RETHINKING REGIONAL INNOVATION AND CHANGE: PATH DEPENDENCY OR REGIONAL BREAKTHROUGH
Economics of Science‚ Technology ...
86 downloads
1128 Views
5MB Size
Report
This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Report copyright / DMCA form
RETHINKING REGIONAL INNOVATION AND CHANGE: PATH DEPENDENCY OR REGIONAL BREAKTHROUGH
Economics of Science‚ Technology and Innovation VOLUME 30
Series Editors Cristiano Antonelli‚ University of Torino‚ Italy Bo Carlsson‚ Case Western Reserve University‚ U.S.A.
Editorial Board: Steven Klepper‚ Carnegie Mellon University‚ U.S.A. Richard Langlois‚ University of Connecticut‚ U.S.A. J.S. Metcalfe‚ University of Manchester‚ U.K. David Mowery‚ University of California‚ Berkeley‚ U.S.A. Pascal Petit‚ CEPREMAP‚ France Luc Soete‚ Maastricht University‚ The Netherlands
The titles published in this series are listed at the end of this volume.
Economics of Science‚ Technology and Innovation
RETHINKING REGIONAL INNOVATION AND CHANGE: PATH DEPENDENCY OR REGIONAL BREAKTHROUGH
edited by
GERHARD FUCHS
and PHILIP SHAPIRA
Springer
eBook ISBN: Print ISBN:
0-387-23002-5 0-387-23001-7
©2005 Springer Science + Business Media, Inc. Print ©2005 Springer Science + Business Media, Inc. Boston All rights reserved No part of this eBook may be reproduced or transmitted in any form or by any means, electronic, mechanical, recording, or otherwise, without written consent from the Publisher Created in the United States of America
Visit Springer's eBookstore at: and the Springer Global Website Online at:
http://www.ebooks.kluweronline.com http://www.springeronline.com
CONTENTS List of Figures
vii
List of Tables
viii
Contributors
ix
Preface
xi
Acknowledgments
1.
2.
3.
4.
5.
6.
Beyond path dependency and competitive convergence: Institutional transfer from a discourse-analytical perspective Christoph Scherrer
xix
1
Tacit knowledge‚ path dependency and local trajectories of growth Meric S. Gertler
23
Regional transformation and regional disequilibrium: New knowledge economies and their discontents Philip Cooke
43
Switching ties‚ recombining teams: Avoiding lock-in through project organization? Gernot Grabher
63
Knowledge-intensive services as a key sector for processes of regional economic innovation: Leapfrogging and path dependency Hans Joachim Kujath
85
Entrepreneurship as a source of path dependency Udo Staber
7. Geographical proximity and the diffusion of knowledge. The case of SME’s in biotechnology Delphine Gallaud and André Torre
107
127
vi
8.
9.
10. 11.
12. 13. 14.
Continuities‚ ruptures‚ and re-bundling of regional development paths: Leipzig’s metamorphosis Harald Bathelt and Jeff Boggs
147
Can less favored regions change their destiny? Lessons from Europe Lena J. Tsipouri
171
Innovation challenges and strategies in catch-up regions Philip Shapira
195
Path dependency in Baden-Württemberg: Lock-in or breakthrough? Gerhard Fuchs and Sandra Wassermann
223
Rethinking regional innovation policy Ron Boschma
249
On the role of global demand in local innovation processes Anders Malmberg and Dominic Power
273
The regionalization of innovation policy: New options for regional change? Knut Koschatzky
291
Index
313
LIST OF FIGURES 8.1
Relationships between regions and technological trajectories 8.2 Sequence of events leading from crisis to re-bundling 8.3 Firms of Leipzig’s book publishing industry‚ 1939 8.4 Firms and institutions of Leipzig’s “new” media industry‚ 2000 10.1 Regional relative to U.S. per capita income‚ BEA regions‚ 1929-1999 10.2 Dispersion in state per capita income in the U.S.‚ annual standard deviation‚ 1929-1999 10.3 State of Georgia – major cities 10.4 Per capita income trends – Georgia and other reference areas‚ 1929-1999 10.5 Corporate strategies and associated returns‚ Georgia manufacturers 1999 10.6 Georgia per capita income trends‚ 1969-1999 11.1 Products with high technology content and BadenWürttemberg’s share of world exports in this category (by sector). 14.1 Elaboration of a regional innovation strategy
152 157 167 163 196 196 202 203 204 213
238 300
LIST OF TABLES 3.1 5.1 5.2 5.3 5.4 7.1 9.1 9.2 9.3 9.4 9.5 9.6 9.7 10.1 10.2
11.1
11.2 11.3 12.1 12.2 14.1
Growth in engineering and science graduates‚ 1975-1995 Types of knowledge conversion within the framework of knowledge services Types of knowledge services in Berlin and Munich‚ 2002 Spatial distribution of customers of knowledge service providers – Berlin‚ Munich Spatial distribution of knowledge sources – Berlin‚ Munich Temporary and permanent geographical proximity in technological co-operation processes The variable geometry of the European “less favored regions” (LFRs) The size of the informal sector‚ selected European countries The “cleaned” less favored regions in descending order of GDP change‚ 1988-1993 and 2000-2006 Regions that grew fast from 1988-1998 Relative position‚ population density‚ and European regional support in the winning regions Levels of R&D inputs in the winning regions Institutional thickness and regional growth Summary of Georgia technology programs (2000) Georgia’s economic development and technology performance Stuttgart - Mittlerer Neckar: Employment and sales turnover of the most important sub-sectors of the manufacturing sector (2000) Importance of high technology commodities in selected countries and regions (Nominal scale) New fields of activities in the region of Stuttgart Two types of evolutionary change Two ideal-types of regional innovation policy Spatial concentration of industrial R&D in selected countries
56 88 96 98
100 137 178 180 181 183 184 187 190 207 211
233 237 243 258 263 304
CONTRIBUTORS Harald Bathelt is Professor of Economic Geography in the Faculty of Geography‚ Philipps-University of Marburg‚ Germany. Jeff Boggs is a Doctoral Candidate in the Department of Geography‚ University of California–Los Angeles‚ USA. Ron Boschma is an Associate Professor of Regional Economics‚ Faculty of Geographical Sciences‚ Utrecht University‚ The Netherlands. Philip Cooke is Professor of Regional Development and Director of the Centre for Advanced Studies at Cardiff University‚ United Kingdom. Gerhard Fuchs is Deputy Director of the Department of Technology‚ Organization‚ and Work‚ Center for Technology Assessment‚ Stuttgart‚ Germany. Delphine Gallaud is a Doctoral Candidate with the Institut National de la Recherche Agronomique (INRA) and Institut de Recherche Interdisciplinaire en Socioéconomie (IRIS)‚ Université Paris Dauphine (Paris IX)‚ France. Meric S. Gertler is Goldring Chair in Canadian Studies and Professor‚ Department of Geography and Program on Globalization and Regional Innovation Systems‚ University of Toronto‚ Canada. Gernot Grabher is Professor of Economic Geography and Head of the Research Area Socio-Economics of Space at the University of Bonn‚ Germany. Knut Koschatzky is Head of the Department of Innovation Services and Regional Development at the Fraunhofer Institute for Systems and Innovation Research‚ Karlsruhe‚ Germany. Hans Joachim Kujath is Head of the Department of Regionalization and Economic Spaces at the Institute for Regional Development and Structural Planning‚ Erkner‚ Germany. Anders Malmberg is Professor in the Department of Social and Economic Geography‚ Uppsala University‚ Sweden.
x
Dominic Power is Associate Professor in the Department of Social and Economic Geography‚ Uppsala University‚ Sweden. Christoph Scherrer is Professor of Global Political Economy‚ Faculty of Social Sciences‚ University of Kassel‚ Germany. Philip Shapira is Professor of Public Policy at Georgia Institute of Technology‚ Atlanta‚ USA. Udo Staber is Professor of Organizational Studies and Chair of the International Management Area at the Stuttgart Institute of Management and Technology‚ Germany. André Torre is Professor of Economics and Research Director at the Institut National de la Recherche Agronomique (INRA)‚ Paris‚ France. Lena J. Tsipouri is Associate Professor in the Department of Economic Sciences‚ University of Athens‚ Greece. Sandra Wassermann is an Assistant Professor in the Department of Political Science at the University of Stuttgart‚ Germany.
Preface RETHINKING REGIONAL INNOVATION AND CHANGE: PATH DEPENDENCY OR REGIONAL BREAKTHROUGH? Gerhard Fuchs and Philip Shapira
The extent to which regions can diverge from established paths of economic development continues to be a matter for heated debate. There are potent arguments that any new constitution of regional economies is determined to a considerable degree by institutional and industrial structures that have emerged in the course of regional industrialization history often spanning hundreds of years. Such structural lineage is often considered to hinder the adjustment to new challenges. The increasing globalization of national and regional economies‚ the rise of supra-national economic blocs‚ ongoing developments in technology‚ and new organizational strategies are among the elements that individually and collectively add new spins to this path dependency debate. Simultaneously‚ these challenges bring along new possibilities of innovation‚ transfer‚ and the adaptation of knowledge. New technologies offer new perspectives and institutional structures‚ learning systems‚ and policy mechanisms might serve as change factors that promise fresh opportunities for regions to pursue new development trajectories. This encourages the intrinsic optimism of many policy makers and local and regional economic developers for regions‚ especially those that are currently less favored‚ to depart from established development paths and to generate higher levels of value-added activity. At the same time‚ change may expose limits to adaptation within regions‚ presenting challenges of industrial restructuring‚ regional crisis‚ continued regional lag‚ or the inability to capitalize on apparent new opportunities – caused‚ at least in part‚ by deep-rooted economic‚ social‚ or political traditions that constrain adjustment and development. The contribution of a volume such as this is to reconsider and advance theories and practices in understanding regional innovation and change in developed societies. We seek to cast light both on the importance of economic-structural and institutional path dependencies as well as on the conditions under which divergence from a path dependent development is viable. In so doing‚ we hope to advance both intellectual debate and policy thinking. Concepts of path dependency and their relationships to structural processes of change at the regional level will be explored and complemented by regional
xii
Gerhard Fuchs and Philip Shapira
case studies from various countries‚ providing important implications for policy and practice. The joint starting points are considerations on how to frame and understand contemporary regional development. All contributions probe the assumption that there are both technological (cf. Dosi‚ 1982) and regional trajectories. Technological knowledge is not only organized in largescale technical systems (Hughes‚ 1987)‚ in branches or in professions but frequently also in regional economic areas‚ where it is embedded in a broad institutional and cultural context. Knowledge incorporated in regional production clusters‚ cooperative relations‚ institutions‚ and policy patterns does not usually develop in great leaps and bounds‚ but incrementally‚ in a mostly evolutionary manner. However‚ prior to discussing the application of path dependency ideas to these processes of regional development‚ it is instructive‚ if not essential‚ to begin by exploring the concept of path dependency and how it developed. Thus‚ in the opening contribution to this volume (chapter 1)‚ Christoph Scherrer provides us with an introduction to the theoretical background of path dependency‚ emphasizing the influence of historical institutionalism. Scherrer further conceptualizes and explores a range of possible trajectories‚ between institutional lock-in and institutional breakthrough. Given today’s global economic conditions and frequently expressed notion of knowledge society‚ the development and institutionalization of collaborative mechanisms of learning is assuming increasing importance – elevating the importance and role of collective levels of reflection. Along this line of thought‚ the book seeks to expose typological classifications concerning preconditions and different courses that processes of learning follow at the regional level. Several contributors to this volume explore how regions can be differentiated with regard to their industrial development paths and their present problems as concrete expressions of regional “learning curves” and “learning requirements.” It has been observed that “technological capabilities ... reflect local‚ regional and national contexts and environments” (Storper‚ 1995: 897). However‚ while we assume that technical knowledge and technological learning are influenced by context and region‚ this still raises issues about the extent to which learning processes track a path dependent course and‚ if so‚ exactly how that learning path correlates with and is bounded by the development paths of technologies and regions. Meric Gertler (chapter 2) examines the notion and concept of tacit knowledge and relates this to the question of whether tacit knowledge can be learned. The chapter points out the limits of reflection: according to Gertler’s arguments‚ actors are not aware of their knowledge‚ since tacit knowledge cannot be separated from its institutional and cultural context. In this perspective‚ knowledge cannot be transferred from one region to another‚ and it is therefore impossible for policy makers to implement regional learning by referring
PREFACE
xiii
to more successful‚ comparable regions. However‚ this argument is challenged by Philip Cooke’s analysis of regional transformation (chapter 3). Cooke highlights both the necessity and the opportunity for integrating innovation and learning in regional systems. In so doing‚ Cooke observes that best practice can be transferred to catch-up regions and thus will generate new opportunities for development. It is apparent that regionally-developed assets that may be embedded in densely woven networks of interactive and exchange relationships must be comprehended as the precondition for understanding regional problems and regional capacities for taking action. This would seem to be one best practice that is transferable. However‚ to what extent does this observation imply that the future development of every region is “pre-structured” by the technological‚ economical‚ cultural‚ political‚ and social history that underpins its current array of assets? Based on the assumption that pre-structuring is not complete‚ several authors explore the dimensions of the “space” for the policyinduced reorganization and development of regional technological capabilities. As mentioned above‚ Cooke sees considerable capacities for regional actors initiating economic development. Similarly‚ Ron Boschma (chapter 12) and Knut Koschatzky (chapter 14) focus on the roles that regional innovation policy can play in transforming regional structures and innovation capacities. Boschma distinguishes between “localized” and “structural” change‚ arguing that different policies are needed to address these varied conditions. According to his typology‚ localized change requires an evolutionary type of policy‚ whereas structural change asks for revolutionary policy methods. In the first case‚ policy makers are charged with reducing the risk of following regional paths that lead to lock-in situations. Whereas in the second case‚ there is the demand for policy makers to actively stimulate diversity and increase returns. Koschatzky’s contribution explores how innovation policy actors at different levels have transformed their objectives throughout the 1990s as a reflection of academic research. He observes that studies and concepts of regional innovation have had decisive effects on policy practice‚ causing the convergence of European‚ federal and state policies into regional innovation policies. Yet new research suggests that some premises of these concepts should be rethought as they may be misleading or‚ at least‚ less crucial than implicitly assumed. According to Anders Malmberg and Dominic Power (chapter 13) there are weaknesses in the geographical theories of the rediscovered economist Alfred Marshall and his present-day adherents that proximity is a precondition for innovation. Instead‚ Malmberg and Powers argue that the economic structures and networks that develop at regional levels do so for other reasons and may not be precursors or promoters of innovation. The authors provide an alternative demand-driven explanation for innovation
xiv
Gerhard Fuchs and Philip Shapira
and point out that‚ in the course of the globalizing economy‚ additional extraregional and international contacts and networks are crucial for regional innovation systems to secure their innovation capabilities‚ as nowadays firms have to respond to global demand. Delphine Gallaud and André Torre (chapter 7) reinforce this critique of current regional development concepts. Like Malmberg and Power‚ they also question frequently stated ideas about the relationship between geographical proximity and regional innovation and cooperation. According to Gallaud and Torre‚ geographical proximity only influences the innovative performance of firms if there is effective interaction between the agents. They argue that the key element in the transmission of knowledge is the characteristics of a firm’s organization rather than its geographic closeness to other enterprises. One means of increasing regional innovative capacity is by strengthening‚ re-orientating‚ or creating institutions – fostering‚ in effect‚ what Ash Amin and Nigel Thrift (1994) have termed “institutional thickness.” They remind us that this concept has multiple implications. First‚ the staying power of institutions is of considerable significance in regional development. Second‚ the local construction and enrichment of a reservoir of shared knowledge is a notable element of regional dynamics. Third‚ the capacity for learning and change is inherent to institutional flexibility within regions. Fourth‚ the innovative capacity of companies is viewed as a shared characteristic of a given region. Fifth‚ regional interactions are firmly based on trust and reciprocity. Finally‚ regions give rise to a consolidated feeling of belonging among their inhabitants‚ reinforcing regional social capital. For Amin and Thrift‚ regional trajectories are secured partly by established economic structures and production clusters‚ and partly by regional institutions. Particularly in innovation research‚ which tends to concentrate on industrial and technological changes‚ the importance of the region as a collective sphere of economic activities‚ institutional and social relationships‚ and political negotiations cannot be underestimated. This line of argument suggests that economic position of a region within the context of global competition is partly the result of path-dependent developments‚ but is also influenced by current institutionally-anchored governance structures that influence the regional innovation capacity. The institutions of the regional innovation system not only serve as a resource for firms to draw on (technological know-how‚ qualified workforce‚ cooperative labor relations‚ etc.)‚ but also fulfill important orientation and governance functions for regional actors in industry‚ science and politics. As institutional thickness increases and gives rise to institutional inertia‚ the industrial development paths of a region also become institutionally stabilized; technological path dependencies are thus accompanied by institutional ones. Given this‚ we have to explore how to deal with the challenges involved in reforming and
PREFACE
xv
developing regions and their industrial‚ innovation‚ learning‚ and institutional systems. The concept of institutional thickness and its relationship to path dependency is empirically tested by Hans Joachim Kujath’s study of the newly developed knowledge economies of the cities of Munich and Berlin (chapter 5). While economic restructuring in Munich still can be interpreted in the context of old economic structures and paths‚ the recent development of the knowledge economy in Berlin is quite different. Following German reunification in 1990‚ Berlin had to overcome its largely de-industrialized starting point and an outmoded institutional and structural context. Kujath views the case of Berlin as an example of regional breakthrough. This study can be contrasted to Lena Tsipouri’s research on European less favored regions (chapter 9). According to the concept of institutional thickness‚ lagging regions that have not yet developed deeply woven institutional structures should have a greater chance of development success since it is easier for them to adapt to new challenges. However‚ in a comparison of traditionally lagging European regions‚ Tsipouri notes that Ireland has experienced more recent success than seen in Greece or Portugal‚ attributing this at least in part to historically-developed educational structures that have enabled Ireland to develop a knowledge-based economy. However‚ it often proves difficult to develop a new regional identity and to generate synergy effects between institutional and technological development paths. This is the task facing especially those regions which have been very successful up to now‚ and which have achieved a high level of technological competence in the so-called mature industrial sectors. For these regions‚ the hitherto established institutions and the institutional thickness thus achieved can even become a problem in themselves‚ since training‚ research and funding facilities tend to stabilize the traditional patterns of industrial development. Against a background of intensified global competition‚ this problem of institutional inertia and restrictions deserves special attention. Research findings reveal interrelationships between technical and industrial development paths and regional institutions. This means that technical innovations usually need to be accompanied by institutional innovations. In some cases‚ institutional innovations – such as initiatives for establishing and facilitating regional innovation networks – may be necessary preconditions for further technical innovations. Within the context of globalization‚ national and regional actors in industry and politics are faced with the challenge of constantly reappraising the functional and operational principles of regional innovation systems and‚ if necessary‚ devising strategies for reforming the institutional and industrial order (cf. Cooke‚ 1998; Drache‚ 1996). Several chapters in the volume focus on the role economic actors play in restructuring regional paths. Gernot Grabher‚ building on his earlier works on
xvi
Gerhard Fuchs and Philip Shapira
regional lock-in‚ explores project ecologies as a strategy of economic actors in order to provide flexible economic structures (chapter 4). Similarly‚ in their study of Baden-Württemberg in southwest Germany‚ Gerhard Fuchs and Sandra Wassermann point out that dominant economic actors are capable of creating guiding models and thus are actively involved in shaping new paths of development (chapter 11). However‚ this argument takes a further twist in the chapter by Udo Staber‚ where he observes that entrepreneurs can and do contribute to regional lock-in (chapter 6). Staber argues that entrepreneurs only rarely leave established paths and‚ in so doing‚ limit the selfdevelopment of innovative structures. According to Staber‚ entrepreneurial activities thus have to be accompanied by efforts from political actors to ensure an adequate institutional framework is available to foster innovation and flexibility. Harald Bathelt and Jeff Boggs offer a structural approach to regional path dependency‚ rejecting the assumption that regional trajectories are to be considered as homogeneous paths (chapter 8). Instead‚ in their view‚ regional development paths consist of bundles of technological trajectories‚ some dominant‚ some operating on the region’s periphery. Sectoral or technological crisis as well as political ruptures may induce regional actors to re-bundle local capital and thus enable the occurrence of a regional breakthrough. Philip Shapira (chapter 10) further takes up the challenge of interpreting and attributing diverse development trends in his assessment of innovation challenges and strategies in catch-up regions. Drawing on the experience of the state of Georgia in the southeastern United States‚ Shapira finds that policy structures and local economic development organizations in some locations can evolve and become more innovative‚ but in other areas these institutions are not only well established but also locked-in to traditional strategies. This reinforces regional development inequities‚ even within a single state. At the same time‚ Shapira notes the role that leadership‚ institutional development‚ and sustained new policies can play in explaining how some previously noninnovative places have broken through. Despite their differences according to the focus (structure-actor; role of economy; role of policy)‚ all concepts and studies collected in this volume have in common the assumption that reflexive practice at the regional level is a crucial tool in (re)orientating regional development. Drawing on empirical information and theoretical interpretations‚ we suggest that there is an axis of regional development along which regional pre-structuring plays out in diverse ways. At one pole‚ we observe regions where new industries‚ economic activities and organizations emerge in successive‚ if not organic‚ ways due to the existence of strong predecessor industries and institutions. These predecessor structures may‚ through spinout and knowledge transfer‚ be influential in creating the supply base for new activities or‚ as they evolve‚ may generate
PREFACE
xvii
demand for new activities. Within this logic‚ the pre-existing industrial and institutional structure in a given region is of crucial importance. As this structure usually develops over long periods of time‚ a strong element of pathdependency comes in to play here. There is also a second pole on the axis of regional development‚ where agency assumes a crucial role over structure. In such circumstances‚ regional development strategies‚ institutional initiatives and other policies seek to promote new economic activities. The desired form of these activities may vary‚ although often policy makers have the final goal of creating a selfsustaining innovative regional industrial cluster. Again‚ policy support may address both supply and demand sides (Eisinger‚ 1988). This second logic is a counter-argument to pure path-dependency: even if the past development and the present economic structure of a given region have created unfavorable preconditions‚ efforts to realize a reorientation may be successful. In practice‚ we recognize that most regions are positioned somewhat between these two poles. Moreover‚ as regions placed more towards the second pole seek to move towards the first‚ those placed towards the first pole frequently recognize that specific kinds of agency interventions are necessary‚ for example to remove bottlenecks on growth or to sustain opportunities for new technology development (see Braczyk‚ Fuchs‚ and Wolf‚ 1999). Recognizing that each region is different‚ we suggest that these complexities of position and change call for a reflexive policy practices (see Schon‚ 1983; Benz et. al.‚ 1998) – through which policy action‚ while informed by wider theory and practice‚ is grounded in (yet in many cases must also challenge) the distinctiveness of regional conditions. Several of the chapters of this volume feature strong cases where such policy practice has led to desired results‚ with the lesson being not necessarily that other regions should pursue the same policy mechanisms but that regions require and should develop customized policies that reflect their own conditions and opportunities for development.
