This article was downloaded by: [University of Saskatchewan Library]On: 25 August 2012, At: 21:53Publisher: RoutledgeInforma Ltd Registered in England and Wales Registered Number: 1072954 Registeredoffice: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK

Economics of Innovation and NewTechnologyPublication details, including instructions for authors andsubscription information:http://www.tandfonline.com/loi/gein20

Sectoral systems of innovation: aframework for linking innovation tothe knowledge base, structure anddynamics of sectorsFranco Malerbaa CESPRI, Università ‘L. Bocconi’, Via Sarfatti 25, 20136, Milano,Italy

Version of record first published: 25 Jan 2007

To cite this article: Franco Malerba (2005): Sectoral systems of innovation: a framework for linkinginnovation to the knowledge base, structure and dynamics of sectors, Economics of Innovation andNew Technology, 14:1-2, 63-82

To link to this article: http://dx.doi.org/10.1080/1043859042000228688

PLEASE SCROLL DOWN FOR ARTICLE

Full terms and conditions of use: http://www.tandfonline.com/page/terms-and-conditions

This article may be used for research, teaching, and private study purposes. Anysubstantial or systematic reproduction, redistribution, reselling, loan, sub-licensing,systematic supply, or distribution in any form to anyone is expressly forbidden.

The publisher does not give any warranty express or implied or make any representationthat the contents will be complete or accurate or up to date. The accuracy of anyinstructions, formulae, and drug doses should be independently verified with primarysources. The publisher shall not be liable for any loss, actions, claims, proceedings,demand, or costs or damages whatsoever or howsoever caused arising directly orindirectly in connection with or arising out of the use of this material.

Econ. Innov. New Techn., 2005, Vol. 14(1–2), January–March, pp. 63–82

SECTORAL SYSTEMS OF INNOVATION:A FRAMEWORK FOR LINKING INNOVATION

TO THE KNOWLEDGE BASE, STRUCTUREAND DYNAMICS OF SECTORS

FRANCO MALERBA∗

CESPRI, Universita ‘L. Bocconi’, Via Sarfatti 25, 20136 Milano, Italy

(Received 1 August 2002; Revised 5 September 2003; In final form 19 January 2004)

This paper proposes a framework for examining factors that affect innovation in sectors: sectoral systems. Sectoralsystems are based on three building blocks: knowledge and technologies, actors and networks, and institutions. Inthe first part of this paper, the concept and the definition of a sectoral systems of innovation are presented. In thesecond part of the paper, the role of knowledge, actors and networks, and institutions in five major sectoral systems isexamined. Then the main focus moves to the analysis of the dynamics and transformation of sectoral systems. Finally,some general conclusions and directions for future research end the paper.

Keywords: Innovation; Sectors; Networks; Institutions

JEL Codes: 030; L10; L60

1 INTRODUCTION

The rate and type of innovation and the organization of innovative activities greatly differacross sectors. Various streams of research have tried to examine patterns and determinants.For the aim of the present paper one could identify three different approaches.

One is the old tradition originally related to Schumpeterian themes. This tradition framedthe issue in terms of what has been termed the ‘market structure and innovation’ approach(Kamien and Schwartz, 1982). Here, the focus was on testing the relationship between therate of innovation and firm size, on the one hand, and monopoly power, on the other. It is nowwidely acknowledged that the early results obtained within this framework suffered of, at least,two main limitations. First, they failed to recognise the mutual causation between innovation,market structure and firm size. Rather these variables are best thought as endogenously code-termined (Dasgupta and Stiglitz, 1980; Nelson and Winter, 1982). Second, starting from theempirical observation that the relevant relationships varied significantly across industries, it wassuggested that other factors, mainly linked to the nature of technology, are important explana-tory variables of the sectoral patterns of innovation. Thus, even the insertion of very roughproxies of opportunity and appropriability conditions significantly improves the performance

∗ Tel.: +39-025836-3397; E-mail: franco.malerba@uni-bocconi.it

ISSN 1043-8599 print; ISSN 1476-8364 online c© 2005 Taylor & Francis LtdDOI: 10.1080/1043859042000228688

Dow

nloa

ded

by [

Uni

vers

ity o

f Sa

skat

chew

an L

ibra

ry]

at 2

1:53

25

Aug

ust 2

012

64 F. MALERBA

of econometric exercises performed in an otherwise conventional approach (Levin et al., 1985;Cohen and Levin, 1989).

A second and more recent tradition takes into account the learning conditions and tech-nological context in which innovation takes place in sectors. The notion of technologicalregime dates back to Nelson and Winter (1977, 1982) who have suggested that the dynamicsof innovation and market structure is driven by processes of market selection and by the natureof technology, which differ greatly across sectors. Technological regimes set the boundariesof what can be achieved in firms’ problem solving activities and identify also the ‘naturaltrajectories’ along which solutions to these problems can be found.1 After Nelson and Winter,various authors – Gort and Klepper (1982), Levin et al., (1985), Cohen and Levin (1989)and Audretsch (1995) among others – have pointed out that, more than firm size or demand,opportunity and appropriability conditions appear as the most relevant factors affecting thedynamics of market structure and innovation. The notion of technological regime providesa synthetic way of representing some of the most important economic properties of tech-nologies and of the characteristics of the learning processes that are involved in innovativeactivities. Thus, it identifies some fundamental structural conditions that contribute to definecompetencies, incentives and dynamic properties of the innovative process. In this line ofresearch, Malerba and Orsenigo (1990, 1993) have proposed that a technological regime isa particular combination of some fundamental properties of technologies: opportunity andappropriability conditions; degrees of cumulativeness of technological knowledge and char-acteristics of the relevant knowledge base. See also Breschi et al. (2000) for an empirical testof these propositions.2

The third line of research has focussed on the sources of innovation and the mechanismsof appropriability, which differ across sectors. Here, the main references are the work byRosenberg (1976, 1982) on the various sources of technological change across a wide varietyof sectors, Levin et al. (1987) on appropriability conditions, Nelson (1993) on universities,Mowery and Nelson (1999) on various industries, Pavitt (1984) on sectoral taxonomies of thesources of innovation and the appropriability mechanisms.3

This paper proposes a different but complementary framework that looks at the rate and typeof innovation and at the organization of innovative activities in sectors: the sectoral systemof innovation approach. This framework has been inspired by evolutionary theory and theinnovation system approach.

Evolutionary theory places dynamics, process and transformation at the centre of theanalysis. Learning and knowledge are key elements in the change of the economic system.‘Boundedly rational’ agents act, learn and search in uncertain and changing environments.Relatedly, competencies correspond to specific ways of packaging knowledge about differentthings and have an intrinsic organisational content. Different agents know how to do differentthings in different ways. Evolutionary theory has placed emphasis on cognitive aspects suchas beliefs, objectives and expectations, in turn affected by previous learning and experience

1 Nelson and Winter (1982) and Winter (1984) identify two different basic technological regimes according tothe relevant knowledge base: an entrepreneurial regime in which the knowledge base is related to science and isnon-cumulative and universal (thus facilitating the entry of new firms), and a routinized regime in which knowledgeis more cumulative and internal to the industry (thus facilitating the innovation by established firms).

2 The notion of technological regime holds some relationship with the concepts of technological paradigms andtrajectories, which capture the idea that technologies differ drastically and that their development retains a strongautonomous internal logic (Dosi, 1982, 1988). Also the ‘bounds’ approach by Sutton (1998) bears some link to thenotion of different learning contexts characterizing the various sectors, because it claims that the relationship betweenmarket structure and innovation is constrained by the specificity of the technology in terms of the diversity of possibletechnological trajectories available to firms and the productivity of R&D investments along each trajectory.

3 Pavitt taxonomy has been tremendously successful in empirical research and has guided also the identificationof firms’ and countries’ advantages in innovation. Refinements and enrichments of the taxonomy have been proposedin the succeeding decades. A very interesting work in this direction is that of Marsili (2001).

Dow

nloa

ded

by [

Uni

vers

ity o

f Sa

skat

chew

an L

ibra

ry]

at 2

1:53

25

Aug

ust 2

012

SECTORAL SYSTEMS OF INNOVATION 65

and by the environment in which agents act (Nelson, 1995; Dosi, 1997; Metcalfe, 1998). Thuslearning, knowledge and behaviour entail agents’ heterogeneity in experience, competenciesand organisation, and their persistent differential performance. A central place in an evolution-ary approach is occupied by three economic processes driving economic change: processesof variety creation in technologies, products, firms and organisations; processes of replica-tion and processes of selection (Nelson, 1995; Metcalfe, 1998). In evolutionary theory, theenvironment and the conditions in which agents operate may differ drastically. Evolutionarytheory stresses major differences in opportunity conditions related to science and technologies.The same holds for the knowledge base underpinning innovative activities, as well as for theinstitutional context. Thus the learning, behaviour and capabilities of agents are constrainedand bounded by the technology, knowledge base and institutional context in which firms act.Heterogeneous firms facing similar technologies, searching around similar knowledge bases,undertaking similar production activities and embedded in the same institutional setting, sharesome common behavioural and organisational traits and develop a similar range of learningpatterns, behaviour and organisational forms (Nelson and Winter, 1982).

