functions in innovation system approach

Upload: soloensis

Post on 06-Apr-2018

214 views

Category:

Documents


0 download

TRANSCRIPT

  • 8/3/2019 Functions in Innovation System Approach

    1/19

    1

    Functions in Innovation System Approaches

    Anna Johnson

    Department of Industrial DynamicsChalmers University of TechnologySE-412 96 Gteborg

    Sweden

    Phone: +46-(0)31-772 12 22Fax: +46-(0)31-772 12 37

    E-mail: [email protected]

    Abstract

    In the last twenty years, a large number of innovation system approaches have emerged.Even though there are similarities between different approaches, they emphasisedifferent aspects of innovation systems. Moreover, within each approach the systemmodels differ in terms of the concepts used and the actors identified and emphasised. Itmight, therefore, be useful to see if there is any agreement between the approaches withrespect to what they claim happens in an innovation system, i.e. what functions are

    served in the system. The purpose of this paper is twofold: (1) to find out if existinginnovation system approaches share an understanding of the functions that are served inan innovation system and (2) to elaborate on the usefulness of the concept of functionin innovation system studies. In summary, there seems to be quite a widely spreadcorrespondence between different innovation system approaches with respect to thefunctions they mention so that a number of common, basic functions can be identified.The concept of function has several benefits. It provides a tool for setting systemborders and may be used as a tool to diagnose the present state of a system or to createan easily grasped picture of innovation system dynamics. It makes it possible to assesssystem performance and may be useful in comparative studies since actors may beuncoupled from what happens in the system.

  • 8/3/2019 Functions in Innovation System Approach

    2/19

    2

    1.IntroductionAn increasing number of researchers in fields such as industrial dynamics, technology

    policy and firm strategy claim that technological development cannot be viewed as an

    isolated phenomenon but has to be studied as a part of a larger system, an innovation

    system.1 A number of system approaches have emerged, including the national

    systems of innovation approach (e.g. Edquist and Johnson (1997), Lundvall (1992) and

    Nelson (1992)), the technological systems approaches (e.g. Hughes (1983) and Carlsson

    and Stankiewicz (1995)), the sociotechnical systems approach (e.g. Bijker (1995)) and

    the network approach (e.g. Hkansson (1990)).

    Even though there are similarities between these approaches, they emphasise different

    aspects of innovation systems, mostly due to underlying differences in the fields of

    research. For example, the industrial network approach focus on how a companys

    relationships to other actors influence its capability to become a successful innovator,

    whereas advocates of the sociotechnical systems approach are interested in the social

    processes influencing the way we perceive technological artefacts.

    Moreover, within each approach authors often focus on describing the characteristics of

    innovation systems on the basis of case studies, resulting in quite case-specific system

    models that differ in terms of the concepts used and the actors identified and

    emphasised.

    The differences between and within approaches make it difficult to compare, or indeed

    combine, the findings of different system approaches. It might, therefore, be useful tolook beneath their surface to see if there is any agreement between the approaches with

    respect to what the claim happens in an innovation system.

    For this purpose, the concept of function may be of use. Inherent in a systems view is

    a notion that all system components contribute to the goal of the system or they would

    1 For reasons of simplicity, all system concepts are here labelled innovation systems.

  • 8/3/2019 Functions in Innovation System Approach

    3/19

    3

    not be considered part of that system.2 The contribution of a component or a set of

    components to the goal is what is here called a function.3 The question is, thus, whether

    the different innovation system approaches agree on the functions that are served by

    components in an innovation system or if there are differences between them also in this

    respect.

    The purpose of this paper is twofold: (1) to find out if existing innovation system

    approaches share an understanding of the functions that are served in an innovation

    system and (2) to elaborate on the usefulness of the concept of function in innovation

    system studies.

    The paper is structured as follows. In section two, some different innovation system

    approaches and some of the authors representing them will be described with a focus on

    the functions they identify. In section three, the approaches views of the functions

    served in innovation systems will be compared. In section four, the analytical usefulness

    of the concept of function will be discussed and section five concludes the paper.

    2.Functions in Innovation System ApproachesIn this section some innovation system approaches will be described with respect to the

    functions they identify.4 The intention has not been to make a complete scrutiny of the

    literature, but rather to study some representatives of the major approaches. Three major

    approaches have been identified: the national systems of innovation approach, the

    technological systems approach and the network/development block approach. The

    classification of authors into these approaches has been made on basis of a perception of

    affinity between authors with regards to level or unit of analysis.5

    The three approacheswill be described in sections 2.2-2.4. However, since functions have been defined as

    contributions to the systems goal, the goal of an innovation system must first be

    discussed.

