chapter in smelser-baltes 2001

Lazarus A A 1997 Brief but Comprehensi�e Psychotherapy: TheMultimodal Way. Springer, New York

Lazarus A A 2000 Multimodal therapy. In: Corsini R J, Wed-ding D (eds.) Current Psychotherapies. Peacock, Itasca, IL

Lazarus A A, Beutler L E 1993 On technical eclecticism. Journalof Counseling and De�elopment 71: 381–5

Lazarus A A, Beutler L E, Norcross J C 1992 The future oftechnical eclecticism. Psychotherapy 29: 11–20

Lazarus A A, Lazarus C N 1991 Multimodal Life HistoryIn�entory. Research Press, Champaign, IL

Norcross J C, Alford B A, DeMichele, J T 1992 The future ofpsychotherapy: Delphi data and concluding observations.Psychotherapy 29: 150–8

Paul G L 1967 Strategy of outcome research in psychotherapy.Journal of Consulting Psychology 31: 109–18

Rotter J B 1954 Social Learning and Clinical Psychology.Prentice-Hall, Englewood Cliff, NJ

Williams T A 1988 A multimodal approach to assessment andintervention with children with learning disabilities. Ph.D.thesis, University of Glasgow

A. A. Lazarus

Multinational Corporations

The multinational corporation is a business organ-ization whose activities are located in more than twocountries and is the organizational form that definesforeign direct investment. This form consists of acountry location where the firm is incorporated and ofthe establishment of branches or subsidiaries in foreigncountries. Multinational companies can, obviously,vary in the extent of their multinational activities intermsof the number of countries inwhich they operate.A large multinational corporation can operate in 100countries, with hundreds of thousands of employeeslocated outside its home country.

The economic definition emphasizes the ability ofowners and their managerial agents in one country tocontrol the operations in foreign countries. There is afrequent confusion that equates the ability to controlwith the flow of capital across national borders. SinceHymer’s thesis (1976), it is recognized widely thatcapital flow is not the distinguishing characteristic of amultinational corporation (see International Business).Capital can flow from one country to another inexpectation of higher rates of return. However, thisflow may be invested in the form of bonds, or in equityamounts too insignificant to grant control to foreignowners. In this case, this type of investment is treatedas a ‘portfolio’ investment. The central aspect of‘direct investment’ is the ownership claim by a partylocated in one country on the operations of a foreignfirm or subsidiary in another. The multinationalcorporation is, thus, the product of foreign directinvestment that is defined as the effective control ofoperations in a country by foreign owners.

The economic definition, however, does not capturethe importance of the multinational corporation as theorganizational mechanism by which different socialand economic systems confront each other. Themultinational corporation, because usually it developsin the cultural and social context of one nation, exportsits organizational baggage from one institutionalsetting to another. In this regard, it plays a powerfulrole as a mechanism by which to transfer organ-izational knowledge across borders. However, whilebeing foreign implies that it might serve the valuablerole of importing new practices, its foreign status alsoimplies that its practices are likely to conflict withexisting institutions and cultural norms. Moreover,since multinational corporations are often large, theypose unusual challenges to national and regionalgovernments who seek to maintain political autonomyand yet are often anxious to seek the investment,technology, and managerial skills of foreign firms.

There are, thus, economic and sociological defini-tions of the multinational corporation that differ, andyet complement, each other. In the economic defini-tion, the multinational corporation is the control offoreign activities through the auspices of the firm. Inthe sociological definition, the multinational corpor-ation is the mechanism by which organizationalpractices are transferred and replicated from onecountry to another.

1. History

The multinational corporation is defined in some sensearbitrarily by where frontiers are drawn. In ancientGreece, these frontiers were the borders amongcity–states. In imperial Rome, the new administrativeunits of an expanding empire and its external bound-aries defined the borders. Political borders correspondonly approximately to the distribution of cultural andethnic dispositions. The history of the multinationalcorporation is tied closely to the origins of trade in andbetween cultural communities, and these communitiesremain important in many sectors in the moderneconomy.

Early trade was often characterized by the difficultyof transacting across borders. Trading has alwaysinstigated from the unequal and varied distribution ofresources across geographies. The trading of salt wasan important factor in most continents and still can befound in rather ancient forms in east Africa. Towns,such as Salzburg, owed their origins to their fortuitousaccess to salt mines.