The other link of the sectoral system of innovation framework is with the innovation systemliterature, in which relationships and networks are key elements of the innovative and pro-duction processes (Edquist, 1997). The innovation system approach considers innovation as aninteractive process among a wide variety of actors. It stresses the point that firms do not innovatein isolation: innovation is seen as a collective process. In the innovative process firms interactwith other firms as well as with non-firm organizations (such as universities, research centres,government agencies, financial institutions and so on). Their action is shaped by institutions(Lundvall, 1993; Carlsson, 1995; Edquist, 1997). This approach places a great deal of emphasison an interdisciplinary approach, emphasises a historical perspective and puts learning as akey determinant of innovation (Edquist, 1997). In particular, the notion of sectoral systems ofinnovation complements other concepts such as national systems of innovation (more focussedon national boundaries and on non-firm organisations and institutions – Freeman, 1987;Nelson, 1993; Lundvall, 1993), regional/local innovation systems (more focussed on theregion – Cooke et al., 1997) and technological systems (focussed on specific technologiesand not on sectors) (Hughes, 1984; Callon, 1992; Carlsson and Stankiewitz, 1995).

In the first part of this paper, the concept and definition of a sectoral system of innovationare presented (Sec. 2) and some general and methodological issues are discussed (Sec. 3). Inthe second part of the paper, the role of knowledge, actors and networks and institutions in fivemajor sectoral systems is examined (Sec. 4). Then the main focus moves to the analysis of thedynamics and transformation of sectoral systems (Sec. 5). Finally, some general conclusionsand directions for future research end the paper (Sec. 6).

2 DEFINITION AND CONCEPTS

A sector is a set of activities which are unified by some related product groups for a given oremerging demand and which share some basic knowledge. In a sector, firms have common-alities and at the same time are heterogeneous. The key point is that innovation has relevantsystemic features so that it is possible to advance the following definition. Sectoral systems ofinnovation have a knowledge base, technologies, inputs and a (potential or existing) demand.They are composed of a set of agents carrying out market and non-market interactions for thecreation, development and diffusion of new sectoral products. These agents are individuals andorganisations at various levels of aggregation, with specific learning processes, competencies,

Dow

nloa

ded

by [

Uni

vers

ity o

f Sa

skat

chew

an L

ibra

ry]

at 2

1:53

25

Aug

ust 2

012

66 F. MALERBA

organisational structure, beliefs, goals and behaviours. They interact through processes of com-munication, exchange, cooperation, competition and command. Their interaction is shaped byinstitutions. A sectoral system undergoes processes of change and transformation through thecoevolution of its various elements.

Thus a sectoral system could be seen as composed by three main building blocks:

– knowledge and technology– actors and networks– institutions

2.1 Knowledge and Technologies

Any sector could be characterised by a specific knowledge base, technologies and inputs. In adynamic way, the focus on knowledge and technology places the issue of sectoral boundaries atthe centre of analysis. In sectors in which innovation is quite rapid, sectoral boundaries are notfixed, but change over time. Knowledge and basic technologies constitute major constraints onthe full range of diversity in the behaviour and organisation of firms. Links and complementar-ities among artefacts and activities also play a major role in defining the real boundaries of asectoral system. These links and complementarities could be static (as input–output links are)or dynamic. Dynamic complementarities take into account interdependencies and feedbacks(both at the demand and at the production levels), are major sources of transformation andgrowth of sectoral systems, and may set in motion virtuous cycles of innovation and change.

2.2 Actors and Networks

A sector is composed of heterogeneous agents that are organisations and individuals (e.g.consumers, entrepreneurs, scientists). Organisations may be firms (e.g. users, producers andinput suppliers) and non-firm organisations (e.g. universities, financial institutions, governmentagencies, trade-unions, or technical associations), including sub-units of larger organisations(e.g. R&D or production departments) and groups of organisations (e.g. industry associa-tions). Agents are characterised by specific learning processes, competencies, beliefs, goals,organisational structures and behaviours. They interact through processes of communication,exchange, cooperation, competition and command. Within sectoral systems, heterogeneousagents are connected in various ways through market and non-market relationships. The typesand structures of relationships and networks differ from sectoral system to sectoral system, asa consequence of the features of the knowledge base, the relevant learning processes, the basictechnologies, the characteristics of demand, the key links and the dynamic complementari-ties. Thus in a sectoral system perspective, innovation is considered a process which involvessystematic interactions among a wide variety of actors for the generation and exchange ofknowledge relevant to innovation and its commercialisation. Interactions include market andnon-market relations that are broader than the market for technological licensing and knowl-edge, interfirm alliances, and formal networks of firms. Often their outcome is not adequatelycaptured by our existing ways of measuring economic output.

2.3 Institutions

Agents’ cognition, actions and interactions are shaped by institutions, which include norms,routines, common habits, established practices, rules, laws, standards and so on. They mayrange from the ones that bind or impose enforcements on agents to the ones that are createdby the interaction among agents (such as contracts); from more binding to less binding; from

Dow

nloa

ded

by [

Uni

vers

ity o

f Sa

skat

chew

an L

ibra

ry]

at 2

1:53

25

Aug

ust 2

012

SECTORAL SYSTEMS OF INNOVATION 67

formal to informal (such as patent laws or specific regulations vs. traditions and conventions).A lot of institutions are national (such as the patent system), while others may be specific tosectoral systems, such as sectoral labour markets or sector-specific financial institutions. Inthis respect, the relationships between national institutions and sectoral systems become quiteimportant in several respects. First, national institutions such as the patent system, propertyrights or antitrust regulations have different effects on innovation in the different sectors.Second, the same institution may take on different features in different countries, and thus mayaffect innovation differently. Third, often the characteristics of national institutions favoursectors that fit their specificities better. In other cases, national institutions may constrain thedevelopment of innovation in specific sectors and mismatches between national institutions andagents and sectoral ones may take place. Fourth, the relationship between national institutionsand sectoral systems may sometimes go from the sector to the national level: the institutionsof a sector, extremely important for a country in terms of employment, competitiveness orstrategic relevance, may end up emerging as national (thus also becoming relevant for othersectors). But in the process of becoming national, they may change some of their originaldistinctive features.

Demand is a key part of a sectoral system. The above mentioned focus on users and oninstitutions puts a different emphasis on the role of demand. Demand is made up of individualconsumers, firms and public agencies, each characterised by knowledge, learning processes,competencies and goals, and affected by social factors and institutions. Thus, in a sectoralsystem demand is not seen as an aggregate set of similar buyers, but as composed of heteroge-neous agents whose interactions with producers are shaped by institutions. The emergence andtransformation of demand play a major role in the dynamics and evolution of sectoral systems.

In general, the sectoral system of innovation framework highlights five key points. First, itfocuses on supply as well as demand and on markets in the innovation process. Second,it examines other types of agents in addition to firms. Third, it places considerable emphasison non-market as well as market interactions. Fourth, it pays attention to institutions. Fifth,it does not consider sectoral boundaries as given and static, but it focuses on the process oftransformation of the system.

3 SOME IMPLICATIONS OF THE SECTORAL SYSTEM OFINNOVATION FRAMEWORK

The discussion in Section 2 has some relevant implications. First, a sectoral system frameworkgives a specific meaning to the concept of the ‘structure’ of a sector. Structure does not simplymean industrial concentration, vertical integration or diversification (as it has been in mostcontributions in industrial economics dealing with innovation). Rather structure relates to thelinks and relationships among agents, knowledge, products and technologies. For example, asfar as agents are concerned, within sectoral systems heterogeneous agents are connected invarious ways through market and non-market relationships. Some of these connections are cap-tured by traditional analyses of industrial organisations which have examined agents involvedin processes of exchange, competition and command (such as vertical integration). Formalcooperation, or informal interactions among firms, or between firms and non-firm organi-sations, have been examined in depth by the recent literature on tacit or explicit collusion,hybrid governance forms and formal R&D cooperation.4 Also the evolutionary approach and

4 This literature has analysed firms with certain market power, suppliers and users facing opportunistic behaviouror asset specificities in transaction, firms with similar knowledge and with appropriability and indivisibility problemsin the R&D process.

Dow

nloa

ded

by [

Uni

vers

ity o

f Sa

skat

chew

an L

ibra

ry]

at 2

1:53

25

Aug

ust 2

012

68 F. MALERBA

the innovation systems literature have paid considerable attention to the wide range of formaland informal interactions among firms. However, according to the perspectives, in uncertainand changing environments networks emerge not because agents are similar, but becausethey are different. Thus networks allow the access to, and the integration of, complementar-ities in knowledge, capabilities and specialisation (see Teubal, et al., 1991; Lundvall, 1993;Nelson, 1995; Edquist, 1997). Therefore, in a sectoral system of innovation framework theterm structure refers also to ‘networks’. It has to be stressed that market and non-market rela-tionships may involve not just firms, but also non-firm organisations. For example, universitiesand public research centres may be a source of innovation and change in several sectors, such aspharmaceuticals and biotechnology, information technology and telecommunications (Nelsonand Rosenberg, 1993). The types and structures of relationships and networks differ from onesectoral system to another, as a consequence of the features of the knowledge base, the rele-vant learning processes, basic technologies, characteristics of demand, key links and dynamiccomplementarities.