    2 In an analytical sense. This does not imply that the system in reality is directed by some overriding principle (see also below).3 To some extent, this view of the concept of function resembles the ones used in political science, although Almond and Powell(1967) include both system capabilities, i.e. what the system does, and the mechanisms by which the capabilities are achieved intheir description of the functioning of a political system. It is also similar to Talcott Parsonss notion of functions as contributions ofelements in social system to the continuing operation of the whole (Cuff et al, 1998).4 The description, thus, includes some analysis; since the authors themselves have not used the concept of function, describing them

    in those terms makes it necessary to reformulate their writings somewhat.

  • 8/3/2019 Functions in Innovation System Approach

    4/19

    4

    2.1The Goal of an Innovation SystemThe goal of an innovation system may be said to be to develop, diffuse and utilise

    innovations.6

    This is, evidently, an analytically constructed goal. Few (if any) innovation systems

    have been consciously created with an explicit goal to develop, diffuse and utilise

    innovations (even though systems may be modified in such a direction by policy or

    strategy). Moreover, actors in the system may very well be driven by individual goals

    that do not correspond with each other or with the systems goal (as here defined), e.g.

    profits or social welfare.

    Nevertheless, for the purpose of identifying functions in innovation system approaches

    it is analytically useful to perceive them as contributions to the development, diffusion

    and utilisation of innovations.7

    2.2The National Systems of Innovation ApproachEven though all national system of innovation agree that the nation or country is an

    appropriate level of analysis, there are many different opinions of how the system

    should be defined more exactly. The definitions range from narrow (including only

    organisations performing R&D) to broad (including basically anything that affects

    learning). The reader should keep this in mind since the definition, of course, affects

    what functions the authors identify.

    Porter (1990) discusses the importance of national characteristics for the competitive

    advantage of individual companies. The basic idea is that in addition to the function

    served by companies in terms of acts of innovation (for example through R&D or small

    advances based on existing knowledge) several functions are served by components in a

    larger system of which the companies form a part.

    5 It should be noted that the authors in some cases have referred to and built on each other's work even though they are presentedhere as separate from each other.6 This definition is inspired by Carlsson and Stankiewicz (1995) and Galli and Teubal (1997).7 The concept of function has been used before in the IS literature in slightly different ways than in this paper. For example, Galliand Teubal (1997) use the concept of function as a synonym to role (e.g. R&D and policy-making). McKelvey (1997) use the

    concept of function as synonymous to three principles of change that should exist in an innovation system: retention andtransmission of information, generation of novelty and selection.

  • 8/3/2019 Functions in Innovation System Approach

    5/19

    5

    One of these functions is guiding innovative activity, i.e. influencing the direction in

    which companies deploy their resources and skills. For example, home demand might

    influence companies perceptions of opportunities, and standards and security

    regulations help (or force) companies to choose specific development paths.

    The role of the government is, among other things, to stimulate markets, provide

    infrastructure and education and give companies incentives to invest in innovation. The

    latter two are of vital importance since the supply of resources, both in terms of

    (venture) capital and competence, is necessary to achieve competitive advantage.

    In contrast to Porter, Lundvall is not primarily interested in industrial competitiveness,

    but in the production, diffusion and use of new, and economically useful, knowledge in

    general. The paper referred to here (Lundvall, 1992) is, however, largely focused on the

    role of learning in the production of new knowledge.

    According to this paper, there are several sources of new knowledge. Whereas some of

    the inputs to the knowledge production process originate from search and explorationefforts in R&D, other arise from learning in connection to everyday activities. Often, the

    identification of bottlenecks in technology affects the direction of problem definition

    and the search for solutions. The rate and direction of innovation are also influenced by

    regulations and standards. Furthermore, providing the resources and competencies

    needed is an important task of the system of innovation.

    As is the case with Lundvall, Edquist and Johnson (1997) see innovations as results ofinteractive learning, where knowledge is combined in new ways or new knowledge is

    created, either in connection with R&D or in relation to everyday activities such as

    production and marketing.

    The actors (organisations) play an important role, as they search for new knowledge,

    absorb knowledge created elsewhere and utilise the expected and unexpected results.

  • 8/3/2019 Functions in Innovation System Approach

    6/19

    6

    They also contribute to the distribution of knowledge, knowledge regulation (e.g.

    standard setting) and the development of the institutional set-up.

    The primary concern of Edquist and Johnson is, however, the institutional set-up,8

    which serves four basic functions with respect to innovation. First, institutions (e.g.

    patent laws, norms for repayment periods etc.) may reduce uncertainty, either by

    providing information about the behaviour of other people or by reducing the amount of

    information needed. Second, institutions manage conflicts and co-operation between

    individuals and groups. For example, practices of co-operation may work as bridges

    between different departments in a company. Another example is social security and

    labour market arrangements that influence conflicts between old products and new,thereby counteracting resistance to change. Third, institutions provide incentives to

    engage in learning and to participate in innovation processes. The incentives can be of

    various kinds, e.g. income taxes, property rights, perceived competitive advantage and

    status norms. Finally, institutions such as tax rules, government subsidies and allocation

    of resources to universities channel resources to innovation activities and also help to

    re-channel resources from ailing activities to new ones.