REFERENCES Amin‚ A.‚ Thrift‚ N.‚ Globalization‚ Institutions‚ and Regional Development in Europe. Oxford: Oxford University Press‚ 1994. Benz‚ A.‚ Fürst‚ D.‚ Kilper‚ H.‚ Rehfeld‚ D.‚ Regionalisierung. Theorie – Praxis – Perspektiven. Opladen: Leske‚ 1998. Braczyk‚ H.-J.‚ Fuchs‚ G.‚ Wolf‚ H.-G.‚ Multimedia and Regional Economic Restructuring. London: Routledge‚ 1999.
xviii
Gerhard Fuchs and Philip Shapira
Cooke‚ P.‚ “Introduction: Origins of the Concept.” In Regional Innovation System‚ H-J Braczyk‚ P. Cooke‚ and M. Heidenreich‚ eds. London: UCL Press‚ 1998: 2-25. Dosi‚ G.‚ Technological Paradigms and Technological Trajectories: A Suggested Interpretation of the Determinants and Directions of Technological Change. Research Policy 1982; 11: 147162. Drache‚ D.‚ Governance and Public Policy in a Global Economy. A Report on Jobs and Investment Strategies in Canada. Toronto‚ Canada: York University (unpublished)‚ 1996. Eisinger‚ P.K.‚ The Rise of the Entrepreneurial State. Madison: University of Wisconsin Press‚ 1988. Hughes‚ T.P.‚ “The Evolution of Large Technical Systems.” In The Social Construction of Technological Systems‚ W.E. Bijker‚ ed. Cambridge‚ MA: MIT Press‚ 1987: 51-82. Schon‚ D.A.‚ The Reflective Practitioner: How Professionals Think in Action. New York: Basic Books‚ 1983. Storper‚ M.‚ Regional Technology Coalitions: an Essential Dimension of National Technology Policy. Research Policy 1995; 24: 895-911.
Acknowledgments The editors wish to thank the Center for Technology Assessment in Stuttgart‚ Germany‚ for providing financial assistance to support the workshop on which this book is based as well for the preparation of the book itself. At the Center for Technology Assessment‚ Sandra Wassermann helped greatly in organizing the initial workshop and in the first initial preparatory steps for publication of this book. Andreas Koch provided additional assistance in preparing the book. Acknowledgements are also due to Monika Baumunk for reading the whole manuscript and correcting typing errors and to Lina Salkauskyte for collating‚ checking and initial indexing of the manuscript and communicating with the authors. At Georgia Tech in Atlanta‚ the editors appreciate the help of Jue Wang in further checking‚ editing‚ and final indexing.
Chapter 1 BEYOND PATH DEPENDENCY AND COMPETITIVE CONVERGENCE Institutional transfer from a discourse-analytical perspective
Christoph Scherrer
1.1 INTRODUCTION The theme of this book – “path dependency or regional breakthrough” – fits very nicely into a lively debate on the possibilities for institutional transfer. This debate was primarily motivated by East German and Eastern European transformation processes. However, the issue of institutional transfer, or “learning” from distant practices, has been raised not only with respect to the “East” but also for the institutions of the so-called “old” Federal Republic of Germany, most recently in the course of the “U.S. job miracle” discussion about labor market institutions. In both discourses, the concepts of path dependency and convergence through “natural selection” to “one best way” play a major role, albeit to differing degrees. In transformation discourse, the empirical observation that key political and economic institutions can evolve differently among countries and that these differences relate to the institutional traditions of the respective country has made the term path dependency increasingly attractive from an analytical point of view. In the discourse on competition between different forms of capitalism, the spectacular job growth under the Clinton administration and the deregulation of various product markets in Europe lend credence to the idea that the Anglo-Saxon brand of capitalism is the “best practice.” Path dependency and competitive convergence represent two poles in the spectrum of answers to whether “learning” from other countries is possible. In its purist form, the path dependency thesis implies that a domestic institutional transformation is affected only slightly by the perception of foreign institutions, and that learning – especially the attempt to implement what has been learned – can take place only within the framework of existing institutions. By contrast, the “one best way” thesis implies that the most efficient practice must be adopted under penalty of ruin.
2
Christoph Scherrer
Which one applies: divergence owing to path dependency or competitive convergence to “one best way”? This chapter1 shows that this question cannot be decided based on prevailing theories, particularly neoclassic economics and new institutional economics, primarily because they inadequately address the “how” of institutional transfer. Instead, the chapter outlines an explanatory model that enriches the key insights of new institutional economics (which can also be traced back to other theories) with the dimensions of discursive strategy and power. The inclusion of these dimensions raises the question of what enables societal actors to convince other relevant actors of the advantages of a foreign model and to initiate the imitation of this model. In the following, this chapter draws on Chantal Mouffe and Ernesto Laclau’s discourse analysis as well as by Gramscian power theory. To avoid misinterpretations, the chapter will not take on the entire spectrum of transnational learning, which besides institutions encompasses mainly ideas and individual political measures. Rather, the focal point lies in discussions on the adoption of foreign institutions. First, however, the next section critically considers the “one best way” and path dependency theses.
1.2 THE ONE BEST WAY THESIS In neoclassical economic theory, transnational competition leads to factor price harmonization and thus to a process of “one best way.” In the long term, only the most efficient production techniques will be maintained on the world market. This line of argumentation has also been applied to the development of economic policy institutions, such as private ownership of the means of production. Indeed, according to new institutional economics pioneer Douglass North, private property rights endured historically because they proved to be the most efficient (North and Thomas, 1973). Today, advocates of this position with a neoclassical or managerial mind-set find that, in the light of international competition, it is imperative for an individual country to facilitate not only the adoption of the most productive techniques but also to orientate its own economic policy institutions on the most efficient model internationally (Siebert, 1996; Ohmae, 1990; Womack et al., 1990; Bartlett and Ghoshal, 1998). A similar line of argumentation can be found in other theories, for example in the modernization theory (Zapf, 1991) and in Marxism (Brenner, 1999). However, this convergence thesis lacks unequivocal empirical evidence. Even between countries with a long tradition of capitalist industrialization, there are profound differences in institutional arrangements. These differences are fundamental for any discussion of national models. Proof of such differences is provided by innumerous empirical examples (Kitschelt et al.,
BEYOND PATH DEPENDENCY AND COMPETITIVE CONVERGENCE
3
1999; Crouch and Streeck, 1997; Hollingsworth et al., 1994; Porter, 1990). Even within the European common market, there has hitherto been scarcely a trace of convergence toward a single best-practice model as regards industrial relations, capital procurement, and education (Traxler, 1999; Cattero, 1999). The same may be said for Eastern Europe (Stark and Bruszt, 1998). Indeed, companies that operate globally or are subject to international competition still reveal national differences in management organization (see review by Eckardt et al., 1999; Lippert, 1999; Pauly and Reich, 1997). The lack of empirical evidence does not necessarily disprove the theory. The discrepancy between empirical data and theory may be explained by the fact that the processes of denationalization (Zürn, 1998) and globalization are still nascent. These processes have only made real headway in recent years. Nonetheless, many observers believe that the process of convergence has accelerated of late. Transnational mergers and capital market liberalization, which lead to similar financial evaluation criteria (“shareholder value”), mean that a greater push toward standardization may be expected in the future (Boyer, 1999:12 et seq.; Ertman, 1999; Cattero, 1999). There is much to substantiate a dependency between the scope and speed of convergence and the degree of market growth. Manufacturers are most often engaged in international competition and are therefore under enormous pressure to conform. Service sector companies are typically less affected by international competition. Thus, while having similar profitdriven efficiency criteria, service companies can deviate more from the “bestpractice” ideal without greatly risking their market position. Yet this insight weakens the postulate of a world-market-induced convergence of economic policy institutions. It raises the question in how far the firms’ pressure to conform is transferred to state institutions. The public choice approach, in line with this tradition of economic theory, sees competition for political office as the definitive transmission belt. Up to now, this approach has mostly been used to show how the creation of economically efficient institutions can be thwarted by interest groups in the electorate (socalled “principals”) and by autonomous politicians (so-called “agents”) (cf. Udehn, 1996:67).2 Without empirical evidence, it cannot be said in this theoretical context that the firms’ pressure to conform is taken up adequately by politics. The convergence thesis thus lacks the crucial political transmission belt to explain convergence. At the national level, where markets are most advanced, divergent corporate cultures may still be found even within the same sector (Dörrenbächer and Wortmann, 1993; Jürgens, 1992). In the light of this empirical phenomenon, neoclassical theory reveals its failings by focusing on price as the decisive factor in market success. Market success cannot be reduced to the ability to offer goods at low prices (Porter, 1990). There are many other contributing
4
Christoph Scherrer
factors, such as reliability, which leave ample room to maneuver for corporate strategies. But even when the price is the sole factor in a product’s market success, market price can be enforced by myriad combinations of input factors with diverging factor endowments. Moreover, neoclassical theory, based on an ideal of equilibrium, is little suited to explain innovations. If all companies were oriented on “best practice,” innovations would soon come to a standstill. On the one hand, there would be no competitive edge to serve as an incentive to innovation. On the other hand, the combinations of factors, that generate promising innovations, would be severely restricted due to the homogeneity of practices. For an innovative climate, it is necessary that a few firms stand apart from their competitors, with the result that they either will eventually fail on the market or will be unexpectedly successful (cf. Hung and Whittington, 1997:553; Bikhchandani et al., 1992).3 However, even if it is assumed that more sectors are subject to marketeconomic logic and that political institutions can quickly react to the demands of their economic subjects, this would not necessarily result in convergence. According to the “theory of comparative institutional advantages”, as elaborated by David Soskice and his research group at the Wissenschaftszentrum Berlin, economic policy institutions can specialize in a manner analogous to product specialization. This theory transposes Ricardo’s idea of mutual gains from world trade to the institutional level. Just as specialization emerges in traded goods, individual nations undergo a specialization in institutional structure, because the product strategy pursued in each instance requires its own institutional setting. For example, German companies focus on incremental innovation because the German labor law and financial system, among other factors, promotes a long-term planning horizon. Meanwhile, American companies use the institutional conditions in the U.S., such as deregulated labor relations and dynamic venture capital markets, to pursue more radical innovation strategies. Therefore, Germany’s adoption of elements of the American model can only be achieved at the cost of the complete dismantling of its models for industry and innovation (Hancké and Callaghan, 1999). Even where convergence is manifest, the impetus is not the competition mechanism. Uniformity can spring from generalizations in interpretative patterns and from similarities in micro-political strategy effects (Ortmann, 1995:285; for “coercive isomorphism,” “mimetic processes” and “normative pressures” see also DiMaggio and Powell, 1983). In the case of management practices, a series of factors could explain alignment tendencies; for example, inter-company management discourse at the level of trade shows, trade journals, and management consultancy agencies (cf. Strang and Soule, 1998). In
BEYOND PATH DEPENDENCY AND COMPETITIVE CONVERGENCE
5
other words, the thesis of competition-based convergence must prove its worth against alternative explanations.
