exist as a result of custom and habit and are thereforeinefcient At any moment in time the world often appears to its actorsas very complex and uncertain For this reason actors often engage incontradictory forms of behavior pursuing different strategies as hedgesagainst a very uncertain world (Lanzara 1998) And their hedging andcontradictory forms of behavior may lead in somewhat different soci-etal directions all constrained by the institutional fabric within whichthe actors are embedded This kind of contradictory behavior occurs inpart because of the contradictions inherent in the dominant congura-tions of institutional arrangements with their contradictory logics (seelsquoSecond level of analysisrsquo above)Despite the emphasis on the logic of institutional continuity this is

not an argument that systems change along some predetermined pathThere are critical turning points in the history of highly industrializedsocieties but with the choices limited by the existing institutional terrainBeing path-dependent a social system of production continues along aparticular logic until or unless a fundamental societal crisis intervenes(Milgrom et al 1991 Krugman 1991 Durlauf 1991 Hollingsworth 1991Pred 1966 David 1988 Hodgson 1998)At this point it is important to confront the question of how social

systems of production evolve And this question gets to the heart of theproblem of building complex societal structures Certainly social systemsof production did not emerge from some process of social engineeringMoreover the various component parts of each social system of produc-tion have often not been designed to be part of a social system ofproduction The component parts of each social system of productionhave emerged more often than not as unintended by-products of goalswhich various actors had in mind at earlier moments in time It is usuallythe case that actors at time lsquotrsquo fail to comprehend the long-term conse-quences of their actions Wolfgang Streeck the author of several brilliantpapers on social systems of production argues that institutional sectordesigns created for one purpose generally address that goal but overtime those designs have quite unintended consequences (Streeck 1997a1997b Wright 1998) Describing the conguration of the German socialsystem of production of the 1970s and 1980s Streeck points out that itwas the unintended by-product of multiple points in history Someelements were pre-Wilhelmian others were introduced by the Alliedpowers after 1945 and others emerged during the years of the GermanFederal Republic All component parts of the German social system ofproduction

were and continue to be changing for their own reasons as wellas in reaction to each other and certainly there can be no presump-tion of a pre-established t between them even though one might

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

618

want to allow for some reinforcement effects of the modelrsquos histor-ically contingent social and economic success

(Streeck 1997b 54)

In sum the emergence of social systems of production is a long-termevolutionary process with each part interacting with its environment andresulting in a congurative whole but with no previous design by eithera single actor or a collectivity of actors As suggested above actors do not recreate their world anew At any moment in history the totalinstitutional complex of a society necessarily contains the resources and legacies of its past Preexisting institutional complexes are nevercompletely wiped out Societies consist of multiple layers of history withtheir diverse logics of action However the cumulative effect of smallperipheral changes in altering particular institutional sectors and insti-tutional arrangements can be substantial (Murmann 1998 Lanzara 1998)lsquoInstitution building is affected more by the ways in which people havecodied the past than by how they envision the future There are manymore potential resources in pre-existing institutional arrangements thanit is usually assumed or suspectedrsquo (Lanzara 1998 30)

Fourth level of analysis

The next level of institutional analysis organizational structures is some-what more controversial As suggested above Douglass North (1990)draws a sharp distinction between institutions and organizations Morerecently many organizational theorists (Powell 1991 Powell andDiMaggio 1991 DiMaggio and Powell 1983 Baum and Oliver 1992Kondra and Hinings 1998 Townley 1997) argue that institutional rulesnorms and conventions unfold in tandem with organizational structuresand this is my positionThe literature which focuses on how institutions inuence organiza-

tions is quite different from two other theoretical literatures which alsoare concerned with how organizational environments shape the behaviorof organizations There are resource dependency theorists (Pfeffer andSalancik 1978) who emphasize the role of environmental resources inshaping organizations while Hannan and Freeman (1977 1984) and otherpopulation ecologists emphasize the survival of organizations givencertain kinds of organizational conditions (Orru et al 1991)The institutionalist perspective as implied above emphasizes the

normative environment in which organizations are embedded It is a perspective which focuses on the way in which organizations in their behavior tend to conform to the institutional rules and normswhich are dominant in the organizational environment However allthree perspectives ndash the institutionalist the population ecology and

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

619

the resource dependency ndash emphasize how the environment inuencesorganizations and how organizations which are subject to the sameenvironment tend to converge in their behavior to have what DiMaggioand Powell have labeled organizational isomorphism (DiMaggio andPowell 1983)There has emerged a vast literature which demonstrates that within

each society there are scal political judicial and other regulatory normswhich limit and shape the culture and structure of organizationalbehavior The normative institutional environment of organizations limits the options of what organizations do in a particular society and inuences the patterns of ownership relations with suppliers andcustomers In short it is the normative environment of organizationswhich denes within a particular society what is socially acceptablebehavior for organizations (Hamilton and Biggart 1988 Hollingsworthand Hollingsworth 2000 Meyer and Rowan 1991 Zucker 1987 1988Orru et al 1991 Townley 1997)Thus far most of the analyses on isomorphic pressures operating on

organizations have been cross-sectional in nature However the entireconcept of isomorphism implies that there are strong environmental pres-sures exerted on organizations and in order to observe this phenomenonit is important that we have longitudinal studies which assess howchanges in an organizationrsquos institutional environment inuence changesin the structure and culture of organizations Moreover the expectationhere is that when one analyzes the patterns of organizational changeover time the isomorphism will be holistic ndash that is within a particularsociety there will be a tendency for the internal structure of particulartypes of organizations to converge in large part because of the pres-sures to conform to the changes in the external norms rules and valuesA close reading of Aoki (1990) and Slack and Hinings (1994) nds thatchanges in the national institutional environment of organizations haveinuenced changes in the internal structure of organizations within aspecic industry of a particular societyHowever it is inappropriate to view organizations as changing only

in relation to exogenous environmental change There is a kind of co-evolution which occurs between organizations and their institutionalenvironments Co-evolution implies nonlinear feedback between orga-nizations and their institutional environments This is of course quitedifferent from the more common way of studying organizations bymodeling relations between independent and dependent variables Ofcourse the more common type of analysis is quite appropriate whenthere are simple relationships which do not involve complex feedbackprocesses But in the strategy for institutional analysis which is proposedherein it is less useful to separate independent from dependent vari-ables and more useful to understand the interacting and co-evolutionary

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

620

processes among all of the levels of institutional analysis discussed inthe various sections above (Baum and Singh 1994a 379ndash80 Mowery andNelson 1999 Murmann 1998)The logic of the perspective herein suggests that within every

society there is variation in the structure and culture of business rmsuniversities and other complex organizations (Kondra and Hinings1998) However this variation takes place within parameters which aresystem specic For example every German rm and every Germanuniversity is different from every other German university and rm but there are a set of cultural and structural characteristics which distin-guish German rms and universities from those in America Similarlyeach Japanese rm and university is unique yet they have a set ofcultural and structural characteristics which are system specic andwhich differentiate them from business rms and universities in other countriesIn all societies each organization has its own distinctive organizational

rules norms and conventions which are subordinate to the meta norms and rules of the larger society within which they are embeddedHowever the strength of the institutional environment within whichorganizations are embedded varies from society to society Some societies have multiple institutional environments and there is hetero-geneity in terms of what constitutes appropriate behavior oforganizations In such societies there can be conict over whatinstitutional logics should regulate specic organizational functions(Friedland and Alford 1991 Townley 1997 262ndash3) In those societies inwhich the institutional norms habits and rules are most developed and in which the institutional pressures to conform are greatest thereis less variation in the structure and culture of business rms and variouskinds of research organizations In such societies the connectednessbetween research organizations and their institutional environments is sufciently strong for organizations to have low autonomy to pursue independent strategies and goals in these societies there is a great deal of organizational isomorphism Conversely the weaker the institutional environment in which research organizations areembedded the greater the variation in the structure and culture of business rms and research organizations Moreover where the institu-tional environments are more weakly developed organizations havegreater autonomy and exibility to respond to the development of new knowledge and to be highly innovative Hence it is in those societies where the institutional environments are most developed and most rigid and when there is less organizational autonomy and exibility that fewer radical innovations in basic and applied science as well as in totally new products and industrial sectors havebeen made

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

621

Fifth level of analysis

We come next to the outputs and performance of the various institutionalcomponents of a society It is at this level that institutional componentsare more pragmatic and exible For example in the legal sector thereare specic statutes and court rulings in the state sector there are specicpolicies in the business sector there are new products new technolo-gies and market strategies It is at this nal level of analysis thatinstitutional spheres are most open and susceptible to change and cross-national mimicry is easiest and most commonThrough the outputs of the society we can obtain some assessment as

to how well a society is performing Moreover we can assess how inno-vative it is how egalitarian it is in the distribution of its resources howegalitarian it is in terms of levels of health education etc Just as it iscomplicated to measure how egalitarian a society is similarly it is dif-cult to assess how innovative it is For example societies vary in makingincremental and radical breakthroughs in basic and applied science in developing totally new products and new kinds of organizations inincremental and process innovations in existing products and organiza-tions in developing new and different forms of marketing ndash bothdomestically and globally (Hage and Hollingsworth 2000) Of coursesocieties vary in their rates of economic growth in their rates of economicproductivity in their quality of life (eg rates of crime life expectancyetc) All of these performance criteria feed back and inuence each ofthe levels of institutional analysis discussed above ndash rules norms valuesetc institutional arrangements institutional sectors the structure andform of organizations Moreover performance measures may inuenceeach other For example if a society has very low rates of growth itmay not be very innovative in some of the types of innovative activitiesmentioned above Good or poor performances of certain kinds inuenceother performance indicators (Hollingsworth et al 1990)Different institutional arrangements and different social systems of

production result in different types of economic performance Hence solong as societies have different social systems of production there areserious constraints on the degree to which they can converge in theirinnovative styles Different social systems of production tend to maxi-mize in a more or less explicit manner different performance criteriausually mixing considerations about static and dynamic efciency protsecurity social peace and economic andor political power In short incontrast to the implications of neoclassical economic theory in real-worldeconomies there are no universal standards all economically rationalactors attempt to maximize Economic history provides numerous exam-ples of how a great variety of principles of rationality are implementedin different societies

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

622

A critical question in institutional analysis is whether an existing socialsystem of production which supports a set of routines for a particularkind of technology and industry can shift from old practices to newones There are numerous historical accounts of how the social systemof production in a particular society which worked so well for a numberof years could not adapt to new technologies (Lazonick 1990 Veblen1915 Nelson 1994) Schumpeter (1983) and Freeman (1991) have devel-oped the idea of a techno-economic paradigm which suggests thatdifferent technologies require specic forms of organizational arrange-ments ndash or what I called social systems of production Because oflsquoinstitutional inertiarsquo the social system of production of a society maynot be able to adapt to a new techno-economic paradigm (Nelson 199458 Lanzara 1998)Whether a social system of production can sustain its performance

standards depends not only on its intrinsic economic rationality but alsoon where it ts into a larger system If a particular social system ofproduction is immune from the innovativeness and competition of analternative system survival can be long lasting But if different socialsystems of production with diverging criteria of good economic perfor-mance meet in the world arena the arbitrariness of nationally imposedconstructed performance standards may be superseded by alternativeperformance criteria as a result of international competitivenessThe world economy is also socially constructed just as are national

economies Even if different social systems of production are competingin the international arena it is not always possible to determine whichis more competitively effective at any moment in time Hegemonicnation-states in the short run shape the rules of trade to favor their indus-trial sectors and rms But the history of hegemonic powers suggeststhat in the longer run such social systems of production sustainedlargely by military and political power eventually give way to moreinnovative and competitive social systems of production In our ownday as nation-states are increasingly integrated into a world economyeven hegemonic powers lose their innovative and competitive advan-tage and their share of world output decreases Such a country mayattempt to restructure its institutional arrangements and to readjust itsperformance preferences But to restructure its system of innovationgenerally calls for a major redistribution of power within a societyLargely for this reason societies have historically had limited capacityto reconstruct a system of innovation in the image of their major competi-tors Each countryrsquos social system of production is a conguration of ahost of norms rules and values as well as of institutional arrangementsEach system is constantly changing and is open to inuence from othersystems And indeed many technologies and practices diffuse from onesociety to another but a societyrsquos capacity to be innovative is constrained

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

623

by the existing social system of production Thus the same technologymay exist in numerous countries but how it is employed and how itinuences societal outcomes and performances varies from one institu-tional conguration or society to another (Hollingsworth 1998)

PERSPECTIVES ON INSTITUTIONAL CHANGE

A fundamental problem which remains unresolved in institutionalanalysis involves the nature of institutional change There is much confu-sion and miscommunication in the social sciences about institutionalchange in large part because we do not even know how to measure therate of institutional change Of course we know that there are bothexternal and internal focuses for change and we have had a good bitof scholarship on this subject What is not clearly understood is thatinstitutions and institutional arrangements within societies are histori-cally rooted that there is a great deal of path-dependency to the waythat institutions evolve Moreover the more intricately linked each sectoris with each other sector and with a societyrsquos rules and norms the lesschoice actors have to devise new institutions and institutional arrange-ments On the other hand if a sector or actor is too isolated it may betoo weak within a society to bring about any kind of effective changeThese considerations are very critical to the problem of how muchfreedom actors have to create new institutions afresh to what extentinstitutions diffuse from one society to another (Murmann 1998 Lanzara1998) This is a fundamental question for students of institutionalanalysis as we attempt to understand how the different institutionallevels are linked togetherThere are a variety of reasons why there is confusion in studying insti-

tutional change There is disagreement over the extent to which actorshave the freedom to build new institutional arrangements and if theydo to what extent the new arrangements and organizational patternsmay depart from past practices Among some institutional economists(Williamson 1985) there is a basic assumption that actors build efcientinstitutions by pursuing their self-interest and by promoting theirstrategic goals Moreover there is the assumption that a mixing of fullrationality with market competition tends to produce all of the optimalinstitutions that are needed in order to coordinate a complex capitalisteconomy (Hollingsworth and Boyer 1997 ch 14 Hodgson 1999) Incontrast the argument developed here is much more complexSocial institutions are historically rooted and there is a great deal of

path-dependency in the way that various institutional componentsevolve The shape of institutional congurations at any moment limitsthe type of options for change Because there is a great deal of institu-tional inertia radical change in the institutional components of a society

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

624

is uncommon (Lanzara 1998) Recognizing the constraints existinginstitutions exert on actors we note that existing congurations of insti-tutional arrangements institutional sectors and organizational propertiescan limit but also inuence the degree to which norms and rules canchange Many scholars (North 1990 Ostrom 1990) have emphasized therole of norms rules ideologies and values in limiting the ability of actorsto develop new institutional arrangements which radically depart fromexisting ones However there is a two-way process of interaction aprocess of co-evolution (Baum and Singh 1994a Nelson 1994) Thus asFriedland and Alford (1991 244) argue actors develop new institutionalarrangements by recognizing organizational failures and low perfor-mance but these new arrangements do not necessarily change a societyrsquosnorms rules habits and its underlying value systemCampbell (1997) advances our understanding of how existing institu-

tions enable actors to construct radically new institutions He recognizesthat most institutions and institutional arrangements embody a degreeof inertia from which it is difcult for actors to depart Moreover thedifferential power relations among actors make it unlikely that lesspowerful actors can change the existing power structure of a societyMost change in the institutional components of a society evolves througha process of constrained selection which reects to a considerable degreethe existing arrangements and power relations This becomes the inter-pretative frame for social actors with institutionalized scripts and ritualswhich tend to be taken for granted and appear to be quite ordinary andnatural This institutionalized scripting of the social world is an extremelyimportant source of social stability and inertiaCampbell has developed a very fruitful explanation of change in terms

of the various institutional components outlined here He distinguishesbetween incremental and radical institutional change and argues thatradical institutional change occurs when social actors with widelydiffering norms cultural scripts and rituals for action engage in intenseinteraction with each other (see also Knight 1992) The more diverse theinteractions the greater the potential for institutional change He arguesthat as shifts in the composition of interaction among social actors occurchanges in the interpretation of problems and interests will follow Ifthere are only minor changes in the extent to which diverse actorsinteract there will be only modest institutional change But the morefundamental the changes in the interaction of diverse social actors thegreater the change in the way that actors interpret their world And thegreater the change in the way that actors interpret their world the morelikely that radical changes in various institutional components willfollow In other words diverse decision making tends to create a widerrange of interpretative frameworks of the social world than is likely tobe the case if social actors operate in isolation

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

625

This perspective of diverse interactions among social actors facilitatesour ability to move beyond the traditional view that existing institutionsconstrain the range of institutional alternatives which actors face Theolder view has tended to suggest that actors engage in institutionalchange by extending existing institutional principles habits and conven-tions Specifying the conditions which facilitate fundamental institutionalchange permits a much richer perspective for institutional analysisCampbellrsquos perspective about the conditions under which fundamentalor radical change in institutional components occurs is consistent withmy own research about the conditions under which major discoveriesor fundamental new knowledge occur the more scholars with diversebackgrounds interact with intensity and frequency the greater the like-lihood that they will develop new and alternative ways of thinking(Hollingsworth and Hollingsworth 2000)These views of institutional change are also consistent with the recent

work of Nelson and his colleagues (Mowery and Nelson 1999 Murmann1998) They report the research on a variety of industrial sectors in anumber of different countries revealing that networks of actors in partic-ular industries business and professional associations universityresearch departments and institutes are frequently able to use their collec-tive power to modify a societyrsquos property rights research system andgovernmental policies It is the intense and frequent interaction of thesediverse complex sets of actors ndash sometimes constituting a technologysystem ndash that often leads to incremental changes in a societyrsquos total socialsystem of production It is this kind of collective action that brings aboutorganizational change change in institutional sectors institutionalarrangements and even in the rules and norms of a society And if weare to understand how the style of innovativeness of a society changeswe must rst understand the changes in its institutional makeup It ishighly unlikely that a society will develop a new style of innovativenesswithout changing its institutional structure

FIRMS THEIR INSTITUTIONAL ENVIRONMENT ANDNATIONAL SYSTEMS OF INNOVATION

Variation in the innovative style of societies is shaped by their insti-tutional makeup ndash the various levels which are discussed aboveinstitutions institutional arrangements the structure and cohesion of institutional sectors which constitute a societyrsquos social system of pro-duction and the structure and culture of its organizations ndash especiallyits business rms and other research organizations By undertaking aninstitutional analysis of a society one can begin to understand in whatkinds of organizations the production of specic kinds of knowledgetakes place and how this is linked to particular kinds of innovativeness

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

626

With such a perspective one can gain a rich understanding of why somesocieties excel in the production of radical breakthroughs in basic andapplied science and in developing radically new products and why othersocieties excel in more incremental innovationsReecting on these issues one cannot help being impressed with the

contrasts between the innovative styles of German and American orga-nizations For example Germany has continued to be very successful in making incremental innovations in many industries in which it wasvery competitive well before World War II paper printing materialsmachinery electrotechnical products motor vehicles chemicals textileyarns and fabrics But the Germans have been much less successful indeveloping totally new products or related innovations in biotechnologyelectronics telecommunications aircraft as well as other newer indus-trial areas In contrast the Americans have been very innovative in these and other industries with products having very short lives andwith technologies which change very rapidly and are very complex(Hollingsworth 1991 Soskice 1997 1999)The importance of institutions and innovativeness for international

competitiveness has led to the recognition that there are national systemsof innovation (Lundvall 1992 Nelson 1993 Edquist 1997 Hage andHollingsworth 2000) But even before the recent studies of nationalsystems of innovation Landes (1969) was a proponent of these viewswhen he made the argument that Germany had developed a nationalsystem of innovation by building on its education and scientic researchsystems which were lacking in Britain and elsewhere However the concept of a national system of innovation immediately poses theproblem of what kind of innovations should be consideredDavid Soskice an economist at the Wissenschaftszentrum Berlin has

addressed a set of problems somewhat complementary to the perspec-tive raised here His work (Soskice 1990 1997 1999) suggests that weare sensitive to the institutional environment in which the followingactors are embedded the employees of rms the owners and nanciersof rms the competitors of rms and nally their collaborators ndash aperspective complementary to Hollingsworthrsquos work on social systemsof production (Hollingsworth 1997 1998) When one reects on the insti-tutional contextsocial system of production within which these actorsare embedded one becomes aware that those countries which have some-what inexible labor markets external to rms but a business system in which rms have long-term commitments to their employees arestrongly associated with incremental but hardly at all with radical inno-vations In the rms of such countries (eg Germany and Japan) thereis much more consensus decision making than in societies where a lsquohireand rersquo set of practices is quite pervasive And this kind of consensusdecision making ndash while highly conducive to incremental innovativeness

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

627

ndash limits the capacity for radical innovations to occur This kind of lsquolongtermismrsquo set of norms and rules occurs not only in the employmentpractices of incrementally innovative rms but also in their nancingand ownership patterns (Hollingsworth 1997) During the past thirtyyears these were characteristics of many rms in Germany and JapanSignicantly in both of these countries venture capital markets werenot highly developed nancing tended to be based more on long-termbank loans than on the equity markets and the relationships of rmstended to be relatively stable And these stable institutional arrange-ments also contributed to a high degree of innovativeness ndash butincremental rather than radical innovationsMeantime we should also pursue Soskicersquos ideas (1999) about how

rmsrsquo relationships with their competitors and collaborators shape theirstyles of innovativeness In those advanced industrial societies in whichbusiness rms are highly mobilized into business associations theretends to be a high degree of standardized quality controls over technicalnorms and a high degree of reliance on legal sanctions as monitoringinstruments vis-agrave-vis association members As a result of this high degreeof integration of rms into associations member rms are more likelyto be associated with incremental innovations than is likely to be the case with rms in those societies where associations are less welldeveloped In the latter type of society radical innovations are morelikely to occur In those advanced industrial societies where rms havelong-term stable relationships with collaborators in product design andproduction the style of innovation is likely to be incremental And it is in those societies in which there is a weaker tradition of strongcollaboration among rms that radical innovations are more likely tooccurObviously even in a country where there may be a great deal of

radical innovativeness most innovations will be of an incremental typefor there are simply fewer radical innovations But Germany and therest of Europe not only have had relatively few radical innovations incomparison with the United States but they also have lagged behindthe USA in developing relatively totally new industries Thus there hasbeen much less innovative activity in Europe in high-technology andnewly emerging industries than in the USAIn complementary research Audretsch and Feldman (1995) have

demonstrated that the USA has been more successful in spawning rmsengaged in the early stages of the industry life cycle but relative toGermany had an environment unfavorable to innovative activity byestablished rms especially in more mature industries German rmshave tended to be associated with a routinized technological regime ndashwhere established enterprises have had strong incentives to engage inincremental innovations but new rms have tended to have fundamental

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

628

innovative disadvantages relative to the Americans (see Audretsch 1995Winter 1982 Casper 2000)On both sides of the Atlantic universities engage in basic and applied

research But signicantly the institutional environment in which univer-sities are embedded inuences the degree to which the research is likelyto represent major breakthroughs or is likely to have more modestimplications Recent research in biomedical science demonstrates that inthose societies in which university professors andor graduate studentshave the following characteristics there are likely to be relatively majorresearch breakthroughs graduate students have a long training periodhigh dependency on a major professor low autonomy in dening thenature of a research project restricted job options and a university envi-ronment where there is a very modest or nonexistent entrepreneurialculture For example in Germany ndash a country with relatively few radicalbreakthroughs ndash the scholar who nishes a habilitation will have spentmany years on a research project before qualifying for a permanent acad-emic position will have been in a high-dependency position with a seniorprofessor will have had limited autonomy to dene a research projectand will be constrained from pursuing a high-risk research project Theyoung German will have an extraordinarily high level of training butwill have faced a number of disincentives to undertake high-risk researchAnd as soon as heshe has completed the habilitation and becomes aprofessor administrative responsibilities constitute another disincentivefrom engaging in high-risk or long-term research projects Moreover theGerman professor is a civil servant with a more stable and secure incomethan hisher American counterpart In a very different institutionalenvironment American professors in large research universities tend tobecome much more entrepreneurial Today many American professorsbehave as though they were operating their own rms within univer-sities and thus devote considerable time to generating funds for researchprojects assistants secretaries supplies etc And the variation in thisentrepreneurial spirit across countries is associated with variation in thetaking of high-risk research strategies and the making of major researchbreakthroughsIn sum societies vary in terms of the institutional environment in

which their organizations are embedded and this variation inuencesthe degree to which they have industrial activities in the early or latestages of the industry life cycle successful new or mature industriesand radical vs incremental innovations And it is this variation whichis encouraging rms in countries with relatively few new industries and with few or no radical innovations to invest in those countries with many new industries and numerous radical innovations ndash as astrategy to be able to transfer new technologies from the latter institu-tional environment to the former

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

629

Table 4 summarizes the way that the norms rules and values of a society are associated with Americarsquos and Germanyrsquos institutionalarrangements their social system of production and their styles of inno-vativeness

CONCLUSION

Several themes deserve to be reemphasized Most importantly institu-tional analysis is emerging as a dominant issue in several social sciencedisciplines The study of various institutional components discussedabove is a complex subject requiring multiple levels of analysis and aknowledge of multiple sectors within societies It requires a theoreticalperspective that evolves from an empirical analysis in different societiesIf we are to advance the study of institutional analysis we need cross-disciplinary collaboration and team research which are not readilyachieved given the structure of most of our research universitiesAt present scholars in each academic discipline have their own distinc-

tive strategies for studying institutional analysis Thus economists tendto use several approaches (North 1990 Williamson 1985 Hodgson 19881998 1999) anthropologists another (Geertz 1995) and sociologistspolitical scientists and historians employ other strategies (Campbell etal 1991 Powell and DiMaggio 1991 Katznelson 1998 Hall and Taylor1996) With practitioners in various disciplines pursuing their distinctiveapproaches to institutional analysis there is low potential for collabo-rative institutional analysis across the social sciences The lessons fromthe biological sciences over the last 40 years are very instructive oncepractitioners from various biological disciplines began to cooperate instudying biological phenomena by using the same concepts at the mole-cular level theoretical advance was very rapid Comparable phenomenatook place in the development of other hybrid elds eg biochemistrybiophysics (Judson 1979 Olby 1979 Kohler 1982 1994) Similarly weresocial scientists in various disciplines to work collaboratively in the areaof institutional analysis there would very likely be an acceleration intheir theoretical knowledgeThis is not a plea that everyone engaged in institutional analysis should

do the same kind of research Indeed just as biologists working at themolecular level study many different kinds of problems similarly socialscientists engaged in institutional analysis would also work on manyseparate problems but in the spirit of a collective enterprise Some wouldconcentrate their attention on the study of rules norms habits conven-tions and values while others would study how these are associatedwith congurations of various institutional arrangements (types ofmarkets hierarchies networks associations communities clans statesetc) Others would work on specic institutional sectors (eg education

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

630

Table 4

Rules norms hab

its an

d values associated w

ith sp

ecic co

ngu

ration

s of coo

rdinating mecha

nism

s soc

ial sy

stem

s of

prod

uction

an

d nationa

l styles of inno

vation

USA

Germany

Norms rules values which provide incentives for

Norms rules values which provide incentives for

short-term horizons

long-term horizons

Con

gu

rative

forms of

Marke

ts c

orpo

rate hierarchies reg

ulatory state

Assoc

iation

s (busine

ss assoc

iation

s la

bor union

s)

coordinating

corp

oratist-type

networks

marke

ts c

orporatist-

mecha

nism

s (listed in

type

state

order of im

portanc

e in

the particu

lar social

system

)

Social sys

tem of

prod

uction

Bus

iness system

Firm

struc

tures

Con

glom

erates b

ut with mov

emen

t toward

Firm

more co

ngru

ent with pa

rticular prod

ucts

elim

ination of produc

tion

in som

e sectors

emph

asis on long

-term strateg

ies rm

s are well

emph

asis on sh

ort-term

strateg

ies high

integrated

differentiation

between divisions

of rms

Firm

s mob

ilized in

Low

High

and in

tegrated

into busine

ss

associations

Indu

strial relations

Low

deg

ree of jo

b secu

rity rigid in

ternal lab

or

High deg

ree of jo

b secu

rity

exible interna

l lab

or

system

marke

t ex

ible externa

l labo

r marke

tsmarke

t rigid externa

l lab

or m

arke

tPatterns of

Dispersed owne

rship frequ

ent turnov

erMore stab

le owne

rship patterns

owne

rship

Table 4

continue

d

USA

Germany

Norms rules values which provide incentives for

Norms rules values which provide incentives for

short-term horizons

long-term horizons

Social system of

Prod

uction

Bus

iness system

Training for labo

r an

dHigh deg

ree of voluntary decision mak

ing by

Collective an

d m

ore co

mpulso

ry d

ecisions

abo

ut

man

agem

ent

individuals as to ho

w m

uch

tim

e an

d ene

rgy

how m

uch ed

ucation

is ne

eded

he

avy em

pha

sis

to inv

est in educ

ation heavy

emph

asis on

on techn

ical training (eg en

gine

ering)

busine

ss sch

ools training

in marke

ting

and

sales accou

nting

Fina

ncial marke

tsEqu

ity an

d ven

ture cap

ital m

arke

ts highly

Equ

ity an

d ven

ture cap

ital m

arke

ts w

eakly

dev

elop

ed na

ncing from

cen

tral gov

ernm

ent

deve

lope

d he

avy em

pha

sis on

corporate loan

sto priva

te rm

s for dev

elop

men

t of new

tech

nologies related

to na

tion

al security

Unive

rsity research

High individual auton

omy high poten

tial for

Low

ind

ividua

l au

tono

my until almost 40

yea

rs of

system

individual creativity u

nive

rsity pr

ofesso

rs

age poten

tial for ind

ividual creativity high

ly

high

ly entreprene

urial

cons

traine

d unive

rsity pr

ofessors not so

entrep

rene

urial

Inno

vative

styles

Highly su

ccessful in m

ajor break

through

s in

High pe

rforman

ce in proc

ess an

d inc

remen

tal

basic an

d app

lied scien

ce go

od in dev

elop

ing

inno

vation

s in existing produ

cts not so su

ccessful

totally

new

lines of pr

oducts an

d in m

ajor break

through

s in basic scien

ce or in

tech

nologies high

deg

ree of entrepr

eneu

rship

deve

loping

totally new

lines of prod

ucts of

amon

g research

ers

tech

nologies low d

egree of entrepr

eneu

rship

amon

g research

ers

Table 4

continue

d

USA

Germany

Norms rules values which provide incentives for

Norms rules values which provide incentives for

short-term horizons

long-term horizons

Prod

ucts in w

hich

Pr

oducts with sh

ort liv

es prod

ucts in w

hich

Older produ

cts with long

history machine

ry pap

er

society excels

tech

nology

is bo

th highly co

mplex

and

pr

inting

textile

s m

otor veh

icles m

achine

too

lsch

ange

s rapidly en

tertainm

ent indus

try

pha

rmaceu

ticals biotechn

olog

y softw

are

aerosp

ace

business systems nancial markets and systems of research) Distinctivewould be the recognition of how these separate studies are linkedtogether how each institutional sector is linked to a societyrsquos normsrules values etc and its conguration of institutional arrangementsSimilarly other institutional analysts might focus on the study of orga-nizations with a major concern for how the institutional environmentof organizations (eg norms rules values congurations of institutionalarrangements institutional sectors) inuenced the structure culture andoutcomes of organizations And nally others might study how all ofthe aforementioned aspects of institutional analysis inuenced a societyrsquosoverall performance Nelson and Sampat (1998) have appropriatelyreminded us that we should engage in institutional analysis in order toaddress specic problems but the view developed in this article is thatwe rst need some mapping of the terrain of institutional analysis inorder for our diverse enterprises to become a meaningful collective enter-prise especially if we are to understand how institutions inuence theinnovative processThe argument developed here is not that all social scientists should engage

in institutional analysis Obviously there are numerous other importantareas of research which lie beyond the eld of institutional analysis Nordoes this article argue that institutional analysis is the most importantsubject for social scientistsSince the structure and culture of our research universities reect an

enormous amount of disciplinary fragmentation what are the prospectsfor promoting the kind of collective and interdisciplinary research agendaproposed above In one respect the prospects are not encouragingDuring the past half century the efforts of American universities to have effective communication across the biological sciences have variedgreatly in their success At the University of California-Berkeley it was not until the 1980s that there was a serious effort to integrate thebiological sciences and at that time the campus abolished seventeendepartments to promote an integrated biological program But at theUniversity of Wisconsin the biological sciences are still very fragmentedinto dozens of different departments in botany plant pathology zoologyand physiology with two departments of biochemistry and two depart-ments of geneticsThere have been other approaches which have worked very well histor-

ically for promoting scientic integration amidst scientic diversity Forexample several summer schools and special institutes were organizedin the United States during the 1940s in an effort to overcome the frag-mentation in the biological sciences These programs made it possiblefor scientists in diverse academic disciplines and from different univer-sities and countries to come together to engage in intense and frequentinteraction Perhaps the most effective program was that held in the

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

634

summers at Cold Spring Harbor New York where a research agendafor the study of molecular biology was developed (Fischer and Lipson1988 Stent and Watson 1966) Scientists went there summer aftersummer and when they returned to their universities many began therestructuring of the biological sciences in their universitiesIn our own day we need summer institutes and other similar programs

so that scholars from diverse backgrounds can come together anddevelop a collective agenda for doing institutional analysis and forstudying the impact of institutions on the performance of societies ndashinnovativeness being one such subject If we can develop such instituteswe will increase the prospects for transforming our universities Thereis reason to be optimistic that once an eclectic group of scholars workwith a common set of concepts we can have genuine theoretical advanceif researchers can bring their diverse expertise to bear on a common setof problems There is considerable evidence that institutional analysisoffers high potential for advancing our knowledge of innovations andtechnological change (Hollingsworth and Hollingsworth 2000) If we aregoing to make substantial theoretical advance it is not likely to occurfrom the vantage point of a multiplicity of disciplines Rather it is likelyto occur as we develop a new intellectual framework and our own acad-emic journals This is a tall agenda which requires us to reassess theway that knowledge is presently produced in our societies

NOTE

This article is part of a much larger agenda involving the study of institutionsorganizations and innovations in which I am involved My debts are many andvaried The rst version was written while I was in residence in the ResearchUnit for Institutional Change and European Integration of the Austrian Academyof Sciences in Vienna in the summer of 1998 I am especially grateful to ProfessorsEgon Matzner and Sonja Puntscher-Riekmann for providing the stimulating envi-ronment in which to think through the ideas developed here Egon Matzner hasengaged in very stimulating discussions with me about the importance of insti-tutions in the general eld of socio-economics The nal version of the articlewas completed during my residence as a fellow at the Neurosciences Institutein La Jolla California I also wish to thank my colleagues in the InnovationTheme Project at the Netherlands Institute for Advanced Study (NIAS) during1998ndash9 and Edgar Grande and Karl Muumlller for stimulating conversations aboutthe interaction of institutions and organizations in facilitating particular typesof innovation I am especially grateful to Jerald Hage for helping me to developa number of issues in the article and for his useful comments on the rst draftFrans van Waarden early on helped me to work through the idea of multiplelevels of institutional analysis Others who have helped to develop the ideas inthe article have been my colleagues in the research project lsquoComparative socialsystems of productionrsquo especially Tom Burns Christel Lane Yoshitaka OkadaWolfgang Streeck and Richard Whitley My colleagues Robert Boyer SteveCasper and David Soskice have taught me much about how incentives providedby the institutional environments of rms inuence individual decision making

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

635

But my most important debt is to Ellen Jane Hollingsworth who has repeatedlyquestioned every single assumption of this article Without her rich and contin-uous tutoring this article would be even less developed than it is at presentFinally without the insightful assistance of David Gear in many phases of thisarticle it would not have been completed

REFERENCES

Aoki Masahiko (1990) lsquoTowards an economic model of the Japanese rmrsquo Journalof Economic Literature March 1ndash27

Arora A Landau Ralph and Rosenberg Nathan (eds) (1998) Chemicals and Long-Term Economic Growth Insights from the Chemical Industry New York JohnWiley amp Sons

Arrow Kenneth (1974) The Limits of Organization New York NortonArthur W B (1988a) lsquoSelf-reinforcing mechanisms in economicsrsquo in P W

Anderson K J Arrow and D Pines (eds) The Economy as an Evolving ComplexSystem Redwood City Calif Addison-Wesley

mdashmdash(1988b) lsquoCompeting technologies an overviewrsquo in G Dosi C Freeman RNelson G Silverberg and L Soete (eds) Technical Change and Economic TheoryLondon Pinter Publishers pp 590ndash607

Audretsch David B (1995) Innovation and Industry Evolution Cambridge MassMIT Press

Audretsch David B and Feldman Maryann P (1995) lsquoInnovative clusters andthe industry life cyclersquo Discussion Paper Wissenschaftszentrum BerlinMarket Processes and Corporate Development Research Unit FS-IV-95-7

Baum Joel A C and Oliver Christine (1992) lsquoInstitutional embeddedness andthe dynamics of organizational populationsrsquo American Sociological Review 57540ndash59

Baum Joel A C and Singh Jitendra V (1994a) lsquoOrganizationndashenvironment coevo-lutionrsquo in Joel A C Baum and Jitendra V Singh (eds) The EvolutionaryDynamics of Organizations New York Oxford University Press pp 379ndash402

mdashmdash(eds) (1994b) The Evolutionary Dynamics of Organizations New York OxfordUniversity Press

Berger Suzanne and Dore Ronald (eds) (1996) National Diversity and GlobalCapitalism Ithaca NY Cornell University Press

Boyer Robert (1997) lsquoThe variety and unequal performance of really existingmarkets farewell to Doctor Panglossrsquo in J Rogers Hollingsworth and RobertBoyer (eds) Contemporary Capitalism the Embeddedness of InstitutionsCambridge and New York Cambridge University Press pp 55ndash93

Burns Tom R and Dietz Thomas (2001 forthcoming) lsquoSocial rule system theorysocial action institutional arrangements and evolutionary processesrsquo in JRogers Hollingsworth and Karl Muumlller (eds) Advances in SocioeconomicsLanham Md Rowman amp Littleeld

Burns Tom R and Flam Helena (1987) The Shaping of Social Organization BeverlyHills Calif Sage

Calvert Randall L (1995) lsquoThe rational choice theory of social institutionsrsquo inJeffrey S Banks and Eric A Hanuskek (eds) Modern Political Economy NewYork Cambridge University Press pp 216ndash66

Camic Charles (1986) lsquoThe matter of habitrsquo American Journal of Sociology 911039ndash87

Campbell John L (1997) lsquoMechanisms of evolutionary change in economicsgovernance interaction interpretation and bricolagersquo in Lars Magnusson

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

636

much closer to the mark in his argument that most societal institutionalarrangements exist as a result of custom and habit and are thereforeinefcient At any moment in time the world often appears to its actorsas very complex and uncertain For this reason actors often engage incontradictory forms of behavior pursuing different strategies as hedgesagainst a very uncertain world (Lanzara 1998) And their hedging andcontradictory forms of behavior may lead in somewhat different soci-etal directions all constrained by the institutional fabric within whichthe actors are embedded This kind of contradictory behavior occurs inpart because of the contradictions inherent in the dominant congura-tions of institutional arrangements with their contradictory logics (seelsquoSecond level of analysisrsquo above)Despite the emphasis on the logic of institutional continuity this is

not an argument that systems change along some predetermined pathThere are critical turning points in the history of highly industrializedsocieties but with the choices limited by the existing institutional terrainBeing path-dependent a social system of production continues along aparticular logic until or unless a fundamental societal crisis intervenes(Milgrom et al 1991 Krugman 1991 Durlauf 1991 Hollingsworth 1991Pred 1966 David 1988 Hodgson 1998)At this point it is important to confront the question of how social

systems of production evolve And this question gets to the heart of theproblem of building complex societal structures Certainly social systemsof production did not emerge from some process of social engineeringMoreover the various component parts of each social system of produc-tion have often not been designed to be part of a social system ofproduction The component parts of each social system of productionhave emerged more often than not as unintended by-products of goalswhich various actors had in mind at earlier moments in time It is usuallythe case that actors at time lsquotrsquo fail to comprehend the long-term conse-quences of their actions Wolfgang Streeck the author of several brilliantpapers on social systems of production argues that institutional sectordesigns created for one purpose generally address that goal but overtime those designs have quite unintended consequences (Streeck 1997a1997b Wright 1998) Describing the conguration of the German socialsystem of production of the 1970s and 1980s Streeck points out that itwas the unintended by-product of multiple points in history Someelements were pre-Wilhelmian others were introduced by the Alliedpowers after 1945 and others emerged during the years of the GermanFederal Republic All component parts of the German social system ofproduction

were and continue to be changing for their own reasons as wellas in reaction to each other and certainly there can be no presump-tion of a pre-established t between them even though one might

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

618

want to allow for some reinforcement effects of the modelrsquos histor-ically contingent social and economic success

(Streeck 1997b 54)

In sum the emergence of social systems of production is a long-termevolutionary process with each part interacting with its environment andresulting in a congurative whole but with no previous design by eithera single actor or a collectivity of actors As suggested above actors do not recreate their world anew At any moment in history the totalinstitutional complex of a society necessarily contains the resources and legacies of its past Preexisting institutional complexes are nevercompletely wiped out Societies consist of multiple layers of history withtheir diverse logics of action However the cumulative effect of smallperipheral changes in altering particular institutional sectors and insti-tutional arrangements can be substantial (Murmann 1998 Lanzara 1998)lsquoInstitution building is affected more by the ways in which people havecodied the past than by how they envision the future There are manymore potential resources in pre-existing institutional arrangements thanit is usually assumed or suspectedrsquo (Lanzara 1998 30)

Fourth level of analysis

The next level of institutional analysis organizational structures is some-what more controversial As suggested above Douglass North (1990)draws a sharp distinction between institutions and organizations Morerecently many organizational theorists (Powell 1991 Powell andDiMaggio 1991 DiMaggio and Powell 1983 Baum and Oliver 1992Kondra and Hinings 1998 Townley 1997) argue that institutional rulesnorms and conventions unfold in tandem with organizational structuresand this is my positionThe literature which focuses on how institutions inuence organiza-

tions is quite different from two other theoretical literatures which alsoare concerned with how organizational environments shape the behaviorof organizations There are resource dependency theorists (Pfeffer andSalancik 1978) who emphasize the role of environmental resources inshaping organizations while Hannan and Freeman (1977 1984) and otherpopulation ecologists emphasize the survival of organizations givencertain kinds of organizational conditions (Orru et al 1991)The institutionalist perspective as implied above emphasizes the

normative environment in which organizations are embedded It is a perspective which focuses on the way in which organizations in their behavior tend to conform to the institutional rules and normswhich are dominant in the organizational environment However allthree perspectives ndash the institutionalist the population ecology and

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

619

the resource dependency ndash emphasize how the environment inuencesorganizations and how organizations which are subject to the sameenvironment tend to converge in their behavior to have what DiMaggioand Powell have labeled organizational isomorphism (DiMaggio andPowell 1983)There has emerged a vast literature which demonstrates that within

each society there are scal political judicial and other regulatory normswhich limit and shape the culture and structure of organizationalbehavior The normative institutional environment of organizations limits the options of what organizations do in a particular society and inuences the patterns of ownership relations with suppliers andcustomers In short it is the normative environment of organizationswhich denes within a particular society what is socially acceptablebehavior for organizations (Hamilton and Biggart 1988 Hollingsworthand Hollingsworth 2000 Meyer and Rowan 1991 Zucker 1987 1988Orru et al 1991 Townley 1997)Thus far most of the analyses on isomorphic pressures operating on

organizations have been cross-sectional in nature However the entireconcept of isomorphism implies that there are strong environmental pres-sures exerted on organizations and in order to observe this phenomenonit is important that we have longitudinal studies which assess howchanges in an organizationrsquos institutional environment inuence changesin the structure and culture of organizations Moreover the expectationhere is that when one analyzes the patterns of organizational changeover time the isomorphism will be holistic ndash that is within a particularsociety there will be a tendency for the internal structure of particulartypes of organizations to converge in large part because of the pres-sures to conform to the changes in the external norms rules and valuesA close reading of Aoki (1990) and Slack and Hinings (1994) nds thatchanges in the national institutional environment of organizations haveinuenced changes in the internal structure of organizations within aspecic industry of a particular societyHowever it is inappropriate to view organizations as changing only

in relation to exogenous environmental change There is a kind of co-evolution which occurs between organizations and their institutionalenvironments Co-evolution implies nonlinear feedback between orga-nizations and their institutional environments This is of course quitedifferent from the more common way of studying organizations bymodeling relations between independent and dependent variables Ofcourse the more common type of analysis is quite appropriate whenthere are simple relationships which do not involve complex feedbackprocesses But in the strategy for institutional analysis which is proposedherein it is less useful to separate independent from dependent vari-ables and more useful to understand the interacting and co-evolutionary

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

620

processes among all of the levels of institutional analysis discussed inthe various sections above (Baum and Singh 1994a 379ndash80 Mowery andNelson 1999 Murmann 1998)The logic of the perspective herein suggests that within every

society there is variation in the structure and culture of business rmsuniversities and other complex organizations (Kondra and Hinings1998) However this variation takes place within parameters which aresystem specic For example every German rm and every Germanuniversity is different from every other German university and rm but there are a set of cultural and structural characteristics which distin-guish German rms and universities from those in America Similarlyeach Japanese rm and university is unique yet they have a set ofcultural and structural characteristics which are system specic andwhich differentiate them from business rms and universities in other countriesIn all societies each organization has its own distinctive organizational

rules norms and conventions which are subordinate to the meta norms and rules of the larger society within which they are embeddedHowever the strength of the institutional environment within whichorganizations are embedded varies from society to society Some societies have multiple institutional environments and there is hetero-geneity in terms of what constitutes appropriate behavior oforganizations In such societies there can be conict over whatinstitutional logics should regulate specic organizational functions(Friedland and Alford 1991 Townley 1997 262ndash3) In those societies inwhich the institutional norms habits and rules are most developed and in which the institutional pressures to conform are greatest thereis less variation in the structure and culture of business rms and variouskinds of research organizations In such societies the connectednessbetween research organizations and their institutional environments is sufciently strong for organizations to have low autonomy to pursue independent strategies and goals in these societies there is a great deal of organizational isomorphism Conversely the weaker the institutional environment in which research organizations areembedded the greater the variation in the structure and culture of business rms and research organizations Moreover where the institu-tional environments are more weakly developed organizations havegreater autonomy and exibility to respond to the development of new knowledge and to be highly innovative Hence it is in those societies where the institutional environments are most developed and most rigid and when there is less organizational autonomy and exibility that fewer radical innovations in basic and applied science as well as in totally new products and industrial sectors havebeen made

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

621

Fifth level of analysis

We come next to the outputs and performance of the various institutionalcomponents of a society It is at this level that institutional componentsare more pragmatic and exible For example in the legal sector thereare specic statutes and court rulings in the state sector there are specicpolicies in the business sector there are new products new technolo-gies and market strategies It is at this nal level of analysis thatinstitutional spheres are most open and susceptible to change and cross-national mimicry is easiest and most commonThrough the outputs of the society we can obtain some assessment as

to how well a society is performing Moreover we can assess how inno-vative it is how egalitarian it is in the distribution of its resources howegalitarian it is in terms of levels of health education etc Just as it iscomplicated to measure how egalitarian a society is similarly it is dif-cult to assess how innovative it is For example societies vary in makingincremental and radical breakthroughs in basic and applied science in developing totally new products and new kinds of organizations inincremental and process innovations in existing products and organiza-tions in developing new and different forms of marketing ndash bothdomestically and globally (Hage and Hollingsworth 2000) Of coursesocieties vary in their rates of economic growth in their rates of economicproductivity in their quality of life (eg rates of crime life expectancyetc) All of these performance criteria feed back and inuence each ofthe levels of institutional analysis discussed above ndash rules norms valuesetc institutional arrangements institutional sectors the structure andform of organizations Moreover performance measures may inuenceeach other For example if a society has very low rates of growth itmay not be very innovative in some of the types of innovative activitiesmentioned above Good or poor performances of certain kinds inuenceother performance indicators (Hollingsworth et al 1990)Different institutional arrangements and different social systems of

production result in different types of economic performance Hence solong as societies have different social systems of production there areserious constraints on the degree to which they can converge in theirinnovative styles Different social systems of production tend to maxi-mize in a more or less explicit manner different performance criteriausually mixing considerations about static and dynamic efciency protsecurity social peace and economic andor political power In short incontrast to the implications of neoclassical economic theory in real-worldeconomies there are no universal standards all economically rationalactors attempt to maximize Economic history provides numerous exam-ples of how a great variety of principles of rationality are implementedin different societies

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

622

A critical question in institutional analysis is whether an existing socialsystem of production which supports a set of routines for a particularkind of technology and industry can shift from old practices to newones There are numerous historical accounts of how the social systemof production in a particular society which worked so well for a numberof years could not adapt to new technologies (Lazonick 1990 Veblen1915 Nelson 1994) Schumpeter (1983) and Freeman (1991) have devel-oped the idea of a techno-economic paradigm which suggests thatdifferent technologies require specic forms of organizational arrange-ments ndash or what I called social systems of production Because oflsquoinstitutional inertiarsquo the social system of production of a society maynot be able to adapt to a new techno-economic paradigm (Nelson 199458 Lanzara 1998)Whether a social system of production can sustain its performance

standards depends not only on its intrinsic economic rationality but alsoon where it ts into a larger system If a particular social system ofproduction is immune from the innovativeness and competition of analternative system survival can be long lasting But if different socialsystems of production with diverging criteria of good economic perfor-mance meet in the world arena the arbitrariness of nationally imposedconstructed performance standards may be superseded by alternativeperformance criteria as a result of international competitivenessThe world economy is also socially constructed just as are national

economies Even if different social systems of production are competingin the international arena it is not always possible to determine whichis more competitively effective at any moment in time Hegemonicnation-states in the short run shape the rules of trade to favor their indus-trial sectors and rms But the history of hegemonic powers suggeststhat in the longer run such social systems of production sustainedlargely by military and political power eventually give way to moreinnovative and competitive social systems of production In our ownday as nation-states are increasingly integrated into a world economyeven hegemonic powers lose their innovative and competitive advan-tage and their share of world output decreases Such a country mayattempt to restructure its institutional arrangements and to readjust itsperformance preferences But to restructure its system of innovationgenerally calls for a major redistribution of power within a societyLargely for this reason societies have historically had limited capacityto reconstruct a system of innovation in the image of their major competi-tors Each countryrsquos social system of production is a conguration of ahost of norms rules and values as well as of institutional arrangementsEach system is constantly changing and is open to inuence from othersystems And indeed many technologies and practices diffuse from onesociety to another but a societyrsquos capacity to be innovative is constrained

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

623

by the existing social system of production Thus the same technologymay exist in numerous countries but how it is employed and how itinuences societal outcomes and performances varies from one institu-tional conguration or society to another (Hollingsworth 1998)

PERSPECTIVES ON INSTITUTIONAL CHANGE

A fundamental problem which remains unresolved in institutionalanalysis involves the nature of institutional change There is much confu-sion and miscommunication in the social sciences about institutionalchange in large part because we do not even know how to measure therate of institutional change Of course we know that there are bothexternal and internal focuses for change and we have had a good bitof scholarship on this subject What is not clearly understood is thatinstitutions and institutional arrangements within societies are histori-cally rooted that there is a great deal of path-dependency to the waythat institutions evolve Moreover the more intricately linked each sectoris with each other sector and with a societyrsquos rules and norms the lesschoice actors have to devise new institutions and institutional arrange-ments On the other hand if a sector or actor is too isolated it may betoo weak within a society to bring about any kind of effective changeThese considerations are very critical to the problem of how muchfreedom actors have to create new institutions afresh to what extentinstitutions diffuse from one society to another (Murmann 1998 Lanzara1998) This is a fundamental question for students of institutionalanalysis as we attempt to understand how the different institutionallevels are linked togetherThere are a variety of reasons why there is confusion in studying insti-

tutional change There is disagreement over the extent to which actorshave the freedom to build new institutional arrangements and if theydo to what extent the new arrangements and organizational patternsmay depart from past practices Among some institutional economists(Williamson 1985) there is a basic assumption that actors build efcientinstitutions by pursuing their self-interest and by promoting theirstrategic goals Moreover there is the assumption that a mixing of fullrationality with market competition tends to produce all of the optimalinstitutions that are needed in order to coordinate a complex capitalisteconomy (Hollingsworth and Boyer 1997 ch 14 Hodgson 1999) Incontrast the argument developed here is much more complexSocial institutions are historically rooted and there is a great deal of

path-dependency in the way that various institutional componentsevolve The shape of institutional congurations at any moment limitsthe type of options for change Because there is a great deal of institu-tional inertia radical change in the institutional components of a society

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

624

is uncommon (Lanzara 1998) Recognizing the constraints existinginstitutions exert on actors we note that existing congurations of insti-tutional arrangements institutional sectors and organizational propertiescan limit but also inuence the degree to which norms and rules canchange Many scholars (North 1990 Ostrom 1990) have emphasized therole of norms rules ideologies and values in limiting the ability of actorsto develop new institutional arrangements which radically depart fromexisting ones However there is a two-way process of interaction aprocess of co-evolution (Baum and Singh 1994a Nelson 1994) Thus asFriedland and Alford (1991 244) argue actors develop new institutionalarrangements by recognizing organizational failures and low perfor-mance but these new arrangements do not necessarily change a societyrsquosnorms rules habits and its underlying value systemCampbell (1997) advances our understanding of how existing institu-

tions enable actors to construct radically new institutions He recognizesthat most institutions and institutional arrangements embody a degreeof inertia from which it is difcult for actors to depart Moreover thedifferential power relations among actors make it unlikely that lesspowerful actors can change the existing power structure of a societyMost change in the institutional components of a society evolves througha process of constrained selection which reects to a considerable degreethe existing arrangements and power relations This becomes the inter-pretative frame for social actors with institutionalized scripts and ritualswhich tend to be taken for granted and appear to be quite ordinary andnatural This institutionalized scripting of the social world is an extremelyimportant source of social stability and inertiaCampbell has developed a very fruitful explanation of change in terms

of the various institutional components outlined here He distinguishesbetween incremental and radical institutional change and argues thatradical institutional change occurs when social actors with widelydiffering norms cultural scripts and rituals for action engage in intenseinteraction with each other (see also Knight 1992) The more diverse theinteractions the greater the potential for institutional change He arguesthat as shifts in the composition of interaction among social actors occurchanges in the interpretation of problems and interests will follow Ifthere are only minor changes in the extent to which diverse actorsinteract there will be only modest institutional change But the morefundamental the changes in the interaction of diverse social actors thegreater the change in the way that actors interpret their world And thegreater the change in the way that actors interpret their world the morelikely that radical changes in various institutional components willfollow In other words diverse decision making tends to create a widerrange of interpretative frameworks of the social world than is likely tobe the case if social actors operate in isolation

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

625

This perspective of diverse interactions among social actors facilitatesour ability to move beyond the traditional view that existing institutionsconstrain the range of institutional alternatives which actors face Theolder view has tended to suggest that actors engage in institutionalchange by extending existing institutional principles habits and conven-tions Specifying the conditions which facilitate fundamental institutionalchange permits a much richer perspective for institutional analysisCampbellrsquos perspective about the conditions under which fundamentalor radical change in institutional components occurs is consistent withmy own research about the conditions under which major discoveriesor fundamental new knowledge occur the more scholars with diversebackgrounds interact with intensity and frequency the greater the like-lihood that they will develop new and alternative ways of thinking(Hollingsworth and Hollingsworth 2000)These views of institutional change are also consistent with the recent

work of Nelson and his colleagues (Mowery and Nelson 1999 Murmann1998) They report the research on a variety of industrial sectors in anumber of different countries revealing that networks of actors in partic-ular industries business and professional associations universityresearch departments and institutes are frequently able to use their collec-tive power to modify a societyrsquos property rights research system andgovernmental policies It is the intense and frequent interaction of thesediverse complex sets of actors ndash sometimes constituting a technologysystem ndash that often leads to incremental changes in a societyrsquos total socialsystem of production It is this kind of collective action that brings aboutorganizational change change in institutional sectors institutionalarrangements and even in the rules and norms of a society And if weare to understand how the style of innovativeness of a society changeswe must rst understand the changes in its institutional makeup It ishighly unlikely that a society will develop a new style of innovativenesswithout changing its institutional structure