    Whereas the papers described so far are largely theoretical in nature, the work of Nelson

    (1992) is based on a comparative study of the national innovation systems of 15

    countries. Among the functions served by different actors in the countries studied, the

    investment in and implementation of R&D are given most attention.

    In this context, the direction of innovative work is said to be an important function, one

    that often is served by customers. Moreover, economic incentives to innovate have to beprovided and the innovative work has to be funded by internal financing of R&D and/or

    the supply of venture capital. Other necessary resources, e.g. competence (knowledge

    and skills), also have to be made available.

    New knowledge may also come from other sources than R&D, e.g. from learning-by-

    doing, learning-by-using or imitation. For learning to occur, it is, however, necessary to

    8 Institutions are defined as sets of common habits, routines, established practices, rules, or laws that regulate the relations andinteractions between individuals and groups (Edquist and Johnson, 1997, pp. 46).

  • 8/3/2019 Functions in Innovation System Approach

    7/19

    7

    co-ordinate departments within companies, to spread knowledge about new

    developments, and to promote co-operation between companies.

    Finally, it might also be necessary to support (or even create) markets. This may involve

    providing companies with incentives to export.

    2.3The Technological Systems ApproachAs in the national systems of innovation approach, the authors that are here considered

    to be part of the technological systems approach define their systems in different ways.

    Moreover, some use other names for their system concepts. However, they all have

    chosen a technology (or product) level of analysis. The issue of interest for these

    authors is, thus, how different actors etc influence the development, diffusion and use of

    a particular technology or product.

    The concept of technological system is said to have been first introduced by Thomas

    Hughes in the book Networks of Power.9

    According to Hughes (1983, 1990) 10, technological systems solve problems identifiedor constructed by the system. The identification of a problem usually involves

    identifying a demand and the available resources that might fill it. Inventions are

    sometimes based on the identification of reverse salients, i.e. components in the

    system that have fallen behind or are out of phase with the others (Hughes, 1983;

    Hughes, 1990), but they can also be based on experiences from using existing

    technology (Hughes, 1983) or on earlier inventions that failed to develop into

    innovations (Hughes, 1990).

    In addition to invention and development of new technology, there is a need to provide

    financial support (Hughes, 1990) and competence (Hughes, 1983) and to raise

    9 However, authors Edward Wenk Jr and Thomas J. Kuehn used the concept already in 1977 as synonymous to their concept of"technological delivery system" (Wenk and Kuehn, 1977).10 The concept of technological system is not well defined in the 1983 book, but in the 1990 paper some system components are

    mentioned: physical and legislative artefacts, organisations, natural resources, books, articles, university teaching and research,inventors, industrial scientists, engineers, managers, financiers and workers.

  • 8/3/2019 Functions in Innovation System Approach

    8/19

    8

    complementary resources (Hughes, 1983). Other important functions are technology

    transfer and adaptation.

    Hughes (1990) also stresses the importance of institutional factors, e.g. values in the

    society and legislation, in the selection of technical, organisational and social solutions.

    For a new technology (or a new technological system) to develop it may be necessary to

    clear political and legislative ground (Hughes, 1983), stimulate enthusiasm for the new

    technology (Hughes, 1983), and through the destruction of alternative systems force

    unity from diversity (Hughes, 1990). It may also be important to reduce uncertainty by

    taking control of the environment (Hughes, 1990).

    In the early 1990s, Carlsson and Stankiewicz developed a notion of a technological

    system largely independent from the one of Hughes, which is described in, e.g.,

    Carlsson and Stankiewiz (1995).11

    They describe innovation in terms of search and experimentation (i.e. learning), which

    makes exchange of information essential. Moreover, for a technological area to grow,

    someone has to perceive the possibility of the technology and identify the potential forgrowth, the resources needed (in terms of funding, competence etc) have to be secured

    and the different activities in the process co-ordinated.

    Furthermore, the institutional framework has an important role to play. Institutions12

    promote stable pattern of social interactions and transactions, reduce social uncertainty

    and prevent or mitigate conflicts. They may also absorb and diffuse some part of the

    risk of individual actors, e.g. by stimulating markets or providing information, and maywork for the creation of efficient selection mechanisms, both at company and market

    level.