1.3 PATH DEPENDENCY In economics, the path dependency concept stems from the attempt to explain the discrepancy between the theoretical assumption of efficient institutional development and the durability of inefficient institutions. The history of technology teaches us that suboptimal technologies can thrive for a relatively long time if they enjoy a head start. The typewriter keyboard is a notable example (Ortmann, 1995:255-261; David, 1985). For a theory based on a rationally calculating, benefit-maximizing individual, one may well ask how such rational decisions can engender something so suboptimal. The answer lies in positive network externalities of technologies. These kinds of externalities occur when a technology application’s utility increases with the number of users or consumers. In such a case, the circumstances of the first successful applications are of major significance. Path dependency is strengthened by: first, high start-up investments that lead to falling per-unit costs with increasing output; second, learning effects in technology application; and third, positive coordination and compatibility effects that proceed from the development of compatible technologies and standards (Arthur, 1994). From the new institutional economics perspective, similar mechanisms come into play at institutions as well. Institutions have high start-up costs; there are learning costs that emerge during organizational set-up and coordination costs that arise in the course of the mutual adaptation of formal and informal rules (Leipold, 1996:97). The path, once chosen, does not lead to a destination because every decision-making situation has alternatives; the number of alternatives, however, is limited by the path. Therefore, it may be rational to hold on to a suboptimal institution. Actors break away from the path only when efficiency losses are greater than the costs for creating a new, more efficient institution (North, 1992; Weinert, 1997:83; see Ruigrok and Tulder, 1995 on globalization strategies of corporations). In transformation research, this microeconomic reasoning has been rarely used for a theoretical foundation of path dependency (exceptions: Murrell, 1992 and 1995; Poznanski, 1996). Gerhard Lehmbruch (1994 and 1995), David Stark (1996) and others, who popularized this concept, share neither the rationality postulate of the new institutional economics nor the idea that the legacy of the past is the sole restricting power. They see the concept as an institutional resource for actors to combine and implement in different ways (see also Nielsen et al., 1995). Path dependency is therefore also characteristic of institutional transformation, because in these situations
6
Christoph Scherrer
actors also exploit the available institutional resources (Stark and Bruszt, 1998:83). In addition, Lehmbruch emphasize traditional interpretations of situations (1995:90). This distancing from new institutional economics is justified, but not for the reason proposed by Helmut Leipold. In his opinion, it makes little sense to understand path dependency through technological development, because in contrast to technologies, institutions are characterized by stagnating or even decreasing profits (1996:100). For Paul Pierson the opposite is true. He advances a number of reasons for why increasing returns to scale are applicable for “politics” (1997:24-36). Considerably more problematic is the new institutional economics assumption of objective efficiency criteria. If efficiency cannot be determined objectively for technology, as David Noble (1984) argued so convincingly (see also, Esser et al., 1997; Ortmann, 1995:260), then this is especially true for political institutions. Efficiency assessments are guided by interests and are dependent on the context. It must be remembered, as even path dependency advocate Paul Pierson emphasizes (1998:21-26), that not every path is characterized by selfintensifying sequences of events. In addition, some sequences of events can be identified that cause unintended reactions. Whereas in the former setting the further path is not inexorably fixed, in the latter the path ends. Accordingly, an observable path does not automatically continue into the future. Unsurprisingly, empirical evidence is ambiguous for the non-economic version of the path dependency concept. Jürgen Beyer and Jan Wilgohls analyzed David Stark’s thesis of path dependency for post-socialist countries. They drew the conclusion that the first free elections had a decisive impact on future privatization policy but, contrary to the path dependency thesis, subsequent changes in political power relations influenced the further course of privatization. In addition, their study showed that different privatization strategies were pursued in countries with similar transformations (Estonia and Lithuania), and that other countries, whose transformations were different, chose similar privatization strategies (East Germany, Estonia; Beyer and Wilgohls, 1998). This criticism indicates a central deficiency in the path dependency concept for societal institutions; namely, the problem of operationalization. There are various ideas of the relevant time frame and key events that determine a given path. In transformation research, some authors highlight the significance of pre-socialist history for current and future developments in Eastern Europe (Janos, 1994), while others see current developments predominately influenced by the recent socialist past (Crawford and Lijphart, 1997; Jowitt, 1992). For Lehmbruch (1998) and Stark (1996), decisions made during the system’s collapse determined the further development path, although the legacy
BEYOND PATH DEPENDENCY AND COMPETITIVE CONVERGENCE
7
of the past influenced these decisions (see also Wollmann, 1996; Nielsen et al., 1995). Within the “U.S.-job-miracle” debate, however, Thomas Ertmann finds 50 years to be too short a time span to define the further development path (specifically, his argument refers to the West German system of industrial relations in comparison to European processes of state formation; 1999). To summarize – in social evolution, every initial condition has a history. For this reason, the path dependency concept is faced with the problem of infinite regression. Yet problems of operationalization plague not only the diachronic but also the synchronic perspective. How can path-critical institutions be isolated from the multitude of institutions in modern complex societies? Key factors include the following: party apparatus institutions, cultural heritage, and informal relationships complementary to planned economy (cf. Bohle, 1999). Moreover, external factors are also path dependency elements; for example, the magnitude of accumulated foreign debt under state socialism (Bohle, 1999:18; cf. Pickel and True, 1999).
1.4 DISCOURSE-ANALYTICAL APPROACH This discussion of the shortcomings of the “one best way” and path dependency concepts should not be misunderstood as a denial of the formative powers of competition and institutional legacy. Obviously, these concepts alone can only partly explain the processes of institutional transfer. Several articles have been published that deal explicitly with the transnational diffusion of political concepts and institutions. They provide important insights into the structural prerequisites for transfer. Accordingly, political measures diffuse more readily than political institutions, and the speed of transfer is accelerated by the existence of international networks and epistemic communities. These empirical works also raise the problem of proving transfer, because not just transfers alone but also endogenous processes can result in similarlooking political measures or institutions (see overview by Stone, 1999). Power relations are rarely treated in these works, unless it is a matter of diffusion within hierarchically structured political institutions (cf. Kern, 2000). Studies on diffusion processes in the environmental field, for instance, are marked by a positive attitude toward innovations, thanks primarily to the so-called “California effect” (i.e. the diffusion of higher environmental standards starting in California). The “progressive” nature of these environmental policy innovations seems to make a critical analysis of the motives and implementation strategies of their protagonists unnecessary (cf. Kern, 2000; Biermann and Simonis, 1998). Because the focus was on environmental measures as objects of transfer, the diffusion of defensive tactics against en-
8
Christoph Scherrer
vironmental regulation among industrial lobbyists did not catch the researchers’ attention. Conversely, studies on the diffusion of concepts in the social policy field, which experienced cuts in many countries over the last decades, highlight far more frequently the power aspects of implementation strategies for these innovations (e.g. Peck, 1999; Plehwe and Walpen, 1999). However, little heed is paid in these works to structural factors; in particular, they neglect the question of compatibility between power-politically driven innovations and the existent institutional set-up. Neo-institutional works highlight this latter aspect (Kitschelt et al., 1999; Döhler, 1991). For a theoretical definition of both structural and actor-oriented instances of institutional transfer, the discourse theory, developed by Ernesto Laclau and Chantal Mouffe (1985) is most helpful (Scherrer, 1995). Their understanding of discourse is not limited – as, in general, most are – to spoken or written text, but is distinguished by an epistemological position. The substance of meaning is not determined by the essence of an object or practice, but rather by discursive articulation: “outside of any discursive context objects do not have being; they only have existence” (Laclau and Mouffe, 1987:85). Such a position theoretically emphasizes the suggestion by David Strang and others that mimetic models do not flow, but interpretations of these boundaries do (“Practices do not flow: Theorized models and careful framings do,” Strang and Soule, 1998; cf. Lillrank, 1995). Accordingly, Laclau and Mouffe reject the idea that societal reality is reducible to an inevitable part of an immanent law. Nevertheless, they do not rule out the existence of structures. If these were absent, then no coherent discourse would be possible because only indeterminacy would prevail (Laclau and Mouffe, 1985:112). Yet structures never achieve a completeness where all elements are defined, but rather are vulnerable to constant interruptions and shifts. Subjects, like structures, never attain a closed identity because this comes about only in relation to other identities. What results is the reciprocal subversion of subject and structure. The subject is the product of a shift in a structure, i.e. the impossibility of a structure to fully constitute itself. The structure results conversely from the impossibility of a subject to continually regenerate everything that is discursive (i.e. all verbalizations and actions, all non-verbalizations and non-actions) (ibid 107). The discourse-analytical assumption of mutual subversion of structure and subject offers a plausible approach to the analysis of institutional transfer. It gives access to how theorists of path dependency, such as David Stark, grasp the significance of institutions for individual and collective action. It also permits an understanding of what protagonists of the “one best way” thesis emphasize to be inevitable instances of competition. At the same time, this discourse-analytical assumption allows an actor-oriented approach. Constant structural shifts cause the creation of constant subjects; the latter are
BEYOND PATH DEPENDENCY AND COMPETITIVE CONVERGENCE
9
compelled through acts of identification to accept new identities and hence to “meaningfully” join the structures of their actions (Laclau, 1990:60-67). These subjects can recreate meanings in the imaginary realm (i.e. a realm illegible by structures) and give structures a “new sense” too, but they cannot as individuals or as subordinate collective actors change these structures voluntarily. As the enforcement of structures is decentralized and structures are not linked with each other “essentially,” they cannot be modified from a privileged position. Their existential conditions would first have to be undermined. The absence of a center does not rule out the existence of centers, of hierarchies among the structures. Centers of societal practices can exist only as long as a structure is not completely closed. In the case of closure, each element of the structure would possess only a relational identity with all other elements (Laclau, 1990:40). Specifically, this means that some practices or bundles of practices (institutions such as the wage or commodity relations) can structurally affect other practices. The extent of this influence rests first on the type of relation they have with the other practices and second on how far they themselves are embedded in society. Generally, the societal availability of practices is contingent on several factors, including: how expansive they are and how long they have endured how self-evident they have become how negative the probable consequences of their change are estimated to be which sanctions will be imposed if attempts at change are made whether actors are ready to defend these practices if the previous mechanisms for maintaining them fall short; what resources they can mobilize in comparison with actors urging change; and how they use these resources. As applied to the question of institutional transfer, these considerations entail searching (a) for temporarily fixed institutions, including their structural elements that either enable or restrict such a transfer, and (b) to a limited extent, for open situations where actors struggle for renewed closures and in so doing become involved in interpretational conflicts. First, regarding structures, one must ask which institutions create a competitive situation, recognized as such, that brings about a “best practice” by “natural selection.” Second, one must analyze which institutions generate institutional legacies, which in turn are also discursively recognized. In both cases, one would need to check how much the competition or the capacity to persist is also discursively grasped, and to what extent there are attempts to modify the institutions that cause these structural effects.
10
Christoph Scherrer
Considerations of the behavior of actors in “open situations” – elaborated to a high level of abstraction in Laclau and Mouffe – may be expanded with Gramscian power theory. On the one hand, this draws attention to nongovernmental collective actors particularly in transnational relations, in the so-called “civil society” (Cox, 1987). On the other hand, it interprets hegemony as a relational equation between collective societal actors, whose reproduction is ensured neither by the “dull compulsion of production relations” nor by coercion, but requires other, non-coercive strategies (Scherrer, 1999:16-33). From this perspective, coalitions must be formed, in addition to active or at least passive consensus, in order to enforce institutional transfer. The discursive strategies of institutional transfer, however, do not take place in a structure- and power-free realm. One must take into account, first, the power relations among the discourse participants and second, the abovementioned structural conditions of concrete institutional transfer.
1.4.1 Power Constellations Discourse on foreign institutions can develop from an “open situation” as regards a domestic institution (e.g. because it apparently does not fulfill the defined objective). Yet, discourse participants may still be attached to the previous political structures of asymmetrical resources and participatory options. In other words, former power positions influence the options for the discursive power of interpretation in relation to a foreign model. If the premise that the transfer process starts with interest groups or political “entrepreneurs” is correct, then their powers in the political process should be the most important factor for a successful model transfer. Research on modernization and democratization of nations (Merkel, 1997:11-15) as well as on the post-communist transformation processes in Eastern Europe (Offe, 1997:216) supports this thesis by emphasizing the role of elites. Hart concludes his international comparison of institutional adaptation to world market pressures with the observation that “movement occurred in statesocietal arrangements within the limits established by the underlying distribution of power among major societal groupings” (Hart, 1992:289). However, because elites will be neither immediately nor as a group convinced of the necessity for imitation, interest formation processes must also be taken into consideration. In the process of achieving understanding amidst divergent interests, actors such as political experts, who otherwise enjoy only limited powers, can play a larger role. This role will undoubtedly grow, as preliminary decision analyses are increasingly being outsourced to independent consultants, not just in the private sector (Micklethwait and Woolridge, 1996) but in policy formulation as well (for privatization, see Strange, 1996:135-
BEYOND PATH DEPENDENCY AND COMPETITIVE CONVERGENCE
11
146; for pension reform, see Blomert, 2001). Moreover, if there is a discrepancy between the promised and actual productivity of an institution, weaker actors will assume more power. Logically enough, the media will play a key role in disseminating proposals for institutional transfer, with their actors very likely pursuing their own interests in the process. The transfer of one country’s practices to another is further influenced by the balance of power between them. The weaker country is usually more willing to learn from the stronger one than vice versa. The stronger country’s apparent success is not the only thing that makes it more attractive; its key actors can use more resources and provide more incentives for others to adopt its institutions. These resources might include its power within international organizations, which might be based on geo-strategic motives (e.g. the United States versus postwar Germany) or on the fear that an opposition group in another country could attain a strategic advantage (for an extreme case – reunification – see Lehmbruch, 1994:29). Yet power alone is not enough. A highhanded use of power can breed resentment and resistance. Hegemony in the Gramscian sense of furthering one’s own interests by integrating those of other groups would be more conducive to model transfer.
1.4.2 Competition as Structure and Discourse Object A proposal for institutional transfer is more persuasive if it is described as a necessary measure for surviving a threatening competitive situation. This leads us back to the “one best way” argument. In contrast to this thesis, however, success of transfer seems to depend less on whether competition really exists but on how much a crisis is seen as the outcome of a competitive situation. If competition is not recognized as such, then it cannot be held up as a basis for institutional change. Naturally, an actual but unrecognized competition mechanism can still have an impact; for example, it can lead to military defeat, to bankruptcy, or to high unemployment. However, whether the defeat is retrospectively associated with the competition mechanism is an open question. Even if a competitive situation is recognized, imitating “best practice” is not the sole option for action. Recognition can lead to efforts to “outrun” the competition or, if catching up seems futile, to discontinue further efforts. Interpreting a crisis as the outcome of a competitive situation will seem more plausible, if supported by everyday experience: as an extreme example by war; or in times of peaceful economic competition by experiences in the consumer world and at the workplace. American industrial trade unions, for instance, long ignored European and Japanese competition, dismissing it as a “foreign competition hoax” dur-
12
Christoph Scherrer
ing the 1959 steel strike, which gave foreign steel producers a breakthrough on the U.S. market. Later, when steel industrialists concluded that it was not possible to catch up with the new competition and made cuts in the investment budget accordingly, unions successfully advocated protectionist policies, using them to obtain wage increases that were out of proportion to productivity gains (Scherrer, 1992:171-182). In this case, there was an “objective” competitive relationship. The institutions of private ownership, trade, the General Agreement on Tariffs and Trade (GATT), and price-sensitive purchasing decisions pitted steel producers against each other in international competition. Yet this competition was not immediately recognized by those involved; once it was generally acknowledged in discourse, one of the institutions held responsible for the situation, the GATT, could, through political power, be circumvented or made ineffective by “voluntary” export restrictions. Although competition could not be fully abolished in this manner (i.e. its effects were manifest in the investment behavior of U.S. firms and, years later, in a major steel crisis), its impact was limited.
1.4.3 Institutional Compatibility Foreign practices are more likely to be adopted if they are compatible with existent values and institutional arrangements. This is the most common argument in the literature for the dissemination of ideas and policies and is most akin to the path dependency argument. It is justified primarily by “interaction requirements” (Scharpf, 1978:363). Not one type of institution creates a system, “but the simultaneous existence and the pattern of interaction of a series of institutions” (Niosi et al., 1993:218). The transfer of a model representing only parts of an entire institutional configuration is therefore faced with the problem of institutional coherence. As industrial geographers Storper and Salais so forcefully argued, “the strength of any successful real world of production is precisely the way in which it is chiseled out of conventions which function together coherently and are made possible by conventions of identity and participation: these elements cannot be mixed and matched à la carte” (Storper and Salais, 1997:172). Furthermore, the adoption of foreign institutions is structurally limited by procedural knowledge. Changing procedural knowledge is difficult and can only be achieved with the passage of time, because such knowledge is implicit. People learn rules without a conscious knowledge of them, and these rules are stored as procedural memory. Procedural knowledge is also rooted in identities that are given by the existent categories defined by the societal division of labor (Kogut, 1997:358; for transformation research in-
BEYOND PATH DEPENDENCY AND COMPETITIVE CONVERGENCE
13
sights on this topic, see Wiesenthal, 1997). Finally, David Strang and John W. Meyer point out that culturally biased objectives, if dissimilar, can hinder institutional transfer (Strang and Meyer, 1993:490-492). The more familiar the model institutions are, the less affected the power relations between the societal actors will be and hence the lower the resistance. If many of the existing institutions need to be changed, then the inadequately fulfilled interaction requirements with other institutions during the implementation process will increase the risk that the initial innovations will fall short of the anticipated efficiency gains or other advantages. This disappointment can cast doubt on further implementation (see above). Institutional restraints on adopting foreign practices can be illustrated by an example from Germany. Despite its great interest in Henry Ford’s production techniques and labor relations, pre-war Germany was not yet ready for Fordism. The introduction of American production methods was limited by the German industry cartel, which enabled small producers to stay in business (Berghahn, 1986:22). The adoption of Taylorist concepts, many of which were indigenous to Germany, was further impeded by a societal order with institutionalized skill identities (the Meister, the Facharbeiter) defended by political actors (Kogut, 1997:360). In 1996, the boards of directors of several German firms supported decentralized wage bargaining in line with the American model. In addition to concerns about the strength of union opposition, other large industrial employers were ambivalent about this demand. As Kathleen Thelen has pointed out, they feared that if works councils were actually to take on more of the bargaining responsibilities traditionally reserved for unions, this would probably undermine the foundations, on which the management’s constructive relations with their works councils were based. This relationship depended on the works councils’ inability to negotiate. Indeed, decentralized bargaining could even open possibilities for plant labor representatives to use works councils’ rather significant legal rights to extract concessions from management over wages (Thelen, 1997). Institutional or mentality restraints, however, can be overcome in the course of time. Either the institutional setting adapts to the new practices, or these new practices are adapted to the old institutions. For Berghahn, who has studied the postwar transformation of West Germany’s industrial structure, the resistance to “Americanization” subsided when a new generation assumed leadership positions (Berghahn, 1986:11). The younger generation openly accepted American methods, primarily because they apparently worked. Even if foreign practices are adopted, the copy is not identical with the original. For example, when Taylorism was finally introduced in Germany, it did not lead to a deskilling of jobs or to impose crude sanctions to the extent
14
Christoph Scherrer
seen elsewhere, because the belief in sustaining the role of the skilled laborer in the production process was retained (Kogut, 1997:362). Besides the problem of adopting all supporting institutions of a foreign practice simultaneously, a principle obstacle is that a prototype is not a fixed template. There is usually ample space for disagreement about the central features of the model to be copied. For instance, many firms claim to have installed Japanese production methods. However, most of them have only adopted a fraction of what makes the Toyota system “best practice,” and usually have adapted this part to fit in the ambient practices. In these cases, transfer may lead to institutional innovations (Jürgens, 1993).