FIRMS THEIR INSTITUTIONAL ENVIRONMENT ANDNATIONAL SYSTEMS OF INNOVATION

Variation in the innovative style of societies is shaped by their insti-tutional makeup ndash the various levels which are discussed aboveinstitutions institutional arrangements the structure and cohesion of institutional sectors which constitute a societyrsquos social system of pro-duction and the structure and culture of its organizations ndash especiallyits business rms and other research organizations By undertaking aninstitutional analysis of a society one can begin to understand in whatkinds of organizations the production of specic kinds of knowledgetakes place and how this is linked to particular kinds of innovativeness

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

626

With such a perspective one can gain a rich understanding of why somesocieties excel in the production of radical breakthroughs in basic andapplied science and in developing radically new products and why othersocieties excel in more incremental innovationsReecting on these issues one cannot help being impressed with the

contrasts between the innovative styles of German and American orga-nizations For example Germany has continued to be very successful in making incremental innovations in many industries in which it wasvery competitive well before World War II paper printing materialsmachinery electrotechnical products motor vehicles chemicals textileyarns and fabrics But the Germans have been much less successful indeveloping totally new products or related innovations in biotechnologyelectronics telecommunications aircraft as well as other newer indus-trial areas In contrast the Americans have been very innovative in these and other industries with products having very short lives andwith technologies which change very rapidly and are very complex(Hollingsworth 1991 Soskice 1997 1999)The importance of institutions and innovativeness for international

competitiveness has led to the recognition that there are national systemsof innovation (Lundvall 1992 Nelson 1993 Edquist 1997 Hage andHollingsworth 2000) But even before the recent studies of nationalsystems of innovation Landes (1969) was a proponent of these viewswhen he made the argument that Germany had developed a nationalsystem of innovation by building on its education and scientic researchsystems which were lacking in Britain and elsewhere However the concept of a national system of innovation immediately poses theproblem of what kind of innovations should be consideredDavid Soskice an economist at the Wissenschaftszentrum Berlin has

addressed a set of problems somewhat complementary to the perspec-tive raised here His work (Soskice 1990 1997 1999) suggests that weare sensitive to the institutional environment in which the followingactors are embedded the employees of rms the owners and nanciersof rms the competitors of rms and nally their collaborators ndash aperspective complementary to Hollingsworthrsquos work on social systemsof production (Hollingsworth 1997 1998) When one reects on the insti-tutional contextsocial system of production within which these actorsare embedded one becomes aware that those countries which have some-what inexible labor markets external to rms but a business system in which rms have long-term commitments to their employees arestrongly associated with incremental but hardly at all with radical inno-vations In the rms of such countries (eg Germany and Japan) thereis much more consensus decision making than in societies where a lsquohireand rersquo set of practices is quite pervasive And this kind of consensusdecision making ndash while highly conducive to incremental innovativeness

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

627

ndash limits the capacity for radical innovations to occur This kind of lsquolongtermismrsquo set of norms and rules occurs not only in the employmentpractices of incrementally innovative rms but also in their nancingand ownership patterns (Hollingsworth 1997) During the past thirtyyears these were characteristics of many rms in Germany and JapanSignicantly in both of these countries venture capital markets werenot highly developed nancing tended to be based more on long-termbank loans than on the equity markets and the relationships of rmstended to be relatively stable And these stable institutional arrange-ments also contributed to a high degree of innovativeness ndash butincremental rather than radical innovationsMeantime we should also pursue Soskicersquos ideas (1999) about how

rmsrsquo relationships with their competitors and collaborators shape theirstyles of innovativeness In those advanced industrial societies in whichbusiness rms are highly mobilized into business associations theretends to be a high degree of standardized quality controls over technicalnorms and a high degree of reliance on legal sanctions as monitoringinstruments vis-agrave-vis association members As a result of this high degreeof integration of rms into associations member rms are more likelyto be associated with incremental innovations than is likely to be the case with rms in those societies where associations are less welldeveloped In the latter type of society radical innovations are morelikely to occur In those advanced industrial societies where rms havelong-term stable relationships with collaborators in product design andproduction the style of innovation is likely to be incremental And it is in those societies in which there is a weaker tradition of strongcollaboration among rms that radical innovations are more likely tooccurObviously even in a country where there may be a great deal of

radical innovativeness most innovations will be of an incremental typefor there are simply fewer radical innovations But Germany and therest of Europe not only have had relatively few radical innovations incomparison with the United States but they also have lagged behindthe USA in developing relatively totally new industries Thus there hasbeen much less innovative activity in Europe in high-technology andnewly emerging industries than in the USAIn complementary research Audretsch and Feldman (1995) have

demonstrated that the USA has been more successful in spawning rmsengaged in the early stages of the industry life cycle but relative toGermany had an environment unfavorable to innovative activity byestablished rms especially in more mature industries German rmshave tended to be associated with a routinized technological regime ndashwhere established enterprises have had strong incentives to engage inincremental innovations but new rms have tended to have fundamental

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

628

innovative disadvantages relative to the Americans (see Audretsch 1995Winter 1982 Casper 2000)On both sides of the Atlantic universities engage in basic and applied

research But signicantly the institutional environment in which univer-sities are embedded inuences the degree to which the research is likelyto represent major breakthroughs or is likely to have more modestimplications Recent research in biomedical science demonstrates that inthose societies in which university professors andor graduate studentshave the following characteristics there are likely to be relatively majorresearch breakthroughs graduate students have a long training periodhigh dependency on a major professor low autonomy in dening thenature of a research project restricted job options and a university envi-ronment where there is a very modest or nonexistent entrepreneurialculture For example in Germany ndash a country with relatively few radicalbreakthroughs ndash the scholar who nishes a habilitation will have spentmany years on a research project before qualifying for a permanent acad-emic position will have been in a high-dependency position with a seniorprofessor will have had limited autonomy to dene a research projectand will be constrained from pursuing a high-risk research project Theyoung German will have an extraordinarily high level of training butwill have faced a number of disincentives to undertake high-risk researchAnd as soon as heshe has completed the habilitation and becomes aprofessor administrative responsibilities constitute another disincentivefrom engaging in high-risk or long-term research projects Moreover theGerman professor is a civil servant with a more stable and secure incomethan hisher American counterpart In a very different institutionalenvironment American professors in large research universities tend tobecome much more entrepreneurial Today many American professorsbehave as though they were operating their own rms within univer-sities and thus devote considerable time to generating funds for researchprojects assistants secretaries supplies etc And the variation in thisentrepreneurial spirit across countries is associated with variation in thetaking of high-risk research strategies and the making of major researchbreakthroughsIn sum societies vary in terms of the institutional environment in

which their organizations are embedded and this variation inuencesthe degree to which they have industrial activities in the early or latestages of the industry life cycle successful new or mature industriesand radical vs incremental innovations And it is this variation whichis encouraging rms in countries with relatively few new industries and with few or no radical innovations to invest in those countries with many new industries and numerous radical innovations ndash as astrategy to be able to transfer new technologies from the latter institu-tional environment to the former

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

629

Table 4 summarizes the way that the norms rules and values of a society are associated with Americarsquos and Germanyrsquos institutionalarrangements their social system of production and their styles of inno-vativeness

CONCLUSION

Several themes deserve to be reemphasized Most importantly institu-tional analysis is emerging as a dominant issue in several social sciencedisciplines The study of various institutional components discussedabove is a complex subject requiring multiple levels of analysis and aknowledge of multiple sectors within societies It requires a theoreticalperspective that evolves from an empirical analysis in different societiesIf we are to advance the study of institutional analysis we need cross-disciplinary collaboration and team research which are not readilyachieved given the structure of most of our research universitiesAt present scholars in each academic discipline have their own distinc-

tive strategies for studying institutional analysis Thus economists tendto use several approaches (North 1990 Williamson 1985 Hodgson 19881998 1999) anthropologists another (Geertz 1995) and sociologistspolitical scientists and historians employ other strategies (Campbell etal 1991 Powell and DiMaggio 1991 Katznelson 1998 Hall and Taylor1996) With practitioners in various disciplines pursuing their distinctiveapproaches to institutional analysis there is low potential for collabo-rative institutional analysis across the social sciences The lessons fromthe biological sciences over the last 40 years are very instructive oncepractitioners from various biological disciplines began to cooperate instudying biological phenomena by using the same concepts at the mole-cular level theoretical advance was very rapid Comparable phenomenatook place in the development of other hybrid elds eg biochemistrybiophysics (Judson 1979 Olby 1979 Kohler 1982 1994) Similarly weresocial scientists in various disciplines to work collaboratively in the areaof institutional analysis there would very likely be an acceleration intheir theoretical knowledgeThis is not a plea that everyone engaged in institutional analysis should

do the same kind of research Indeed just as biologists working at themolecular level study many different kinds of problems similarly socialscientists engaged in institutional analysis would also work on manyseparate problems but in the spirit of a collective enterprise Some wouldconcentrate their attention on the study of rules norms habits conven-tions and values while others would study how these are associatedwith congurations of various institutional arrangements (types ofmarkets hierarchies networks associations communities clans statesetc) Others would work on specic institutional sectors (eg education

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

630

Table 4

Rules norms hab

its an

d values associated w

ith sp

ecic co

ngu

ration

s of coo

rdinating mecha

nism

s soc

ial sy

stem

s of

prod

uction

an

d nationa

l styles of inno

vation

USA

Germany

Norms rules values which provide incentives for

Norms rules values which provide incentives for

short-term horizons

long-term horizons

Con

gu

rative

forms of

Marke

ts c

orpo

rate hierarchies reg

ulatory state

Assoc

iation

s (busine

ss assoc

iation

s la

bor union

s)

coordinating

corp

oratist-type

networks

marke

ts c

orporatist-

mecha

nism

s (listed in

type

state

order of im

portanc

e in

the particu

lar social

system

)

Social sys

tem of

prod

uction

Bus

iness system

Firm

struc

tures

Con

glom

erates b

ut with mov

emen

t toward

Firm

more co

ngru

ent with pa

rticular prod

ucts

elim

ination of produc

tion

in som

e sectors

emph

asis on long

-term strateg

ies rm

s are well

emph

asis on sh

ort-term

strateg

ies high

integrated

differentiation

between divisions

of rms

Firm

s mob

ilized in

Low

High

and in

tegrated

into busine

ss

associations

Indu

strial relations

Low

deg

ree of jo

b secu

rity rigid in

ternal lab

or

High deg

ree of jo

b secu

rity

exible interna

l lab

or

system

marke

t ex

ible externa

l labo

r marke

tsmarke

t rigid externa

l lab

or m

arke

tPatterns of

Dispersed owne

rship frequ

ent turnov

erMore stab

le owne

rship patterns

owne

rship

Table 4

continue

d

USA

Germany

Norms rules values which provide incentives for

Norms rules values which provide incentives for

short-term horizons

long-term horizons

Social system of

Prod

uction

Bus

iness system

Training for labo

r an

dHigh deg

ree of voluntary decision mak

ing by

Collective an

d m

ore co

mpulso

ry d

ecisions

abo

ut

man

agem

ent

individuals as to ho

w m

uch

tim

e an

d ene

rgy

how m

uch ed

ucation

is ne

eded

he

avy em

pha

sis

to inv

est in educ

ation heavy

emph

asis on

on techn

ical training (eg en

gine

ering)

busine

ss sch

ools training

in marke

ting

and

sales accou

nting

Fina

ncial marke

tsEqu

ity an

d ven

ture cap

ital m

arke

ts highly

Equ

ity an

d ven

ture cap

ital m

arke

ts w

eakly

dev

elop

ed na

ncing from

cen

tral gov

ernm

ent

deve

lope

d he

avy em

pha

sis on

corporate loan

sto priva

te rm

s for dev

elop

men

t of new

tech

nologies related

to na

tion

al security

Unive

rsity research

High individual auton

omy high poten

tial for

Low

ind

ividua

l au

tono

my until almost 40

yea

rs of

system

individual creativity u

nive

rsity pr

ofesso

rs

age poten

tial for ind

ividual creativity high

ly

high

ly entreprene

urial

cons

traine

d unive

rsity pr

ofessors not so

entrep

rene

urial

Inno

vative

styles

Highly su

ccessful in m

ajor break

through

s in

High pe

rforman

ce in proc

ess an

d inc

remen

tal

basic an

d app

lied scien

ce go

od in dev

elop

ing

inno

vation

s in existing produ

cts not so su

ccessful

totally

new

lines of pr

oducts an

d in m

ajor break

through

s in basic scien

ce or in

tech

nologies high

deg

ree of entrepr

eneu

rship

deve

loping

totally new

lines of prod

ucts of

amon

g research

ers

tech

nologies low d

egree of entrepr

eneu

rship

amon

g research

ers

Table 4

continue

d

USA

Germany

Norms rules values which provide incentives for

Norms rules values which provide incentives for

short-term horizons

long-term horizons

Prod

ucts in w

hich

Pr

oducts with sh

ort liv

es prod

ucts in w

hich

Older produ

cts with long

history machine

ry pap

er

society excels

tech

nology

is bo

th highly co

mplex

and

pr

inting

textile

s m

otor veh

icles m

achine

too

lsch

ange

s rapidly en

tertainm

ent indus

try

pha

rmaceu

ticals biotechn

olog

y softw

are

aerosp

ace

business systems nancial markets and systems of research) Distinctivewould be the recognition of how these separate studies are linkedtogether how each institutional sector is linked to a societyrsquos normsrules values etc and its conguration of institutional arrangementsSimilarly other institutional analysts might focus on the study of orga-nizations with a major concern for how the institutional environmentof organizations (eg norms rules values congurations of institutionalarrangements institutional sectors) inuenced the structure culture andoutcomes of organizations And nally others might study how all ofthe aforementioned aspects of institutional analysis inuenced a societyrsquosoverall performance Nelson and Sampat (1998) have appropriatelyreminded us that we should engage in institutional analysis in order toaddress specic problems but the view developed in this article is thatwe rst need some mapping of the terrain of institutional analysis inorder for our diverse enterprises to become a meaningful collective enter-prise especially if we are to understand how institutions inuence theinnovative processThe argument developed here is not that all social scientists should engage

in institutional analysis Obviously there are numerous other importantareas of research which lie beyond the eld of institutional analysis Nordoes this article argue that institutional analysis is the most importantsubject for social scientistsSince the structure and culture of our research universities reect an

enormous amount of disciplinary fragmentation what are the prospectsfor promoting the kind of collective and interdisciplinary research agendaproposed above In one respect the prospects are not encouragingDuring the past half century the efforts of American universities to have effective communication across the biological sciences have variedgreatly in their success At the University of California-Berkeley it was not until the 1980s that there was a serious effort to integrate thebiological sciences and at that time the campus abolished seventeendepartments to promote an integrated biological program But at theUniversity of Wisconsin the biological sciences are still very fragmentedinto dozens of different departments in botany plant pathology zoologyand physiology with two departments of biochemistry and two depart-ments of geneticsThere have been other approaches which have worked very well histor-

ically for promoting scientic integration amidst scientic diversity Forexample several summer schools and special institutes were organizedin the United States during the 1940s in an effort to overcome the frag-mentation in the biological sciences These programs made it possiblefor scientists in diverse academic disciplines and from different univer-sities and countries to come together to engage in intense and frequentinteraction Perhaps the most effective program was that held in the

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

634

summers at Cold Spring Harbor New York where a research agendafor the study of molecular biology was developed (Fischer and Lipson1988 Stent and Watson 1966) Scientists went there summer aftersummer and when they returned to their universities many began therestructuring of the biological sciences in their universitiesIn our own day we need summer institutes and other similar programs

so that scholars from diverse backgrounds can come together anddevelop a collective agenda for doing institutional analysis and forstudying the impact of institutions on the performance of societies ndashinnovativeness being one such subject If we can develop such instituteswe will increase the prospects for transforming our universities Thereis reason to be optimistic that once an eclectic group of scholars workwith a common set of concepts we can have genuine theoretical advanceif researchers can bring their diverse expertise to bear on a common setof problems There is considerable evidence that institutional analysisoffers high potential for advancing our knowledge of innovations andtechnological change (Hollingsworth and Hollingsworth 2000) If we aregoing to make substantial theoretical advance it is not likely to occurfrom the vantage point of a multiplicity of disciplines Rather it is likelyto occur as we develop a new intellectual framework and our own acad-emic journals This is a tall agenda which requires us to reassess theway that knowledge is presently produced in our societies

NOTE

This article is part of a much larger agenda involving the study of institutionsorganizations and innovations in which I am involved My debts are many andvaried The rst version was written while I was in residence in the ResearchUnit for Institutional Change and European Integration of the Austrian Academyof Sciences in Vienna in the summer of 1998 I am especially grateful to ProfessorsEgon Matzner and Sonja Puntscher-Riekmann for providing the stimulating envi-ronment in which to think through the ideas developed here Egon Matzner hasengaged in very stimulating discussions with me about the importance of insti-tutions in the general eld of socio-economics The nal version of the articlewas completed during my residence as a fellow at the Neurosciences Institutein La Jolla California I also wish to thank my colleagues in the InnovationTheme Project at the Netherlands Institute for Advanced Study (NIAS) during1998ndash9 and Edgar Grande and Karl Muumlller for stimulating conversations aboutthe interaction of institutions and organizations in facilitating particular typesof innovation I am especially grateful to Jerald Hage for helping me to developa number of issues in the article and for his useful comments on the rst draftFrans van Waarden early on helped me to work through the idea of multiplelevels of institutional analysis Others who have helped to develop the ideas inthe article have been my colleagues in the research project lsquoComparative socialsystems of productionrsquo especially Tom Burns Christel Lane Yoshitaka OkadaWolfgang Streeck and Richard Whitley My colleagues Robert Boyer SteveCasper and David Soskice have taught me much about how incentives providedby the institutional environments of rms inuence individual decision making

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

635

But my most important debt is to Ellen Jane Hollingsworth who has repeatedlyquestioned every single assumption of this article Without her rich and contin-uous tutoring this article would be even less developed than it is at presentFinally without the insightful assistance of David Gear in many phases of thisarticle it would not have been completed

REFERENCES

Aoki Masahiko (1990) lsquoTowards an economic model of the Japanese rmrsquo Journalof Economic Literature March 1ndash27

Arora A Landau Ralph and Rosenberg Nathan (eds) (1998) Chemicals and Long-Term Economic Growth Insights from the Chemical Industry New York JohnWiley amp Sons

Arrow Kenneth (1974) The Limits of Organization New York NortonArthur W B (1988a) lsquoSelf-reinforcing mechanisms in economicsrsquo in P W

Anderson K J Arrow and D Pines (eds) The Economy as an Evolving ComplexSystem Redwood City Calif Addison-Wesley

mdashmdash(1988b) lsquoCompeting technologies an overviewrsquo in G Dosi C Freeman RNelson G Silverberg and L Soete (eds) Technical Change and Economic TheoryLondon Pinter Publishers pp 590ndash607

Audretsch David B (1995) Innovation and Industry Evolution Cambridge MassMIT Press

Audretsch David B and Feldman Maryann P (1995) lsquoInnovative clusters andthe industry life cyclersquo Discussion Paper Wissenschaftszentrum BerlinMarket Processes and Corporate Development Research Unit FS-IV-95-7

Baum Joel A C and Oliver Christine (1992) lsquoInstitutional embeddedness andthe dynamics of organizational populationsrsquo American Sociological Review 57540ndash59

Baum Joel A C and Singh Jitendra V (1994a) lsquoOrganizationndashenvironment coevo-lutionrsquo in Joel A C Baum and Jitendra V Singh (eds) The EvolutionaryDynamics of Organizations New York Oxford University Press pp 379ndash402

mdashmdash(eds) (1994b) The Evolutionary Dynamics of Organizations New York OxfordUniversity Press

Berger Suzanne and Dore Ronald (eds) (1996) National Diversity and GlobalCapitalism Ithaca NY Cornell University Press

Boyer Robert (1997) lsquoThe variety and unequal performance of really existingmarkets farewell to Doctor Panglossrsquo in J Rogers Hollingsworth and RobertBoyer (eds) Contemporary Capitalism the Embeddedness of InstitutionsCambridge and New York Cambridge University Press pp 55ndash93

Burns Tom R and Dietz Thomas (2001 forthcoming) lsquoSocial rule system theorysocial action institutional arrangements and evolutionary processesrsquo in JRogers Hollingsworth and Karl Muumlller (eds) Advances in SocioeconomicsLanham Md Rowman amp Littleeld

Burns Tom R and Flam Helena (1987) The Shaping of Social Organization BeverlyHills Calif Sage

Calvert Randall L (1995) lsquoThe rational choice theory of social institutionsrsquo inJeffrey S Banks and Eric A Hanuskek (eds) Modern Political Economy NewYork Cambridge University Press pp 216ndash66

Camic Charles (1986) lsquoThe matter of habitrsquo American Journal of Sociology 911039ndash87

Campbell John L (1997) lsquoMechanisms of evolutionary change in economicsgovernance interaction interpretation and bricolagersquo in Lars Magnusson

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

636

want to allow for some reinforcement effects of the modelrsquos histor-ically contingent social and economic success

(Streeck 1997b 54)

In sum the emergence of social systems of production is a long-termevolutionary process with each part interacting with its environment andresulting in a congurative whole but with no previous design by eithera single actor or a collectivity of actors As suggested above actors do not recreate their world anew At any moment in history the totalinstitutional complex of a society necessarily contains the resources and legacies of its past Preexisting institutional complexes are nevercompletely wiped out Societies consist of multiple layers of history withtheir diverse logics of action However the cumulative effect of smallperipheral changes in altering particular institutional sectors and insti-tutional arrangements can be substantial (Murmann 1998 Lanzara 1998)lsquoInstitution building is affected more by the ways in which people havecodied the past than by how they envision the future There are manymore potential resources in pre-existing institutional arrangements thanit is usually assumed or suspectedrsquo (Lanzara 1998 30)

Fourth level of analysis

The next level of institutional analysis organizational structures is some-what more controversial As suggested above Douglass North (1990)draws a sharp distinction between institutions and organizations Morerecently many organizational theorists (Powell 1991 Powell andDiMaggio 1991 DiMaggio and Powell 1983 Baum and Oliver 1992Kondra and Hinings 1998 Townley 1997) argue that institutional rulesnorms and conventions unfold in tandem with organizational structuresand this is my positionThe literature which focuses on how institutions inuence organiza-

tions is quite different from two other theoretical literatures which alsoare concerned with how organizational environments shape the behaviorof organizations There are resource dependency theorists (Pfeffer andSalancik 1978) who emphasize the role of environmental resources inshaping organizations while Hannan and Freeman (1977 1984) and otherpopulation ecologists emphasize the survival of organizations givencertain kinds of organizational conditions (Orru et al 1991)The institutionalist perspective as implied above emphasizes the

normative environment in which organizations are embedded It is a perspective which focuses on the way in which organizations in their behavior tend to conform to the institutional rules and normswhich are dominant in the organizational environment However allthree perspectives ndash the institutionalist the population ecology and

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

619

the resource dependency ndash emphasize how the environment inuencesorganizations and how organizations which are subject to the sameenvironment tend to converge in their behavior to have what DiMaggioand Powell have labeled organizational isomorphism (DiMaggio andPowell 1983)There has emerged a vast literature which demonstrates that within

each society there are scal political judicial and other regulatory normswhich limit and shape the culture and structure of organizationalbehavior The normative institutional environment of organizations limits the options of what organizations do in a particular society and inuences the patterns of ownership relations with suppliers andcustomers In short it is the normative environment of organizationswhich denes within a particular society what is socially acceptablebehavior for organizations (Hamilton and Biggart 1988 Hollingsworthand Hollingsworth 2000 Meyer and Rowan 1991 Zucker 1987 1988Orru et al 1991 Townley 1997)Thus far most of the analyses on isomorphic pressures operating on

organizations have been cross-sectional in nature However the entireconcept of isomorphism implies that there are strong environmental pres-sures exerted on organizations and in order to observe this phenomenonit is important that we have longitudinal studies which assess howchanges in an organizationrsquos institutional environment inuence changesin the structure and culture of organizations Moreover the expectationhere is that when one analyzes the patterns of organizational changeover time the isomorphism will be holistic ndash that is within a particularsociety there will be a tendency for the internal structure of particulartypes of organizations to converge in large part because of the pres-sures to conform to the changes in the external norms rules and valuesA close reading of Aoki (1990) and Slack and Hinings (1994) nds thatchanges in the national institutional environment of organizations haveinuenced changes in the internal structure of organizations within aspecic industry of a particular societyHowever it is inappropriate to view organizations as changing only

in relation to exogenous environmental change There is a kind of co-evolution which occurs between organizations and their institutionalenvironments Co-evolution implies nonlinear feedback between orga-nizations and their institutional environments This is of course quitedifferent from the more common way of studying organizations bymodeling relations between independent and dependent variables Ofcourse the more common type of analysis is quite appropriate whenthere are simple relationships which do not involve complex feedbackprocesses But in the strategy for institutional analysis which is proposedherein it is less useful to separate independent from dependent vari-ables and more useful to understand the interacting and co-evolutionary

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

620

processes among all of the levels of institutional analysis discussed inthe various sections above (Baum and Singh 1994a 379ndash80 Mowery andNelson 1999 Murmann 1998)The logic of the perspective herein suggests that within every

society there is variation in the structure and culture of business rmsuniversities and other complex organizations (Kondra and Hinings1998) However this variation takes place within parameters which aresystem specic For example every German rm and every Germanuniversity is different from every other German university and rm but there are a set of cultural and structural characteristics which distin-guish German rms and universities from those in America Similarlyeach Japanese rm and university is unique yet they have a set ofcultural and structural characteristics which are system specic andwhich differentiate them from business rms and universities in other countriesIn all societies each organization has its own distinctive organizational

rules norms and conventions which are subordinate to the meta norms and rules of the larger society within which they are embeddedHowever the strength of the institutional environment within whichorganizations are embedded varies from society to society Some societies have multiple institutional environments and there is hetero-geneity in terms of what constitutes appropriate behavior oforganizations In such societies there can be conict over whatinstitutional logics should regulate specic organizational functions(Friedland and Alford 1991 Townley 1997 262ndash3) In those societies inwhich the institutional norms habits and rules are most developed and in which the institutional pressures to conform are greatest thereis less variation in the structure and culture of business rms and variouskinds of research organizations In such societies the connectednessbetween research organizations and their institutional environments is sufciently strong for organizations to have low autonomy to pursue independent strategies and goals in these societies there is a great deal of organizational isomorphism Conversely the weaker the institutional environment in which research organizations areembedded the greater the variation in the structure and culture of business rms and research organizations Moreover where the institu-tional environments are more weakly developed organizations havegreater autonomy and exibility to respond to the development of new knowledge and to be highly innovative Hence it is in those societies where the institutional environments are most developed and most rigid and when there is less organizational autonomy and exibility that fewer radical innovations in basic and applied science as well as in totally new products and industrial sectors havebeen made

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

621

Fifth level of analysis

We come next to the outputs and performance of the various institutionalcomponents of a society It is at this level that institutional componentsare more pragmatic and exible For example in the legal sector thereare specic statutes and court rulings in the state sector there are specicpolicies in the business sector there are new products new technolo-gies and market strategies It is at this nal level of analysis thatinstitutional spheres are most open and susceptible to change and cross-national mimicry is easiest and most commonThrough the outputs of the society we can obtain some assessment as

to how well a society is performing Moreover we can assess how inno-vative it is how egalitarian it is in the distribution of its resources howegalitarian it is in terms of levels of health education etc Just as it iscomplicated to measure how egalitarian a society is similarly it is dif-cult to assess how innovative it is For example societies vary in makingincremental and radical breakthroughs in basic and applied science in developing totally new products and new kinds of organizations inincremental and process innovations in existing products and organiza-tions in developing new and different forms of marketing ndash bothdomestically and globally (Hage and Hollingsworth 2000) Of coursesocieties vary in their rates of economic growth in their rates of economicproductivity in their quality of life (eg rates of crime life expectancyetc) All of these performance criteria feed back and inuence each ofthe levels of institutional analysis discussed above ndash rules norms valuesetc institutional arrangements institutional sectors the structure andform of organizations Moreover performance measures may inuenceeach other For example if a society has very low rates of growth itmay not be very innovative in some of the types of innovative activitiesmentioned above Good or poor performances of certain kinds inuenceother performance indicators (Hollingsworth et al 1990)Different institutional arrangements and different social systems of

production result in different types of economic performance Hence solong as societies have different social systems of production there areserious constraints on the degree to which they can converge in theirinnovative styles Different social systems of production tend to maxi-mize in a more or less explicit manner different performance criteriausually mixing considerations about static and dynamic efciency protsecurity social peace and economic andor political power In short incontrast to the implications of neoclassical economic theory in real-worldeconomies there are no universal standards all economically rationalactors attempt to maximize Economic history provides numerous exam-ples of how a great variety of principles of rationality are implementedin different societies

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

622

A critical question in institutional analysis is whether an existing socialsystem of production which supports a set of routines for a particularkind of technology and industry can shift from old practices to newones There are numerous historical accounts of how the social systemof production in a particular society which worked so well for a numberof years could not adapt to new technologies (Lazonick 1990 Veblen1915 Nelson 1994) Schumpeter (1983) and Freeman (1991) have devel-oped the idea of a techno-economic paradigm which suggests thatdifferent technologies require specic forms of organizational arrange-ments ndash or what I called social systems of production Because oflsquoinstitutional inertiarsquo the social system of production of a society maynot be able to adapt to a new techno-economic paradigm (Nelson 199458 Lanzara 1998)Whether a social system of production can sustain its performance

standards depends not only on its intrinsic economic rationality but alsoon where it ts into a larger system If a particular social system ofproduction is immune from the innovativeness and competition of analternative system survival can be long lasting But if different socialsystems of production with diverging criteria of good economic perfor-mance meet in the world arena the arbitrariness of nationally imposedconstructed performance standards may be superseded by alternativeperformance criteria as a result of international competitivenessThe world economy is also socially constructed just as are national

economies Even if different social systems of production are competingin the international arena it is not always possible to determine whichis more competitively effective at any moment in time Hegemonicnation-states in the short run shape the rules of trade to favor their indus-trial sectors and rms But the history of hegemonic powers suggeststhat in the longer run such social systems of production sustainedlargely by military and political power eventually give way to moreinnovative and competitive social systems of production In our ownday as nation-states are increasingly integrated into a world economyeven hegemonic powers lose their innovative and competitive advan-tage and their share of world output decreases Such a country mayattempt to restructure its institutional arrangements and to readjust itsperformance preferences But to restructure its system of innovationgenerally calls for a major redistribution of power within a societyLargely for this reason societies have historically had limited capacityto reconstruct a system of innovation in the image of their major competi-tors Each countryrsquos social system of production is a conguration of ahost of norms rules and values as well as of institutional arrangementsEach system is constantly changing and is open to inuence from othersystems And indeed many technologies and practices diffuse from onesociety to another but a societyrsquos capacity to be innovative is constrained

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

623

by the existing social system of production Thus the same technologymay exist in numerous countries but how it is employed and how itinuences societal outcomes and performances varies from one institu-tional conguration or society to another (Hollingsworth 1998)

PERSPECTIVES ON INSTITUTIONAL CHANGE

A fundamental problem which remains unresolved in institutionalanalysis involves the nature of institutional change There is much confu-sion and miscommunication in the social sciences about institutionalchange in large part because we do not even know how to measure therate of institutional change Of course we know that there are bothexternal and internal focuses for change and we have had a good bitof scholarship on this subject What is not clearly understood is thatinstitutions and institutional arrangements within societies are histori-cally rooted that there is a great deal of path-dependency to the waythat institutions evolve Moreover the more intricately linked each sectoris with each other sector and with a societyrsquos rules and norms the lesschoice actors have to devise new institutions and institutional arrange-ments On the other hand if a sector or actor is too isolated it may betoo weak within a society to bring about any kind of effective changeThese considerations are very critical to the problem of how muchfreedom actors have to create new institutions afresh to what extentinstitutions diffuse from one society to another (Murmann 1998 Lanzara1998) This is a fundamental question for students of institutionalanalysis as we attempt to understand how the different institutionallevels are linked togetherThere are a variety of reasons why there is confusion in studying insti-

tutional change There is disagreement over the extent to which actorshave the freedom to build new institutional arrangements and if theydo to what extent the new arrangements and organizational patternsmay depart from past practices Among some institutional economists(Williamson 1985) there is a basic assumption that actors build efcientinstitutions by pursuing their self-interest and by promoting theirstrategic goals Moreover there is the assumption that a mixing of fullrationality with market competition tends to produce all of the optimalinstitutions that are needed in order to coordinate a complex capitalisteconomy (Hollingsworth and Boyer 1997 ch 14 Hodgson 1999) Incontrast the argument developed here is much more complexSocial institutions are historically rooted and there is a great deal of

path-dependency in the way that various institutional componentsevolve The shape of institutional congurations at any moment limitsthe type of options for change Because there is a great deal of institu-tional inertia radical change in the institutional components of a society

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

624

is uncommon (Lanzara 1998) Recognizing the constraints existinginstitutions exert on actors we note that existing congurations of insti-tutional arrangements institutional sectors and organizational propertiescan limit but also inuence the degree to which norms and rules canchange Many scholars (North 1990 Ostrom 1990) have emphasized therole of norms rules ideologies and values in limiting the ability of actorsto develop new institutional arrangements which radically depart fromexisting ones However there is a two-way process of interaction aprocess of co-evolution (Baum and Singh 1994a Nelson 1994) Thus asFriedland and Alford (1991 244) argue actors develop new institutionalarrangements by recognizing organizational failures and low perfor-mance but these new arrangements do not necessarily change a societyrsquosnorms rules habits and its underlying value systemCampbell (1997) advances our understanding of how existing institu-

tions enable actors to construct radically new institutions He recognizesthat most institutions and institutional arrangements embody a degreeof inertia from which it is difcult for actors to depart Moreover thedifferential power relations among actors make it unlikely that lesspowerful actors can change the existing power structure of a societyMost change in the institutional components of a society evolves througha process of constrained selection which reects to a considerable degreethe existing arrangements and power relations This becomes the inter-pretative frame for social actors with institutionalized scripts and ritualswhich tend to be taken for granted and appear to be quite ordinary andnatural This institutionalized scripting of the social world is an extremelyimportant source of social stability and inertiaCampbell has developed a very fruitful explanation of change in terms

of the various institutional components outlined here He distinguishesbetween incremental and radical institutional change and argues thatradical institutional change occurs when social actors with widelydiffering norms cultural scripts and rituals for action engage in intenseinteraction with each other (see also Knight 1992) The more diverse theinteractions the greater the potential for institutional change He arguesthat as shifts in the composition of interaction among social actors occurchanges in the interpretation of problems and interests will follow Ifthere are only minor changes in the extent to which diverse actorsinteract there will be only modest institutional change But the morefundamental the changes in the interaction of diverse social actors thegreater the change in the way that actors interpret their world And thegreater the change in the way that actors interpret their world the morelikely that radical changes in various institutional components willfollow In other words diverse decision making tends to create a widerrange of interpretative frameworks of the social world than is likely tobe the case if social actors operate in isolation

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

625

This perspective of diverse interactions among social actors facilitatesour ability to move beyond the traditional view that existing institutionsconstrain the range of institutional alternatives which actors face Theolder view has tended to suggest that actors engage in institutionalchange by extending existing institutional principles habits and conven-tions Specifying the conditions which facilitate fundamental institutionalchange permits a much richer perspective for institutional analysisCampbellrsquos perspective about the conditions under which fundamentalor radical change in institutional components occurs is consistent withmy own research about the conditions under which major discoveriesor fundamental new knowledge occur the more scholars with diversebackgrounds interact with intensity and frequency the greater the like-lihood that they will develop new and alternative ways of thinking(Hollingsworth and Hollingsworth 2000)These views of institutional change are also consistent with the recent

work of Nelson and his colleagues (Mowery and Nelson 1999 Murmann1998) They report the research on a variety of industrial sectors in anumber of different countries revealing that networks of actors in partic-ular industries business and professional associations universityresearch departments and institutes are frequently able to use their collec-tive power to modify a societyrsquos property rights research system andgovernmental policies It is the intense and frequent interaction of thesediverse complex sets of actors ndash sometimes constituting a technologysystem ndash that often leads to incremental changes in a societyrsquos total socialsystem of production It is this kind of collective action that brings aboutorganizational change change in institutional sectors institutionalarrangements and even in the rules and norms of a society And if weare to understand how the style of innovativeness of a society changeswe must rst understand the changes in its institutional makeup It ishighly unlikely that a society will develop a new style of innovativenesswithout changing its institutional structure

FIRMS THEIR INSTITUTIONAL ENVIRONMENT ANDNATIONAL SYSTEMS OF INNOVATION

Variation in the innovative style of societies is shaped by their insti-tutional makeup ndash the various levels which are discussed aboveinstitutions institutional arrangements the structure and cohesion of institutional sectors which constitute a societyrsquos social system of pro-duction and the structure and culture of its organizations ndash especiallyits business rms and other research organizations By undertaking aninstitutional analysis of a society one can begin to understand in whatkinds of organizations the production of specic kinds of knowledgetakes place and how this is linked to particular kinds of innovativeness

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

626

With such a perspective one can gain a rich understanding of why somesocieties excel in the production of radical breakthroughs in basic andapplied science and in developing radically new products and why othersocieties excel in more incremental innovationsReecting on these issues one cannot help being impressed with the

contrasts between the innovative styles of German and American orga-nizations For example Germany has continued to be very successful in making incremental innovations in many industries in which it wasvery competitive well before World War II paper printing materialsmachinery electrotechnical products motor vehicles chemicals textileyarns and fabrics But the Germans have been much less successful indeveloping totally new products or related innovations in biotechnologyelectronics telecommunications aircraft as well as other newer indus-trial areas In contrast the Americans have been very innovative in these and other industries with products having very short lives andwith technologies which change very rapidly and are very complex(Hollingsworth 1991 Soskice 1997 1999)The importance of institutions and innovativeness for international

competitiveness has led to the recognition that there are national systemsof innovation (Lundvall 1992 Nelson 1993 Edquist 1997 Hage andHollingsworth 2000) But even before the recent studies of nationalsystems of innovation Landes (1969) was a proponent of these viewswhen he made the argument that Germany had developed a nationalsystem of innovation by building on its education and scientic researchsystems which were lacking in Britain and elsewhere However the concept of a national system of innovation immediately poses theproblem of what kind of innovations should be consideredDavid Soskice an economist at the Wissenschaftszentrum Berlin has

addressed a set of problems somewhat complementary to the perspec-tive raised here His work (Soskice 1990 1997 1999) suggests that weare sensitive to the institutional environment in which the followingactors are embedded the employees of rms the owners and nanciersof rms the competitors of rms and nally their collaborators ndash aperspective complementary to Hollingsworthrsquos work on social systemsof production (Hollingsworth 1997 1998) When one reects on the insti-tutional contextsocial system of production within which these actorsare embedded one becomes aware that those countries which have some-what inexible labor markets external to rms but a business system in which rms have long-term commitments to their employees arestrongly associated with incremental but hardly at all with radical inno-vations In the rms of such countries (eg Germany and Japan) thereis much more consensus decision making than in societies where a lsquohireand rersquo set of practices is quite pervasive And this kind of consensusdecision making ndash while highly conducive to incremental innovativeness

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

627

ndash limits the capacity for radical innovations to occur This kind of lsquolongtermismrsquo set of norms and rules occurs not only in the employmentpractices of incrementally innovative rms but also in their nancingand ownership patterns (Hollingsworth 1997) During the past thirtyyears these were characteristics of many rms in Germany and JapanSignicantly in both of these countries venture capital markets werenot highly developed nancing tended to be based more on long-termbank loans than on the equity markets and the relationships of rmstended to be relatively stable And these stable institutional arrange-ments also contributed to a high degree of innovativeness ndash butincremental rather than radical innovationsMeantime we should also pursue Soskicersquos ideas (1999) about how

rmsrsquo relationships with their competitors and collaborators shape theirstyles of innovativeness In those advanced industrial societies in whichbusiness rms are highly mobilized into business associations theretends to be a high degree of standardized quality controls over technicalnorms and a high degree of reliance on legal sanctions as monitoringinstruments vis-agrave-vis association members As a result of this high degreeof integration of rms into associations member rms are more likelyto be associated with incremental innovations than is likely to be the case with rms in those societies where associations are less welldeveloped In the latter type of society radical innovations are morelikely to occur In those advanced industrial societies where rms havelong-term stable relationships with collaborators in product design andproduction the style of innovation is likely to be incremental And it is in those societies in which there is a weaker tradition of strongcollaboration among rms that radical innovations are more likely tooccurObviously even in a country where there may be a great deal of

radical innovativeness most innovations will be of an incremental typefor there are simply fewer radical innovations But Germany and therest of Europe not only have had relatively few radical innovations incomparison with the United States but they also have lagged behindthe USA in developing relatively totally new industries Thus there hasbeen much less innovative activity in Europe in high-technology andnewly emerging industries than in the USAIn complementary research Audretsch and Feldman (1995) have

demonstrated that the USA has been more successful in spawning rmsengaged in the early stages of the industry life cycle but relative toGermany had an environment unfavorable to innovative activity byestablished rms especially in more mature industries German rmshave tended to be associated with a routinized technological regime ndashwhere established enterprises have had strong incentives to engage inincremental innovations but new rms have tended to have fundamental

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

628

innovative disadvantages relative to the Americans (see Audretsch 1995Winter 1982 Casper 2000)On both sides of the Atlantic universities engage in basic and applied

research But signicantly the institutional environment in which univer-sities are embedded inuences the degree to which the research is likelyto represent major breakthroughs or is likely to have more modestimplications Recent research in biomedical science demonstrates that inthose societies in which university professors andor graduate studentshave the following characteristics there are likely to be relatively majorresearch breakthroughs graduate students have a long training periodhigh dependency on a major professor low autonomy in dening thenature of a research project restricted job options and a university envi-ronment where there is a very modest or nonexistent entrepreneurialculture For example in Germany ndash a country with relatively few radicalbreakthroughs ndash the scholar who nishes a habilitation will have spentmany years on a research project before qualifying for a permanent acad-emic position will have been in a high-dependency position with a seniorprofessor will have had limited autonomy to dene a research projectand will be constrained from pursuing a high-risk research project Theyoung German will have an extraordinarily high level of training butwill have faced a number of disincentives to undertake high-risk researchAnd as soon as heshe has completed the habilitation and becomes aprofessor administrative responsibilities constitute another disincentivefrom engaging in high-risk or long-term research projects Moreover theGerman professor is a civil servant with a more stable and secure incomethan hisher American counterpart In a very different institutionalenvironment American professors in large research universities tend tobecome much more entrepreneurial Today many American professorsbehave as though they were operating their own rms within univer-sities and thus devote considerable time to generating funds for researchprojects assistants secretaries supplies etc And the variation in thisentrepreneurial spirit across countries is associated with variation in thetaking of high-risk research strategies and the making of major researchbreakthroughsIn sum societies vary in terms of the institutional environment in

which their organizations are embedded and this variation inuencesthe degree to which they have industrial activities in the early or latestages of the industry life cycle successful new or mature industriesand radical vs incremental innovations And it is this variation whichis encouraging rms in countries with relatively few new industries and with few or no radical innovations to invest in those countries with many new industries and numerous radical innovations ndash as astrategy to be able to transfer new technologies from the latter institu-tional environment to the former

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

629

Table 4 summarizes the way that the norms rules and values of a society are associated with Americarsquos and Germanyrsquos institutionalarrangements their social system of production and their styles of inno-vativeness

CONCLUSION

Several themes deserve to be reemphasized Most importantly institu-tional analysis is emerging as a dominant issue in several social sciencedisciplines The study of various institutional components discussedabove is a complex subject requiring multiple levels of analysis and aknowledge of multiple sectors within societies It requires a theoreticalperspective that evolves from an empirical analysis in different societiesIf we are to advance the study of institutional analysis we need cross-disciplinary collaboration and team research which are not readilyachieved given the structure of most of our research universitiesAt present scholars in each academic discipline have their own distinc-

tive strategies for studying institutional analysis Thus economists tendto use several approaches (North 1990 Williamson 1985 Hodgson 19881998 1999) anthropologists another (Geertz 1995) and sociologistspolitical scientists and historians employ other strategies (Campbell etal 1991 Powell and DiMaggio 1991 Katznelson 1998 Hall and Taylor1996) With practitioners in various disciplines pursuing their distinctiveapproaches to institutional analysis there is low potential for collabo-rative institutional analysis across the social sciences The lessons fromthe biological sciences over the last 40 years are very instructive oncepractitioners from various biological disciplines began to cooperate instudying biological phenomena by using the same concepts at the mole-cular level theoretical advance was very rapid Comparable phenomenatook place in the development of other hybrid elds eg biochemistrybiophysics (Judson 1979 Olby 1979 Kohler 1982 1994) Similarly weresocial scientists in various disciplines to work collaboratively in the areaof institutional analysis there would very likely be an acceleration intheir theoretical knowledgeThis is not a plea that everyone engaged in institutional analysis should

do the same kind of research Indeed just as biologists working at themolecular level study many different kinds of problems similarly socialscientists engaged in institutional analysis would also work on manyseparate problems but in the spirit of a collective enterprise Some wouldconcentrate their attention on the study of rules norms habits conven-tions and values while others would study how these are associatedwith congurations of various institutional arrangements (types ofmarkets hierarchies networks associations communities clans statesetc) Others would work on specic institutional sectors (eg education

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

630

Table 4

Rules norms hab

its an

d values associated w

ith sp

ecic co

ngu

ration

s of coo

rdinating mecha

nism

s soc

ial sy

stem

s of

prod

uction

an

d nationa

l styles of inno

vation

USA

Germany

Norms rules values which provide incentives for

Norms rules values which provide incentives for

short-term horizons

long-term horizons

Con

gu

rative

forms of

Marke

ts c

orpo

rate hierarchies reg

ulatory state

Assoc

iation

s (busine

ss assoc

iation

s la

bor union

s)

coordinating

corp

oratist-type

networks

marke

ts c

orporatist-

mecha

nism

s (listed in

type

state

order of im

portanc

e in

the particu

lar social

system

)

Social sys

tem of

prod

uction

Bus

iness system

Firm

struc

tures

Con

glom

erates b

ut with mov

emen

t toward

Firm

more co

ngru

ent with pa

rticular prod

ucts

elim

ination of produc

tion

in som

e sectors

emph

asis on long

-term strateg

ies rm

s are well

emph

asis on sh

ort-term

strateg

ies high

integrated

differentiation

between divisions

of rms

Firm

s mob

ilized in

Low

High

and in

tegrated

into busine

ss

associations

Indu

strial relations

Low

deg

ree of jo

b secu

rity rigid in

ternal lab

or

High deg

ree of jo

b secu

rity

exible interna

l lab

or

system

marke

t ex

ible externa

l labo

r marke

tsmarke

t rigid externa

l lab

or m

arke

tPatterns of

Dispersed owne

rship frequ

ent turnov

erMore stab

le owne

rship patterns

owne

rship

Table 4

continue

d

USA

Germany

Norms rules values which provide incentives for

Norms rules values which provide incentives for

short-term horizons

long-term horizons

Social system of

Prod

uction

Bus

iness system

Training for labo

r an

dHigh deg

ree of voluntary decision mak

ing by

Collective an

d m

ore co

mpulso

ry d

ecisions

abo

ut

man

agem

ent

individuals as to ho

w m

uch

tim

e an

d ene

rgy

how m

uch ed

ucation

is ne

eded

he

avy em

pha

sis

to inv

est in educ

ation heavy

emph

asis on

on techn

ical training (eg en

gine

ering)

busine

ss sch

ools training

in marke

ting

and

sales accou

nting

Fina

ncial marke

tsEqu

ity an

d ven

ture cap

ital m

arke

ts highly

Equ

ity an

d ven

ture cap

ital m

arke

ts w

eakly

dev

elop

ed na

ncing from

cen

tral gov

ernm

ent

deve

lope

d he

avy em

pha

sis on

corporate loan

sto priva

te rm

s for dev

elop

men

t of new

tech

nologies related

to na

tion

al security

Unive

rsity research

High individual auton

omy high poten

tial for

Low

ind

ividua

l au

tono

my until almost 40

yea

rs of

system

individual creativity u

nive

rsity pr

ofesso

rs

age poten

tial for ind

ividual creativity high

ly

high

ly entreprene

urial

cons

traine

d unive

rsity pr

ofessors not so

entrep

rene

urial

Inno

vative

styles

Highly su

ccessful in m

ajor break

through

s in

High pe

rforman

ce in proc

ess an

d inc

remen

tal

basic an

d app

lied scien

ce go

od in dev

elop

ing

inno

vation

s in existing produ

cts not so su

ccessful

totally

new

lines of pr

oducts an

d in m

ajor break

through

s in basic scien

ce or in

tech

nologies high

deg

ree of entrepr

eneu

rship

deve

loping

totally new

lines of prod

ucts of

amon

g research

ers

tech

nologies low d

egree of entrepr

eneu

rship

amon

g research

ers

Table 4

continue

d

USA

Germany

Norms rules values which provide incentives for

Norms rules values which provide incentives for

short-term horizons

long-term horizons

Prod

ucts in w

hich

Pr

oducts with sh

ort liv

es prod

ucts in w

hich

Older produ

cts with long

history machine

ry pap

er

society excels

tech

nology

is bo

th highly co

mplex

and

pr

inting

textile

s m

otor veh

icles m

achine

too

lsch

ange

s rapidly en

tertainm

ent indus

try

pha

rmaceu

ticals biotechn

olog

y softw

are

aerosp

ace

business systems nancial markets and systems of research) Distinctivewould be the recognition of how these separate studies are linkedtogether how each institutional sector is linked to a societyrsquos normsrules values etc and its conguration of institutional arrangementsSimilarly other institutional analysts might focus on the study of orga-nizations with a major concern for how the institutional environmentof organizations (eg norms rules values congurations of institutionalarrangements institutional sectors) inuenced the structure culture andoutcomes of organizations And nally others might study how all ofthe aforementioned aspects of institutional analysis inuenced a societyrsquosoverall performance Nelson and Sampat (1998) have appropriatelyreminded us that we should engage in institutional analysis in order toaddress specic problems but the view developed in this article is thatwe rst need some mapping of the terrain of institutional analysis inorder for our diverse enterprises to become a meaningful collective enter-prise especially if we are to understand how institutions inuence theinnovative processThe argument developed here is not that all social scientists should engage

in institutional analysis Obviously there are numerous other importantareas of research which lie beyond the eld of institutional analysis Nordoes this article argue that institutional analysis is the most importantsubject for social scientistsSince the structure and culture of our research universities reect an

enormous amount of disciplinary fragmentation what are the prospectsfor promoting the kind of collective and interdisciplinary research agendaproposed above In one respect the prospects are not encouragingDuring the past half century the efforts of American universities to have effective communication across the biological sciences have variedgreatly in their success At the University of California-Berkeley it was not until the 1980s that there was a serious effort to integrate thebiological sciences and at that time the campus abolished seventeendepartments to promote an integrated biological program But at theUniversity of Wisconsin the biological sciences are still very fragmentedinto dozens of different departments in botany plant pathology zoologyand physiology with two departments of biochemistry and two depart-ments of geneticsThere have been other approaches which have worked very well histor-

ically for promoting scientic integration amidst scientic diversity Forexample several summer schools and special institutes were organizedin the United States during the 1940s in an effort to overcome the frag-mentation in the biological sciences These programs made it possiblefor scientists in diverse academic disciplines and from different univer-sities and countries to come together to engage in intense and frequentinteraction Perhaps the most effective program was that held in the

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

634

summers at Cold Spring Harbor New York where a research agendafor the study of molecular biology was developed (Fischer and Lipson1988 Stent and Watson 1966) Scientists went there summer aftersummer and when they returned to their universities many began therestructuring of the biological sciences in their universitiesIn our own day we need summer institutes and other similar programs

so that scholars from diverse backgrounds can come together anddevelop a collective agenda for doing institutional analysis and forstudying the impact of institutions on the performance of societies ndashinnovativeness being one such subject If we can develop such instituteswe will increase the prospects for transforming our universities Thereis reason to be optimistic that once an eclectic group of scholars workwith a common set of concepts we can have genuine theoretical advanceif researchers can bring their diverse expertise to bear on a common setof problems There is considerable evidence that institutional analysisoffers high potential for advancing our knowledge of innovations andtechnological change (Hollingsworth and Hollingsworth 2000) If we aregoing to make substantial theoretical advance it is not likely to occurfrom the vantage point of a multiplicity of disciplines Rather it is likelyto occur as we develop a new intellectual framework and our own acad-emic journals This is a tall agenda which requires us to reassess theway that knowledge is presently produced in our societies

NOTE

This article is part of a much larger agenda involving the study of institutionsorganizations and innovations in which I am involved My debts are many andvaried The rst version was written while I was in residence in the ResearchUnit for Institutional Change and European Integration of the Austrian Academyof Sciences in Vienna in the summer of 1998 I am especially grateful to ProfessorsEgon Matzner and Sonja Puntscher-Riekmann for providing the stimulating envi-ronment in which to think through the ideas developed here Egon Matzner hasengaged in very stimulating discussions with me about the importance of insti-tutions in the general eld of socio-economics The nal version of the articlewas completed during my residence as a fellow at the Neurosciences Institutein La Jolla California I also wish to thank my colleagues in the InnovationTheme Project at the Netherlands Institute for Advanced Study (NIAS) during1998ndash9 and Edgar Grande and Karl Muumlller for stimulating conversations aboutthe interaction of institutions and organizations in facilitating particular typesof innovation I am especially grateful to Jerald Hage for helping me to developa number of issues in the article and for his useful comments on the rst draftFrans van Waarden early on helped me to work through the idea of multiplelevels of institutional analysis Others who have helped to develop the ideas inthe article have been my colleagues in the research project lsquoComparative socialsystems of productionrsquo especially Tom Burns Christel Lane Yoshitaka OkadaWolfgang Streeck and Richard Whitley My colleagues Robert Boyer SteveCasper and David Soskice have taught me much about how incentives providedby the institutional environments of rms inuence individual decision making

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

635

But my most important debt is to Ellen Jane Hollingsworth who has repeatedlyquestioned every single assumption of this article Without her rich and contin-uous tutoring this article would be even less developed than it is at presentFinally without the insightful assistance of David Gear in many phases of thisarticle it would not have been completed

REFERENCES

Aoki Masahiko (1990) lsquoTowards an economic model of the Japanese rmrsquo Journalof Economic Literature March 1ndash27

Arora A Landau Ralph and Rosenberg Nathan (eds) (1998) Chemicals and Long-Term Economic Growth Insights from the Chemical Industry New York JohnWiley amp Sons

Arrow Kenneth (1974) The Limits of Organization New York NortonArthur W B (1988a) lsquoSelf-reinforcing mechanisms in economicsrsquo in P W

Anderson K J Arrow and D Pines (eds) The Economy as an Evolving ComplexSystem Redwood City Calif Addison-Wesley

mdashmdash(1988b) lsquoCompeting technologies an overviewrsquo in G Dosi C Freeman RNelson G Silverberg and L Soete (eds) Technical Change and Economic TheoryLondon Pinter Publishers pp 590ndash607

Audretsch David B (1995) Innovation and Industry Evolution Cambridge MassMIT Press

Audretsch David B and Feldman Maryann P (1995) lsquoInnovative clusters andthe industry life cyclersquo Discussion Paper Wissenschaftszentrum BerlinMarket Processes and Corporate Development Research Unit FS-IV-95-7

Baum Joel A C and Oliver Christine (1992) lsquoInstitutional embeddedness andthe dynamics of organizational populationsrsquo American Sociological Review 57540ndash59

Baum Joel A C and Singh Jitendra V (1994a) lsquoOrganizationndashenvironment coevo-lutionrsquo in Joel A C Baum and Jitendra V Singh (eds) The EvolutionaryDynamics of Organizations New York Oxford University Press pp 379ndash402

mdashmdash(eds) (1994b) The Evolutionary Dynamics of Organizations New York OxfordUniversity Press

Berger Suzanne and Dore Ronald (eds) (1996) National Diversity and GlobalCapitalism Ithaca NY Cornell University Press

Boyer Robert (1997) lsquoThe variety and unequal performance of really existingmarkets farewell to Doctor Panglossrsquo in J Rogers Hollingsworth and RobertBoyer (eds) Contemporary Capitalism the Embeddedness of InstitutionsCambridge and New York Cambridge University Press pp 55ndash93

Burns Tom R and Dietz Thomas (2001 forthcoming) lsquoSocial rule system theorysocial action institutional arrangements and evolutionary processesrsquo in JRogers Hollingsworth and Karl Muumlller (eds) Advances in SocioeconomicsLanham Md Rowman amp Littleeld