    In addition to the authors that explicitly speak of technological systems, there are others

    using very similar concepts, e.g. Anders Lundgren, Gunnar Eliasson and Wiebe Bijker.

    11 They define a technological system as a network of agents interacting in a specific economic/industrial area under a particularinstitutional infrastructure or set of infrastructures and involved in the generation, diffusion, and utilization of technology

    (Carlsson and Stankiewicz, 1995, pp. 49). This concept is very similar to the one developed by Wenk and Kuehn (1977), althoughno reference is made to their work.

  • 8/3/2019 Functions in Innovation System Approach

    9/19

    9

    Lundgren (1993) uses the concept of industrial networks,13 defined as technical

    systems14 and networks of relationships between actors. Within industrial networks,

    innovation is often driven by the identification of imbalances or bottlenecks in the

    technological system or the relationships between actors. However, since problems are

    defined differently by different actors, due to the fact that their historical background

    influence the way they perceive the situation at hand, the resulting research and product

    development will follow different routes.

    For networks of relationships to emerge, it is necessary that some actor identifies the

    technical system, i.e. perceives the complementarities of different products andtechnologies, and spreads this notion to other actors. At this stage, the role of networks

    of relationships is to co-ordinate specialisation and division of labour between actors,

    develop routine transactions and distribute the economic surplus. The actors also have

    to attract resources and work to legitimise the activities of the network to the

    environment. Later on, the technical system has to be adapted to other systems through

    complementary investments in other fields. New problems may be identified through a

    process of learning-by-using and standards are developed.

    Another related concept is the competence bloc, developed by Gunnar Eliasson and

    described in, e.g., Eliasson (1997).15 The first requirement of a competence bloc is the

    presence of innovators who integrate different new and old technologies in an

    innovative, i.e. unexpected. Knowledge is transferred by, e.g., movement of people and

    imitation. Second, entrepreneurship is needed to identify and select the commercially

    viable innovations and move them to the market.

    During the innovation process, it is important that venture capitalists recognise and

    finance commercially viable opportunities. It is also crucial that there are incentives for

    12 Defined as both organisations and regimes.13 This concept has been developed based on a study of the emergence of digital image technology in Sweden.14 In the paper referred to, Lundgren uses the concept of technological system for complementary products of differenttechnologies and pieces of knowledge (Lundgren, 1993, pp. 147), which obviously is not the same as the technological systemsconcept of Hughes and Carlsson and Stankiewicz (described above). In fact, Lundgren (19??) in Swedish uses a concept which ismore correctly translated to technical system, a translation that will be used here in order to avoid misunderstandings.

  • 8/3/2019 Functions in Innovation System Approach

    10/19

    10

    actors to involve themselves in the process, i.e. that actors who succeed are rewarded by

    getting reasonable or better returns on their investments. Moreover, the selection

    process demands competent actors who have the capability of recognising and

    evaluating opportunities. Therefore, organisations supplying educated people have to be

    present.

    Finally, since technological development results in reallocation of factors of production

    over markets, there is also an acceptance (social) problem to be solved. It is therefore

    important to have institutions making the necessary social adjustments (e.g. in social

    insurance and labour market policy), so that the society is prepared to accept the change

    accompanying economic growth.

    The last related concept described here is the sociotechnical system concept used by

    Bijker (1995).16 According to Bijker, the process of innovation consists of variation and

    selection in three layers. First, relevant social groups identify a variety of problems,

    based on perceived functional failures or presumptive anomalies17, of which some are

    selected for further attention. Second, a variety of solutions are generated and some of

    these solutions yield new artefacts. Third, one of the artefacts becomes dominant overall relevant social groups, partly as a result of one relevant social groups ability to

    convince the other groups about the superiority of its problem definition and resulting

    artefact.

    Different mechanisms might provide the necessary input to innovation in different

    situations. In the case of the social construction of Bakelite, Bijker describes how

    research and development activities were guided in a specific direction by monetaryincentives and how learning-by-doing and learning-by-using gave important inputs to

    15 According to Eliasson (1997), the main differences between this concept and that of Carlsson and Stankiewicz (1995) are that it isnot input determined but defined in terms of its end resultsfunctionally related products in the market and that it is concernedwith the selection of successful innovations. These differences are, however, not easily noticed in the work of the different authors.16 Bijker stresses the need to view every artefact as a sociotechnical ensemble, since it is impossible to separate the technical fromthe social, economical or political. The resulting notion of a sociotechnical system, consisting of different artefacts and actors,strongly resembles the technological system concepts described above. However, in contrast to the other authors Bijker makes astrong point of not assuming a fixed artefact after the invention artefacts are gradually constructed in the social interactionsbetween and within so called relevant social groups and the point of interest is the process by which artefacts attain or fail to attaina stable interpretation.17 Assumptions that under some future conditions the system will fail or function badly or that a radically different system will do amuch better job.