This unequal distribution incited traders to travellong distances and undergo unusual risks for the hopeof gain. The dilemma is quite apparent. How canmarkets for trade develop between distant cultures?Indeed, there is good evidence that early trading wasquite precarious. Brown (1969) notes that Hermes wasnot only the god of messengers but also of theft and

10197


trade. He was often worshipped at borders, markingthe place of trade. Herodotus describes the practice ofthe ‘silent trade’ of goods at boundary markings inwhich one party would deposit their goods at theborder, the second party would later deposit a cor-responding offer. The original party might accept thisbid, take the goods, and leave their initial offer.Clearly, only a strong demand for unusual foreigngoods could sustain such a treacherous institution oftransaction.

Not surprisingly, Curtin (1984) doubts the accuracyof this account. Instead, he stresses the importance ofethnic groups, such as the Gujerati or overseasChinese, or the force of imperialism and militarypower as ways in which ‘cross-cultural trading’ waspermitted to develop. But such solutions remainedfragile, with foreign enclaves and ethnic groups subjectto local hatred and military rule often confrontingcompetition from other powers.

Yet, trade persisted through the centuries. Indeed,the demand for foreign goods was so strong thatincredible profits could be realized by internationaltrade. Since transportation was poor, and technologyto hold inventories rudimentary, an agent in the rightspot at the right time could earn fortunes. And in thislaid the great problem. For in times of great demand,prices could rise astronomically, as Braudel (1973)documents in his history of material life. With suchdistances between the agent and the principal in theexporting country, an agent could easily, and oftendid, disappear with the profits.

The early solutions were many and provide insightinto one of the great properties of the multinationalcorporation: its ability to organize transactions withinits own organizational boundaries. Great fairs wereone solution, where principals met their customers inone place. But fairs were intermittent and often distantfrom final markets. The further development ofRoman commercial law during the medieval andRenaissance periods strengthened the liability of theprincipal and also legal recourse in principal–agentconflict. However, the application of law between twoparties in different legal jurisdictions has always beena difficult form of dispute resolution. In his study ofthe Jewish Magrebi traders of the twelfth century,Greif (1989) places considerable emphasis upon thevalue of membership in ethnic ‘trading communities.’Since ethnic enclaves in foreign sites could monitor theactivities of agents, principals could be informed ofmalfeasance. Agents could then be excluded fromfuture contracts within the community, thus deterringdishonest behavior. Partnerships were also anotherway in which agents were also bonded to the firm.

However, another solution was the company, aword whose roots come from the Latin meaning ‘withbread’ which implies the familial origins of thisbusiness organization. The company permitted in timethe investing of capital in its ownership by outsiders inpromise of future dividends. The capital might be

raised for a single voyage or, eventually, could bemaintained in the company in the form of stockownership. With the evolution of limited liability lawin the nineteenth century, and the diminished role ofthe state in restricting the growth of the joint stockcompany, this organizational form expanded rapidlyin many European countries as well as in the UnitedStates. Along with the growth of equity markets, thefinancial resources to invest overseas became moreavailable due to advances in the banking system andalso in bond markets. Indeed, the great capital needsof the railway industry created an international marketfor the sale and purchase of railroad bonds. However,international loans and bonds were risky and it wasnot uncommon for sovereign governments to default,including state governments within the United States.

The increasing wealth of western countries, alongwith constraints on the speed by which nationalindustries could absorb new loans, encouraged mass-ive foreign investments by the end of the nineteenthcentury and early twentieth century. This was a periodof globalization in terms of the percentages of capitaloutflows to total capital accumulation. Great Britainis estimated to have exported some 25 percent of itscapital prior to World War I and French capitalexports were often greater. This capital went tocountries hungry to finance their industrial expansion,including the newly industrializing countries of Russiaand especially America.

The British were unusual in directing a considerableamount of their outward capital flows in the form ofdirect investments. A great deal of this investmentwent into their colonies. Prior to the mid-1800s,companies operating overseas engaged largely inwholesale operations; they did not run factories,operate mines, or own agriculture. The British directinvestmentswere, thus, different than before, as Britishcame to own and run local operations. Moreover, theBritish invested in South and North America.

The ‘free-standing’ company, as labeled by Wilkins(1988), was a peculiar form of investment that wasquite prominent in many British colonies. This firmraised capital in the domestic financial market whereits administrative office was located, but it operated nodomestic activities. All operations were overseas. Thisform of company represents the advantage of westerncountries in being able to raise capital from efficientfinancialmarkets.However, it posed the same problemas facing the medieval agent relationship, namely, howcould foreign investors be confident that they wouldrecover profitably their investments if the assets werelocated overseas, often in countries with very poorlegal infrastructures? Wilkins found that oversightoccurred through the executive boards of the com-panies, where prominent people located in the homemarket faced the loss of reputation if the companyshould prove to be dishonest.