Burns Tom R and Flam Helena (1987) The Shaping of Social Organization BeverlyHills Calif Sage

Calvert Randall L (1995) lsquoThe rational choice theory of social institutionsrsquo inJeffrey S Banks and Eric A Hanuskek (eds) Modern Political Economy NewYork Cambridge University Press pp 216ndash66

Camic Charles (1986) lsquoThe matter of habitrsquo American Journal of Sociology 911039ndash87

Campbell John L (1997) lsquoMechanisms of evolutionary change in economicsgovernance interaction interpretation and bricolagersquo in Lars Magnusson

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

636

the resource dependency ndash emphasize how the environment inuencesorganizations and how organizations which are subject to the sameenvironment tend to converge in their behavior to have what DiMaggioand Powell have labeled organizational isomorphism (DiMaggio andPowell 1983)There has emerged a vast literature which demonstrates that within

each society there are scal political judicial and other regulatory normswhich limit and shape the culture and structure of organizationalbehavior The normative institutional environment of organizations limits the options of what organizations do in a particular society and inuences the patterns of ownership relations with suppliers andcustomers In short it is the normative environment of organizationswhich denes within a particular society what is socially acceptablebehavior for organizations (Hamilton and Biggart 1988 Hollingsworthand Hollingsworth 2000 Meyer and Rowan 1991 Zucker 1987 1988Orru et al 1991 Townley 1997)Thus far most of the analyses on isomorphic pressures operating on

organizations have been cross-sectional in nature However the entireconcept of isomorphism implies that there are strong environmental pres-sures exerted on organizations and in order to observe this phenomenonit is important that we have longitudinal studies which assess howchanges in an organizationrsquos institutional environment inuence changesin the structure and culture of organizations Moreover the expectationhere is that when one analyzes the patterns of organizational changeover time the isomorphism will be holistic ndash that is within a particularsociety there will be a tendency for the internal structure of particulartypes of organizations to converge in large part because of the pres-sures to conform to the changes in the external norms rules and valuesA close reading of Aoki (1990) and Slack and Hinings (1994) nds thatchanges in the national institutional environment of organizations haveinuenced changes in the internal structure of organizations within aspecic industry of a particular societyHowever it is inappropriate to view organizations as changing only

in relation to exogenous environmental change There is a kind of co-evolution which occurs between organizations and their institutionalenvironments Co-evolution implies nonlinear feedback between orga-nizations and their institutional environments This is of course quitedifferent from the more common way of studying organizations bymodeling relations between independent and dependent variables Ofcourse the more common type of analysis is quite appropriate whenthere are simple relationships which do not involve complex feedbackprocesses But in the strategy for institutional analysis which is proposedherein it is less useful to separate independent from dependent vari-ables and more useful to understand the interacting and co-evolutionary

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

620

processes among all of the levels of institutional analysis discussed inthe various sections above (Baum and Singh 1994a 379ndash80 Mowery andNelson 1999 Murmann 1998)The logic of the perspective herein suggests that within every

society there is variation in the structure and culture of business rmsuniversities and other complex organizations (Kondra and Hinings1998) However this variation takes place within parameters which aresystem specic For example every German rm and every Germanuniversity is different from every other German university and rm but there are a set of cultural and structural characteristics which distin-guish German rms and universities from those in America Similarlyeach Japanese rm and university is unique yet they have a set ofcultural and structural characteristics which are system specic andwhich differentiate them from business rms and universities in other countriesIn all societies each organization has its own distinctive organizational

rules norms and conventions which are subordinate to the meta norms and rules of the larger society within which they are embeddedHowever the strength of the institutional environment within whichorganizations are embedded varies from society to society Some societies have multiple institutional environments and there is hetero-geneity in terms of what constitutes appropriate behavior oforganizations In such societies there can be conict over whatinstitutional logics should regulate specic organizational functions(Friedland and Alford 1991 Townley 1997 262ndash3) In those societies inwhich the institutional norms habits and rules are most developed and in which the institutional pressures to conform are greatest thereis less variation in the structure and culture of business rms and variouskinds of research organizations In such societies the connectednessbetween research organizations and their institutional environments is sufciently strong for organizations to have low autonomy to pursue independent strategies and goals in these societies there is a great deal of organizational isomorphism Conversely the weaker the institutional environment in which research organizations areembedded the greater the variation in the structure and culture of business rms and research organizations Moreover where the institu-tional environments are more weakly developed organizations havegreater autonomy and exibility to respond to the development of new knowledge and to be highly innovative Hence it is in those societies where the institutional environments are most developed and most rigid and when there is less organizational autonomy and exibility that fewer radical innovations in basic and applied science as well as in totally new products and industrial sectors havebeen made

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

621

Fifth level of analysis

We come next to the outputs and performance of the various institutionalcomponents of a society It is at this level that institutional componentsare more pragmatic and exible For example in the legal sector thereare specic statutes and court rulings in the state sector there are specicpolicies in the business sector there are new products new technolo-gies and market strategies It is at this nal level of analysis thatinstitutional spheres are most open and susceptible to change and cross-national mimicry is easiest and most commonThrough the outputs of the society we can obtain some assessment as

to how well a society is performing Moreover we can assess how inno-vative it is how egalitarian it is in the distribution of its resources howegalitarian it is in terms of levels of health education etc Just as it iscomplicated to measure how egalitarian a society is similarly it is dif-cult to assess how innovative it is For example societies vary in makingincremental and radical breakthroughs in basic and applied science in developing totally new products and new kinds of organizations inincremental and process innovations in existing products and organiza-tions in developing new and different forms of marketing ndash bothdomestically and globally (Hage and Hollingsworth 2000) Of coursesocieties vary in their rates of economic growth in their rates of economicproductivity in their quality of life (eg rates of crime life expectancyetc) All of these performance criteria feed back and inuence each ofthe levels of institutional analysis discussed above ndash rules norms valuesetc institutional arrangements institutional sectors the structure andform of organizations Moreover performance measures may inuenceeach other For example if a society has very low rates of growth itmay not be very innovative in some of the types of innovative activitiesmentioned above Good or poor performances of certain kinds inuenceother performance indicators (Hollingsworth et al 1990)Different institutional arrangements and different social systems of

production result in different types of economic performance Hence solong as societies have different social systems of production there areserious constraints on the degree to which they can converge in theirinnovative styles Different social systems of production tend to maxi-mize in a more or less explicit manner different performance criteriausually mixing considerations about static and dynamic efciency protsecurity social peace and economic andor political power In short incontrast to the implications of neoclassical economic theory in real-worldeconomies there are no universal standards all economically rationalactors attempt to maximize Economic history provides numerous exam-ples of how a great variety of principles of rationality are implementedin different societies

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

622

A critical question in institutional analysis is whether an existing socialsystem of production which supports a set of routines for a particularkind of technology and industry can shift from old practices to newones There are numerous historical accounts of how the social systemof production in a particular society which worked so well for a numberof years could not adapt to new technologies (Lazonick 1990 Veblen1915 Nelson 1994) Schumpeter (1983) and Freeman (1991) have devel-oped the idea of a techno-economic paradigm which suggests thatdifferent technologies require specic forms of organizational arrange-ments ndash or what I called social systems of production Because oflsquoinstitutional inertiarsquo the social system of production of a society maynot be able to adapt to a new techno-economic paradigm (Nelson 199458 Lanzara 1998)Whether a social system of production can sustain its performance

standards depends not only on its intrinsic economic rationality but alsoon where it ts into a larger system If a particular social system ofproduction is immune from the innovativeness and competition of analternative system survival can be long lasting But if different socialsystems of production with diverging criteria of good economic perfor-mance meet in the world arena the arbitrariness of nationally imposedconstructed performance standards may be superseded by alternativeperformance criteria as a result of international competitivenessThe world economy is also socially constructed just as are national

economies Even if different social systems of production are competingin the international arena it is not always possible to determine whichis more competitively effective at any moment in time Hegemonicnation-states in the short run shape the rules of trade to favor their indus-trial sectors and rms But the history of hegemonic powers suggeststhat in the longer run such social systems of production sustainedlargely by military and political power eventually give way to moreinnovative and competitive social systems of production In our ownday as nation-states are increasingly integrated into a world economyeven hegemonic powers lose their innovative and competitive advan-tage and their share of world output decreases Such a country mayattempt to restructure its institutional arrangements and to readjust itsperformance preferences But to restructure its system of innovationgenerally calls for a major redistribution of power within a societyLargely for this reason societies have historically had limited capacityto reconstruct a system of innovation in the image of their major competi-tors Each countryrsquos social system of production is a conguration of ahost of norms rules and values as well as of institutional arrangementsEach system is constantly changing and is open to inuence from othersystems And indeed many technologies and practices diffuse from onesociety to another but a societyrsquos capacity to be innovative is constrained

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

623

by the existing social system of production Thus the same technologymay exist in numerous countries but how it is employed and how itinuences societal outcomes and performances varies from one institu-tional conguration or society to another (Hollingsworth 1998)

PERSPECTIVES ON INSTITUTIONAL CHANGE

A fundamental problem which remains unresolved in institutionalanalysis involves the nature of institutional change There is much confu-sion and miscommunication in the social sciences about institutionalchange in large part because we do not even know how to measure therate of institutional change Of course we know that there are bothexternal and internal focuses for change and we have had a good bitof scholarship on this subject What is not clearly understood is thatinstitutions and institutional arrangements within societies are histori-cally rooted that there is a great deal of path-dependency to the waythat institutions evolve Moreover the more intricately linked each sectoris with each other sector and with a societyrsquos rules and norms the lesschoice actors have to devise new institutions and institutional arrange-ments On the other hand if a sector or actor is too isolated it may betoo weak within a society to bring about any kind of effective changeThese considerations are very critical to the problem of how muchfreedom actors have to create new institutions afresh to what extentinstitutions diffuse from one society to another (Murmann 1998 Lanzara1998) This is a fundamental question for students of institutionalanalysis as we attempt to understand how the different institutionallevels are linked togetherThere are a variety of reasons why there is confusion in studying insti-

tutional change There is disagreement over the extent to which actorshave the freedom to build new institutional arrangements and if theydo to what extent the new arrangements and organizational patternsmay depart from past practices Among some institutional economists(Williamson 1985) there is a basic assumption that actors build efcientinstitutions by pursuing their self-interest and by promoting theirstrategic goals Moreover there is the assumption that a mixing of fullrationality with market competition tends to produce all of the optimalinstitutions that are needed in order to coordinate a complex capitalisteconomy (Hollingsworth and Boyer 1997 ch 14 Hodgson 1999) Incontrast the argument developed here is much more complexSocial institutions are historically rooted and there is a great deal of

path-dependency in the way that various institutional componentsevolve The shape of institutional congurations at any moment limitsthe type of options for change Because there is a great deal of institu-tional inertia radical change in the institutional components of a society

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

624

is uncommon (Lanzara 1998) Recognizing the constraints existinginstitutions exert on actors we note that existing congurations of insti-tutional arrangements institutional sectors and organizational propertiescan limit but also inuence the degree to which norms and rules canchange Many scholars (North 1990 Ostrom 1990) have emphasized therole of norms rules ideologies and values in limiting the ability of actorsto develop new institutional arrangements which radically depart fromexisting ones However there is a two-way process of interaction aprocess of co-evolution (Baum and Singh 1994a Nelson 1994) Thus asFriedland and Alford (1991 244) argue actors develop new institutionalarrangements by recognizing organizational failures and low perfor-mance but these new arrangements do not necessarily change a societyrsquosnorms rules habits and its underlying value systemCampbell (1997) advances our understanding of how existing institu-

tions enable actors to construct radically new institutions He recognizesthat most institutions and institutional arrangements embody a degreeof inertia from which it is difcult for actors to depart Moreover thedifferential power relations among actors make it unlikely that lesspowerful actors can change the existing power structure of a societyMost change in the institutional components of a society evolves througha process of constrained selection which reects to a considerable degreethe existing arrangements and power relations This becomes the inter-pretative frame for social actors with institutionalized scripts and ritualswhich tend to be taken for granted and appear to be quite ordinary andnatural This institutionalized scripting of the social world is an extremelyimportant source of social stability and inertiaCampbell has developed a very fruitful explanation of change in terms

of the various institutional components outlined here He distinguishesbetween incremental and radical institutional change and argues thatradical institutional change occurs when social actors with widelydiffering norms cultural scripts and rituals for action engage in intenseinteraction with each other (see also Knight 1992) The more diverse theinteractions the greater the potential for institutional change He arguesthat as shifts in the composition of interaction among social actors occurchanges in the interpretation of problems and interests will follow Ifthere are only minor changes in the extent to which diverse actorsinteract there will be only modest institutional change But the morefundamental the changes in the interaction of diverse social actors thegreater the change in the way that actors interpret their world And thegreater the change in the way that actors interpret their world the morelikely that radical changes in various institutional components willfollow In other words diverse decision making tends to create a widerrange of interpretative frameworks of the social world than is likely tobe the case if social actors operate in isolation

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

625

This perspective of diverse interactions among social actors facilitatesour ability to move beyond the traditional view that existing institutionsconstrain the range of institutional alternatives which actors face Theolder view has tended to suggest that actors engage in institutionalchange by extending existing institutional principles habits and conven-tions Specifying the conditions which facilitate fundamental institutionalchange permits a much richer perspective for institutional analysisCampbellrsquos perspective about the conditions under which fundamentalor radical change in institutional components occurs is consistent withmy own research about the conditions under which major discoveriesor fundamental new knowledge occur the more scholars with diversebackgrounds interact with intensity and frequency the greater the like-lihood that they will develop new and alternative ways of thinking(Hollingsworth and Hollingsworth 2000)These views of institutional change are also consistent with the recent

work of Nelson and his colleagues (Mowery and Nelson 1999 Murmann1998) They report the research on a variety of industrial sectors in anumber of different countries revealing that networks of actors in partic-ular industries business and professional associations universityresearch departments and institutes are frequently able to use their collec-tive power to modify a societyrsquos property rights research system andgovernmental policies It is the intense and frequent interaction of thesediverse complex sets of actors ndash sometimes constituting a technologysystem ndash that often leads to incremental changes in a societyrsquos total socialsystem of production It is this kind of collective action that brings aboutorganizational change change in institutional sectors institutionalarrangements and even in the rules and norms of a society And if weare to understand how the style of innovativeness of a society changeswe must rst understand the changes in its institutional makeup It ishighly unlikely that a society will develop a new style of innovativenesswithout changing its institutional structure

FIRMS THEIR INSTITUTIONAL ENVIRONMENT ANDNATIONAL SYSTEMS OF INNOVATION

Variation in the innovative style of societies is shaped by their insti-tutional makeup ndash the various levels which are discussed aboveinstitutions institutional arrangements the structure and cohesion of institutional sectors which constitute a societyrsquos social system of pro-duction and the structure and culture of its organizations ndash especiallyits business rms and other research organizations By undertaking aninstitutional analysis of a society one can begin to understand in whatkinds of organizations the production of specic kinds of knowledgetakes place and how this is linked to particular kinds of innovativeness

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

626

With such a perspective one can gain a rich understanding of why somesocieties excel in the production of radical breakthroughs in basic andapplied science and in developing radically new products and why othersocieties excel in more incremental innovationsReecting on these issues one cannot help being impressed with the

contrasts between the innovative styles of German and American orga-nizations For example Germany has continued to be very successful in making incremental innovations in many industries in which it wasvery competitive well before World War II paper printing materialsmachinery electrotechnical products motor vehicles chemicals textileyarns and fabrics But the Germans have been much less successful indeveloping totally new products or related innovations in biotechnologyelectronics telecommunications aircraft as well as other newer indus-trial areas In contrast the Americans have been very innovative in these and other industries with products having very short lives andwith technologies which change very rapidly and are very complex(Hollingsworth 1991 Soskice 1997 1999)The importance of institutions and innovativeness for international

competitiveness has led to the recognition that there are national systemsof innovation (Lundvall 1992 Nelson 1993 Edquist 1997 Hage andHollingsworth 2000) But even before the recent studies of nationalsystems of innovation Landes (1969) was a proponent of these viewswhen he made the argument that Germany had developed a nationalsystem of innovation by building on its education and scientic researchsystems which were lacking in Britain and elsewhere However the concept of a national system of innovation immediately poses theproblem of what kind of innovations should be consideredDavid Soskice an economist at the Wissenschaftszentrum Berlin has

addressed a set of problems somewhat complementary to the perspec-tive raised here His work (Soskice 1990 1997 1999) suggests that weare sensitive to the institutional environment in which the followingactors are embedded the employees of rms the owners and nanciersof rms the competitors of rms and nally their collaborators ndash aperspective complementary to Hollingsworthrsquos work on social systemsof production (Hollingsworth 1997 1998) When one reects on the insti-tutional contextsocial system of production within which these actorsare embedded one becomes aware that those countries which have some-what inexible labor markets external to rms but a business system in which rms have long-term commitments to their employees arestrongly associated with incremental but hardly at all with radical inno-vations In the rms of such countries (eg Germany and Japan) thereis much more consensus decision making than in societies where a lsquohireand rersquo set of practices is quite pervasive And this kind of consensusdecision making ndash while highly conducive to incremental innovativeness

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

627

ndash limits the capacity for radical innovations to occur This kind of lsquolongtermismrsquo set of norms and rules occurs not only in the employmentpractices of incrementally innovative rms but also in their nancingand ownership patterns (Hollingsworth 1997) During the past thirtyyears these were characteristics of many rms in Germany and JapanSignicantly in both of these countries venture capital markets werenot highly developed nancing tended to be based more on long-termbank loans than on the equity markets and the relationships of rmstended to be relatively stable And these stable institutional arrange-ments also contributed to a high degree of innovativeness ndash butincremental rather than radical innovationsMeantime we should also pursue Soskicersquos ideas (1999) about how

rmsrsquo relationships with their competitors and collaborators shape theirstyles of innovativeness In those advanced industrial societies in whichbusiness rms are highly mobilized into business associations theretends to be a high degree of standardized quality controls over technicalnorms and a high degree of reliance on legal sanctions as monitoringinstruments vis-agrave-vis association members As a result of this high degreeof integration of rms into associations member rms are more likelyto be associated with incremental innovations than is likely to be the case with rms in those societies where associations are less welldeveloped In the latter type of society radical innovations are morelikely to occur In those advanced industrial societies where rms havelong-term stable relationships with collaborators in product design andproduction the style of innovation is likely to be incremental And it is in those societies in which there is a weaker tradition of strongcollaboration among rms that radical innovations are more likely tooccurObviously even in a country where there may be a great deal of

radical innovativeness most innovations will be of an incremental typefor there are simply fewer radical innovations But Germany and therest of Europe not only have had relatively few radical innovations incomparison with the United States but they also have lagged behindthe USA in developing relatively totally new industries Thus there hasbeen much less innovative activity in Europe in high-technology andnewly emerging industries than in the USAIn complementary research Audretsch and Feldman (1995) have

demonstrated that the USA has been more successful in spawning rmsengaged in the early stages of the industry life cycle but relative toGermany had an environment unfavorable to innovative activity byestablished rms especially in more mature industries German rmshave tended to be associated with a routinized technological regime ndashwhere established enterprises have had strong incentives to engage inincremental innovations but new rms have tended to have fundamental

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

628

innovative disadvantages relative to the Americans (see Audretsch 1995Winter 1982 Casper 2000)On both sides of the Atlantic universities engage in basic and applied

research But signicantly the institutional environment in which univer-sities are embedded inuences the degree to which the research is likelyto represent major breakthroughs or is likely to have more modestimplications Recent research in biomedical science demonstrates that inthose societies in which university professors andor graduate studentshave the following characteristics there are likely to be relatively majorresearch breakthroughs graduate students have a long training periodhigh dependency on a major professor low autonomy in dening thenature of a research project restricted job options and a university envi-ronment where there is a very modest or nonexistent entrepreneurialculture For example in Germany ndash a country with relatively few radicalbreakthroughs ndash the scholar who nishes a habilitation will have spentmany years on a research project before qualifying for a permanent acad-emic position will have been in a high-dependency position with a seniorprofessor will have had limited autonomy to dene a research projectand will be constrained from pursuing a high-risk research project Theyoung German will have an extraordinarily high level of training butwill have faced a number of disincentives to undertake high-risk researchAnd as soon as heshe has completed the habilitation and becomes aprofessor administrative responsibilities constitute another disincentivefrom engaging in high-risk or long-term research projects Moreover theGerman professor is a civil servant with a more stable and secure incomethan hisher American counterpart In a very different institutionalenvironment American professors in large research universities tend tobecome much more entrepreneurial Today many American professorsbehave as though they were operating their own rms within univer-sities and thus devote considerable time to generating funds for researchprojects assistants secretaries supplies etc And the variation in thisentrepreneurial spirit across countries is associated with variation in thetaking of high-risk research strategies and the making of major researchbreakthroughsIn sum societies vary in terms of the institutional environment in

which their organizations are embedded and this variation inuencesthe degree to which they have industrial activities in the early or latestages of the industry life cycle successful new or mature industriesand radical vs incremental innovations And it is this variation whichis encouraging rms in countries with relatively few new industries and with few or no radical innovations to invest in those countries with many new industries and numerous radical innovations ndash as astrategy to be able to transfer new technologies from the latter institu-tional environment to the former

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

629

Table 4 summarizes the way that the norms rules and values of a society are associated with Americarsquos and Germanyrsquos institutionalarrangements their social system of production and their styles of inno-vativeness

CONCLUSION

Several themes deserve to be reemphasized Most importantly institu-tional analysis is emerging as a dominant issue in several social sciencedisciplines The study of various institutional components discussedabove is a complex subject requiring multiple levels of analysis and aknowledge of multiple sectors within societies It requires a theoreticalperspective that evolves from an empirical analysis in different societiesIf we are to advance the study of institutional analysis we need cross-disciplinary collaboration and team research which are not readilyachieved given the structure of most of our research universitiesAt present scholars in each academic discipline have their own distinc-

tive strategies for studying institutional analysis Thus economists tendto use several approaches (North 1990 Williamson 1985 Hodgson 19881998 1999) anthropologists another (Geertz 1995) and sociologistspolitical scientists and historians employ other strategies (Campbell etal 1991 Powell and DiMaggio 1991 Katznelson 1998 Hall and Taylor1996) With practitioners in various disciplines pursuing their distinctiveapproaches to institutional analysis there is low potential for collabo-rative institutional analysis across the social sciences The lessons fromthe biological sciences over the last 40 years are very instructive oncepractitioners from various biological disciplines began to cooperate instudying biological phenomena by using the same concepts at the mole-cular level theoretical advance was very rapid Comparable phenomenatook place in the development of other hybrid elds eg biochemistrybiophysics (Judson 1979 Olby 1979 Kohler 1982 1994) Similarly weresocial scientists in various disciplines to work collaboratively in the areaof institutional analysis there would very likely be an acceleration intheir theoretical knowledgeThis is not a plea that everyone engaged in institutional analysis should

do the same kind of research Indeed just as biologists working at themolecular level study many different kinds of problems similarly socialscientists engaged in institutional analysis would also work on manyseparate problems but in the spirit of a collective enterprise Some wouldconcentrate their attention on the study of rules norms habits conven-tions and values while others would study how these are associatedwith congurations of various institutional arrangements (types ofmarkets hierarchies networks associations communities clans statesetc) Others would work on specic institutional sectors (eg education

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

630

Table 4

Rules norms hab

its an

d values associated w

ith sp

ecic co

ngu

ration

s of coo

rdinating mecha

nism

s soc

ial sy

stem

s of

prod

uction

an

d nationa

l styles of inno

vation

USA

Germany

Norms rules values which provide incentives for

Norms rules values which provide incentives for

short-term horizons

long-term horizons

Con

gu

rative

forms of

Marke

ts c

orpo

rate hierarchies reg

ulatory state

Assoc

iation

s (busine

ss assoc

iation

s la

bor union

s)

coordinating

corp

oratist-type

networks

marke

ts c

orporatist-

mecha

nism

s (listed in

type

state

order of im

portanc

e in

the particu

lar social

system

)

Social sys

tem of

prod

uction

Bus

iness system

Firm

struc

tures

Con

glom

erates b

ut with mov

emen

t toward

Firm

more co

ngru

ent with pa

rticular prod

ucts

elim

ination of produc

tion

in som

e sectors

emph

asis on long

-term strateg

ies rm

s are well

emph

asis on sh

ort-term

strateg

ies high

integrated

differentiation

between divisions

of rms

Firm

s mob

ilized in

Low

High

and in

tegrated

into busine

ss

associations

Indu

strial relations

Low

deg

ree of jo

b secu

rity rigid in

ternal lab

or

High deg

ree of jo

b secu

rity

exible interna

l lab

or

system

marke

t ex

ible externa

l labo

r marke

tsmarke

t rigid externa

l lab

or m

arke

tPatterns of

Dispersed owne

rship frequ

ent turnov

erMore stab

le owne

rship patterns

owne

rship

Table 4

continue

d

USA

Germany

Norms rules values which provide incentives for

Norms rules values which provide incentives for

short-term horizons

long-term horizons

Social system of

Prod

uction

Bus

iness system

Training for labo

r an

dHigh deg

ree of voluntary decision mak

ing by

Collective an

d m

ore co

mpulso

ry d

ecisions

abo

ut

man

agem

ent

individuals as to ho

w m

uch

tim

e an

d ene

rgy

how m

uch ed

ucation

is ne

eded

he

avy em

pha

sis

to inv

est in educ

ation heavy

emph

asis on

on techn

ical training (eg en

gine

ering)

busine

ss sch

ools training

in marke

ting

and

sales accou

nting

Fina

ncial marke

tsEqu

ity an

d ven

ture cap

ital m

arke

ts highly

Equ

ity an

d ven

ture cap

ital m

arke

ts w

eakly

dev

elop

ed na

ncing from

cen

tral gov

ernm

ent

deve

lope

d he

avy em

pha

sis on

corporate loan

sto priva

te rm

s for dev

elop

men

t of new

tech

nologies related

to na

tion

al security

Unive

rsity research

High individual auton

omy high poten

tial for

Low

ind

ividua

l au

tono

my until almost 40

yea

rs of

system

individual creativity u

nive

rsity pr

ofesso

rs

age poten

tial for ind

ividual creativity high

ly

high

ly entreprene

urial

cons

traine

d unive

rsity pr

ofessors not so

entrep

rene

urial

Inno

vative

styles

Highly su

ccessful in m

ajor break

through

s in

High pe

rforman

ce in proc

ess an

d inc

remen

tal

basic an

d app

lied scien

ce go

od in dev

elop

ing

inno

vation

s in existing produ

cts not so su

ccessful

totally

new

lines of pr

oducts an

d in m

ajor break

through

s in basic scien

ce or in

tech

nologies high

deg

ree of entrepr

eneu

rship

deve

loping

totally new

lines of prod

ucts of

amon

g research

ers

tech

nologies low d

egree of entrepr

eneu

rship

amon

g research

ers

Table 4

continue

d

USA

Germany

Norms rules values which provide incentives for

Norms rules values which provide incentives for

short-term horizons

long-term horizons

Prod

ucts in w

hich

Pr

oducts with sh

ort liv

es prod

ucts in w

hich

Older produ

cts with long

history machine

ry pap

er

society excels

tech

nology

is bo

th highly co

mplex

and

pr

inting

textile

s m

otor veh

icles m

achine

too

lsch

ange

s rapidly en

tertainm

ent indus

try

pha

rmaceu

ticals biotechn

olog

y softw

are

aerosp

ace

business systems nancial markets and systems of research) Distinctivewould be the recognition of how these separate studies are linkedtogether how each institutional sector is linked to a societyrsquos normsrules values etc and its conguration of institutional arrangementsSimilarly other institutional analysts might focus on the study of orga-nizations with a major concern for how the institutional environmentof organizations (eg norms rules values congurations of institutionalarrangements institutional sectors) inuenced the structure culture andoutcomes of organizations And nally others might study how all ofthe aforementioned aspects of institutional analysis inuenced a societyrsquosoverall performance Nelson and Sampat (1998) have appropriatelyreminded us that we should engage in institutional analysis in order toaddress specic problems but the view developed in this article is thatwe rst need some mapping of the terrain of institutional analysis inorder for our diverse enterprises to become a meaningful collective enter-prise especially if we are to understand how institutions inuence theinnovative processThe argument developed here is not that all social scientists should engage

in institutional analysis Obviously there are numerous other importantareas of research which lie beyond the eld of institutional analysis Nordoes this article argue that institutional analysis is the most importantsubject for social scientistsSince the structure and culture of our research universities reect an

enormous amount of disciplinary fragmentation what are the prospectsfor promoting the kind of collective and interdisciplinary research agendaproposed above In one respect the prospects are not encouragingDuring the past half century the efforts of American universities to have effective communication across the biological sciences have variedgreatly in their success At the University of California-Berkeley it was not until the 1980s that there was a serious effort to integrate thebiological sciences and at that time the campus abolished seventeendepartments to promote an integrated biological program But at theUniversity of Wisconsin the biological sciences are still very fragmentedinto dozens of different departments in botany plant pathology zoologyand physiology with two departments of biochemistry and two depart-ments of geneticsThere have been other approaches which have worked very well histor-

ically for promoting scientic integration amidst scientic diversity Forexample several summer schools and special institutes were organizedin the United States during the 1940s in an effort to overcome the frag-mentation in the biological sciences These programs made it possiblefor scientists in diverse academic disciplines and from different univer-sities and countries to come together to engage in intense and frequentinteraction Perhaps the most effective program was that held in the

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

634

summers at Cold Spring Harbor New York where a research agendafor the study of molecular biology was developed (Fischer and Lipson1988 Stent and Watson 1966) Scientists went there summer aftersummer and when they returned to their universities many began therestructuring of the biological sciences in their universitiesIn our own day we need summer institutes and other similar programs

so that scholars from diverse backgrounds can come together anddevelop a collective agenda for doing institutional analysis and forstudying the impact of institutions on the performance of societies ndashinnovativeness being one such subject If we can develop such instituteswe will increase the prospects for transforming our universities Thereis reason to be optimistic that once an eclectic group of scholars workwith a common set of concepts we can have genuine theoretical advanceif researchers can bring their diverse expertise to bear on a common setof problems There is considerable evidence that institutional analysisoffers high potential for advancing our knowledge of innovations andtechnological change (Hollingsworth and Hollingsworth 2000) If we aregoing to make substantial theoretical advance it is not likely to occurfrom the vantage point of a multiplicity of disciplines Rather it is likelyto occur as we develop a new intellectual framework and our own acad-emic journals This is a tall agenda which requires us to reassess theway that knowledge is presently produced in our societies

NOTE

This article is part of a much larger agenda involving the study of institutionsorganizations and innovations in which I am involved My debts are many andvaried The rst version was written while I was in residence in the ResearchUnit for Institutional Change and European Integration of the Austrian Academyof Sciences in Vienna in the summer of 1998 I am especially grateful to ProfessorsEgon Matzner and Sonja Puntscher-Riekmann for providing the stimulating envi-ronment in which to think through the ideas developed here Egon Matzner hasengaged in very stimulating discussions with me about the importance of insti-tutions in the general eld of socio-economics The nal version of the articlewas completed during my residence as a fellow at the Neurosciences Institutein La Jolla California I also wish to thank my colleagues in the InnovationTheme Project at the Netherlands Institute for Advanced Study (NIAS) during1998ndash9 and Edgar Grande and Karl Muumlller for stimulating conversations aboutthe interaction of institutions and organizations in facilitating particular typesof innovation I am especially grateful to Jerald Hage for helping me to developa number of issues in the article and for his useful comments on the rst draftFrans van Waarden early on helped me to work through the idea of multiplelevels of institutional analysis Others who have helped to develop the ideas inthe article have been my colleagues in the research project lsquoComparative socialsystems of productionrsquo especially Tom Burns Christel Lane Yoshitaka OkadaWolfgang Streeck and Richard Whitley My colleagues Robert Boyer SteveCasper and David Soskice have taught me much about how incentives providedby the institutional environments of rms inuence individual decision making

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

635

But my most important debt is to Ellen Jane Hollingsworth who has repeatedlyquestioned every single assumption of this article Without her rich and contin-uous tutoring this article would be even less developed than it is at presentFinally without the insightful assistance of David Gear in many phases of thisarticle it would not have been completed

REFERENCES

Aoki Masahiko (1990) lsquoTowards an economic model of the Japanese rmrsquo Journalof Economic Literature March 1ndash27

Arora A Landau Ralph and Rosenberg Nathan (eds) (1998) Chemicals and Long-Term Economic Growth Insights from the Chemical Industry New York JohnWiley amp Sons

Arrow Kenneth (1974) The Limits of Organization New York NortonArthur W B (1988a) lsquoSelf-reinforcing mechanisms in economicsrsquo in P W

Anderson K J Arrow and D Pines (eds) The Economy as an Evolving ComplexSystem Redwood City Calif Addison-Wesley

mdashmdash(1988b) lsquoCompeting technologies an overviewrsquo in G Dosi C Freeman RNelson G Silverberg and L Soete (eds) Technical Change and Economic TheoryLondon Pinter Publishers pp 590ndash607

Audretsch David B (1995) Innovation and Industry Evolution Cambridge MassMIT Press

Audretsch David B and Feldman Maryann P (1995) lsquoInnovative clusters andthe industry life cyclersquo Discussion Paper Wissenschaftszentrum BerlinMarket Processes and Corporate Development Research Unit FS-IV-95-7

Baum Joel A C and Oliver Christine (1992) lsquoInstitutional embeddedness andthe dynamics of organizational populationsrsquo American Sociological Review 57540ndash59

Baum Joel A C and Singh Jitendra V (1994a) lsquoOrganizationndashenvironment coevo-lutionrsquo in Joel A C Baum and Jitendra V Singh (eds) The EvolutionaryDynamics of Organizations New York Oxford University Press pp 379ndash402

mdashmdash(eds) (1994b) The Evolutionary Dynamics of Organizations New York OxfordUniversity Press

Berger Suzanne and Dore Ronald (eds) (1996) National Diversity and GlobalCapitalism Ithaca NY Cornell University Press

Boyer Robert (1997) lsquoThe variety and unequal performance of really existingmarkets farewell to Doctor Panglossrsquo in J Rogers Hollingsworth and RobertBoyer (eds) Contemporary Capitalism the Embeddedness of InstitutionsCambridge and New York Cambridge University Press pp 55ndash93

Burns Tom R and Dietz Thomas (2001 forthcoming) lsquoSocial rule system theorysocial action institutional arrangements and evolutionary processesrsquo in JRogers Hollingsworth and Karl Muumlller (eds) Advances in SocioeconomicsLanham Md Rowman amp Littleeld

Burns Tom R and Flam Helena (1987) The Shaping of Social Organization BeverlyHills Calif Sage

Calvert Randall L (1995) lsquoThe rational choice theory of social institutionsrsquo inJeffrey S Banks and Eric A Hanuskek (eds) Modern Political Economy NewYork Cambridge University Press pp 216ndash66

Camic Charles (1986) lsquoThe matter of habitrsquo American Journal of Sociology 911039ndash87

Campbell John L (1997) lsquoMechanisms of evolutionary change in economicsgovernance interaction interpretation and bricolagersquo in Lars Magnusson

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

636

processes among all of the levels of institutional analysis discussed inthe various sections above (Baum and Singh 1994a 379ndash80 Mowery andNelson 1999 Murmann 1998)The logic of the perspective herein suggests that within every

society there is variation in the structure and culture of business rmsuniversities and other complex organizations (Kondra and Hinings1998) However this variation takes place within parameters which aresystem specic For example every German rm and every Germanuniversity is different from every other German university and rm but there are a set of cultural and structural characteristics which distin-guish German rms and universities from those in America Similarlyeach Japanese rm and university is unique yet they have a set ofcultural and structural characteristics which are system specic andwhich differentiate them from business rms and universities in other countriesIn all societies each organization has its own distinctive organizational

rules norms and conventions which are subordinate to the meta norms and rules of the larger society within which they are embeddedHowever the strength of the institutional environment within whichorganizations are embedded varies from society to society Some societies have multiple institutional environments and there is hetero-geneity in terms of what constitutes appropriate behavior oforganizations In such societies there can be conict over whatinstitutional logics should regulate specic organizational functions(Friedland and Alford 1991 Townley 1997 262ndash3) In those societies inwhich the institutional norms habits and rules are most developed and in which the institutional pressures to conform are greatest thereis less variation in the structure and culture of business rms and variouskinds of research organizations In such societies the connectednessbetween research organizations and their institutional environments is sufciently strong for organizations to have low autonomy to pursue independent strategies and goals in these societies there is a great deal of organizational isomorphism Conversely the weaker the institutional environment in which research organizations areembedded the greater the variation in the structure and culture of business rms and research organizations Moreover where the institu-tional environments are more weakly developed organizations havegreater autonomy and exibility to respond to the development of new knowledge and to be highly innovative Hence it is in those societies where the institutional environments are most developed and most rigid and when there is less organizational autonomy and exibility that fewer radical innovations in basic and applied science as well as in totally new products and industrial sectors havebeen made

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

621

Fifth level of analysis

We come next to the outputs and performance of the various institutionalcomponents of a society It is at this level that institutional componentsare more pragmatic and exible For example in the legal sector thereare specic statutes and court rulings in the state sector there are specicpolicies in the business sector there are new products new technolo-gies and market strategies It is at this nal level of analysis thatinstitutional spheres are most open and susceptible to change and cross-national mimicry is easiest and most commonThrough the outputs of the society we can obtain some assessment as

to how well a society is performing Moreover we can assess how inno-vative it is how egalitarian it is in the distribution of its resources howegalitarian it is in terms of levels of health education etc Just as it iscomplicated to measure how egalitarian a society is similarly it is dif-cult to assess how innovative it is For example societies vary in makingincremental and radical breakthroughs in basic and applied science in developing totally new products and new kinds of organizations inincremental and process innovations in existing products and organiza-tions in developing new and different forms of marketing ndash bothdomestically and globally (Hage and Hollingsworth 2000) Of coursesocieties vary in their rates of economic growth in their rates of economicproductivity in their quality of life (eg rates of crime life expectancyetc) All of these performance criteria feed back and inuence each ofthe levels of institutional analysis discussed above ndash rules norms valuesetc institutional arrangements institutional sectors the structure andform of organizations Moreover performance measures may inuenceeach other For example if a society has very low rates of growth itmay not be very innovative in some of the types of innovative activitiesmentioned above Good or poor performances of certain kinds inuenceother performance indicators (Hollingsworth et al 1990)Different institutional arrangements and different social systems of

production result in different types of economic performance Hence solong as societies have different social systems of production there areserious constraints on the degree to which they can converge in theirinnovative styles Different social systems of production tend to maxi-mize in a more or less explicit manner different performance criteriausually mixing considerations about static and dynamic efciency protsecurity social peace and economic andor political power In short incontrast to the implications of neoclassical economic theory in real-worldeconomies there are no universal standards all economically rationalactors attempt to maximize Economic history provides numerous exam-ples of how a great variety of principles of rationality are implementedin different societies

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

622

A critical question in institutional analysis is whether an existing socialsystem of production which supports a set of routines for a particularkind of technology and industry can shift from old practices to newones There are numerous historical accounts of how the social systemof production in a particular society which worked so well for a numberof years could not adapt to new technologies (Lazonick 1990 Veblen1915 Nelson 1994) Schumpeter (1983) and Freeman (1991) have devel-oped the idea of a techno-economic paradigm which suggests thatdifferent technologies require specic forms of organizational arrange-ments ndash or what I called social systems of production Because oflsquoinstitutional inertiarsquo the social system of production of a society maynot be able to adapt to a new techno-economic paradigm (Nelson 199458 Lanzara 1998)Whether a social system of production can sustain its performance

standards depends not only on its intrinsic economic rationality but alsoon where it ts into a larger system If a particular social system ofproduction is immune from the innovativeness and competition of analternative system survival can be long lasting But if different socialsystems of production with diverging criteria of good economic perfor-mance meet in the world arena the arbitrariness of nationally imposedconstructed performance standards may be superseded by alternativeperformance criteria as a result of international competitivenessThe world economy is also socially constructed just as are national

economies Even if different social systems of production are competingin the international arena it is not always possible to determine whichis more competitively effective at any moment in time Hegemonicnation-states in the short run shape the rules of trade to favor their indus-trial sectors and rms But the history of hegemonic powers suggeststhat in the longer run such social systems of production sustainedlargely by military and political power eventually give way to moreinnovative and competitive social systems of production In our ownday as nation-states are increasingly integrated into a world economyeven hegemonic powers lose their innovative and competitive advan-tage and their share of world output decreases Such a country mayattempt to restructure its institutional arrangements and to readjust itsperformance preferences But to restructure its system of innovationgenerally calls for a major redistribution of power within a societyLargely for this reason societies have historically had limited capacityto reconstruct a system of innovation in the image of their major competi-tors Each countryrsquos social system of production is a conguration of ahost of norms rules and values as well as of institutional arrangementsEach system is constantly changing and is open to inuence from othersystems And indeed many technologies and practices diffuse from onesociety to another but a societyrsquos capacity to be innovative is constrained

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

623

by the existing social system of production Thus the same technologymay exist in numerous countries but how it is employed and how itinuences societal outcomes and performances varies from one institu-tional conguration or society to another (Hollingsworth 1998)

PERSPECTIVES ON INSTITUTIONAL CHANGE

A fundamental problem which remains unresolved in institutionalanalysis involves the nature of institutional change There is much confu-sion and miscommunication in the social sciences about institutionalchange in large part because we do not even know how to measure therate of institutional change Of course we know that there are bothexternal and internal focuses for change and we have had a good bitof scholarship on this subject What is not clearly understood is thatinstitutions and institutional arrangements within societies are histori-cally rooted that there is a great deal of path-dependency to the waythat institutions evolve Moreover the more intricately linked each sectoris with each other sector and with a societyrsquos rules and norms the lesschoice actors have to devise new institutions and institutional arrange-ments On the other hand if a sector or actor is too isolated it may betoo weak within a society to bring about any kind of effective changeThese considerations are very critical to the problem of how muchfreedom actors have to create new institutions afresh to what extentinstitutions diffuse from one society to another (Murmann 1998 Lanzara1998) This is a fundamental question for students of institutionalanalysis as we attempt to understand how the different institutionallevels are linked togetherThere are a variety of reasons why there is confusion in studying insti-

tutional change There is disagreement over the extent to which actorshave the freedom to build new institutional arrangements and if theydo to what extent the new arrangements and organizational patternsmay depart from past practices Among some institutional economists(Williamson 1985) there is a basic assumption that actors build efcientinstitutions by pursuing their self-interest and by promoting theirstrategic goals Moreover there is the assumption that a mixing of fullrationality with market competition tends to produce all of the optimalinstitutions that are needed in order to coordinate a complex capitalisteconomy (Hollingsworth and Boyer 1997 ch 14 Hodgson 1999) Incontrast the argument developed here is much more complexSocial institutions are historically rooted and there is a great deal of

path-dependency in the way that various institutional componentsevolve The shape of institutional congurations at any moment limitsthe type of options for change Because there is a great deal of institu-tional inertia radical change in the institutional components of a society

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

624

is uncommon (Lanzara 1998) Recognizing the constraints existinginstitutions exert on actors we note that existing congurations of insti-tutional arrangements institutional sectors and organizational propertiescan limit but also inuence the degree to which norms and rules canchange Many scholars (North 1990 Ostrom 1990) have emphasized therole of norms rules ideologies and values in limiting the ability of actorsto develop new institutional arrangements which radically depart fromexisting ones However there is a two-way process of interaction aprocess of co-evolution (Baum and Singh 1994a Nelson 1994) Thus asFriedland and Alford (1991 244) argue actors develop new institutionalarrangements by recognizing organizational failures and low perfor-mance but these new arrangements do not necessarily change a societyrsquosnorms rules habits and its underlying value systemCampbell (1997) advances our understanding of how existing institu-

tions enable actors to construct radically new institutions He recognizesthat most institutions and institutional arrangements embody a degreeof inertia from which it is difcult for actors to depart Moreover thedifferential power relations among actors make it unlikely that lesspowerful actors can change the existing power structure of a societyMost change in the institutional components of a society evolves througha process of constrained selection which reects to a considerable degreethe existing arrangements and power relations This becomes the inter-pretative frame for social actors with institutionalized scripts and ritualswhich tend to be taken for granted and appear to be quite ordinary andnatural This institutionalized scripting of the social world is an extremelyimportant source of social stability and inertiaCampbell has developed a very fruitful explanation of change in terms

of the various institutional components outlined here He distinguishesbetween incremental and radical institutional change and argues thatradical institutional change occurs when social actors with widelydiffering norms cultural scripts and rituals for action engage in intenseinteraction with each other (see also Knight 1992) The more diverse theinteractions the greater the potential for institutional change He arguesthat as shifts in the composition of interaction among social actors occurchanges in the interpretation of problems and interests will follow Ifthere are only minor changes in the extent to which diverse actorsinteract there will be only modest institutional change But the morefundamental the changes in the interaction of diverse social actors thegreater the change in the way that actors interpret their world And thegreater the change in the way that actors interpret their world the morelikely that radical changes in various institutional components willfollow In other words diverse decision making tends to create a widerrange of interpretative frameworks of the social world than is likely tobe the case if social actors operate in isolation

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

625

This perspective of diverse interactions among social actors facilitatesour ability to move beyond the traditional view that existing institutionsconstrain the range of institutional alternatives which actors face Theolder view has tended to suggest that actors engage in institutionalchange by extending existing institutional principles habits and conven-tions Specifying the conditions which facilitate fundamental institutionalchange permits a much richer perspective for institutional analysisCampbellrsquos perspective about the conditions under which fundamentalor radical change in institutional components occurs is consistent withmy own research about the conditions under which major discoveriesor fundamental new knowledge occur the more scholars with diversebackgrounds interact with intensity and frequency the greater the like-lihood that they will develop new and alternative ways of thinking(Hollingsworth and Hollingsworth 2000)These views of institutional change are also consistent with the recent

work of Nelson and his colleagues (Mowery and Nelson 1999 Murmann1998) They report the research on a variety of industrial sectors in anumber of different countries revealing that networks of actors in partic-ular industries business and professional associations universityresearch departments and institutes are frequently able to use their collec-tive power to modify a societyrsquos property rights research system andgovernmental policies It is the intense and frequent interaction of thesediverse complex sets of actors ndash sometimes constituting a technologysystem ndash that often leads to incremental changes in a societyrsquos total socialsystem of production It is this kind of collective action that brings aboutorganizational change change in institutional sectors institutionalarrangements and even in the rules and norms of a society And if weare to understand how the style of innovativeness of a society changeswe must rst understand the changes in its institutional makeup It ishighly unlikely that a society will develop a new style of innovativenesswithout changing its institutional structure

FIRMS THEIR INSTITUTIONAL ENVIRONMENT ANDNATIONAL SYSTEMS OF INNOVATION

Variation in the innovative style of societies is shaped by their insti-tutional makeup ndash the various levels which are discussed aboveinstitutions institutional arrangements the structure and cohesion of institutional sectors which constitute a societyrsquos social system of pro-duction and the structure and culture of its organizations ndash especiallyits business rms and other research organizations By undertaking aninstitutional analysis of a society one can begin to understand in whatkinds of organizations the production of specic kinds of knowledgetakes place and how this is linked to particular kinds of innovativeness

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

626

With such a perspective one can gain a rich understanding of why somesocieties excel in the production of radical breakthroughs in basic andapplied science and in developing radically new products and why othersocieties excel in more incremental innovationsReecting on these issues one cannot help being impressed with the

contrasts between the innovative styles of German and American orga-nizations For example Germany has continued to be very successful in making incremental innovations in many industries in which it wasvery competitive well before World War II paper printing materialsmachinery electrotechnical products motor vehicles chemicals textileyarns and fabrics But the Germans have been much less successful indeveloping totally new products or related innovations in biotechnologyelectronics telecommunications aircraft as well as other newer indus-trial areas In contrast the Americans have been very innovative in these and other industries with products having very short lives andwith technologies which change very rapidly and are very complex(Hollingsworth 1991 Soskice 1997 1999)The importance of institutions and innovativeness for international

competitiveness has led to the recognition that there are national systemsof innovation (Lundvall 1992 Nelson 1993 Edquist 1997 Hage andHollingsworth 2000) But even before the recent studies of nationalsystems of innovation Landes (1969) was a proponent of these viewswhen he made the argument that Germany had developed a nationalsystem of innovation by building on its education and scientic researchsystems which were lacking in Britain and elsewhere However the concept of a national system of innovation immediately poses theproblem of what kind of innovations should be consideredDavid Soskice an economist at the Wissenschaftszentrum Berlin has

addressed a set of problems somewhat complementary to the perspec-tive raised here His work (Soskice 1990 1997 1999) suggests that weare sensitive to the institutional environment in which the followingactors are embedded the employees of rms the owners and nanciersof rms the competitors of rms and nally their collaborators ndash aperspective complementary to Hollingsworthrsquos work on social systemsof production (Hollingsworth 1997 1998) When one reects on the insti-tutional contextsocial system of production within which these actorsare embedded one becomes aware that those countries which have some-what inexible labor markets external to rms but a business system in which rms have long-term commitments to their employees arestrongly associated with incremental but hardly at all with radical inno-vations In the rms of such countries (eg Germany and Japan) thereis much more consensus decision making than in societies where a lsquohireand rersquo set of practices is quite pervasive And this kind of consensusdecision making ndash while highly conducive to incremental innovativeness

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

627

ndash limits the capacity for radical innovations to occur This kind of lsquolongtermismrsquo set of norms and rules occurs not only in the employmentpractices of incrementally innovative rms but also in their nancingand ownership patterns (Hollingsworth 1997) During the past thirtyyears these were characteristics of many rms in Germany and JapanSignicantly in both of these countries venture capital markets werenot highly developed nancing tended to be based more on long-termbank loans than on the equity markets and the relationships of rmstended to be relatively stable And these stable institutional arrange-ments also contributed to a high degree of innovativeness ndash butincremental rather than radical innovationsMeantime we should also pursue Soskicersquos ideas (1999) about how

rmsrsquo relationships with their competitors and collaborators shape theirstyles of innovativeness In those advanced industrial societies in whichbusiness rms are highly mobilized into business associations theretends to be a high degree of standardized quality controls over technicalnorms and a high degree of reliance on legal sanctions as monitoringinstruments vis-agrave-vis association members As a result of this high degreeof integration of rms into associations member rms are more likelyto be associated with incremental innovations than is likely to be the case with rms in those societies where associations are less welldeveloped In the latter type of society radical innovations are morelikely to occur In those advanced industrial societies where rms havelong-term stable relationships with collaborators in product design andproduction the style of innovation is likely to be incremental And it is in those societies in which there is a weaker tradition of strongcollaboration among rms that radical innovations are more likely tooccurObviously even in a country where there may be a great deal of

radical innovativeness most innovations will be of an incremental typefor there are simply fewer radical innovations But Germany and therest of Europe not only have had relatively few radical innovations incomparison with the United States but they also have lagged behindthe USA in developing relatively totally new industries Thus there hasbeen much less innovative activity in Europe in high-technology andnewly emerging industries than in the USAIn complementary research Audretsch and Feldman (1995) have

demonstrated that the USA has been more successful in spawning rmsengaged in the early stages of the industry life cycle but relative toGermany had an environment unfavorable to innovative activity byestablished rms especially in more mature industries German rmshave tended to be associated with a routinized technological regime ndashwhere established enterprises have had strong incentives to engage inincremental innovations but new rms have tended to have fundamental

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

628

innovative disadvantages relative to the Americans (see Audretsch 1995Winter 1982 Casper 2000)On both sides of the Atlantic universities engage in basic and applied

research But signicantly the institutional environment in which univer-sities are embedded inuences the degree to which the research is likelyto represent major breakthroughs or is likely to have more modestimplications Recent research in biomedical science demonstrates that inthose societies in which university professors andor graduate studentshave the following characteristics there are likely to be relatively majorresearch breakthroughs graduate students have a long training periodhigh dependency on a major professor low autonomy in dening thenature of a research project restricted job options and a university envi-ronment where there is a very modest or nonexistent entrepreneurialculture For example in Germany ndash a country with relatively few radicalbreakthroughs ndash the scholar who nishes a habilitation will have spentmany years on a research project before qualifying for a permanent acad-emic position will have been in a high-dependency position with a seniorprofessor will have had limited autonomy to dene a research projectand will be constrained from pursuing a high-risk research project Theyoung German will have an extraordinarily high level of training butwill have faced a number of disincentives to undertake high-risk researchAnd as soon as heshe has completed the habilitation and becomes aprofessor administrative responsibilities constitute another disincentivefrom engaging in high-risk or long-term research projects Moreover theGerman professor is a civil servant with a more stable and secure incomethan hisher American counterpart In a very different institutionalenvironment American professors in large research universities tend tobecome much more entrepreneurial Today many American professorsbehave as though they were operating their own rms within univer-sities and thus devote considerable time to generating funds for researchprojects assistants secretaries supplies etc And the variation in thisentrepreneurial spirit across countries is associated with variation in thetaking of high-risk research strategies and the making of major researchbreakthroughsIn sum societies vary in terms of the institutional environment in

which their organizations are embedded and this variation inuencesthe degree to which they have industrial activities in the early or latestages of the industry life cycle successful new or mature industriesand radical vs incremental innovations And it is this variation whichis encouraging rms in countries with relatively few new industries and with few or no radical innovations to invest in those countries with many new industries and numerous radical innovations ndash as astrategy to be able to transfer new technologies from the latter institu-tional environment to the former

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

629

Table 4 summarizes the way that the norms rules and values of a society are associated with Americarsquos and Germanyrsquos institutionalarrangements their social system of production and their styles of inno-vativeness

CONCLUSION

Several themes deserve to be reemphasized Most importantly institu-tional analysis is emerging as a dominant issue in several social sciencedisciplines The study of various institutional components discussedabove is a complex subject requiring multiple levels of analysis and aknowledge of multiple sectors within societies It requires a theoreticalperspective that evolves from an empirical analysis in different societiesIf we are to advance the study of institutional analysis we need cross-disciplinary collaboration and team research which are not readilyachieved given the structure of most of our research universitiesAt present scholars in each academic discipline have their own distinc-

tive strategies for studying institutional analysis Thus economists tendto use several approaches (North 1990 Williamson 1985 Hodgson 19881998 1999) anthropologists another (Geertz 1995) and sociologistspolitical scientists and historians employ other strategies (Campbell etal 1991 Powell and DiMaggio 1991 Katznelson 1998 Hall and Taylor1996) With practitioners in various disciplines pursuing their distinctiveapproaches to institutional analysis there is low potential for collabo-rative institutional analysis across the social sciences The lessons fromthe biological sciences over the last 40 years are very instructive oncepractitioners from various biological disciplines began to cooperate instudying biological phenomena by using the same concepts at the mole-cular level theoretical advance was very rapid Comparable phenomenatook place in the development of other hybrid elds eg biochemistrybiophysics (Judson 1979 Olby 1979 Kohler 1982 1994) Similarly weresocial scientists in various disciplines to work collaboratively in the areaof institutional analysis there would very likely be an acceleration intheir theoretical knowledgeThis is not a plea that everyone engaged in institutional analysis should

do the same kind of research Indeed just as biologists working at themolecular level study many different kinds of problems similarly socialscientists engaged in institutional analysis would also work on manyseparate problems but in the spirit of a collective enterprise Some wouldconcentrate their attention on the study of rules norms habits conven-tions and values while others would study how these are associatedwith congurations of various institutional arrangements (types ofmarkets hierarchies networks associations communities clans statesetc) Others would work on specic institutional sectors (eg education

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

630

Table 4

Rules norms hab

its an

d values associated w

ith sp

ecic co

ngu

ration

s of coo

rdinating mecha

nism

s soc

ial sy

stem

s of

prod

uction

an

d nationa

l styles of inno

vation

USA

Germany

Norms rules values which provide incentives for

Norms rules values which provide incentives for

short-term horizons

long-term horizons

Con

gu

rative

forms of

Marke

ts c

orpo

rate hierarchies reg

ulatory state

Assoc

iation

s (busine

ss assoc

iation

s la

bor union

s)

coordinating

corp

oratist-type

networks

marke

ts c

orporatist-

mecha

nism

s (listed in

type

state

order of im

portanc

e in

the particu

lar social

system

)