  • 8/3/2019 Functions in Innovation System Approach

    11/19

    11

    the innovation process. Moreover, the patent system provided both efficient conflict-

    solving and supplied incentives for innovation.

    2.4The Network Approach and the Development Block ApproachWhereas the authors described above are concerned with analysing entire innovation

    systems, other authors have chosen other units of analysis. Since these approaches are

    focused on smaller parts of the total system, the functions identified are fewer (but

    maybe better understood).

    Hkansson (1990) focuses on the relationships between actors, especially on the

    importance and functions of relationships with different counterparts in the development

    process. He identifies three reasons why relationships ought to be important: (1)

    interactions with companies that have knowledge in other areas can generate technical

    questions and new knowledge to solve them, (2) the evaluation and acceptance of a new

    technology or product is dependent on support from several actors and (3) companies

    often have to supplement their resources with those of others.

    The unit of analysis of Dahmn (1987) is the development block which refers to asequence of complementarities which by way of a series of structural tensions, i.e.

    disequilibria, may result in a balanced situation. Economic success in a development

    process might require the realisation of one or more specific complementary stage, and

    in situations where stages are missing, the development potential will be released as

    soon as the missing stages have come in place.

    The primary function of actors is to identify the development block in advance and tofill the gaps within it by active search, both for new technical solutions and for actors

    that can invest in complementary stages, for example market identification or creation.

    3.Comparison and Identification of Basic FunctionsIt is already evident that the different approaches are similar with respect to the

    functions they identify, but the questions of how similar and in what ways still remain.

  • 8/3/2019 Functions in Innovation System Approach

    12/19

    12

    In this section, the approaches will be compared more systematically. The comparison

    will be built around some basic functions that most authors seem to agree on.

    One group of functions, described by all authors, consists of the functions directly

    related to the innovation process. The first of these is the function identify problem

    (Bijker, 1995; Dahmn, 1987; Hughes, 1983 and 1990; Hkansson, 1990; Lundgren,

    1993)18. Almost all of these authors recognise the importance of identifying bottlenecks

    in the system, or in the words of the authors themselves functional failures (Bijker,

    1995), missing complementary stages (Dahmn, 1987), reverse salients (Hughes,

    1983 and 1990), imbalances or bottlenecks (Lundgren, 1993) and bottlenecks in

    technology (Lundvall, 1992). Most considered only bottlenecks of a technical nature,but they might according to Dahmn (1987) and Lundgren (1993) also appear in other

    parts of the innovation system (e.g. in the relationships between actors).

    For most authors, the next step in the innovation process is to develop a solution to the

    identified problem,19 often a new technology or product, i.e. to create new knowledge.20

    Although some of the authors (Edquist and Johnson, 1997; Eliasson, 1997; Hughes,

    19XX; Porter, 1990), recognise the possibility to create new knowledge by combiningold and new knowledge in an innovative way, most of them speak primarily of the

    production of entirely new knowledge. They also very much agree on the possible

    sources of new knowledge: R&D (Bijker, 1995; Edquist and Johnson, 1997; Hughes,

    19XX; Lundgren, 1993; Nelson, 1992; Porter, 1990), search and experimentation

    (Carlsson and Stankiewicz, 1995; Lundvall, 1993),21 learning in connection to everyday

    activities, i.e. learning-by-doing and learning-by-using (Bijker, 1995; Edquist and

    Johnson, 1997; Hughes, 19XX; Lundgren, 1993; Lundvall, 1992; Nelson, 1992) andimitation22 (Edquist and Johnson, 1997; Eliasson, 1997; Nelson, 1992).

    18 In this section, references indicate that the identified function is mentioned by the authors in question.19 It should be noted that although the authors describe the innovation process in stages, several of them emphasise the feedbackloops in the process. It is, thus, possible for sub-functions creating new knowledge to contribute to the problem identificationfunction and vice versa.20 According to Bijker and Hughes, however, there is an intermediary function: to develop a solution idea. They both argue that thesearch for solutions is not unconstrained, but is determined by the actors perceptions of viable solutions.21 Search and experimentation is of course a much more general concept than the others, but in the papers it is used much in thesame way as R&D, i.e. as a deliberate way of producing new knowledge.22 Imitation may create knowledge that is new to an individual actor although it might not be new to the innovation system as a

    whole. Moreover, even though the authors do not mention the possibility, imitation might give important inputs to the creation ofknowledge that is new to the system as well.

  • 8/3/2019 Functions in Innovation System Approach

    13/19

    13

    In addition to the functions directly concerned with the innovation process, the authors

    identify several functions that support the innovation process indirectly. Since the

    support might vary between innovations, these functions might also work to promote

    specific innovations.