The United States offers an unusual case. A rapidlygrowing country, it imported more capital than it

10198


exported up to World War I. However, whereas itimported largely portfolio investments, its outwardflows were dominated by foreign direct investments. Inother words, American companies showed an earlypenchant for expanding overseas (Wilkins 1970). TheSinger Company built within two decades of itsfounding a large factory employing thousands ofworkers in Scotland. Oil companies, Kodak,Westinghouse, Ford, and mining and agriculturalcompanies all invested overseas. Companies in oil,mining, and agriculture often invested in poorermarkets where there were resources to be found. Theseearly investors were often involved inextricably in thepolitics of the foreign governments, and the Americanmilitary itself intervened aggressively numerous timesin the Caribbean, Mexico and Central America, andSouth America. As in the British case, the history ofAmerican direct investment occurred in the context ofan expanding military presence of the home govern-ment. Moreover, since many of these countries werepoor, the multinational corporations responded to thedemands of the host nation, especially in the form ofconcessionary contracts, to provide public services,e.g., hospitals, roads, and power (Robinson 1964).This complicated legacy of the early history of themultinational corporation created hostility on the partof the local population that persisted throughout mostof the twentieth century.

It is important to underscore that the multinationalcorporation usually evolved in the context of specificnational institutions. As many others have pointedout, the multinational corporation is a growing firmwhose organizational borders have spilled acrossborders. Moreover, since this large firm is usually tiedto a larger domestic network of suppliers andcustomers, its expansion overseas is accompanied bythe co-investments of these other members. This is apattern seen in American investments in the UnitedKingdom in the 1950s and repeated by Japanesemultinational corporations investing in the UnitedKingdom in the 1980s and 1990s (Dunning 1993).

Chandler (1990) noted that these multinationalcorporations reflected the national characteristics ofmanagement. In comparing the cases of the largestfirms in the United Kingdom, Germany, and theUnited States, Chandler found that differences inmanagerial capabilities, reflecting national institu-tions, explain their success and failure patterns. Heparticularly criticized the managerial capabilities ofBritish firms, a point not shared by some Britishhistorians (Hannah 1999). But more importantly,Chandler’s thesis assumed that size itself constitutedthe realization of scale and scope economies instead ofthe outcome of success and growth. This observationis especially important for understanding the lack oflarge multinational corporations in Italy or in Taiwan,both of which have very successful small firm econ-omies but do not have multinational corporationscomparable to other countries of similar levels of

economic wealth. Yet, both countries are relativelywealthy and successful, and their many small com-panies have achieved high rates of exporting. Even inthe case of the United States, the evidence implies thatAmerican firms, large and small, came to Europe,riding on the back of the national organizing principlesof standardization in work methods (Kogut 1992).Chandler’s larger point of the effect of nationalsystems on firm capabilities is largely accepted; hisbelief that large firms reflect better managementbecause they achieve scale is disputed far more.

2. Organization of Multinational Corporations

The national origins of the multinational corporationinfluence their subsequent organizational evolution.In the United States, the great post-World War IIexpansion of multinational corporations coincidedwith the diffusion of their adoption of the multi-divisional structure. Indeed, the refinement of‘organizational technologies’ permitted Americanfirms to manage their rapidly growing operations on aworldwide basis.

The young American multinational corporationbegan with little knowledge and experience of theforeign market. Initial foreign sales are exportsexecuted within the existing organizational structure.These structures could be functional (e.g., organizedby production and sales) or divisional (e.g., organizedby geographic area or product division). As foreignsales increased, an international division was createdto make the sale and to provide customer support. Inconsequence, the domestic organizational unit wouldsell to the international division its products at aninternal ‘transfer price.’ As this price was set by thedomestic market, it was unlikely to reflect the com-petitive conditions in the foreign country. Moreover,the products were often not designed to the tastes offoreign demand. Because of this internal conflict, theAmerican multinational corporation would replacethe international division by transferring responsi-bilities to area divisions (e.g., Europe, Asia, SouthAmerica) or to product divisions (Stopford and Wells1972). While these structures diminished the internalconflict, they ran the hazard that the divisionalmanagers across areas did not cooperate or thatmanagers across product divisions tended to fall backupon focusing on the home market.