Social sys

tem of

prod

uction

Bus

iness system

Firm

struc

tures

Con

glom

erates b

ut with mov

emen

t toward

Firm

more co

ngru

ent with pa

rticular prod

ucts

elim

ination of produc

tion

in som

e sectors

emph

asis on long

-term strateg

ies rm

s are well

emph

asis on sh

ort-term

strateg

ies high

integrated

differentiation

between divisions

of rms

Firm

s mob

ilized in

Low

High

and in

tegrated

into busine

ss

associations

Indu

strial relations

Low

deg

ree of jo

b secu

rity rigid in

ternal lab

or

High deg

ree of jo

b secu

rity

exible interna

l lab

or

system

marke

t ex

ible externa

l labo

r marke

tsmarke

t rigid externa

l lab

or m

arke

tPatterns of

Dispersed owne

rship frequ

ent turnov

erMore stab

le owne

rship patterns

owne

rship

Table 4

continue

d

USA

Germany

Norms rules values which provide incentives for

Norms rules values which provide incentives for

short-term horizons

long-term horizons

Social system of

Prod

uction

Bus

iness system

Training for labo

r an

dHigh deg

ree of voluntary decision mak

ing by

Collective an

d m

ore co

mpulso

ry d

ecisions

abo

ut

man

agem

ent

individuals as to ho

w m

uch

tim

e an

d ene

rgy

how m

uch ed

ucation

is ne

eded

he

avy em

pha

sis

to inv

est in educ

ation heavy

emph

asis on

on techn

ical training (eg en

gine

ering)

busine

ss sch

ools training

in marke

ting

and

sales accou

nting

Fina

ncial marke

tsEqu

ity an

d ven

ture cap

ital m

arke

ts highly

Equ

ity an

d ven

ture cap

ital m

arke

ts w

eakly

dev

elop

ed na

ncing from

cen

tral gov

ernm

ent

deve

lope

d he

avy em

pha

sis on

corporate loan

sto priva

te rm

s for dev

elop

men

t of new

tech

nologies related

to na

tion

al security

Unive

rsity research

High individual auton

omy high poten

tial for

Low

ind

ividua

l au

tono

my until almost 40

yea

rs of

system

individual creativity u

nive

rsity pr

ofesso

rs

age poten

tial for ind

ividual creativity high

ly

high

ly entreprene

urial

cons

traine

d unive

rsity pr

ofessors not so

entrep

rene

urial

Inno

vative

styles

Highly su

ccessful in m

ajor break

through

s in

High pe

rforman

ce in proc

ess an

d inc

remen

tal

basic an

d app

lied scien

ce go

od in dev

elop

ing

inno

vation

s in existing produ

cts not so su

ccessful

totally

new

lines of pr

oducts an

d in m

ajor break

through

s in basic scien

ce or in

tech

nologies high

deg

ree of entrepr

eneu

rship

deve

loping

totally new

lines of prod

ucts of

amon

g research

ers

tech

nologies low d

egree of entrepr

eneu

rship

amon

g research

ers

Table 4

continue

d

USA

Germany

Norms rules values which provide incentives for

Norms rules values which provide incentives for

short-term horizons

long-term horizons

Prod

ucts in w

hich

Pr

oducts with sh

ort liv

es prod

ucts in w

hich

Older produ

cts with long

history machine

ry pap

er

society excels

tech

nology

is bo

th highly co

mplex

and

pr

inting

textile

s m

otor veh

icles m

achine

too

lsch

ange

s rapidly en

tertainm

ent indus

try

pha

rmaceu

ticals biotechn

olog

y softw

are

aerosp

ace

business systems nancial markets and systems of research) Distinctivewould be the recognition of how these separate studies are linkedtogether how each institutional sector is linked to a societyrsquos normsrules values etc and its conguration of institutional arrangementsSimilarly other institutional analysts might focus on the study of orga-nizations with a major concern for how the institutional environmentof organizations (eg norms rules values congurations of institutionalarrangements institutional sectors) inuenced the structure culture andoutcomes of organizations And nally others might study how all ofthe aforementioned aspects of institutional analysis inuenced a societyrsquosoverall performance Nelson and Sampat (1998) have appropriatelyreminded us that we should engage in institutional analysis in order toaddress specic problems but the view developed in this article is thatwe rst need some mapping of the terrain of institutional analysis inorder for our diverse enterprises to become a meaningful collective enter-prise especially if we are to understand how institutions inuence theinnovative processThe argument developed here is not that all social scientists should engage

in institutional analysis Obviously there are numerous other importantareas of research which lie beyond the eld of institutional analysis Nordoes this article argue that institutional analysis is the most importantsubject for social scientistsSince the structure and culture of our research universities reect an

enormous amount of disciplinary fragmentation what are the prospectsfor promoting the kind of collective and interdisciplinary research agendaproposed above In one respect the prospects are not encouragingDuring the past half century the efforts of American universities to have effective communication across the biological sciences have variedgreatly in their success At the University of California-Berkeley it was not until the 1980s that there was a serious effort to integrate thebiological sciences and at that time the campus abolished seventeendepartments to promote an integrated biological program But at theUniversity of Wisconsin the biological sciences are still very fragmentedinto dozens of different departments in botany plant pathology zoologyand physiology with two departments of biochemistry and two depart-ments of geneticsThere have been other approaches which have worked very well histor-

ically for promoting scientic integration amidst scientic diversity Forexample several summer schools and special institutes were organizedin the United States during the 1940s in an effort to overcome the frag-mentation in the biological sciences These programs made it possiblefor scientists in diverse academic disciplines and from different univer-sities and countries to come together to engage in intense and frequentinteraction Perhaps the most effective program was that held in the

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

634

summers at Cold Spring Harbor New York where a research agendafor the study of molecular biology was developed (Fischer and Lipson1988 Stent and Watson 1966) Scientists went there summer aftersummer and when they returned to their universities many began therestructuring of the biological sciences in their universitiesIn our own day we need summer institutes and other similar programs

so that scholars from diverse backgrounds can come together anddevelop a collective agenda for doing institutional analysis and forstudying the impact of institutions on the performance of societies ndashinnovativeness being one such subject If we can develop such instituteswe will increase the prospects for transforming our universities Thereis reason to be optimistic that once an eclectic group of scholars workwith a common set of concepts we can have genuine theoretical advanceif researchers can bring their diverse expertise to bear on a common setof problems There is considerable evidence that institutional analysisoffers high potential for advancing our knowledge of innovations andtechnological change (Hollingsworth and Hollingsworth 2000) If we aregoing to make substantial theoretical advance it is not likely to occurfrom the vantage point of a multiplicity of disciplines Rather it is likelyto occur as we develop a new intellectual framework and our own acad-emic journals This is a tall agenda which requires us to reassess theway that knowledge is presently produced in our societies

NOTE

This article is part of a much larger agenda involving the study of institutionsorganizations and innovations in which I am involved My debts are many andvaried The rst version was written while I was in residence in the ResearchUnit for Institutional Change and European Integration of the Austrian Academyof Sciences in Vienna in the summer of 1998 I am especially grateful to ProfessorsEgon Matzner and Sonja Puntscher-Riekmann for providing the stimulating envi-ronment in which to think through the ideas developed here Egon Matzner hasengaged in very stimulating discussions with me about the importance of insti-tutions in the general eld of socio-economics The nal version of the articlewas completed during my residence as a fellow at the Neurosciences Institutein La Jolla California I also wish to thank my colleagues in the InnovationTheme Project at the Netherlands Institute for Advanced Study (NIAS) during1998ndash9 and Edgar Grande and Karl Muumlller for stimulating conversations aboutthe interaction of institutions and organizations in facilitating particular typesof innovation I am especially grateful to Jerald Hage for helping me to developa number of issues in the article and for his useful comments on the rst draftFrans van Waarden early on helped me to work through the idea of multiplelevels of institutional analysis Others who have helped to develop the ideas inthe article have been my colleagues in the research project lsquoComparative socialsystems of productionrsquo especially Tom Burns Christel Lane Yoshitaka OkadaWolfgang Streeck and Richard Whitley My colleagues Robert Boyer SteveCasper and David Soskice have taught me much about how incentives providedby the institutional environments of rms inuence individual decision making

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

635

But my most important debt is to Ellen Jane Hollingsworth who has repeatedlyquestioned every single assumption of this article Without her rich and contin-uous tutoring this article would be even less developed than it is at presentFinally without the insightful assistance of David Gear in many phases of thisarticle it would not have been completed

REFERENCES

Aoki Masahiko (1990) lsquoTowards an economic model of the Japanese rmrsquo Journalof Economic Literature March 1ndash27

Arora A Landau Ralph and Rosenberg Nathan (eds) (1998) Chemicals and Long-Term Economic Growth Insights from the Chemical Industry New York JohnWiley amp Sons

Arrow Kenneth (1974) The Limits of Organization New York NortonArthur W B (1988a) lsquoSelf-reinforcing mechanisms in economicsrsquo in P W

Anderson K J Arrow and D Pines (eds) The Economy as an Evolving ComplexSystem Redwood City Calif Addison-Wesley

mdashmdash(1988b) lsquoCompeting technologies an overviewrsquo in G Dosi C Freeman RNelson G Silverberg and L Soete (eds) Technical Change and Economic TheoryLondon Pinter Publishers pp 590ndash607

Audretsch David B (1995) Innovation and Industry Evolution Cambridge MassMIT Press

Audretsch David B and Feldman Maryann P (1995) lsquoInnovative clusters andthe industry life cyclersquo Discussion Paper Wissenschaftszentrum BerlinMarket Processes and Corporate Development Research Unit FS-IV-95-7

Baum Joel A C and Oliver Christine (1992) lsquoInstitutional embeddedness andthe dynamics of organizational populationsrsquo American Sociological Review 57540ndash59

Baum Joel A C and Singh Jitendra V (1994a) lsquoOrganizationndashenvironment coevo-lutionrsquo in Joel A C Baum and Jitendra V Singh (eds) The EvolutionaryDynamics of Organizations New York Oxford University Press pp 379ndash402

mdashmdash(eds) (1994b) The Evolutionary Dynamics of Organizations New York OxfordUniversity Press

Berger Suzanne and Dore Ronald (eds) (1996) National Diversity and GlobalCapitalism Ithaca NY Cornell University Press

Boyer Robert (1997) lsquoThe variety and unequal performance of really existingmarkets farewell to Doctor Panglossrsquo in J Rogers Hollingsworth and RobertBoyer (eds) Contemporary Capitalism the Embeddedness of InstitutionsCambridge and New York Cambridge University Press pp 55ndash93

Burns Tom R and Dietz Thomas (2001 forthcoming) lsquoSocial rule system theorysocial action institutional arrangements and evolutionary processesrsquo in JRogers Hollingsworth and Karl Muumlller (eds) Advances in SocioeconomicsLanham Md Rowman amp Littleeld

Burns Tom R and Flam Helena (1987) The Shaping of Social Organization BeverlyHills Calif Sage

Calvert Randall L (1995) lsquoThe rational choice theory of social institutionsrsquo inJeffrey S Banks and Eric A Hanuskek (eds) Modern Political Economy NewYork Cambridge University Press pp 216ndash66

Camic Charles (1986) lsquoThe matter of habitrsquo American Journal of Sociology 911039ndash87

Campbell John L (1997) lsquoMechanisms of evolutionary change in economicsgovernance interaction interpretation and bricolagersquo in Lars Magnusson

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

636

Fifth level of analysis

We come next to the outputs and performance of the various institutionalcomponents of a society It is at this level that institutional componentsare more pragmatic and exible For example in the legal sector thereare specic statutes and court rulings in the state sector there are specicpolicies in the business sector there are new products new technolo-gies and market strategies It is at this nal level of analysis thatinstitutional spheres are most open and susceptible to change and cross-national mimicry is easiest and most commonThrough the outputs of the society we can obtain some assessment as

to how well a society is performing Moreover we can assess how inno-vative it is how egalitarian it is in the distribution of its resources howegalitarian it is in terms of levels of health education etc Just as it iscomplicated to measure how egalitarian a society is similarly it is dif-cult to assess how innovative it is For example societies vary in makingincremental and radical breakthroughs in basic and applied science in developing totally new products and new kinds of organizations inincremental and process innovations in existing products and organiza-tions in developing new and different forms of marketing ndash bothdomestically and globally (Hage and Hollingsworth 2000) Of coursesocieties vary in their rates of economic growth in their rates of economicproductivity in their quality of life (eg rates of crime life expectancyetc) All of these performance criteria feed back and inuence each ofthe levels of institutional analysis discussed above ndash rules norms valuesetc institutional arrangements institutional sectors the structure andform of organizations Moreover performance measures may inuenceeach other For example if a society has very low rates of growth itmay not be very innovative in some of the types of innovative activitiesmentioned above Good or poor performances of certain kinds inuenceother performance indicators (Hollingsworth et al 1990)Different institutional arrangements and different social systems of

production result in different types of economic performance Hence solong as societies have different social systems of production there areserious constraints on the degree to which they can converge in theirinnovative styles Different social systems of production tend to maxi-mize in a more or less explicit manner different performance criteriausually mixing considerations about static and dynamic efciency protsecurity social peace and economic andor political power In short incontrast to the implications of neoclassical economic theory in real-worldeconomies there are no universal standards all economically rationalactors attempt to maximize Economic history provides numerous exam-ples of how a great variety of principles of rationality are implementedin different societies

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

622

A critical question in institutional analysis is whether an existing socialsystem of production which supports a set of routines for a particularkind of technology and industry can shift from old practices to newones There are numerous historical accounts of how the social systemof production in a particular society which worked so well for a numberof years could not adapt to new technologies (Lazonick 1990 Veblen1915 Nelson 1994) Schumpeter (1983) and Freeman (1991) have devel-oped the idea of a techno-economic paradigm which suggests thatdifferent technologies require specic forms of organizational arrange-ments ndash or what I called social systems of production Because oflsquoinstitutional inertiarsquo the social system of production of a society maynot be able to adapt to a new techno-economic paradigm (Nelson 199458 Lanzara 1998)Whether a social system of production can sustain its performance

standards depends not only on its intrinsic economic rationality but alsoon where it ts into a larger system If a particular social system ofproduction is immune from the innovativeness and competition of analternative system survival can be long lasting But if different socialsystems of production with diverging criteria of good economic perfor-mance meet in the world arena the arbitrariness of nationally imposedconstructed performance standards may be superseded by alternativeperformance criteria as a result of international competitivenessThe world economy is also socially constructed just as are national

economies Even if different social systems of production are competingin the international arena it is not always possible to determine whichis more competitively effective at any moment in time Hegemonicnation-states in the short run shape the rules of trade to favor their indus-trial sectors and rms But the history of hegemonic powers suggeststhat in the longer run such social systems of production sustainedlargely by military and political power eventually give way to moreinnovative and competitive social systems of production In our ownday as nation-states are increasingly integrated into a world economyeven hegemonic powers lose their innovative and competitive advan-tage and their share of world output decreases Such a country mayattempt to restructure its institutional arrangements and to readjust itsperformance preferences But to restructure its system of innovationgenerally calls for a major redistribution of power within a societyLargely for this reason societies have historically had limited capacityto reconstruct a system of innovation in the image of their major competi-tors Each countryrsquos social system of production is a conguration of ahost of norms rules and values as well as of institutional arrangementsEach system is constantly changing and is open to inuence from othersystems And indeed many technologies and practices diffuse from onesociety to another but a societyrsquos capacity to be innovative is constrained

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

623

by the existing social system of production Thus the same technologymay exist in numerous countries but how it is employed and how itinuences societal outcomes and performances varies from one institu-tional conguration or society to another (Hollingsworth 1998)

PERSPECTIVES ON INSTITUTIONAL CHANGE

A fundamental problem which remains unresolved in institutionalanalysis involves the nature of institutional change There is much confu-sion and miscommunication in the social sciences about institutionalchange in large part because we do not even know how to measure therate of institutional change Of course we know that there are bothexternal and internal focuses for change and we have had a good bitof scholarship on this subject What is not clearly understood is thatinstitutions and institutional arrangements within societies are histori-cally rooted that there is a great deal of path-dependency to the waythat institutions evolve Moreover the more intricately linked each sectoris with each other sector and with a societyrsquos rules and norms the lesschoice actors have to devise new institutions and institutional arrange-ments On the other hand if a sector or actor is too isolated it may betoo weak within a society to bring about any kind of effective changeThese considerations are very critical to the problem of how muchfreedom actors have to create new institutions afresh to what extentinstitutions diffuse from one society to another (Murmann 1998 Lanzara1998) This is a fundamental question for students of institutionalanalysis as we attempt to understand how the different institutionallevels are linked togetherThere are a variety of reasons why there is confusion in studying insti-

tutional change There is disagreement over the extent to which actorshave the freedom to build new institutional arrangements and if theydo to what extent the new arrangements and organizational patternsmay depart from past practices Among some institutional economists(Williamson 1985) there is a basic assumption that actors build efcientinstitutions by pursuing their self-interest and by promoting theirstrategic goals Moreover there is the assumption that a mixing of fullrationality with market competition tends to produce all of the optimalinstitutions that are needed in order to coordinate a complex capitalisteconomy (Hollingsworth and Boyer 1997 ch 14 Hodgson 1999) Incontrast the argument developed here is much more complexSocial institutions are historically rooted and there is a great deal of

path-dependency in the way that various institutional componentsevolve The shape of institutional congurations at any moment limitsthe type of options for change Because there is a great deal of institu-tional inertia radical change in the institutional components of a society

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

624

is uncommon (Lanzara 1998) Recognizing the constraints existinginstitutions exert on actors we note that existing congurations of insti-tutional arrangements institutional sectors and organizational propertiescan limit but also inuence the degree to which norms and rules canchange Many scholars (North 1990 Ostrom 1990) have emphasized therole of norms rules ideologies and values in limiting the ability of actorsto develop new institutional arrangements which radically depart fromexisting ones However there is a two-way process of interaction aprocess of co-evolution (Baum and Singh 1994a Nelson 1994) Thus asFriedland and Alford (1991 244) argue actors develop new institutionalarrangements by recognizing organizational failures and low perfor-mance but these new arrangements do not necessarily change a societyrsquosnorms rules habits and its underlying value systemCampbell (1997) advances our understanding of how existing institu-

tions enable actors to construct radically new institutions He recognizesthat most institutions and institutional arrangements embody a degreeof inertia from which it is difcult for actors to depart Moreover thedifferential power relations among actors make it unlikely that lesspowerful actors can change the existing power structure of a societyMost change in the institutional components of a society evolves througha process of constrained selection which reects to a considerable degreethe existing arrangements and power relations This becomes the inter-pretative frame for social actors with institutionalized scripts and ritualswhich tend to be taken for granted and appear to be quite ordinary andnatural This institutionalized scripting of the social world is an extremelyimportant source of social stability and inertiaCampbell has developed a very fruitful explanation of change in terms

of the various institutional components outlined here He distinguishesbetween incremental and radical institutional change and argues thatradical institutional change occurs when social actors with widelydiffering norms cultural scripts and rituals for action engage in intenseinteraction with each other (see also Knight 1992) The more diverse theinteractions the greater the potential for institutional change He arguesthat as shifts in the composition of interaction among social actors occurchanges in the interpretation of problems and interests will follow Ifthere are only minor changes in the extent to which diverse actorsinteract there will be only modest institutional change But the morefundamental the changes in the interaction of diverse social actors thegreater the change in the way that actors interpret their world And thegreater the change in the way that actors interpret their world the morelikely that radical changes in various institutional components willfollow In other words diverse decision making tends to create a widerrange of interpretative frameworks of the social world than is likely tobe the case if social actors operate in isolation

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

625

This perspective of diverse interactions among social actors facilitatesour ability to move beyond the traditional view that existing institutionsconstrain the range of institutional alternatives which actors face Theolder view has tended to suggest that actors engage in institutionalchange by extending existing institutional principles habits and conven-tions Specifying the conditions which facilitate fundamental institutionalchange permits a much richer perspective for institutional analysisCampbellrsquos perspective about the conditions under which fundamentalor radical change in institutional components occurs is consistent withmy own research about the conditions under which major discoveriesor fundamental new knowledge occur the more scholars with diversebackgrounds interact with intensity and frequency the greater the like-lihood that they will develop new and alternative ways of thinking(Hollingsworth and Hollingsworth 2000)These views of institutional change are also consistent with the recent

work of Nelson and his colleagues (Mowery and Nelson 1999 Murmann1998) They report the research on a variety of industrial sectors in anumber of different countries revealing that networks of actors in partic-ular industries business and professional associations universityresearch departments and institutes are frequently able to use their collec-tive power to modify a societyrsquos property rights research system andgovernmental policies It is the intense and frequent interaction of thesediverse complex sets of actors ndash sometimes constituting a technologysystem ndash that often leads to incremental changes in a societyrsquos total socialsystem of production It is this kind of collective action that brings aboutorganizational change change in institutional sectors institutionalarrangements and even in the rules and norms of a society And if weare to understand how the style of innovativeness of a society changeswe must rst understand the changes in its institutional makeup It ishighly unlikely that a society will develop a new style of innovativenesswithout changing its institutional structure

FIRMS THEIR INSTITUTIONAL ENVIRONMENT ANDNATIONAL SYSTEMS OF INNOVATION

Variation in the innovative style of societies is shaped by their insti-tutional makeup ndash the various levels which are discussed aboveinstitutions institutional arrangements the structure and cohesion of institutional sectors which constitute a societyrsquos social system of pro-duction and the structure and culture of its organizations ndash especiallyits business rms and other research organizations By undertaking aninstitutional analysis of a society one can begin to understand in whatkinds of organizations the production of specic kinds of knowledgetakes place and how this is linked to particular kinds of innovativeness

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

626

With such a perspective one can gain a rich understanding of why somesocieties excel in the production of radical breakthroughs in basic andapplied science and in developing radically new products and why othersocieties excel in more incremental innovationsReecting on these issues one cannot help being impressed with the

contrasts between the innovative styles of German and American orga-nizations For example Germany has continued to be very successful in making incremental innovations in many industries in which it wasvery competitive well before World War II paper printing materialsmachinery electrotechnical products motor vehicles chemicals textileyarns and fabrics But the Germans have been much less successful indeveloping totally new products or related innovations in biotechnologyelectronics telecommunications aircraft as well as other newer indus-trial areas In contrast the Americans have been very innovative in these and other industries with products having very short lives andwith technologies which change very rapidly and are very complex(Hollingsworth 1991 Soskice 1997 1999)The importance of institutions and innovativeness for international

competitiveness has led to the recognition that there are national systemsof innovation (Lundvall 1992 Nelson 1993 Edquist 1997 Hage andHollingsworth 2000) But even before the recent studies of nationalsystems of innovation Landes (1969) was a proponent of these viewswhen he made the argument that Germany had developed a nationalsystem of innovation by building on its education and scientic researchsystems which were lacking in Britain and elsewhere However the concept of a national system of innovation immediately poses theproblem of what kind of innovations should be consideredDavid Soskice an economist at the Wissenschaftszentrum Berlin has

addressed a set of problems somewhat complementary to the perspec-tive raised here His work (Soskice 1990 1997 1999) suggests that weare sensitive to the institutional environment in which the followingactors are embedded the employees of rms the owners and nanciersof rms the competitors of rms and nally their collaborators ndash aperspective complementary to Hollingsworthrsquos work on social systemsof production (Hollingsworth 1997 1998) When one reects on the insti-tutional contextsocial system of production within which these actorsare embedded one becomes aware that those countries which have some-what inexible labor markets external to rms but a business system in which rms have long-term commitments to their employees arestrongly associated with incremental but hardly at all with radical inno-vations In the rms of such countries (eg Germany and Japan) thereis much more consensus decision making than in societies where a lsquohireand rersquo set of practices is quite pervasive And this kind of consensusdecision making ndash while highly conducive to incremental innovativeness

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

627

ndash limits the capacity for radical innovations to occur This kind of lsquolongtermismrsquo set of norms and rules occurs not only in the employmentpractices of incrementally innovative rms but also in their nancingand ownership patterns (Hollingsworth 1997) During the past thirtyyears these were characteristics of many rms in Germany and JapanSignicantly in both of these countries venture capital markets werenot highly developed nancing tended to be based more on long-termbank loans than on the equity markets and the relationships of rmstended to be relatively stable And these stable institutional arrange-ments also contributed to a high degree of innovativeness ndash butincremental rather than radical innovationsMeantime we should also pursue Soskicersquos ideas (1999) about how

rmsrsquo relationships with their competitors and collaborators shape theirstyles of innovativeness In those advanced industrial societies in whichbusiness rms are highly mobilized into business associations theretends to be a high degree of standardized quality controls over technicalnorms and a high degree of reliance on legal sanctions as monitoringinstruments vis-agrave-vis association members As a result of this high degreeof integration of rms into associations member rms are more likelyto be associated with incremental innovations than is likely to be the case with rms in those societies where associations are less welldeveloped In the latter type of society radical innovations are morelikely to occur In those advanced industrial societies where rms havelong-term stable relationships with collaborators in product design andproduction the style of innovation is likely to be incremental And it is in those societies in which there is a weaker tradition of strongcollaboration among rms that radical innovations are more likely tooccurObviously even in a country where there may be a great deal of

radical innovativeness most innovations will be of an incremental typefor there are simply fewer radical innovations But Germany and therest of Europe not only have had relatively few radical innovations incomparison with the United States but they also have lagged behindthe USA in developing relatively totally new industries Thus there hasbeen much less innovative activity in Europe in high-technology andnewly emerging industries than in the USAIn complementary research Audretsch and Feldman (1995) have

demonstrated that the USA has been more successful in spawning rmsengaged in the early stages of the industry life cycle but relative toGermany had an environment unfavorable to innovative activity byestablished rms especially in more mature industries German rmshave tended to be associated with a routinized technological regime ndashwhere established enterprises have had strong incentives to engage inincremental innovations but new rms have tended to have fundamental

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

628

innovative disadvantages relative to the Americans (see Audretsch 1995Winter 1982 Casper 2000)On both sides of the Atlantic universities engage in basic and applied

research But signicantly the institutional environment in which univer-sities are embedded inuences the degree to which the research is likelyto represent major breakthroughs or is likely to have more modestimplications Recent research in biomedical science demonstrates that inthose societies in which university professors andor graduate studentshave the following characteristics there are likely to be relatively majorresearch breakthroughs graduate students have a long training periodhigh dependency on a major professor low autonomy in dening thenature of a research project restricted job options and a university envi-ronment where there is a very modest or nonexistent entrepreneurialculture For example in Germany ndash a country with relatively few radicalbreakthroughs ndash the scholar who nishes a habilitation will have spentmany years on a research project before qualifying for a permanent acad-emic position will have been in a high-dependency position with a seniorprofessor will have had limited autonomy to dene a research projectand will be constrained from pursuing a high-risk research project Theyoung German will have an extraordinarily high level of training butwill have faced a number of disincentives to undertake high-risk researchAnd as soon as heshe has completed the habilitation and becomes aprofessor administrative responsibilities constitute another disincentivefrom engaging in high-risk or long-term research projects Moreover theGerman professor is a civil servant with a more stable and secure incomethan hisher American counterpart In a very different institutionalenvironment American professors in large research universities tend tobecome much more entrepreneurial Today many American professorsbehave as though they were operating their own rms within univer-sities and thus devote considerable time to generating funds for researchprojects assistants secretaries supplies etc And the variation in thisentrepreneurial spirit across countries is associated with variation in thetaking of high-risk research strategies and the making of major researchbreakthroughsIn sum societies vary in terms of the institutional environment in

which their organizations are embedded and this variation inuencesthe degree to which they have industrial activities in the early or latestages of the industry life cycle successful new or mature industriesand radical vs incremental innovations And it is this variation whichis encouraging rms in countries with relatively few new industries and with few or no radical innovations to invest in those countries with many new industries and numerous radical innovations ndash as astrategy to be able to transfer new technologies from the latter institu-tional environment to the former

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

629

Table 4 summarizes the way that the norms rules and values of a society are associated with Americarsquos and Germanyrsquos institutionalarrangements their social system of production and their styles of inno-vativeness

CONCLUSION

Several themes deserve to be reemphasized Most importantly institu-tional analysis is emerging as a dominant issue in several social sciencedisciplines The study of various institutional components discussedabove is a complex subject requiring multiple levels of analysis and aknowledge of multiple sectors within societies It requires a theoreticalperspective that evolves from an empirical analysis in different societiesIf we are to advance the study of institutional analysis we need cross-disciplinary collaboration and team research which are not readilyachieved given the structure of most of our research universitiesAt present scholars in each academic discipline have their own distinc-

tive strategies for studying institutional analysis Thus economists tendto use several approaches (North 1990 Williamson 1985 Hodgson 19881998 1999) anthropologists another (Geertz 1995) and sociologistspolitical scientists and historians employ other strategies (Campbell etal 1991 Powell and DiMaggio 1991 Katznelson 1998 Hall and Taylor1996) With practitioners in various disciplines pursuing their distinctiveapproaches to institutional analysis there is low potential for collabo-rative institutional analysis across the social sciences The lessons fromthe biological sciences over the last 40 years are very instructive oncepractitioners from various biological disciplines began to cooperate instudying biological phenomena by using the same concepts at the mole-cular level theoretical advance was very rapid Comparable phenomenatook place in the development of other hybrid elds eg biochemistrybiophysics (Judson 1979 Olby 1979 Kohler 1982 1994) Similarly weresocial scientists in various disciplines to work collaboratively in the areaof institutional analysis there would very likely be an acceleration intheir theoretical knowledgeThis is not a plea that everyone engaged in institutional analysis should

do the same kind of research Indeed just as biologists working at themolecular level study many different kinds of problems similarly socialscientists engaged in institutional analysis would also work on manyseparate problems but in the spirit of a collective enterprise Some wouldconcentrate their attention on the study of rules norms habits conven-tions and values while others would study how these are associatedwith congurations of various institutional arrangements (types ofmarkets hierarchies networks associations communities clans statesetc) Others would work on specic institutional sectors (eg education

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

630

Table 4

Rules norms hab

its an

d values associated w

ith sp

ecic co

ngu

ration

s of coo

rdinating mecha

nism

s soc

ial sy

stem

s of

prod

uction

an

d nationa

l styles of inno

vation

USA

Germany

Norms rules values which provide incentives for

Norms rules values which provide incentives for

short-term horizons

long-term horizons

Con

gu

rative

forms of

Marke

ts c

orpo

rate hierarchies reg

ulatory state

Assoc

iation

s (busine

ss assoc

iation

s la

bor union

s)

coordinating

corp

oratist-type

networks

marke

ts c

orporatist-

mecha

nism

s (listed in

type

state

order of im

portanc

e in

the particu

lar social

system

)

Social sys

tem of

prod

uction

Bus

iness system

Firm

struc

tures

Con

glom

erates b

ut with mov

emen

t toward

Firm

more co

ngru

ent with pa

rticular prod

ucts

elim

ination of produc

tion

in som

e sectors

emph

asis on long

-term strateg

ies rm

s are well

emph

asis on sh

ort-term

strateg

ies high

integrated

differentiation

between divisions

of rms

Firm

s mob

ilized in

Low

High

and in

tegrated

into busine

ss

associations

Indu

strial relations

Low

deg

ree of jo

b secu

rity rigid in

ternal lab

or

High deg

ree of jo

b secu

rity

exible interna

l lab

or

system

marke

t ex

ible externa

l labo

r marke

tsmarke

t rigid externa

l lab

or m

arke

tPatterns of

Dispersed owne

rship frequ

ent turnov

erMore stab

le owne

rship patterns

owne

rship

Table 4

continue

d

USA

Germany

Norms rules values which provide incentives for

Norms rules values which provide incentives for

short-term horizons

long-term horizons

Social system of

Prod

uction

Bus

iness system

Training for labo

r an

dHigh deg

ree of voluntary decision mak

ing by

Collective an

d m

ore co

mpulso

ry d

ecisions

abo

ut

man

agem

ent

individuals as to ho

w m

uch

tim

e an

d ene

rgy

how m

uch ed

ucation

is ne

eded

he

avy em

pha

sis

to inv

est in educ

ation heavy

emph

asis on

on techn

ical training (eg en

gine

ering)

busine

ss sch

ools training

in marke

ting

and

sales accou

nting

Fina

ncial marke

tsEqu

ity an

d ven

ture cap

ital m

arke

ts highly

Equ

ity an

d ven

ture cap

ital m

arke

ts w

eakly

dev

elop

ed na

ncing from

cen

tral gov

ernm

ent

deve

lope

d he

avy em

pha

sis on

corporate loan

sto priva

te rm

s for dev

elop

men

t of new

tech

nologies related

to na

tion

al security

Unive

rsity research

High individual auton

omy high poten

tial for

Low

ind

ividua

l au

tono

my until almost 40

yea

rs of

system

individual creativity u

nive

rsity pr

ofesso

rs

age poten

tial for ind

ividual creativity high

ly

high

ly entreprene

urial

cons

traine

d unive

rsity pr

ofessors not so

entrep

rene

urial

Inno

vative

styles

Highly su

ccessful in m

ajor break

through

s in

High pe

rforman

ce in proc

ess an

d inc

remen

tal

basic an

d app

lied scien

ce go

od in dev

elop

ing

inno

vation

s in existing produ

cts not so su

ccessful

totally

new

lines of pr

oducts an

d in m

ajor break

through

s in basic scien

ce or in

tech

nologies high

deg

ree of entrepr

eneu

rship

deve

loping

totally new

lines of prod

ucts of

amon

g research

ers

tech

nologies low d

egree of entrepr

eneu

rship

amon

g research

ers

Table 4

continue

d

USA

Germany

Norms rules values which provide incentives for

Norms rules values which provide incentives for

short-term horizons

long-term horizons

Prod

ucts in w

hich

Pr

oducts with sh

ort liv

es prod

ucts in w

hich

Older produ

cts with long

history machine

ry pap

er

society excels

tech

nology

is bo

th highly co

mplex

and

pr

inting

textile

s m

otor veh

icles m

achine

too

lsch

ange

s rapidly en

tertainm

ent indus

try

pha

rmaceu

ticals biotechn

olog

y softw

are

aerosp

ace

business systems nancial markets and systems of research) Distinctivewould be the recognition of how these separate studies are linkedtogether how each institutional sector is linked to a societyrsquos normsrules values etc and its conguration of institutional arrangementsSimilarly other institutional analysts might focus on the study of orga-nizations with a major concern for how the institutional environmentof organizations (eg norms rules values congurations of institutionalarrangements institutional sectors) inuenced the structure culture andoutcomes of organizations And nally others might study how all ofthe aforementioned aspects of institutional analysis inuenced a societyrsquosoverall performance Nelson and Sampat (1998) have appropriatelyreminded us that we should engage in institutional analysis in order toaddress specic problems but the view developed in this article is thatwe rst need some mapping of the terrain of institutional analysis inorder for our diverse enterprises to become a meaningful collective enter-prise especially if we are to understand how institutions inuence theinnovative processThe argument developed here is not that all social scientists should engage

in institutional analysis Obviously there are numerous other importantareas of research which lie beyond the eld of institutional analysis Nordoes this article argue that institutional analysis is the most importantsubject for social scientistsSince the structure and culture of our research universities reect an

enormous amount of disciplinary fragmentation what are the prospectsfor promoting the kind of collective and interdisciplinary research agendaproposed above In one respect the prospects are not encouragingDuring the past half century the efforts of American universities to have effective communication across the biological sciences have variedgreatly in their success At the University of California-Berkeley it was not until the 1980s that there was a serious effort to integrate thebiological sciences and at that time the campus abolished seventeendepartments to promote an integrated biological program But at theUniversity of Wisconsin the biological sciences are still very fragmentedinto dozens of different departments in botany plant pathology zoologyand physiology with two departments of biochemistry and two depart-ments of geneticsThere have been other approaches which have worked very well histor-

ically for promoting scientic integration amidst scientic diversity Forexample several summer schools and special institutes were organizedin the United States during the 1940s in an effort to overcome the frag-mentation in the biological sciences These programs made it possiblefor scientists in diverse academic disciplines and from different univer-sities and countries to come together to engage in intense and frequentinteraction Perhaps the most effective program was that held in the

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

634

summers at Cold Spring Harbor New York where a research agendafor the study of molecular biology was developed (Fischer and Lipson1988 Stent and Watson 1966) Scientists went there summer aftersummer and when they returned to their universities many began therestructuring of the biological sciences in their universitiesIn our own day we need summer institutes and other similar programs

so that scholars from diverse backgrounds can come together anddevelop a collective agenda for doing institutional analysis and forstudying the impact of institutions on the performance of societies ndashinnovativeness being one such subject If we can develop such instituteswe will increase the prospects for transforming our universities Thereis reason to be optimistic that once an eclectic group of scholars workwith a common set of concepts we can have genuine theoretical advanceif researchers can bring their diverse expertise to bear on a common setof problems There is considerable evidence that institutional analysisoffers high potential for advancing our knowledge of innovations andtechnological change (Hollingsworth and Hollingsworth 2000) If we aregoing to make substantial theoretical advance it is not likely to occurfrom the vantage point of a multiplicity of disciplines Rather it is likelyto occur as we develop a new intellectual framework and our own acad-emic journals This is a tall agenda which requires us to reassess theway that knowledge is presently produced in our societies

NOTE

This article is part of a much larger agenda involving the study of institutionsorganizations and innovations in which I am involved My debts are many andvaried The rst version was written while I was in residence in the ResearchUnit for Institutional Change and European Integration of the Austrian Academyof Sciences in Vienna in the summer of 1998 I am especially grateful to ProfessorsEgon Matzner and Sonja Puntscher-Riekmann for providing the stimulating envi-ronment in which to think through the ideas developed here Egon Matzner hasengaged in very stimulating discussions with me about the importance of insti-tutions in the general eld of socio-economics The nal version of the articlewas completed during my residence as a fellow at the Neurosciences Institutein La Jolla California I also wish to thank my colleagues in the InnovationTheme Project at the Netherlands Institute for Advanced Study (NIAS) during1998ndash9 and Edgar Grande and Karl Muumlller for stimulating conversations aboutthe interaction of institutions and organizations in facilitating particular typesof innovation I am especially grateful to Jerald Hage for helping me to developa number of issues in the article and for his useful comments on the rst draftFrans van Waarden early on helped me to work through the idea of multiplelevels of institutional analysis Others who have helped to develop the ideas inthe article have been my colleagues in the research project lsquoComparative socialsystems of productionrsquo especially Tom Burns Christel Lane Yoshitaka OkadaWolfgang Streeck and Richard Whitley My colleagues Robert Boyer SteveCasper and David Soskice have taught me much about how incentives providedby the institutional environments of rms inuence individual decision making

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

635

But my most important debt is to Ellen Jane Hollingsworth who has repeatedlyquestioned every single assumption of this article Without her rich and contin-uous tutoring this article would be even less developed than it is at presentFinally without the insightful assistance of David Gear in many phases of thisarticle it would not have been completed

REFERENCES

Aoki Masahiko (1990) lsquoTowards an economic model of the Japanese rmrsquo Journalof Economic Literature March 1ndash27

Arora A Landau Ralph and Rosenberg Nathan (eds) (1998) Chemicals and Long-Term Economic Growth Insights from the Chemical Industry New York JohnWiley amp Sons

Arrow Kenneth (1974) The Limits of Organization New York NortonArthur W B (1988a) lsquoSelf-reinforcing mechanisms in economicsrsquo in P W

Anderson K J Arrow and D Pines (eds) The Economy as an Evolving ComplexSystem Redwood City Calif Addison-Wesley

mdashmdash(1988b) lsquoCompeting technologies an overviewrsquo in G Dosi C Freeman RNelson G Silverberg and L Soete (eds) Technical Change and Economic TheoryLondon Pinter Publishers pp 590ndash607

Audretsch David B (1995) Innovation and Industry Evolution Cambridge MassMIT Press

Audretsch David B and Feldman Maryann P (1995) lsquoInnovative clusters andthe industry life cyclersquo Discussion Paper Wissenschaftszentrum BerlinMarket Processes and Corporate Development Research Unit FS-IV-95-7

Baum Joel A C and Oliver Christine (1992) lsquoInstitutional embeddedness andthe dynamics of organizational populationsrsquo American Sociological Review 57540ndash59

Baum Joel A C and Singh Jitendra V (1994a) lsquoOrganizationndashenvironment coevo-lutionrsquo in Joel A C Baum and Jitendra V Singh (eds) The EvolutionaryDynamics of Organizations New York Oxford University Press pp 379ndash402

mdashmdash(eds) (1994b) The Evolutionary Dynamics of Organizations New York OxfordUniversity Press

Berger Suzanne and Dore Ronald (eds) (1996) National Diversity and GlobalCapitalism Ithaca NY Cornell University Press

Boyer Robert (1997) lsquoThe variety and unequal performance of really existingmarkets farewell to Doctor Panglossrsquo in J Rogers Hollingsworth and RobertBoyer (eds) Contemporary Capitalism the Embeddedness of InstitutionsCambridge and New York Cambridge University Press pp 55ndash93

Burns Tom R and Dietz Thomas (2001 forthcoming) lsquoSocial rule system theorysocial action institutional arrangements and evolutionary processesrsquo in JRogers Hollingsworth and Karl Muumlller (eds) Advances in SocioeconomicsLanham Md Rowman amp Littleeld

Burns Tom R and Flam Helena (1987) The Shaping of Social Organization BeverlyHills Calif Sage

Calvert Randall L (1995) lsquoThe rational choice theory of social institutionsrsquo inJeffrey S Banks and Eric A Hanuskek (eds) Modern Political Economy NewYork Cambridge University Press pp 216ndash66

Camic Charles (1986) lsquoThe matter of habitrsquo American Journal of Sociology 911039ndash87

Campbell John L (1997) lsquoMechanisms of evolutionary change in economicsgovernance interaction interpretation and bricolagersquo in Lars Magnusson

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

636

A critical question in institutional analysis is whether an existing socialsystem of production which supports a set of routines for a particularkind of technology and industry can shift from old practices to newones There are numerous historical accounts of how the social systemof production in a particular society which worked so well for a numberof years could not adapt to new technologies (Lazonick 1990 Veblen1915 Nelson 1994) Schumpeter (1983) and Freeman (1991) have devel-oped the idea of a techno-economic paradigm which suggests thatdifferent technologies require specic forms of organizational arrange-ments ndash or what I called social systems of production Because oflsquoinstitutional inertiarsquo the social system of production of a society maynot be able to adapt to a new techno-economic paradigm (Nelson 199458 Lanzara 1998)Whether a social system of production can sustain its performance

standards depends not only on its intrinsic economic rationality but alsoon where it ts into a larger system If a particular social system ofproduction is immune from the innovativeness and competition of analternative system survival can be long lasting But if different socialsystems of production with diverging criteria of good economic perfor-mance meet in the world arena the arbitrariness of nationally imposedconstructed performance standards may be superseded by alternativeperformance criteria as a result of international competitivenessThe world economy is also socially constructed just as are national

economies Even if different social systems of production are competingin the international arena it is not always possible to determine whichis more competitively effective at any moment in time Hegemonicnation-states in the short run shape the rules of trade to favor their indus-trial sectors and rms But the history of hegemonic powers suggeststhat in the longer run such social systems of production sustainedlargely by military and political power eventually give way to moreinnovative and competitive social systems of production In our ownday as nation-states are increasingly integrated into a world economyeven hegemonic powers lose their innovative and competitive advan-tage and their share of world output decreases Such a country mayattempt to restructure its institutional arrangements and to readjust itsperformance preferences But to restructure its system of innovationgenerally calls for a major redistribution of power within a societyLargely for this reason societies have historically had limited capacityto reconstruct a system of innovation in the image of their major competi-tors Each countryrsquos social system of production is a conguration of ahost of norms rules and values as well as of institutional arrangementsEach system is constantly changing and is open to inuence from othersystems And indeed many technologies and practices diffuse from onesociety to another but a societyrsquos capacity to be innovative is constrained

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

623

by the existing social system of production Thus the same technologymay exist in numerous countries but how it is employed and how itinuences societal outcomes and performances varies from one institu-tional conguration or society to another (Hollingsworth 1998)

PERSPECTIVES ON INSTITUTIONAL CHANGE

A fundamental problem which remains unresolved in institutionalanalysis involves the nature of institutional change There is much confu-sion and miscommunication in the social sciences about institutionalchange in large part because we do not even know how to measure therate of institutional change Of course we know that there are bothexternal and internal focuses for change and we have had a good bitof scholarship on this subject What is not clearly understood is thatinstitutions and institutional arrangements within societies are histori-cally rooted that there is a great deal of path-dependency to the waythat institutions evolve Moreover the more intricately linked each sectoris with each other sector and with a societyrsquos rules and norms the lesschoice actors have to devise new institutions and institutional arrange-ments On the other hand if a sector or actor is too isolated it may betoo weak within a society to bring about any kind of effective changeThese considerations are very critical to the problem of how muchfreedom actors have to create new institutions afresh to what extentinstitutions diffuse from one society to another (Murmann 1998 Lanzara1998) This is a fundamental question for students of institutionalanalysis as we attempt to understand how the different institutionallevels are linked togetherThere are a variety of reasons why there is confusion in studying insti-

tutional change There is disagreement over the extent to which actorshave the freedom to build new institutional arrangements and if theydo to what extent the new arrangements and organizational patternsmay depart from past practices Among some institutional economists(Williamson 1985) there is a basic assumption that actors build efcientinstitutions by pursuing their self-interest and by promoting theirstrategic goals Moreover there is the assumption that a mixing of fullrationality with market competition tends to produce all of the optimalinstitutions that are needed in order to coordinate a complex capitalisteconomy (Hollingsworth and Boyer 1997 ch 14 Hodgson 1999) Incontrast the argument developed here is much more complexSocial institutions are historically rooted and there is a great deal of

path-dependency in the way that various institutional componentsevolve The shape of institutional congurations at any moment limitsthe type of options for change Because there is a great deal of institu-tional inertia radical change in the institutional components of a society

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

624

is uncommon (Lanzara 1998) Recognizing the constraints existinginstitutions exert on actors we note that existing congurations of insti-tutional arrangements institutional sectors and organizational propertiescan limit but also inuence the degree to which norms and rules canchange Many scholars (North 1990 Ostrom 1990) have emphasized therole of norms rules ideologies and values in limiting the ability of actorsto develop new institutional arrangements which radically depart fromexisting ones However there is a two-way process of interaction aprocess of co-evolution (Baum and Singh 1994a Nelson 1994) Thus asFriedland and Alford (1991 244) argue actors develop new institutionalarrangements by recognizing organizational failures and low perfor-mance but these new arrangements do not necessarily change a societyrsquosnorms rules habits and its underlying value systemCampbell (1997) advances our understanding of how existing institu-

tions enable actors to construct radically new institutions He recognizesthat most institutions and institutional arrangements embody a degreeof inertia from which it is difcult for actors to depart Moreover thedifferential power relations among actors make it unlikely that lesspowerful actors can change the existing power structure of a societyMost change in the institutional components of a society evolves througha process of constrained selection which reects to a considerable degreethe existing arrangements and power relations This becomes the inter-pretative frame for social actors with institutionalized scripts and ritualswhich tend to be taken for granted and appear to be quite ordinary andnatural This institutionalized scripting of the social world is an extremelyimportant source of social stability and inertiaCampbell has developed a very fruitful explanation of change in terms

of the various institutional components outlined here He distinguishesbetween incremental and radical institutional change and argues thatradical institutional change occurs when social actors with widelydiffering norms cultural scripts and rituals for action engage in intenseinteraction with each other (see also Knight 1992) The more diverse theinteractions the greater the potential for institutional change He arguesthat as shifts in the composition of interaction among social actors occurchanges in the interpretation of problems and interests will follow Ifthere are only minor changes in the extent to which diverse actorsinteract there will be only modest institutional change But the morefundamental the changes in the interaction of diverse social actors thegreater the change in the way that actors interpret their world And thegreater the change in the way that actors interpret their world the morelikely that radical changes in various institutional components willfollow In other words diverse decision making tends to create a widerrange of interpretative frameworks of the social world than is likely tobe the case if social actors operate in isolation

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

625

This perspective of diverse interactions among social actors facilitatesour ability to move beyond the traditional view that existing institutionsconstrain the range of institutional alternatives which actors face Theolder view has tended to suggest that actors engage in institutionalchange by extending existing institutional principles habits and conven-tions Specifying the conditions which facilitate fundamental institutionalchange permits a much richer perspective for institutional analysisCampbellrsquos perspective about the conditions under which fundamentalor radical change in institutional components occurs is consistent withmy own research about the conditions under which major discoveriesor fundamental new knowledge occur the more scholars with diversebackgrounds interact with intensity and frequency the greater the like-lihood that they will develop new and alternative ways of thinking(Hollingsworth and Hollingsworth 2000)These views of institutional change are also consistent with the recent

work of Nelson and his colleagues (Mowery and Nelson 1999 Murmann1998) They report the research on a variety of industrial sectors in anumber of different countries revealing that networks of actors in partic-ular industries business and professional associations universityresearch departments and institutes are frequently able to use their collec-tive power to modify a societyrsquos property rights research system andgovernmental policies It is the intense and frequent interaction of thesediverse complex sets of actors ndash sometimes constituting a technologysystem ndash that often leads to incremental changes in a societyrsquos total socialsystem of production It is this kind of collective action that brings aboutorganizational change change in institutional sectors institutionalarrangements and even in the rules and norms of a society And if weare to understand how the style of innovativeness of a society changeswe must rst understand the changes in its institutional makeup It ishighly unlikely that a society will develop a new style of innovativenesswithout changing its institutional structure

FIRMS THEIR INSTITUTIONAL ENVIRONMENT ANDNATIONAL SYSTEMS OF INNOVATION

Variation in the innovative style of societies is shaped by their insti-tutional makeup ndash the various levels which are discussed aboveinstitutions institutional arrangements the structure and cohesion of institutional sectors which constitute a societyrsquos social system of pro-duction and the structure and culture of its organizations ndash especiallyits business rms and other research organizations By undertaking aninstitutional analysis of a society one can begin to understand in whatkinds of organizations the production of specic kinds of knowledgetakes place and how this is linked to particular kinds of innovativeness

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

626

With such a perspective one can gain a rich understanding of why somesocieties excel in the production of radical breakthroughs in basic andapplied science and in developing radically new products and why othersocieties excel in more incremental innovationsReecting on these issues one cannot help being impressed with the

contrasts between the innovative styles of German and American orga-nizations For example Germany has continued to be very successful in making incremental innovations in many industries in which it wasvery competitive well before World War II paper printing materialsmachinery electrotechnical products motor vehicles chemicals textileyarns and fabrics But the Germans have been much less successful indeveloping totally new products or related innovations in biotechnologyelectronics telecommunications aircraft as well as other newer indus-trial areas In contrast the Americans have been very innovative in these and other industries with products having very short lives andwith technologies which change very rapidly and are very complex(Hollingsworth 1991 Soskice 1997 1999)The importance of institutions and innovativeness for international

competitiveness has led to the recognition that there are national systemsof innovation (Lundvall 1992 Nelson 1993 Edquist 1997 Hage andHollingsworth 2000) But even before the recent studies of nationalsystems of innovation Landes (1969) was a proponent of these viewswhen he made the argument that Germany had developed a nationalsystem of innovation by building on its education and scientic researchsystems which were lacking in Britain and elsewhere However the concept of a national system of innovation immediately poses theproblem of what kind of innovations should be consideredDavid Soskice an economist at the Wissenschaftszentrum Berlin has

addressed a set of problems somewhat complementary to the perspec-tive raised here His work (Soskice 1990 1997 1999) suggests that weare sensitive to the institutional environment in which the followingactors are embedded the employees of rms the owners and nanciersof rms the competitors of rms and nally their collaborators ndash aperspective complementary to Hollingsworthrsquos work on social systemsof production (Hollingsworth 1997 1998) When one reects on the insti-tutional contextsocial system of production within which these actorsare embedded one becomes aware that those countries which have some-what inexible labor markets external to rms but a business system in which rms have long-term commitments to their employees arestrongly associated with incremental but hardly at all with radical inno-vations In the rms of such countries (eg Germany and Japan) thereis much more consensus decision making than in societies where a lsquohireand rersquo set of practices is quite pervasive And this kind of consensusdecision making ndash while highly conducive to incremental innovativeness

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

627

ndash limits the capacity for radical innovations to occur This kind of lsquolongtermismrsquo set of norms and rules occurs not only in the employmentpractices of incrementally innovative rms but also in their nancingand ownership patterns (Hollingsworth 1997) During the past thirtyyears these were characteristics of many rms in Germany and JapanSignicantly in both of these countries venture capital markets werenot highly developed nancing tended to be based more on long-termbank loans than on the equity markets and the relationships of rmstended to be relatively stable And these stable institutional arrange-ments also contributed to a high degree of innovativeness ndash butincremental rather than radical innovationsMeantime we should also pursue Soskicersquos ideas (1999) about how

rmsrsquo relationships with their competitors and collaborators shape theirstyles of innovativeness In those advanced industrial societies in whichbusiness rms are highly mobilized into business associations theretends to be a high degree of standardized quality controls over technicalnorms and a high degree of reliance on legal sanctions as monitoringinstruments vis-agrave-vis association members As a result of this high degreeof integration of rms into associations member rms are more likelyto be associated with incremental innovations than is likely to be the case with rms in those societies where associations are less welldeveloped In the latter type of society radical innovations are morelikely to occur In those advanced industrial societies where rms havelong-term stable relationships with collaborators in product design andproduction the style of innovation is likely to be incremental And it is in those societies in which there is a weaker tradition of strongcollaboration among rms that radical innovations are more likely tooccurObviously even in a country where there may be a great deal of

radical innovativeness most innovations will be of an incremental typefor there are simply fewer radical innovations But Germany and therest of Europe not only have had relatively few radical innovations incomparison with the United States but they also have lagged behindthe USA in developing relatively totally new industries Thus there hasbeen much less innovative activity in Europe in high-technology andnewly emerging industries than in the USAIn complementary research Audretsch and Feldman (1995) have

demonstrated that the USA has been more successful in spawning rmsengaged in the early stages of the industry life cycle but relative toGermany had an environment unfavorable to innovative activity byestablished rms especially in more mature industries German rmshave tended to be associated with a routinized technological regime ndashwhere established enterprises have had strong incentives to engage inincremental innovations but new rms have tended to have fundamental

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

628

innovative disadvantages relative to the Americans (see Audretsch 1995Winter 1982 Casper 2000)On both sides of the Atlantic universities engage in basic and applied

research But signicantly the institutional environment in which univer-sities are embedded inuences the degree to which the research is likelyto represent major breakthroughs or is likely to have more modestimplications Recent research in biomedical science demonstrates that inthose societies in which university professors andor graduate studentshave the following characteristics there are likely to be relatively majorresearch breakthroughs graduate students have a long training periodhigh dependency on a major professor low autonomy in dening thenature of a research project restricted job options and a university envi-ronment where there is a very modest or nonexistent entrepreneurialculture For example in Germany ndash a country with relatively few radicalbreakthroughs ndash the scholar who nishes a habilitation will have spentmany years on a research project before qualifying for a permanent acad-emic position will have been in a high-dependency position with a seniorprofessor will have had limited autonomy to dene a research projectand will be constrained from pursuing a high-risk research project Theyoung German will have an extraordinarily high level of training butwill have faced a number of disincentives to undertake high-risk researchAnd as soon as heshe has completed the habilitation and becomes aprofessor administrative responsibilities constitute another disincentivefrom engaging in high-risk or long-term research projects Moreover theGerman professor is a civil servant with a more stable and secure incomethan hisher American counterpart In a very different institutionalenvironment American professors in large research universities tend tobecome much more entrepreneurial Today many American professorsbehave as though they were operating their own rms within univer-sities and thus devote considerable time to generating funds for researchprojects assistants secretaries supplies etc And the variation in thisentrepreneurial spirit across countries is associated with variation in thetaking of high-risk research strategies and the making of major researchbreakthroughsIn sum societies vary in terms of the institutional environment in

which their organizations are embedded and this variation inuencesthe degree to which they have industrial activities in the early or latestages of the industry life cycle successful new or mature industriesand radical vs incremental innovations And it is this variation whichis encouraging rms in countries with relatively few new industries and with few or no radical innovations to invest in those countries with many new industries and numerous radical innovations ndash as astrategy to be able to transfer new technologies from the latter institu-tional environment to the former

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

629

Table 4 summarizes the way that the norms rules and values of a society are associated with Americarsquos and Germanyrsquos institutionalarrangements their social system of production and their styles of inno-vativeness