Second, the focus on actors and networks implies also that in a dynamic perspective asectoral system is a collective emergent outcome of the interaction and coevolution of its vari-ous elements. This process involves technology, demand, knowledge base, learning processes,firms, non-firm organisations and institutions. Nelson (1994) and Metcalfe (1998) have dis-cussed these processes at the general level. More broadly, for evolutionary theory aggregatephenomena are emergent properties of far from equilibrium interaction and have a meta-stablenature (Lane, 1993a, b). In a sectoral system perspective, these processes are sector-specific.

Third, the focus on dynamics implies also that the transformation of existing sectoral systemsand the emergence of new sectoral systems become a major part of the analysis. The emergenceof new clusters that span over several sectors (such as internet–software–telecom, biotechnol-ogy–pharmaceutical or new materials) is a particularly interesting theme. Here, transformationmeans the integration and fusion of previously separated knowledge and technologies as wellas new relations and dynamics among different types of users and consumers, firms with differ-ent specialisation and competencies, and non-firm organisations and institutions (all of themgrounded in previously separated sectors).

In terms of methodology, a key issue refers to the level of agents aggregation and to thegeographical and product boundaries of sectoral systems. In terms of level of aggregationagents, the analysis may consider agents at lower or higher levels of aggregation comparedto firms. So one may examine the individual firms’ sub-units, as well as groups of firms andof non-firm organisations. Flexibility has to be used in the choice of the unit of analysis, thevariables to be examined and the level of details in the study to be conducted.

In terms of geographical boundaries, national boundaries are not always the most appropriateones for an examination of the structure, agents and dynamics of sectoral systems. Oftenthe boundaries are local, and the sectoral specialisation defines the specialisation of the wholearea. For example, machinery is concentrated in regional areas, traditional sectors define thespecialisation of industrial districts in Italy, sectoral specialisation and local agglomerationoverlap in Route 128 (for minicomputers) and in Silicon Valley (for personal computers,software and microelectronics) (Saxenian, 1994). More often in a sectoral system, one mayfind the coexistence of local, national and global boundaries: global for knowledge interaction;local for the labour market and national for some key institutions.

In terms of products, sectoral systems may be delimited in different ways, depending on thegoal of the analysis. So, sectors may be defined broadly (as in this paper): pharmaceuticals;chemicals; telecommunications; software and machine tools. This broad definition allows toemphasise interdependencies, linkages and transformations spanning over a wide range ofproducts, actors and functions. However, in some other cases, a more disaggregated levelmay be used. In this case, one may find the coexistence of quite different innovation systems

Dow

nloa

ded

by [

Uni

vers

ity o

f Sa

skat

chew

an L

ibra

ry]

at 2

1:53

25

Aug

ust 2

012

SECTORAL SYSTEMS OF INNOVATION 69

within broadly defined sectors such as telecommunications or pharmaceuticals. In this case,one may use the term ‘sector’ for the broad aggregations mentioned above, and the terms ‘sub-sectors’, ‘product groups’ and ‘product segments’ for more narrowly defined aggregationswithin broad sectors. Again, the aim of the analysis has to drive the delimitation of the sectoralsystem. Sometimes, it is necessary to analyse very broad sectoral systems, such as computerhardware and software. Other times it is not, as in the case of custom software. Particularlywith respect to the emergence of new clusters such as software–internet–telecommunications,new materials, and pharmaceutical–biotechnology, a high level of aggregation is important ifone has to identify interdependencies and linkages. In any case, the goal and the objectives ofthe analysis should dictate the appropriate level of aggregation.

One final remark has to be advanced about the impossibility of identifying an ‘optimal’structure and an ‘optimal’ working for sectoral systems. In reality some coherence amongthe various elements of a sectoral system does occur and it develops over time as a resultof both conscious design and unplanned processes. And mismatches among the various partsand variables of sectoral systems could be identified and eventually eliminated. But the actualcoherence is far from being ‘optimal’. Sectoral systems may develop different features indifferent countries, and at different times. This is so because they emerge and develop incontinuously changing environments, are characterized by path-dependent processes and areembedded in different socio-economic contexts.

In general, sectoral systems may prove a useful tool for descriptive analyses of the innovationprocess in sectors; for the recognition of the factors affecting innovation; for studies of the rela-tionship between innovation and the changing boundaries of sectors; for a full understandingof the short-term and long-term dynamics and transformation of sectors; for the identificationof the factors affecting the international performance of firms and countries in the differentsectors and for the development of new public policy indications.

4 A CHARACTERISATION OF SOME KEY SECTORAL SYSTEMS IN TERMSOF KNOWLEDGE, ACTORS, NETWORKS AND INSTITUTIONS

This paper examines innovation in five major sectors in Europe and in other advanced countries:pharmaceutical and biotechnology, telecommunication equipment and services, chemicals,software and machine tools. These sectors have been chosen because technological change isquite rapid and innovation plays a major role in fostering growth and in affecting the competi-tiveness of firms and countries. In addition, broad sectoral boundaries have been considered, sothat linkages in the structure and interdependencies in the transformation processes spanningover a large set of products, actors and functions can be assessed. However, in the discussionsometimes a more disaggregated level has also been used in order to show that within thebroadly defined sectors different innovation systems may coexist. The countries examined aremainly Europe and US, and whenever possible, some specific European countries has beenmentioned. The following discussion draws from the more extensive results of the project‘Sectoral Systems in Europe–Innovation, Competitiveness and Growth’ (ESSY)5 and fromMalerba (2004).

5 ESSY [Project financed within the TSER Programme – Contract No. SOE1-CT 98-1116] was a 3 year projectconducted by 10 research centres in Europe – CESPRI (Universita’ Bocconi), SPRU (University of Sussex), WZB(Berlin), S.S. Sant’Anna (Pisa), CRIC (Manchester University), CREII (Paris XIII), TEMA (Linkoping University),Pompeu Fabra (Barcelona), ISI (Karlsruhe) and IKE (Alborg University). It was supported by the European Union.I wish to thank the main participants in ESSY: R. O’Brien (the EU officer responsible for the project), B. Coriat, G.Dosi, C. Edquist, S. Metcalfe, D. Soskice , W.E. Steinmueller, B. Dalum, W. Garcia, J. Wengel, F. Montobbio, S.Breschi, M. Harvey, F. Lissoni, M. McKelvey, L. Orsenigo, F. Pammolli, O. Weistein, L. D’Adderio, G. Bottazzi, F.Cesaroni, N. Corrocher, P. Geoffron, L. Hommen, A. James, H. Kettler, M. Riccaboni, D. Rivaud-Danset, B. Tether.

Dow

nloa

ded

by [

Uni

vers

ity o

f Sa

skat

chew

an L

ibra

ry]

at 2

1:53

25

Aug

ust 2

012

70 F. MALERBA

In the following section, each sectoral system has been examined according to the threebroad dimensions discussed above: knowledge and technologies; actors and networks andinstitutions.

4.1 Pharmaceuticals and Biotechnology

In the pharmaceutical sectoral system, the knowledge base and the learning processes havegreatly affected innovation and the organisation of innovative activities. In the early stages(1850–1945), the industry was close to chemicals, with little formal research until the 1930sand a major use of licenses. The following period (1945–early 1980s) was characterised by theintroduction of the random screening of natural and chemically derived compounds. This led toan explosion of R&D. A few blockbusters were discovered in every period: each one had highgrowth. The advent of molecular biology since the 1980s led to a new learning regime basedon molecular genetics and DNA technology, with two research processes: one concerningcospecialised technologies and the other generic technologies. Nowadays, no individual firmcan gain control of more than a subset of the search space. Innovation increasingly depend onstrong scientific capabilities and on the ability to interact with science and scientific institutionsin order to explore the search space (Henderson et al., 1999; McKelvey et al., 2004).

The change in the knowledge base discussed above has led to a different organisation ofinnovative activity within and across firms. Division of labour has taken place between newbiotechnology firms (NBFs) which lacked experience in clinical testing and established compa-nies that (with time) adopted molecular biology. Networks of collaborative relations (facilitatedby the science base and by the abstract and codified nature of knowledge generated by the NBF)emerged in the sector. Further, mergers and acquisitions allowed established firms to obtaincomplementary knowledge for the development of innovative products. As of now, the phar-maceutical–biotechnology sectoral system has a structure of innovative actors which includeslarge firms, NBFs, small firms and individuals (such as scientists or NBF entrepreneurs). Inaddition, a very rich set of non-firm organisations and institutions greatly affect innovation,ranging from universities to public and private research organisations, the financial system andventure capital, the legal system and IPR. Demand channelled through agencies, physiciansand the health system, and institutions such as regulation played a significant role in the dif-fusion of new drugs. Nowadays, no individual firm can hope to gain control of more than asubset of the search space. Even, the innovativeness and competitiveness of the largest phar-maceutical firms depend on strong scientific capabilities and on the ability to interact on theone hand with science and scientific institutions (in order to explore such a complex space) andon the other with specialised innovative firms (in order to develop new products) (McKelveyet al., 2004).