European firms grew up in a different environment.Europe was a highly unstable continent during thetwentieth century, with significant political and econ-omic conflicts. Borders sometimes changed, and theyalso marked the necessity to pay commercial duties.The gradual creation of the European Union in thelast third of the twentieth century eliminated tariffsand lessened national differences. This history markedtheir firms. The organizational technologies in most of

10199


thesecountrieswerequitedifferent to those found in theUnited States. Holding company structures werecommon, banks or other financial institutions playedpowerful roles in ownership and cross holdings, andfirms were tied together in complex financial webs.These structures, often called ‘mother–daughter’organizations, reflect the linguistic terms used innorthern Europe to describe headquarters and sub-sidiary. These mother–daughter structures consistedoften of a headquarters that held a portfolio consistingof ownership control over dozens of companies, manyof them in the same sector but located in differentcountries. Reporting relationships among thesecompanies and headquarters was less formal thanfound in American companies, and control was oftenexercised through the movement of managers whowere lifetime employees. These structures were oftenseen as inefficient, especially compared to the size andorganization of American multinationals that wereinvesting rapidly in Europe at this time. Europeancritics called for the restructuring of European firmsand industry along American lines.

The structure of Japanese companies reflected alsotheir historical origins. In Japan, the trading companyperformed an extraordinary role in the exports duringthe first two-thirds of the twentieth century. Thesecompanies started primarily as wholesalers, but overtime developed their own industrial companies locatedin Japan and eventually overseas. Though the forcedimposition of American antitrust law after World WarII disrupted these patterns, trading companies andother affiliated company networks, called keiretsu,reconstituted the earlier pattern of constellations ofdomestic companies competing in home and, throughtheir trading companies, in foreign markets. As in theAmerican and European cases, Japanese firms variedin their internal structures, with functional and evenfactory-based organization being the predominantstructures. Still, a dominant trend emerged over timeas the role of trading companies declined, especiallyfor the exportation of industrial and high technologygoods. This trend consisted of the appendage of aninternational division to the existing structure (Suzuki1991).

These different national structures confronted anincreasingly more globally integrated environment inthe course of the twentieth century. This environmentposed the classic organizational problem of balancingintegration and differentiation. For the multinationalcorporation, this problem posed itself as meeting thedemands to achieve global scale against the needs ofnational markets and governments. Given differentnational principles of organizing the activities ofmultinational corporations, there was not a rapidconvergence to a global structure. The growth ofmultinational corporations, especially in Europe, wasstymied by these traditions. Continuing nationaldifferences and the failure to develop harmonizedEuropean corporate law deterred the emergence of

pan-European multinational corporations. Many ofthe most important cross-European mergers in the1970s failed within a decade. Instead, Americanmultinational corporations, less bound by these sep-arate national traditions, were able to create integratedEuropean strategies.

Nevertheless, it was in Europe where new organ-izational structures developed to resolve the conflictbetween integration and differentiation. Multinationalcorporations from smaller countries offered the lab-oratory experiments that influenced the evolutionof multinational corporations from larger nations.Because of the small size of many European countries,large firms in Sweden, Switzerland, and theNetherlands quickly became multinational corpor-ations, and these corporations had far more assets andemployees outside their home countries than within.Over the course of the last two decades of the twentiethcentury, organizational structures developed, labeledas transnational corporations, that distributed globalresponsibilities to national subsidiaries, created globalteams and projects, and encouraged the transfer of‘best practices’ across borders.

3. Diffusion of Organizational Knowledge

In its evolution the multinational corporation is notwithout serious contradictions. Evolving from itsnational context, the multinational corporation em-ploys large numbers of employees of diversenationalities and ethnicities. Westney (1993) notesthat a subsidiary is, thus, caught between the in-stitutional pressures to conform to the company normsand values, as well as to the cultural and socialinfluences of its local national environment. At theheart of the evolution of the multination corporation,thus, lies the tension between national institutions andthe fragile emergence of a global culture.

The international evolution of the organizationalstructures of American multinationals mirrored, as wenoted above, the broader diffusion of organizationaltechnologies in the home market. The initial invest-ments by a firm took place, often, on the basis ofopportunity and the extension to familiar countries.Much like the ethnic trading communities dating backto the earliest times, the inexperienced multinationalcorporation preferred countries that are culturallysimilar to what their managers know at home(Johanson and Vahlne 1978). In these countries, theyoften established foreign enclaves where their ex-patriate managers could live in the simulatedfamiliarity of their home environments.

The relevance of the second definition of foreigndirect investment as the transfer of organizationalknowledge is critical to understanding the powerfulconflicts posed by the multinational corporation. Themultinational corporation, competing often on su-perior technologies and managerial capabilities, serves

10200


as a conduit of knowledge across borders. Some kindsof technologies can be purchased on markets and theflow of licensing payments attached to the sale of theright to use a technology constitutes a nontrivial flowin international balance of payments. Technology canalso flow by the movement of people. Just as Englishcraftsmen were imported by continental Europe in theearly industrial period, there exists an internationalmarket for managers and skilled workers.