1.5 BEYOND PATH DEPENDENCY AND “ONE BEST WAY” The result of the discussion is that neither the thesis of competition-induced “one best way” nor path dependency can adequately explain whether transnational institutional transfer is possible or not. The thesis by which, under penalty of ruin, the most efficient practice has to be adopted has an unstable foundation both empirically and theoretically. Empirically, no clear trend toward convergence is discernible, price alone is not the decisive factor in competition (identical prices could conceal very different combinations of input factors or institutional settings), and many economic activities are not subject to direct competition. According to the “theory of comparative institutional advantages,” it is precisely competitive conditions that form the groundwork for the specialization of economic policy institutions. There is no clear-cut empirical evidence for the path dependency thesis, which sees the legacy of existing institutions to be a restraining factor in the possibilities for institutional transfer. In addition, the thesis can be difficult to operationalize. There are various contrasting ideas about the relevant time span and the major events that shape the respective path. In summary, the spectrum between the poles of “one best way” and path dependency leaves room for a multiplicity of institutional combinations. To determine their concrete shape, I have chosen a discourse-analytical approach which, without having to deny the power of competition and institutional legacy, increases an awareness of the political conflicts surrounding institutional transfer. This approach cannot provide any simple explanations, however, as it deals with the “radical contingency” of all relations between societal interests, identities, and positions. A good starting point for a prognosis, based on the above-mentioned considerations, would be the relative power relations between advocates and
BEYOND PATH DEPENDENCY AND COMPETITIVE CONVERGENCE
15
opponents of institutional transfer. These relations, defined by the availability of economic, political, and media resources, represent just one approach, however. The proposal to copy a foreign model can achieve a much greater response, if it offers a solution to a problem estimated to be urgent by the majority of people. The work of persuasion can be facilitated by the following factors: prevailing consciousness of strong competitive pressure; extensive compatibility of the envisaged policies and institutions with existent values and institutional structures; a powerful and interested model country or international organizations; and a model that proves itself according to the expectations of its supporters.
NOTES 1.
This chapter emerged in the context of my work in the DFG-funded research project “From the Chandlerian Business Model to Wintelism” in the Work Regulation Division of the WZB in Berlin. For their helpful suggestions, I would like to thank Reinhard Blomert, Bernhard Ebbinghaus, Ulrich Jürgens, Kristine Kern, Thomas Sablowski and the participants in Michael Schumann’s research colloquium at the Sociological Seminar of the Georg-August-Universität, Göttingen, especially my commentator Ulrich Voskamp.
2.
Among the Public Choice authors, Mancur Olson (1982) has focused most on competition between individual polities. For Olsen, “distribution coalitions” prevent political agreement as to best practice, which leads to the ruin of the nation in question. In his eyes, “learning” does not take place.
3.
This argumentation could be refuted by the fact that it is based on a static version of “best practice.” From a dynamic viewpoint, “best practice” will continually repeat itself – innovations will be taken up by everyone else in turn. However, this viewpoint implies that the institutional transformation does not strictly orientate itself on the best model but can be influenced by other factors. If so, there would be far less pressure to adopt the “best practice” and more room for a variety of practices.
REFERENCES Arthur, W.B., Increasing Returns and Path Dependence in the Economy. Ann Arbor: University of Michigan Press, 1994. Bartlett, C., Ghoshal, S., Managing across Borders: The Transnational Solution. London: Cambridge, MA: Harvard Business School Press, 1998. Berghahn, V., The Americanization of West German Industry, 1945-1973. Cambridge, MA: Cambridge University Press, 1986.
16
Christoph Scherrer
Beyer, J., Wielgohs, J., Path-Dependency Approaches and National Differences in postSocialist Institution Building: The Case of Large Privatization Policies, Paper presented at International Conference on Socio-Economics, July 13-16, 1998, Vienna. Biermann, F., Simonis, U.E., Institutionelles Lernen in der Weltumweltpolitik. Wissenschaftszentrum Berlin TAU Papers, 1998, FS II: 98-404. Bikhchandani, S., Hirshleifer, D., Welch, I., A Theory of Fads, Fashion, Custom, and Cultural Change as Informational Cascades. Journal of Political Economy 1992; 100: 992-1026. Blomert, R., Herbstliche Finanzblüten. Neue Rundschau 2001; 1:23-48. Bohle, D., Der Pfad in die Abhängigkeit? Eine kritische Bewertung institutionalistischer Beiträge in der Transformationsdebatte. Wissenschaftszentrum Berlin für Sozialforschung Discussion Paper 1999; FS I: 99-103. Boyer, R., “The Diversity and Future of Capitalisms: A ‘Regulationist’ Analysis.” In Capitalism in Evolution, G. Hodgson, N. Yokokawa, eds. Aldershot, UK: Edward Elgar, 1999. Brenner, R., Turbulence in the World Economy. London: Verso, 1999. Cattero, B., “Jenseits von Konvergenz und Pfadabhängigkeit – Über die Europäisierung der industriellen Beziehungen in Zeiten der Globalisierung.” In Ökonomische und soziale Herausforderungen am Ende des zwanzigsten Jahrhunderts, G. Schmidt, R. Trinczek, eds. BadenBaden: Nomos, 1999. Cox, R. W., Power, Production, and World Order. New York: Columbia University Press, 1987. Crawford, B., Lijphart, A., “Old Legacies, New Institutions: Explaining Political and Economic Trajectories in Post-Communist Regimes.” In Liberalization and Leninist Legacies: Comparative Perspectives on Democratic Transition, B. Crawford, A. Lijphart, eds. Berkeley: University of California Press, 1997. Crouch, C., Streeck, W. eds., Political Economy of Modern Capitalism: Mapping Convergence and Diversity. London: Sage, 1997. David, P., Clio and the Economics of QWERTY. American Economic Review 1985; 75:332337. DiMaggio, P.J., Powell, W.W., The Iron Cage Revisited: Institutional Isomorphism and Collective Rationality in Organizational Fields. American Sociological Review 1983; 48:147-160. Döhler, M., “Policy Networks, Opportunity Structures and Neo-Conservative Reform Strategies in Health Policy.” In Empirical Evidence and Theoretical Considerations, B. Marin, R. Mayntz, eds. Frankfurt am Main: Campus-Verlag, 1991. Dörrenbächer, C., Wortmann, M., Unternehmensstrategien auf dem europäischen Markt für Hygienepapierwaren. Informationen über Multinationale Konzerne 1993; 3:30-35. Eckardt, A., Köhler, H.-D., Pries, L. eds., Global Players in lokalen Bindungen: Unternehmensglobalisierung in soziologischer Perspektive. Berlin: edition Sigma, 1999.
BEYOND PATH DEPENDENCY AND COMPETITIVE CONVERGENCE
17
Ertman, Th., “Pfadabhängige Entwicklung und systemimmanente Zwänge: Sind die Möglichkeiten des Lernens begrenzt?” In Jobwunder USA – Modell für Deutschland?, S. Lang, M. Mayer, Ch. Scherrer, eds. Münster: Verlag Westfälisches Dampfboot, 1999. Esser, J., Fleischmann, G., Heimer, T. eds., Soziale Schließung im Prozeß der Technologieentwicklung: Leitbild, Paradigma, Standard. Frankfurt am Main: Campus Verlag, 1997. Hancké, B., Callaghan, H., “Systemwettbewerb oder -komplementarität? Deutsche und amerikanische Institutionen und Innovationsstrategien im Globalisierungszeitalter.” In Jobwunder USA – Modell für Deutschland?, S. Lang, M. Mayer, Ch. Scherrer, eds. Münster: Verlag Westfälisches Dampfboot, 1999. Hart, J.A., Rival Capitalists: International Competitiveness in the United States, Japan, and Western Europe. Ithaca: Cornell University Press, 1992. Hollingsworth, J.R., Schmitter, Ph.C., Streeck, W. eds., Governing Capitalist Economies: Performance and Control of Economic Sectors. New York: Oxford University Press, 1994. Hung, S., Whittington, R., Strategies and Institutions: A Pluralistic Account of Strategies in the Taiwanese Computer Industry. Organization Studies 1997; 4:551-575. Janos, A.C., Continuity and Change in Eastern Europe: Strategies of Post-Communist Politics. East European Politics and Societies 1994; 8:1-31. Jowitt, K., “The Leninist Legacy.” In New World Disorder: The Leninist Extinction, K. Jowitt, ed. Berkeley/London: University of California Press, 1992. Jürgens, U., “Internationalization Strategies of Japanese and German Automobile Companies.” In New Impact on Industrial Relations: Internationalization and Changing Production Strategies, S. Tokunaga, N. Altmann, and H. Demes, eds. München: München: Ludicium, 1992. Jürgens, U., “Teams als universelles Konzept? Einige vergleichende Anmerkungen zur Teamwork-Debatte in Japan, Deutschland und den USA.” In Jenseits des Sozialpakts: Neue Unternehmensstrategien, Gewerkschaften und Arbeitskämpfe in den USA, B. Lüthje, Ch. Scherrer, eds. Münster: Westfälisches Dampfboot, 1993. Kern, K., Die Diffusion von Politikinnovationen: Umweltpolitische Innovationen im Mehrebenensystem der USA. Opladen: Leske + Budrich, 2000. Kitschelt, H. et al., “Convergence and Divergence in Advanced Capitalist Democracies.” In Continuity and Change in Contemporary Capitalism, H. Kitschelt, P. Lange, G. Marks, and J.D. Stephens, eds. Cambridge: Cambridge University Press, 1999. Kogut, B., “Identity, Procedural Knowledge, and Institutions: Functional and Historical Explanations for Institutional Change.” In Ökonomische Leistungsfähigkeit und institutionelle Innovation: Das deutsche Produktions- und Politikregime im globalen Wettbewerb, F. Naschold et al., eds. Berlin: Cambridge University Press, 1997. Laclau, E., Mouffe, Ch., Hegemonie und radikale Demokratie. Wien: Passagen Verlag, 1991 (Original: Hegemony and Socialist Strategy: Towards a Radical Democratic Politics. London: 1985).
18
Christoph Scherrer
Laclau, E., Mouffe, Ch., Post-Marxism without Apologies. New Left Review 1987; 166:79106. Laclau, E., New Reflections on the Revolution of our Time. London: Verso, 1990. Lehmbruch, G., “Zwischen Institutionentransfer und Eigendynamik: Sektorale Transformationspfade und ihre Bestimmungsgründe.” In Transformationspfade in Ostdeutschland: Beiträge zur sektoralen Vereinigungspolitik, R. Czada, G. Lehmbruch, eds. Frankfurt am Main/New York: Campus, 1998. Lehmbruch, G., Die Rolle der Spitzenverbände im Transformationsprozeß: Eine neoinstitutionalistische Perspektive. Berliner Debatte INITIAL 1995; 6:89-105. Lehmbruch, G., Institutionen, Interessen und sektorale Variationen in der Transformationsdynamik der politischen Ökonomie Ostdeutschlands. Journal für Sozialforschung 1994; 34:2144. Leipold, H., “Zur Pfadabhängigkeit der institutionellen Entwicklung: Erklärungsansätze des Wandels von Ordnungen.” In Entstehung und Wettbewerb von Systemen, D. Cassel, ed. Berlin: Duncker & Humblot, 1996. Lillrank, P., The Transfer of Management Innovations from Japan. Organisation Studies 1995; 16:971-89. Lippert, I., Zwischen Pfadabhängigkeit und radikalem Wandel – Neuordnung von Prozeßketten im internationalem Maschinenbau. Berlin: edition sigma, 1999. Merkel, W., “Die Rolle von Eliten und Massen beim Übergang von autokratischen zu demokratischen Herrschaftssystemen.” In Einheit und Differenz: Die Transformation Ostdeutschlands in vergleichender Perspektive, J. Wielgohs, H. Wiesenthal, eds. Berlin: Berliner Debatte Wissenschaftsverlag, 1997. Micklethwait, J., Wooldridge, A., The Witch Doctors. New York: Random, 1996. Murrell, P., Evolutionary and Radical Approaches to Economic Reform. Economies of Planning 1992; 25:79-95. Murrell, P., The Transition According to Cambridge, Mass. Journal of Economic Literature 1995; 33:164-178. Nielsen, K., Jessop, B., Hausner, J., “Introduction.” In Strategic Choice and Path-Dependency in Post-Socialism: Institutional Dynamics in the Transformation Process, J. Hausner, B. Jessop, K. Nielsen, eds. Aldershot: Edward Elgar, 1995. Niosi, J. et al., National Systems of Innovation: In Search of a Workable Concept. Technology in Society 1993; 15:207-227. Noble, D., Forces of Production: A Social History of Industrial Automation. New York: Alfred A. Knopf, 1984. North, D.C., Institutionen, institutioneller Wandel und Wirtschaftsleistung. Tübingen: JCB Mohr, 1992.
BEYOND PATH DEPENDENCY AND COMPETITIVE CONVERGENCE
19
North, D.C., Thomas, R. P., The Rise of the Western World: A New Economic History. Cambridge: Cambridge University Press, 1973. Offe, C., “Die politisch-kulturelle Innenseite der Konsolidierung: Eine Anmerkung über Besonderheiten der postkommunistischen Transformation.” In Einheit und Differenz: Die Transformation Ostdeutschlands in vergleichender Perspektive, J. Wielgohs, H. Wiesenthal , eds. Berlin: Berliner Debatte Wissenschaftsverlag, 1997. Ohmae, K., The Borderless World. New York: Harper Collins Publishers, 1990. Olson, M., The Rise and Decline of Nations: Economic Growth, Stagflation and Social Rigidities. New Haven, Conn.: Yale University Press, 1982. Ortmann, G., Formen der Produktion: Organisation und Rekursivität. Wiesbaden: Westdeutscher Verlag, 1995. Pauly, L.W., Reich, S., National Structures and Multinational Corporate Behavior: Enduring Differences in the Age of Globalization. International Organization 1997; 51:1-30. Peck, J., Workfare States. New York: Guilford, 1999. Pickel, A., True, J., Global Forces, Transnational Linkages and Local Actors: Towards a New Political Economy of Post-Socialist Transformation. Paper Prepared for Presentation at the Annual Conference on Socio-Economics, University of Wisconsin-Madison, 1999. Pierson, P., Increasing Returns, Path Dependence and the Study of Politics. Jean Monnet Chair Papers 44, The Robert Schuman Centre at the European University Institute, 1997. Pierson, P., Not Just What, But When: Issues of Timing and Sequence in Comparative Politics, prepared for presentation at the American Political Science Association Meeting. Boston: 1998. Plehwe, D., Walpen, B., Wissenschaftliche und wissenschaftspolitische Produktionsweisen im Neoliberalismus: Beiträge der Mont Pelerin Society und marktradikaler Think Tanks zur Hegemoniegewinnung und –erhaltung. Prokla-Zeitschrift für kritische Sozialwissenschaft 1999; 29:203-236. Porter, M.E., The Competitive Advantage of Nations. New York: The Free Press, 1990. Poznanski, K. Z., Poland’s Protracted Transition: Institutional Change and Economic Growth 1970-1994. Cambridge: Cambridge University Press, 1996. Ruigrok, W., van Tulder, R., The Logic of International Restructuring, London: Routledge, 1995. Scharpf, F.W., “Interorganizational Policy Studies: Issues, Concepts and Perspectives.” In Interorganizational Policy Making, K. Hanf, F. W. Scharpf, eds. London: Sage, 1978. Scherrer, Ch., Eine diskursanalytische Kritik der Regulationstheorie. Prokla 1995; 25:457-482. Scherrer, Ch., Globalisierung wider Willen? Die Durchsetzung liberaler Außenwirtschaftspolitik in den USA. Berlin: Sigma Verlag, 1999.
20
Christoph Scherrer
Scherrer, Ch., Im Bann des Fordismus: Der Konkurrenzkampf der Auto- und Stahlindustrie in den USA. Berlin: edition sigma, 1992. Siebert, H., Labor Market Rigidities: At the Root of Unemployment in Europe. Journal of Economic Perspectives 1996; 11:37-54. Stark, D., Recombinant Property in East European Capitalism. American Journal of Sociology 1996; 101:993-1027. Stark, D., Bruszt, L., Postsocialist Pathways: Transforming Politics and Property in East Central Europe. Cambridge: Cambridge University Press, 1998. Stone, D., Learning Lessons and Transferring Policy Across Time, Space and Disciplines. Politics 1999; 19:51-59. Storper, M., Salais, R., Worlds of Production: The Action Frameworks of the Economy. Cambridge, MA: Harvard University Press, 1997. Strang, D., Meyer, J.W., Institutional Conditions for Diffusion. Theory and Society 1993; 22:487-511. Strang, D., Soule, S. A., Diffusion in Organizations and Social Movements: From Hybrid Corn to Poison Pills. Annual Review of Sociology 1998; 24:265-291. Strange, S., The Retreat of the State: The Diffusion of Power in the World Economy. Cambridge: Cambridge University Press, 1996. Thelen, K., Why German Employers Cannot Bring Themselves to Abandon the German Model. Manuscript, 1997. Traxler, F., “Wirtschaftliche Internationalisierung und die (Dis)Organisierung der Arbeitsbeziehungen.” In Ökonomische und soziale Herausforderungen am Ende des zwanzigsten Jahrhunderts, G. Schmidt, R. Trinczek, eds. Baden-Baden: Nomos, 1999. Udehn, L., The Limits of Public Choice: A Sociological Critique of the Economic Theory of Politics. London: Routledge, 1996. Weinert, R., “Institutionenwandel und Gesellschaftstheorie: Modernisierung, Differenzierung und Neuer Ökonomischer Institutionalismus.” In Institutionenwandel, G. Göhler, ed. Opladen: Westdeutscher Verlag, 1997. Wiesenthal, H., “Probleme der Tramsformationssteuerung – Eine Perspektive politikwissenschaftlicher Forschung.” In Einheit und Differenz: Die Transformation Ostdeutschlands in vergleichender Perspektive, J. Wielgohs, H. Wiesenthal, eds. Berlin: Berliner Debatte Wissenschaftsverlag, 1997. Wollmann, H., “Institutionenbildung in Ostdeutschland: Rezeption, Eigenentwicklung oder Innovation?” In Institutionenbildung in Ostdeutschland, A. Eisen, H. Wollmann, eds. Opladen: Leske + Budrich, 1996. Womack, J.P., Jones, D.T., Roos, D., The Machine that Changed the World: MIT-Study on the Future of the Automobile. New York & Oxford: Rawson Macmillan, 1990.