Summing up: in pharmaceuticals and biotechnology, a wide variety of science and engi-neering fields are relevant important roles in renewing the search space. Universities, venturecapital and national health systems play a major role in the innovative process. There are sev-eral relevant actors: large firms; small firms and NBFs. An extensive division of labour throughnetworks is present. NBFs have entered the sector, competing as well as cooperating with (orbeing bought up by) the established large pharmaceutical firms. In this sector, demand andinstitutions (such as regulation, IPR and national health systems) affect the innovation process.

P. Caracostas, D. Mowery, R. Nelson, F. Onida, S. Torrisi, F. Gianfrate, B. Lamborghini, E. Hoffman, V. Maglia andD. Speroni commented the final papers of the ESSY project during the final Conference at Bocconi University. All thesupporting papers of the whole project ‘Sectoral Systems in Europe’– the Working Papers ESSY – can be downloadedfrom the ESSY website at http://www.cespri.it/ricerca/es wp.htm.

Dow

nloa

ded

by [

Uni

vers

ity o

f Sa

skat

chew

an L

ibra

ry]

at 2

1:53

25

Aug

ust 2

012

SECTORAL SYSTEMS OF INNOVATION 71

4.2 Chemicals

In chemicals learning processes based on formal search processes have been present since thebeginning of the history of the industry with the diffusion of the ‘synthetic–dyestuff model’(which introduced a scientific base to innovation), and later on with the development of organicchemistry (related to the understanding of the chemical structure of new molecules and thepossibility of exploiting economies of scope in knowledge for the development of differentorganic products). This has led to the presence of firms with large R&D departments (some ofwhich have been active since the beginning of the industry) and to a greater role for universitiesand other scientific organizations. Changes in knowledge and learning processes have beenaccompanied by the development of new products which were quite different from previousones, and by the emergence of different actors and organisations. For example, the secondmajor change in the industry, polymer chemistry (1920s), based upon the idea that materialsconsist of long chains of molecules – polymers – linked together by chemical bounds led tothe development of materials by design, in which the scientific understanding of chemicalcomposites is the base for different product applications. Polymer chemistry provided a com-mon technological base for developing applications and product differentiation in five distinctmarkets: plastics, fibres, rubbers, surface coatings and adhesives. The other major change inthe industry, the development of chemical engineering and the concept of unit operation (1915)broke down chemical processes into a limited number of basic components, common to manyproduct lines. This development became the general purpose technology of the chemical sector.It allowed the separation of process innovation from product innovation: process innovationbecame a commodity that could be traded. In general, one could claim that these changesled to a transformation of firms’ learning processes away from trial and error procedures to ascience-based approach to industrial research. The advances in chemical disciplines such aspolymer chemistry and chemical engineering have created the base for greater codificabilityof knowledge. At the same time firms’ behaviour has enhanced the transferability of chemicaltechnologies. Separability and transferability made possible the transaction of technology inthe chemical industry and the emergence of new markets for engineering and process designservices for chemical plants. This type of knowledge base has implied that internal R&Dhas been complemented by external links and knowledge. Nowadays in chemicals innova-tion requires the interaction between R&D capabilities and external sources of scientific andtechnological knowledge (Arora et al., 1999; Cesaroni et al., 2004).

In chemicals, the structure of the sectoral system has been centred around large firms, whichhave been the major source of innovation over a long period of time. Large R&D expendi-tures, economies of scale and scope (Chandler, 1990), cumulativeness of technical advanceand commercialisation capabilities have given these firms major innovative and commercialadvantages (Arora et al., 1999). The changes in the knowledge base discussed above haveaffected the types of actors and networks. As mentioned previously, with the diffusion of thesynthetic–dyestuff model, firms scaled up their R&D departments and the role of universi-ties increased. The introduction of polymer chemistry (1920s) affected the structure of theindustry because knowledge about the characteristics of different market segments becameimportant so that firms had to develop extensive linkages with downstream markets. The othermajor change related to the development of chemical engineering, and the concept of unit ofoperation led to an increasing division of labour between chemical companies and technologysuppliers, with the rise of the specialised engineering firms (SEFs), which developed verticallinks with chemical companies. In this period, university research continued to be importantfor the development of innovations, and links between university and industry increased. Inaddition, advances in chemical disciplines and the separability of knowledge increased thetransferability of chemical technologies. Thus, there has been a greater role for licensing also

Dow

nloa

ded

by [

Uni

vers

ity o

f Sa

skat

chew

an L

ibra

ry]

at 2

1:53

25

Aug

ust 2

012

72 F. MALERBA

by large firms, which in turn increased knowledge diffusion. It must be noted that large firmsalso licensed process technology and that SEFs did not develop radically new processes. Rather,they acted as independent licensors of technology on behalf of other firms. The increasingreliance on external links for complementary scientific and technological knowledge has led tothe emergence of networks of three types: interfirms, university–industry and user–producersin specialty segments. However, the relevant networks have changed in relation with the type ofknowledge base. In the synthetic–dyestuff model, firms developed links with universities andwith users. In polymer chemistry and with the diffusion of chemical engineering, networksbetween producers and users, industry–university networks, and vertical networks betweenchemical companies and engineering contractors have been common, with the use of merg-ers and acquisitions to related and unrelated sectors in order to acquire capabilities (Cesaroniet al., 2004). In general, however, the inventive capacity of firms within a country dependsheavily upon the strength of the underlying universities and public research organizations.

Institutions have played a critical role concerning two different situations: the restructuringprocesses and patent policy. Concerning industry’s restructuring processes, in the past (duringand after WWI) national governments allowed or promoted the creations of cartels and nationalgiants. Germany and Great Britain are clear example in this direction.6 Both in Britain andin Germany different trade associations and alliances among firms emerged. In Britain, thechemical industry organised itself into theAssociation of Chemical Manufacturers. In Germany,the eight largest dye producers formed a ‘quasi-cartel.’ Since the 1980s, the chemical industryhas entered a new phase of restructuring, in which public policy has played a role as well. Inthis period, governments have managed the restructuring process to a good extent, especiallyin France and Italy. The second important role of institutions in chemicals is related to patentpolicies, especially relevant to small firms.

Indeed, proper forms of intellectual property rights and sufficiently strong patent protectionsupported the activity of smaller technology-based firms. In turn, this created the bases for adivision of labour between technology suppliers and users, and allowed the development ofmarkets for technology.This pattern was particularly evident in US, where patent protection wasproperly defined. By contrast, European markets for technology are far from being developed.This requires policy support for intellectual property rights.

Summing up: the chemicals sectoral system is characterised by the continuity ininnovativeness by large multinational firms through R&D, economies of scale, scope,cumulativeness of advance and commercialisation capabilities. Firms’ internal R&D has beencomplemented by external links and by the capability of absorbing external scientific andtechnological knowledge.

4.3 Telecommunication Equipment and Services

In telecommunications equipment and services, the knowledge base has been quite diversifiedbecause the sectoral system encompasses fixed communications, mobile phones, internet andother services.All these product groups present different features, but they are related technolo-gies in some way or another. Moreover, this broad sectoral system has been recently affected byprocesses of convergence between information and communication technologies and betweenICT and broadcasting-audio-visual technologies. Until the advent of the internet, the telecom

6 While in Germany the presence of chemical trade associations made it easier to create a link between thegovernment and the individual firms, in Britain the absence of such associations imposed a deeper intervention bystate authorities. The British state reorganised the chemical industry (traditionally independent of the government) tosupply chemicals for war needs. As a consequence of this ‘forced’ co-ordination, the leaders of the largest chemicalfirms came to know one another.

Dow

nloa

ded

by [

Uni

vers

ity o

f Sa

skat

chew

an L

ibra

ry]

at 2

1:53

25

Aug

ust 2

012

SECTORAL SYSTEMS OF INNOVATION 73

service industry did not experience major technological and market discontinuities. With theinternet and its open network architecture, modular components and distributed intelligence,both the knowledge base and the types of actors and competencies have changed significantly.