    The first support function is to supply incentives for companies to engage in innovative

    work (Porter, 1990; Edquist and Johnson, 1997; Nelson, 1992; Eliasson, 1997, Bijker,

    1995); companies have to feel that they get reasonable returns on their investments in

    R&D etc.

    The second support function is to supply resources. The most mentioned resources areindubitably funding (Porter, 1990; Nelson, 1992, Hughes, 1990, Carlsson and

    Stankiewicz, 1995, Eliasson, 1997, Dahmn, 1987) and competence (Porter, 1990;

    Lundvall, 1990; Nelson, 1992; Hughes, 1983; Carlsson and Stankiewicz, 1995;

    Eliasson, 1997). However, some authors also leave the door open for other resources to

    be considered (Lundvall, 1990; Nelson, 1992; Hughes, 1983; Carlsson and Stankiewicz,

    1995) or do not specify what resources they have in mind (Edquist and Johnson, 1997;

    Hkansson, 1990; Lundgren, 1993).

    The third support function is to guide the direction of search, i.e. influence the direction

    in which actors deploy their resources. Several authors (Porter, 1990; Lundvall, 1992;

    Nelson, 1992; Hughes, 1983; Lundgren, 1993; Bijker, 1995) mention this function,

    although in different ways. Some authors (Bijker, 1995; Hughes, 1983; Lundgren, 1993)

    recognise the guidance inherent in problem identification and identification of solution

    idea. Other authors (Porter, 1990, Lundvall, 1992; Nelson, 1992) perceive the guidanceof innovative activity as a function not necessarily served by the companies themselves.

    For example, Lundvall (1992) and Porter (1990) mention the role of standards and

    regulations. Both groups of authors seem to refer primarily to guidance in a technical

    sense, i.e. in terms of choice of product design or specific technology within a product

    or technology area, but the possibility to also include guidance towards new

    technological areas or perhaps different markets should not be neglected. These latter

  • 8/3/2019 Functions in Innovation System Approach

    14/19

    14

    types of guidance are very much related to the function provide incentives for //

    innovative work and to the fourth support function.

    The fourth, support function is to recognise the potential for growth of the innovation,

    which is necessary for it to attract resources and be brought to the market. This may

    come in terms of identifying technological possibility (Carlsson and Stankiewicz, 1995),

    commercial viability (Eliasson, 1997) and/or complementary resources (Lundgren,

    1993; Dahmn, 1987).

    The fifth support function is to facilitate the exchange of information and knowledge

    (Carlsson and Stankiewicz, 1995; Edquist and Johnson, 1997; Nelson, 1992; Lundgren,1993).23 This may be important both for providing feedback between system

    performance and goals (Hughes, 1990) and for diffusion of technology and products on

    the market. This function is related to co-ordination of different departments within

    companies (Edquist and Johnson, 1997; Nelson, 1992, Carlsson and Stankiewicz, 1995),

    to promotion of co-operation between actors (Edquist and Johnson, 1997; Nelson, 1992,

    Bijker, 1995; Hkansson, 1990) and to division of labour between actors once co-

    operation is established (Lundgren, 1993).

    The sixth support function is to stimulate/create markets (Porter, 1990; Nelson, 1992;

    Carlsson and Stankiewicz, 1995; Dahmn, 1987) since markets do not necessarily

    develop spontaneously.24 As already mentioned, diffusion is also connected to the

    facilitation of information exchange and the transfer of knowledge/technology.

    The seventh support function is to reduce social uncertainty, i.e. uncertainty about howothers will act and react (Edquist and Johnson, 1997; Hughes, 1990; Carlsson and

    Stankiewicz, 1995). Related to this function is also the function to prevent or solve

    conflicts between companies or individuals (Edquist and Johnson, 1997; Carlsson and

    Stankiewicz, 1995; Bijker, 1995), since conflicts often arise when actors do not

    understand each other. According to Edquist and Johnson (1997) social uncertainty can

    23 Some of these authors use the word transfer instead of exchange and/or technology instead of knowledge.24 In this context, it is interesting to note that few authors seem to consider market introduction and diffusion as direct parts of theinnovation process; these steps are only dealt with in terms of supporting functions such as this one.

  • 8/3/2019 Functions in Innovation System Approach

    15/19

    15

    be reduced either by providing information about the behaviour of other people or by

    reducing the amount of information needed (e.g. through patent laws). The latter

    function is similar to one mentioned by Carlsson and Stankiewicz (1995), i.e. the

    promotion of stable patterns of interaction and transactions, the purpose of which can be

    said to reduce the need for information.