BEYOND PATH DEPENDENCY AND COMPETITIVE CONVERGENCE
21
Zapf, W., “Modernisierung und Modernisierungstheorien.” In Die Modernisierung moderner Gesellschaften: Verhandlungen des 25. Deutschen Soziologentages in Frankfurt am Main, W. Zapf, ed. Frankfurt am Main: Campus-Verlag, 1991. Zürn, M., Regieren jenseits des Nationalstaates: Globalisierung und Denationalisierung als Chance. Frankfurt am Main: Suhrkamp, 1998.
Chapter 2 TACIT KNOWLEDGE, PATH DEPENDENCY AND LOCAL TRAJECTORIES OF GROWTH
Meric S. Gertler
2.1 INTRODUCTION It is now widely agreed that the transition to a knowledge-based or learning economy has elevated the status and importance of regions in the world economy. In particular, regions are now acknowledged as the principal sites of innovation and innovative production, at a time when the competitive success of firms, industries, and nations relies increasingly on the creative use of knowledge and ideas. It is also widely understood that regions are highly uneven in their ability to support learning and knowledge-based production: that is, we are witnessing the emergence of an increasingly uneven geography of innovation and production, both within and between nations. As this geography of innovative activity continues to evolve, there are strong tendencies for winners to keep winning, and losers to keep losing, exacerbating already established disparities in local economic opportunity. The prospects for overcoming this process of growing economic polarization depend on the ability of regional economies to “change paths” – that is, to alter the nature and direction of their economic trajectories over time. However, what are the prospects of achieving this kind of fundamental reorientation? On the one hand, the history of capitalist economic development has been shown to produce an “inconstant geography” – epochal changes in the macro-economy repeatedly generate major ruptures, reversals and sea changes in the geography of capitalism, in which previously peripheral and disadvantaged regions rise up to challenge the established order (Storper and Walker, 1989). On the other hand, recent ideas from the field of evolutionary economics seem to point to another conclusion. The core argument emerging from this literature emphasizes that economic systems – whether at the level of the nation, the region, or the firm – have a marked tendency to change slowly.
24
Meric S. Gertler
Moreover, the direction of this change is strongly shaped by past experiences, decisions, practices, relationships and accidents of history. The central argument of this chapter is that – apart from epochal shifts regional economies do indeed change slowly, in path-dependent ways. Nevertheless, the origins or foundations of this path dependency are still not well understood. Indeed, it is argued that considerable confusion has arisen around this question, in part because of a lack of conceptual clarity around the notion of tacit knowledge and its role in generating the uneven geography of innovation referred to above. So long as we labor under these misconceptions, we will fail to achieve a deeper understanding of the sources of regional path dependency and the prospects for changing long-established development trajectories. The following section provides a brief overview of the core ideas underlying the evolutionary economics of path dependency, emphasizing the role of institutions in creating regional and national path dependency. Following this, the chapter examines the possible or probable pathways to regional economic change and the insights arising from several different strands of literature concerning the feasibility of effecting such change. In the process, the somewhat slippery idea of tacit knowledge is subject to systematic scrutiny. Finally, there is a fuller elaboration on the institutional impediments to fundamental change at the regional level, owing to the rooted trajectories of regional growth and decline.
2.2 THE EVOLUTION OF NATIONAL AND REGIONAL ECONOMIES: KEY CONCEPTS Evolutionary economics has become a vibrant field since the publication of Nelson and Winter’s (1982) seminal work. A few key concepts have been especially influential to economic geographers’ study of production systems (Barnes, 1997). First, economic systems change over time, but they do so in ways that are to some extent shaped and constrained by past decisions, random events, and accidents of history. Current decisions and events are not determined by past ones, but they are conditioned by them. As a result of past events and choices, certain choices today are easier to pursue, others less so. This is the key idea of path dependency. Walker captures the idea succinctly: One of the most exciting ideas in contemporary economic geography is that industrial history is literally embodied in the present. That is, choices made in the past – technologies embodied in machinery and product design, firm assets gained as patents or specific competencies, or labor skills
TACIT KNOWLEDGE, PATH DEPENDENCY, AND LOCAL GROWTH
25
acquired through learning – influence subsequent choices of methods, designs, and practices. (2000:126)
Walker’s Marxist interpretation of the concept of path dependency is remarkably microeconomic in focus, emphasizing technology, machinery, products, and competencies of the firm and the skills of workers. Hence, the past may be embodied within material objects such as machinery, buildings, and physical infrastructure, or through the experiences of individuals (alone or in groups). However, an equally – if not more – important part of this past is also embodied in institutions – social structures that shape the attitudes, norms, rules, expectations, and practices of individuals and firms through formal or informal means of regulation – meaning that path dependency has a strong social dimension. In essence then, what we might describe as national or regional “production cultures” or “embeddedness” can be thought of as a real and significant component of “industrial history...literally embodied in the present.” These cultures become a part of the historical baggage (sometimes useful, sometimes a liability) associated with particular regions. Moreover, they evolve within a larger macro framework of institutional architecture that strongly shapes incentives, decisions, and practices of individual economic actors (David, 1994; Zysman, 1994; Hodgson, 1997; Hall and Soskice, 2001). We shall explore the consequences of this key idea for our understanding of tacit knowledge and regional growth dynamics below. There is another potent idea within the evolutionary approach, closely associated with the path dependency notion – increasing returns – which has been appropriated from the work of economists (Young, 1928; Myrdal, 1957) working well before Nelson and Winter formulated their influential ideas. It refers to the process in which, once a particular economic change occurs, it becomes self-reinforcing. A particular technological design, once adopted by a critical mass of early users, becomes a standard. After this happens, the market for this design will expand even further. Initial growth begets further growth. Moreover, even though a particular technology has become a standard in its field, this will not necessarily indicate that it is unequivocally superior to available alternatives. Its dominance may instead be because it was the first viable technology in the marketplace, that many supplying businesses, complementary technologies and institutions, plus a large community of developers and users, have been created to support its use. Once this unfolds to the point where perfectly viable – if not superior – alternatives cannot easily be adopted, a situation of “lock-in” is said to have been reached (David, 1985; Arthur, 1989). The twin concepts of path dependency and increasing returns have obvious relevance in understanding the historical paths taken by production
26
Meric S. Gertler
regions. Once a region establishes itself as an early success in a particular set of production activities, its chances for continued growth are very good indeed. While this may be to some extent reducible to the success of dominant “lead” firms in the region (e.g. Microsoft in Seattle), the interesting aspects of this process have more to do with the collective processes and forces at work: local social and economic institutions and culture. Similarly, as cases such as Germany’s Ruhr Valley or Route 128 in Massachusetts suggest, ailing places may also be difficult to turn around for the same reasons. Once a path-dependent trajectory of decline becomes firmly established, institutional and cultural lock-in will make deviation from this path a serious challenge. In the following section, we consider the options for altering regional economic trajectories, and the prospects for success.
2.3 PROSPECTS FOR OVERCOMING REGIONAL PATH DEPENDENCY THROUGH KNOWLEDGE TRANSFER The perspective offered above is undoubtedly too pessimistic for policy makers concerned with turning around the fortunes of lagging or less favored regions. In recent years, there has been considerable interest in exploiting learning opportunities to assist such regions. The goal is to change the behavior of local economic actors and firms by transferring knowledge and practices from more successful firms, regions or nations to less successful ones (Gertler, 2001 and 2002). This kind of knowledge transfer might be effected from within (in which managers and workers would seek to emulate “best practice” methods imported from extra-regional locations), or through explicit intervention of external actors. In the latter scenario, inbound foreign direct investment by multinational firms would create opportunities for local actors to learn new practices – either through simple demonstration effects, or because of incoming firms imposing new standards for quality and performance on their local suppliers. Alternatively, knowledge transfer may occur because of local firms setting up branch operations abroad, and then bringing home the lessons they might have learned in order to transform their local operations. How feasible are such learning-based strategies for altering regional development pathways? I shall argue below that this approach is considerably more difficult to implement than is widely appreciated. The reasons for this difficulty stem from the tacit component of economic knowledge, which is always present but is especially prominent when firms seek to innovate. The concept of tacit knowledge has attracted much attention in the recent litera-
TACIT KNOWLEDGE, PATH DEPENDENCY, AND LOCAL GROWTH
27
ture on regional innovation systems, clusters and localized learning, where the distinction between tacit (“know how”) and codified (“know that”) knowledge emerges as central. The touchstone for this distinction is Michael Polanyi’s oft-quoted phrase that “we can know more than we can tell” (1966:4) – alluding to a form of knowledge that is “imperfectly accessible to conscious thought” (Nelson and Winter, 1982:79). Underlying this is a deeper distinction between forms of knowledge (explicit or codified) that can be effectively expressed and shared using symbolic forms of representation and other forms of knowledge (tacit) that defy such representation. Hence, economic action depends on the effective combination of both codified and tacit knowledge. Yet, the tacit component of the knowledge required for successful performance of a skill defies codification or articulation either because the performer herself is not fully conscious of all the “secrets” of successful performance, or because the codes of language are not well enough developed or sufficiently universal to permit explication and translation. Given the attention recently focused on tacit knowledge, it should come as no surprise that there is now a lively debate and lack of consensus over the prospects for transferring economic knowledge of this sort over long distances. At least three distinct positions are evident within the literature, each of which produces its own distinctive vision of the geography of tacit knowledge. As we shall see, learning over long distances turns out to be anything but straightforward.
2.3.1 Learning and Regions The “learning regions” thesis is now well established in economic geography, regional economic planning, and the (regional) innovation systems literature (see Lundvall and Johnson, 1994; Florida, 1995; Asheim, 1996; Morgan, 1997; Cooke and Morgan, 1998; Maskell and Malmberg, 1999; Lundvall and Maskell, 2000). Briefly, it argues that tacit knowledge does not “travel” easily. This is because its transmission is best shared through face-to-face interaction between partners who already share some basic similarities: the same language; common “codes” of communication; shared conventions and norms; personal knowledge of each other based on a history of successful collaboration or informal interaction. These commonalities are said to serve the vital purpose of building trust between partners, which in turn facilitates the local flow of tacit knowledge between partners.1 While this analysis appears, on the surface, to be solely concerned with the sharing of tacit knowledge, it also implicitly addresses the issue of how tacit knowledge is produced, identified, and appropriated. Because this approach has adopted the learning-by-interacting model as the cornerstone of its
28
Meric S. Gertler
conceptual framework, it argues that the production of tacit knowledge occurs simultaneously with the act of transmission – primarily through the mechanism of user-producer interaction (Lundvall, 1988; Gertler, 1995). According to this perspective, knowledge does not flow unidirectionally from technology producers to users. Instead, users provide tacit knowledge to producers in order to enable the latter to devise innovative solutions to users’ practical problems. At the same time, by supplying users with innovative technologies, producers are also sharing their tacit knowledge with their customers. The end product arising from this close interaction benefits both users and producers, and embodies within it new tacit knowledge that could not have been produced by either party working in isolation. This, in effect, describes a social process of joint innovation and tacit knowledge production. On the related issue of finding and appropriating tacit knowledge, the learning region approach at least implies that firms will search locally first, for reasons well justified in economic terms. Their intimate knowledge of other local firms and their capabilities is built up through past interaction and/or word-of-mouth referrals and local reputation (i.e. network) effects. This “local knowledge” greatly improves their odds of finding the right “match.” Moreover, as is clear from the above discussion, appropriation (access to and successful absorption) of another firm’s tacit knowledge is greatly facilitated by the bonds of trust that have developed between such local partners over time, or which are supported by strong locally-grounded deterrents to opportunistic behavior by potential innovation partners with respect to the use of one another’s intellectual property. The tacit knowledge geography associated with this perspective is clear and unequivocal: since spatial proximity is key to the effective production and transmission/sharing of tacit knowledge, this reinforces the importance of innovative clusters, districts, and regions. Moreover, as Maskell and Malmberg (1999) point out, these regions also benefit from the presence of localized capabilities and intangible assets that further strengthen their centripetal pull. Many of these are social assets – i.e. they exist between rather than within firms (Leonard and Swap, 2000). Although they are therefore not fully appropriable by individual firms, only local firms can enjoy their benefits. These assets also include the region’s unique institutional endowment, which can act to support and reinforce local advantage. Because such assets exhibit strong tendencies of path-dependent development, they may prove to be very difficult to emulate by would-be imitators in other regions, thereby preserving the initial advantage of “first mover” regions. Maskell and Malmberg argue (1999:181): It is the region’s distinct institutional endowment that embeds knowledge and allows for knowledge creation which – through interaction with available physical and human resources – constitutes its capabilities and en-
TACIT KNOWLEDGE, PATH DEPENDENCY, AND LOCAL GROWTH
29
hances or abates the competitiveness of the firms in the region. The pathdependent nature of such localised capabilities makes them difficult to imitate and they thereby establish the basis of sustainable competitive advantage.
Allen (2000) has raised some important questions about the “selfevident truths” at the heart of the learning region thesis. In particular, he doubts the validity of the underlying assumption that tacit knowledge can only be connected to the local scale, while codified knowledge is necessarily global in reach or availability. He sees this dualism as unhelpful to our understanding of the geography of knowledge flows, considering it “a flawed, if not spurious, exercise” (30). Allen suggests “distanciated contacts” and “‘thick’ relationships may span organizational and industry boundaries, as can the puzzles and performances which constitute them” (28). In asserting that “the translation of ideas and practices...[is] likely to involve people moving to and through ‘local’ contexts, to which they bring their own blend of tacit and codified knowledge, ways of doing things and ways of judging things” (28), his arguments anticipate a second position evident in the literature.
2.3.2 Communities of Practice A more recent literature, some of which also traces its intellectual lineage back to a competence-based view of the firm, emphasizes the central role of “communities of practice” as key entities driving the firm’s knowledgeprocessing activities. This literature argues that routines and established practices shaped by organizations (or subset communities within organizations) promote the production and sharing of tacit knowledge (Brown and Duguid, 1996 and 2000; Wenger, 1998; Wenger and Snyder, 2000). Communities of practice are defined as groups of workers informally bound together by shared experience, expertise, and commitment to a joint enterprise. These communities normally self-organize for solving practical problems facing the larger organization and, in the process, they produce innovations (both product and process). The commonalities shared by members of the community facilitate the identification, joint production and sharing of tacit knowledge through collaborative problem solving assisted by story telling and other narrative devices for circulating tacit knowledge. Thus, according to this approach, organizational or relational proximity and occupational similarity are more important than geographical proximity in supporting the production, identification, appropriation, and flow of tacit knowledge (Amin, 2000; Amin and Cohendet, 1999 and 2000). The resulting
30
Meric S. Gertler
tacit knowledge geography is distinctly different from that which is envisioned by adherents to the learning region approach. In this view, the joint production and diffusion/transmission of tacit knowledge across intraorganizational boundaries is possible, so long as it is mediated within these communities. Moreover, because communities of practice may extend outside the single firm to include customers or suppliers, tacit knowledge can also flow across the boundaries of individual organizations. To this point, the argument does not appear to differ substantially from the user-producer interaction perspective inherent in the learning region approach. However, the communities of practice literature asserts plainly that tacit knowledge may also flow across regional and national boundaries if organizational or “virtual community” proximity is strong enough – a phenomenon that Bunnell and Coe (2001) refer to as the “de-territorialisation of closeness.” In other words, learning (and the sharing of tacit knowledge) need not be subject to the “friction of distance” if relational proximity is present. Amin (2000:14) states the case with characteristic eloquence: Is it not relational proximity – more specifically, ongoing organisational routines and the social practices of collectives implicated in a common venture – rather than geographical proximity that constitutes the “soft” architecture of learning? Such relational proximity might, of course, draw on face-to-face contact, but it can also be achieved at a distance (isn’t this what the communications revolution and global business travel are all about?). More importantly, relational proximity does not in any way implicate, a priori, local clustering or any of the other properties of place that economic geographers and geographical economists have come to stress in recent years.
Rather than seeing “the local as a unique source of tacit knowledge for competitive advantage,” Amin argues that “it is within organisational spaces, with their complex geographies blending action at a distance and local practices, that codified and tacit knowledge are mobilized for competitive advantage” (2000:14). Allen (2000:28) concurs, noting that: What matters in such situations is not the fact of local embeddedness, but the existence of relationships in which people are able to internalize shared understandings or are able to translate particular performances on the basis of their own tacit and codified understandings.