However, some knowledge is embedded in organ-izations and can only be transferred as an organizationacross borders. The issue is more than whetherknowledge is tacit or explicit, for this pertains to theknowledge held by individuals too. The central featureof the diffusion of knowledge by the multinational isthat it transfers the knowledge of how to organize andhow to coordinate people and across divisions. Theseorganizing principles then provide the capabilities forthe firm to achieve quality, speed products to themarket, or lower costs (Kogut 1992). For this themultinational corporation is required, which throughits own activities or in joint ventures with hostcompanies, transfers the organizational knowledge.

The global market for knowledge extends also toconsultants. The large American consulting practicesoften have their origins in their acquisition ofAmerican organizing principles that they then trans-ferred around the world. The international diffusionof the divisional structure, for example, reflects theknowledge acquired in American consulting practicesthat was then sold in abroad. Channon (1973), forexample, observes that half of the firms he observed asadopting the divisional structure relied upon theconsulting services of the same American firm. Simi-larly, British and then American banks spreadthroughout the world on servicing the needs of theirexpanding home clients, and then on transferring theirpractices into these countries (Jones 1993).

It is important to realize that once this transitionperiod passed, the multinational corporation andmultinational service firms, such as those engaged inconsulting, banking, and advertising, had permanentlyand historically changed the global economy. Nolonger was the flow from the US to Europe, or fromone ethnic community to another. The network ofconnections developed by the multinational corpo-ration permitted the flow of ideas and practices amongand between countries. Thus, whereas it took half acentury for American practices to diffuse to Europe,the introduction of quality circles from Japan hap-pened within a decade or two of their originalinnovation. In Europe, Japanese innovations wereoften diffused by American firms, including consultingcompanies.

However, this image of the free flow of knowledgeneeds to be strongly conditioned on the continuingimportance of national institutions. Nations consist ofdefined cultures and economic and social institutions,such as unions, financial systems, and religious values.

These institutions interact and their complex inter-actions causally influence behaviors. For example, theGerman system of centralized bargaining, enterpriseclubs, strong banks, and social welfare has, untilrecent times, composed a national configuration ofinstitutions that influences the capabilities of residentfirms to be able to manufacture high quality goods forexport markets (Streeck 1995, Soskice 1990). Theintroduction of Japanese and American teams con-fronts in Germany the presence of existing workcouncils. These councils are fairly rigid features of theGerman environment. Since Japanese methods maynot be effective without teams as complementaryfactors in organization, this institutional refusal caneffectively deter the diffusion of these methods. Inshort, national organizing principles are embedded inthe wider social institutions. Inconsistency betweenthese institutions and principles, then, is an empiricalquestion of the bargaining strength of vested powers.

4. Economics and Politics of MultinationalCorporations

Since the multinational corporation is definitionallyequivalent to foreign direct investment, theories offoreign direct investment must account for why onecountry invests in another and why this investment iscarried out within organizational boundaries of a firm(see Buckley and Casson 1976, see Foreign In�estment:Direct). In distinguishing between portfolio and directinvestment, Hymer noted that firms operate at adisadvantage in foreign markets and hence they musthave an offsetting competitive advantage to competeoverseas. These advantages for overseas investmentsare the same ones that allow a firm to compete andgrow in the home market. These observations haveimportant implications. The first is that direct invest-ment is the growth of the firm across borders andhence the firm expands internationally on what it haslearned at home. This observation is the basis for theevolutionary theory of the firm. The second obser-vation that Hymer made is that firms that expandoverseas, because they have competitive resources, arealso likely to be large and to belong to oligopolisticindustries.

In these observations, we can understand the am-bivalence expressed in popular and policy debatesregarding the multinational corporation. Competitionamong multinational corporations often is the ex-tension of their home domestic and oligopolisticrivalry that spills across national borders. In manyglobal industries, the same company names dominateeach country’s list of the largest firms inside theirnational frontiers. No matter if it is Poland or France,Singapore or Mexico, the same multinational cor-porations will be found in the local oligopolisticindustries (e.g., consumer goods or automobiles).

10201


Because they are large even in their home markets,investments by multinational corporations can have alarge impact on a host country (Caves 1974).

As a consequence, the multinational corporationhas often been the subject of debates concerningnational sovereignty and welfare. In recent decades,acquisitions have generally been the primary way bywhich multinationals invest in wealthy foreigncountries, where the vast proportion of direct in-vestment is concentrated. Given the size of a multi-national corporation and occasional nationalimportance of the targeted acquisition, even wealthycountries frequently evidence discomfort, if not out-right public hostility, to multinational investments.Moreover, multinational corporations are sometimesthe vehicles for foreign policies of their home or hostcountry. The decision, for example, of the US toembargo technology and investment flows to Cuba,the former Soviet Union, Iran, and other countriesperiodically has caused conflict with other countries.