CONCLUSION

Several themes deserve to be reemphasized Most importantly institu-tional analysis is emerging as a dominant issue in several social sciencedisciplines The study of various institutional components discussedabove is a complex subject requiring multiple levels of analysis and aknowledge of multiple sectors within societies It requires a theoreticalperspective that evolves from an empirical analysis in different societiesIf we are to advance the study of institutional analysis we need cross-disciplinary collaboration and team research which are not readilyachieved given the structure of most of our research universitiesAt present scholars in each academic discipline have their own distinc-

tive strategies for studying institutional analysis Thus economists tendto use several approaches (North 1990 Williamson 1985 Hodgson 19881998 1999) anthropologists another (Geertz 1995) and sociologistspolitical scientists and historians employ other strategies (Campbell etal 1991 Powell and DiMaggio 1991 Katznelson 1998 Hall and Taylor1996) With practitioners in various disciplines pursuing their distinctiveapproaches to institutional analysis there is low potential for collabo-rative institutional analysis across the social sciences The lessons fromthe biological sciences over the last 40 years are very instructive oncepractitioners from various biological disciplines began to cooperate instudying biological phenomena by using the same concepts at the mole-cular level theoretical advance was very rapid Comparable phenomenatook place in the development of other hybrid elds eg biochemistrybiophysics (Judson 1979 Olby 1979 Kohler 1982 1994) Similarly weresocial scientists in various disciplines to work collaboratively in the areaof institutional analysis there would very likely be an acceleration intheir theoretical knowledgeThis is not a plea that everyone engaged in institutional analysis should

do the same kind of research Indeed just as biologists working at themolecular level study many different kinds of problems similarly socialscientists engaged in institutional analysis would also work on manyseparate problems but in the spirit of a collective enterprise Some wouldconcentrate their attention on the study of rules norms habits conven-tions and values while others would study how these are associatedwith congurations of various institutional arrangements (types ofmarkets hierarchies networks associations communities clans statesetc) Others would work on specic institutional sectors (eg education

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

630

Table 4

Rules norms hab

its an

d values associated w

ith sp

ecic co

ngu

ration

s of coo

rdinating mecha

nism

s soc

ial sy

stem

s of

prod

uction

an

d nationa

l styles of inno

vation

USA

Germany

Norms rules values which provide incentives for

Norms rules values which provide incentives for

short-term horizons

long-term horizons

Con

gu

rative

forms of

Marke

ts c

orpo

rate hierarchies reg

ulatory state

Assoc

iation

s (busine

ss assoc

iation

s la

bor union

s)

coordinating

corp

oratist-type

networks

marke

ts c

orporatist-

mecha

nism

s (listed in

type

state

order of im

portanc

e in

the particu

lar social

system

)

Social sys

tem of

prod

uction

Bus

iness system

Firm

struc

tures

Con

glom

erates b

ut with mov

emen

t toward

Firm

more co

ngru

ent with pa

rticular prod

ucts

elim

ination of produc

tion

in som

e sectors

emph

asis on long

-term strateg

ies rm

s are well

emph

asis on sh

ort-term

strateg

ies high

integrated

differentiation

between divisions

of rms

Firm

s mob

ilized in

Low

High

and in

tegrated

into busine

ss

associations

Indu

strial relations

Low

deg

ree of jo

b secu

rity rigid in

ternal lab

or

High deg

ree of jo

b secu

rity

exible interna

l lab

or

system

marke

t ex

ible externa

l labo

r marke

tsmarke

t rigid externa

l lab

or m

arke

tPatterns of

Dispersed owne

rship frequ

ent turnov

erMore stab

le owne

rship patterns

owne

rship

Table 4

continue

d

USA

Germany

Norms rules values which provide incentives for

Norms rules values which provide incentives for

short-term horizons

long-term horizons

Social system of

Prod

uction

Bus

iness system

Training for labo

r an

dHigh deg

ree of voluntary decision mak

ing by

Collective an

d m

ore co

mpulso

ry d

ecisions

abo

ut

man

agem

ent

individuals as to ho

w m

uch

tim

e an

d ene

rgy

how m

uch ed

ucation

is ne

eded

he

avy em

pha

sis

to inv

est in educ

ation heavy

emph

asis on

on techn

ical training (eg en

gine

ering)

busine

ss sch

ools training

in marke

ting

and

sales accou

nting

Fina

ncial marke

tsEqu

ity an

d ven

ture cap

ital m

arke

ts highly

Equ

ity an

d ven

ture cap

ital m

arke

ts w

eakly

dev

elop

ed na

ncing from

cen

tral gov

ernm

ent

deve

lope

d he

avy em

pha

sis on

corporate loan

sto priva

te rm

s for dev

elop

men

t of new

tech

nologies related

to na

tion

al security

Unive

rsity research

High individual auton

omy high poten

tial for

Low

ind

ividua

l au

tono

my until almost 40

yea

rs of

system

individual creativity u

nive

rsity pr

ofesso

rs

age poten

tial for ind

ividual creativity high

ly

high

ly entreprene

urial

cons

traine

d unive

rsity pr

ofessors not so

entrep

rene

urial

Inno

vative

styles

Highly su

ccessful in m

ajor break

through

s in

High pe

rforman

ce in proc

ess an

d inc

remen

tal

basic an

d app

lied scien

ce go

od in dev

elop

ing

inno

vation

s in existing produ

cts not so su

ccessful

totally

new

lines of pr

oducts an

d in m

ajor break

through

s in basic scien

ce or in

tech

nologies high

deg

ree of entrepr

eneu

rship

deve

loping

totally new

lines of prod

ucts of

amon

g research

ers

tech

nologies low d

egree of entrepr

eneu

rship

amon

g research

ers

Table 4

continue

d

USA

Germany

Norms rules values which provide incentives for

Norms rules values which provide incentives for

short-term horizons

long-term horizons

Prod

ucts in w

hich

Pr

oducts with sh

ort liv

es prod

ucts in w

hich

Older produ

cts with long

history machine

ry pap

er

society excels

tech

nology

is bo

th highly co

mplex

and

pr

inting

textile

s m

otor veh

icles m

achine

too

lsch

ange

s rapidly en

tertainm

ent indus

try

pha

rmaceu

ticals biotechn

olog

y softw

are

aerosp

ace

business systems nancial markets and systems of research) Distinctivewould be the recognition of how these separate studies are linkedtogether how each institutional sector is linked to a societyrsquos normsrules values etc and its conguration of institutional arrangementsSimilarly other institutional analysts might focus on the study of orga-nizations with a major concern for how the institutional environmentof organizations (eg norms rules values congurations of institutionalarrangements institutional sectors) inuenced the structure culture andoutcomes of organizations And nally others might study how all ofthe aforementioned aspects of institutional analysis inuenced a societyrsquosoverall performance Nelson and Sampat (1998) have appropriatelyreminded us that we should engage in institutional analysis in order toaddress specic problems but the view developed in this article is thatwe rst need some mapping of the terrain of institutional analysis inorder for our diverse enterprises to become a meaningful collective enter-prise especially if we are to understand how institutions inuence theinnovative processThe argument developed here is not that all social scientists should engage

in institutional analysis Obviously there are numerous other importantareas of research which lie beyond the eld of institutional analysis Nordoes this article argue that institutional analysis is the most importantsubject for social scientistsSince the structure and culture of our research universities reect an

enormous amount of disciplinary fragmentation what are the prospectsfor promoting the kind of collective and interdisciplinary research agendaproposed above In one respect the prospects are not encouragingDuring the past half century the efforts of American universities to have effective communication across the biological sciences have variedgreatly in their success At the University of California-Berkeley it was not until the 1980s that there was a serious effort to integrate thebiological sciences and at that time the campus abolished seventeendepartments to promote an integrated biological program But at theUniversity of Wisconsin the biological sciences are still very fragmentedinto dozens of different departments in botany plant pathology zoologyand physiology with two departments of biochemistry and two depart-ments of geneticsThere have been other approaches which have worked very well histor-

ically for promoting scientic integration amidst scientic diversity Forexample several summer schools and special institutes were organizedin the United States during the 1940s in an effort to overcome the frag-mentation in the biological sciences These programs made it possiblefor scientists in diverse academic disciplines and from different univer-sities and countries to come together to engage in intense and frequentinteraction Perhaps the most effective program was that held in the

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

634

summers at Cold Spring Harbor New York where a research agendafor the study of molecular biology was developed (Fischer and Lipson1988 Stent and Watson 1966) Scientists went there summer aftersummer and when they returned to their universities many began therestructuring of the biological sciences in their universitiesIn our own day we need summer institutes and other similar programs

so that scholars from diverse backgrounds can come together anddevelop a collective agenda for doing institutional analysis and forstudying the impact of institutions on the performance of societies ndashinnovativeness being one such subject If we can develop such instituteswe will increase the prospects for transforming our universities Thereis reason to be optimistic that once an eclectic group of scholars workwith a common set of concepts we can have genuine theoretical advanceif researchers can bring their diverse expertise to bear on a common setof problems There is considerable evidence that institutional analysisoffers high potential for advancing our knowledge of innovations andtechnological change (Hollingsworth and Hollingsworth 2000) If we aregoing to make substantial theoretical advance it is not likely to occurfrom the vantage point of a multiplicity of disciplines Rather it is likelyto occur as we develop a new intellectual framework and our own acad-emic journals This is a tall agenda which requires us to reassess theway that knowledge is presently produced in our societies

NOTE

This article is part of a much larger agenda involving the study of institutionsorganizations and innovations in which I am involved My debts are many andvaried The rst version was written while I was in residence in the ResearchUnit for Institutional Change and European Integration of the Austrian Academyof Sciences in Vienna in the summer of 1998 I am especially grateful to ProfessorsEgon Matzner and Sonja Puntscher-Riekmann for providing the stimulating envi-ronment in which to think through the ideas developed here Egon Matzner hasengaged in very stimulating discussions with me about the importance of insti-tutions in the general eld of socio-economics The nal version of the articlewas completed during my residence as a fellow at the Neurosciences Institutein La Jolla California I also wish to thank my colleagues in the InnovationTheme Project at the Netherlands Institute for Advanced Study (NIAS) during1998ndash9 and Edgar Grande and Karl Muumlller for stimulating conversations aboutthe interaction of institutions and organizations in facilitating particular typesof innovation I am especially grateful to Jerald Hage for helping me to developa number of issues in the article and for his useful comments on the rst draftFrans van Waarden early on helped me to work through the idea of multiplelevels of institutional analysis Others who have helped to develop the ideas inthe article have been my colleagues in the research project lsquoComparative socialsystems of productionrsquo especially Tom Burns Christel Lane Yoshitaka OkadaWolfgang Streeck and Richard Whitley My colleagues Robert Boyer SteveCasper and David Soskice have taught me much about how incentives providedby the institutional environments of rms inuence individual decision making

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

635

But my most important debt is to Ellen Jane Hollingsworth who has repeatedlyquestioned every single assumption of this article Without her rich and contin-uous tutoring this article would be even less developed than it is at presentFinally without the insightful assistance of David Gear in many phases of thisarticle it would not have been completed

REFERENCES

Aoki Masahiko (1990) lsquoTowards an economic model of the Japanese rmrsquo Journalof Economic Literature March 1ndash27

Arora A Landau Ralph and Rosenberg Nathan (eds) (1998) Chemicals and Long-Term Economic Growth Insights from the Chemical Industry New York JohnWiley amp Sons

Arrow Kenneth (1974) The Limits of Organization New York NortonArthur W B (1988a) lsquoSelf-reinforcing mechanisms in economicsrsquo in P W

Anderson K J Arrow and D Pines (eds) The Economy as an Evolving ComplexSystem Redwood City Calif Addison-Wesley

mdashmdash(1988b) lsquoCompeting technologies an overviewrsquo in G Dosi C Freeman RNelson G Silverberg and L Soete (eds) Technical Change and Economic TheoryLondon Pinter Publishers pp 590ndash607

Audretsch David B (1995) Innovation and Industry Evolution Cambridge MassMIT Press

Audretsch David B and Feldman Maryann P (1995) lsquoInnovative clusters andthe industry life cyclersquo Discussion Paper Wissenschaftszentrum BerlinMarket Processes and Corporate Development Research Unit FS-IV-95-7

Baum Joel A C and Oliver Christine (1992) lsquoInstitutional embeddedness andthe dynamics of organizational populationsrsquo American Sociological Review 57540ndash59

Baum Joel A C and Singh Jitendra V (1994a) lsquoOrganizationndashenvironment coevo-lutionrsquo in Joel A C Baum and Jitendra V Singh (eds) The EvolutionaryDynamics of Organizations New York Oxford University Press pp 379ndash402

mdashmdash(eds) (1994b) The Evolutionary Dynamics of Organizations New York OxfordUniversity Press

Berger Suzanne and Dore Ronald (eds) (1996) National Diversity and GlobalCapitalism Ithaca NY Cornell University Press

Boyer Robert (1997) lsquoThe variety and unequal performance of really existingmarkets farewell to Doctor Panglossrsquo in J Rogers Hollingsworth and RobertBoyer (eds) Contemporary Capitalism the Embeddedness of InstitutionsCambridge and New York Cambridge University Press pp 55ndash93

Burns Tom R and Dietz Thomas (2001 forthcoming) lsquoSocial rule system theorysocial action institutional arrangements and evolutionary processesrsquo in JRogers Hollingsworth and Karl Muumlller (eds) Advances in SocioeconomicsLanham Md Rowman amp Littleeld

Burns Tom R and Flam Helena (1987) The Shaping of Social Organization BeverlyHills Calif Sage

Calvert Randall L (1995) lsquoThe rational choice theory of social institutionsrsquo inJeffrey S Banks and Eric A Hanuskek (eds) Modern Political Economy NewYork Cambridge University Press pp 216ndash66

Camic Charles (1986) lsquoThe matter of habitrsquo American Journal of Sociology 911039ndash87

Campbell John L (1997) lsquoMechanisms of evolutionary change in economicsgovernance interaction interpretation and bricolagersquo in Lars Magnusson

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

636

by the existing social system of production Thus the same technologymay exist in numerous countries but how it is employed and how itinuences societal outcomes and performances varies from one institu-tional conguration or society to another (Hollingsworth 1998)

PERSPECTIVES ON INSTITUTIONAL CHANGE

A fundamental problem which remains unresolved in institutionalanalysis involves the nature of institutional change There is much confu-sion and miscommunication in the social sciences about institutionalchange in large part because we do not even know how to measure therate of institutional change Of course we know that there are bothexternal and internal focuses for change and we have had a good bitof scholarship on this subject What is not clearly understood is thatinstitutions and institutional arrangements within societies are histori-cally rooted that there is a great deal of path-dependency to the waythat institutions evolve Moreover the more intricately linked each sectoris with each other sector and with a societyrsquos rules and norms the lesschoice actors have to devise new institutions and institutional arrange-ments On the other hand if a sector or actor is too isolated it may betoo weak within a society to bring about any kind of effective changeThese considerations are very critical to the problem of how muchfreedom actors have to create new institutions afresh to what extentinstitutions diffuse from one society to another (Murmann 1998 Lanzara1998) This is a fundamental question for students of institutionalanalysis as we attempt to understand how the different institutionallevels are linked togetherThere are a variety of reasons why there is confusion in studying insti-

tutional change There is disagreement over the extent to which actorshave the freedom to build new institutional arrangements and if theydo to what extent the new arrangements and organizational patternsmay depart from past practices Among some institutional economists(Williamson 1985) there is a basic assumption that actors build efcientinstitutions by pursuing their self-interest and by promoting theirstrategic goals Moreover there is the assumption that a mixing of fullrationality with market competition tends to produce all of the optimalinstitutions that are needed in order to coordinate a complex capitalisteconomy (Hollingsworth and Boyer 1997 ch 14 Hodgson 1999) Incontrast the argument developed here is much more complexSocial institutions are historically rooted and there is a great deal of

path-dependency in the way that various institutional componentsevolve The shape of institutional congurations at any moment limitsthe type of options for change Because there is a great deal of institu-tional inertia radical change in the institutional components of a society

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

624

is uncommon (Lanzara 1998) Recognizing the constraints existinginstitutions exert on actors we note that existing congurations of insti-tutional arrangements institutional sectors and organizational propertiescan limit but also inuence the degree to which norms and rules canchange Many scholars (North 1990 Ostrom 1990) have emphasized therole of norms rules ideologies and values in limiting the ability of actorsto develop new institutional arrangements which radically depart fromexisting ones However there is a two-way process of interaction aprocess of co-evolution (Baum and Singh 1994a Nelson 1994) Thus asFriedland and Alford (1991 244) argue actors develop new institutionalarrangements by recognizing organizational failures and low perfor-mance but these new arrangements do not necessarily change a societyrsquosnorms rules habits and its underlying value systemCampbell (1997) advances our understanding of how existing institu-

tions enable actors to construct radically new institutions He recognizesthat most institutions and institutional arrangements embody a degreeof inertia from which it is difcult for actors to depart Moreover thedifferential power relations among actors make it unlikely that lesspowerful actors can change the existing power structure of a societyMost change in the institutional components of a society evolves througha process of constrained selection which reects to a considerable degreethe existing arrangements and power relations This becomes the inter-pretative frame for social actors with institutionalized scripts and ritualswhich tend to be taken for granted and appear to be quite ordinary andnatural This institutionalized scripting of the social world is an extremelyimportant source of social stability and inertiaCampbell has developed a very fruitful explanation of change in terms

of the various institutional components outlined here He distinguishesbetween incremental and radical institutional change and argues thatradical institutional change occurs when social actors with widelydiffering norms cultural scripts and rituals for action engage in intenseinteraction with each other (see also Knight 1992) The more diverse theinteractions the greater the potential for institutional change He arguesthat as shifts in the composition of interaction among social actors occurchanges in the interpretation of problems and interests will follow Ifthere are only minor changes in the extent to which diverse actorsinteract there will be only modest institutional change But the morefundamental the changes in the interaction of diverse social actors thegreater the change in the way that actors interpret their world And thegreater the change in the way that actors interpret their world the morelikely that radical changes in various institutional components willfollow In other words diverse decision making tends to create a widerrange of interpretative frameworks of the social world than is likely tobe the case if social actors operate in isolation

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

625

This perspective of diverse interactions among social actors facilitatesour ability to move beyond the traditional view that existing institutionsconstrain the range of institutional alternatives which actors face Theolder view has tended to suggest that actors engage in institutionalchange by extending existing institutional principles habits and conven-tions Specifying the conditions which facilitate fundamental institutionalchange permits a much richer perspective for institutional analysisCampbellrsquos perspective about the conditions under which fundamentalor radical change in institutional components occurs is consistent withmy own research about the conditions under which major discoveriesor fundamental new knowledge occur the more scholars with diversebackgrounds interact with intensity and frequency the greater the like-lihood that they will develop new and alternative ways of thinking(Hollingsworth and Hollingsworth 2000)These views of institutional change are also consistent with the recent

work of Nelson and his colleagues (Mowery and Nelson 1999 Murmann1998) They report the research on a variety of industrial sectors in anumber of different countries revealing that networks of actors in partic-ular industries business and professional associations universityresearch departments and institutes are frequently able to use their collec-tive power to modify a societyrsquos property rights research system andgovernmental policies It is the intense and frequent interaction of thesediverse complex sets of actors ndash sometimes constituting a technologysystem ndash that often leads to incremental changes in a societyrsquos total socialsystem of production It is this kind of collective action that brings aboutorganizational change change in institutional sectors institutionalarrangements and even in the rules and norms of a society And if weare to understand how the style of innovativeness of a society changeswe must rst understand the changes in its institutional makeup It ishighly unlikely that a society will develop a new style of innovativenesswithout changing its institutional structure

FIRMS THEIR INSTITUTIONAL ENVIRONMENT ANDNATIONAL SYSTEMS OF INNOVATION

Variation in the innovative style of societies is shaped by their insti-tutional makeup ndash the various levels which are discussed aboveinstitutions institutional arrangements the structure and cohesion of institutional sectors which constitute a societyrsquos social system of pro-duction and the structure and culture of its organizations ndash especiallyits business rms and other research organizations By undertaking aninstitutional analysis of a society one can begin to understand in whatkinds of organizations the production of specic kinds of knowledgetakes place and how this is linked to particular kinds of innovativeness

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

626

With such a perspective one can gain a rich understanding of why somesocieties excel in the production of radical breakthroughs in basic andapplied science and in developing radically new products and why othersocieties excel in more incremental innovationsReecting on these issues one cannot help being impressed with the

contrasts between the innovative styles of German and American orga-nizations For example Germany has continued to be very successful in making incremental innovations in many industries in which it wasvery competitive well before World War II paper printing materialsmachinery electrotechnical products motor vehicles chemicals textileyarns and fabrics But the Germans have been much less successful indeveloping totally new products or related innovations in biotechnologyelectronics telecommunications aircraft as well as other newer indus-trial areas In contrast the Americans have been very innovative in these and other industries with products having very short lives andwith technologies which change very rapidly and are very complex(Hollingsworth 1991 Soskice 1997 1999)The importance of institutions and innovativeness for international

competitiveness has led to the recognition that there are national systemsof innovation (Lundvall 1992 Nelson 1993 Edquist 1997 Hage andHollingsworth 2000) But even before the recent studies of nationalsystems of innovation Landes (1969) was a proponent of these viewswhen he made the argument that Germany had developed a nationalsystem of innovation by building on its education and scientic researchsystems which were lacking in Britain and elsewhere However the concept of a national system of innovation immediately poses theproblem of what kind of innovations should be consideredDavid Soskice an economist at the Wissenschaftszentrum Berlin has

addressed a set of problems somewhat complementary to the perspec-tive raised here His work (Soskice 1990 1997 1999) suggests that weare sensitive to the institutional environment in which the followingactors are embedded the employees of rms the owners and nanciersof rms the competitors of rms and nally their collaborators ndash aperspective complementary to Hollingsworthrsquos work on social systemsof production (Hollingsworth 1997 1998) When one reects on the insti-tutional contextsocial system of production within which these actorsare embedded one becomes aware that those countries which have some-what inexible labor markets external to rms but a business system in which rms have long-term commitments to their employees arestrongly associated with incremental but hardly at all with radical inno-vations In the rms of such countries (eg Germany and Japan) thereis much more consensus decision making than in societies where a lsquohireand rersquo set of practices is quite pervasive And this kind of consensusdecision making ndash while highly conducive to incremental innovativeness

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

627

ndash limits the capacity for radical innovations to occur This kind of lsquolongtermismrsquo set of norms and rules occurs not only in the employmentpractices of incrementally innovative rms but also in their nancingand ownership patterns (Hollingsworth 1997) During the past thirtyyears these were characteristics of many rms in Germany and JapanSignicantly in both of these countries venture capital markets werenot highly developed nancing tended to be based more on long-termbank loans than on the equity markets and the relationships of rmstended to be relatively stable And these stable institutional arrange-ments also contributed to a high degree of innovativeness ndash butincremental rather than radical innovationsMeantime we should also pursue Soskicersquos ideas (1999) about how

rmsrsquo relationships with their competitors and collaborators shape theirstyles of innovativeness In those advanced industrial societies in whichbusiness rms are highly mobilized into business associations theretends to be a high degree of standardized quality controls over technicalnorms and a high degree of reliance on legal sanctions as monitoringinstruments vis-agrave-vis association members As a result of this high degreeof integration of rms into associations member rms are more likelyto be associated with incremental innovations than is likely to be the case with rms in those societies where associations are less welldeveloped In the latter type of society radical innovations are morelikely to occur In those advanced industrial societies where rms havelong-term stable relationships with collaborators in product design andproduction the style of innovation is likely to be incremental And it is in those societies in which there is a weaker tradition of strongcollaboration among rms that radical innovations are more likely tooccurObviously even in a country where there may be a great deal of

radical innovativeness most innovations will be of an incremental typefor there are simply fewer radical innovations But Germany and therest of Europe not only have had relatively few radical innovations incomparison with the United States but they also have lagged behindthe USA in developing relatively totally new industries Thus there hasbeen much less innovative activity in Europe in high-technology andnewly emerging industries than in the USAIn complementary research Audretsch and Feldman (1995) have

demonstrated that the USA has been more successful in spawning rmsengaged in the early stages of the industry life cycle but relative toGermany had an environment unfavorable to innovative activity byestablished rms especially in more mature industries German rmshave tended to be associated with a routinized technological regime ndashwhere established enterprises have had strong incentives to engage inincremental innovations but new rms have tended to have fundamental

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

628

innovative disadvantages relative to the Americans (see Audretsch 1995Winter 1982 Casper 2000)On both sides of the Atlantic universities engage in basic and applied

research But signicantly the institutional environment in which univer-sities are embedded inuences the degree to which the research is likelyto represent major breakthroughs or is likely to have more modestimplications Recent research in biomedical science demonstrates that inthose societies in which university professors andor graduate studentshave the following characteristics there are likely to be relatively majorresearch breakthroughs graduate students have a long training periodhigh dependency on a major professor low autonomy in dening thenature of a research project restricted job options and a university envi-ronment where there is a very modest or nonexistent entrepreneurialculture For example in Germany ndash a country with relatively few radicalbreakthroughs ndash the scholar who nishes a habilitation will have spentmany years on a research project before qualifying for a permanent acad-emic position will have been in a high-dependency position with a seniorprofessor will have had limited autonomy to dene a research projectand will be constrained from pursuing a high-risk research project Theyoung German will have an extraordinarily high level of training butwill have faced a number of disincentives to undertake high-risk researchAnd as soon as heshe has completed the habilitation and becomes aprofessor administrative responsibilities constitute another disincentivefrom engaging in high-risk or long-term research projects Moreover theGerman professor is a civil servant with a more stable and secure incomethan hisher American counterpart In a very different institutionalenvironment American professors in large research universities tend tobecome much more entrepreneurial Today many American professorsbehave as though they were operating their own rms within univer-sities and thus devote considerable time to generating funds for researchprojects assistants secretaries supplies etc And the variation in thisentrepreneurial spirit across countries is associated with variation in thetaking of high-risk research strategies and the making of major researchbreakthroughsIn sum societies vary in terms of the institutional environment in

which their organizations are embedded and this variation inuencesthe degree to which they have industrial activities in the early or latestages of the industry life cycle successful new or mature industriesand radical vs incremental innovations And it is this variation whichis encouraging rms in countries with relatively few new industries and with few or no radical innovations to invest in those countries with many new industries and numerous radical innovations ndash as astrategy to be able to transfer new technologies from the latter institu-tional environment to the former

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

629

Table 4 summarizes the way that the norms rules and values of a society are associated with Americarsquos and Germanyrsquos institutionalarrangements their social system of production and their styles of inno-vativeness

CONCLUSION

Several themes deserve to be reemphasized Most importantly institu-tional analysis is emerging as a dominant issue in several social sciencedisciplines The study of various institutional components discussedabove is a complex subject requiring multiple levels of analysis and aknowledge of multiple sectors within societies It requires a theoreticalperspective that evolves from an empirical analysis in different societiesIf we are to advance the study of institutional analysis we need cross-disciplinary collaboration and team research which are not readilyachieved given the structure of most of our research universitiesAt present scholars in each academic discipline have their own distinc-

tive strategies for studying institutional analysis Thus economists tendto use several approaches (North 1990 Williamson 1985 Hodgson 19881998 1999) anthropologists another (Geertz 1995) and sociologistspolitical scientists and historians employ other strategies (Campbell etal 1991 Powell and DiMaggio 1991 Katznelson 1998 Hall and Taylor1996) With practitioners in various disciplines pursuing their distinctiveapproaches to institutional analysis there is low potential for collabo-rative institutional analysis across the social sciences The lessons fromthe biological sciences over the last 40 years are very instructive oncepractitioners from various biological disciplines began to cooperate instudying biological phenomena by using the same concepts at the mole-cular level theoretical advance was very rapid Comparable phenomenatook place in the development of other hybrid elds eg biochemistrybiophysics (Judson 1979 Olby 1979 Kohler 1982 1994) Similarly weresocial scientists in various disciplines to work collaboratively in the areaof institutional analysis there would very likely be an acceleration intheir theoretical knowledgeThis is not a plea that everyone engaged in institutional analysis should

do the same kind of research Indeed just as biologists working at themolecular level study many different kinds of problems similarly socialscientists engaged in institutional analysis would also work on manyseparate problems but in the spirit of a collective enterprise Some wouldconcentrate their attention on the study of rules norms habits conven-tions and values while others would study how these are associatedwith congurations of various institutional arrangements (types ofmarkets hierarchies networks associations communities clans statesetc) Others would work on specic institutional sectors (eg education

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

630

Table 4

Rules norms hab

its an

d values associated w

ith sp

ecic co

ngu

ration

s of coo

rdinating mecha

nism

s soc

ial sy

stem

s of

prod

uction

an

d nationa

l styles of inno

vation

USA

Germany

Norms rules values which provide incentives for

Norms rules values which provide incentives for

short-term horizons

long-term horizons

Con

gu

rative

forms of

Marke

ts c

orpo

rate hierarchies reg

ulatory state

Assoc

iation

s (busine

ss assoc

iation

s la

bor union

s)

coordinating

corp

oratist-type

networks

marke

ts c

orporatist-

mecha

nism

s (listed in

type

state

order of im

portanc

e in

the particu

lar social

system

)

Social sys

tem of

prod

uction

Bus

iness system

Firm

struc

tures

Con

glom

erates b

ut with mov

emen

t toward

Firm

more co

ngru

ent with pa

rticular prod

ucts

elim

ination of produc

tion

in som

e sectors

emph

asis on long

-term strateg

ies rm

s are well

emph

asis on sh

ort-term

strateg

ies high

integrated

differentiation

between divisions

of rms

Firm

s mob

ilized in

Low

High

and in

tegrated

into busine

ss

associations

Indu

strial relations

Low

deg

ree of jo

b secu

rity rigid in

ternal lab

or

High deg

ree of jo

b secu

rity

exible interna

l lab

or

system

marke

t ex

ible externa

l labo

r marke

tsmarke

t rigid externa

l lab

or m

arke

tPatterns of

Dispersed owne

rship frequ

ent turnov

erMore stab

le owne

rship patterns

owne

rship

Table 4

continue

d

USA

Germany

Norms rules values which provide incentives for

Norms rules values which provide incentives for

short-term horizons

long-term horizons

Social system of

Prod

uction

Bus

iness system

Training for labo

r an

dHigh deg

ree of voluntary decision mak

ing by

Collective an

d m

ore co

mpulso

ry d

ecisions

abo

ut

man

agem

ent

individuals as to ho

w m

uch

tim

e an

d ene

rgy

how m

uch ed

ucation

is ne

eded

he

avy em

pha

sis

to inv

est in educ

ation heavy

emph

asis on

on techn

ical training (eg en

gine

ering)

busine

ss sch

ools training

in marke

ting

and

sales accou

nting

Fina

ncial marke

tsEqu

ity an

d ven

ture cap

ital m

arke

ts highly

Equ

ity an

d ven

ture cap

ital m

arke

ts w

eakly

dev

elop

ed na

ncing from

cen

tral gov

ernm

ent

deve

lope

d he

avy em

pha

sis on

corporate loan

sto priva

te rm

s for dev

elop

men

t of new

tech

nologies related

to na

tion

al security

Unive

rsity research

High individual auton

omy high poten

tial for

Low

ind

ividua

l au

tono

my until almost 40

yea

rs of

system

individual creativity u

nive

rsity pr

ofesso

rs

age poten

tial for ind

ividual creativity high

ly

high

ly entreprene

urial

cons

traine

d unive

rsity pr

ofessors not so

entrep

rene

urial

Inno

vative

styles

Highly su

ccessful in m

ajor break

through

s in

High pe

rforman

ce in proc

ess an

d inc

remen

tal

basic an

d app

lied scien

ce go

od in dev

elop

ing

inno

vation

s in existing produ

cts not so su

ccessful

totally

new

lines of pr

oducts an

d in m

ajor break

through

s in basic scien

ce or in

tech

nologies high

deg

ree of entrepr

eneu

rship

deve

loping

totally new

lines of prod

ucts of

amon

g research

ers

tech

nologies low d

egree of entrepr

eneu

rship

amon

g research

ers

Table 4

continue

d

USA

Germany

Norms rules values which provide incentives for

Norms rules values which provide incentives for

short-term horizons

long-term horizons

Prod

ucts in w

hich

Pr

oducts with sh

ort liv

es prod

ucts in w

hich

Older produ

cts with long

history machine

ry pap

er

society excels

tech

nology

is bo

th highly co

mplex

and

pr

inting

textile

s m

otor veh

icles m

achine

too

lsch

ange

s rapidly en

tertainm

ent indus

try

pha

rmaceu

ticals biotechn

olog

y softw

are

aerosp

ace

business systems nancial markets and systems of research) Distinctivewould be the recognition of how these separate studies are linkedtogether how each institutional sector is linked to a societyrsquos normsrules values etc and its conguration of institutional arrangementsSimilarly other institutional analysts might focus on the study of orga-nizations with a major concern for how the institutional environmentof organizations (eg norms rules values congurations of institutionalarrangements institutional sectors) inuenced the structure culture andoutcomes of organizations And nally others might study how all ofthe aforementioned aspects of institutional analysis inuenced a societyrsquosoverall performance Nelson and Sampat (1998) have appropriatelyreminded us that we should engage in institutional analysis in order toaddress specic problems but the view developed in this article is thatwe rst need some mapping of the terrain of institutional analysis inorder for our diverse enterprises to become a meaningful collective enter-prise especially if we are to understand how institutions inuence theinnovative processThe argument developed here is not that all social scientists should engage

in institutional analysis Obviously there are numerous other importantareas of research which lie beyond the eld of institutional analysis Nordoes this article argue that institutional analysis is the most importantsubject for social scientistsSince the structure and culture of our research universities reect an

enormous amount of disciplinary fragmentation what are the prospectsfor promoting the kind of collective and interdisciplinary research agendaproposed above In one respect the prospects are not encouragingDuring the past half century the efforts of American universities to have effective communication across the biological sciences have variedgreatly in their success At the University of California-Berkeley it was not until the 1980s that there was a serious effort to integrate thebiological sciences and at that time the campus abolished seventeendepartments to promote an integrated biological program But at theUniversity of Wisconsin the biological sciences are still very fragmentedinto dozens of different departments in botany plant pathology zoologyand physiology with two departments of biochemistry and two depart-ments of geneticsThere have been other approaches which have worked very well histor-

ically for promoting scientic integration amidst scientic diversity Forexample several summer schools and special institutes were organizedin the United States during the 1940s in an effort to overcome the frag-mentation in the biological sciences These programs made it possiblefor scientists in diverse academic disciplines and from different univer-sities and countries to come together to engage in intense and frequentinteraction Perhaps the most effective program was that held in the

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

634

summers at Cold Spring Harbor New York where a research agendafor the study of molecular biology was developed (Fischer and Lipson1988 Stent and Watson 1966) Scientists went there summer aftersummer and when they returned to their universities many began therestructuring of the biological sciences in their universitiesIn our own day we need summer institutes and other similar programs

so that scholars from diverse backgrounds can come together anddevelop a collective agenda for doing institutional analysis and forstudying the impact of institutions on the performance of societies ndashinnovativeness being one such subject If we can develop such instituteswe will increase the prospects for transforming our universities Thereis reason to be optimistic that once an eclectic group of scholars workwith a common set of concepts we can have genuine theoretical advanceif researchers can bring their diverse expertise to bear on a common setof problems There is considerable evidence that institutional analysisoffers high potential for advancing our knowledge of innovations andtechnological change (Hollingsworth and Hollingsworth 2000) If we aregoing to make substantial theoretical advance it is not likely to occurfrom the vantage point of a multiplicity of disciplines Rather it is likelyto occur as we develop a new intellectual framework and our own acad-emic journals This is a tall agenda which requires us to reassess theway that knowledge is presently produced in our societies

NOTE

This article is part of a much larger agenda involving the study of institutionsorganizations and innovations in which I am involved My debts are many andvaried The rst version was written while I was in residence in the ResearchUnit for Institutional Change and European Integration of the Austrian Academyof Sciences in Vienna in the summer of 1998 I am especially grateful to ProfessorsEgon Matzner and Sonja Puntscher-Riekmann for providing the stimulating envi-ronment in which to think through the ideas developed here Egon Matzner hasengaged in very stimulating discussions with me about the importance of insti-tutions in the general eld of socio-economics The nal version of the articlewas completed during my residence as a fellow at the Neurosciences Institutein La Jolla California I also wish to thank my colleagues in the InnovationTheme Project at the Netherlands Institute for Advanced Study (NIAS) during1998ndash9 and Edgar Grande and Karl Muumlller for stimulating conversations aboutthe interaction of institutions and organizations in facilitating particular typesof innovation I am especially grateful to Jerald Hage for helping me to developa number of issues in the article and for his useful comments on the rst draftFrans van Waarden early on helped me to work through the idea of multiplelevels of institutional analysis Others who have helped to develop the ideas inthe article have been my colleagues in the research project lsquoComparative socialsystems of productionrsquo especially Tom Burns Christel Lane Yoshitaka OkadaWolfgang Streeck and Richard Whitley My colleagues Robert Boyer SteveCasper and David Soskice have taught me much about how incentives providedby the institutional environments of rms inuence individual decision making

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

635

But my most important debt is to Ellen Jane Hollingsworth who has repeatedlyquestioned every single assumption of this article Without her rich and contin-uous tutoring this article would be even less developed than it is at presentFinally without the insightful assistance of David Gear in many phases of thisarticle it would not have been completed

REFERENCES

Aoki Masahiko (1990) lsquoTowards an economic model of the Japanese rmrsquo Journalof Economic Literature March 1ndash27

Arora A Landau Ralph and Rosenberg Nathan (eds) (1998) Chemicals and Long-Term Economic Growth Insights from the Chemical Industry New York JohnWiley amp Sons

Arrow Kenneth (1974) The Limits of Organization New York NortonArthur W B (1988a) lsquoSelf-reinforcing mechanisms in economicsrsquo in P W

Anderson K J Arrow and D Pines (eds) The Economy as an Evolving ComplexSystem Redwood City Calif Addison-Wesley

mdashmdash(1988b) lsquoCompeting technologies an overviewrsquo in G Dosi C Freeman RNelson G Silverberg and L Soete (eds) Technical Change and Economic TheoryLondon Pinter Publishers pp 590ndash607

Audretsch David B (1995) Innovation and Industry Evolution Cambridge MassMIT Press

Audretsch David B and Feldman Maryann P (1995) lsquoInnovative clusters andthe industry life cyclersquo Discussion Paper Wissenschaftszentrum BerlinMarket Processes and Corporate Development Research Unit FS-IV-95-7

Baum Joel A C and Oliver Christine (1992) lsquoInstitutional embeddedness andthe dynamics of organizational populationsrsquo American Sociological Review 57540ndash59

Baum Joel A C and Singh Jitendra V (1994a) lsquoOrganizationndashenvironment coevo-lutionrsquo in Joel A C Baum and Jitendra V Singh (eds) The EvolutionaryDynamics of Organizations New York Oxford University Press pp 379ndash402

mdashmdash(eds) (1994b) The Evolutionary Dynamics of Organizations New York OxfordUniversity Press

Berger Suzanne and Dore Ronald (eds) (1996) National Diversity and GlobalCapitalism Ithaca NY Cornell University Press

Boyer Robert (1997) lsquoThe variety and unequal performance of really existingmarkets farewell to Doctor Panglossrsquo in J Rogers Hollingsworth and RobertBoyer (eds) Contemporary Capitalism the Embeddedness of InstitutionsCambridge and New York Cambridge University Press pp 55ndash93

Burns Tom R and Dietz Thomas (2001 forthcoming) lsquoSocial rule system theorysocial action institutional arrangements and evolutionary processesrsquo in JRogers Hollingsworth and Karl Muumlller (eds) Advances in SocioeconomicsLanham Md Rowman amp Littleeld

Burns Tom R and Flam Helena (1987) The Shaping of Social Organization BeverlyHills Calif Sage

Calvert Randall L (1995) lsquoThe rational choice theory of social institutionsrsquo inJeffrey S Banks and Eric A Hanuskek (eds) Modern Political Economy NewYork Cambridge University Press pp 216ndash66

Camic Charles (1986) lsquoThe matter of habitrsquo American Journal of Sociology 911039ndash87

Campbell John L (1997) lsquoMechanisms of evolutionary change in economicsgovernance interaction interpretation and bricolagersquo in Lars Magnusson

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

636

is uncommon (Lanzara 1998) Recognizing the constraints existinginstitutions exert on actors we note that existing congurations of insti-tutional arrangements institutional sectors and organizational propertiescan limit but also inuence the degree to which norms and rules canchange Many scholars (North 1990 Ostrom 1990) have emphasized therole of norms rules ideologies and values in limiting the ability of actorsto develop new institutional arrangements which radically depart fromexisting ones However there is a two-way process of interaction aprocess of co-evolution (Baum and Singh 1994a Nelson 1994) Thus asFriedland and Alford (1991 244) argue actors develop new institutionalarrangements by recognizing organizational failures and low perfor-mance but these new arrangements do not necessarily change a societyrsquosnorms rules habits and its underlying value systemCampbell (1997) advances our understanding of how existing institu-

tions enable actors to construct radically new institutions He recognizesthat most institutions and institutional arrangements embody a degreeof inertia from which it is difcult for actors to depart Moreover thedifferential power relations among actors make it unlikely that lesspowerful actors can change the existing power structure of a societyMost change in the institutional components of a society evolves througha process of constrained selection which reects to a considerable degreethe existing arrangements and power relations This becomes the inter-pretative frame for social actors with institutionalized scripts and ritualswhich tend to be taken for granted and appear to be quite ordinary andnatural This institutionalized scripting of the social world is an extremelyimportant source of social stability and inertiaCampbell has developed a very fruitful explanation of change in terms

of the various institutional components outlined here He distinguishesbetween incremental and radical institutional change and argues thatradical institutional change occurs when social actors with widelydiffering norms cultural scripts and rituals for action engage in intenseinteraction with each other (see also Knight 1992) The more diverse theinteractions the greater the potential for institutional change He arguesthat as shifts in the composition of interaction among social actors occurchanges in the interpretation of problems and interests will follow Ifthere are only minor changes in the extent to which diverse actorsinteract there will be only modest institutional change But the morefundamental the changes in the interaction of diverse social actors thegreater the change in the way that actors interpret their world And thegreater the change in the way that actors interpret their world the morelikely that radical changes in various institutional components willfollow In other words diverse decision making tends to create a widerrange of interpretative frameworks of the social world than is likely tobe the case if social actors operate in isolation

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

625

This perspective of diverse interactions among social actors facilitatesour ability to move beyond the traditional view that existing institutionsconstrain the range of institutional alternatives which actors face Theolder view has tended to suggest that actors engage in institutionalchange by extending existing institutional principles habits and conven-tions Specifying the conditions which facilitate fundamental institutionalchange permits a much richer perspective for institutional analysisCampbellrsquos perspective about the conditions under which fundamentalor radical change in institutional components occurs is consistent withmy own research about the conditions under which major discoveriesor fundamental new knowledge occur the more scholars with diversebackgrounds interact with intensity and frequency the greater the like-lihood that they will develop new and alternative ways of thinking(Hollingsworth and Hollingsworth 2000)These views of institutional change are also consistent with the recent

work of Nelson and his colleagues (Mowery and Nelson 1999 Murmann1998) They report the research on a variety of industrial sectors in anumber of different countries revealing that networks of actors in partic-ular industries business and professional associations universityresearch departments and institutes are frequently able to use their collec-tive power to modify a societyrsquos property rights research system andgovernmental policies It is the intense and frequent interaction of thesediverse complex sets of actors ndash sometimes constituting a technologysystem ndash that often leads to incremental changes in a societyrsquos total socialsystem of production It is this kind of collective action that brings aboutorganizational change change in institutional sectors institutionalarrangements and even in the rules and norms of a society And if weare to understand how the style of innovativeness of a society changeswe must rst understand the changes in its institutional makeup It ishighly unlikely that a society will develop a new style of innovativenesswithout changing its institutional structure

FIRMS THEIR INSTITUTIONAL ENVIRONMENT ANDNATIONAL SYSTEMS OF INNOVATION

Variation in the innovative style of societies is shaped by their insti-tutional makeup ndash the various levels which are discussed aboveinstitutions institutional arrangements the structure and cohesion of institutional sectors which constitute a societyrsquos social system of pro-duction and the structure and culture of its organizations ndash especiallyits business rms and other research organizations By undertaking aninstitutional analysis of a society one can begin to understand in whatkinds of organizations the production of specic kinds of knowledgetakes place and how this is linked to particular kinds of innovativeness

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

626

With such a perspective one can gain a rich understanding of why somesocieties excel in the production of radical breakthroughs in basic andapplied science and in developing radically new products and why othersocieties excel in more incremental innovationsReecting on these issues one cannot help being impressed with the

contrasts between the innovative styles of German and American orga-nizations For example Germany has continued to be very successful in making incremental innovations in many industries in which it wasvery competitive well before World War II paper printing materialsmachinery electrotechnical products motor vehicles chemicals textileyarns and fabrics But the Germans have been much less successful indeveloping totally new products or related innovations in biotechnologyelectronics telecommunications aircraft as well as other newer indus-trial areas In contrast the Americans have been very innovative in these and other industries with products having very short lives andwith technologies which change very rapidly and are very complex(Hollingsworth 1991 Soskice 1997 1999)The importance of institutions and innovativeness for international

competitiveness has led to the recognition that there are national systemsof innovation (Lundvall 1992 Nelson 1993 Edquist 1997 Hage andHollingsworth 2000) But even before the recent studies of nationalsystems of innovation Landes (1969) was a proponent of these viewswhen he made the argument that Germany had developed a nationalsystem of innovation by building on its education and scientic researchsystems which were lacking in Britain and elsewhere However the concept of a national system of innovation immediately poses theproblem of what kind of innovations should be consideredDavid Soskice an economist at the Wissenschaftszentrum Berlin has

addressed a set of problems somewhat complementary to the perspec-tive raised here His work (Soskice 1990 1997 1999) suggests that weare sensitive to the institutional environment in which the followingactors are embedded the employees of rms the owners and nanciersof rms the competitors of rms and nally their collaborators ndash aperspective complementary to Hollingsworthrsquos work on social systemsof production (Hollingsworth 1997 1998) When one reects on the insti-tutional contextsocial system of production within which these actorsare embedded one becomes aware that those countries which have some-what inexible labor markets external to rms but a business system in which rms have long-term commitments to their employees arestrongly associated with incremental but hardly at all with radical inno-vations In the rms of such countries (eg Germany and Japan) thereis much more consensus decision making than in societies where a lsquohireand rersquo set of practices is quite pervasive And this kind of consensusdecision making ndash while highly conducive to incremental innovativeness

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

627

ndash limits the capacity for radical innovations to occur This kind of lsquolongtermismrsquo set of norms and rules occurs not only in the employmentpractices of incrementally innovative rms but also in their nancingand ownership patterns (Hollingsworth 1997) During the past thirtyyears these were characteristics of many rms in Germany and JapanSignicantly in both of these countries venture capital markets werenot highly developed nancing tended to be based more on long-termbank loans than on the equity markets and the relationships of rmstended to be relatively stable And these stable institutional arrange-ments also contributed to a high degree of innovativeness ndash butincremental rather than radical innovationsMeantime we should also pursue Soskicersquos ideas (1999) about how

rmsrsquo relationships with their competitors and collaborators shape theirstyles of innovativeness In those advanced industrial societies in whichbusiness rms are highly mobilized into business associations theretends to be a high degree of standardized quality controls over technicalnorms and a high degree of reliance on legal sanctions as monitoringinstruments vis-agrave-vis association members As a result of this high degreeof integration of rms into associations member rms are more likelyto be associated with incremental innovations than is likely to be the case with rms in those societies where associations are less welldeveloped In the latter type of society radical innovations are morelikely to occur In those advanced industrial societies where rms havelong-term stable relationships with collaborators in product design andproduction the style of innovation is likely to be incremental And it is in those societies in which there is a weaker tradition of strongcollaboration among rms that radical innovations are more likely tooccurObviously even in a country where there may be a great deal of

radical innovativeness most innovations will be of an incremental typefor there are simply fewer radical innovations But Germany and therest of Europe not only have had relatively few radical innovations incomparison with the United States but they also have lagged behindthe USA in developing relatively totally new industries Thus there hasbeen much less innovative activity in Europe in high-technology andnewly emerging industries than in the USAIn complementary research Audretsch and Feldman (1995) have

demonstrated that the USA has been more successful in spawning rmsengaged in the early stages of the industry life cycle but relative toGermany had an environment unfavorable to innovative activity byestablished rms especially in more mature industries German rmshave tended to be associated with a routinized technological regime ndashwhere established enterprises have had strong incentives to engage inincremental innovations but new rms have tended to have fundamental

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

628

innovative disadvantages relative to the Americans (see Audretsch 1995Winter 1982 Casper 2000)On both sides of the Atlantic universities engage in basic and applied

research But signicantly the institutional environment in which univer-sities are embedded inuences the degree to which the research is likelyto represent major breakthroughs or is likely to have more modestimplications Recent research in biomedical science demonstrates that inthose societies in which university professors andor graduate studentshave the following characteristics there are likely to be relatively majorresearch breakthroughs graduate students have a long training periodhigh dependency on a major professor low autonomy in dening thenature of a research project restricted job options and a university envi-ronment where there is a very modest or nonexistent entrepreneurialculture For example in Germany ndash a country with relatively few radicalbreakthroughs ndash the scholar who nishes a habilitation will have spentmany years on a research project before qualifying for a permanent acad-emic position will have been in a high-dependency position with a seniorprofessor will have had limited autonomy to dene a research projectand will be constrained from pursuing a high-risk research project Theyoung German will have an extraordinarily high level of training butwill have faced a number of disincentives to undertake high-risk researchAnd as soon as heshe has completed the habilitation and becomes aprofessor administrative responsibilities constitute another disincentivefrom engaging in high-risk or long-term research projects Moreover theGerman professor is a civil servant with a more stable and secure incomethan hisher American counterpart In a very different institutionalenvironment American professors in large research universities tend tobecome much more entrepreneurial Today many American professorsbehave as though they were operating their own rms within univer-sities and thus devote considerable time to generating funds for researchprojects assistants secretaries supplies etc And the variation in thisentrepreneurial spirit across countries is associated with variation in thetaking of high-risk research strategies and the making of major researchbreakthroughsIn sum societies vary in terms of the institutional environment in

which their organizations are embedded and this variation inuencesthe degree to which they have industrial activities in the early or latestages of the industry life cycle successful new or mature industriesand radical vs incremental innovations And it is this variation whichis encouraging rms in countries with relatively few new industries and with few or no radical innovations to invest in those countries with many new industries and numerous radical innovations ndash as astrategy to be able to transfer new technologies from the latter institu-tional environment to the former

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

629

Table 4 summarizes the way that the norms rules and values of a society are associated with Americarsquos and Germanyrsquos institutionalarrangements their social system of production and their styles of inno-vativeness

CONCLUSION

Several themes deserve to be reemphasized Most importantly institu-tional analysis is emerging as a dominant issue in several social sciencedisciplines The study of various institutional components discussedabove is a complex subject requiring multiple levels of analysis and aknowledge of multiple sectors within societies It requires a theoreticalperspective that evolves from an empirical analysis in different societiesIf we are to advance the study of institutional analysis we need cross-disciplinary collaboration and team research which are not readilyachieved given the structure of most of our research universitiesAt present scholars in each academic discipline have their own distinc-

tive strategies for studying institutional analysis Thus economists tendto use several approaches (North 1990 Williamson 1985 Hodgson 19881998 1999) anthropologists another (Geertz 1995) and sociologistspolitical scientists and historians employ other strategies (Campbell etal 1991 Powell and DiMaggio 1991 Katznelson 1998 Hall and Taylor1996) With practitioners in various disciplines pursuing their distinctiveapproaches to institutional analysis there is low potential for collabo-rative institutional analysis across the social sciences The lessons fromthe biological sciences over the last 40 years are very instructive oncepractitioners from various biological disciplines began to cooperate instudying biological phenomena by using the same concepts at the mole-cular level theoretical advance was very rapid Comparable phenomenatook place in the development of other hybrid elds eg biochemistrybiophysics (Judson 1979 Olby 1979 Kohler 1982 1994) Similarly weresocial scientists in various disciplines to work collaboratively in the areaof institutional analysis there would very likely be an acceleration intheir theoretical knowledgeThis is not a plea that everyone engaged in institutional analysis should

do the same kind of research Indeed just as biologists working at themolecular level study many different kinds of problems similarly socialscientists engaged in institutional analysis would also work on manyseparate problems but in the spirit of a collective enterprise Some wouldconcentrate their attention on the study of rules norms habits conven-tions and values while others would study how these are associatedwith congurations of various institutional arrangements (types ofmarkets hierarchies networks associations communities clans statesetc) Others would work on specic institutional sectors (eg education

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

630

Table 4

Rules norms hab

its an

d values associated w

ith sp

ecic co

ngu

ration

s of coo

rdinating mecha

nism

s soc

ial sy

stem

s of

prod

uction

an

d nationa

l styles of inno

vation

USA

Germany

Norms rules values which provide incentives for

Norms rules values which provide incentives for

short-term horizons

long-term horizons

Con

gu

rative

forms of

Marke

ts c

orpo

rate hierarchies reg

ulatory state

Assoc

iation

s (busine

ss assoc

iation

s la

bor union

s)

coordinating

corp

oratist-type

networks

marke

ts c

orporatist-

mecha

nism

s (listed in

type

state

order of im

portanc

e in

the particu

lar social

system

)

Social sys

tem of

prod

uction

Bus

iness system

Firm

struc

tures

Con

glom

erates b

ut with mov

emen

t toward

Firm

more co

ngru

ent with pa

rticular prod

ucts

elim

ination of produc

tion

in som

e sectors

emph

asis on long

-term strateg

ies rm

s are well

emph

asis on sh

ort-term

strateg

ies high

integrated

differentiation

between divisions

of rms

Firm

s mob

ilized in

Low

High

and in

tegrated

into busine

ss

associations

Indu

strial relations

Low

deg

ree of jo

b secu

rity rigid in

ternal lab

or

High deg

ree of jo

b secu

rity

exible interna

l lab

or

system

marke

t ex

ible externa

l labo

r marke

tsmarke

t rigid externa

l lab

or m

arke

tPatterns of

Dispersed owne

rship frequ

ent turnov

erMore stab

le owne

rship patterns

owne

rship

Table 4

continue

d

USA

Germany

Norms rules values which provide incentives for

Norms rules values which provide incentives for

short-term horizons

long-term horizons

Social system of

Prod

uction

Bus

iness system

Training for labo

r an

dHigh deg

ree of voluntary decision mak

ing by

Collective an

d m

ore co

mpulso

ry d

ecisions

abo

ut

man

agem

ent

individuals as to ho

w m

uch

tim

e an

d ene

rgy

how m

uch ed

ucation

is ne

eded

he

avy em

pha

sis

to inv

est in educ

ation heavy

emph

asis on

on techn

ical training (eg en

gine

ering)

busine

ss sch

ools training

in marke

ting

and

sales accou

nting

Fina

ncial marke

tsEqu

ity an

d ven

ture cap

ital m

arke

ts highly

Equ

ity an

d ven

ture cap

ital m

arke

ts w

eakly

dev

elop

ed na

ncing from

cen

tral gov

ernm

ent

deve

lope

d he

avy em

pha

sis on

corporate loan

sto priva

te rm

s for dev

elop

men

t of new

tech

nologies related

to na

tion

al security

Unive

rsity research

High individual auton

omy high poten

tial for

Low

ind

ividua

l au

tono

my until almost 40

yea

rs of

system

individual creativity u

nive

rsity pr

ofesso

rs

age poten

tial for ind

ividual creativity high

ly

high

ly entreprene

urial

cons

traine

d unive

rsity pr

ofessors not so

entrep

rene

urial

Inno

vative

styles

Highly su

ccessful in m

ajor break

through

s in

High pe

rforman

ce in proc

ess an

d inc

remen

tal

basic an

d app

lied scien

ce go

od in dev

elop

ing

inno

vation

s in existing produ

cts not so su

ccessful

totally

new

lines of pr

oducts an

d in m

ajor break

through

s in basic scien

ce or in

tech

nologies high

deg

ree of entrepr

eneu

rship

deve

loping

totally new

lines of prod

ucts of

amon

g research

ers

tech

nologies low d

egree of entrepr

eneu

rship

amon

g research

ers

Table 4

continue

d

USA

Germany

Norms rules values which provide incentives for

Norms rules values which provide incentives for

short-term horizons

long-term horizons

Prod

ucts in w

hich

Pr

oducts with sh

ort liv

es prod

ucts in w

hich

Older produ

cts with long

history machine

ry pap

er

society excels

tech

nology

is bo

th highly co

mplex

and

pr

inting

textile

s m

otor veh

icles m

achine

too

lsch

ange

s rapidly en

tertainm

ent indus

try

pha

rmaceu

ticals biotechn

olog

y softw

are

aerosp

ace

business systems nancial markets and systems of research) Distinctivewould be the recognition of how these separate studies are linkedtogether how each institutional sector is linked to a societyrsquos normsrules values etc and its conguration of institutional arrangementsSimilarly other institutional analysts might focus on the study of orga-nizations with a major concern for how the institutional environmentof organizations (eg norms rules values congurations of institutionalarrangements institutional sectors) inuenced the structure culture andoutcomes of organizations And nally others might study how all ofthe aforementioned aspects of institutional analysis inuenced a societyrsquosoverall performance Nelson and Sampat (1998) have appropriatelyreminded us that we should engage in institutional analysis in order toaddress specic problems but the view developed in this article is thatwe rst need some mapping of the terrain of institutional analysis inorder for our diverse enterprises to become a meaningful collective enter-prise especially if we are to understand how institutions inuence theinnovative processThe argument developed here is not that all social scientists should engage

in institutional analysis Obviously there are numerous other importantareas of research which lie beyond the eld of institutional analysis Nordoes this article argue that institutional analysis is the most importantsubject for social scientistsSince the structure and culture of our research universities reect an

enormous amount of disciplinary fragmentation what are the prospectsfor promoting the kind of collective and interdisciplinary research agendaproposed above In one respect the prospects are not encouragingDuring the past half century the efforts of American universities to have effective communication across the biological sciences have variedgreatly in their success At the University of California-Berkeley it was not until the 1980s that there was a serious effort to integrate thebiological sciences and at that time the campus abolished seventeendepartments to promote an integrated biological program But at theUniversity of Wisconsin the biological sciences are still very fragmentedinto dozens of different departments in botany plant pathology zoologyand physiology with two departments of biochemistry and two depart-ments of geneticsThere have been other approaches which have worked very well histor-

ically for promoting scientic integration amidst scientic diversity Forexample several summer schools and special institutes were organizedin the United States during the 1940s in an effort to overcome the frag-mentation in the biological sciences These programs made it possiblefor scientists in diverse academic disciplines and from different univer-sities and countries to come together to engage in intense and frequentinteraction Perhaps the most effective program was that held in the