In other words, in place of local context, this perspective substitutes organizational context as the crucial social environment shaping tacit knowledge production, identification, appropriation, absorption, and circulation. These arguments are useful reminders of the importance of relationships and the strength of underlying similarities rather than geographical proximity per
TACIT KNOWLEDGE, PATH DEPENDENCY, AND LOCAL GROWTH
31
se in determining the effectiveness of knowledge sharing between economic actors. However, they do beg one very important question: what forces shape or define this “relational proximity,” enabling it to transcend physical, cultural, and institutional divides? How are “shared understandings” produced? Much of the communities of practice literature is largely silent on this question, although the work of Brown and Duguid (1996 and 2000) stands out as a clear exception. Brown and Duguid freely acknowledge a potential problem when they argue that tacit knowledge cannot be assumed to circulate freely just because the technology to support its circulation is available. This conclusion is especially noteworthy considering that Brown holds the post of Chief Scientist and Director of the Xerox Palo Alto Research Center (PARC), an organization that Schoenberger (1997:187) describes as “one of the most productive and creative computer research centers in the country,” established in the heart of Silicon Valley in 1970.2 In sharp contrast to the arguments reviewed above, Brown and Duguid (2000) stake out a very different position on the spatial reach of communities of practice (143): They are relatively tight-knit groups of people who know each other and work together directly. They are usually face-to-face communities that continually negotiate with, communicate with, and coordinate with each other directly in the course of work. And this negotiation, communication, and coordination is highly implicit, part of work practice,...work chat. ... Groups like this cultivate their own style, their own sense of taste, judgment, and appropriateness, their own slang and in-terms. ... In these groups, the demands of direct coordination inevitably limit reach. You can only work closely with so many people. ... Ideas and knowledge may be distributed across the group, not held individually. These groups allow for highly productive and creative work to develop collaboratively.
On the use of information technologies per se, they are equally unequivocal (146): Yet for the sort of implicit communication, negotiation, and collective improvisation that we have described as part of practice, learning, and knowledge sharing, it’s clear that there are advantages to working together, however well people may be connected by technology. Indeed, one of the most powerful uses of information technology seems to be to support people who do work together directly and to allow them to schedule efficient face-to-face encounters.
In their view, the narratives and social ties so crucial to the flow of knowledge within communities of practice are deeply embedded within the social systems in which they arise. This view provides a convenient bridge to
32
Meric S. Gertler
the third perspective on the potential for tacit knowledge to be transferred over long distances.
2.3.3 Knowledge Enablers In contrast to the optimistic outlook of the communities of practice literature (Brown and Duguid excepted), another recent body of work begins from a rather different premise. It argues that while organizations may be able to produce tacit knowledge effectively (implicitly, using spatially concentrated resources to achieve this), it is devilishly difficult to disseminate or share it more widely (“harness” it) within the organization. This problem has become the focus of a huge effort by firms (especially large ones), and has come to be recognized – even by those who originally promoted the idea of a “knowledge-creating company” – as a very significant obstacle to greater innovativeness (Ichijo, von Krogh and Nonaka, 1998; von Krogh, Ichijo and Nonaka, 2000). This literature sets about to document some of the creative ways in which some firms have responded to this situation, emphasizing the key role of knowledge enablers – that is, “knowledge activists” who aim to span boundaries within the large organization, acting as agents for the diffusion of tacit knowledge – normally with at least partial codification in the process of transmission. The boundary-spanning strategies of these knowledge activists make heavy use of story telling as a key mode of tacit knowledge transfer. Even this can only work when supported by direct, face-to-face interaction and communication between people. For this reason, another key element of a knowledge-enabling strategy is the circulation of key personnel between head office and branch locations (or between different branches) around the globe. The geography of tacit knowledge implicit in this approach is subtly but importantly different from either of the first two positions reviewed above. While the production of tacit knowledge remains strongly localized, the possibilities for its dissemination – once produced – create large spread effects within multi-divisional and multi-locational organizations. There is also at least the potential for wider diffusion of this knowledge outside the organization, if the appropriate enablers are in place. In their recent book, von Krogh et al. (2000) argue that “microcommunities of knowledge” play a key role to ensure the success of this tacit knowledge circulation within large organizations. These are small groups (typically no more than five to seven people) whose members are strongly bound together through common work histories and who employ face-to-face interaction as their most important modus operandi. Of course, von Krogh et al. recognize that “geography” makes all of this more difficult and challenging.
TACIT KNOWLEDGE, PATH DEPENDENCY, AND LOCAL GROWTH
33
It is clear from their discussion that for them, “geography” signifies both physical separation and local cultural differences. While they emphasize the importance of a common or shared “social context” in facilitating the flow of tacit knowledge, they view the creation and shaping of this context as primarily within the purview of the firm.3 Furthermore, although they offer detailed renderings of fascinating case studies involving the transmission of tacit technological knowledge between culturally (as well as physically) distant sites (e.g. advanced intercity rail technology being transposed from Switzerland to India), they provide no real insights into how “local culture” is produced.
2.4 THE ROOTED TRAJECTORIES OF REGIONAL ECONOMIC CHANGE What seems clear from the preceding discussion is that considerable disagreement and confusion persists concerning the nature of tacit knowledge and its geography. I will argue in this section that the principal reason for this confusion stems from the relatively limited and superficial interpretation of the concept adopted by most scholars up until now. Furthermore, these limitations arise from a too literal reliance on Michael Polanyi’s own particular conception of tacit knowledge. My goal here is to reconstruct our understanding of this notion based on more robust foundations which allow us, in particular, to develop a deeper understanding of the true meaning and significance of local context or culture. Recalling our earlier discussion, Polanyi’s original conception of tacit knowledge is primarily experiential and cognitive in nature. It is experiential in the sense that he conceives of tacit knowledge as an understanding of “know-how,” acquired through experience. It is cognitive in the sense that it defies conscious articulation – meaning that (i) we may not even be aware of it, and (ii) when we try to articulate or explain it to someone else, communicating this knowledge in verbal, written, or diagrammatic form will never be equal to the task. Hence, tacit knowledge must be learned by demonstration, imitation,4 performance and shared experience: “I cannot explain very well how I do this, so let me show you.” Polanyi asserts that tacit knowledge is context-dependent in the sense that common rules shared between one person and another are important for the successful transmission of tacit knowledge. However, perhaps the reason so many have gone astray in applying Polanyi’s concept is that he never fully specifies how “context” and “rules” are produced. Instead, they remain primarily idiosyncratic and “cultural” in origin, in line with his larger project to legitimize the personal, passionate pursuit of
34
Meric S. Gertler
knowledge by individual scholars as the driving force behind scientific advance.5 This suggests to us that we cannot sort out the geography of tacit knowledge without inquiring more systematically into the fundamental nature of “culture” and the institutional underpinnings of economic activity. Ironically, the source for a more helpful perspective on this issue is another wellknown Polanyi – Michael’s older brother, Karl – whose classic work (1944) yields the central insight that markets and the behavior of economic actors are socially constructed, embedded, and governed. This has important implications for the tacit knowledge problems discussed above. It suggests that the ability of individual workers or firms to produce and share tacit knowledge depends on much more than spatial proximity or cultural affinity. In particular, it depends on institutional proximity – that is, the shared norms, conventions, values, expectations, and routines arising from commonly experienced frameworks of institutions. This form of proximity or affinity may override organizational or relational proximity if the organization in question extends geographically across institutional divides. Karl Polanyi’s work also suggests that the ability of firms to find and appropriate the tacit knowledge produced by individual workers on the shop (or office, or laboratory) floor, or by teams (communities) will be highly sensitive to the institutions governing the employment relation, which, themselves, vary geographically by nation and region (Kochan and Osterman, 1994; Wever, 1995). These insights suggest that we need to devise an alternative conception of tacit knowledge, since (Michael) Polanyi’s famous aphorism “we can know more than we can tell” seems to lead us – unproductively, in my view – into the realm of psychology, cognitive science, sensory perception, and the psychomotor foundations of skilled performance, without paying equivalent attention to the broader (geographically delimited) context within which such performances are situated.6 Instead, we might find it more useful to reject this phrase in favor of alternative formulations such as: “we do not understand what we know, or why we do things the way we do.” Following Karl’s inspiration, then, we can more productively interpret the origins of routines, characteristic practices, “settled habits of thought,” and “second nature” as arising from concrete institutional origins: while corporate agency and the distinctive “culture” of the firm undoubtedly play a major role, they do not exist within a vacuum (Schoenberger, 1997; Glasmeier, 2001) and, contrary to the underlying premise of much of the knowledge management literature, managers do not fully (or even largely) shape their own destiny. They operate within a possibility set that is constrained by larger forces – particularly the institutional and regulatory frameworks at the national and regional scales (Lam, 1998; Lam and Lundvall, 2000; Whitley, 1999).7
TACIT KNOWLEDGE, PATH DEPENDENCY, AND LOCAL GROWTH
35
These frameworks influence firms’ practices, values, and expectations in fundamental ways. The structure of capital markets has a profound influence over firms’ time horizons, which in turn shape engineers’ choices of production technologies, the design characteristics of their products, but also firms’ interest and willingness to engage in long-term, close interaction with potential local innovation partners (Gertler, 1997 and 2002). The structure of corporate governance systems shapes and constrains the firm’s strategic priorities and time horizons, as well as its managers’ abilities to organize the workplace in order to “reap” workers’ tacit knowledge (O’Sullivan, 2000). The shape of labor market legislation and industrial relations regulation affects crucial variables such as training regimes, rates of labor force turnover, inter-firm mobility, and the ability of employers to retain “talent” (Wever, 1995). There are two general points arising from this analysis. The first is that “context” is, to a very large degree, defined by such institutional features. Of course, as Maskell and Malmberg (1999:173) point out, the institutional endowment of a region or nation accumulates over time, and “thus represents the intricate contemporary interaction between elements of different ages ... – from the very old (religion, beliefs, values) to the recent/current (contemporary industry standards, current regulations, etc.).” The second point – and the explicit link to tacit knowledge – is that such institutional influences are subtle but pervasive: indeed, often so subtle that firms and individuals are not even conscious of the impact they exert over their own choices, practices, attitudes, values, and expectations. That firms remain almost completely oblivious to the influence of these institutional forces becomes readily apparent whenever they attempt to engage in learning which spans institutionally defined contextual divides – whether this is between different firms or between different divisions or branches of the same firm situated in different institutional settings (Gertler, 2001 and 2002). Consider some brief examples to illustrate this point. Returning to the case of Xerox, Schoenberger documents how the parent corporation, based on the east coast of the United States, was singularly incapable of reaping the commercial benefits of the innovations produced in its “peripheral” research sites – Xerox PARC in Palo Alto or its foreign subsidiary Fuji Xerox in Japan. Her careful analysis of this sorry tale reveals that the principal impediment to successful commercialization was the fact that these peripheral research sites operated under radically distinct cultures and contexts, defined by divergent institutional frameworks at the regional or national scale. Hence, despite the stellar track record of innovation at PARC, other firms such as Apple and Microsoft appropriated its most important new product ideas and system innovations. Similarly, the innovations developed at Fuji Xerox were significant enough to have offered the parent firm a viable new product tech-
36
Meric S. Gertler
nology at a time when its existing products were quickly losing market share. However, the larger corporation proved itself incapable of learning from its smaller “outpost” operations, despite unequivocal evidence of clearly superior technologies (Schoenberger, 1999). These cases demonstrate that learning, when attempted across major institutional-contextual boundaries, will be subject to formidable obstacles, even in the presence of substantial corporate wealth and resources. The inevitable geographical variations in institutionally defined local context are endemic to large, global organizations, meaning that fully “knowing” what some key employee, situated in a far-flung corner of the corporation, knows will be all but impossible. In addition, even if one could know this, the ability or inclination of central management to act on this knowledge would surely be limited. The upshot is that transcending the bonds of spatial proximity may be possible, but it will also be difficult and expensive, because of the fundamentally different institutional environments involved – what we might understand as the distinctive and uneven (though systematic) “economic geography of context.” Technological fixes and corporate willpower alone may not be sufficient to overcome these obstacles. In conclusion, an important implication arising from the arguments outlined above is that the geographically specific, multi-level institutional foundations of tacit knowledge production also act to shape regional trajectories of growth in strongly path-dependent ways (Zysman, 1994). This is not to say that regional economies remain forever captive to inherited institutional features and frameworks – only that attempts to alter the course of their evolution over time are more likely to succeed when development strategies are conceived in ways that acknowledge and exploit the region’s institutionally shaped competitive advantages. In other words, regional policy makers concerned with changing the behavior of managers, workers and firms should think very carefully before selecting role models from which to learn. While there is still considerable scope for the agency of the individual, the firm, or the civic entrepreneur here, the ease with which particular choices can be pursued will continue to be shaped by larger forces of governance.
ACKNOWLEDGEMENTS The financial support of the Social Sciences and Humanities Research Council of Canada and the Goldring Chair in Canadian Studies at the University of Toronto is gratefully acknowledged. I would also like to acknowledge the helpful comments of the participants of the workshop on “Rethinking Re-
TACIT KNOWLEDGE, PATH DEPENDENCY, AND LOCAL GROWTH
37
gional Innovation and Change: Path Dependency or Regional Breakthrough?” in Stuttgart, February 28-March 1, 2002.
NOTES 1.
In a recent paper, Cohen and Fields (1999) question the extent of similarity between Silicon Valley and the “paradigmatic” European cases of learning regions and industrial districts. They contend that the social cohesion within Silicon Valley is based more on widely circulated knowledge concerning the reputations and reliability of individual firms, as well as inter-firm mobility of skilled workers, rather than on deep cultural similarities borne of decades (if not centuries) of close social interaction between firms. Consistent with this view, Florida’s (1995) conception of a learning region is considerably more modest than those espoused by European authors on the subject.
2.
As evidence of her claim, Schoenberger notes (188) that “PARC staffers, in a remarkably short time, developed many of the essential hardware and software components of...[a new, open information architecture]. These included personal computers, graphics interfaces, laser printers, WYSIWYG...word processing programs, and networking capability.”
3.
In doing so, they seem to accept uncritically the typical corporate hubris that represents “corporate culture” as the all-powerful but easily malleable plaything of senior company executives, operating independently of a wider social or regulatory context. For penetrating critiques of this position, see Schoenberger (1997) and Glasmeier (2001).
4.
Brown and Duguid (2000:136) liken this to theft, referring to “stolen knowledge” – that which is acquired by unobtrusive or covert observation of expert performances, rather than through formal instruction.
5.
“Tacit assent and intellectual passions, the sharing of an idiom and of a cultural heritage, affiliation to a like-minded community: such are the impulses which shape our vision of the nature of things on which we rely for our mastery of things” (Polanyi, 1958:266).
6.
One warning of the potentially futile direction in which this leads us comes from the curious nature of virtually all of the examples used to convey the meaning of tacit knowledge à la Michael Polanyi: recognizing a face, riding a bicycle, swimming, playing a piano, or more recently, landing an airplane, or baking bread. The connections between examples such as these, which emphasize the psychomotor properties of performance of physical skills, and innovative practices in regional and national economies seem remote.
38
Meric S. Gertler
7.
In contrast to this analysis of Karl Polanyi’s contribution to our understanding of the origins of tacit knowledge and context, Allen (2000:27) argues that the problematic dualism of local/tacit versus global/codified knowledge arises from the widespread tendency to blend Karl’s central idea of embeddedness with Michael’s idea of tacitness. My own interpretation differs from Allen’s: where he sees “embeddedness” as the culprit, I would suggest instead a less than careful reading of Karl’s more fundamental argument concerning the institutional underpinnings of economic action. In my view, the real confusion has arisen from excessive reliance on the ideas of Michael, to the neglect of Karl’s insights.
REFERENCES Allen, J., “Power/economic knowledge: Symbolic and spatial formations.” In Knowledge, Space, Economy, J.R. Bryson, P.W. Daniels, N. Henry and J. Pollard, eds. London: Routledge, 2000. Amin, A. and Cohendet, P., Learning and adaptation in decentralised business networks. Environment and Planning D: Society and Space 1999; 17:87-104. Amin, A. and Cohendet, P., Organisational learning and governance through embedded practices. Journal of Management and Governance 2000; 4:93-116. Amin, A., Organisational learning through communities of practice. Paper presented at the Workshop on The Firm in Economic Geography, University of Portsmouth, UK, 9-11 March, 2000. Arthur, B., Competing technologies, increasing returns and lock-in by historical events. Economic Journal 1989; 99:116-131. Asheim, B., Industrial districts as ‘learning regions’: A condition for prosperity? European Planning Studies 1996; 4:379-400. Barnes, T.J., “Theories of accumulation and regulation: bringing life back into economic geography – introduction to section three.” In Geographies of Economies, R. Lee and J. Wills, eds. London: Arnold, 1997. Brown, J.S. and Duguid, P., “Organizational learning and communities-of-practice.” In Organizational Learning, M. Cohen and L. Sproull eds. London: Sage, 1996. Brown, J.S. and Duguid, P., The Social Life of Information. Boston: Harvard Business School Press, 2000. Bunnell, T. and Coe, N., Spaces and scales of innovation. Progress in Human Geography 2001; 25:569-89. Cohen, S. and Fields, G., Social capital and capital gains in Silicon Valley. California Management Review 1999; 41:108-30. Cooke, P. and Morgan, K., The Associational Economy: Firms, Regions and Innovation. Oxford: Oxford University Press, 1998.