Multinational corporations are especially problem-atic in developing countries. By definition, developingcountries are relatively poor, thus both in need ofcapital and yet concerned over their loss of inde-pendence. As discussed above, the history of multi-national corporations in developing countries ismarked by its origins in policies of imperialism andcolonialism. Especially in Latin America, where aschool of thought labeled Dependencia has been in-fluential, the concern over dependence on the UnitedStates resulted in efforts to curb the power of themultinational corporations by restricting the amountof equity ownership a foreign firm could hold in adomestic company or by prohibiting investment incertain sectors. Mexico’s constitution forbids foreigninvestment in the oil industry; Brazil pursued for along time a policy to restrict foreign participation inthe electronics industry.

The other side of the coin is that multinationalcorporations bring investment and technology to theforeign country. Vernon (1966) hypothesized thatinnovations start in wealthy countries. As the marketis saturated and as oligopolistic rivalry increases,multinational corporations are pushed out from theirhome markets to expand abroad in new markets andto locate less expensive places. Thus, Vernon seizedboth sides of the debate, recognizing the value of thetransfer of technology but also emphasizing theoligopolistic nature of multinational investment.

It is, in fact, difficult to draw simple conclusionsregarding the relationship of foreign investment andnational growth. Countries such as Singapore,Malaysia, and Thailand have encouraged foreigndirect investment actively. The growth in China’scoastal sector is indisputably linked to the massiveinvestments by multinational corporations. However,historically Japan and Korea have pursued morecautious policies regarding investments by multi-national corporations. In these countries, the state has

often negotiated the terms for entry by multinationalcorporations, sometimes requiring licensing to dom-estic competitors as a price. The efficacy of suchpolicies for these countries is much disputed. However,for many other countries, the intervention of thegovernment in demanding licenses unquestionablyleads to internal corruptionand to insufficientdomesticcompetition.

There are many channels by which a country canabsorb foreign technology and managerial techniques.Most of the evidence shows, however, that prohibi-tions on the in-flows of direct investment can be verycostly for many countries. With their domestic in-dustries still to be developed, a developing countryrequires substantial investment. Some countries, pri-marily in Asia, have been able to achieve very highsavings rates to finance their industries without directinvestment. Moreover, high savings rates, plus pol-itical stability, create growth, and growth attractsforeign portfolio capital. A poor country that pro-hibits foreign direct investment but does not have highrates of saving is entirely dependent upon portfoliocapital. The history of debt and currency crises in the1990s convinced many poor countries that foreigndirect investment was a preferable means of attractingcapital, because it could not be easily pulled out of acountry on short-notice in response to a financialcrisis.

However, multinational corporations also respondto the volatility in the global market. This volatilityderives from changes in exchange rates, politics, andproductivity. Once having achieved sufficient experi-ence and having established subsidiaries around theworld, the multinational corporation might choose toclose a plant in one location and open plants in newlocations. Of course, such actions might provoke aresponse by labor, but historically, labor has beenorganized by national, not by international, organ-izations (Martinelli 1975). Yet, there is also thepossibility that locations lose some kinds of plants butgain more sophisticated investments. Cantwell (1999)proposed that some regions and countries pullmultinational investments. Yet, it has long beennoticed that foreign direct investment among devel-oped countries flows to high cost locations. Regionssuch as Silicon Valley, Baden-Wuertemberg, andSingapore attract multinational investments not be-cause wages are low, but because productivity levelsare high and workers are well trained. In many cases,developing countries have given rise to their ownmultinational corporations acting in the region andsometimes globally (Lall 1983). In this sense, themultinational corporation acts as a training center inthe developmental strategies of emerging economies.

5. Globalization

The peculiar conflict in the world economy is thegrowing trend toward economic integration without a

10202


concomitant growth in global political and socialinstitutions to regulate and arbitrate this trend.Multinational corporations are rapidly adopting ad-vanced information technologies to increase theefficiency and capabilities of their operations. Valuableinformation, such as software or financial services, istransmitted digitally. Governments are often unableto tax the value of these services or control theircontent. Moreover, information technologies, oftensupported by private telecommunication networks,permit, for example, a unit in Germany to control themanufacturing operations located in Brazil.

Digital technologies permit a high degree of in-tegration and, in the short run, aggravate the gapbetween countries that are rich and those that are poorand do not have the infrastructural capabilities.However, these technologies are beginning to haveprofound effects on some regional economies. In partsof China, India, Israel, and elsewhere, advancedsatellite transmission transmits digitally encoded workbetween their sites and other organizational units ofmultinational corporations located elsewhere.Whereas before the Indian engineer from Bangaloremight have tried to migrate to the US to bring hishuman capital to a more attractive labor market, theincreasing wages and job prospects in India areencouraging many to stay at home and participatedigitally in the world economy.