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

634

summers at Cold Spring Harbor New York where a research agendafor the study of molecular biology was developed (Fischer and Lipson1988 Stent and Watson 1966) Scientists went there summer aftersummer and when they returned to their universities many began therestructuring of the biological sciences in their universitiesIn our own day we need summer institutes and other similar programs

so that scholars from diverse backgrounds can come together anddevelop a collective agenda for doing institutional analysis and forstudying the impact of institutions on the performance of societies ndashinnovativeness being one such subject If we can develop such instituteswe will increase the prospects for transforming our universities Thereis reason to be optimistic that once an eclectic group of scholars workwith a common set of concepts we can have genuine theoretical advanceif researchers can bring their diverse expertise to bear on a common setof problems There is considerable evidence that institutional analysisoffers high potential for advancing our knowledge of innovations andtechnological change (Hollingsworth and Hollingsworth 2000) If we aregoing to make substantial theoretical advance it is not likely to occurfrom the vantage point of a multiplicity of disciplines Rather it is likelyto occur as we develop a new intellectual framework and our own acad-emic journals This is a tall agenda which requires us to reassess theway that knowledge is presently produced in our societies

NOTE

This article is part of a much larger agenda involving the study of institutionsorganizations and innovations in which I am involved My debts are many andvaried The rst version was written while I was in residence in the ResearchUnit for Institutional Change and European Integration of the Austrian Academyof Sciences in Vienna in the summer of 1998 I am especially grateful to ProfessorsEgon Matzner and Sonja Puntscher-Riekmann for providing the stimulating envi-ronment in which to think through the ideas developed here Egon Matzner hasengaged in very stimulating discussions with me about the importance of insti-tutions in the general eld of socio-economics The nal version of the articlewas completed during my residence as a fellow at the Neurosciences Institutein La Jolla California I also wish to thank my colleagues in the InnovationTheme Project at the Netherlands Institute for Advanced Study (NIAS) during1998ndash9 and Edgar Grande and Karl Muumlller for stimulating conversations aboutthe interaction of institutions and organizations in facilitating particular typesof innovation I am especially grateful to Jerald Hage for helping me to developa number of issues in the article and for his useful comments on the rst draftFrans van Waarden early on helped me to work through the idea of multiplelevels of institutional analysis Others who have helped to develop the ideas inthe article have been my colleagues in the research project lsquoComparative socialsystems of productionrsquo especially Tom Burns Christel Lane Yoshitaka OkadaWolfgang Streeck and Richard Whitley My colleagues Robert Boyer SteveCasper and David Soskice have taught me much about how incentives providedby the institutional environments of rms inuence individual decision making

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

635

But my most important debt is to Ellen Jane Hollingsworth who has repeatedlyquestioned every single assumption of this article Without her rich and contin-uous tutoring this article would be even less developed than it is at presentFinally without the insightful assistance of David Gear in many phases of thisarticle it would not have been completed

REFERENCES

Aoki Masahiko (1990) lsquoTowards an economic model of the Japanese rmrsquo Journalof Economic Literature March 1ndash27

Arora A Landau Ralph and Rosenberg Nathan (eds) (1998) Chemicals and Long-Term Economic Growth Insights from the Chemical Industry New York JohnWiley amp Sons

Arrow Kenneth (1974) The Limits of Organization New York NortonArthur W B (1988a) lsquoSelf-reinforcing mechanisms in economicsrsquo in P W

Anderson K J Arrow and D Pines (eds) The Economy as an Evolving ComplexSystem Redwood City Calif Addison-Wesley

mdashmdash(1988b) lsquoCompeting technologies an overviewrsquo in G Dosi C Freeman RNelson G Silverberg and L Soete (eds) Technical Change and Economic TheoryLondon Pinter Publishers pp 590ndash607

Audretsch David B (1995) Innovation and Industry Evolution Cambridge MassMIT Press

Audretsch David B and Feldman Maryann P (1995) lsquoInnovative clusters andthe industry life cyclersquo Discussion Paper Wissenschaftszentrum BerlinMarket Processes and Corporate Development Research Unit FS-IV-95-7

Baum Joel A C and Oliver Christine (1992) lsquoInstitutional embeddedness andthe dynamics of organizational populationsrsquo American Sociological Review 57540ndash59

Baum Joel A C and Singh Jitendra V (1994a) lsquoOrganizationndashenvironment coevo-lutionrsquo in Joel A C Baum and Jitendra V Singh (eds) The EvolutionaryDynamics of Organizations New York Oxford University Press pp 379ndash402

mdashmdash(eds) (1994b) The Evolutionary Dynamics of Organizations New York OxfordUniversity Press

Berger Suzanne and Dore Ronald (eds) (1996) National Diversity and GlobalCapitalism Ithaca NY Cornell University Press

Boyer Robert (1997) lsquoThe variety and unequal performance of really existingmarkets farewell to Doctor Panglossrsquo in J Rogers Hollingsworth and RobertBoyer (eds) Contemporary Capitalism the Embeddedness of InstitutionsCambridge and New York Cambridge University Press pp 55ndash93

Burns Tom R and Dietz Thomas (2001 forthcoming) lsquoSocial rule system theorysocial action institutional arrangements and evolutionary processesrsquo in JRogers Hollingsworth and Karl Muumlller (eds) Advances in SocioeconomicsLanham Md Rowman amp Littleeld

Burns Tom R and Flam Helena (1987) The Shaping of Social Organization BeverlyHills Calif Sage

Calvert Randall L (1995) lsquoThe rational choice theory of social institutionsrsquo inJeffrey S Banks and Eric A Hanuskek (eds) Modern Political Economy NewYork Cambridge University Press pp 216ndash66

Camic Charles (1986) lsquoThe matter of habitrsquo American Journal of Sociology 911039ndash87

Campbell John L (1997) lsquoMechanisms of evolutionary change in economicsgovernance interaction interpretation and bricolagersquo in Lars Magnusson

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

636

This perspective of diverse interactions among social actors facilitatesour ability to move beyond the traditional view that existing institutionsconstrain the range of institutional alternatives which actors face Theolder view has tended to suggest that actors engage in institutionalchange by extending existing institutional principles habits and conven-tions Specifying the conditions which facilitate fundamental institutionalchange permits a much richer perspective for institutional analysisCampbellrsquos perspective about the conditions under which fundamentalor radical change in institutional components occurs is consistent withmy own research about the conditions under which major discoveriesor fundamental new knowledge occur the more scholars with diversebackgrounds interact with intensity and frequency the greater the like-lihood that they will develop new and alternative ways of thinking(Hollingsworth and Hollingsworth 2000)These views of institutional change are also consistent with the recent

work of Nelson and his colleagues (Mowery and Nelson 1999 Murmann1998) They report the research on a variety of industrial sectors in anumber of different countries revealing that networks of actors in partic-ular industries business and professional associations universityresearch departments and institutes are frequently able to use their collec-tive power to modify a societyrsquos property rights research system andgovernmental policies It is the intense and frequent interaction of thesediverse complex sets of actors ndash sometimes constituting a technologysystem ndash that often leads to incremental changes in a societyrsquos total socialsystem of production It is this kind of collective action that brings aboutorganizational change change in institutional sectors institutionalarrangements and even in the rules and norms of a society And if weare to understand how the style of innovativeness of a society changeswe must rst understand the changes in its institutional makeup It ishighly unlikely that a society will develop a new style of innovativenesswithout changing its institutional structure

FIRMS THEIR INSTITUTIONAL ENVIRONMENT ANDNATIONAL SYSTEMS OF INNOVATION

Variation in the innovative style of societies is shaped by their insti-tutional makeup ndash the various levels which are discussed aboveinstitutions institutional arrangements the structure and cohesion of institutional sectors which constitute a societyrsquos social system of pro-duction and the structure and culture of its organizations ndash especiallyits business rms and other research organizations By undertaking aninstitutional analysis of a society one can begin to understand in whatkinds of organizations the production of specic kinds of knowledgetakes place and how this is linked to particular kinds of innovativeness

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

626

With such a perspective one can gain a rich understanding of why somesocieties excel in the production of radical breakthroughs in basic andapplied science and in developing radically new products and why othersocieties excel in more incremental innovationsReecting on these issues one cannot help being impressed with the

contrasts between the innovative styles of German and American orga-nizations For example Germany has continued to be very successful in making incremental innovations in many industries in which it wasvery competitive well before World War II paper printing materialsmachinery electrotechnical products motor vehicles chemicals textileyarns and fabrics But the Germans have been much less successful indeveloping totally new products or related innovations in biotechnologyelectronics telecommunications aircraft as well as other newer indus-trial areas In contrast the Americans have been very innovative in these and other industries with products having very short lives andwith technologies which change very rapidly and are very complex(Hollingsworth 1991 Soskice 1997 1999)The importance of institutions and innovativeness for international

competitiveness has led to the recognition that there are national systemsof innovation (Lundvall 1992 Nelson 1993 Edquist 1997 Hage andHollingsworth 2000) But even before the recent studies of nationalsystems of innovation Landes (1969) was a proponent of these viewswhen he made the argument that Germany had developed a nationalsystem of innovation by building on its education and scientic researchsystems which were lacking in Britain and elsewhere However the concept of a national system of innovation immediately poses theproblem of what kind of innovations should be consideredDavid Soskice an economist at the Wissenschaftszentrum Berlin has

addressed a set of problems somewhat complementary to the perspec-tive raised here His work (Soskice 1990 1997 1999) suggests that weare sensitive to the institutional environment in which the followingactors are embedded the employees of rms the owners and nanciersof rms the competitors of rms and nally their collaborators ndash aperspective complementary to Hollingsworthrsquos work on social systemsof production (Hollingsworth 1997 1998) When one reects on the insti-tutional contextsocial system of production within which these actorsare embedded one becomes aware that those countries which have some-what inexible labor markets external to rms but a business system in which rms have long-term commitments to their employees arestrongly associated with incremental but hardly at all with radical inno-vations In the rms of such countries (eg Germany and Japan) thereis much more consensus decision making than in societies where a lsquohireand rersquo set of practices is quite pervasive And this kind of consensusdecision making ndash while highly conducive to incremental innovativeness

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

627

ndash limits the capacity for radical innovations to occur This kind of lsquolongtermismrsquo set of norms and rules occurs not only in the employmentpractices of incrementally innovative rms but also in their nancingand ownership patterns (Hollingsworth 1997) During the past thirtyyears these were characteristics of many rms in Germany and JapanSignicantly in both of these countries venture capital markets werenot highly developed nancing tended to be based more on long-termbank loans than on the equity markets and the relationships of rmstended to be relatively stable And these stable institutional arrange-ments also contributed to a high degree of innovativeness ndash butincremental rather than radical innovationsMeantime we should also pursue Soskicersquos ideas (1999) about how

rmsrsquo relationships with their competitors and collaborators shape theirstyles of innovativeness In those advanced industrial societies in whichbusiness rms are highly mobilized into business associations theretends to be a high degree of standardized quality controls over technicalnorms and a high degree of reliance on legal sanctions as monitoringinstruments vis-agrave-vis association members As a result of this high degreeof integration of rms into associations member rms are more likelyto be associated with incremental innovations than is likely to be the case with rms in those societies where associations are less welldeveloped In the latter type of society radical innovations are morelikely to occur In those advanced industrial societies where rms havelong-term stable relationships with collaborators in product design andproduction the style of innovation is likely to be incremental And it is in those societies in which there is a weaker tradition of strongcollaboration among rms that radical innovations are more likely tooccurObviously even in a country where there may be a great deal of

radical innovativeness most innovations will be of an incremental typefor there are simply fewer radical innovations But Germany and therest of Europe not only have had relatively few radical innovations incomparison with the United States but they also have lagged behindthe USA in developing relatively totally new industries Thus there hasbeen much less innovative activity in Europe in high-technology andnewly emerging industries than in the USAIn complementary research Audretsch and Feldman (1995) have

demonstrated that the USA has been more successful in spawning rmsengaged in the early stages of the industry life cycle but relative toGermany had an environment unfavorable to innovative activity byestablished rms especially in more mature industries German rmshave tended to be associated with a routinized technological regime ndashwhere established enterprises have had strong incentives to engage inincremental innovations but new rms have tended to have fundamental

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

628

innovative disadvantages relative to the Americans (see Audretsch 1995Winter 1982 Casper 2000)On both sides of the Atlantic universities engage in basic and applied

research But signicantly the institutional environment in which univer-sities are embedded inuences the degree to which the research is likelyto represent major breakthroughs or is likely to have more modestimplications Recent research in biomedical science demonstrates that inthose societies in which university professors andor graduate studentshave the following characteristics there are likely to be relatively majorresearch breakthroughs graduate students have a long training periodhigh dependency on a major professor low autonomy in dening thenature of a research project restricted job options and a university envi-ronment where there is a very modest or nonexistent entrepreneurialculture For example in Germany ndash a country with relatively few radicalbreakthroughs ndash the scholar who nishes a habilitation will have spentmany years on a research project before qualifying for a permanent acad-emic position will have been in a high-dependency position with a seniorprofessor will have had limited autonomy to dene a research projectand will be constrained from pursuing a high-risk research project Theyoung German will have an extraordinarily high level of training butwill have faced a number of disincentives to undertake high-risk researchAnd as soon as heshe has completed the habilitation and becomes aprofessor administrative responsibilities constitute another disincentivefrom engaging in high-risk or long-term research projects Moreover theGerman professor is a civil servant with a more stable and secure incomethan hisher American counterpart In a very different institutionalenvironment American professors in large research universities tend tobecome much more entrepreneurial Today many American professorsbehave as though they were operating their own rms within univer-sities and thus devote considerable time to generating funds for researchprojects assistants secretaries supplies etc And the variation in thisentrepreneurial spirit across countries is associated with variation in thetaking of high-risk research strategies and the making of major researchbreakthroughsIn sum societies vary in terms of the institutional environment in

which their organizations are embedded and this variation inuencesthe degree to which they have industrial activities in the early or latestages of the industry life cycle successful new or mature industriesand radical vs incremental innovations And it is this variation whichis encouraging rms in countries with relatively few new industries and with few or no radical innovations to invest in those countries with many new industries and numerous radical innovations ndash as astrategy to be able to transfer new technologies from the latter institu-tional environment to the former

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

629

Table 4 summarizes the way that the norms rules and values of a society are associated with Americarsquos and Germanyrsquos institutionalarrangements their social system of production and their styles of inno-vativeness

CONCLUSION

Several themes deserve to be reemphasized Most importantly institu-tional analysis is emerging as a dominant issue in several social sciencedisciplines The study of various institutional components discussedabove is a complex subject requiring multiple levels of analysis and aknowledge of multiple sectors within societies It requires a theoreticalperspective that evolves from an empirical analysis in different societiesIf we are to advance the study of institutional analysis we need cross-disciplinary collaboration and team research which are not readilyachieved given the structure of most of our research universitiesAt present scholars in each academic discipline have their own distinc-

tive strategies for studying institutional analysis Thus economists tendto use several approaches (North 1990 Williamson 1985 Hodgson 19881998 1999) anthropologists another (Geertz 1995) and sociologistspolitical scientists and historians employ other strategies (Campbell etal 1991 Powell and DiMaggio 1991 Katznelson 1998 Hall and Taylor1996) With practitioners in various disciplines pursuing their distinctiveapproaches to institutional analysis there is low potential for collabo-rative institutional analysis across the social sciences The lessons fromthe biological sciences over the last 40 years are very instructive oncepractitioners from various biological disciplines began to cooperate instudying biological phenomena by using the same concepts at the mole-cular level theoretical advance was very rapid Comparable phenomenatook place in the development of other hybrid elds eg biochemistrybiophysics (Judson 1979 Olby 1979 Kohler 1982 1994) Similarly weresocial scientists in various disciplines to work collaboratively in the areaof institutional analysis there would very likely be an acceleration intheir theoretical knowledgeThis is not a plea that everyone engaged in institutional analysis should

do the same kind of research Indeed just as biologists working at themolecular level study many different kinds of problems similarly socialscientists engaged in institutional analysis would also work on manyseparate problems but in the spirit of a collective enterprise Some wouldconcentrate their attention on the study of rules norms habits conven-tions and values while others would study how these are associatedwith congurations of various institutional arrangements (types ofmarkets hierarchies networks associations communities clans statesetc) Others would work on specic institutional sectors (eg education

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

630

Table 4

Rules norms hab

its an

d values associated w

ith sp

ecic co

ngu

ration

s of coo

rdinating mecha

nism

s soc

ial sy

stem

s of

prod

uction

an

d nationa

l styles of inno

vation

USA

Germany

Norms rules values which provide incentives for

Norms rules values which provide incentives for

short-term horizons

long-term horizons

Con

gu

rative

forms of

Marke

ts c

orpo

rate hierarchies reg

ulatory state

Assoc

iation

s (busine

ss assoc

iation

s la

bor union

s)

coordinating

corp

oratist-type

networks

marke

ts c

orporatist-

mecha

nism

s (listed in

type

state

order of im

portanc

e in

the particu

lar social

system

)

Social sys

tem of

prod

uction

Bus

iness system

Firm

struc

tures

Con

glom

erates b

ut with mov

emen

t toward

Firm

more co

ngru

ent with pa

rticular prod

ucts

elim

ination of produc

tion

in som

e sectors

emph

asis on long

-term strateg

ies rm

s are well

emph

asis on sh

ort-term

strateg

ies high

integrated

differentiation

between divisions

of rms

Firm

s mob

ilized in

Low

High

and in

tegrated

into busine

ss

associations

Indu

strial relations

Low

deg

ree of jo

b secu

rity rigid in

ternal lab

or

High deg

ree of jo

b secu

rity

exible interna

l lab

or

system

marke

t ex

ible externa

l labo

r marke

tsmarke

t rigid externa

l lab

or m

arke

tPatterns of

Dispersed owne

rship frequ

ent turnov

erMore stab

le owne

rship patterns

owne

rship

Table 4

continue

d

USA

Germany

Norms rules values which provide incentives for

Norms rules values which provide incentives for

short-term horizons

long-term horizons

Social system of

Prod

uction

Bus

iness system

Training for labo

r an

dHigh deg

ree of voluntary decision mak

ing by

Collective an

d m

ore co

mpulso

ry d

ecisions

abo

ut

man

agem

ent

individuals as to ho

w m

uch

tim

e an

d ene

rgy

how m

uch ed

ucation

is ne

eded

he

avy em

pha

sis

to inv

est in educ

ation heavy

emph

asis on

on techn

ical training (eg en

gine

ering)

busine

ss sch

ools training

in marke

ting

and

sales accou

nting

Fina

ncial marke

tsEqu

ity an

d ven

ture cap

ital m

arke

ts highly

Equ

ity an

d ven

ture cap

ital m

arke

ts w

eakly

dev

elop

ed na

ncing from

cen

tral gov

ernm

ent

deve

lope

d he

avy em

pha

sis on

corporate loan

sto priva

te rm

s for dev

elop

men

t of new

tech

nologies related

to na

tion

al security

Unive

rsity research

High individual auton

omy high poten

tial for

Low

ind

ividua

l au

tono

my until almost 40

yea

rs of

system

individual creativity u

nive

rsity pr

ofesso

rs

age poten

tial for ind

ividual creativity high

ly

high

ly entreprene

urial

cons

traine

d unive

rsity pr

ofessors not so

entrep

rene

urial

Inno

vative

styles

Highly su

ccessful in m

ajor break

through

s in

High pe

rforman

ce in proc

ess an

d inc

remen

tal

basic an

d app

lied scien

ce go

od in dev

elop

ing

inno

vation

s in existing produ

cts not so su

ccessful

totally

new

lines of pr

oducts an

d in m

ajor break

through

s in basic scien

ce or in

tech

nologies high

deg

ree of entrepr

eneu

rship

deve

loping

totally new

lines of prod

ucts of

amon

g research

ers

tech

nologies low d

egree of entrepr

eneu

rship

amon

g research

ers

Table 4

continue

d

USA

Germany

Norms rules values which provide incentives for

Norms rules values which provide incentives for

short-term horizons

long-term horizons

Prod

ucts in w

hich

Pr

oducts with sh

ort liv

es prod

ucts in w

hich

Older produ

cts with long

history machine

ry pap

er

society excels

tech

nology

is bo

th highly co

mplex

and

pr

inting

textile

s m

otor veh

icles m

achine

too

lsch

ange

s rapidly en

tertainm

ent indus

try

pha

rmaceu

ticals biotechn

olog

y softw

are

aerosp

ace

business systems nancial markets and systems of research) Distinctivewould be the recognition of how these separate studies are linkedtogether how each institutional sector is linked to a societyrsquos normsrules values etc and its conguration of institutional arrangementsSimilarly other institutional analysts might focus on the study of orga-nizations with a major concern for how the institutional environmentof organizations (eg norms rules values congurations of institutionalarrangements institutional sectors) inuenced the structure culture andoutcomes of organizations And nally others might study how all ofthe aforementioned aspects of institutional analysis inuenced a societyrsquosoverall performance Nelson and Sampat (1998) have appropriatelyreminded us that we should engage in institutional analysis in order toaddress specic problems but the view developed in this article is thatwe rst need some mapping of the terrain of institutional analysis inorder for our diverse enterprises to become a meaningful collective enter-prise especially if we are to understand how institutions inuence theinnovative processThe argument developed here is not that all social scientists should engage

in institutional analysis Obviously there are numerous other importantareas of research which lie beyond the eld of institutional analysis Nordoes this article argue that institutional analysis is the most importantsubject for social scientistsSince the structure and culture of our research universities reect an

enormous amount of disciplinary fragmentation what are the prospectsfor promoting the kind of collective and interdisciplinary research agendaproposed above In one respect the prospects are not encouragingDuring the past half century the efforts of American universities to have effective communication across the biological sciences have variedgreatly in their success At the University of California-Berkeley it was not until the 1980s that there was a serious effort to integrate thebiological sciences and at that time the campus abolished seventeendepartments to promote an integrated biological program But at theUniversity of Wisconsin the biological sciences are still very fragmentedinto dozens of different departments in botany plant pathology zoologyand physiology with two departments of biochemistry and two depart-ments of geneticsThere have been other approaches which have worked very well histor-

ically for promoting scientic integration amidst scientic diversity Forexample several summer schools and special institutes were organizedin the United States during the 1940s in an effort to overcome the frag-mentation in the biological sciences These programs made it possiblefor scientists in diverse academic disciplines and from different univer-sities and countries to come together to engage in intense and frequentinteraction Perhaps the most effective program was that held in the

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

634

summers at Cold Spring Harbor New York where a research agendafor the study of molecular biology was developed (Fischer and Lipson1988 Stent and Watson 1966) Scientists went there summer aftersummer and when they returned to their universities many began therestructuring of the biological sciences in their universitiesIn our own day we need summer institutes and other similar programs

so that scholars from diverse backgrounds can come together anddevelop a collective agenda for doing institutional analysis and forstudying the impact of institutions on the performance of societies ndashinnovativeness being one such subject If we can develop such instituteswe will increase the prospects for transforming our universities Thereis reason to be optimistic that once an eclectic group of scholars workwith a common set of concepts we can have genuine theoretical advanceif researchers can bring their diverse expertise to bear on a common setof problems There is considerable evidence that institutional analysisoffers high potential for advancing our knowledge of innovations andtechnological change (Hollingsworth and Hollingsworth 2000) If we aregoing to make substantial theoretical advance it is not likely to occurfrom the vantage point of a multiplicity of disciplines Rather it is likelyto occur as we develop a new intellectual framework and our own acad-emic journals This is a tall agenda which requires us to reassess theway that knowledge is presently produced in our societies

NOTE

This article is part of a much larger agenda involving the study of institutionsorganizations and innovations in which I am involved My debts are many andvaried The rst version was written while I was in residence in the ResearchUnit for Institutional Change and European Integration of the Austrian Academyof Sciences in Vienna in the summer of 1998 I am especially grateful to ProfessorsEgon Matzner and Sonja Puntscher-Riekmann for providing the stimulating envi-ronment in which to think through the ideas developed here Egon Matzner hasengaged in very stimulating discussions with me about the importance of insti-tutions in the general eld of socio-economics The nal version of the articlewas completed during my residence as a fellow at the Neurosciences Institutein La Jolla California I also wish to thank my colleagues in the InnovationTheme Project at the Netherlands Institute for Advanced Study (NIAS) during1998ndash9 and Edgar Grande and Karl Muumlller for stimulating conversations aboutthe interaction of institutions and organizations in facilitating particular typesof innovation I am especially grateful to Jerald Hage for helping me to developa number of issues in the article and for his useful comments on the rst draftFrans van Waarden early on helped me to work through the idea of multiplelevels of institutional analysis Others who have helped to develop the ideas inthe article have been my colleagues in the research project lsquoComparative socialsystems of productionrsquo especially Tom Burns Christel Lane Yoshitaka OkadaWolfgang Streeck and Richard Whitley My colleagues Robert Boyer SteveCasper and David Soskice have taught me much about how incentives providedby the institutional environments of rms inuence individual decision making

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

635

But my most important debt is to Ellen Jane Hollingsworth who has repeatedlyquestioned every single assumption of this article Without her rich and contin-uous tutoring this article would be even less developed than it is at presentFinally without the insightful assistance of David Gear in many phases of thisarticle it would not have been completed

REFERENCES

Aoki Masahiko (1990) lsquoTowards an economic model of the Japanese rmrsquo Journalof Economic Literature March 1ndash27

Arora A Landau Ralph and Rosenberg Nathan (eds) (1998) Chemicals and Long-Term Economic Growth Insights from the Chemical Industry New York JohnWiley amp Sons

Arrow Kenneth (1974) The Limits of Organization New York NortonArthur W B (1988a) lsquoSelf-reinforcing mechanisms in economicsrsquo in P W

Anderson K J Arrow and D Pines (eds) The Economy as an Evolving ComplexSystem Redwood City Calif Addison-Wesley

mdashmdash(1988b) lsquoCompeting technologies an overviewrsquo in G Dosi C Freeman RNelson G Silverberg and L Soete (eds) Technical Change and Economic TheoryLondon Pinter Publishers pp 590ndash607

Audretsch David B (1995) Innovation and Industry Evolution Cambridge MassMIT Press

Audretsch David B and Feldman Maryann P (1995) lsquoInnovative clusters andthe industry life cyclersquo Discussion Paper Wissenschaftszentrum BerlinMarket Processes and Corporate Development Research Unit FS-IV-95-7

Baum Joel A C and Oliver Christine (1992) lsquoInstitutional embeddedness andthe dynamics of organizational populationsrsquo American Sociological Review 57540ndash59

Baum Joel A C and Singh Jitendra V (1994a) lsquoOrganizationndashenvironment coevo-lutionrsquo in Joel A C Baum and Jitendra V Singh (eds) The EvolutionaryDynamics of Organizations New York Oxford University Press pp 379ndash402

mdashmdash(eds) (1994b) The Evolutionary Dynamics of Organizations New York OxfordUniversity Press

Berger Suzanne and Dore Ronald (eds) (1996) National Diversity and GlobalCapitalism Ithaca NY Cornell University Press

Boyer Robert (1997) lsquoThe variety and unequal performance of really existingmarkets farewell to Doctor Panglossrsquo in J Rogers Hollingsworth and RobertBoyer (eds) Contemporary Capitalism the Embeddedness of InstitutionsCambridge and New York Cambridge University Press pp 55ndash93

Burns Tom R and Dietz Thomas (2001 forthcoming) lsquoSocial rule system theorysocial action institutional arrangements and evolutionary processesrsquo in JRogers Hollingsworth and Karl Muumlller (eds) Advances in SocioeconomicsLanham Md Rowman amp Littleeld

Burns Tom R and Flam Helena (1987) The Shaping of Social Organization BeverlyHills Calif Sage

Calvert Randall L (1995) lsquoThe rational choice theory of social institutionsrsquo inJeffrey S Banks and Eric A Hanuskek (eds) Modern Political Economy NewYork Cambridge University Press pp 216ndash66

Camic Charles (1986) lsquoThe matter of habitrsquo American Journal of Sociology 911039ndash87

Campbell John L (1997) lsquoMechanisms of evolutionary change in economicsgovernance interaction interpretation and bricolagersquo in Lars Magnusson

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

636

With such a perspective one can gain a rich understanding of why somesocieties excel in the production of radical breakthroughs in basic andapplied science and in developing radically new products and why othersocieties excel in more incremental innovationsReecting on these issues one cannot help being impressed with the

contrasts between the innovative styles of German and American orga-nizations For example Germany has continued to be very successful in making incremental innovations in many industries in which it wasvery competitive well before World War II paper printing materialsmachinery electrotechnical products motor vehicles chemicals textileyarns and fabrics But the Germans have been much less successful indeveloping totally new products or related innovations in biotechnologyelectronics telecommunications aircraft as well as other newer indus-trial areas In contrast the Americans have been very innovative in these and other industries with products having very short lives andwith technologies which change very rapidly and are very complex(Hollingsworth 1991 Soskice 1997 1999)The importance of institutions and innovativeness for international

competitiveness has led to the recognition that there are national systemsof innovation (Lundvall 1992 Nelson 1993 Edquist 1997 Hage andHollingsworth 2000) But even before the recent studies of nationalsystems of innovation Landes (1969) was a proponent of these viewswhen he made the argument that Germany had developed a nationalsystem of innovation by building on its education and scientic researchsystems which were lacking in Britain and elsewhere However the concept of a national system of innovation immediately poses theproblem of what kind of innovations should be consideredDavid Soskice an economist at the Wissenschaftszentrum Berlin has

addressed a set of problems somewhat complementary to the perspec-tive raised here His work (Soskice 1990 1997 1999) suggests that weare sensitive to the institutional environment in which the followingactors are embedded the employees of rms the owners and nanciersof rms the competitors of rms and nally their collaborators ndash aperspective complementary to Hollingsworthrsquos work on social systemsof production (Hollingsworth 1997 1998) When one reects on the insti-tutional contextsocial system of production within which these actorsare embedded one becomes aware that those countries which have some-what inexible labor markets external to rms but a business system in which rms have long-term commitments to their employees arestrongly associated with incremental but hardly at all with radical inno-vations In the rms of such countries (eg Germany and Japan) thereis much more consensus decision making than in societies where a lsquohireand rersquo set of practices is quite pervasive And this kind of consensusdecision making ndash while highly conducive to incremental innovativeness

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

627

ndash limits the capacity for radical innovations to occur This kind of lsquolongtermismrsquo set of norms and rules occurs not only in the employmentpractices of incrementally innovative rms but also in their nancingand ownership patterns (Hollingsworth 1997) During the past thirtyyears these were characteristics of many rms in Germany and JapanSignicantly in both of these countries venture capital markets werenot highly developed nancing tended to be based more on long-termbank loans than on the equity markets and the relationships of rmstended to be relatively stable And these stable institutional arrange-ments also contributed to a high degree of innovativeness ndash butincremental rather than radical innovationsMeantime we should also pursue Soskicersquos ideas (1999) about how

rmsrsquo relationships with their competitors and collaborators shape theirstyles of innovativeness In those advanced industrial societies in whichbusiness rms are highly mobilized into business associations theretends to be a high degree of standardized quality controls over technicalnorms and a high degree of reliance on legal sanctions as monitoringinstruments vis-agrave-vis association members As a result of this high degreeof integration of rms into associations member rms are more likelyto be associated with incremental innovations than is likely to be the case with rms in those societies where associations are less welldeveloped In the latter type of society radical innovations are morelikely to occur In those advanced industrial societies where rms havelong-term stable relationships with collaborators in product design andproduction the style of innovation is likely to be incremental And it is in those societies in which there is a weaker tradition of strongcollaboration among rms that radical innovations are more likely tooccurObviously even in a country where there may be a great deal of

radical innovativeness most innovations will be of an incremental typefor there are simply fewer radical innovations But Germany and therest of Europe not only have had relatively few radical innovations incomparison with the United States but they also have lagged behindthe USA in developing relatively totally new industries Thus there hasbeen much less innovative activity in Europe in high-technology andnewly emerging industries than in the USAIn complementary research Audretsch and Feldman (1995) have

demonstrated that the USA has been more successful in spawning rmsengaged in the early stages of the industry life cycle but relative toGermany had an environment unfavorable to innovative activity byestablished rms especially in more mature industries German rmshave tended to be associated with a routinized technological regime ndashwhere established enterprises have had strong incentives to engage inincremental innovations but new rms have tended to have fundamental

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

628

innovative disadvantages relative to the Americans (see Audretsch 1995Winter 1982 Casper 2000)On both sides of the Atlantic universities engage in basic and applied

research But signicantly the institutional environment in which univer-sities are embedded inuences the degree to which the research is likelyto represent major breakthroughs or is likely to have more modestimplications Recent research in biomedical science demonstrates that inthose societies in which university professors andor graduate studentshave the following characteristics there are likely to be relatively majorresearch breakthroughs graduate students have a long training periodhigh dependency on a major professor low autonomy in dening thenature of a research project restricted job options and a university envi-ronment where there is a very modest or nonexistent entrepreneurialculture For example in Germany ndash a country with relatively few radicalbreakthroughs ndash the scholar who nishes a habilitation will have spentmany years on a research project before qualifying for a permanent acad-emic position will have been in a high-dependency position with a seniorprofessor will have had limited autonomy to dene a research projectand will be constrained from pursuing a high-risk research project Theyoung German will have an extraordinarily high level of training butwill have faced a number of disincentives to undertake high-risk researchAnd as soon as heshe has completed the habilitation and becomes aprofessor administrative responsibilities constitute another disincentivefrom engaging in high-risk or long-term research projects Moreover theGerman professor is a civil servant with a more stable and secure incomethan hisher American counterpart In a very different institutionalenvironment American professors in large research universities tend tobecome much more entrepreneurial Today many American professorsbehave as though they were operating their own rms within univer-sities and thus devote considerable time to generating funds for researchprojects assistants secretaries supplies etc And the variation in thisentrepreneurial spirit across countries is associated with variation in thetaking of high-risk research strategies and the making of major researchbreakthroughsIn sum societies vary in terms of the institutional environment in

which their organizations are embedded and this variation inuencesthe degree to which they have industrial activities in the early or latestages of the industry life cycle successful new or mature industriesand radical vs incremental innovations And it is this variation whichis encouraging rms in countries with relatively few new industries and with few or no radical innovations to invest in those countries with many new industries and numerous radical innovations ndash as astrategy to be able to transfer new technologies from the latter institu-tional environment to the former

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

629

Table 4 summarizes the way that the norms rules and values of a society are associated with Americarsquos and Germanyrsquos institutionalarrangements their social system of production and their styles of inno-vativeness

CONCLUSION

Several themes deserve to be reemphasized Most importantly institu-tional analysis is emerging as a dominant issue in several social sciencedisciplines The study of various institutional components discussedabove is a complex subject requiring multiple levels of analysis and aknowledge of multiple sectors within societies It requires a theoreticalperspective that evolves from an empirical analysis in different societiesIf we are to advance the study of institutional analysis we need cross-disciplinary collaboration and team research which are not readilyachieved given the structure of most of our research universitiesAt present scholars in each academic discipline have their own distinc-

tive strategies for studying institutional analysis Thus economists tendto use several approaches (North 1990 Williamson 1985 Hodgson 19881998 1999) anthropologists another (Geertz 1995) and sociologistspolitical scientists and historians employ other strategies (Campbell etal 1991 Powell and DiMaggio 1991 Katznelson 1998 Hall and Taylor1996) With practitioners in various disciplines pursuing their distinctiveapproaches to institutional analysis there is low potential for collabo-rative institutional analysis across the social sciences The lessons fromthe biological sciences over the last 40 years are very instructive oncepractitioners from various biological disciplines began to cooperate instudying biological phenomena by using the same concepts at the mole-cular level theoretical advance was very rapid Comparable phenomenatook place in the development of other hybrid elds eg biochemistrybiophysics (Judson 1979 Olby 1979 Kohler 1982 1994) Similarly weresocial scientists in various disciplines to work collaboratively in the areaof institutional analysis there would very likely be an acceleration intheir theoretical knowledgeThis is not a plea that everyone engaged in institutional analysis should

do the same kind of research Indeed just as biologists working at themolecular level study many different kinds of problems similarly socialscientists engaged in institutional analysis would also work on manyseparate problems but in the spirit of a collective enterprise Some wouldconcentrate their attention on the study of rules norms habits conven-tions and values while others would study how these are associatedwith congurations of various institutional arrangements (types ofmarkets hierarchies networks associations communities clans statesetc) Others would work on specic institutional sectors (eg education

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

630

Table 4

Rules norms hab

its an

d values associated w

ith sp

ecic co

ngu

ration

s of coo

rdinating mecha

nism

s soc

ial sy

stem

s of

prod

uction

an

d nationa

l styles of inno

vation

USA

Germany

Norms rules values which provide incentives for

Norms rules values which provide incentives for

short-term horizons

long-term horizons

Con

gu

rative

forms of

Marke

ts c

orpo

rate hierarchies reg

ulatory state

Assoc

iation

s (busine

ss assoc

iation

s la

bor union

s)

coordinating

corp

oratist-type

networks

marke

ts c

orporatist-

mecha

nism

s (listed in

type

state

order of im

portanc

e in

the particu

lar social

system

)

Social sys

tem of

prod

uction

Bus

iness system

Firm

struc

tures

Con

glom

erates b

ut with mov

emen

t toward

Firm

more co

ngru

ent with pa

rticular prod

ucts

elim

ination of produc

tion

in som

e sectors

emph

asis on long

-term strateg

ies rm

s are well

emph

asis on sh

ort-term

strateg

ies high

integrated

differentiation

between divisions

of rms

Firm

s mob

ilized in

Low

High

and in

tegrated

into busine

ss

associations

Indu

strial relations

Low

deg

ree of jo

b secu

rity rigid in

ternal lab

or

High deg

ree of jo

b secu

rity

exible interna

l lab

or

system

marke

t ex

ible externa

l labo

r marke

tsmarke

t rigid externa

l lab

or m

arke

tPatterns of

Dispersed owne

rship frequ

ent turnov

erMore stab

le owne

rship patterns

owne

rship

Table 4

continue

d

USA

Germany

Norms rules values which provide incentives for

Norms rules values which provide incentives for

short-term horizons

long-term horizons

Social system of

Prod

uction

Bus

iness system

Training for labo

r an

dHigh deg

ree of voluntary decision mak

ing by

Collective an

d m

ore co

mpulso

ry d

ecisions

abo

ut

man

agem

ent

individuals as to ho

w m

uch

tim

e an

d ene

rgy

how m

uch ed

ucation

is ne

eded

he

avy em

pha

sis

to inv

est in educ

ation heavy

emph

asis on

on techn

ical training (eg en

gine

ering)

busine

ss sch

ools training

in marke

ting

and

sales accou

nting

Fina

ncial marke

tsEqu

ity an

d ven

ture cap

ital m

arke

ts highly

Equ

ity an

d ven

ture cap

ital m

arke

ts w

eakly

dev

elop

ed na

ncing from

cen

tral gov

ernm

ent

deve

lope

d he

avy em

pha

sis on

corporate loan

sto priva

te rm

s for dev

elop

men

t of new

tech

nologies related

to na

tion

al security

Unive

rsity research

High individual auton

omy high poten

tial for

Low

ind

ividua

l au

tono

my until almost 40

yea

rs of

system

individual creativity u

nive

rsity pr

ofesso

rs

age poten

tial for ind

ividual creativity high

ly

high

ly entreprene

urial

cons

traine

d unive

rsity pr

ofessors not so

entrep

rene

urial

Inno

vative

styles

Highly su

ccessful in m

ajor break

through

s in

High pe

rforman

ce in proc

ess an

d inc

remen

tal

basic an

d app

lied scien

ce go

od in dev

elop

ing

inno

vation

s in existing produ

cts not so su

ccessful

totally

new

lines of pr

oducts an

d in m

ajor break

through

s in basic scien

ce or in

tech

nologies high

deg

ree of entrepr

eneu

rship

deve

loping

totally new

lines of prod

ucts of

amon

g research

ers

tech

nologies low d

egree of entrepr

eneu

rship

amon

g research

ers

Table 4

continue

d

USA

Germany

Norms rules values which provide incentives for

Norms rules values which provide incentives for

short-term horizons

long-term horizons

Prod

ucts in w

hich

Pr

oducts with sh

ort liv

es prod

ucts in w

hich

Older produ

cts with long

history machine

ry pap

er

society excels

tech

nology

is bo

th highly co

mplex

and

pr

inting

textile

s m

otor veh

icles m

achine

too

lsch

ange

s rapidly en

tertainm

ent indus

try

pha

rmaceu

ticals biotechn

olog

y softw

are

aerosp

ace

business systems nancial markets and systems of research) Distinctivewould be the recognition of how these separate studies are linkedtogether how each institutional sector is linked to a societyrsquos normsrules values etc and its conguration of institutional arrangementsSimilarly other institutional analysts might focus on the study of orga-nizations with a major concern for how the institutional environmentof organizations (eg norms rules values congurations of institutionalarrangements institutional sectors) inuenced the structure culture andoutcomes of organizations And nally others might study how all ofthe aforementioned aspects of institutional analysis inuenced a societyrsquosoverall performance Nelson and Sampat (1998) have appropriatelyreminded us that we should engage in institutional analysis in order toaddress specic problems but the view developed in this article is thatwe rst need some mapping of the terrain of institutional analysis inorder for our diverse enterprises to become a meaningful collective enter-prise especially if we are to understand how institutions inuence theinnovative processThe argument developed here is not that all social scientists should engage

in institutional analysis Obviously there are numerous other importantareas of research which lie beyond the eld of institutional analysis Nordoes this article argue that institutional analysis is the most importantsubject for social scientistsSince the structure and culture of our research universities reect an

enormous amount of disciplinary fragmentation what are the prospectsfor promoting the kind of collective and interdisciplinary research agendaproposed above In one respect the prospects are not encouragingDuring the past half century the efforts of American universities to have effective communication across the biological sciences have variedgreatly in their success At the University of California-Berkeley it was not until the 1980s that there was a serious effort to integrate thebiological sciences and at that time the campus abolished seventeendepartments to promote an integrated biological program But at theUniversity of Wisconsin the biological sciences are still very fragmentedinto dozens of different departments in botany plant pathology zoologyand physiology with two departments of biochemistry and two depart-ments of geneticsThere have been other approaches which have worked very well histor-

ically for promoting scientic integration amidst scientic diversity Forexample several summer schools and special institutes were organizedin the United States during the 1940s in an effort to overcome the frag-mentation in the biological sciences These programs made it possiblefor scientists in diverse academic disciplines and from different univer-sities and countries to come together to engage in intense and frequentinteraction Perhaps the most effective program was that held in the

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

634

summers at Cold Spring Harbor New York where a research agendafor the study of molecular biology was developed (Fischer and Lipson1988 Stent and Watson 1966) Scientists went there summer aftersummer and when they returned to their universities many began therestructuring of the biological sciences in their universitiesIn our own day we need summer institutes and other similar programs

so that scholars from diverse backgrounds can come together anddevelop a collective agenda for doing institutional analysis and forstudying the impact of institutions on the performance of societies ndashinnovativeness being one such subject If we can develop such instituteswe will increase the prospects for transforming our universities Thereis reason to be optimistic that once an eclectic group of scholars workwith a common set of concepts we can have genuine theoretical advanceif researchers can bring their diverse expertise to bear on a common setof problems There is considerable evidence that institutional analysisoffers high potential for advancing our knowledge of innovations andtechnological change (Hollingsworth and Hollingsworth 2000) If we aregoing to make substantial theoretical advance it is not likely to occurfrom the vantage point of a multiplicity of disciplines Rather it is likelyto occur as we develop a new intellectual framework and our own acad-emic journals This is a tall agenda which requires us to reassess theway that knowledge is presently produced in our societies

NOTE

This article is part of a much larger agenda involving the study of institutionsorganizations and innovations in which I am involved My debts are many andvaried The rst version was written while I was in residence in the ResearchUnit for Institutional Change and European Integration of the Austrian Academyof Sciences in Vienna in the summer of 1998 I am especially grateful to ProfessorsEgon Matzner and Sonja Puntscher-Riekmann for providing the stimulating envi-ronment in which to think through the ideas developed here Egon Matzner hasengaged in very stimulating discussions with me about the importance of insti-tutions in the general eld of socio-economics The nal version of the articlewas completed during my residence as a fellow at the Neurosciences Institutein La Jolla California I also wish to thank my colleagues in the InnovationTheme Project at the Netherlands Institute for Advanced Study (NIAS) during1998ndash9 and Edgar Grande and Karl Muumlller for stimulating conversations aboutthe interaction of institutions and organizations in facilitating particular typesof innovation I am especially grateful to Jerald Hage for helping me to developa number of issues in the article and for his useful comments on the rst draftFrans van Waarden early on helped me to work through the idea of multiplelevels of institutional analysis Others who have helped to develop the ideas inthe article have been my colleagues in the research project lsquoComparative socialsystems of productionrsquo especially Tom Burns Christel Lane Yoshitaka OkadaWolfgang Streeck and Richard Whitley My colleagues Robert Boyer SteveCasper and David Soskice have taught me much about how incentives providedby the institutional environments of rms inuence individual decision making

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

635

But my most important debt is to Ellen Jane Hollingsworth who has repeatedlyquestioned every single assumption of this article Without her rich and contin-uous tutoring this article would be even less developed than it is at presentFinally without the insightful assistance of David Gear in many phases of thisarticle it would not have been completed

REFERENCES

Aoki Masahiko (1990) lsquoTowards an economic model of the Japanese rmrsquo Journalof Economic Literature March 1ndash27

Arora A Landau Ralph and Rosenberg Nathan (eds) (1998) Chemicals and Long-Term Economic Growth Insights from the Chemical Industry New York JohnWiley amp Sons

Arrow Kenneth (1974) The Limits of Organization New York NortonArthur W B (1988a) lsquoSelf-reinforcing mechanisms in economicsrsquo in P W

Anderson K J Arrow and D Pines (eds) The Economy as an Evolving ComplexSystem Redwood City Calif Addison-Wesley

mdashmdash(1988b) lsquoCompeting technologies an overviewrsquo in G Dosi C Freeman RNelson G Silverberg and L Soete (eds) Technical Change and Economic TheoryLondon Pinter Publishers pp 590ndash607

Audretsch David B (1995) Innovation and Industry Evolution Cambridge MassMIT Press

Audretsch David B and Feldman Maryann P (1995) lsquoInnovative clusters andthe industry life cyclersquo Discussion Paper Wissenschaftszentrum BerlinMarket Processes and Corporate Development Research Unit FS-IV-95-7

Baum Joel A C and Oliver Christine (1992) lsquoInstitutional embeddedness andthe dynamics of organizational populationsrsquo American Sociological Review 57540ndash59

Baum Joel A C and Singh Jitendra V (1994a) lsquoOrganizationndashenvironment coevo-lutionrsquo in Joel A C Baum and Jitendra V Singh (eds) The EvolutionaryDynamics of Organizations New York Oxford University Press pp 379ndash402

mdashmdash(eds) (1994b) The Evolutionary Dynamics of Organizations New York OxfordUniversity Press

Berger Suzanne and Dore Ronald (eds) (1996) National Diversity and GlobalCapitalism Ithaca NY Cornell University Press

Boyer Robert (1997) lsquoThe variety and unequal performance of really existingmarkets farewell to Doctor Panglossrsquo in J Rogers Hollingsworth and RobertBoyer (eds) Contemporary Capitalism the Embeddedness of InstitutionsCambridge and New York Cambridge University Press pp 55ndash93

Burns Tom R and Dietz Thomas (2001 forthcoming) lsquoSocial rule system theorysocial action institutional arrangements and evolutionary processesrsquo in JRogers Hollingsworth and Karl Muumlller (eds) Advances in SocioeconomicsLanham Md Rowman amp Littleeld

Burns Tom R and Flam Helena (1987) The Shaping of Social Organization BeverlyHills Calif Sage

Calvert Randall L (1995) lsquoThe rational choice theory of social institutionsrsquo inJeffrey S Banks and Eric A Hanuskek (eds) Modern Political Economy NewYork Cambridge University Press pp 216ndash66

Camic Charles (1986) lsquoThe matter of habitrsquo American Journal of Sociology 911039ndash87

Campbell John L (1997) lsquoMechanisms of evolutionary change in economicsgovernance interaction interpretation and bricolagersquo in Lars Magnusson

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

636

ndash limits the capacity for radical innovations to occur This kind of lsquolongtermismrsquo set of norms and rules occurs not only in the employmentpractices of incrementally innovative rms but also in their nancingand ownership patterns (Hollingsworth 1997) During the past thirtyyears these were characteristics of many rms in Germany and JapanSignicantly in both of these countries venture capital markets werenot highly developed nancing tended to be based more on long-termbank loans than on the equity markets and the relationships of rmstended to be relatively stable And these stable institutional arrange-ments also contributed to a high degree of innovativeness ndash butincremental rather than radical innovationsMeantime we should also pursue Soskicersquos ideas (1999) about how

rmsrsquo relationships with their competitors and collaborators shape theirstyles of innovativeness In those advanced industrial societies in whichbusiness rms are highly mobilized into business associations theretends to be a high degree of standardized quality controls over technicalnorms and a high degree of reliance on legal sanctions as monitoringinstruments vis-agrave-vis association members As a result of this high degreeof integration of rms into associations member rms are more likelyto be associated with incremental innovations than is likely to be the case with rms in those societies where associations are less welldeveloped In the latter type of society radical innovations are morelikely to occur In those advanced industrial societies where rms havelong-term stable relationships with collaborators in product design andproduction the style of innovation is likely to be incremental And it is in those societies in which there is a weaker tradition of strongcollaboration among rms that radical innovations are more likely tooccurObviously even in a country where there may be a great deal of

radical innovativeness most innovations will be of an incremental typefor there are simply fewer radical innovations But Germany and therest of Europe not only have had relatively few radical innovations incomparison with the United States but they also have lagged behindthe USA in developing relatively totally new industries Thus there hasbeen much less innovative activity in Europe in high-technology andnewly emerging industries than in the USAIn complementary research Audretsch and Feldman (1995) have

demonstrated that the USA has been more successful in spawning rmsengaged in the early stages of the industry life cycle but relative toGermany had an environment unfavorable to innovative activity byestablished rms especially in more mature industries German rmshave tended to be associated with a routinized technological regime ndashwhere established enterprises have had strong incentives to engage inincremental innovations but new rms have tended to have fundamental

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

628

innovative disadvantages relative to the Americans (see Audretsch 1995Winter 1982 Casper 2000)On both sides of the Atlantic universities engage in basic and applied

research But signicantly the institutional environment in which univer-sities are embedded inuences the degree to which the research is likelyto represent major breakthroughs or is likely to have more modestimplications Recent research in biomedical science demonstrates that inthose societies in which university professors andor graduate studentshave the following characteristics there are likely to be relatively majorresearch breakthroughs graduate students have a long training periodhigh dependency on a major professor low autonomy in dening thenature of a research project restricted job options and a university envi-ronment where there is a very modest or nonexistent entrepreneurialculture For example in Germany ndash a country with relatively few radicalbreakthroughs ndash the scholar who nishes a habilitation will have spentmany years on a research project before qualifying for a permanent acad-emic position will have been in a high-dependency position with a seniorprofessor will have had limited autonomy to dene a research projectand will be constrained from pursuing a high-risk research project Theyoung German will have an extraordinarily high level of training butwill have faced a number of disincentives to undertake high-risk researchAnd as soon as heshe has completed the habilitation and becomes aprofessor administrative responsibilities constitute another disincentivefrom engaging in high-risk or long-term research projects Moreover theGerman professor is a civil servant with a more stable and secure incomethan hisher American counterpart In a very different institutionalenvironment American professors in large research universities tend tobecome much more entrepreneurial Today many American professorsbehave as though they were operating their own rms within univer-sities and thus devote considerable time to generating funds for researchprojects assistants secretaries supplies etc And the variation in thisentrepreneurial spirit across countries is associated with variation in thetaking of high-risk research strategies and the making of major researchbreakthroughsIn sum societies vary in terms of the institutional environment in

which their organizations are embedded and this variation inuencesthe degree to which they have industrial activities in the early or latestages of the industry life cycle successful new or mature industriesand radical vs incremental innovations And it is this variation whichis encouraging rms in countries with relatively few new industries and with few or no radical innovations to invest in those countries with many new industries and numerous radical innovations ndash as astrategy to be able to transfer new technologies from the latter institu-tional environment to the former

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

629

Table 4 summarizes the way that the norms rules and values of a society are associated with Americarsquos and Germanyrsquos institutionalarrangements their social system of production and their styles of inno-vativeness

CONCLUSION

Several themes deserve to be reemphasized Most importantly institu-tional analysis is emerging as a dominant issue in several social sciencedisciplines The study of various institutional components discussedabove is a complex subject requiring multiple levels of analysis and aknowledge of multiple sectors within societies It requires a theoreticalperspective that evolves from an empirical analysis in different societiesIf we are to advance the study of institutional analysis we need cross-disciplinary collaboration and team research which are not readilyachieved given the structure of most of our research universitiesAt present scholars in each academic discipline have their own distinc-

tive strategies for studying institutional analysis Thus economists tendto use several approaches (North 1990 Williamson 1985 Hodgson 19881998 1999) anthropologists another (Geertz 1995) and sociologistspolitical scientists and historians employ other strategies (Campbell etal 1991 Powell and DiMaggio 1991 Katznelson 1998 Hall and Taylor1996) With practitioners in various disciplines pursuing their distinctiveapproaches to institutional analysis there is low potential for collabo-rative institutional analysis across the social sciences The lessons fromthe biological sciences over the last 40 years are very instructive oncepractitioners from various biological disciplines began to cooperate instudying biological phenomena by using the same concepts at the mole-cular level theoretical advance was very rapid Comparable phenomenatook place in the development of other hybrid elds eg biochemistrybiophysics (Judson 1979 Olby 1979 Kohler 1982 1994) Similarly weresocial scientists in various disciplines to work collaboratively in the areaof institutional analysis there would very likely be an acceleration intheir theoretical knowledgeThis is not a plea that everyone engaged in institutional analysis should

do the same kind of research Indeed just as biologists working at themolecular level study many different kinds of problems similarly socialscientists engaged in institutional analysis would also work on manyseparate problems but in the spirit of a collective enterprise Some wouldconcentrate their attention on the study of rules norms habits conven-tions and values while others would study how these are associatedwith congurations of various institutional arrangements (types ofmarkets hierarchies networks associations communities clans statesetc) Others would work on specic institutional sectors (eg education