TACIT KNOWLEDGE, PATH DEPENDENCY, AND LOCAL GROWTH
39
David, P., Clio and the Economics of QWERTY. American Economic Review 1985; 75:332337. David, P., Why are institutions the ‘carriers of history?’ Path dependence and the evolution of conventions, organizations and institutions. Structural Change and Economic Dynamics 1994; 5:205-20. Florida, R., Toward the learning region. Futures 1995; 27:527-36. Gertler, M.S., ‘Being there’: Proximity, organization, and culture in the development and adoption of advanced manufacturing technologies. Economic Geography 1995; 71:1-26. Gertler, M.S., “Technology, culture and social learning: regional and national institutions of governance.” In Innovation and Social Learning, M.S. Gertler and D.A. Wolfe, eds. Basingstoke: Palgrave/Macmillan, 2002. Gertler, M.S., “The invention of regional culture.” In Geographies of Economies, R. Lee and J. Wills, eds. London: Edward Arnold, 1997. Gertler, M.S., Best practice? Geography, learning and the institutional limits to strong convergence. Journal of Economic Geography 2001; 1:5-26. Glasmeier, A.K., Manufacturing Time: Global Competition in the Watch Industry, 1795-2000. New York: Guilford Press, 2001. Hall, P. and Soskice, D., eds., Varieties of Capitalism. Oxford: Oxford University Press, 2001. Hodgson, G., The ubiquity of habits and rules. Cambridge Journal of Economics 1997; 21:663-84. Ichijo, K., von Krogh, G. and Nonaka, I., “Knowledge enablers.” In Knowing in Firms, G. von Krogh, J. Roos, and D. Kleine, eds. London: Sage, 1998. Kochan, T.A. and Osterman, P., The Mutual Gains Enterprise. Boston: Harvard Business School Press, 1994. Lam, A. and Lundvall, B-Å., Innovation policy and knowledge management in the learning economy: the interplay between firm strategies and national systems of competence building and innovation. Paper presented at the OECD High Level Forum on Knowledge Management: The New Challenge for Firms and Organizations, Ottawa, September 21, 2000. Lam, A., Tacit knowledge, organisational learning and innovation: a societal perspective. DRUID Working Paper No. 98-22, Aalborg University, Denmark, 1998. Leonard, D. and Swap, W., Gurus in the garage. Harvard Business Review 2000; 78:71-82. Lundvall, B.-Å. and Maskell, P., “Nation states and economic development – from national systems of production to national systems of knowledge creation and learning.” In The Oxford Handbook of Economic Geography, G.L. Clark, M.P. Feldman, and M.S. Gertler, eds. Oxford: Oxford University Press, 2000. Lundvall, B-Å and Johnson, B., The learning economy. Journal of Industry Studies 1994; 1:23-42.
40
Meric S. Gertler
Lundvall, B-Å., “Innovation as an interactive process: from user-producer interaction to the national system of innovation.” In Technical Change and Economic Theory, G. Dosi, C. Freeman, O. Silverberg, and L. Soete, eds. London: Frances Pinter, 1988. Maskell, P. and Malmberg, A., Localised learning and industrial competitiveness. Cambridge Journal of Economics 1999; 23:167-86. Morgan, K., The learning region: institutions, innovation and regional renewal. Regional Studies 1997; 31:491-504. Myrdal, G., Economic Theory and Under-Developed Regions. New York: Harper and Row, 1957. Nelson, R.R. and Winter, S.G., An Evolutionary Theory of Economic Change. Cambridge, MA: Harvard University Press, 1982. O’Sullivan, M., Contests for Corporate Control: Corporate Governance in the United States and Germany. Oxford: Oxford University Press, 2000. Polanyi, K., The Great Transformation. New York: Rinehart, 1944. Polanyi, M., Personal Knowledge: Towards a Post-Critical Philosophy. London: Routledge and Keegan Paul, 1958. Polanyi, M., The Tacit Dimension. New York: Doubleday, 1966. Schoenberger, E., “The firm in the region and the region in the firm.” In The New Industrial Geography: Regions, Regulation and Institutions, T.J. Barnes and M.S. Gertler, eds. London: Routledge, 1999. Schoenberger, E., The Cultural Crisis of the Firm. Oxford: Blackwell, 1997. Storper, M. and Walker, R.A., The Capitalist Imperative. Oxford: Blackwell, 1989. von Krogh, G., Ichijo, K. and Nonaka, I., Enabling Knowledge Creation: How to Unlock the Mystery of Tacit Knowledge and Release the Power of Innovation. Oxford: Oxford University Press, 2000. Walker, R.A., “The geography of production.” In A Companion to Economic Geography, E. Sheppard and T. J. Barnes, eds. Oxford: Blackwell, 2000. Wenger, E. and Snyder, W.H., Communities of practice: The organizational frontier. Harvard Business Review 2000; 78:139-45. Wenger, E., Communities of Practice: Learning, Meaning and Identity. Cambridge: Cambridge University Press, 1998. Wever, K.S., Negotiating Competitiveness: Employment Relations and Organizational Innovation in Germany and the United States. Boston: Harvard Business School Press, 1995. Whitley, R., Divergent Capitalisms: The Social Structuring and Change of Business Systems. Oxford: Oxford University Press, 1999.
TACIT KNOWLEDGE, PATH DEPENDENCY, AND LOCAL GROWTH
41
Young, A.A., Increasing returns and economic progress. The Economic Journal 1928; 38:52742. Zysman, J., How institutions create historically rooted trajectories of growth. Industrial and Corporate Change 1994; 3:243-83.
Chapter 3 REGIONAL TRANSFORMATION AND REGIONAL DISEQUILIBRIUM: NEW KNOWLEDGE ECONOMIES AND THEIR DISCONTENTS
Philip Cooke
3.1 INTRODUCTION Today, it can be argued with some confidence that we are living through a period of disruptive economic change, driven by new knowledge creation and innovation. Since Paul David’s (1975) introduction of “path dependence” as a developmental concept, we became used to the idea that even radical innovation is conditioned significantly by pre-existing sunk costs. This was given support by research on national systems of innovation showing most innovation to be marginal change not breakthroughs (Edquist, 1997). There is clearly a strong technological determinism at work in these observations despite awareness in both research traditions of the social nature of innovation. We must be willing to explore, in pondering regional transformation, the differences between regional and technological trajectories and path dependencies even though technology is a major driver of regional economic change. One important difference is that competitive advantage is not only technological but involves factors such as knowledge “talent” (Florida, 2002).
3.2 KNOWLEDGE ECONOMY AS A HISTORICAL RECONSTRUCTION Most accounts of the rise of Silicon Valley focus on technology rather than organization. An extreme example of this is an account that traces the area’s rise to pre-electronic times. Sturgeon (2000) supplies an etymology that also displaces Silicon Valley’s lineage to San Francisco while placing inventor
44
Philip Cooke
Lee de Forrest’s work on wireless valves (in the US vacuum tubes) in the context of San Francisco’s Federal Telegraph Corporation, and their navy defense contracts for ship-to-shore radio. Much of the period from around 1906 to 1939 (Hewlett Packard) then 1971 (Intel) is filled with references to the likes of Philo Farnsworth, an early patentee of television and various other more or less obscure inventors located in San Francisco. Thus, both significant spatial and technological admixtures are utilized to give modern day Silicon Valley a possibly spurious pre-history. The reason for this has as much to do with status as with science, for how can a place without history be taken seriously? Each downturn in its actual history was perceived by bankers, governments and academics as fitting the presumption of many that Silicon Valley would probably not last but would migrate or simply disappear according to principles of product life cycle theory like the minicomputer cluster in Boston (Saxenian, 1994). In other words, the efforts to invent tradition for Silicon Valley, a place without an administration, with an identity literally built on sand, are to embed it symbolically in the minds of power brokers for whom, it is feared, a west coast location must always seem peripheral, if not marginal. Despite its tendentiousness, Sturgeon’s approach at least avoids that of a leading exponent of path dependence, Brian Arthur who is reduced to “explaining” the existence of Silicon Valley as pure chance (Arthur, 1994). The counter-argument proposed here is that something new happened that had less to do with Lee de Forrest or Hewlett Packard and much more to do with venture capitalists like Arthur Rock and Gene Kleiner. That is, Silicon Valley was the first place to systematize the process of interactive innovation, in contrast to the linear innovation practices of the likes of GE and RCA that sought to impose hierarchical control over the innovation process by squeezing out the likes of Philo Farnsworth from the history of technology. As Sturgeon usefully shows, if there was one collective source of entrepreneurial energy underpinning the emergence of what became Silicon Valley “it was opposition to the domination of the field by RCA” (Sturgeon, 2000:28). In other words, what is of particular interest in Silicon Valley is not the technology but the means by which “exploration” knowledge moves to “exploitation” knowledge, from the laboratory bench to the market through a chain reaction involving multiple intermediaries, and in particular who those intermediaries are and what services they provide. Despite the major downturn in information and communications technologies (ICT) in Silicon Valley as everywhere else in 2001, sufficient has been written about the processes fuelling the Internet communications boom for there to be now a reasonably clear picture of the most recent form taken by its “innovation growth engine.” Summarizing, we can say that the following are central institutional features:
REGIONAL TRANSFORMATION AND REGIONAL DISEQUILIBRIUM
45
Basic research, knowledge generation, and application capability of the kind normally found centered in advanced private research or leading edge public research laboratories. Venture capital is crucial as the means by which ideas that have been screened and selected are given a chance to fly as commercial products or services. Law firms are important as gatekeepers, advising firms on appropriate investors, counselors assisting entrepreneurs to access other services, and sources of contacts for many things ranging from recruitment to contract manufacturing. Specialist consultants in business and technological services ranging from management accountants to auditing and accountancy services, head hunting services, and specialist technical services. A local value chain of firms that can conduct, for example, contract manufacturing, design and fabrication, and various prosaic supplies like logistics, or exhibition organization and specialized catering services (Bahrami and Evans, 2000). It is notable that the overwhelming majority of these elements of the innovation support infrastructure are private. This raises the question as to why Silicon Valley is not fully operational as a competitive, arm’s-length exchange-driven system. This is because of certain key conventions in the sense conveyed by Douglass North (1993) as “ways of being,” expected behaviors and forms of practice. These intersperse and facilitate adhesion of the marketdriven character of the cluster with forms of social capital that draw on aspects of shared culture that are more collective than purely competitive. In the downturn, sadly, social capital is the first casualty. There has been the consequential wreckage of hundreds of start-up companies and the human waste involved in a business culture in which jobs are tied to revenues. These layoffs are structured so that freelances and other short-term contract workers, otherwise the lower ranks, are fired first. Even before the downturn occurred, negatives like difficult commutes, high living costs, pollution, working conditions involving 60-hour working weeks, tiny workspaces or “cubicles,” modest wages, and the routine nature of most work were compensated for by the anticipation of high returns from stock options. Once the value of these stock options declined, sometimes to zero, and the system of “rating and ranking” workers on performance turned into a job-culling mechanism, the sacrifices no longer seemed worthwhile. In the first half of 2001, 107,000 persons were made redundant in California, compared to 14,300 in the equivalent period of 2000 (Gardner, 2001). Without naively wishing for “the airplane that cannot crash” as a Silicon Valley venture capitalist put it to this
46
Philip Cooke
author in 2002 (see Isaacs, 2002) it is not unreasonable to look for models that offer better than “crash and burn” control mechanisms.
3.3 LEARNING THE KNOWLEDGE ECONOMY MODEL: ACCIDENT OR DESIGN? It has been argued thus far that the Silicon Valley cluster experience is underpinned by a simple model of innovation whereby venture capital induces start-up businesses from research laboratories. From a regional policy perspective, it is this rather than the technology that grew from the early semiconductor designs that was really a breakthrough. However, it has taken two decades for that to be realized, about the same time it took for another organizational innovation – “lean production” – to be understood in the West. There it was finally successfully deployed in the global manufacturing industry contest that has seen the question “Can Japan Compete?” replacing “Regaining the Productive Edge” as a rising slogan (Dertouzos, Lester and Solow, 1989; Porter, Takeuchi and Sakakibara, 2000). It can be argued that, just as leading firms eventually learnt how to adapt to lean production, leading regions learnt first about the importance of exploiting knowledge-based industry and having the key investors in geographical or functional (distant but strong networks) proximity to facilitate innovation and commercialization. They were learning economies that quickly understood the changed nature of what came to be called “structural competitiveness” (Cooke and Schienstock, 2000). This meant the following: Successful just-in-time management of production stocks and flows by firms. Effective firm cross-functional integration of marketing, R&D, design and production. Linkage between internal and external, university, and research laboratory research. Effective elaboration of local and global value chains. Recognition of value of investment in skills upgrading. Prioritizing and implementing total quality management schemes. Regions with strong markets that could supply support services to customers in these fields had a competitive advantage. Regions with leading animateur firms that were also “demanding customers” – as Porter (1990) put it – were similarly privileged. Failing that, regions with innovative public innovation support systems were advantaged over those without any of these supports, but not over the other two “systems.” In brief, the successful regions made a breakthrough from a previous existence or path dependence by
REGIONAL TRANSFORMATION AND REGIONAL DISEQUILIBRIUM
47
realizing the nature of the new rules of the competitive game. Importantly, there was not a single pathway to achieving this, but nor was chance the explanation for why regional breakthrough happened where it did. Such locations create profound social and spatial disequilibria. The designed, rather than accidental, aspect of regional breakthrough occurred because of regional innovation and learning capabilities. In common, these show capability in “crossing boundaries” from knowledge exploration to knowledge exploitation. As Castells (1996) argues, this is at the heart of a certain kind of “knowledge economy” that adds value by the interaction of these two types of knowledge (Cooke, 2002). Elaborating on Latour’s (1998) differentiation of “research” from “science” this is “symbolic” research knowledge valorization (software, genomics, new media), compared to “scientific” knowledge application for testing, examination, quality control etc. (e.g. food processing industries). Such capabilities are expressed, as we shall see, in a variety of forms ranging from prioritizing strategies for academic entrepreneurship, to promoting equity investment, to skills upgrading. We can next examine cases of these, beginning with the example of academic entrepreneurship in Sweden.
3.4 KNOWLEDGE ECONOMY BREAKTHROUGH EXEMPLARS 3.4.1 Academic Entrepreneurship in Sweden Although 1980s legislation paved the way, Swedish academic entrepreneurship grew with the passing of the 1996 Higher Education Act, which officially gave Sweden’s universities this “third task” obligation. It also built on previous initiatives such as allowing professors entrepreneurship rights, establishment since 1989 of university-centered science parks, technology link foundations, and technopoles. In many of these settings, the approach has been to physically combine on a single site, the university research facilities, one or more of Sweden’s largest technology businesses, and a host of university spinout firms. Thus, Kista – an industry science park since 1972 located between Stockholm and Uppsala – has Ericsson as the centerpiece, employing some 12,000 of the 29,000 on site. Mjärdevi Science Park at Linköping has Ericsson and Saab, while Ideon at Lund has AstraZeneca research laboratories employing some 1,200 R&D staff (Brown-Humes, 2001; Jones-Evans and Klofsten, 1997; Jonsson, 2001).
48
Philip Cooke
In Kista, Nokia has 300 employees and global lead firms like Microsoft, Intel, IBM, Sun, Novell, Adobe, Siemens, Oracle, and Hewlett Packard are joined by some 400 mostly academic spinout small and medium-size enterprises (SMEs). Arriving later than Ericsson and IBM were numerous academic research institutes like The Royal Institute of Technology, University of Stockholm, Swedish Institute of Computer Science, Swedish System Development Institute, and the Industrial Center of Microelectronics. They interact as a system of innovation linked also through learning networks like Soft Center to other systems and software science parks in Sweden and the US. Kista is, however, not a fully-fledged cluster because it is substantially vertically networked into Ericsson’s global mobile telephony value chain, nor does it possess a full range of innovation support services on site. Although it is also experiencing the telecommunications downturn and Ericsson announced 1,000 redundancies for Kista in May 2001, it is not as badly affected as Silicon Valley because it was not dependent on Internet businesses and is far more of a services than manufacturing center. Interestingly, it has smaller income differentials and surprisingly low wage costs for engineers and programmers ($24,000 and $22,000 respectively in Kista, compared to $68,000 and $64,000 in Japan and $60,000 and $54,000 per year in the US in 2000) in an otherwise high living cost economy. Comparatively generous welfare payments and working time legislation mean the stress factor is less than Silicon Valley for those laid off as well as those in employment in Kista. Thus, Kista offers what appear to be superior social integration possibilities, reflecting Sweden’s higher prioritization of those aspects in the wider society. Despite this, it is worth noting that by 1998, European Venture Capital Association statistics showed Swedish venture capital investment as a percentage of GDP was third in Europe at 0.17 percent, behind only the UK (0.39 percent) and The Netherlands (0.24 percent). Other insightful Swedish cases include Mjärdevi (Linköping) and Ideon (Lund), both of which are to some extent peripheral to Stockholm. In Mjärdevi, the SMIL (Business Development in Linköping) network and the Centre for Innovation and Entrepreneurship (CIE) at Linköping University both involve public investors such as NUTEK (the Swedish Business Development Agency) and the Technology Links foundation as well as private funds and business angels. SMIL began in 1984 as a joint program by business leaders from earlier university spinout firms and academic personnel to boost technology-based industry. SMIL is a network of some 150 firms supported with training from CIE, funded through NUTEK. It offers three types of formal training. The first is entrepreneurship and new business development aimed at graduates and researchers wishing to begin a firm. Training lasts a year and would-be firms are mentored to the business plan stage with costs met by the network. Seven or eight firms are established annually
REGIONAL TRANSFORMATION AND REGIONAL DISEQUILIBRIUM
49
through this program. The second initiative is a development program for established firms. These firms are grouped and represented by three personnel each and engage in interactive self-learning with CIE tutoring. Similar numbers of firms to those in the start-up program are involved each year. In the third program, management groups focus on problem-centered real-time learning involving SMIL facilitation and external advisors. Modest membership fees are charged to firms involved in these programs. This is designed as a methodology for new technology business formation in a regional context where the resources of a large capital city are absent. In this case, networking supplements the focused resource of the single university and a few large firms in the area. It is highly dependent on public funding, without which it would not exist. Moreover, spinouts are also dependent on contracts from the large firm clients rather than venture capital, though they may also be networked to external investment angels. Ideon is different in its sectoral emphasis. It is one of Sweden’s stronger centers of healthcare and biotechnology enterprises with, it will be recalled, AstraZeneca R&D located in close proximity. Interestingly, in a study of inter-firm relations among fifteen innovative firms in the science park “...venture capitalist firms had not taken any active part in the innovation processes among the case firms” (Jonsson, 2001:17). Although most firms are Lund University spinouts, only a minority had close and long-established links to research undertaken there. Most were present to access scarce skills from the labor educated at the university. Market linkages with firms sharing the science park location exist but these are a minority in the sample, most are present to advertise the status of the science park address. This “symbolic capital” is seen to attract customers and contracts. Most innovation partners and customers or suppliers are distant, many abroad. Thus in Ideon, “untraded interdependencies” are rare, project work on short-term contracts is common, and arm’s-length exchange is the main medium of interaction (for similar conclusions about biotechnology “clusters” in the US, see Zucker et al., 1998). To the extent that systemic relations are found, they are of three kinds. First, limited numbers of intense, regular interactions between research centers and advanced knowledge-based firms in which tacit knowledge, proximate relationships, and social capital are strong. Second, supply chain interactions of a contractual nature over codified knowledge exchange but with modest social capital and limited proximity requirements. Third, loose, proximate supplier networks of low value market products and services of a routine nature. Further examination of innovation interactions from Lund-based biotechnology researchers by Vinnova (2001) reveals strong national network linkages between key centers of expertise. Lund is a prominent center for this kind of activity. Moreover, scientific publishing links to Sweden’s two larg-
50
Philip Cooke
est biotechnology firms, AstraZeneca, where it is second, and Amersham Pharmacia where it is third, show that it is also strong at the key innovation applications interface. Finally, with the construction of the Öresund Bridge (linking Sweden with Denmark), Lund gains opportunities to develop as part of an international healthcare research and technological complex with Copenhagen on the Danish side. An organization to promote a cross-border regional innovation research, Øforsk exists and meetings have been held. This builds upon the Medicon Valley Academy, an association of industry, academic, and hospital medical and biotechnologists from Lund, Malmö, and Copenhagen that sustains networks, and a marketing and lobbying forum to promote a possible future trans-national cluster. To conclude this section on Sweden’s academic entrepreneurship model of grasping opportunities to switch away from path dependence towards the “breakthrough” sectors represented by modern computing, communications, and biotechnologies, there are clear contrasts with the Silicon Valley model. Most obviously, it is the product of public policy designed to stimulate high technology spinout firms through academic entrepreneurship in science parks. New firm formation is also largely dependent upon targeted public support. Additionally, and consistent with a national institutional system of governance in which social partnership remains pronounced, the localized innovation systems involve at least one major Swedish corporation that, in reality, animates a local value chain. Despite this, it is not entirely clear that firms that are successfully brought into being act as firms in fully functioning clusters, if Porter’s (1998) definition involving strong horizontal as well as vertical linkages is taken seriously. Finally, social partnership means less wage polarization and, probably, less brutal treatment of enterprises and workforces in downturns than appears to be the case in Silicon Valley. Even so, Asheim and Clark (2001) report the cost of stress to the Swedish economy as being Euro 4 billion per year. The important feature of the Swedish academic entrepreneurship model is that it demonstrates that significant shifts within broad technological path dependence, taking advantage of breakthrough innovation, are possible by policy design once the fundamental requirements of ideas, finance, and innovation are met. Three questions arise: first, are the resulting firms sustainable in pure market terms? In other words, are they competitive and attractive to equity investment and the initial public offering or market flotation that signifies success? Second, can the model of successfully attracting private venture capital to the regions evolve or will innovative regions continue to require the presence of public investment to sustain innovation? Third, will these “new economy” complexes develop as “generative” clusters rather than weakly integrated value chains, and to what extent may that resolve the first two
REGIONAL TRANSFORMATION AND REGIONAL DISEQUILIBRIUM
51
problems? If so, what policy or market steps are needed to secure such an outcome?