These trends have unclear effects on multinationalcorporations. They permit more easily the develop-ment of projects that are in continual development, aswork passes from one unit to the next as the dayadvances. They also allow more easily the coupling ofless expensive labor in one country with more ex-pensive skilled labor in another.

At the same time, the origins of the multinationalcorporation laid in its ability to organize labor on aworldwide basis on principles other than ethnic andcultural identities. This organization has never beenwithout cost and risk. The growth of a world digitaleconomy permits alternative ways by which labor cancooperate and be coordinated on a world basis. Theintriguing question at this point in history is whetherthe multinational corporation, though still a vitalpresence in the world economy, nevertheless willrecede relatively in importance as information tech-nologies reduces the meaning of geographic distance.

See also: Development and the State; Diaspora;Globalization: Geographical Aspects; Globalization:Legal Aspects; Globalization: Political Aspects;Globalization, Subsuming Pluralism, TransnationalOrganizations, Diaspora, and Postmodernity; Inter-national Business; International Law and Treaties;International Marketing; International Organization;International Trade: Economic Integration

Bibliography

Braudel F 1973 Capitalism and Material Life, 1400–1800 (trans.Kochan M). Harper and Row, New York

Brown N O 1969 Hermes the Thief: The E�olution of a Myth, 2ndedn. University of Wisconsin Press, Madison, WI

Buckley P J, Casson M 1976 The Future of the MultinationalEnterprise. MacMillan, London

Cantwell J 1989 Technological Inno�ations and MultinationalCorporations. Blackwell, London

CavesR E1974Causesofdirect investment:Foreignfirms.Sharesin Canadian and United Kingdom manufacturing industries.

Re�iew of Economic Statistics 56: 79–93Chandler A D 1990 Scale and Scope: The Dynamics of Industrial

Capitalism. Belknap Press, Cambridge, MAChannon D F 1973 The Strategy and Structure of British

Enterprises. Harvard University, BostonCurtinP D1984Cross-culturalTrade inWorldHistory. Princeton

University Press, Princeton, NJDunning J 1993 The governance of Japanese and US manu-

facturing affiliates in the UK: Some country-specific differ-ences. In: Kogut B (ed.) Country Competiti�eness: Technologyand the Organizing of Work. Oxford University Press, Oxford,UK

Greif A 1989 Reputation and coalitions in medieval trade:Evidence on the Maghribi traders. Journal of EconomicHistory 39: 857–82

Hannah L 1999 Marshal’s ‘trees’ and the global ‘forest’: Were‘Giant Redwoods’ different? In: Lamereaux N R, RaffD M G, Temin P (eds.) Learning by Doing in Markets, Firms,

and Countries. University of Chicago Press, Chicagopp. 253–86

Hymer S 1976 The international operations of national firms: Astudy of direct investment. Ph.D. thesis, MassachusettsInstitute of Technology

Johanson J, Vahlne J-E 1978 A model for the decision makingprocess affecting pattern and pace of the internationalization

of the firm. In: Ghertman M, Leontiades J (eds.) EuropeanResearch in International Business. North Holland,Amsterdam pp. 9–27

Jones G 1993 British Multinational Banking, 1830–1990. OxfordUniversity Press, New York

Kogut B 1992 National organizing principles of work and the

erstwhile dominance of the American multinational corpor-ation. Industrial and Corporate Change 1: 285–325

Lall S 1983 The New Multinationals: The Spread of Third WorldEnterprise. Wiley, Chichester, UK

Martinelli A 1975 Multinational corporations, national econ-omic policies and labor unions. In: Lindberg L, Alford R,

Crouch C, Offe C (eds.) Stress and Contradiction in ModernCapitalism. Heath, Lexington, MA pp. 425–43

RobinsonR D1964InternationalBusinessPolicy.Holt,Rinehart,and Winston, New York

Soskice D 1990 Wage determination: The changing role ofinstitutions in advanced industrialized countries. Oxford

Re�iew of Economic Policy 6: 36–61Stopford J M, Wells L T 1972 Managing the Multinational:

Organization of the Firm and Organization of the Subsidiaries.Basic Books, New York

Streeck W 1995 German capitalism: Does it exist? Can it survive?In: Crouch C, Streeck W (eds.) Modern Capitalism or Modern

Capitalisms? Pinter, LondonSuzuki Y 1991 Japanese Management Structures, 1920–1980. St.