    The eighth and final support function is to counteract the resistance to change that may

    arise in society when an innovation is introduced or, in other words, to provide

    legitimacy forthe innovation (and the activities of the system surrounding it) in the eyes

    of the outside world (Lundgren, 1993). Resistance may arise because of conflicts

    between new and old products (Edquist and Johnson, 1997) or because of the effects thenew product may have on employment etc (Eliasson, 1997). Some possible means to

    counteract the resistance are to stimulate the enthusiasm for the new technology

    (Hughes, 1983), to clear political and legislative ground in favour of the new technology

    (Hughes, 1983; Eliasson, 199725). Moreover, stimulating relationships between actors

    may be needed in order to create support for a new technology or product (Hkansson,

    1990; Bijker, 1995)

    Obviously, all these functions are very much related to each other. In fact, much of the

    system dynamics is created by the interaction of functions (see, e.g., Johnson and

    Jacobsson (2000)). For example, the interaction makes it possible for cumulative and

    circular causation to appear. A system in which one function is not served (or served in

    a way which is not good for the system) may, therefore, be expected to malfunction also

    in other ways. On the other hand, virtuous circles may appear as well if the functions

    strengthen each other through feedback loops.

    To conclude, there seems to be quite a widely spread correspondence between different

    innovation system approaches with respect to the functions they identify. The basic

    functions described above cover the ones identified in the different approaches to a very

    large extent and most functions are mentioned by a majority of the authors. Thus, the

    first purpose of the paper has been fulfilled. In the next section, the other purpose of the

    25 Eliasson (1997) focus on the possibility to make adjustments in, for example, the social welfare system and labour marketregulations, so that society is prepared to accept new technology even though it might entail unemployment.

  • 8/3/2019 Functions in Innovation System Approach

    16/19

    16

    paper, to discuss the possible benefits of using the concept of function in innovation

    system studies, will be covered.

    4.Benefits of the Concept of Function in Innovation System StudiesEven though it apparently is possible to identify basic functions that, according to many

    authors in the field, are (or should be) served in an innovation system, the benefits of

    doing so may be less obvious at a first glance. The concept of function may, however,

    contribute to innovation system studies in a number of ways.

    First, it provides a tool for setting system borders, which is a problem in many existing

    innovation system approaches. The innovation system would then include all

    components that influence one or more of the identified functions for the object of study

    (e.g. a product or technology).26 This means that the borders are not set a priori to

    nation, region or technology and that different levels of analysis may be combined.

    This type of definition could be especially useful in cases such as wind power

    technology, which is influenced by factors on the technological level (e.g. technology-

    specific knowledge and subsidies), the sectoral level (e.g. energy sector incumbents,substitutes and energy taxes) andthe national level (e.g. norms and values in society).

    Second, the concept of function may be used as a tool to describe the present state of a

    system. The mechanisms that in a particular situation induce or block the functions may

    be identified and possibly stimulated and removed respectively (by policy and/or

    strategy).

    Third, it may be useful when studying innovation system dynamics. Mapping the

    functional pattern, i.e. how functions have been served, over time gives an easily

    grasped picture of the way in which the system has emerged. Thereby, the concept may

    provide some structure to a process which is often difficult to describe and may, thus,

    possibly contribute to the understanding of how innovation systems emerge and change.

    26 This type of extended technological system was used in a study of the Swedish innovation system for renewable energy

    technology (Johnson and Jacobsson, 1999) and in a study of the German, Dutch and Swedish wind turbine industries (Johnson andJacobsson, 2000).

  • 8/3/2019 Functions in Innovation System Approach

    17/19

    17

    Fourth, the concept of function allows us to assess the performance of an innovation

    system, for example in terms of how it has supported the development of a new

    industry. This may be done by analysing the functionality of the system, i.e.how well

    the functions have been served, which, of course, demands a definition of what well

    served means in the particular case of interest.27

    Finally, by focussing on functions actors may be uncoupled from what happens in an

    innovation system. This may be useful in comparative studies since it reduces the risk of

    comparing system structure instead of system functionality; two systems may function

    equally well even though their structure is totally different (i.e. functions may be servedin many different ways). Of course, the purpose of a study may be to compare structure

    instead of, or in combination with, functionality. In such cases, the concept of function

    may still be of use, for example in an analysis of the relationship between how and how

    well functions are (or have been) served.

    5.Concluding remarksThe existing innovation system approaches seem to have a shared understanding of anumber of basic functions that are (or should be) served in innovation systems. The

    differences between approaches thus seem to be less profound than the confusion in

    terms of levels of analysis and concepts used indicates.

    There also seem to be a number of benefits of using the concept of functions in

    innovation system studies. So far, the concept has, however, only been used in a couple

    of very similar studies. The usefulness for other types of studies is, thus, still to bedemonstrated.