The process of convergence has generated the entry of several new actors coming fromvarious previously separated industries, each one emphasizing different sets of competencies.For example, in telecommunication equipment and networks firms may range from incum-bent telecom equipment suppliers and incumbent network operators, to new entrant telecomoperators, cable TV operators and alternative network providers. In internet services, firmsmay range from internet service providers to internet content providers, e-commerce compa-nies, and software and internet specialised consulting companies. Specialised competenciesand specific knowledge have increasingly become a key asset for firms survival and growth.Even more important in the new telecom environment is the combination of existing and newcompetencies – software programming, network management and content provision – whichtraditionally belonged to different companies (Corrocher, 2002). Networks among a variety ofactors (not only firms, but also standard-setting organisations and research organisations) arerelevant.7 Demand plays a key role in innovation not just in terms of user–producer interaction,but also in terms of emerging characteristics. This is particularly true in the internet servicessector, where the changing requirements of the final users – from standardised services likeinternet access and e-mails, to more complex applications such as intranets, extranets andplatforms for electronic commerce – have stimulated firms to upgrade the quality of services(Edquist, 2004).

Regulation, liberalisation/privatisation and standards have played a key role in the organi-sation and performance of the sector. They had major effects on the behaviour of incumbentsand have transformed the structure of the industry.

In summary, in telecommunication equipment and services a convergence of differenttechnologies, demand and industries has taken place. This required processes of knowledgeintegration by the actors in the sectors. This convergence has been associated with the creationof a wide variety of different specialised and integrated actors, ranging from large equipmentproducers to new service firms. In this broad sector, innovation is very much affected bystandards, the institutional setting and the processes of privatisation and liberalisation.

4.4 Software

The software sectoral system has a quite differentiated knowledge base, with extended com-plementarities. Here, knowledge refers both to the control of the operations of the computersystem providing the platform for the different functionalities and to the software employingthese functionalities. However, the boundaries between operating systems and application soft-ware are becoming blurred, because of the dynamics of the inward (from software designersto the definition of system resources) and outward (from system level software to the userinterface) integration of software functions (Steinmueller, 2004). The strength of the forcesfavouring the creation of generic platforms (and therefore favouring internationally dominantplatform suppliers) is moderated by other forces: the continuing need for variety generation inthe organisations producing the sub-systems (that allow these platforms to be customised); thepotential for new methods for ‘platform’ creation (based upon the use of the internet as a toolfor collaborative innovation and the distribution of software products); the identification ofemerging areas where dominance in ‘platform’ creation remains contestable (such as embed-ded software); the identification of areas of the software industry that remain in a pre-dominant

7 For example networks played a role in the case of GSM (Hommen and Manninen, 2002).

Dow

nloa

ded

by [

Uni

vers

ity o

f Sa

skat

chew

an L

ibra

ry]

at 2

1:53

25

Aug

ust 2

012

74 F. MALERBA

design state of variety generation (such as multimedia software). Much of the innovative chal-lenge of the software industry, therefore, involves design innovation, not only in the basicoperations of the information processing ‘machine’ defined by software, but also in the veryconceptualisation of the information that needs to be processed. Nowadays, the three broadsoftware product groups (global package, situated software and middleware software) requiredifferent types of knowledge and learning processes. Global package software is characterisedby search for generic solutions, experience as a major input for innovation and a key role of pro-cess innovation. Situated and embedded software on the other hand has knowledge related tospecific contexts and specialised purposes. Middleware software and integrated software solu-tions – such as product data managers (PDM) and enterprise resource planning (ERP) – aimto reach many users but focus on situated specific applications (Mowery, 1996; Torrisi, 1998;Steinmueller, 2004).

In software, the changing knowledge base and the blurring boundaries between operationsystems and application software have created an evolving division of labour among users,‘platform’ developers and specialised software vendors, and a further tension between hor-izontal integration and specialisation. The historical role of computer producers has largelybeen displaced by a division of labour between software and hardware ‘platform’ producers,which is governed by the needs of the other as well as by the aim to preserve market positions.The sectoral system of innovation in software, however, is incomplete without the additionof companies that utilise these platforms to deliver enterprise-critical applications. Many ofthese applications continue to be self-produced by organisations that use the tools providedas part of the platform or available from the development tools markets. This process, how-ever, is creating a market for specialised software producers whose outputs are aimed at thecustomisation of the needs of a particular class of users.

IPRs play a major role in strengthening appropriability, but have been greatly affectedby the emerging open source movement. In addition, standards play a major role. Standarddevelopment organisations, country and industry consortia, and standards setting alliances arevery important. Networks of users also play an increasingly important function. Users alsooften gather around user mailing lists: these are used as vehicles to test and compare theperformance and capabilities of competing software products (Steinmueller, 2004).

In summary: software has a highly differentiated knowledge base in which the contextof application is relevant. This has created several different and distinctive product groups.The role of large computer suppliers in developing integrated hardware and software systemshas been displaced by a lot of specialised software companies innovating either in packagesoftware or in customised software. User–producer interaction, global and local networks andhigh mobility of skilled human capital are present. The role of the university has becomeimportant in the open source domain. IPR, standards and standard setting alliances play amajor role in innovation, diffusion and competition.

4.5 Machine Tools

In machine tools innovation is incremental and now also increasingly systemic. Knowledgeabout applications is very important: therefore user–producer relationships as well as partner-ships with customers are common. The knowledge base is embodied in skilled personnel withapplied technical qualifications on the shop floor level and in design engineers with a long-termemployment in the company. Internal training (particularly apprenticeship) is quite relevant.In small firms, R&D is not done extensively and R&D cooperation is not common. The sectoris characterised by national differences in the structure of demand which have in turn led tointernational differences in the rate and direction of technical change. Recently, however, theknowledge base has shifted from purely mechanical to mechanical as well as microelectronic

Dow

nloa

ded

by [

Uni

vers

ity o

f Sa

skat

chew

an L

ibra

ry]

at 2

1:53

25

Aug

ust 2

012

SECTORAL SYSTEMS OF INNOVATION 75

and information intensive, with increasing codification and an increasing use of formal R&D.Products have increasingly been modularised and standardised. A key role is also played byinformation flows about components among producers of different inputs and technologies,such as lasers, materials or measurement and control devices (Mazzoleni, 1999; Wengel andShapira, 2004).

Firms are highly specialised, and often focused on specific vertical segments. Networkshere differ from country to country, because the types of products and the different users anddemand structures have led to different innovation systems. In any case, local financial organi-sations and vertical links with users play a major role. Organisations and networks engaged intechnology transfer in a broad sense have developed. Market mechanisms increasingly showup in previously ‘non-market’ relationships, such as the cooperation in industrial/professionalassociations, or special customer–supplier interactions. And public–private industry consortiaincreasingly complement the latter (Wengel and Shapira, 2004).

Internal and regional labour markets and local institutions (e.g. local banks) play a major rolein influencing the international advantages of specific areas. Trust-based, close relationshipsat the regional level have over a long time ensured sufficient financing of the innovationand expansion plans of family businesses in Germany and Italy. The consequence has beenthat other more risky or expensive ways were rarely used and more radical changes seldomtook place. In Germany, vocational training has greatly fostered the development of skills inthe machine tools industry. Fairly stable employment conditions and company employmentstrategies (internal labour markets) formed the background for cumulative knowledge buildingand incremental innovations. Standards have a long tradition not only with respect to healthand safety but also with respect to economies of scale. They built a basis for the share ofdevelopment tasks between the machine tool makers and the suppliers of components andperiphery equipment. This adds again to a predominantly incremental innovation regime. TheEU machine directive was fundamental for the realisation of a common market, particularlyin the machine tool industry (Wengel and Shapira, 2004).

In summary, in machine tools an application-specific knowledge base has been associatedwith firms specialisation. Here user–producer interaction, local networks of innovators andin-house experienced human capital are key factors for innovation. However, recently prod-ucts are increasingly being modularised and standardised and suppliers of components areincreasingly involved in innovation.

4.6 What are the Main General Conclusions on the Role of Knowledge, Actors andNetworks and Institutions?

From the previous discussion, one general conclusion which can be drawn is that knowledgeat the base of innovative activities has been tremendously different from sector to sector. Theknowledge base has changed over time and has affected the boundaries and structure of sectoralsystems. In general, in several sectors a rich, multidisciplinary and multi-source knowledgebase and a rapid technological change have implied a great heterogeneity of actors. In additionto firms within a sector, some actors have proven particularly important for innovation. In par-ticular, suppliers and users have become relevant in the organisation of innovative activities.Suppliers and users have also affected the boundaries of sectoral systems by greatly affect-ing sectoral linkages and interdependencies. Demand has often proven important in severalrespects: a major cause in the redefinition of the boundaries of a sectoral system; a stimulusfor innovation and a factor shaping the organisation of innovative and production activities. Inaddition, the emergence of new demand or the transformation of existing demand has been oneof the major elements of change in sectoral systems over time. Often universities have playeda key role in basic research and human capital formation. In biotechnology and software, they

Dow

nloa

ded

by [

Uni

vers

ity o

f Sa

skat

chew

an L

ibra

ry]

at 2

1:53

25

Aug

ust 2

012

76 F. MALERBA

have also been a source of start-ups and even innovation. In these sectors, new actors such asventure capital have emerged over time.