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

630

Table 4

Rules norms hab

its an

d values associated w

ith sp

ecic co

ngu

ration

s of coo

rdinating mecha

nism

s soc

ial sy

stem

s of

prod

uction

an

d nationa

l styles of inno

vation

USA

Germany

Norms rules values which provide incentives for

Norms rules values which provide incentives for

short-term horizons

long-term horizons

Con

gu

rative

forms of

Marke

ts c

orpo

rate hierarchies reg

ulatory state

Assoc

iation

s (busine

ss assoc

iation

s la

bor union

s)

coordinating

corp

oratist-type

networks

marke

ts c

orporatist-

mecha

nism

s (listed in

type

state

order of im

portanc

e in

the particu

lar social

system

)

Social sys

tem of

prod

uction

Bus

iness system

Firm

struc

tures

Con

glom

erates b

ut with mov

emen

t toward

Firm

more co

ngru

ent with pa

rticular prod

ucts

elim

ination of produc

tion

in som

e sectors

emph

asis on long

-term strateg

ies rm

s are well

emph

asis on sh

ort-term

strateg

ies high

integrated

differentiation

between divisions

of rms

Firm

s mob

ilized in

Low

High

and in

tegrated

into busine

ss

associations

Indu

strial relations

Low

deg

ree of jo

b secu

rity rigid in

ternal lab

or

High deg

ree of jo

b secu

rity

exible interna

l lab

or

system

marke

t ex

ible externa

l labo

r marke

tsmarke

t rigid externa

l lab

or m

arke

tPatterns of

Dispersed owne

rship frequ

ent turnov

erMore stab

le owne

rship patterns

owne

rship

Table 4

continue

d

USA

Germany

Norms rules values which provide incentives for

Norms rules values which provide incentives for

short-term horizons

long-term horizons

Social system of

Prod

uction

Bus

iness system

Training for labo

r an

dHigh deg

ree of voluntary decision mak

ing by

Collective an

d m

ore co

mpulso

ry d

ecisions

abo

ut

man

agem

ent

individuals as to ho

w m

uch

tim

e an

d ene

rgy

how m

uch ed

ucation

is ne

eded

he

avy em

pha

sis

to inv

est in educ

ation heavy

emph

asis on

on techn

ical training (eg en

gine

ering)

busine

ss sch

ools training

in marke

ting

and

sales accou

nting

Fina

ncial marke

tsEqu

ity an

d ven

ture cap

ital m

arke

ts highly

Equ

ity an

d ven

ture cap

ital m

arke

ts w

eakly

dev

elop

ed na

ncing from

cen

tral gov

ernm

ent

deve

lope

d he

avy em

pha

sis on

corporate loan

sto priva

te rm

s for dev

elop

men

t of new

tech

nologies related

to na

tion

al security

Unive

rsity research

High individual auton

omy high poten

tial for

Low

ind

ividua

l au

tono

my until almost 40

yea

rs of

system

individual creativity u

nive

rsity pr

ofesso

rs

age poten

tial for ind

ividual creativity high

ly

high

ly entreprene

urial

cons

traine

d unive

rsity pr

ofessors not so

entrep

rene

urial

Inno

vative

styles

Highly su

ccessful in m

ajor break

through

s in

High pe

rforman

ce in proc

ess an

d inc

remen

tal

basic an

d app

lied scien

ce go

od in dev

elop

ing

inno

vation

s in existing produ

cts not so su

ccessful

totally

new

lines of pr

oducts an

d in m

ajor break

through

s in basic scien

ce or in

tech

nologies high

deg

ree of entrepr

eneu

rship

deve

loping

totally new

lines of prod

ucts of

amon

g research

ers

tech

nologies low d

egree of entrepr

eneu

rship

amon

g research

ers

Table 4

continue

d

USA

Germany

Norms rules values which provide incentives for

Norms rules values which provide incentives for

short-term horizons

long-term horizons

Prod

ucts in w

hich

Pr

oducts with sh

ort liv

es prod

ucts in w

hich

Older produ

cts with long

history machine

ry pap

er

society excels

tech

nology

is bo

th highly co

mplex

and

pr

inting

textile

s m

otor veh

icles m

achine

too

lsch

ange

s rapidly en

tertainm

ent indus

try

pha

rmaceu

ticals biotechn

olog

y softw

are

aerosp

ace

business systems nancial markets and systems of research) Distinctivewould be the recognition of how these separate studies are linkedtogether how each institutional sector is linked to a societyrsquos normsrules values etc and its conguration of institutional arrangementsSimilarly other institutional analysts might focus on the study of orga-nizations with a major concern for how the institutional environmentof organizations (eg norms rules values congurations of institutionalarrangements institutional sectors) inuenced the structure culture andoutcomes of organizations And nally others might study how all ofthe aforementioned aspects of institutional analysis inuenced a societyrsquosoverall performance Nelson and Sampat (1998) have appropriatelyreminded us that we should engage in institutional analysis in order toaddress specic problems but the view developed in this article is thatwe rst need some mapping of the terrain of institutional analysis inorder for our diverse enterprises to become a meaningful collective enter-prise especially if we are to understand how institutions inuence theinnovative processThe argument developed here is not that all social scientists should engage

in institutional analysis Obviously there are numerous other importantareas of research which lie beyond the eld of institutional analysis Nordoes this article argue that institutional analysis is the most importantsubject for social scientistsSince the structure and culture of our research universities reect an

enormous amount of disciplinary fragmentation what are the prospectsfor promoting the kind of collective and interdisciplinary research agendaproposed above In one respect the prospects are not encouragingDuring the past half century the efforts of American universities to have effective communication across the biological sciences have variedgreatly in their success At the University of California-Berkeley it was not until the 1980s that there was a serious effort to integrate thebiological sciences and at that time the campus abolished seventeendepartments to promote an integrated biological program But at theUniversity of Wisconsin the biological sciences are still very fragmentedinto dozens of different departments in botany plant pathology zoologyand physiology with two departments of biochemistry and two depart-ments of geneticsThere have been other approaches which have worked very well histor-

ically for promoting scientic integration amidst scientic diversity Forexample several summer schools and special institutes were organizedin the United States during the 1940s in an effort to overcome the frag-mentation in the biological sciences These programs made it possiblefor scientists in diverse academic disciplines and from different univer-sities and countries to come together to engage in intense and frequentinteraction Perhaps the most effective program was that held in the

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

634

summers at Cold Spring Harbor New York where a research agendafor the study of molecular biology was developed (Fischer and Lipson1988 Stent and Watson 1966) Scientists went there summer aftersummer and when they returned to their universities many began therestructuring of the biological sciences in their universitiesIn our own day we need summer institutes and other similar programs

so that scholars from diverse backgrounds can come together anddevelop a collective agenda for doing institutional analysis and forstudying the impact of institutions on the performance of societies ndashinnovativeness being one such subject If we can develop such instituteswe will increase the prospects for transforming our universities Thereis reason to be optimistic that once an eclectic group of scholars workwith a common set of concepts we can have genuine theoretical advanceif researchers can bring their diverse expertise to bear on a common setof problems There is considerable evidence that institutional analysisoffers high potential for advancing our knowledge of innovations andtechnological change (Hollingsworth and Hollingsworth 2000) If we aregoing to make substantial theoretical advance it is not likely to occurfrom the vantage point of a multiplicity of disciplines Rather it is likelyto occur as we develop a new intellectual framework and our own acad-emic journals This is a tall agenda which requires us to reassess theway that knowledge is presently produced in our societies

NOTE

This article is part of a much larger agenda involving the study of institutionsorganizations and innovations in which I am involved My debts are many andvaried The rst version was written while I was in residence in the ResearchUnit for Institutional Change and European Integration of the Austrian Academyof Sciences in Vienna in the summer of 1998 I am especially grateful to ProfessorsEgon Matzner and Sonja Puntscher-Riekmann for providing the stimulating envi-ronment in which to think through the ideas developed here Egon Matzner hasengaged in very stimulating discussions with me about the importance of insti-tutions in the general eld of socio-economics The nal version of the articlewas completed during my residence as a fellow at the Neurosciences Institutein La Jolla California I also wish to thank my colleagues in the InnovationTheme Project at the Netherlands Institute for Advanced Study (NIAS) during1998ndash9 and Edgar Grande and Karl Muumlller for stimulating conversations aboutthe interaction of institutions and organizations in facilitating particular typesof innovation I am especially grateful to Jerald Hage for helping me to developa number of issues in the article and for his useful comments on the rst draftFrans van Waarden early on helped me to work through the idea of multiplelevels of institutional analysis Others who have helped to develop the ideas inthe article have been my colleagues in the research project lsquoComparative socialsystems of productionrsquo especially Tom Burns Christel Lane Yoshitaka OkadaWolfgang Streeck and Richard Whitley My colleagues Robert Boyer SteveCasper and David Soskice have taught me much about how incentives providedby the institutional environments of rms inuence individual decision making

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

635

But my most important debt is to Ellen Jane Hollingsworth who has repeatedlyquestioned every single assumption of this article Without her rich and contin-uous tutoring this article would be even less developed than it is at presentFinally without the insightful assistance of David Gear in many phases of thisarticle it would not have been completed

REFERENCES

Aoki Masahiko (1990) lsquoTowards an economic model of the Japanese rmrsquo Journalof Economic Literature March 1ndash27

Arora A Landau Ralph and Rosenberg Nathan (eds) (1998) Chemicals and Long-Term Economic Growth Insights from the Chemical Industry New York JohnWiley amp Sons

Arrow Kenneth (1974) The Limits of Organization New York NortonArthur W B (1988a) lsquoSelf-reinforcing mechanisms in economicsrsquo in P W

Anderson K J Arrow and D Pines (eds) The Economy as an Evolving ComplexSystem Redwood City Calif Addison-Wesley

mdashmdash(1988b) lsquoCompeting technologies an overviewrsquo in G Dosi C Freeman RNelson G Silverberg and L Soete (eds) Technical Change and Economic TheoryLondon Pinter Publishers pp 590ndash607

Audretsch David B (1995) Innovation and Industry Evolution Cambridge MassMIT Press

Audretsch David B and Feldman Maryann P (1995) lsquoInnovative clusters andthe industry life cyclersquo Discussion Paper Wissenschaftszentrum BerlinMarket Processes and Corporate Development Research Unit FS-IV-95-7

Baum Joel A C and Oliver Christine (1992) lsquoInstitutional embeddedness andthe dynamics of organizational populationsrsquo American Sociological Review 57540ndash59

Baum Joel A C and Singh Jitendra V (1994a) lsquoOrganizationndashenvironment coevo-lutionrsquo in Joel A C Baum and Jitendra V Singh (eds) The EvolutionaryDynamics of Organizations New York Oxford University Press pp 379ndash402

mdashmdash(eds) (1994b) The Evolutionary Dynamics of Organizations New York OxfordUniversity Press

Berger Suzanne and Dore Ronald (eds) (1996) National Diversity and GlobalCapitalism Ithaca NY Cornell University Press

Boyer Robert (1997) lsquoThe variety and unequal performance of really existingmarkets farewell to Doctor Panglossrsquo in J Rogers Hollingsworth and RobertBoyer (eds) Contemporary Capitalism the Embeddedness of InstitutionsCambridge and New York Cambridge University Press pp 55ndash93

Burns Tom R and Dietz Thomas (2001 forthcoming) lsquoSocial rule system theorysocial action institutional arrangements and evolutionary processesrsquo in JRogers Hollingsworth and Karl Muumlller (eds) Advances in SocioeconomicsLanham Md Rowman amp Littleeld

Burns Tom R and Flam Helena (1987) The Shaping of Social Organization BeverlyHills Calif Sage

Calvert Randall L (1995) lsquoThe rational choice theory of social institutionsrsquo inJeffrey S Banks and Eric A Hanuskek (eds) Modern Political Economy NewYork Cambridge University Press pp 216ndash66

Camic Charles (1986) lsquoThe matter of habitrsquo American Journal of Sociology 911039ndash87

Campbell John L (1997) lsquoMechanisms of evolutionary change in economicsgovernance interaction interpretation and bricolagersquo in Lars Magnusson

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

636

innovative disadvantages relative to the Americans (see Audretsch 1995Winter 1982 Casper 2000)On both sides of the Atlantic universities engage in basic and applied

research But signicantly the institutional environment in which univer-sities are embedded inuences the degree to which the research is likelyto represent major breakthroughs or is likely to have more modestimplications Recent research in biomedical science demonstrates that inthose societies in which university professors andor graduate studentshave the following characteristics there are likely to be relatively majorresearch breakthroughs graduate students have a long training periodhigh dependency on a major professor low autonomy in dening thenature of a research project restricted job options and a university envi-ronment where there is a very modest or nonexistent entrepreneurialculture For example in Germany ndash a country with relatively few radicalbreakthroughs ndash the scholar who nishes a habilitation will have spentmany years on a research project before qualifying for a permanent acad-emic position will have been in a high-dependency position with a seniorprofessor will have had limited autonomy to dene a research projectand will be constrained from pursuing a high-risk research project Theyoung German will have an extraordinarily high level of training butwill have faced a number of disincentives to undertake high-risk researchAnd as soon as heshe has completed the habilitation and becomes aprofessor administrative responsibilities constitute another disincentivefrom engaging in high-risk or long-term research projects Moreover theGerman professor is a civil servant with a more stable and secure incomethan hisher American counterpart In a very different institutionalenvironment American professors in large research universities tend tobecome much more entrepreneurial Today many American professorsbehave as though they were operating their own rms within univer-sities and thus devote considerable time to generating funds for researchprojects assistants secretaries supplies etc And the variation in thisentrepreneurial spirit across countries is associated with variation in thetaking of high-risk research strategies and the making of major researchbreakthroughsIn sum societies vary in terms of the institutional environment in

which their organizations are embedded and this variation inuencesthe degree to which they have industrial activities in the early or latestages of the industry life cycle successful new or mature industriesand radical vs incremental innovations And it is this variation whichis encouraging rms in countries with relatively few new industries and with few or no radical innovations to invest in those countries with many new industries and numerous radical innovations ndash as astrategy to be able to transfer new technologies from the latter institu-tional environment to the former

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

629

Table 4 summarizes the way that the norms rules and values of a society are associated with Americarsquos and Germanyrsquos institutionalarrangements their social system of production and their styles of inno-vativeness

CONCLUSION

Several themes deserve to be reemphasized Most importantly institu-tional analysis is emerging as a dominant issue in several social sciencedisciplines The study of various institutional components discussedabove is a complex subject requiring multiple levels of analysis and aknowledge of multiple sectors within societies It requires a theoreticalperspective that evolves from an empirical analysis in different societiesIf we are to advance the study of institutional analysis we need cross-disciplinary collaboration and team research which are not readilyachieved given the structure of most of our research universitiesAt present scholars in each academic discipline have their own distinc-

tive strategies for studying institutional analysis Thus economists tendto use several approaches (North 1990 Williamson 1985 Hodgson 19881998 1999) anthropologists another (Geertz 1995) and sociologistspolitical scientists and historians employ other strategies (Campbell etal 1991 Powell and DiMaggio 1991 Katznelson 1998 Hall and Taylor1996) With practitioners in various disciplines pursuing their distinctiveapproaches to institutional analysis there is low potential for collabo-rative institutional analysis across the social sciences The lessons fromthe biological sciences over the last 40 years are very instructive oncepractitioners from various biological disciplines began to cooperate instudying biological phenomena by using the same concepts at the mole-cular level theoretical advance was very rapid Comparable phenomenatook place in the development of other hybrid elds eg biochemistrybiophysics (Judson 1979 Olby 1979 Kohler 1982 1994) Similarly weresocial scientists in various disciplines to work collaboratively in the areaof institutional analysis there would very likely be an acceleration intheir theoretical knowledgeThis is not a plea that everyone engaged in institutional analysis should

do the same kind of research Indeed just as biologists working at themolecular level study many different kinds of problems similarly socialscientists engaged in institutional analysis would also work on manyseparate problems but in the spirit of a collective enterprise Some wouldconcentrate their attention on the study of rules norms habits conven-tions and values while others would study how these are associatedwith congurations of various institutional arrangements (types ofmarkets hierarchies networks associations communities clans statesetc) Others would work on specic institutional sectors (eg education

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

630

Table 4

Rules norms hab

its an

d values associated w

ith sp

ecic co

ngu

ration

s of coo

rdinating mecha

nism

s soc

ial sy

stem

s of

prod

uction

an

d nationa

l styles of inno

vation

USA

Germany

Norms rules values which provide incentives for

Norms rules values which provide incentives for

short-term horizons

long-term horizons

Con

gu

rative

forms of

Marke

ts c

orpo

rate hierarchies reg

ulatory state

Assoc

iation

s (busine

ss assoc

iation

s la

bor union

s)

coordinating

corp

oratist-type

networks

marke

ts c

orporatist-

mecha

nism

s (listed in

type

state

order of im

portanc

e in

the particu

lar social

system

)

Social sys

tem of

prod

uction

Bus

iness system

Firm

struc

tures

Con

glom

erates b

ut with mov

emen

t toward

Firm

more co

ngru

ent with pa

rticular prod

ucts

elim

ination of produc

tion

in som

e sectors

emph

asis on long

-term strateg

ies rm

s are well

emph

asis on sh

ort-term

strateg

ies high

integrated

differentiation

between divisions

of rms

Firm

s mob

ilized in

Low

High

and in

tegrated

into busine

ss

associations

Indu

strial relations

Low

deg

ree of jo

b secu

rity rigid in

ternal lab

or

High deg

ree of jo

b secu

rity

exible interna

l lab

or

system

marke

t ex

ible externa

l labo

r marke

tsmarke

t rigid externa

l lab

or m

arke

tPatterns of

Dispersed owne

rship frequ

ent turnov

erMore stab

le owne

rship patterns

owne

rship

Table 4

continue

d

USA

Germany

Norms rules values which provide incentives for

Norms rules values which provide incentives for

short-term horizons

long-term horizons

Social system of

Prod

uction

Bus

iness system

Training for labo

r an

dHigh deg

ree of voluntary decision mak

ing by

Collective an

d m

ore co

mpulso

ry d

ecisions

abo

ut

man

agem

ent

individuals as to ho

w m

uch

tim

e an

d ene

rgy

how m

uch ed

ucation

is ne

eded

he

avy em

pha

sis

to inv

est in educ

ation heavy

emph

asis on

on techn

ical training (eg en

gine

ering)

busine

ss sch

ools training

in marke

ting

and

sales accou

nting

Fina

ncial marke

tsEqu

ity an

d ven

ture cap

ital m

arke

ts highly

Equ

ity an

d ven

ture cap

ital m

arke

ts w

eakly

dev

elop

ed na

ncing from

cen

tral gov

ernm

ent

deve

lope

d he

avy em

pha

sis on

corporate loan

sto priva

te rm

s for dev

elop

men

t of new

tech

nologies related

to na

tion

al security

Unive

rsity research

High individual auton

omy high poten

tial for

Low

ind

ividua

l au

tono

my until almost 40

yea

rs of

system

individual creativity u

nive

rsity pr

ofesso

rs

age poten

tial for ind

ividual creativity high

ly

high

ly entreprene

urial

cons

traine

d unive

rsity pr

ofessors not so

entrep

rene

urial

Inno

vative

styles

Highly su

ccessful in m

ajor break

through

s in

High pe

rforman

ce in proc

ess an

d inc

remen

tal

basic an

d app

lied scien

ce go

od in dev

elop

ing

inno

vation

s in existing produ

cts not so su

ccessful

totally

new

lines of pr

oducts an

d in m

ajor break

through

s in basic scien

ce or in

tech

nologies high

deg

ree of entrepr

eneu

rship

deve

loping

totally new

lines of prod

ucts of

amon

g research

ers

tech

nologies low d

egree of entrepr

eneu

rship

amon

g research

ers

Table 4

continue

d

USA

Germany

Norms rules values which provide incentives for

Norms rules values which provide incentives for

short-term horizons

long-term horizons

Prod

ucts in w

hich

Pr

oducts with sh

ort liv

es prod

ucts in w

hich

Older produ

cts with long

history machine

ry pap

er

society excels

tech

nology

is bo

th highly co

mplex

and

pr

inting

textile

s m

otor veh

icles m

achine

too

lsch

ange

s rapidly en

tertainm

ent indus

try

pha

rmaceu

ticals biotechn

olog

y softw

are

aerosp

ace

business systems nancial markets and systems of research) Distinctivewould be the recognition of how these separate studies are linkedtogether how each institutional sector is linked to a societyrsquos normsrules values etc and its conguration of institutional arrangementsSimilarly other institutional analysts might focus on the study of orga-nizations with a major concern for how the institutional environmentof organizations (eg norms rules values congurations of institutionalarrangements institutional sectors) inuenced the structure culture andoutcomes of organizations And nally others might study how all ofthe aforementioned aspects of institutional analysis inuenced a societyrsquosoverall performance Nelson and Sampat (1998) have appropriatelyreminded us that we should engage in institutional analysis in order toaddress specic problems but the view developed in this article is thatwe rst need some mapping of the terrain of institutional analysis inorder for our diverse enterprises to become a meaningful collective enter-prise especially if we are to understand how institutions inuence theinnovative processThe argument developed here is not that all social scientists should engage

in institutional analysis Obviously there are numerous other importantareas of research which lie beyond the eld of institutional analysis Nordoes this article argue that institutional analysis is the most importantsubject for social scientistsSince the structure and culture of our research universities reect an

enormous amount of disciplinary fragmentation what are the prospectsfor promoting the kind of collective and interdisciplinary research agendaproposed above In one respect the prospects are not encouragingDuring the past half century the efforts of American universities to have effective communication across the biological sciences have variedgreatly in their success At the University of California-Berkeley it was not until the 1980s that there was a serious effort to integrate thebiological sciences and at that time the campus abolished seventeendepartments to promote an integrated biological program But at theUniversity of Wisconsin the biological sciences are still very fragmentedinto dozens of different departments in botany plant pathology zoologyand physiology with two departments of biochemistry and two depart-ments of geneticsThere have been other approaches which have worked very well histor-

ically for promoting scientic integration amidst scientic diversity Forexample several summer schools and special institutes were organizedin the United States during the 1940s in an effort to overcome the frag-mentation in the biological sciences These programs made it possiblefor scientists in diverse academic disciplines and from different univer-sities and countries to come together to engage in intense and frequentinteraction Perhaps the most effective program was that held in the

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

634

summers at Cold Spring Harbor New York where a research agendafor the study of molecular biology was developed (Fischer and Lipson1988 Stent and Watson 1966) Scientists went there summer aftersummer and when they returned to their universities many began therestructuring of the biological sciences in their universitiesIn our own day we need summer institutes and other similar programs

so that scholars from diverse backgrounds can come together anddevelop a collective agenda for doing institutional analysis and forstudying the impact of institutions on the performance of societies ndashinnovativeness being one such subject If we can develop such instituteswe will increase the prospects for transforming our universities Thereis reason to be optimistic that once an eclectic group of scholars workwith a common set of concepts we can have genuine theoretical advanceif researchers can bring their diverse expertise to bear on a common setof problems There is considerable evidence that institutional analysisoffers high potential for advancing our knowledge of innovations andtechnological change (Hollingsworth and Hollingsworth 2000) If we aregoing to make substantial theoretical advance it is not likely to occurfrom the vantage point of a multiplicity of disciplines Rather it is likelyto occur as we develop a new intellectual framework and our own acad-emic journals This is a tall agenda which requires us to reassess theway that knowledge is presently produced in our societies

NOTE

This article is part of a much larger agenda involving the study of institutionsorganizations and innovations in which I am involved My debts are many andvaried The rst version was written while I was in residence in the ResearchUnit for Institutional Change and European Integration of the Austrian Academyof Sciences in Vienna in the summer of 1998 I am especially grateful to ProfessorsEgon Matzner and Sonja Puntscher-Riekmann for providing the stimulating envi-ronment in which to think through the ideas developed here Egon Matzner hasengaged in very stimulating discussions with me about the importance of insti-tutions in the general eld of socio-economics The nal version of the articlewas completed during my residence as a fellow at the Neurosciences Institutein La Jolla California I also wish to thank my colleagues in the InnovationTheme Project at the Netherlands Institute for Advanced Study (NIAS) during1998ndash9 and Edgar Grande and Karl Muumlller for stimulating conversations aboutthe interaction of institutions and organizations in facilitating particular typesof innovation I am especially grateful to Jerald Hage for helping me to developa number of issues in the article and for his useful comments on the rst draftFrans van Waarden early on helped me to work through the idea of multiplelevels of institutional analysis Others who have helped to develop the ideas inthe article have been my colleagues in the research project lsquoComparative socialsystems of productionrsquo especially Tom Burns Christel Lane Yoshitaka OkadaWolfgang Streeck and Richard Whitley My colleagues Robert Boyer SteveCasper and David Soskice have taught me much about how incentives providedby the institutional environments of rms inuence individual decision making

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

635

But my most important debt is to Ellen Jane Hollingsworth who has repeatedlyquestioned every single assumption of this article Without her rich and contin-uous tutoring this article would be even less developed than it is at presentFinally without the insightful assistance of David Gear in many phases of thisarticle it would not have been completed

REFERENCES

Aoki Masahiko (1990) lsquoTowards an economic model of the Japanese rmrsquo Journalof Economic Literature March 1ndash27

Arora A Landau Ralph and Rosenberg Nathan (eds) (1998) Chemicals and Long-Term Economic Growth Insights from the Chemical Industry New York JohnWiley amp Sons

Arrow Kenneth (1974) The Limits of Organization New York NortonArthur W B (1988a) lsquoSelf-reinforcing mechanisms in economicsrsquo in P W

Anderson K J Arrow and D Pines (eds) The Economy as an Evolving ComplexSystem Redwood City Calif Addison-Wesley

mdashmdash(1988b) lsquoCompeting technologies an overviewrsquo in G Dosi C Freeman RNelson G Silverberg and L Soete (eds) Technical Change and Economic TheoryLondon Pinter Publishers pp 590ndash607

Audretsch David B (1995) Innovation and Industry Evolution Cambridge MassMIT Press

Audretsch David B and Feldman Maryann P (1995) lsquoInnovative clusters andthe industry life cyclersquo Discussion Paper Wissenschaftszentrum BerlinMarket Processes and Corporate Development Research Unit FS-IV-95-7

Baum Joel A C and Oliver Christine (1992) lsquoInstitutional embeddedness andthe dynamics of organizational populationsrsquo American Sociological Review 57540ndash59

Baum Joel A C and Singh Jitendra V (1994a) lsquoOrganizationndashenvironment coevo-lutionrsquo in Joel A C Baum and Jitendra V Singh (eds) The EvolutionaryDynamics of Organizations New York Oxford University Press pp 379ndash402

mdashmdash(eds) (1994b) The Evolutionary Dynamics of Organizations New York OxfordUniversity Press

Berger Suzanne and Dore Ronald (eds) (1996) National Diversity and GlobalCapitalism Ithaca NY Cornell University Press

Boyer Robert (1997) lsquoThe variety and unequal performance of really existingmarkets farewell to Doctor Panglossrsquo in J Rogers Hollingsworth and RobertBoyer (eds) Contemporary Capitalism the Embeddedness of InstitutionsCambridge and New York Cambridge University Press pp 55ndash93

Burns Tom R and Dietz Thomas (2001 forthcoming) lsquoSocial rule system theorysocial action institutional arrangements and evolutionary processesrsquo in JRogers Hollingsworth and Karl Muumlller (eds) Advances in SocioeconomicsLanham Md Rowman amp Littleeld

Burns Tom R and Flam Helena (1987) The Shaping of Social Organization BeverlyHills Calif Sage

Calvert Randall L (1995) lsquoThe rational choice theory of social institutionsrsquo inJeffrey S Banks and Eric A Hanuskek (eds) Modern Political Economy NewYork Cambridge University Press pp 216ndash66

Camic Charles (1986) lsquoThe matter of habitrsquo American Journal of Sociology 911039ndash87

Campbell John L (1997) lsquoMechanisms of evolutionary change in economicsgovernance interaction interpretation and bricolagersquo in Lars Magnusson

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

636

Table 4 summarizes the way that the norms rules and values of a society are associated with Americarsquos and Germanyrsquos institutionalarrangements their social system of production and their styles of inno-vativeness

CONCLUSION

Several themes deserve to be reemphasized Most importantly institu-tional analysis is emerging as a dominant issue in several social sciencedisciplines The study of various institutional components discussedabove is a complex subject requiring multiple levels of analysis and aknowledge of multiple sectors within societies It requires a theoreticalperspective that evolves from an empirical analysis in different societiesIf we are to advance the study of institutional analysis we need cross-disciplinary collaboration and team research which are not readilyachieved given the structure of most of our research universitiesAt present scholars in each academic discipline have their own distinc-

tive strategies for studying institutional analysis Thus economists tendto use several approaches (North 1990 Williamson 1985 Hodgson 19881998 1999) anthropologists another (Geertz 1995) and sociologistspolitical scientists and historians employ other strategies (Campbell etal 1991 Powell and DiMaggio 1991 Katznelson 1998 Hall and Taylor1996) With practitioners in various disciplines pursuing their distinctiveapproaches to institutional analysis there is low potential for collabo-rative institutional analysis across the social sciences The lessons fromthe biological sciences over the last 40 years are very instructive oncepractitioners from various biological disciplines began to cooperate instudying biological phenomena by using the same concepts at the mole-cular level theoretical advance was very rapid Comparable phenomenatook place in the development of other hybrid elds eg biochemistrybiophysics (Judson 1979 Olby 1979 Kohler 1982 1994) Similarly weresocial scientists in various disciplines to work collaboratively in the areaof institutional analysis there would very likely be an acceleration intheir theoretical knowledgeThis is not a plea that everyone engaged in institutional analysis should

do the same kind of research Indeed just as biologists working at themolecular level study many different kinds of problems similarly socialscientists engaged in institutional analysis would also work on manyseparate problems but in the spirit of a collective enterprise Some wouldconcentrate their attention on the study of rules norms habits conven-tions and values while others would study how these are associatedwith congurations of various institutional arrangements (types ofmarkets hierarchies networks associations communities clans statesetc) Others would work on specic institutional sectors (eg education

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

630

Table 4

Rules norms hab

its an

d values associated w

ith sp

ecic co

ngu

ration

s of coo

rdinating mecha

nism

s soc

ial sy

stem

s of

prod

uction

an

d nationa

l styles of inno

vation

USA

Germany

Norms rules values which provide incentives for

Norms rules values which provide incentives for

short-term horizons

long-term horizons

Con

gu

rative

forms of

Marke

ts c

orpo

rate hierarchies reg

ulatory state

Assoc

iation

s (busine

ss assoc

iation

s la

bor union

s)

coordinating

corp

oratist-type

networks

marke

ts c

orporatist-

mecha

nism

s (listed in

type

state

order of im

portanc

e in

the particu

lar social

system

)

Social sys

tem of

prod

uction

Bus

iness system

Firm

struc

tures

Con

glom

erates b

ut with mov

emen

t toward

Firm

more co

ngru

ent with pa

rticular prod

ucts

elim

ination of produc

tion

in som

e sectors

emph

asis on long

-term strateg

ies rm

s are well

emph

asis on sh

ort-term

strateg

ies high

integrated

differentiation

between divisions

of rms

Firm

s mob

ilized in

Low

High

and in

tegrated

into busine

ss

associations

Indu

strial relations

Low

deg

ree of jo

b secu

rity rigid in

ternal lab

or

High deg

ree of jo

b secu

rity

exible interna

l lab

or

system

marke

t ex

ible externa

l labo

r marke

tsmarke

t rigid externa

l lab

or m

arke

tPatterns of

Dispersed owne

rship frequ

ent turnov

erMore stab

le owne

rship patterns

owne

rship

Table 4

continue

d

USA

Germany

Norms rules values which provide incentives for

Norms rules values which provide incentives for

short-term horizons

long-term horizons

Social system of

Prod

uction

Bus

iness system

Training for labo

r an

dHigh deg

ree of voluntary decision mak

ing by

Collective an

d m

ore co

mpulso

ry d

ecisions

abo

ut

man

agem

ent

individuals as to ho

w m

uch

tim

e an

d ene

rgy

how m

uch ed

ucation

is ne

eded

he

avy em

pha

sis

to inv

est in educ

ation heavy

emph

asis on

on techn

ical training (eg en

gine

ering)

busine

ss sch

ools training

in marke

ting

and

sales accou

nting

Fina

ncial marke

tsEqu

ity an

d ven

ture cap

ital m

arke

ts highly

Equ

ity an

d ven

ture cap

ital m

arke

ts w

eakly

dev

elop

ed na

ncing from

cen

tral gov

ernm

ent

deve

lope

d he

avy em

pha

sis on

corporate loan

sto priva

te rm

s for dev

elop

men

t of new

tech

nologies related

to na

tion

al security

Unive

rsity research

High individual auton

omy high poten

tial for

Low

ind

ividua

l au

tono

my until almost 40

yea

rs of

system

individual creativity u

nive

rsity pr

ofesso

rs

age poten

tial for ind

ividual creativity high

ly

high

ly entreprene

urial

cons

traine

d unive

rsity pr

ofessors not so

entrep

rene

urial

Inno

vative

styles

Highly su

ccessful in m

ajor break

through

s in

High pe

rforman

ce in proc

ess an

d inc

remen

tal

basic an

d app

lied scien

ce go

od in dev

elop

ing

inno

vation

s in existing produ

cts not so su

ccessful

totally

new

lines of pr

oducts an

d in m

ajor break

through

s in basic scien

ce or in

tech

nologies high

deg

ree of entrepr

eneu

rship

deve

loping

totally new

lines of prod

ucts of

amon

g research

ers

tech

nologies low d

egree of entrepr

eneu

rship

amon

g research

ers

Table 4

continue

d

USA

Germany

Norms rules values which provide incentives for

Norms rules values which provide incentives for

short-term horizons

long-term horizons

Prod

ucts in w

hich

Pr

oducts with sh

ort liv

es prod

ucts in w

hich

Older produ

cts with long

history machine

ry pap

er

society excels

tech

nology

is bo

th highly co

mplex

and

pr

inting

textile

s m

otor veh

icles m

achine

too

lsch

ange

s rapidly en

tertainm

ent indus

try

pha

rmaceu

ticals biotechn

olog

y softw

are

aerosp

ace

business systems nancial markets and systems of research) Distinctivewould be the recognition of how these separate studies are linkedtogether how each institutional sector is linked to a societyrsquos normsrules values etc and its conguration of institutional arrangementsSimilarly other institutional analysts might focus on the study of orga-nizations with a major concern for how the institutional environmentof organizations (eg norms rules values congurations of institutionalarrangements institutional sectors) inuenced the structure culture andoutcomes of organizations And nally others might study how all ofthe aforementioned aspects of institutional analysis inuenced a societyrsquosoverall performance Nelson and Sampat (1998) have appropriatelyreminded us that we should engage in institutional analysis in order toaddress specic problems but the view developed in this article is thatwe rst need some mapping of the terrain of institutional analysis inorder for our diverse enterprises to become a meaningful collective enter-prise especially if we are to understand how institutions inuence theinnovative processThe argument developed here is not that all social scientists should engage

in institutional analysis Obviously there are numerous other importantareas of research which lie beyond the eld of institutional analysis Nordoes this article argue that institutional analysis is the most importantsubject for social scientistsSince the structure and culture of our research universities reect an

enormous amount of disciplinary fragmentation what are the prospectsfor promoting the kind of collective and interdisciplinary research agendaproposed above In one respect the prospects are not encouragingDuring the past half century the efforts of American universities to have effective communication across the biological sciences have variedgreatly in their success At the University of California-Berkeley it was not until the 1980s that there was a serious effort to integrate thebiological sciences and at that time the campus abolished seventeendepartments to promote an integrated biological program But at theUniversity of Wisconsin the biological sciences are still very fragmentedinto dozens of different departments in botany plant pathology zoologyand physiology with two departments of biochemistry and two depart-ments of geneticsThere have been other approaches which have worked very well histor-

ically for promoting scientic integration amidst scientic diversity Forexample several summer schools and special institutes were organizedin the United States during the 1940s in an effort to overcome the frag-mentation in the biological sciences These programs made it possiblefor scientists in diverse academic disciplines and from different univer-sities and countries to come together to engage in intense and frequentinteraction Perhaps the most effective program was that held in the

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

634

summers at Cold Spring Harbor New York where a research agendafor the study of molecular biology was developed (Fischer and Lipson1988 Stent and Watson 1966) Scientists went there summer aftersummer and when they returned to their universities many began therestructuring of the biological sciences in their universitiesIn our own day we need summer institutes and other similar programs

so that scholars from diverse backgrounds can come together anddevelop a collective agenda for doing institutional analysis and forstudying the impact of institutions on the performance of societies ndashinnovativeness being one such subject If we can develop such instituteswe will increase the prospects for transforming our universities Thereis reason to be optimistic that once an eclectic group of scholars workwith a common set of concepts we can have genuine theoretical advanceif researchers can bring their diverse expertise to bear on a common setof problems There is considerable evidence that institutional analysisoffers high potential for advancing our knowledge of innovations andtechnological change (Hollingsworth and Hollingsworth 2000) If we aregoing to make substantial theoretical advance it is not likely to occurfrom the vantage point of a multiplicity of disciplines Rather it is likelyto occur as we develop a new intellectual framework and our own acad-emic journals This is a tall agenda which requires us to reassess theway that knowledge is presently produced in our societies

NOTE

This article is part of a much larger agenda involving the study of institutionsorganizations and innovations in which I am involved My debts are many andvaried The rst version was written while I was in residence in the ResearchUnit for Institutional Change and European Integration of the Austrian Academyof Sciences in Vienna in the summer of 1998 I am especially grateful to ProfessorsEgon Matzner and Sonja Puntscher-Riekmann for providing the stimulating envi-ronment in which to think through the ideas developed here Egon Matzner hasengaged in very stimulating discussions with me about the importance of insti-tutions in the general eld of socio-economics The nal version of the articlewas completed during my residence as a fellow at the Neurosciences Institutein La Jolla California I also wish to thank my colleagues in the InnovationTheme Project at the Netherlands Institute for Advanced Study (NIAS) during1998ndash9 and Edgar Grande and Karl Muumlller for stimulating conversations aboutthe interaction of institutions and organizations in facilitating particular typesof innovation I am especially grateful to Jerald Hage for helping me to developa number of issues in the article and for his useful comments on the rst draftFrans van Waarden early on helped me to work through the idea of multiplelevels of institutional analysis Others who have helped to develop the ideas inthe article have been my colleagues in the research project lsquoComparative socialsystems of productionrsquo especially Tom Burns Christel Lane Yoshitaka OkadaWolfgang Streeck and Richard Whitley My colleagues Robert Boyer SteveCasper and David Soskice have taught me much about how incentives providedby the institutional environments of rms inuence individual decision making

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

635

But my most important debt is to Ellen Jane Hollingsworth who has repeatedlyquestioned every single assumption of this article Without her rich and contin-uous tutoring this article would be even less developed than it is at presentFinally without the insightful assistance of David Gear in many phases of thisarticle it would not have been completed

REFERENCES

Aoki Masahiko (1990) lsquoTowards an economic model of the Japanese rmrsquo Journalof Economic Literature March 1ndash27

Arora A Landau Ralph and Rosenberg Nathan (eds) (1998) Chemicals and Long-Term Economic Growth Insights from the Chemical Industry New York JohnWiley amp Sons

Arrow Kenneth (1974) The Limits of Organization New York NortonArthur W B (1988a) lsquoSelf-reinforcing mechanisms in economicsrsquo in P W

Anderson K J Arrow and D Pines (eds) The Economy as an Evolving ComplexSystem Redwood City Calif Addison-Wesley

mdashmdash(1988b) lsquoCompeting technologies an overviewrsquo in G Dosi C Freeman RNelson G Silverberg and L Soete (eds) Technical Change and Economic TheoryLondon Pinter Publishers pp 590ndash607

Audretsch David B (1995) Innovation and Industry Evolution Cambridge MassMIT Press

Audretsch David B and Feldman Maryann P (1995) lsquoInnovative clusters andthe industry life cyclersquo Discussion Paper Wissenschaftszentrum BerlinMarket Processes and Corporate Development Research Unit FS-IV-95-7

Baum Joel A C and Oliver Christine (1992) lsquoInstitutional embeddedness andthe dynamics of organizational populationsrsquo American Sociological Review 57540ndash59

Baum Joel A C and Singh Jitendra V (1994a) lsquoOrganizationndashenvironment coevo-lutionrsquo in Joel A C Baum and Jitendra V Singh (eds) The EvolutionaryDynamics of Organizations New York Oxford University Press pp 379ndash402

mdashmdash(eds) (1994b) The Evolutionary Dynamics of Organizations New York OxfordUniversity Press

Berger Suzanne and Dore Ronald (eds) (1996) National Diversity and GlobalCapitalism Ithaca NY Cornell University Press

Boyer Robert (1997) lsquoThe variety and unequal performance of really existingmarkets farewell to Doctor Panglossrsquo in J Rogers Hollingsworth and RobertBoyer (eds) Contemporary Capitalism the Embeddedness of InstitutionsCambridge and New York Cambridge University Press pp 55ndash93

Burns Tom R and Dietz Thomas (2001 forthcoming) lsquoSocial rule system theorysocial action institutional arrangements and evolutionary processesrsquo in JRogers Hollingsworth and Karl Muumlller (eds) Advances in SocioeconomicsLanham Md Rowman amp Littleeld

Burns Tom R and Flam Helena (1987) The Shaping of Social Organization BeverlyHills Calif Sage

Calvert Randall L (1995) lsquoThe rational choice theory of social institutionsrsquo inJeffrey S Banks and Eric A Hanuskek (eds) Modern Political Economy NewYork Cambridge University Press pp 216ndash66

Camic Charles (1986) lsquoThe matter of habitrsquo American Journal of Sociology 911039ndash87

Campbell John L (1997) lsquoMechanisms of evolutionary change in economicsgovernance interaction interpretation and bricolagersquo in Lars Magnusson

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

636

Table 4

Rules norms hab

its an

d values associated w

ith sp

ecic co

ngu

ration

s of coo

rdinating mecha

nism

s soc

ial sy

stem

s of

prod

uction

an

d nationa

l styles of inno

vation

USA

Germany

Norms rules values which provide incentives for

Norms rules values which provide incentives for

short-term horizons

long-term horizons

Con

gu

rative

forms of

Marke

ts c

orpo

rate hierarchies reg

ulatory state

Assoc

iation

s (busine

ss assoc

iation

s la

bor union

s)

coordinating

corp

oratist-type

networks

marke

ts c

orporatist-

mecha

nism

s (listed in

type

state

order of im

portanc

e in

the particu

lar social

system

)

Social sys

tem of

prod

uction

Bus

iness system

Firm

struc

tures

Con

glom

erates b

ut with mov

emen

t toward

Firm

more co

ngru

ent with pa

rticular prod

ucts

elim

ination of produc

tion

in som

e sectors

emph

asis on long

-term strateg

ies rm

s are well

emph

asis on sh

ort-term

strateg

ies high

integrated

differentiation

between divisions

of rms

Firm

s mob

ilized in

Low

High

and in

tegrated

into busine

ss

associations

Indu

strial relations

Low

deg

ree of jo

b secu

rity rigid in

ternal lab

or

High deg

ree of jo

b secu

rity

exible interna

l lab

or

system

marke

t ex

ible externa

l labo

r marke

tsmarke

t rigid externa

l lab

or m

arke

tPatterns of

Dispersed owne

rship frequ

ent turnov

erMore stab

le owne

rship patterns

owne

rship

Table 4

continue

d

USA

Germany

Norms rules values which provide incentives for

Norms rules values which provide incentives for

short-term horizons

long-term horizons

Social system of

Prod

uction

Bus

iness system

Training for labo

r an

dHigh deg

ree of voluntary decision mak

ing by

Collective an

d m

ore co

mpulso

ry d

ecisions

abo

ut

man

agem

ent

individuals as to ho

w m

uch

tim

e an

d ene

rgy

how m

uch ed

ucation

is ne

eded

he

avy em

pha

sis

to inv

est in educ

ation heavy

emph

asis on

on techn

ical training (eg en

gine

ering)

busine

ss sch

ools training

in marke

ting

and

sales accou

nting

Fina

ncial marke

tsEqu

ity an

d ven

ture cap

ital m

arke

ts highly

Equ

ity an

d ven

ture cap

ital m

arke

ts w

eakly

dev

elop

ed na

ncing from

cen

tral gov

ernm

ent

deve

lope

d he

avy em

pha

sis on

corporate loan

sto priva

te rm

s for dev

elop

men

t of new

tech

nologies related

to na

tion

al security

Unive

rsity research

High individual auton

omy high poten

tial for

Low

ind

ividua

l au

tono

my until almost 40

yea

rs of

system

individual creativity u

nive

rsity pr

ofesso

rs

age poten

tial for ind

ividual creativity high

ly

high

ly entreprene

urial

cons

traine

d unive

rsity pr

ofessors not so

entrep

rene

urial

Inno

vative

styles

Highly su

ccessful in m

ajor break

through

s in

High pe

rforman

ce in proc

ess an

d inc

remen

tal

basic an

d app

lied scien

ce go

od in dev

elop

ing

inno

vation

s in existing produ

cts not so su

ccessful

totally

new

lines of pr

oducts an

d in m

ajor break

through

s in basic scien

ce or in

tech

nologies high

deg

ree of entrepr

eneu

rship

deve

loping

totally new

lines of prod

ucts of

amon

g research

ers

tech

nologies low d

egree of entrepr

eneu

rship

amon

g research

ers

Table 4

continue

d

USA

Germany

Norms rules values which provide incentives for

Norms rules values which provide incentives for

short-term horizons

long-term horizons

Prod

ucts in w

hich

Pr

oducts with sh

ort liv

es prod

ucts in w

hich

Older produ

cts with long

history machine

ry pap

er

society excels

tech

nology

is bo

th highly co

mplex

and

pr

inting

textile

s m

otor veh

icles m

achine

too

lsch

ange

s rapidly en

tertainm

ent indus

try

pha

rmaceu

ticals biotechn

olog

y softw

are

aerosp

ace

business systems nancial markets and systems of research) Distinctivewould be the recognition of how these separate studies are linkedtogether how each institutional sector is linked to a societyrsquos normsrules values etc and its conguration of institutional arrangementsSimilarly other institutional analysts might focus on the study of orga-nizations with a major concern for how the institutional environmentof organizations (eg norms rules values congurations of institutionalarrangements institutional sectors) inuenced the structure culture andoutcomes of organizations And nally others might study how all ofthe aforementioned aspects of institutional analysis inuenced a societyrsquosoverall performance Nelson and Sampat (1998) have appropriatelyreminded us that we should engage in institutional analysis in order toaddress specic problems but the view developed in this article is thatwe rst need some mapping of the terrain of institutional analysis inorder for our diverse enterprises to become a meaningful collective enter-prise especially if we are to understand how institutions inuence theinnovative processThe argument developed here is not that all social scientists should engage

in institutional analysis Obviously there are numerous other importantareas of research which lie beyond the eld of institutional analysis Nordoes this article argue that institutional analysis is the most importantsubject for social scientistsSince the structure and culture of our research universities reect an

enormous amount of disciplinary fragmentation what are the prospectsfor promoting the kind of collective and interdisciplinary research agendaproposed above In one respect the prospects are not encouragingDuring the past half century the efforts of American universities to have effective communication across the biological sciences have variedgreatly in their success At the University of California-Berkeley it was not until the 1980s that there was a serious effort to integrate thebiological sciences and at that time the campus abolished seventeendepartments to promote an integrated biological program But at theUniversity of Wisconsin the biological sciences are still very fragmentedinto dozens of different departments in botany plant pathology zoologyand physiology with two departments of biochemistry and two depart-ments of geneticsThere have been other approaches which have worked very well histor-

ically for promoting scientic integration amidst scientic diversity Forexample several summer schools and special institutes were organizedin the United States during the 1940s in an effort to overcome the frag-mentation in the biological sciences These programs made it possiblefor scientists in diverse academic disciplines and from different univer-sities and countries to come together to engage in intense and frequentinteraction Perhaps the most effective program was that held in the

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

634

summers at Cold Spring Harbor New York where a research agendafor the study of molecular biology was developed (Fischer and Lipson1988 Stent and Watson 1966) Scientists went there summer aftersummer and when they returned to their universities many began therestructuring of the biological sciences in their universitiesIn our own day we need summer institutes and other similar programs

so that scholars from diverse backgrounds can come together anddevelop a collective agenda for doing institutional analysis and forstudying the impact of institutions on the performance of societies ndashinnovativeness being one such subject If we can develop such instituteswe will increase the prospects for transforming our universities Thereis reason to be optimistic that once an eclectic group of scholars workwith a common set of concepts we can have genuine theoretical advanceif researchers can bring their diverse expertise to bear on a common setof problems There is considerable evidence that institutional analysisoffers high potential for advancing our knowledge of innovations andtechnological change (Hollingsworth and Hollingsworth 2000) If we aregoing to make substantial theoretical advance it is not likely to occurfrom the vantage point of a multiplicity of disciplines Rather it is likelyto occur as we develop a new intellectual framework and our own acad-emic journals This is a tall agenda which requires us to reassess theway that knowledge is presently produced in our societies

NOTE

This article is part of a much larger agenda involving the study of institutionsorganizations and innovations in which I am involved My debts are many andvaried The rst version was written while I was in residence in the ResearchUnit for Institutional Change and European Integration of the Austrian Academyof Sciences in Vienna in the summer of 1998 I am especially grateful to ProfessorsEgon Matzner and Sonja Puntscher-Riekmann for providing the stimulating envi-ronment in which to think through the ideas developed here Egon Matzner hasengaged in very stimulating discussions with me about the importance of insti-tutions in the general eld of socio-economics The nal version of the articlewas completed during my residence as a fellow at the Neurosciences Institutein La Jolla California I also wish to thank my colleagues in the InnovationTheme Project at the Netherlands Institute for Advanced Study (NIAS) during1998ndash9 and Edgar Grande and Karl Muumlller for stimulating conversations aboutthe interaction of institutions and organizations in facilitating particular typesof innovation I am especially grateful to Jerald Hage for helping me to developa number of issues in the article and for his useful comments on the rst draftFrans van Waarden early on helped me to work through the idea of multiplelevels of institutional analysis Others who have helped to develop the ideas inthe article have been my colleagues in the research project lsquoComparative socialsystems of productionrsquo especially Tom Burns Christel Lane Yoshitaka OkadaWolfgang Streeck and Richard Whitley My colleagues Robert Boyer SteveCasper and David Soskice have taught me much about how incentives providedby the institutional environments of rms inuence individual decision making

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

635

But my most important debt is to Ellen Jane Hollingsworth who has repeatedlyquestioned every single assumption of this article Without her rich and contin-uous tutoring this article would be even less developed than it is at presentFinally without the insightful assistance of David Gear in many phases of thisarticle it would not have been completed

REFERENCES

Aoki Masahiko (1990) lsquoTowards an economic model of the Japanese rmrsquo Journalof Economic Literature March 1ndash27

Arora A Landau Ralph and Rosenberg Nathan (eds) (1998) Chemicals and Long-Term Economic Growth Insights from the Chemical Industry New York JohnWiley amp Sons

Arrow Kenneth (1974) The Limits of Organization New York NortonArthur W B (1988a) lsquoSelf-reinforcing mechanisms in economicsrsquo in P W

Anderson K J Arrow and D Pines (eds) The Economy as an Evolving ComplexSystem Redwood City Calif Addison-Wesley

mdashmdash(1988b) lsquoCompeting technologies an overviewrsquo in G Dosi C Freeman RNelson G Silverberg and L Soete (eds) Technical Change and Economic TheoryLondon Pinter Publishers pp 590ndash607

Audretsch David B (1995) Innovation and Industry Evolution Cambridge MassMIT Press

Audretsch David B and Feldman Maryann P (1995) lsquoInnovative clusters andthe industry life cyclersquo Discussion Paper Wissenschaftszentrum BerlinMarket Processes and Corporate Development Research Unit FS-IV-95-7

Baum Joel A C and Oliver Christine (1992) lsquoInstitutional embeddedness andthe dynamics of organizational populationsrsquo American Sociological Review 57540ndash59

Baum Joel A C and Singh Jitendra V (1994a) lsquoOrganizationndashenvironment coevo-lutionrsquo in Joel A C Baum and Jitendra V Singh (eds) The EvolutionaryDynamics of Organizations New York Oxford University Press pp 379ndash402

mdashmdash(eds) (1994b) The Evolutionary Dynamics of Organizations New York OxfordUniversity Press

Berger Suzanne and Dore Ronald (eds) (1996) National Diversity and GlobalCapitalism Ithaca NY Cornell University Press

Boyer Robert (1997) lsquoThe variety and unequal performance of really existingmarkets farewell to Doctor Panglossrsquo in J Rogers Hollingsworth and RobertBoyer (eds) Contemporary Capitalism the Embeddedness of InstitutionsCambridge and New York Cambridge University Press pp 55ndash93

Burns Tom R and Dietz Thomas (2001 forthcoming) lsquoSocial rule system theorysocial action institutional arrangements and evolutionary processesrsquo in JRogers Hollingsworth and Karl Muumlller (eds) Advances in SocioeconomicsLanham Md Rowman amp Littleeld

Burns Tom R and Flam Helena (1987) The Shaping of Social Organization BeverlyHills Calif Sage

Calvert Randall L (1995) lsquoThe rational choice theory of social institutionsrsquo inJeffrey S Banks and Eric A Hanuskek (eds) Modern Political Economy NewYork Cambridge University Press pp 216ndash66

Camic Charles (1986) lsquoThe matter of habitrsquo American Journal of Sociology 911039ndash87

Campbell John L (1997) lsquoMechanisms of evolutionary change in economicsgovernance interaction interpretation and bricolagersquo in Lars Magnusson

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

636

Table 4

continue

d

USA

Germany

Norms rules values which provide incentives for

Norms rules values which provide incentives for

short-term horizons

long-term horizons

Social system of

Prod

uction

Bus

iness system

Training for labo

r an

dHigh deg

ree of voluntary decision mak

ing by

Collective an

d m

ore co

mpulso

ry d

ecisions

abo

ut

man

agem

ent

individuals as to ho

w m

uch

tim

e an

d ene

rgy

how m

uch ed

ucation

is ne

eded

he

avy em

pha

sis

to inv

est in educ

ation heavy

emph

asis on

on techn

ical training (eg en

gine

ering)

busine

ss sch

ools training

in marke

ting

and

sales accou

nting

Fina

ncial marke

tsEqu

ity an

d ven

ture cap

ital m

arke

ts highly

Equ

ity an

d ven

ture cap

ital m

arke

ts w

eakly

dev

elop

ed na

ncing from

cen

tral gov

ernm

ent

deve

lope

d he

avy em

pha

sis on

corporate loan

sto priva

te rm

s for dev

elop

men

t of new

tech

nologies related

to na

tion

al security

Unive

rsity research

High individual auton

omy high poten

tial for

Low

ind

ividua

l au

tono

my until almost 40

yea

rs of

system

individual creativity u

nive

rsity pr

ofesso

rs

age poten

tial for ind

ividual creativity high

ly

high

ly entreprene

urial

cons

traine

d unive

rsity pr

ofessors not so

entrep

rene

urial

Inno

vative

styles

Highly su

ccessful in m

ajor break

through

s in

High pe

rforman

ce in proc

ess an

d inc

remen

tal

basic an

d app

lied scien

ce go

od in dev

elop

ing

inno

vation

s in existing produ

cts not so su

ccessful

totally

new

lines of pr

oducts an

d in m

ajor break

through

s in basic scien

ce or in

tech

nologies high

deg

ree of entrepr

eneu

rship

deve

loping

totally new

lines of prod

ucts of

amon

g research

ers

tech

nologies low d

egree of entrepr

eneu

rship

amon

g research

ers

Table 4

continue

d

USA

Germany

Norms rules values which provide incentives for

Norms rules values which provide incentives for

short-term horizons

long-term horizons

Prod

ucts in w

hich

Pr

oducts with sh

ort liv

es prod

ucts in w

hich

Older produ

cts with long

history machine

ry pap

er

society excels

tech

nology

is bo

th highly co

mplex

and

pr

inting

textile

s m

otor veh

icles m

achine

too

lsch

ange

s rapidly en

tertainm

ent indus

try

pha

rmaceu

ticals biotechn

olog

y softw

are

aerosp

ace

business systems nancial markets and systems of research) Distinctivewould be the recognition of how these separate studies are linkedtogether how each institutional sector is linked to a societyrsquos normsrules values etc and its conguration of institutional arrangementsSimilarly other institutional analysts might focus on the study of orga-nizations with a major concern for how the institutional environmentof organizations (eg norms rules values congurations of institutionalarrangements institutional sectors) inuenced the structure culture andoutcomes of organizations And nally others might study how all ofthe aforementioned aspects of institutional analysis inuenced a societyrsquosoverall performance Nelson and Sampat (1998) have appropriatelyreminded us that we should engage in institutional analysis in order toaddress specic problems but the view developed in this article is thatwe rst need some mapping of the terrain of institutional analysis inorder for our diverse enterprises to become a meaningful collective enter-prise especially if we are to understand how institutions inuence theinnovative processThe argument developed here is not that all social scientists should engage

in institutional analysis Obviously there are numerous other importantareas of research which lie beyond the eld of institutional analysis Nordoes this article argue that institutional analysis is the most importantsubject for social scientistsSince the structure and culture of our research universities reect an

enormous amount of disciplinary fragmentation what are the prospectsfor promoting the kind of collective and interdisciplinary research agendaproposed above In one respect the prospects are not encouragingDuring the past half century the efforts of American universities to have effective communication across the biological sciences have variedgreatly in their success At the University of California-Berkeley it was not until the 1980s that there was a serious effort to integrate thebiological sciences and at that time the campus abolished seventeendepartments to promote an integrated biological program But at theUniversity of Wisconsin the biological sciences are still very fragmentedinto dozens of different departments in botany plant pathology zoologyand physiology with two departments of biochemistry and two depart-ments of geneticsThere have been other approaches which have worked very well histor-

ically for promoting scientic integration amidst scientic diversity Forexample several summer schools and special institutes were organizedin the United States during the 1940s in an effort to overcome the frag-mentation in the biological sciences These programs made it possiblefor scientists in diverse academic disciplines and from different univer-sities and countries to come together to engage in intense and frequentinteraction Perhaps the most effective program was that held in the