    27 This type of analysis may be found in Johnson and Jacobsson (2000), where the difference in success between the Swedish,

    German and Dutch wind turbine industries is argued to be due mostly to differences in how the functions were served in the threeinnovation systems in different industry life-cycle phases.

  • 8/3/2019 Functions in Innovation System Approach

    18/19

    18

    6.ReferencesAlmond, G. A. and Powell, G. B. (1967): Comparative Politics: A Developmental

    Approach. Little, Brown and Company, Boston.

    Bijker, W. E. (1995): Of Bicycles, Bakelites, and Bulbs: Toward a Theory ofSociotechnical Change. The MIT Press. Cambridge, Massachusetts.

    Carlsson, B. and Stankiewicz, R. (1995): On the nature, function and composition oftechnological systems. In: Carlsson, B. (ed.): Technological Systems and EconomicPerformance: The Case of Factory Automation. Kluwer Academic Publishers.Dordrecht, pp. 21-56.

    Cuff, E. C., Sharrock, W. W. and Francis, D. W. (1998): Perspectives in Sociology (4th

    ed.). Routledge, London.

    Dahmn, E. (1987): Development Blocks in Industrial Economics. ScandinavianEconomic History Review.

    Edquist, C. and Johnson, B. (1997): Institutions and Organizations in Systems ofInnovation. In: Edquist, C. (ed): Systems of Innovation: Technologies, Institutions andOrganizations. Pinter Publishers. London, pp. 41-63.

    Eliasson, G. (1997): Competence Blocs and Industrial Policy in the Knowledge BasedEconomy. TRITA-IEO R 1997-04. Department of Industrial Economics and

    Management. Stockholm.

    Galli, R. and Teubal, M. (1997): Paradigmatic Shifts in National Innovation Systems.In: Edquist, C. (ed): Systems of Innovation: Technologies, Institutions andOrganizations. Pinter Publishers. London, pp. 342-370.

    Hughes, T. P. (1983): Networks of Power: Electrification in Western Society, 1880-1930. The Johns Hopkins University Press. Baltimore.

    Hughes, T. P. (1990): The Evolution of Large Technological Systems. In: Bijker, W.E. (ed.); Hughes, T. P. (ed.) & Pinch, T. J. (ed.): The Social construction of

    Technological Systems: New Directions in the Sociology and History of Technology.The MIT Press. Cambridge, Massachusetts.

    Hkansson, H. (1990): Technological Collaboration in Industrial Networks. EMJ,vol.3, pp. 371-379.

    Johnson, A. and Jacobsson, S. (1999): Inducement and Blocking Mechanisms in theDevelopment of a New Industry. To appear in: Coombs, R., Green, K., Walsh, V. andRichards, A. (eds): Technology and the Market: Demand, Users and Innovation.Edward Elgar. Cheltenham and Northhampton, Massachusetts.

  • 8/3/2019 Functions in Innovation System Approach

    19/19

    Johnson, A. and Jacobsson, S. (2000): The Emergence of a Growth Industry: AComparative Analysis of the German, Dutch and Swedish Wind Turbine Industries.Mimeo. The department of Industrial Dynamics, Chalmers University of Technology,Sweden. (To be published in the Schumpeter Conference 2000 proceedings.)

    McKelvey, M. (1997): Using Evolutionary Theory to Define Systems of Innovation.In: Edquist, C. (ed): Systems of Innovation: Technologies, Institutions andOrganizations. Pinter Publishers. London, pp. 200-222.

    Lundgren, A. (19XX): Ett ntverks frdelse Utvecklingen av digital bildteknik iSverige. In: Hkansson, H., Laage-Hellman, J., Lundgren, A. and Waluszewski, A.:Teknikutveckling i fretaget: Ett ntverksperspektiv. Studentlitteratur. (In Swedish).

    Lundgren, A. (1993): Technological innovation and the emergence and evolution ofindustrial networks: the case of digital image technology in Sweden. Advances in

    International Marketing, vol. 5, pp. 145-170.Lundvall, B.-. (1992): Introduction. In: Lundvall, B.-. (ed.): National Systems of

    Innovation toward a Theory of Innovation and Interactive Learning. PinterPublishers. London, pp. 1-19.

    Nelson, R. R. (1992): National Innovation Systems: A Retrospective on a Study.Industrial and Corporate Change, no. 2, pp.347-374.

    Porter, M. (1990): The Competitive Advantage of Nations. Harvard Business Review.March-April.

    Wenk, E. and Kuehn, T. J. (1977): Interinstitutional Networks in TechnologicalDelivery Systems. In: Haberer, J. (ed): Science and Technology Policy. LexingtonBooks, Lexington, Massachusetts, pp. 153-175.