In all the sectors examined, institutions have played a major role in affecting the rate oftechnological change, the organisation of innovative activity and the performance of sectoralsystems. Each sector, however, has seen the presence of a different set of relevant institutions.Some of these institutions are national and are present in all the sectors (and in all Europeancountries). Other institutions are sector-specific (i.e. they are present only in one sector). Insome cases, the relevant institutions in a sector have been the outcome of the interplay betweensectoral and national variables (Casper and Soskice, 2004; Coriat and Weinstein, 2004).

5 THE DYNAMICS AND TRANSFORMATION OF SECTORAL SYSTEMS

The dynamics and transformation of sectoral systems are the result of several differentprocesses. At a very general level, it is possible to identify two basic evolutionary pro-cesses – variety creation and selection (Nelson, 1995; Metcalfe, 1998). Processes of varietycreation in products, technologies, firms, institutions as well as strategies and behaviour maybe related to entry, R&D, innovation and so on. These mechanisms interact at various levels.For example, the emergence and growth of new sectoral institutions and organisations such asnew specialised departments within universities and new scientific, technological and educa-tional fields increase variety and can be associated with the emergence of new technologiesand new knowledge (Rosenberg and Nelson, 1993). This is the case of the chemical industrywith the emergence of new departments and engineering degrees in universities in response tonew technological developments in industry (Arora et al., 1999).

The creation of new agents – both new firms and non-firm organizations – is particularlyimportant for the dynamics of sectoral systems. For example, new firms bring a variety ofapproaches, specialisation and knowledge in the innovation and production processes. Theycontribute to the major changes in the population of agents and to the transformation of tech-nologies and products in a sector. As examined by Audretsch (1996) and Geroski (1995)among others, the role of new firms differs drastically from sector to sector (in terms of entryrates, composition and origin), and thus has quite different effects on the features of sectoralsystems and their degree of change. Sectoral differences in the level and type of entry seemto be closely related to differences in the knowledge base, in the level, diffusion and distri-bution of competencies and in the presence of non-firm organisations (such as universitiesand venture capital) and institutions (such as regulations or labour markets) (Geroski, 1995;Audretsch, 1996; Malerba and Orsenigo, 1999).

Processes of selection play the key role of reducing heterogeneity in terms of firms, products,activities, technologies, and so on. In addition to market selection, in several sectoral systemsnon-market selection processes are also at work (as in the cases of the involvement of themilitary, the health system and so on). In general, selection affects the growth and decline ofthe various groups of agents and the range of viable behaviours and organizations in a sectoralsystem (Metcalfe, 1998).

Change and transformation in sectoral systems is the result of the coevolution of variouselements: technology, knowledge base, learning, demand, firms, non-firm organisations andinstitutions. Nelson (1994) and Metcalfe (1998) have discussed these processes by focusingspecifically on the interaction between technology, industrial structure, institutions anddemand. In sectoral systems, changes in the knowledge base or in demand have affectedthe characteristics of the actors, the organisation of R&D and of the innovative process, thetype of networks, the structure of the market and the relevant institutions. All these variables

Dow

nloa

ded

by [

Uni

vers

ity o

f Sa

skat

chew

an L

ibra

ry]

at 2

1:53

25

Aug

ust 2

012

SECTORAL SYSTEMS OF INNOVATION 77

have in turn led to further modifications in the technology, the knowledge base, demand andso on.

These processes are sector-specific. For example, just looking at three elements such astechnology, demand and firms, in sectors characterised by a system product and consumerswith a quite homogeneous demand, coevolution leads to the emergence of a dominant designand industrial concentration (Klepper, 1996). However, in sectors with either a heterogeneousdemand, or competing technologies with lock-ins, or network externalities and standards,specialised products and a more fragmented market structure may emerge.

Often coevolution is related to path-dependent processes (David, 1985; Arthur, 1988). Herelocal learning, interaction among agents and networks may generate increasing returns andirreversibilities that may lock sectoral systems into inferior technologies. The cases of sectorswith competing technologies such as nuclear energy (Cowan, 1990), cars (and their powersources – Foreman and Peck, 1996), metallurgy (ferrous casting – Foray and Grubler, 1990)and multimedia (VCR – Cusumano, 1991) are interesting examples of path-dependent pro-cesses. Recent work such as Mowery and Nelson (1999) on the long-term evolution of sectorssuch as semiconductors, computers, software, pharmaceuticals and biotechnology, chemicals,medical devices and machinery show that these coevolutionary processes clearly differ amongsectors. The example of the computer industry is a case in point: its long-term developmentcannot just be described in terms of sales’ growth and the introduction over time of radicallynew products (such as the minicomputer, the microcomputer and the computer networks) withdifferent features and demand. Rather, in this sector complementarities between changes incomponents and changes in computer systems and coevolution among technology, demand,institutions and firms’ organisation and strategies have characterised the whole history of theindustry (Bresnahan and Malerba, 1999).

The transformation of sectors may also be related to the emergence of new clusters that spanover several sectors, such as internet–software–telecom, biotechnology–pharmaceuticals andnew materials. Here, transformation means the integration and fusion of previously separatedknowledge and technologies as well as new relations and dynamics among different typesof users and consumers, firms with different specialisation and competencies, and non-firmorganisations and institutions (all of them grounded in previously separated sectors).

In pharmaceuticals, changes in the knowledge base and in the relevant learning processesof firms have induced deep transformations in the behaviour and structure of the agents andin their relationships with each other. The specific ways these transformations have occurredacross countries have been profoundly different, due to the details of the institutional structureof each country. For example, the nature of the process of drug discovery and drug developmenthad an important impact on the patterns of competition and on market structure. In turn marketcompetition and market structure were dependent on the strategies and fortunes of individualcompanies, which were linked to different national contexts and international performance.Firms had diverse reactions in order to try to increase their fit and to survive in their particularenvironment. These environments kept changing, not least due to innovations and choices madeby all the constituent competitors. However, while these environments previously could be saidto be national, now the defining characteristics are increasingly international. Product approvalregulations inserted an incentive for more innovative strategies, at least for those firms andcountries which had the capabilities to invest in the new technologies. Similarly, weak patentprotection induced imitative strategies, but this effect was much less important for firms andcountries which had developed strong technological and scientific capabilities (for exampleGermany until the advent of the molecular biology revolution). Conversely, the introduction ofstronger patent protection might have contributed to the practical disappearance of the Italianindustry, which was until the mid-1970s one the most successful producer of generics. Afinal example is the molecular biology revolution which by creating new competencies and

Dow

nloa

ded

by [

Uni

vers

ity o

f Sa

skat

chew

an L

ibra

ry]

at 2

1:53

25

Aug

ust 2

012

78 F. MALERBA

a new technological regime, induced deep changes in the incentive structures within firms,universities, etc. In this process of adaptation and change, different dynamic processes lead todifferential patterns of competition and performance (McKelvey et al., 2004).

Also in chemicals processes of coevolution of technology, demand, markets, agents andinstitutions have been present. One interesting example of coevolutionary process in chem-icals is related to the environmental issue. The chemical industry has often been accused ofhaving considerable responsibility for pollution, and chemical firms, before others, have beenhighly committed to solve environmental problems. Some significant accidents (e.g. Seveso,Bhopal) have contributed to generate a widespread suspicion against chemical firms and theindustry as a whole. This greater attention paid by consumers to pollution and environmen-tal problems resulted in three different, but related consequences. First, all developed coun-tries have observed the rise of new markets for environmentally-safe, less pollutant products.Second, governments have paid greater attention to pollution, and have subsequently tried toimpose regulations and define appropriate control measures, in order to reduce waste produc-tion and pollution. Third, as a consequence of both forces, chemical firms have had to developand adopt new production technologies (environmental technologies and green processes), andnew products (e.g., less polluting solvents and paints). Moreover, rigid environmental standardsand strong public pressure have had a positive influence on the environmental innovations ofchemical firms. Indeed, another consequence of the growing attention to environmental issueshas been the birth of an intermediate market for environmental technologies and engineer-ing services related to environmental technologies. Similarly to the birth of SEFs providingprocess technologies in chemicals, new environmentally-related SEFs have started to operate(especially in US), and a new market for environmental technologies and engineering servicesis about to emerge (Arduini and Cesaroni, 2001).

In telecom equipment and services the early separation of the radio spectrum for use inone-way broadcasting and two-way telephony has given rise to the oligopolistic structure ofthe industry that persisted for quite a long time (Dalum and Villumsen, 2002). The conver-gence first within ICT and then between ICT and broadcasting-audio-visual, together with theemergence of internet, has originated a more fluid market structure with a lot of different actorswith different specialisation and capabilities, and new types of users. This in turn has greatlyexpanded the boundaries of the sector by creating new segments and new opportunities, andalso national differences in the organisation of innovation. Moreover, the emergence of theinternet has generated more pressure in favour of open standards and has led to the rise of newactors such as ISPs and content providers.