Martin’s Press, New York

10203


Vernon R 1966 International investment and international tradein the product life cycle. Quarterly Journal of Economics 80:190–207

Westney D E 1993 Institutional theory and the multinationalcorporation. In: Westney D E, Ghoshal S (eds.) OrganizationTheory and the Multinational Corporation. St. Martin’s Press,New York

Wilkins M 1970 The Emergence of Multinational Enterprise:American Business Abroad from the Colonial Era to 1914.Harvard University Press, Cambridge, MA

Wilkins M 1988 The free-standing company, 1870–1914: Animportant type of British foreign direct investment. EconomicHistory Re�iew 39: 259–82

B. Kogut

Multiple Imputation

Missing data are ubiquitous in social science research.For example, they occur in a survey on people’sattitudes when some people do not respond to all ofthe survey questions, or in a longitudinal survey whensome people drop out. A common technique forhandling missing data is to impute, that is, fill in, avalue for each missing datum. This results in acompleted data set, so that standard methods thathave been developed for analyzing complete data canbe applied immediately (see Nonsampling Errorsand Statistical Data, Missing).

Imputation has other advantages in the context ofthe production of a data set for general use, such as apublic-use file. For example, the data producer can usespecialized knowledge about the reasons for missingdata, including confidential information that cannotbe released to the public, to create the imputations;and imputation by the data producer fixes the missingdata problem in the same way for all users, so thatconsistency of analyses across users is ensured. Whenthe missing-data problem is left to the user, theknowledge of the data producer can fail to beincorporated, analyses are not typically consistentacross users, and all users expend resources addressingthe missing-data problem.

Although imputing just one value for each missingdatum satisfies critical data-processing objectives andcan incorporate knowledge from the data producer, itfails to achieve statistical validity for the resultinginferences based on the completed data. Specifically,for validity, the resulting estimates based on the datacompleted by imputation should be approximatelyunbiased for their population estimands, confidenceintervals should attain at least their nominal cover-ages, and tests of null hypotheses should not reject truenull hypotheses more frequently than their nominallevels. But a single imputed value cannot reflect any ofthe uncertainty about the true underlying value, andso analyses that treat imputed values just like observedvalues systematically underestimate uncertainty.Thus, using standard complete-data techniques will

result in standard error estimates that are too small,confidence intervals that undercover, and P-valuesthat are too significant; even if the modeling forimputation is carried out carefully. For example, large-sample results by Rubin and Schenker (1986) showthat for simple situations with 30 percent of the datamissing, single imputation under the correct modelfollowed by the standard complete-data analysisresults in nominal 90 percent confidence intervalshaving actual coverages below 80 percent. The inac-curacy of nominal levels is even more extreme inmultiparameter problems (Rubin 1987, Chap. 4),where nominal 5 percent tests can easily have rejectionrates of 50 percent or more when the null hypothesis istrue.

Multiple imputation (Rubin 1987) retains the ad-vantages of single imputation while allowing the dataanalyst to obtain valid assessments of uncertainty. Thekey idea is to impute two or more times for the missingdata using independent draws of the missing valuesfrom a distribution that is appropriate under theposited assumptions about the data and the mech-anism that creates missing data, resulting in two ormore completed data sets, each of which is analyzedusing the same standard complete-data method. Theanalyses are then combined in a simple generic waythat reflects the extra uncertainty due to havingimputed rather than actual data. Multiple imputationscan also be created under several different modelsto display sensitivity to the choice of missing-datamodel.

1. Theoretical Moti�ation for MultipleImputation

The theoretical motivation for multiple imputation isBayesian (see Bayesian Statistics), although the pro-cedure has excellent properties from a frequentistperspective. Formally, let Q be the population quan-tity of interest, and suppose the data can be partitionedinto the observed values Xobs and the missing valuesXmis. If Xmis had been observed, inferences for Q wouldhave been based on the complete-data posteriordensity p(Q rXobs, Xmis). Because Xmis is not observed,inferences are based on the actual posterior densityp(Q rXobs), which can be expressed as

p (Q rXobs)¯& p(Q rXobs, Xmis) p (Xmis rXobs)dXmis (1)

Equation (1) shows that the actual posterior density ofQ can be obtained by averaging the complete-dataposterior density over the posterior predictive dis-tribution of Xmis. In principle, the set of multipleimputations under one model are repeated indepen-dent draws from p(Xmis rXobs). Thus, multiple im-putation allows the data analyst to approximate Eqn.(1) by separately analyzing each data set completed byimputation and then combining the results of theseparate analyses.

10204


Copyright # 2001 Elsevier Science Ltd. All rights reserved.International Encyclopedia of the Social & Behavioral Sciences ISBN: 0-08-043076-7

chapter in smelser-baltes 2001

Documents