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

634

summers at Cold Spring Harbor New York where a research agendafor the study of molecular biology was developed (Fischer and Lipson1988 Stent and Watson 1966) Scientists went there summer aftersummer and when they returned to their universities many began therestructuring of the biological sciences in their universitiesIn our own day we need summer institutes and other similar programs

so that scholars from diverse backgrounds can come together anddevelop a collective agenda for doing institutional analysis and forstudying the impact of institutions on the performance of societies ndashinnovativeness being one such subject If we can develop such instituteswe will increase the prospects for transforming our universities Thereis reason to be optimistic that once an eclectic group of scholars workwith a common set of concepts we can have genuine theoretical advanceif researchers can bring their diverse expertise to bear on a common setof problems There is considerable evidence that institutional analysisoffers high potential for advancing our knowledge of innovations andtechnological change (Hollingsworth and Hollingsworth 2000) If we aregoing to make substantial theoretical advance it is not likely to occurfrom the vantage point of a multiplicity of disciplines Rather it is likelyto occur as we develop a new intellectual framework and our own acad-emic journals This is a tall agenda which requires us to reassess theway that knowledge is presently produced in our societies

NOTE

This article is part of a much larger agenda involving the study of institutionsorganizations and innovations in which I am involved My debts are many andvaried The rst version was written while I was in residence in the ResearchUnit for Institutional Change and European Integration of the Austrian Academyof Sciences in Vienna in the summer of 1998 I am especially grateful to ProfessorsEgon Matzner and Sonja Puntscher-Riekmann for providing the stimulating envi-ronment in which to think through the ideas developed here Egon Matzner hasengaged in very stimulating discussions with me about the importance of insti-tutions in the general eld of socio-economics The nal version of the articlewas completed during my residence as a fellow at the Neurosciences Institutein La Jolla California I also wish to thank my colleagues in the InnovationTheme Project at the Netherlands Institute for Advanced Study (NIAS) during1998ndash9 and Edgar Grande and Karl Muumlller for stimulating conversations aboutthe interaction of institutions and organizations in facilitating particular typesof innovation I am especially grateful to Jerald Hage for helping me to developa number of issues in the article and for his useful comments on the rst draftFrans van Waarden early on helped me to work through the idea of multiplelevels of institutional analysis Others who have helped to develop the ideas inthe article have been my colleagues in the research project lsquoComparative socialsystems of productionrsquo especially Tom Burns Christel Lane Yoshitaka OkadaWolfgang Streeck and Richard Whitley My colleagues Robert Boyer SteveCasper and David Soskice have taught me much about how incentives providedby the institutional environments of rms inuence individual decision making

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

635

But my most important debt is to Ellen Jane Hollingsworth who has repeatedlyquestioned every single assumption of this article Without her rich and contin-uous tutoring this article would be even less developed than it is at presentFinally without the insightful assistance of David Gear in many phases of thisarticle it would not have been completed

REFERENCES

Aoki Masahiko (1990) lsquoTowards an economic model of the Japanese rmrsquo Journalof Economic Literature March 1ndash27

Arora A Landau Ralph and Rosenberg Nathan (eds) (1998) Chemicals and Long-Term Economic Growth Insights from the Chemical Industry New York JohnWiley amp Sons

Arrow Kenneth (1974) The Limits of Organization New York NortonArthur W B (1988a) lsquoSelf-reinforcing mechanisms in economicsrsquo in P W

Anderson K J Arrow and D Pines (eds) The Economy as an Evolving ComplexSystem Redwood City Calif Addison-Wesley

mdashmdash(1988b) lsquoCompeting technologies an overviewrsquo in G Dosi C Freeman RNelson G Silverberg and L Soete (eds) Technical Change and Economic TheoryLondon Pinter Publishers pp 590ndash607

Audretsch David B (1995) Innovation and Industry Evolution Cambridge MassMIT Press

Audretsch David B and Feldman Maryann P (1995) lsquoInnovative clusters andthe industry life cyclersquo Discussion Paper Wissenschaftszentrum BerlinMarket Processes and Corporate Development Research Unit FS-IV-95-7

Baum Joel A C and Oliver Christine (1992) lsquoInstitutional embeddedness andthe dynamics of organizational populationsrsquo American Sociological Review 57540ndash59

Baum Joel A C and Singh Jitendra V (1994a) lsquoOrganizationndashenvironment coevo-lutionrsquo in Joel A C Baum and Jitendra V Singh (eds) The EvolutionaryDynamics of Organizations New York Oxford University Press pp 379ndash402

mdashmdash(eds) (1994b) The Evolutionary Dynamics of Organizations New York OxfordUniversity Press

Berger Suzanne and Dore Ronald (eds) (1996) National Diversity and GlobalCapitalism Ithaca NY Cornell University Press

Boyer Robert (1997) lsquoThe variety and unequal performance of really existingmarkets farewell to Doctor Panglossrsquo in J Rogers Hollingsworth and RobertBoyer (eds) Contemporary Capitalism the Embeddedness of InstitutionsCambridge and New York Cambridge University Press pp 55ndash93

Burns Tom R and Dietz Thomas (2001 forthcoming) lsquoSocial rule system theorysocial action institutional arrangements and evolutionary processesrsquo in JRogers Hollingsworth and Karl Muumlller (eds) Advances in SocioeconomicsLanham Md Rowman amp Littleeld

Burns Tom R and Flam Helena (1987) The Shaping of Social Organization BeverlyHills Calif Sage

Calvert Randall L (1995) lsquoThe rational choice theory of social institutionsrsquo inJeffrey S Banks and Eric A Hanuskek (eds) Modern Political Economy NewYork Cambridge University Press pp 216ndash66

Camic Charles (1986) lsquoThe matter of habitrsquo American Journal of Sociology 911039ndash87

Campbell John L (1997) lsquoMechanisms of evolutionary change in economicsgovernance interaction interpretation and bricolagersquo in Lars Magnusson

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

636

Table 4

continue

d

USA

Germany

Norms rules values which provide incentives for

Norms rules values which provide incentives for

short-term horizons

long-term horizons

Prod

ucts in w

hich

Pr

oducts with sh

ort liv

es prod

ucts in w

hich

Older produ

cts with long

history machine

ry pap

er

society excels

tech

nology

is bo

th highly co

mplex

and

pr

inting

textile

s m

otor veh

icles m

achine

too

lsch

ange

s rapidly en

tertainm

ent indus

try

pha

rmaceu

ticals biotechn

olog

y softw

are

aerosp

ace

business systems nancial markets and systems of research) Distinctivewould be the recognition of how these separate studies are linkedtogether how each institutional sector is linked to a societyrsquos normsrules values etc and its conguration of institutional arrangementsSimilarly other institutional analysts might focus on the study of orga-nizations with a major concern for how the institutional environmentof organizations (eg norms rules values congurations of institutionalarrangements institutional sectors) inuenced the structure culture andoutcomes of organizations And nally others might study how all ofthe aforementioned aspects of institutional analysis inuenced a societyrsquosoverall performance Nelson and Sampat (1998) have appropriatelyreminded us that we should engage in institutional analysis in order toaddress specic problems but the view developed in this article is thatwe rst need some mapping of the terrain of institutional analysis inorder for our diverse enterprises to become a meaningful collective enter-prise especially if we are to understand how institutions inuence theinnovative processThe argument developed here is not that all social scientists should engage

in institutional analysis Obviously there are numerous other importantareas of research which lie beyond the eld of institutional analysis Nordoes this article argue that institutional analysis is the most importantsubject for social scientistsSince the structure and culture of our research universities reect an

enormous amount of disciplinary fragmentation what are the prospectsfor promoting the kind of collective and interdisciplinary research agendaproposed above In one respect the prospects are not encouragingDuring the past half century the efforts of American universities to have effective communication across the biological sciences have variedgreatly in their success At the University of California-Berkeley it was not until the 1980s that there was a serious effort to integrate thebiological sciences and at that time the campus abolished seventeendepartments to promote an integrated biological program But at theUniversity of Wisconsin the biological sciences are still very fragmentedinto dozens of different departments in botany plant pathology zoologyand physiology with two departments of biochemistry and two depart-ments of geneticsThere have been other approaches which have worked very well histor-

ically for promoting scientic integration amidst scientic diversity Forexample several summer schools and special institutes were organizedin the United States during the 1940s in an effort to overcome the frag-mentation in the biological sciences These programs made it possiblefor scientists in diverse academic disciplines and from different univer-sities and countries to come together to engage in intense and frequentinteraction Perhaps the most effective program was that held in the

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

634

summers at Cold Spring Harbor New York where a research agendafor the study of molecular biology was developed (Fischer and Lipson1988 Stent and Watson 1966) Scientists went there summer aftersummer and when they returned to their universities many began therestructuring of the biological sciences in their universitiesIn our own day we need summer institutes and other similar programs

so that scholars from diverse backgrounds can come together anddevelop a collective agenda for doing institutional analysis and forstudying the impact of institutions on the performance of societies ndashinnovativeness being one such subject If we can develop such instituteswe will increase the prospects for transforming our universities Thereis reason to be optimistic that once an eclectic group of scholars workwith a common set of concepts we can have genuine theoretical advanceif researchers can bring their diverse expertise to bear on a common setof problems There is considerable evidence that institutional analysisoffers high potential for advancing our knowledge of innovations andtechnological change (Hollingsworth and Hollingsworth 2000) If we aregoing to make substantial theoretical advance it is not likely to occurfrom the vantage point of a multiplicity of disciplines Rather it is likelyto occur as we develop a new intellectual framework and our own acad-emic journals This is a tall agenda which requires us to reassess theway that knowledge is presently produced in our societies

NOTE

This article is part of a much larger agenda involving the study of institutionsorganizations and innovations in which I am involved My debts are many andvaried The rst version was written while I was in residence in the ResearchUnit for Institutional Change and European Integration of the Austrian Academyof Sciences in Vienna in the summer of 1998 I am especially grateful to ProfessorsEgon Matzner and Sonja Puntscher-Riekmann for providing the stimulating envi-ronment in which to think through the ideas developed here Egon Matzner hasengaged in very stimulating discussions with me about the importance of insti-tutions in the general eld of socio-economics The nal version of the articlewas completed during my residence as a fellow at the Neurosciences Institutein La Jolla California I also wish to thank my colleagues in the InnovationTheme Project at the Netherlands Institute for Advanced Study (NIAS) during1998ndash9 and Edgar Grande and Karl Muumlller for stimulating conversations aboutthe interaction of institutions and organizations in facilitating particular typesof innovation I am especially grateful to Jerald Hage for helping me to developa number of issues in the article and for his useful comments on the rst draftFrans van Waarden early on helped me to work through the idea of multiplelevels of institutional analysis Others who have helped to develop the ideas inthe article have been my colleagues in the research project lsquoComparative socialsystems of productionrsquo especially Tom Burns Christel Lane Yoshitaka OkadaWolfgang Streeck and Richard Whitley My colleagues Robert Boyer SteveCasper and David Soskice have taught me much about how incentives providedby the institutional environments of rms inuence individual decision making

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

635

But my most important debt is to Ellen Jane Hollingsworth who has repeatedlyquestioned every single assumption of this article Without her rich and contin-uous tutoring this article would be even less developed than it is at presentFinally without the insightful assistance of David Gear in many phases of thisarticle it would not have been completed

REFERENCES

Aoki Masahiko (1990) lsquoTowards an economic model of the Japanese rmrsquo Journalof Economic Literature March 1ndash27

Arora A Landau Ralph and Rosenberg Nathan (eds) (1998) Chemicals and Long-Term Economic Growth Insights from the Chemical Industry New York JohnWiley amp Sons

Arrow Kenneth (1974) The Limits of Organization New York NortonArthur W B (1988a) lsquoSelf-reinforcing mechanisms in economicsrsquo in P W

Anderson K J Arrow and D Pines (eds) The Economy as an Evolving ComplexSystem Redwood City Calif Addison-Wesley

mdashmdash(1988b) lsquoCompeting technologies an overviewrsquo in G Dosi C Freeman RNelson G Silverberg and L Soete (eds) Technical Change and Economic TheoryLondon Pinter Publishers pp 590ndash607

Audretsch David B (1995) Innovation and Industry Evolution Cambridge MassMIT Press

Audretsch David B and Feldman Maryann P (1995) lsquoInnovative clusters andthe industry life cyclersquo Discussion Paper Wissenschaftszentrum BerlinMarket Processes and Corporate Development Research Unit FS-IV-95-7

Baum Joel A C and Oliver Christine (1992) lsquoInstitutional embeddedness andthe dynamics of organizational populationsrsquo American Sociological Review 57540ndash59

Baum Joel A C and Singh Jitendra V (1994a) lsquoOrganizationndashenvironment coevo-lutionrsquo in Joel A C Baum and Jitendra V Singh (eds) The EvolutionaryDynamics of Organizations New York Oxford University Press pp 379ndash402

mdashmdash(eds) (1994b) The Evolutionary Dynamics of Organizations New York OxfordUniversity Press

Berger Suzanne and Dore Ronald (eds) (1996) National Diversity and GlobalCapitalism Ithaca NY Cornell University Press

Boyer Robert (1997) lsquoThe variety and unequal performance of really existingmarkets farewell to Doctor Panglossrsquo in J Rogers Hollingsworth and RobertBoyer (eds) Contemporary Capitalism the Embeddedness of InstitutionsCambridge and New York Cambridge University Press pp 55ndash93

Burns Tom R and Dietz Thomas (2001 forthcoming) lsquoSocial rule system theorysocial action institutional arrangements and evolutionary processesrsquo in JRogers Hollingsworth and Karl Muumlller (eds) Advances in SocioeconomicsLanham Md Rowman amp Littleeld

Burns Tom R and Flam Helena (1987) The Shaping of Social Organization BeverlyHills Calif Sage

Calvert Randall L (1995) lsquoThe rational choice theory of social institutionsrsquo inJeffrey S Banks and Eric A Hanuskek (eds) Modern Political Economy NewYork Cambridge University Press pp 216ndash66

Camic Charles (1986) lsquoThe matter of habitrsquo American Journal of Sociology 911039ndash87

Campbell John L (1997) lsquoMechanisms of evolutionary change in economicsgovernance interaction interpretation and bricolagersquo in Lars Magnusson

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

636

business systems nancial markets and systems of research) Distinctivewould be the recognition of how these separate studies are linkedtogether how each institutional sector is linked to a societyrsquos normsrules values etc and its conguration of institutional arrangementsSimilarly other institutional analysts might focus on the study of orga-nizations with a major concern for how the institutional environmentof organizations (eg norms rules values congurations of institutionalarrangements institutional sectors) inuenced the structure culture andoutcomes of organizations And nally others might study how all ofthe aforementioned aspects of institutional analysis inuenced a societyrsquosoverall performance Nelson and Sampat (1998) have appropriatelyreminded us that we should engage in institutional analysis in order toaddress specic problems but the view developed in this article is thatwe rst need some mapping of the terrain of institutional analysis inorder for our diverse enterprises to become a meaningful collective enter-prise especially if we are to understand how institutions inuence theinnovative processThe argument developed here is not that all social scientists should engage

in institutional analysis Obviously there are numerous other importantareas of research which lie beyond the eld of institutional analysis Nordoes this article argue that institutional analysis is the most importantsubject for social scientistsSince the structure and culture of our research universities reect an

enormous amount of disciplinary fragmentation what are the prospectsfor promoting the kind of collective and interdisciplinary research agendaproposed above In one respect the prospects are not encouragingDuring the past half century the efforts of American universities to have effective communication across the biological sciences have variedgreatly in their success At the University of California-Berkeley it was not until the 1980s that there was a serious effort to integrate thebiological sciences and at that time the campus abolished seventeendepartments to promote an integrated biological program But at theUniversity of Wisconsin the biological sciences are still very fragmentedinto dozens of different departments in botany plant pathology zoologyand physiology with two departments of biochemistry and two depart-ments of geneticsThere have been other approaches which have worked very well histor-

ically for promoting scientic integration amidst scientic diversity Forexample several summer schools and special institutes were organizedin the United States during the 1940s in an effort to overcome the frag-mentation in the biological sciences These programs made it possiblefor scientists in diverse academic disciplines and from different univer-sities and countries to come together to engage in intense and frequentinteraction Perhaps the most effective program was that held in the

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

634

summers at Cold Spring Harbor New York where a research agendafor the study of molecular biology was developed (Fischer and Lipson1988 Stent and Watson 1966) Scientists went there summer aftersummer and when they returned to their universities many began therestructuring of the biological sciences in their universitiesIn our own day we need summer institutes and other similar programs

so that scholars from diverse backgrounds can come together anddevelop a collective agenda for doing institutional analysis and forstudying the impact of institutions on the performance of societies ndashinnovativeness being one such subject If we can develop such instituteswe will increase the prospects for transforming our universities Thereis reason to be optimistic that once an eclectic group of scholars workwith a common set of concepts we can have genuine theoretical advanceif researchers can bring their diverse expertise to bear on a common setof problems There is considerable evidence that institutional analysisoffers high potential for advancing our knowledge of innovations andtechnological change (Hollingsworth and Hollingsworth 2000) If we aregoing to make substantial theoretical advance it is not likely to occurfrom the vantage point of a multiplicity of disciplines Rather it is likelyto occur as we develop a new intellectual framework and our own acad-emic journals This is a tall agenda which requires us to reassess theway that knowledge is presently produced in our societies

NOTE

This article is part of a much larger agenda involving the study of institutionsorganizations and innovations in which I am involved My debts are many andvaried The rst version was written while I was in residence in the ResearchUnit for Institutional Change and European Integration of the Austrian Academyof Sciences in Vienna in the summer of 1998 I am especially grateful to ProfessorsEgon Matzner and Sonja Puntscher-Riekmann for providing the stimulating envi-ronment in which to think through the ideas developed here Egon Matzner hasengaged in very stimulating discussions with me about the importance of insti-tutions in the general eld of socio-economics The nal version of the articlewas completed during my residence as a fellow at the Neurosciences Institutein La Jolla California I also wish to thank my colleagues in the InnovationTheme Project at the Netherlands Institute for Advanced Study (NIAS) during1998ndash9 and Edgar Grande and Karl Muumlller for stimulating conversations aboutthe interaction of institutions and organizations in facilitating particular typesof innovation I am especially grateful to Jerald Hage for helping me to developa number of issues in the article and for his useful comments on the rst draftFrans van Waarden early on helped me to work through the idea of multiplelevels of institutional analysis Others who have helped to develop the ideas inthe article have been my colleagues in the research project lsquoComparative socialsystems of productionrsquo especially Tom Burns Christel Lane Yoshitaka OkadaWolfgang Streeck and Richard Whitley My colleagues Robert Boyer SteveCasper and David Soskice have taught me much about how incentives providedby the institutional environments of rms inuence individual decision making

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

635

But my most important debt is to Ellen Jane Hollingsworth who has repeatedlyquestioned every single assumption of this article Without her rich and contin-uous tutoring this article would be even less developed than it is at presentFinally without the insightful assistance of David Gear in many phases of thisarticle it would not have been completed

REFERENCES

Aoki Masahiko (1990) lsquoTowards an economic model of the Japanese rmrsquo Journalof Economic Literature March 1ndash27

Arora A Landau Ralph and Rosenberg Nathan (eds) (1998) Chemicals and Long-Term Economic Growth Insights from the Chemical Industry New York JohnWiley amp Sons

Arrow Kenneth (1974) The Limits of Organization New York NortonArthur W B (1988a) lsquoSelf-reinforcing mechanisms in economicsrsquo in P W

Anderson K J Arrow and D Pines (eds) The Economy as an Evolving ComplexSystem Redwood City Calif Addison-Wesley

mdashmdash(1988b) lsquoCompeting technologies an overviewrsquo in G Dosi C Freeman RNelson G Silverberg and L Soete (eds) Technical Change and Economic TheoryLondon Pinter Publishers pp 590ndash607

Audretsch David B (1995) Innovation and Industry Evolution Cambridge MassMIT Press

Audretsch David B and Feldman Maryann P (1995) lsquoInnovative clusters andthe industry life cyclersquo Discussion Paper Wissenschaftszentrum BerlinMarket Processes and Corporate Development Research Unit FS-IV-95-7

Baum Joel A C and Oliver Christine (1992) lsquoInstitutional embeddedness andthe dynamics of organizational populationsrsquo American Sociological Review 57540ndash59

Baum Joel A C and Singh Jitendra V (1994a) lsquoOrganizationndashenvironment coevo-lutionrsquo in Joel A C Baum and Jitendra V Singh (eds) The EvolutionaryDynamics of Organizations New York Oxford University Press pp 379ndash402

mdashmdash(eds) (1994b) The Evolutionary Dynamics of Organizations New York OxfordUniversity Press

Berger Suzanne and Dore Ronald (eds) (1996) National Diversity and GlobalCapitalism Ithaca NY Cornell University Press

Boyer Robert (1997) lsquoThe variety and unequal performance of really existingmarkets farewell to Doctor Panglossrsquo in J Rogers Hollingsworth and RobertBoyer (eds) Contemporary Capitalism the Embeddedness of InstitutionsCambridge and New York Cambridge University Press pp 55ndash93

Burns Tom R and Dietz Thomas (2001 forthcoming) lsquoSocial rule system theorysocial action institutional arrangements and evolutionary processesrsquo in JRogers Hollingsworth and Karl Muumlller (eds) Advances in SocioeconomicsLanham Md Rowman amp Littleeld

Burns Tom R and Flam Helena (1987) The Shaping of Social Organization BeverlyHills Calif Sage

Calvert Randall L (1995) lsquoThe rational choice theory of social institutionsrsquo inJeffrey S Banks and Eric A Hanuskek (eds) Modern Political Economy NewYork Cambridge University Press pp 216ndash66

Camic Charles (1986) lsquoThe matter of habitrsquo American Journal of Sociology 911039ndash87

Campbell John L (1997) lsquoMechanisms of evolutionary change in economicsgovernance interaction interpretation and bricolagersquo in Lars Magnusson

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

636

summers at Cold Spring Harbor New York where a research agendafor the study of molecular biology was developed (Fischer and Lipson1988 Stent and Watson 1966) Scientists went there summer aftersummer and when they returned to their universities many began therestructuring of the biological sciences in their universitiesIn our own day we need summer institutes and other similar programs

so that scholars from diverse backgrounds can come together anddevelop a collective agenda for doing institutional analysis and forstudying the impact of institutions on the performance of societies ndashinnovativeness being one such subject If we can develop such instituteswe will increase the prospects for transforming our universities Thereis reason to be optimistic that once an eclectic group of scholars workwith a common set of concepts we can have genuine theoretical advanceif researchers can bring their diverse expertise to bear on a common setof problems There is considerable evidence that institutional analysisoffers high potential for advancing our knowledge of innovations andtechnological change (Hollingsworth and Hollingsworth 2000) If we aregoing to make substantial theoretical advance it is not likely to occurfrom the vantage point of a multiplicity of disciplines Rather it is likelyto occur as we develop a new intellectual framework and our own acad-emic journals This is a tall agenda which requires us to reassess theway that knowledge is presently produced in our societies

NOTE

This article is part of a much larger agenda involving the study of institutionsorganizations and innovations in which I am involved My debts are many andvaried The rst version was written while I was in residence in the ResearchUnit for Institutional Change and European Integration of the Austrian Academyof Sciences in Vienna in the summer of 1998 I am especially grateful to ProfessorsEgon Matzner and Sonja Puntscher-Riekmann for providing the stimulating envi-ronment in which to think through the ideas developed here Egon Matzner hasengaged in very stimulating discussions with me about the importance of insti-tutions in the general eld of socio-economics The nal version of the articlewas completed during my residence as a fellow at the Neurosciences Institutein La Jolla California I also wish to thank my colleagues in the InnovationTheme Project at the Netherlands Institute for Advanced Study (NIAS) during1998ndash9 and Edgar Grande and Karl Muumlller for stimulating conversations aboutthe interaction of institutions and organizations in facilitating particular typesof innovation I am especially grateful to Jerald Hage for helping me to developa number of issues in the article and for his useful comments on the rst draftFrans van Waarden early on helped me to work through the idea of multiplelevels of institutional analysis Others who have helped to develop the ideas inthe article have been my colleagues in the research project lsquoComparative socialsystems of productionrsquo especially Tom Burns Christel Lane Yoshitaka OkadaWolfgang Streeck and Richard Whitley My colleagues Robert Boyer SteveCasper and David Soskice have taught me much about how incentives providedby the institutional environments of rms inuence individual decision making

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

635

But my most important debt is to Ellen Jane Hollingsworth who has repeatedlyquestioned every single assumption of this article Without her rich and contin-uous tutoring this article would be even less developed than it is at presentFinally without the insightful assistance of David Gear in many phases of thisarticle it would not have been completed

REFERENCES

Aoki Masahiko (1990) lsquoTowards an economic model of the Japanese rmrsquo Journalof Economic Literature March 1ndash27

Arora A Landau Ralph and Rosenberg Nathan (eds) (1998) Chemicals and Long-Term Economic Growth Insights from the Chemical Industry New York JohnWiley amp Sons

Arrow Kenneth (1974) The Limits of Organization New York NortonArthur W B (1988a) lsquoSelf-reinforcing mechanisms in economicsrsquo in P W

Anderson K J Arrow and D Pines (eds) The Economy as an Evolving ComplexSystem Redwood City Calif Addison-Wesley

mdashmdash(1988b) lsquoCompeting technologies an overviewrsquo in G Dosi C Freeman RNelson G Silverberg and L Soete (eds) Technical Change and Economic TheoryLondon Pinter Publishers pp 590ndash607

Audretsch David B (1995) Innovation and Industry Evolution Cambridge MassMIT Press

Audretsch David B and Feldman Maryann P (1995) lsquoInnovative clusters andthe industry life cyclersquo Discussion Paper Wissenschaftszentrum BerlinMarket Processes and Corporate Development Research Unit FS-IV-95-7

Baum Joel A C and Oliver Christine (1992) lsquoInstitutional embeddedness andthe dynamics of organizational populationsrsquo American Sociological Review 57540ndash59

Baum Joel A C and Singh Jitendra V (1994a) lsquoOrganizationndashenvironment coevo-lutionrsquo in Joel A C Baum and Jitendra V Singh (eds) The EvolutionaryDynamics of Organizations New York Oxford University Press pp 379ndash402

mdashmdash(eds) (1994b) The Evolutionary Dynamics of Organizations New York OxfordUniversity Press

Berger Suzanne and Dore Ronald (eds) (1996) National Diversity and GlobalCapitalism Ithaca NY Cornell University Press

Boyer Robert (1997) lsquoThe variety and unequal performance of really existingmarkets farewell to Doctor Panglossrsquo in J Rogers Hollingsworth and RobertBoyer (eds) Contemporary Capitalism the Embeddedness of InstitutionsCambridge and New York Cambridge University Press pp 55ndash93

Burns Tom R and Dietz Thomas (2001 forthcoming) lsquoSocial rule system theorysocial action institutional arrangements and evolutionary processesrsquo in JRogers Hollingsworth and Karl Muumlller (eds) Advances in SocioeconomicsLanham Md Rowman amp Littleeld

Burns Tom R and Flam Helena (1987) The Shaping of Social Organization BeverlyHills Calif Sage

Calvert Randall L (1995) lsquoThe rational choice theory of social institutionsrsquo inJeffrey S Banks and Eric A Hanuskek (eds) Modern Political Economy NewYork Cambridge University Press pp 216ndash66

Camic Charles (1986) lsquoThe matter of habitrsquo American Journal of Sociology 911039ndash87

Campbell John L (1997) lsquoMechanisms of evolutionary change in economicsgovernance interaction interpretation and bricolagersquo in Lars Magnusson

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

636

But my most important debt is to Ellen Jane Hollingsworth who has repeatedlyquestioned every single assumption of this article Without her rich and contin-uous tutoring this article would be even less developed than it is at presentFinally without the insightful assistance of David Gear in many phases of thisarticle it would not have been completed

REFERENCES

Aoki Masahiko (1990) lsquoTowards an economic model of the Japanese rmrsquo Journalof Economic Literature March 1ndash27

Arora A Landau Ralph and Rosenberg Nathan (eds) (1998) Chemicals and Long-Term Economic Growth Insights from the Chemical Industry New York JohnWiley amp Sons

Arrow Kenneth (1974) The Limits of Organization New York NortonArthur W B (1988a) lsquoSelf-reinforcing mechanisms in economicsrsquo in P W

Anderson K J Arrow and D Pines (eds) The Economy as an Evolving ComplexSystem Redwood City Calif Addison-Wesley

mdashmdash(1988b) lsquoCompeting technologies an overviewrsquo in G Dosi C Freeman RNelson G Silverberg and L Soete (eds) Technical Change and Economic TheoryLondon Pinter Publishers pp 590ndash607

Audretsch David B (1995) Innovation and Industry Evolution Cambridge MassMIT Press

Audretsch David B and Feldman Maryann P (1995) lsquoInnovative clusters andthe industry life cyclersquo Discussion Paper Wissenschaftszentrum BerlinMarket Processes and Corporate Development Research Unit FS-IV-95-7

Baum Joel A C and Oliver Christine (1992) lsquoInstitutional embeddedness andthe dynamics of organizational populationsrsquo American Sociological Review 57540ndash59

Baum Joel A C and Singh Jitendra V (1994a) lsquoOrganizationndashenvironment coevo-lutionrsquo in Joel A C Baum and Jitendra V Singh (eds) The EvolutionaryDynamics of Organizations New York Oxford University Press pp 379ndash402

mdashmdash(eds) (1994b) The Evolutionary Dynamics of Organizations New York OxfordUniversity Press

Berger Suzanne and Dore Ronald (eds) (1996) National Diversity and GlobalCapitalism Ithaca NY Cornell University Press

Boyer Robert (1997) lsquoThe variety and unequal performance of really existingmarkets farewell to Doctor Panglossrsquo in J Rogers Hollingsworth and RobertBoyer (eds) Contemporary Capitalism the Embeddedness of InstitutionsCambridge and New York Cambridge University Press pp 55ndash93

Burns Tom R and Dietz Thomas (2001 forthcoming) lsquoSocial rule system theorysocial action institutional arrangements and evolutionary processesrsquo in JRogers Hollingsworth and Karl Muumlller (eds) Advances in SocioeconomicsLanham Md Rowman amp Littleeld

Burns Tom R and Flam Helena (1987) The Shaping of Social Organization BeverlyHills Calif Sage

Calvert Randall L (1995) lsquoThe rational choice theory of social institutionsrsquo inJeffrey S Banks and Eric A Hanuskek (eds) Modern Political Economy NewYork Cambridge University Press pp 216ndash66

Camic Charles (1986) lsquoThe matter of habitrsquo American Journal of Sociology 911039ndash87

Campbell John L (1997) lsquoMechanisms of evolutionary change in economicsgovernance interaction interpretation and bricolagersquo in Lars Magnusson

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

636

and Jan Ottosson (eds) Evolutionary Economics and Path DependenceCheltenham Edward Elgar pp 10ndash32

Campbell John and Lindberg Leon (1990) lsquoProperty rights and the organizationof economic activity by the statersquo American Sociological Review 55

Campbell John L Hollingsworth J Rogers and Lindberg Leon (eds) (1991) TheGovernance of the American Economy Cambridge and New York CambridgeUniversity Press

Casper Steven (2000 forthcoming) lsquoHigh technology governance and institut-ional adaptiveness do technology policies usefully promote commercial innovation within the German biotechnology industryrsquo Organization Studies21

Chandler Alfred (1962) Strategy and Structure Cambridge Mass MIT Pressmdashmdash(1977) The Visible Hand the Managerial Revolution in American Business

Cambridge Mass Harvard University Pressmdashmdash(1990) Scale and Scope the Dynamics of Industrial Capitalism Cambridge Mass

Harvard University PressCoase Ronald H (1960) lsquoThe problem of social costrsquo Journal of Law and Economics

3 1ndash44mdashmdash(1981) lsquoThe Coase theorem and the empty core a commentrsquo Journal of Law

and Economics 24 183ndash7David Paul (1988) lsquoPath-dependence putting the past in the future of economicsrsquo

IMSS Technical Report No 533 Stanford University CalifDiMaggio Paul and Powell Walter W (1983) lsquoThe iron cage revisited institu-

tional isomorphism and collective rationality in organizational eldsrsquoAmerican Sociological Review 48 147ndash60

mdashmdash(1991) lsquoIntroductionrsquo in Walter W Powell and Paul J DiMaggio (eds) TheNew Institutionalism in Organizational Analysis Chicago University of ChicagoPress pp 1ndash38

Dosi Giovanni (1988) lsquoSources procedures and microeconomic effects of inno-vationrsquo Journal of Economic Literature 36 1126ndash71

Douglas Mary (1987) How Institutions Think Syracuse NY Syracuse UniversityPress

Durlauf Steven N (1991) lsquoPath dependence in economics the invisible hand inthe grip of the pastrsquo American Economics Association Papers and Proceedings81 70ndash4

Edquist Charles (ed) (1997) Systems of Innovation Technologies Institutions andOrganizations London Pinter

Eggertsson Thrainn (1990) Economic Behavior and Institutions CambridgeCambridge University Press

Elster Jon (1989) The Cement of Society a Study of Social Order CambridgeCambridge University Press

Etzioni Amitai (1988) The Moral Dimension Toward a New Economics New YorkThe Free Press

Field Alexander (1984) lsquoMicroeconomics norms and rationalityrsquo EconomicDevelopment and Cultural Change 32 683ndash711

Finnemore Martha (1996) lsquoNorms culture and world politics insights from soci-ologyrsquos institutionalismrsquo International Organization 50 325ndash47

Fischer Ernst Peter and Lipson Carol (1988) Thinking About Science Max Delbruckand the Origins of Molecular Biology New York W W Norton

Fransman Martin (1994) lsquoInformation knowledge vision and theories of thermrsquo Industrial and Corporate Change 3 713ndash57

Freeman Christopher (1991) lsquoThe nature of innovation and the evolution of theproductive systemrsquo Technology and Productivity Paris OECD

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

637

Friedland Roger and Alford Robert A (1991) lsquoBringing society back in symbolspractices and institutional contradictionsrsquo in Walter W Powell and Paul JDiMaggio (eds) The New Institutionalism in Organizational Analysis ChicagoUniversity of Chicago Press pp 232ndash63

Fukuyama Francis (1995) Trust Social Virtues and the Creation of Prosperity NewYork The Free Press

Gambetta Diego (ed) (1988) Trust Making and Breaking Cooperative RelationsNew York Basil Blackwell

Geertz Clifford (1995) After the Fact Cambridge Mass Harvard University PressGrafstein Robert (1992) Institutional Realism Social and Political Constraints on

Rational Actors New Haven Conn Yale University PressGramsci Antonio (1971) lsquoAmericanism and Fordismrsquo in Quintin Hoare and

Geoffrey Nowell Smith (eds and trans) Selections from the Prison NotebooksNew York International Publishers pp 277ndash320

Granovetter Mark (1985) lsquoEconomic action and social structure the problem ofembeddednessrsquo American Journal of Sociology 91 481ndash510

Habermas J (1975) Legitimation Crisis Boston Mass Beacon PressHage Jerald and Alter Catherine (1997) lsquoA typology of interorganizational rela-

tionships and networksrsquo in J Rogers Hollingsworth and Robert Boyer (eds)Contemporary Capitalism the Embeddedness of Institutions Cambridge and NewYork Cambridge University Press pp 94ndash126

Hage Jerald and Hollingsworth J Rogers (2000) lsquoA strategy for analysis of ideainnovation networks and institutionsrsquo Organization Studies 21

Haringkansson H and Lundgren A (1997) lsquoPaths in time and spacendashpath dependencein industrial networksrsquo in L Magnusson and J Ottosson (eds) EvolutionaryEconomics and Path Dependence Cheltenham Edward Elgar pp 119ndash37

Hall Peter A and Taylor Rosemary C R (1996) lsquoPolitical science and the threenew institutionalismsrsquo Cologne MPIG Max-Planck Institut fuumlrGesellschaftsforchung Discussion Paper 966

Hamilton Gary G and Biggart Nichole Woolsey (1988) lsquoMarket culture andauthority a comparative analysis of management and organization in theFar Eastrsquo American Journal of Sociology 94 S52ndashS94

Hannan Michael T and Freeman John H (1977) lsquoThe population ecology oforganizationsrsquo American Journal of Sociology 82 929ndash64

mdashmdash(1984) lsquoStructural inertia and organizational changersquo American SociologicalReview 49 149ndash64

Hechter Michael and Kanazawa Satoshi (1997) lsquoSociological and rational choicetheoryrsquo Annual Review of Sociology 23 191ndash214

Hodgson Geoffrey M (1988) Economics and Institutions a Manifesto for a ModernInstitutional Economics Philadelphia University of Pennsylvania Press

mdashmdash(1989) lsquoInstitutional rigidities and economic growthrsquo Cambridge Journal ofEconomics 13 79ndash101

mdashmdash(1997) lsquoThe ubiquity of habits and rulesrsquo Cambridge Journal of Economics 21663ndash84

mdashmdash(1998) lsquoThe approach of institutional economicsrsquo Journal of Economic Literature36 166ndash92

mdashmdash (1999) Economics and Utopia London RoutledgeHollingsworth J Rogers (1984) lsquoThe snare of specializationrsquo Bulletin of the Atomic

Scientists 40 34ndash7mdashmdash(1991) lsquoThe logic of coordinating American manufacturing sectorsrsquo in John

L Campbell J Rogers Hollingsworth and Leon Lindberg (eds) The Governanceof the American Economy Cambridge and New York Cambridge UniversityPress pp 35ndash73

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

638

mdashmdash (1997) lsquoContinuities and changes in social systems of production the casesof Japan Germany and the United Statesrsquo in J Rogers Hollingsworth andRobert Boyer (eds) Contemporary Capitalism the Embeddedness of InstitutionsCambridge and New York Cambridge University Press pp 265ndash310

mdashmdash(1998) lsquoNew perspectives on the spatial dimensions of economic coordina-tion tensions between globalization and social systems of productionrsquo Reviewof International Political Economy 5 482ndash507

Hollingsworth J Rogers and Boyer Robert (eds) (1997) Contemporary Capitalismthe Embeddedness of Institutions Cambridge and New York CambridgeUniversity Press

Hollingsworth J Rogers and Hanneman Robert (1982) lsquoWorking-class powerand the political economy of Western capitalist societiesrsquo Comparative SocialResearch 5 61ndash80

Hollingsworth J Rogers and Hollingsworth Ellen Jane (2000) lsquoMajor discoveriesand biomedical research organizations perspectives on interdisciplinaritynurturing leadership and integrated structure and culturesrsquo in PeterWeingart and Nico Stehr (eds) Practising Interdisciplinarity TorontoUniversity of Toronto Press pp 215ndash44

Hollingsworth J Rogers and Lindberg Leon (1985) lsquoThe governance of theAmerican economy the role of markets clans hierarchies and associativebehaviorrsquo in Wolfgang Streeck and Philippe C Schmitter (eds) Private InterestGovernment Beyond Market and State London and Beverly Hills Calif SagePublications pp 221ndash54

Hollingsworth J Rogers and Streeck Wolfgang (1994) lsquoCountries and sectorsperformance convergence and competitivenessrsquo in J Rogers HollingsworthPhilippe Schmitter and Wolfgang Streeck (eds) Governing Capitalist EconomiesPerformance and Control of Economic Sectors New York Oxford UniversityPress pp 270ndash300

Hollingsworth J Rogers Hage Jerald and Hanneman Robert A (1990) StateIntervention in Medical Care Consequences for Britain France Sweden and theUnited States 1890ndash1970 Ithaca NY Cornell University Press

Hollingsworth J Rogers Hollingsworth Ellen Jane and Hage Jerald (2001 forth-coming) Organizations and Performance in Bio-Medical Science

Hollingsworth J Rogers Schmitter Philippe and Streeck Wolfgang (eds) (1994)Governing Capitalist Economies Performance and Control of Economic SectorsNew York Oxford University Press

Immergut Ellen M (1998) lsquoThe theoretical core of the new institutionalismrsquoPolitics and Society 26 5ndash34

Jepperson Ronald L (1991) lsquoInstitutions institutional effects and institutional-ismrsquo in Walter W Powell and Paul J DiMaggio (eds) The New Institutionalismin Organizational Analysis Chicago University of Chicago Press pp 143ndash63

Johnson Bjoumlrn (1992) lsquoInstitutional learningrsquo in Bengt-Ake Lundvall (ed) NationalSystems of Innovation Towards a Theory of Innovation and Interactive LearningLondon Pinter pp 23ndash44

Judson Horace F (1979) The Eighth Day of Creation New York Simon amp SchusterKatznelson Ira (1998) lsquoThe doleful dance of politics and policy can historical

institutionalism make a differencersquo American Political Science Review 92191ndash7

Kenney Martin and Florida Richard (1993) Beyond Mass Production the JapaneseSystem and Its Transfer to the US New York Oxford University Press

Kim Eun Mee (1997) Big Business Strong State Collusion and Conict in SouthKorean Development 1960ndash1990 Albany NY State University of New YorkPress

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

639

Knight David (1992) Institutions and Social Conict Cambridge CambridgeUniversity Press

Kohler Robert E (1982) From Medical Chemistry to Biochemistry CambridgeCambridge University Press

mdashmdash(1994) The Lords of the Fly Chicago University of Chicago PressKondra Alex Z and Hinings C R (1998) lsquoOrganizational diversity and change

in institutional theoryrsquo Organization Studies 19 743ndash67Krugman Paul (1991) lsquoHistory and industry location the case of the manufac-

turing beltrsquo American Economic Association Papers and Proceedings 81 80ndash3Landes David S (1969) The Unbound Prometheus New York Cambridge

University Pressmdashmdash(1998) The Wealth and Poverty of Nations Why Some Are So Rich and Some Are

So Poor New York W W NortonLanglois Richard (ed) (1986) Economics as a Process Essays in the New Institutional

Economics Cambridge Cambridge University Pressmdashmdash(1989) lsquoWhat is wrong with the old institutional economicsrsquo Review of

Political Economy 1 270ndash98Langlois Richard and Robertson Paul L (eds) (1995) Firms Markets and Economic

Change London RoutledgeLanzara Giovan Francesco (1998) lsquoSelf-destructive processes in institution

building and some modest countervailing mechanismsrsquo European Journal ofPolitical Research 33 1ndash39

Lazonick William (1990) Competitive Advantage on the Shop Floor CambridgeMass Harvard University Press

Legro Jeffrey (1997) lsquoWhich norms matter Revisiting the ldquofailurerdquo of interna-tionalismrsquo International Organization 51 31ndash63

Leibenstein Harvey (1966) lsquoAllocative efciency versus X-efciencyrsquo AmericanEconomic Review 66 392ndash415

mdashmdash(1976) Beyond Economic Man a New Foundation in Microeconomics CambridgeMass Harvard University Press

Lindberg Leon and Campbell John L (1991) lsquoThe State and the organization ofeconomic activityrsquo in John L Campbell J Roger Hollingsworth and LeonLindberg (eds) The Governance of the American economy New York CambridgeUniversity Press pp 356ndash95

Lindberg Leon Campbell John L and Hollingsworth J Rogers (1991) lsquoEconomicgovernance and the analysis of structural change in the American economyrsquo inJohn L Campbell J Rogers Hollingsworth and Leon Lindberg (eds) Governanceof the American Economy New York Cambridge University Press pp 3ndash34

Lundvall Bengt-Ake (ed) (1992) National Systems of Innovation Towards a Theoryof Innovation and Interactive Learning London Pinter

Magnusson Lars and Ottosson Jan (1997) Evolutionary Economics and PathDependence Cheltenham Edward Elgar

March James and Olsen Johan P (1989) Rediscovering Institutions New YorkThe Free Press

Maurice M Sorge A and Warner M (1980) lsquoSocietal differences in organizingmanufacturing units a comparison of France West Germany and GreatBritainrsquo Organization Studies I 59ndash86

Maynard-Smith J (1982) Evolution and the Theory of Games Cambridge and NewYork Cambridge University Press

Meyer John and Rowan B (1991) lsquoInstitutionalized organizations formal struc-ture as myth and ceremonyrsquo reprinted in W Powell and Paul DiMaggio(eds) The New Institutionalism in Organizational Analysis Chicago Universityof Chicago Press pp 41ndash62

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

640

Milgrom Paul Qian Yingyi and Roberts John (1991) lsquoComplementaritiesmomentum and the evolution of modern manufacturingrsquo American EconomicsAssociation Papers and Proceedings 81 84

Mowery David and Nelson Richard R (eds) (1999) The Sources of IndustrialLeadership New York Cambridge University Press

Murmann Johann Peter (1998) lsquoKnowledge and competitive advantage in thesynthetic dye industry 1850ndash1914rsquo PhD dissertation Columbia University

Nelson Richard R (ed) (1993) National Innovation Systems a Comparative AnalysisNew York and Oxford Oxford University Press

mdashmdash(1994) lsquoThe co-evolution of technology industrial structure and supportinginstitutionsrsquo Industrial and Corporate Change 3 417ndash19

mdashmdash(1995a) lsquoCo-evolution of industry structure technology and supporting insti-tution and the making of comparative advantagersquo International Journal of theEconomics of Business 2 171ndash84

mdashmdash(1995b) lsquoWhy should managers be thinking about technology policyrsquoStrategic Management Journal 16 581ndash8

mdashmdash(1996) lsquoThe evolution of competitive or comparative advantage a prelimi-nary report on a studyrsquo Industrial and Corporate Change 5 597ndash618

mdashmdash(1999) lsquoThe sources of industrial leadership a perspective on industrialpolicyrsquo in David Mowery and Richard R Nelson (eds) The Sources of IndustrialLeadership New York Cambridge University Press

Nelson Richard R and Sampat Bhaven N (1998) lsquoMaking sense of institutionsas a factor shaping economic performancersquo unpublished paper

Nelson Richard R and Winter S G (1982) An Evolutionary Theory of EconomicChange Cambridge Mass Harvard University Press

North Douglass (1981) Structure and Change in Economic History New YorkNorton

mdashmdash(1991) Institutions Institutional Change and Economic Performance Cambridgeand New York Cambridge University Press

Olby Robert (1979) The Path to the Double Helix Seattle University of WashingtonPress

Oliver Nick and Wilkinson Barry (1988) The Japanization of British IndustryOxford Basil Blackwell

Orren K and Skowronek S (1991) lsquoBeyond the iconography of order notes fora ldquonew institutionalismrdquorsquo in L C Dodd and C Jillson (eds) The Dynamicsof American Politics Boulder Colo Westview Press pp 311ndash30

Orru Marcu Woolsey Biggart Nichole and Hamilton Gary G (1991)lsquoOrganizational isomorphism in East Asiarsquo in Paul J DiMaggio and WalterW Powell (eds) The Institutionalism in Organizational Analysis ChicagoUniversity of Chicago Press pp 361ndash89

Ostrom Elinor (1986) lsquoAn agenda for the study of institutionsrsquo Public Choice 483ndash25

mdashmdash(1990) Governing the Commons the Evolution of Institutions for Collective ActionNew York Cambridge University Press

Pfeffer Jeffrey and Salancik Gerald (1978) The External Control of Organizationsa Resource Dependence Perspective New York Harper amp Row

Polanyi Karl (1944) The Great Transformation the Political and Economic Origins ofOur Time (reprinted 1957) Boston Mass Beacon Press

Popper Karl (1961) The Poverty of Historicism London Routledge amp Kegan PaulPorter Michael (1990) The Competitive Advantage of Nations New York The Free

PressPosner R (1992) Economic Analysis of Law Boston Mass Little Brown

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

641

Powell Walter W (1991) lsquoExpanding the scope of institutional analysisrsquo in WalterW Powell and Paul J DiMaggio (eds) The New Institutionalism inOrganizational Analysis Chicago and London University of Chicago Presspp 183ndash203

Powell Walter W and DiMaggio Paul J (eds) (1991) The New Institutionalismin Organizational Analysis Chicago University of Chicago Press

Putnam Robert D (1993) Making Democracy Work Princeton NJ PrincetonUniversity Press

Roland Gerard (1990) lsquoGorbachev and the common European home the conver-gence debate revisitedrsquo Kyklos 43 385ndash409

Sabel Charles F (1992) lsquoStudied trust building new forms of cooperation in avolatile economyrsquo in Frank Pyke and Werner Sengenberger (eds) IndustrialDistricts and Local Economic Regeneration Geneva International Institute forLabor Studies pp 215ndash50

Schneiberg Marc and Hollingsworth J Rogers (1990) lsquoCan transaction costeconomics explain trade associationsrsquo in Masahiko Aoki Bo Gustafsson andOliver E Williamson (eds) The Firm as a Nexus of Treaties London and BeverlyHills Calif Sage Publications pp 320ndash460

Schotter Andrew (1981) The Economic Theory of Social Institutions Cambridge andNew York Cambridge University Press

Schumpeter J A (1983) The Theory of Economic Development New BrunswickNJ Transaction Books

Scott W Richard (1994) lsquoInstitutions and organizations towards a theoreticalsynthesisrsquo in W Richard Scott and John W Meyer and Associates (eds)Institutional Environment and Organizations Structural Complexity andIndividualism Thousand Oaks Calif Sage pp 55ndash80

Scott W Richard Meyer John W and Associates (eds) (1994) InstitutionalEnvironment and Organizations Structural Complexity and IndividualismThousand Oaks Calif Sage

Shepsle Kenneth A (1986) lsquoInstitutional equilibrium and equilibrium institu-tionsrsquo in Herbert Weisberg (ed) Political Science the Science of Politics NewYork Agathon pp 51ndash81

mdashmdash(1989) lsquoStudying institutions some lessons from the rational choiceapproachrsquo Journal of Theoretical Politics 1 131ndash47

Shimanoff S B (1980) Communication Rules Theory and Research Beverly HillsCalif Sage Publications

Slack Trevor and Hinings Bob (1994) lsquoInstitutional pressures and isomorphicchange an empire testrsquo Organization Studies 15 803ndash27

Somit Albert and Peterson Steven A (eds) (1992) The Dynamics of Evolution thePunctuated Equilibrium Debate in the Natural and Social Sciences Ithaca NYCornell University Press

Sorge A (1989) lsquoAn essay on technical change its dimensions and social andstrategic contextrsquo Organization Studies 10(1) 23ndash44

Sorge A and Streeck W (1988) lsquoIndustrial relations and technical change thecase for an extended perspectiversquo in R Hyman and W Streeck (eds) NewTechnology and Industrial Relations New York and Oxford Basil Blackwellpp 19ndash47

Soskice David (1990) lsquoWage determination the changing role of institutions inadvanced industrialized countriesrsquo Oxford Review of Economic Policy 6 36ndash61

mdashmdash(1997) lsquoGerman technology policy innovation and national institutionalframeworksrsquo Industry and Innovation 1 (June) 75ndash95

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

642

mdashmdash(1999) lsquoDivergent production regimes coordinated and uncoordinatedmarket economies in the 1980s and 1990srsquo in Herbert Kitschelt Peter LangeGary Marks and John D Stephens (eds) Continuity and Change in ContemporaryCapitalism Cambridge Cambridge University Press

Steinmo Sven Thelen Kathleen and Longstreth Frank (eds) (1992) StructuringPolitics Historical Institutionalism in Comparative Analysis New YorkCambridge University Press

Stent Gunther and Watson James D (eds) (1966) Phage and the Origins of MolecularBiology Cold Spring Harbor NY Cold Spring Harbor Laboratory ofQuantitative Biology

Stinchcombe Arthur L (1997) lsquoOn the virtue of the old institutionalismrsquo AnnualReview of Sociology 23 1ndash18

Streeck Wolfgang (1992) Social Institutions and Economic Performance NewburyPark Calif Sage

mdashmdash(1997a) lsquoBenecial constraints on the economic limits of rational volun-tarismrsquo in J Rogers Hollingsworth and Robert Boyer (eds) ContemporaryCapitalism the Embeddedness of Institutions New York and CambridgeCambridge University Press pp 197ndash219

mdashmdash(1997b) lsquoGerman capitalism does it exist can it surviversquo in Colin Crouchand Wolfgang Streeck (eds) Political Economy of Modern Capitalism MappingConvergence and Divergence Thousand Oaks Calif Sage pp 33ndash54

Streeck Wolfgang and Schmitter Philippe C (1985a) lsquoCommunity market state ndash and associations The prospective contribution of interest governanceto social orderrsquo in Wolfgang Streeck and Philippe C Schmitter (eds) PrivateInterest Government Beyond Market and State Beverly Hills Calif Sage pp1ndash29

mdashmdash(eds) (1985b) Private Interest Government Beyond Market and State BeverlyHills Calif Sage

Teece D (1993) lsquoPerspectives on Alfred Chandlerrsquos scale and scopersquo Journal ofEconomic Literature March 199ndash225

Townley Barbara (1997) lsquoThe institutional logic of performance appraisalrsquoOrganization Studies 18 261ndash85

Veblen T (1899) The Theory of the Leisure Class an Economic Study of InstitutionsNew York Macmillan

mdashmdash(1915) Imperial Germany and the Industrial Revolution New York MacmillanWhitley Richard (2000) lsquoThe institutional structuring of innovation strategies

business systems rm types and patterns of technical change in differentmarket economiesrsquo Organization Studies 21

Williamson Oliver E (1975) Markets and Hierarchies Analysis and AntitrustImplications New York The Free Press

mdashmdash(1985) The Economic Institutions of Capitalism New York The Free PressWinter Sidney (1982) lsquoSchumpeterian competition in alternative technological

regimesrsquo Journal of Economic Behavior and Organization 5 137ndash58Wright Erik (1998) lsquoA brief comment on Wolfgang Streeckrsquos essay ldquoBenecial

constraints on the economic limits of rational voluntarismrdquorsquo unpublishedpaper

Zucker Lynne G (1987) lsquoInstitutional theories of organizationsrsquo Annual Reviewof Sociology 13 443ndash64

mdashmdash(1988) Institutional Patterns and Organizations Culture and EnvironmentCambridge Mass Ballinger

HOLLINGSWORTH DOING INSTITUTIONAL ANALYSIS

643

mdashmdash(1991) lsquoThe role of institutionalism in cultural persistencersquo in Walter WPowell and Paul J DiMaggio (eds) The New Institutionalism in OrganizationalAnalysis Chicago University of Chicago Press pp 83ndash107

Zysman John (1994) lsquoHow institutions create historically rooted trajectories ofgrowthrsquo Industrial and Corporate Change 3 243ndash83

REVIEW OF INTERNATIONAL POLITICAL ECONOMY

644