In software, since the early 1980s, the spread of networked computing, embedded software,the internet, the development of open system architectures and open source, and the growthof web-based network computing has led to the decline of large computer producers as devel-opers of integrated hardware and software systems, to the emergence of numerous specialisedsoftware companies innovative either in package software, or in customised software, and toan increasing role of the university in open source. This in turn has led to the expansion andgrowth of several software product groups, each of which has different types of products, firmsand capabilities. Moreover, software distribution has also changed accordingly, from licensingagreements in the early days, to the rise of independent software vendors, to price discountfor package software, and (with the diffusion of the CD-ROM and the internet) to sharewareand freeware (the latter is particularly relevant with Linux) (Steinmueller, 2004). In enterprisesoftware higher demand for integration by user organisations has reinforced the role of existingactors (i.e. large producers of standardised integrated software solutions) as well as has createdscope for new actors (i.e. systems integrators, specialised niche applications producers and soft-ware implementation consultants) (D’Adderio, 2001). The increasingly generic nature of largesystems has also introduced a greater need for customisation whereby customer knowledge and

Dow

nloa

ded

by [

Uni

vers

ity o

f Sa

skat

chew

an L

ibra

ry]

at 2

1:53

25

Aug

ust 2

012

SECTORAL SYSTEMS OF INNOVATION 79

requirements (expressed by global or industry-specific user group associations) have becomean important source of input into the development of new or revised modules, with user groupsattempting to directly influence the shaping of a system. In response to the increasing need forcustomisation, large software producers are pursuing a higher level of internal specialisationby creating sub-units that address a specific market segment and compete for resources withother units.

In machine tools a major driving force for coevolutionary processes is the demand fromadvanced customer sectors, namely automotive, aeronautics and defence. Incremental inno-vation has remained dominant, some internationalisation of production has taken place, anduser–producer relationships have become more market driven. A coevolutionary process canbe observed in the context of technological developments, namely in electronics but also withrespect to new materials. Electronic devices have an increasing share of the value of machinetools. IT technologies (PC, operating systems and internet) often determine technical solutionson how to control machine tools and on how to integrate them in company production sys-tems. As a consequence, besides electrical engineers, computer scientists have partly replacedmechanical engineers in the design departments of machine tool manufacturers and broughtwith them other ways of working. Some firms have followed strategies of outsourcing or sep-arating such units. On the shop-floor level a related change in required qualifications has takenplace. New apprenticeships have developed, others have disappeared. Links to basic researchare now looked for and patenting has been growing strongly in recent years.

6 THE CHALLENGES AHEAD

This paper has proposed a framework for examining factors that affect innovation in sectors:sectoral systems. Sectoral systems are based on three building blocks: knowledge and tech-nologies, actors and networks, and institutions. On that basis it has examined several sectorsin terms of both their basic features and their dynamics and transformation.

The sectoral system framework can be very useful for research on the internationalperformance of countries in specific sectors. One can claim that in several sectoral systemsdifferences between Europe, US and Japan in the sources of knowledge, types and competenciesof actors, networks and institutions have greatly affected countries’ international performanceand that the lack of success of some European countries in some sectors has been due toproblems and deficiencies in their sectoral systems. Even within the sectors in which Europedoes not fare well, those European countries that specialise in product groups with products,knowledge and institutional requirements that match their specific institutional framework aresuccessful (see for example the analysis in Coriat et al., 2004).

Also for technology and innovation policy a sectoral system of innovation approach mayprove useful. A sectoral system approach provides the identification of ‘system failures’ andthe related variables which should be policy targets. Sectoral analyses should focus on sys-temic features in relation to knowledge and boundaries, heterogeneity of actors and networks,institutions and transformation through coevolutionary processes.As a consequence, the under-standing of these dimensions becomes a prerequisite for any policy addressed to a specificsector. Given the major differences among sectoral systems, the impact of general or horizon-tal policies may drastically differ across sectors, because the channels and ways policies havetheir effects differ from sector to sector. For example, networks and non-firm organisationshave different relevance in different sectors. Therefore, policies affecting networks or non-firmorganizations such as transfer agencies have to take these differences into account. In addition,a sectoral system framework emphasises that for fostering innovation and diffusion in a sector,

Dow

nloa

ded

by [

Uni

vers

ity o

f Sa

skat

chew

an L

ibra

ry]

at 2

1:53

25

Aug

ust 2

012

80 F. MALERBA

technology and innovation policies may not be enough. A wide range of other policies may benecessary. Innovation and technology policy could be supplemented by other types of policies(such as science policy, industrial policy, policies related to standards and IPR and competitionpolicy). This point highlights also the importance of the interdependencies, links and feedbacksamong all of these policies, and their combined effects on the dynamics and transformation ofsectors (see for example the analysis by Edquist et al., 2004).

Acknowledgements

Parts of this paper draw from Malerba (2004) which has a much longer and detailed discussionof the sectors examined. I thank two anonymous referees for helpful comments.

References

Arduini, R. and Cesaroni, F. (2001) Environmental Technologies in the European Chemical Industry, LEM WorkingPaper 2001/09, Pisa: Scuola Superiore Sant’Anna.

Arora, A., Landau, R. and Rosenberg, N. (1999) Dynamics of Comparative Advantages in the Chemical Industry. InMowery, D. and Nelson, R. (eds) The Sources of Industrial Leadership. Cambridge: Cambridge University Press.

Arthur, B. (1988) Competing Technologies, Increasing Returns and Lock-Ins by Historical Events. Economic Journal,99(394), 116–131.

Audretsch, D. (1995) Innovation and Industry Evolution. Cambridge, MA: MIT Press.Breschi, S., Malerba, F. and Orsenigo, L. (2000) Technological Regimes and Schumpeterian Patterns of Innovation.

Economic Journal, 110(463), 388–410.Bresnahan, T. and Malerba, F. (1999) Industrial Dynamics and the Evolution of Firms and Nations Competitive

Capabilities in the World Computer Industry. In Mowery, D. and Nelson, R. (eds) The Sources of Industrial Lead-ership. Cambridge: Cambridge University Press.

Callon, M. (1992) The Dynamics of Techno-Economic Networks. In Coombs, R., Saviotti, P. and Walsh, V. (eds)Technical Change and Company Strategies. London: Academy Press.

Carlsson, B. (1995) Technological Systems and Economic Performance: The Case of Factory Automation. Dordrecht:Kluwer.

Carlsson, B. and Stankiewitz, R. (1995) On the Nature, Function and Composition of Technological Systems. InCarlsson, B. (ed.) Technological Systems and Economic Performance. Dordrecht: Kluwer.

Casper, S. and Soskice, D. (2004) Sectoral Systems of Innovation and Varieties of Capitalism: Explaining the Devel-opment of High-Technology Entrepreneurship in Europe. In Malerba, F. (ed.) Sectoral Systems of Innovation.Concepts, Issues and Analyses of Six Major Sectors in Europe. Cambridge: Cambridge University Press.

Cesaroni, F., Gambardella, A., Garcia-Fontes, W. and Mariani, M. (2004) The Chemical Sectoral System. Firms,Markets, Institutions and the Processes of Knowledge Creation and Diffusion. In Malerba, F. (ed.) SectoralSystems of Innovation. Concepts, Issues and Analyses of Six Major Sectors in Europe. Cambridge: CambridgeUniversity Press.

Chandler, A. (1990) Scale and Scope: The Dynamics of Industrial Capitalism. Cambridge, MA and London: HarvardUniversity Press.

Cohen, W.M. and Levin, R. (1989) Empirical Studies of Innovation and Market Structure. In Schmalensee, R. andWillig, R. (eds) Handbook of Industrial Organization. New York: North Holland.

Cooke, P., Gomez Urange, M. and Extebarria (1997) Regional Innovation Systems: Institutional and Organizationaldimensions. Research Policy, 4–5, 475–493.

Coriat, B. and Weinstein, O. (2004) National Institutional Frameworks, Institutional Complementarities and SectoralSystems of Innovation. In Malerba, F. (ed.) Sectoral Systems of Innovation. Concepts, Issues and Analyses of SixMajor Sectors in Europe. Cambridge: Cambridge University Press.

Coriat, B., Malerba, F. and Montobbio, F. (2004) The International Performance of European Sectoral Systems. InMalerba, F. (ed.) Sectoral Systems of Innovation. Concepts, Issues and Analyses of Six Major Sectors in Europe.Cambridge: Cambridge University Press.

Corrocher, N. (2002) The Internet Services Industry: Sectoral Dynamics of Innovation and Production and Country-Specific Trends in Italy and in the UK. Working Paper ESSY. Available at http://www.cespri.it/ricerca/es wp.htm.

Cowan, R. (1990) Nuclear Power Reactors a Study in Technological Lock. Journal of Economic History, 50(3),541–567.

Cusmano, L. (1991) Japan’s Software Factory. Oxford: Oxford University Press.D’Adderio, L. (2001) Inside the Virtual Product: The Influence of Integrated Software Systems on Organisational

Knowledge Dynamics. SPRU, Brighton: University of Sussex.Dalum, B. and Villumsen, G. (2002) Fixed Data Communications – Challenges for Europe. Working Paper ESSY.

Available at http://www.cespri.it/ricerca/es wp.htm.

Dow

nloa

ded

by [

Uni

vers

ity o

f Sa

skat

chew

an L

ibra

ry]

at 2

1:53

25

Aug

ust 2

012

SECTORAL SYSTEMS OF INNOVATION 81

Dasgupta, P. and Stigliz, J. (1980) Industrial Structure and the Nature of Innovation Activity. Economic Journal,90(358), 266–293.

David, P. (1985) Clio and the Economic of the QWERTY. American Economic Review, 75(2), 332–337.Dosi, G. (1982) Technological Paradigms and Technological Trajectories. A Suggested Interpretation of the Determi-

nants and Directions of Technical Change. Research Policy, 11(3), 147–208.Dosi, G. (1988) Sources, Procedures and Microeconomic Effects of Innovation. Journal of Economic Literature, 26(3),

1120–1171.Dosi, G. (1997) Opportunities, Incentives and the Collective Patterns of Technological Change. Economic Journal,

107(444), 1530–1547.Edquist, C. (ed.) (1997) Systems of Innovation. London: Frances Pinter.Edquist, C. (2004) The Fixed Internet and Mobile Telecommunications Sectoral System of Innovation: Equipment

Production, Access Provision and Content Provision. In Malerba, F. (ed.) Sectoral Systems of Innovation. Concepts,Issues and Analyses of Six Major Sectors in Europe. Cambridge: Cambridge University Press.

Edquist, C., Malerba, F., Metcalfe, S., Montobbio, F. and Steinmueller, E. (2004) Sectoral Systems: Implications forEuropean Technology Policy. In Malerba, F. (ed.) Sectoral Systems of Innovation. Concepts, Issues and Analysesof Six Major Sectors in Europe. Cambridge: Cambridge University Press.

Foray, D. and Grubler, A. (1990) Morphological Analysis, Diffusion and Lock-Out of Technologies: Ferrous Castingin France and the FRG. Research Policy, 19(6), 535–550.

Foreman-Peck, J. (1996) Technological Lock-In and the Power Source for the Motor Car. University of Oxford,Discussion Paper in Economics.

Freeman, C. (1987) Technology Policy and Economic Performance: Lessons from Japan. London: Frances Pinter.Geroski, P. (1995) What Do We Know About Entry? International Journal of Industrial Organization, 13(4), 421–440.Gort, M. and Klepper, S. (1982) Time Paths in the Diffusion of Product Innovations. Economic Journal, 92(367),

630–653.Henderson, R., Orsenigo, L. and Pisano, G. (1999) In Mowery, D. and Nelson, R. (eds) The Pharmaceutical Industry

and the Revolution in Molecular Biology. cit. 267–312.Hommen, L. and Manninen, E. (2001) GSM (Global System for Mobile Telecommunications). Working Paper ESSY.

Available at http://www.cespri.it/ricerca/es wp.htm.Hughes, T.P. (1984) The Evolution of Large Technological Systems. In Bijker, W., Hughes, T. and Pinch, T. (eds) The

Social Construction of Technological Systems. Cambridge, MA: MIT Press.Kamien, M. and Schwartz, N. (1982) Market Structure and Innovation. Cambridge: Cambridge University Press.Klepper, S. (1996) Entry, Exit, Growth and Innovation Over the Product Life Cycle. American Economic Review,

86(3), 562–583.Lane, D. (1993a) Artificial Worlds and Economics Part I. Journal of Evolutionary Economics, 3(1), 89–107.Lane, D. (1993b) Artificial Worlds and Economics Part II. Journal of Evolutionary Economics, 3(1), 177–197.Levin, R., Cohen, W.M. and Mowery, D. (1985) R&D Appropriability, Opportunity and Market Structure. New

Evidence on Some Schumpeterian Hypotheses. American Economic Review, Papers and Proceedings, 75(2), 20–24.Levin, R., Klevorik, A., Nelson, R.R. and Winter, S. (1987) Appropriating the Returns from Industrial Research and

Development. Brookings Paper on Economic Activity, 19(2), 783–831.Lundvall, B.A. (1993) National Systems of Innovation. London: Frances Pinter.Malerba, F. (2004) Sectoral Systems of Innovation. Concepts, Issues and Analyses of Six Major Sectors in Europe.

Cambridge: Cambridge University Press.Malerba, F. and Orsenigo, L. (1990) Technological Regimes and Patterns of Innovation: A Theoretical and Empirical

Investigation of the Italian Case. In Heertje, A. (ed.) Evolving Industries and Market Structures. Ann Arbor, MI:University of Michigan Press.

Malerba, F. and Orsenigo, L. (1993) Technological Regimes and Firm Behavior. Industrial and Corporate Change,2(1), 45–74.

Malerba, F. and Orsenigo, L. (1996) Schumpeterian Patterns of Innovation, Cambridge. Journal of Economics, 19(1),47–65.

Malerba, F. and Orsenigo, L. (1999) Technological Entry, Exit and Survival: An Empirical Analysis of Patent Data.Research Policy, 28(6), 643–660.

Marsili, O. (2001) The Anatomy and Evolution of Industries: Technological Change and Industry Dynamics.Cheltenham: Edward Elger.

Mazzoleni, R. (1999) Innovation in the Machine Tool Industry: A Historical Perspective on the Dynamics of Com-parative Advantage. In Mowery, D. and Nelson, R. (eds) Sources of Industrial Leadership. Cambridge: CambridgeUniversity Press.

McKelvey, M., Orsenigo, L. and Pammolli, F. (2004) Pharmaceuticals Analyzed Through the Lens of a SectoralInnovation System. In Malerba, F. (ed.) Sectoral Systems of Innovation. Concepts, Issues and Analyses of Six MajorSectors in Europe. Cambridge: Cambridge University Press.

Metcalfe, S. (1998) Evolutionary Economics and Creative Destruction. London: Routledge.Mowery, D. (ed.) (1996) The International Computer Software Industry: A Comparative Study of Industry Evolution

and Structure. Oxford: Oxford University Press.Mowery, D. and Nelson, R. (1999) The Sources of Industrial Leadership. Cambridge: Cambridge University Press.Nelson, R. (1993) National Innovation Systems: A Comparative Study. Oxford: Oxford University Press.Nelson, R. (1994) The Coevolution of Technology, Industrial Structure and Supporting Institutions. Industrial and

Corporate Change, 3(1), 47–63.

Dow

nloa

ded

by [

Uni

vers

ity o

f Sa

skat

chew

an L

ibra

ry]

at 2

1:53

25

Aug

ust 2

012

82 F. MALERBA

Nelson, R. (1995) Recent Evolutionary Theorizing about Economic Change. Journal of Economic Literature, 33(1),48–90.

Nelson, R. and Rosenberg, N. (1993) Technical Innovation and National Systems. In Nelson, R. (ed.) NationalInnovation Systems. Oxford: Oxford University Press.

Nelson, R. and Winter, S. (1977) In Search of Useful Theories of Innovation. Research Policy, 6(1), 36–76.Nelson, R. and Winter, S. (1982) An Evolutionary Theory of Economic Change. Harvard, MA: The Belknapp Press

of Harvard University Press.Pavitt, K. (1984) Sectoral Patterns of Technical Change: Towards a Taxonomy and a Theory. Research Policy, 13(6),

343–375.Rosenberg, N. (1976) Perspectives on Technology. Cambridge University Press.Rosenberg, N. (1982) Inside the Black Box: Technology and Economics. Cambridge: Cambridge University Press.Saxenian, A. (1994) Regional Advantages. Cambridge, MA: Harvard University Press.Steinmueller, E. (2004) The European Software Sectoral System of Innovation. In Malerba, F. (ed.) Sectoral Systems

of Innovation. Concepts, Issues and Analyses of Six Major Sectors in Europe. Cambridge: Cambridge UniversityPress.

Sutton, J. (1998) Technology and Market Structure. Cambridge, MA: MIT Press.Teubal, M., Yinnon, T. and Zuscovitch, E. (1991) Networks and Market Creation. Research Policy, 20(5), 381–392.Torrisi, S. (1998) Industrial Organisation and Innovation: An International Study of the Software Industry. Chel-

tenham: Edward Elgar.Wengel, J. and Shapira, P. (2004) Machine Tools: the Remarking of a Traditional Sectoral Innovation System. In

Malerba, F. (ed.) Sectoral Systems of Innovation. Concepts, Issues and Analyses of Six Major Sectors in Europe.Cambridge: Cambridge University Press.

Winter, S. (1984) Schumpeterian Competition in Alternative Technological Regimes. Journal of Economic Behaviourand Organisation, 5(3–4), 287–320.

Dow

nloa

ded

by [

Uni

vers

ity o

f Sa

skat

chew

an L

ibra

ry]

at 2

1:53

25

Aug

ust 2

012