3.4.2 Equity Investment Models in Israel and India A model owing more to market processes – though not without either shortlived or indirect public sector influence before equity investment came swiftly on-stream – is that responsible for the development of globally successful software clusters in Israel and India. The specific locations are, in Israel, Tel Aviv to Haifa, particularly the Herzliah Corridor and, in India, Bangalore and, lately, Hyderabad. Both India (where employment quadrupled to 400,000) and Israel where employment grew from little above zero to many thousands (Teubal, 2000) registered major employment increments in the software industry during the 1990s. In Israel the following process occurred. The model involves as a key theme the role of “Silicon Valley offshore” but it is “generative” because the crucial knowledge behind the core software technology was intrinsic to Israeli research and military applications. Hence, the software application was also indigenous, being in the data security niche, otherwise known as Internet “firewalls.” The Weizmann Institute of Science was key as the invention source of what became a world standard encryption algorithm. Israelis then pioneered the technology, mainly in the Israel Defense Forces and Mossad, the security service. A combination of military downsizing, following the Camp David accords in the early 1990s, and an influx of post-Soviet refugee scientists and entrepreneurs led to first-mover entrepreneurs being former military personnel made redundant with scaling back of Israeli defense expenditure, second mover entrepreneurs being immigrants among whom technical labor was also abundant. First-mover firms specialized in anti-virus, software protection and encryption technologies and included Carmel Software, Iris, BRM and Eliashim. At this stage, there was no venture capital in Israel. Funding for entrepreneurship was accessed through the Office of the Chief Scientist’s Industrial R&D Fund supported by a 1984 R&D law that allowed 50 percent grants for commercial exploitation of scientific research. Firms like Algorithmic Research and NDS were set up in the emerging Herzliah Corridor. A second development stage (1990-96) saw the emergence of “firewall” firms. With the arrival of the Internet in this period, demand for encryption engines mushroomed, and world-leader firms like Checkpoint, Memco, and Aladdin were founded. At this time also (1991), the Israeli government set up a public venture capital firm called Yozma, which was instrumental in the evolution of the Israeli venture capital industry. Yozma was introduced in
52
Philip Cooke
recognition of a key weakness in the Israeli system of innovation. It was perceived that without the substantial presence of a vehicle to stimulate the emergence of private venture capital there could be no breakthrough from old path dependence. Yozma was designed as fitting into the wider business sectors support structure in Israel with a specific brief. This was to fund through equity investment the clear demand from new security software firms, and strengthen connections with and promote the opportunities of investment to the US venture capital industry, especially that part most familiar with high technology, in Silicon Valley. This complemented incubation facilities, research institutes, universities, business associations, and policy agencies that already existed. Lack of venture capital was the main barrier to innovation and providing conditions for supplying and augmenting it was a strategic innovation support decision in a context where a globally competitive window of opportunity was in danger of being closed off to Israeli businesses. The third stage of development coincided with the privatization in 1997 of Yozma. By that date, the number of venture funds had risen from one to seventy, many originating in the US, and some one billion dollars were available for start-up and later stage equity investment. As well as ties with the US venture capital industry becoming more pronounced, this was the point from which those that had reached the stage of making a successful initial public offering (IPO) began to be listed on NASDAQ, while US companies also acquired many non-IPO firms. This represented the fulfillment of two key strategic priorities articulated within Israel’s system of innovation. As noted, one related to a perceived need to enhance the effectiveness with which firms moved from start-up to subsequent “implementation” phases. The second imperative was to enhance the creation of global high technology firms. Firms also function as a cluster, with a specific security software sub-cluster. Of 19 such firms studied by Teubal, Avnimelech, and Gayego (2000), 47 percent were in Tel Aviv, and 41 percent in Haifa. They comment that: “The share of these companies in the overall information security sub-cluster is enormous and so is their impact on this [‘firewalls’] sub-cluster (they are leading firms and even ‘key agents’ in the sub-cluster creation process)” (Teubal et al., 2000:23). Thus, a model of “generative” growth was massively enhanced by initiating a public equity investment vehicle to facilitate exploitation and commercialization of an indigenous technology. A cluster with strong local and global value chain linkages as well as horizontal spillover and formal knowledge-exchange capabilities was established swiftly, in a breakthrough industry, in a spatial setting that had hitherto sustained basic research and transfer within the public realm (state military) but little spinout. As noted in Teubal and von Tunzelmann (2000), if the successful Silicon Valley commercialization process co-exists with significant poverty and social exclusion that they
REGIONAL TRANSFORMATION AND REGIONAL DISEQUILIBRIUM
53
paraphrase as a “High Tech & Homelessness” model, while Sweden limits the un-checked negative aspects of growth by social inclusion, then Israeli high tech clustering success seems to have been bought at the expense of serious social deterioration and something approaching civil war at home with horrific global implications abroad. In some ways, the same processes explain the major success of cluster development in the Indian software industry, though so far without the catastrophic social conflict coinciding with implementation of the breakthrough model in Israel. In answer to the question of how software took off in Bangalore, Balasubramanyam and Balasubramanyam (2000) offer the following insights. Bangalore has no clear path dependence (if that concept is to retain meaning) to its current situation. The city’s origins are a product, dating from the days of the British Raj, of urban environment and science infrastructure. At 3,000 feet above sea level, it was designed as a retirement resort for former army officers and civil servants. The British administration established there the Indian Institute of Science, specializing from the beginning in engineering and physics. It is home to Bangalore University, possessing fourteen engineering colleges, including nowadays software and computer engineering. There are also numerous stateowned technology-intensive defense and telecommunications facilities, attracted by human capital and distance from security-threatened frontiers. This more modern infrastructure, including state-owned corporations such as Hindustan Aeronautics has promoted software through the research and learning culture embedded in the city (Srinivas, 1997). Software expertise and localized demand led to the origins of the cluster, but it became globalized through the emigration of Bangalore software engineers to work in Silicon Valley and other centers, and the arrival of inward investment from such destinations to take advantage of extremely cheap software intellectual labor of high caliber. Thus, IBM, Hewlett-Packard, Oracle, Siemens, and Motorola all followed the pioneering investor, Texas Instruments. A national Software Technology Park was set up in 1991, beginning with a handful of firms, many that subsequently grew rapidly. By 1998, there were 183 units on site and over 200 software firms in Bangalore. By 2001, the US accounted for 60 percent of Indian software exports in an industry worth $8.3 billion in total. Reliance on selling to the US implies a high risk of induced downturn, but also the opportunity of increased US outsourcing. Dependence on the US market has changed from that in the mid-1990s when the US absorbed 90 percent of Indian software exports. The next stage of development, based on software industry interaction between India and the US, gave rise to a so-called “to-and-fro brain drain” involving temporary and permanent return migration. The latter typically resulted in the establishment of small software “bodyshops” contracted to
54
Philip Cooke
projects in customer premises, something declining with the growth of locally owned, larger software houses like Infosys, first listed on NASDAQ in 1999, and joint venture company PSI Data Systems. Another of these firms is Wipro, India’s largest measured by market capitalization. It listed on the New York Stock Exchange in 2000. These early, fast growth businesses thus became the leaders of the Bangalore cluster. They had a foreign stock-market listing, a global customer base, and a small but growing non-Indian workforce. Wipro and Infosys have operations in North America but also subcontract locally in Bangalore. The third stage of cluster evolution means diversification. For example, Bangalore hosts customer relationship management (CRM) software customer call centers such as 24/7 customer.com with customers like local Internet portal firm Rediff and US companies that include AltaVista and Compaq. In 2000, 24/7 raised $3 million from Indian Silicon Valley business angels and venture capitalists. Annual salary levels of $4,000-$6,000 give the cluster a significant competitive advantage. By 1998, the Bangalore Software Technology Park was responsible for 53 percent of Indian software exports. However, in true cluster-development style, Bangalore has begun to experience some bottlenecks or “diseconomies of agglomeration.” The first of these concerns bandwidth, since software and telecom data traffic increasingly requires greater amounts of this. As a less developed country, India has to prioritize investment and rely on its corporations to respond to demand rather than build for stock. Fears that the software industry could lose thousands of jobs led to a re-appraisal in 2000 and twelve new Internet gateways were established. There is a further infrastructural problem, which is that power supply is no longer adequate to meet growing demands from the software sector. Investment from US power-generating companies is being sought but so far, no new power supply deals have been signed. Finally, Bangalore’s international outlook means that human capital is perceived to be reaching a point when inward or indigenous investors cannot necessarily rely upon supply. This is one reason why Hyderabad is seen as a rising software cluster and Bangalore as possibly having reached a peak. The Indian co-founder of netdecisions.com – a UK headquartered international digital solutions provider – explained a decision to locate a second main center at Visakhapatnam on the east coast of Andhra Pradesh, where Hyderabad is state capital, in the following terms: “One of the challenges of setting up in Bangalore is that while it is a hothouse for skills and capabilities, people retention is a real problem. A lot of people go to Bangalore as a stepping-stone to moving overseas. If you are trying to build a sustainable presence in India, it actually doesn’t always serve you very well” (Counsell, 2001). Hence Bangalore’s early experience of two-way globalization and the “to and fro brain drain” remains part of its
REGIONAL TRANSFORMATION AND REGIONAL DISEQUILIBRIUM
55
economy culture, and just as close linkages with Silicon Valley-located, frequently Indian-owned equity investors and ICT firms were a notable force in the software breakthrough in India, so they are now so much part of its development trajectory that they may require some re-appraisal. Although the core elements of regional innovation, which are basic research and associated skills, entrepreneurship for commercialization and finance to translate innovation into productive businesses were all present in these two software breakthrough clusters in Israel and India, it is clear that the growth process was “generative” rather than “induced” in the Swedish manner. Thus, the role of state governance of innovation was directly facilitative in establishing Yozma in Israel and establishing a software science park in Bangalore as part of an India-wide program. However, there the direct support ended. In neither case were large indigenous firms the “anchors” of the developments that occurred, although inward investors established plants to access labor in Bangalore and acquired some of the Herzliah security software firms after they had reached second stage growth and venture capital was seeking to realize returns. Basic research was of greater importance technologically in Israel and for scarce, cheap high-grade human capital in India than it was initially in Sweden where large firms were more important knowledge generators and exploiters. Hence, in both cases, spinout firms formed cluster-like horizontal and vertical linkages in ways those in Swedish science parks have yet to achieve to the same degree. To conclude, both are more like Silicon Valley in their reliance on markets although they are, for different reasons also good examples of the operation of a “Silicon Valley offshore” model because of the lack of sufficient indigenous equity investment to make the key innovation breakthrough.
3.4.3 Skills Upgrading Approaches in Ireland If there is to be a shift from path dependence on a development trajectory that is visibly losing competitive advantage there has to be a breakthrough in the skills profile of the regional economy. This is especially so when the rising paradigm involves intense learning and knowledge generation of the technological and organizational kind discussed thus far. It can be said with some confidence that whenever rapid catch-up strategies have been pursued in economic history, the transformation of skills has been a key part of that process. Thus a lecture by Richard Lipsey referring to Bismarck’s education reforms, noted that: “Many observers credit the German trade schools, that educate the majority of German youths who do not go on to higher academic education, with providing the German comparative advantage in high quality standard consumer goods” (Best, 2000). The southeast Asian “tigers” clearly
56
Philip Cooke
learnt that lesson in substantially increasing the numbers of awards of natural science and engineering degrees, something also done by Ireland as Table 3.1 shows.
The Republic of Ireland’s surge in human capital investment did not have an immediate but a lagged effect upon economic performance. Thus Fuente & Vives (1997) show that although the point at which secondary and university enrolment as a percentage of the labor force rises significantly above the OECD average in 1975 the impact of the labor market factor upon Ireland’s differential growth rate did not become positive until 1990, though it was approaching neutral by 1985. The year 1990 is usually taken as the beginning of Ireland’s massive increase in GDP per capita performance over the OECD average. Other factors must also be accounted for, such as macroeconomic stability contingent on social partnership agreement on an employment and incomes pact, liberal inward investment policies, and the contribution of EU transfers, a large portion of which were devoted to human capital investment. Among the institutions invested in were new regional technology colleges where a deliberate policy of concentrating upon the education of young people in information technology (IT) and languages was pursued. This, added to increases in the training numbers for graduate engineers and scientists at upgraded and existing universities, meant Ireland’s supply of ITliterate labor began to outstrip its capacity to employ it. This led to a revival in out-migration of young people, a politically unpalatable result in an economy that had suffered such losses on a massive scale historically. A renewed effort was made to attract high technology inward investment assisted by a liberal inward investment incentive scheme. From 1990, Ireland’s annual rate of increase in exports rose from an annual average of 8 percent up to 1989 to 11 percent per year thereafter. This was largely accounted for by exports from inward investors. However, as we shall see, in one sector, again software, this also stimulated generative growth of indige-
REGIONAL TRANSFORMATION AND REGIONAL DISEQUILIBRIUM
57
nous capability that, more recently has also made a substantial economic contribution. Key to this is Dublin where much of Ireland’s new software industry clusters. The city has become one of Europe’s leading centers for software production, design, and export. From about 1985 Ireland became a major host to software inward investment with 9,100 jobs by 1997, of which 76 percent were located in Dublin. In per capita terms, Ireland became the world’s second largest exporter of software products after the USA. This stimulated growth in indigenous software production, now employing 9,200, mainly (71 percent) in Dublin. Much of the American product is adapted (including translated) to European markets. The Irish government strategy of establishing new and expanding traditional higher education institutions to create a workforce highly qualified in ICT engineering and languages has been enormously important to the development of the software industry. Thus, as many as 70 percent of personnel employed in the indigenous industry have graduate qualifications. This in turn created the basis for demand from niche firms for more customized than standard package software solutions and a round of new firm formation began in the 1990s. Private venture capital was initially less pronounced, because not widely available, than public venture capital and related enterprise funding. Thus 80 percent of indigenous software firms in Dublin are recipients of state aids for business development. Approximately half of that total has also accessed private investment, though the rest reported that private investment was not easy to find (O’Malley and O’Gorman, 2001). The fiscal environment in Ireland is favorable at only a 10 percent corporate tax rate for internationally traded services. Telecommunications infrastructure is modern and efficient for software transmission. Three specialist Programme in Advanced Technology (PAT) centers exist at Irish universities aimed at enhancing applied research and technology centers, and are used by some software firms. Interaction in the supply chain to exacting US customers like Oracle, Lotus, and Microsoft, based in Dublin, is propulsive in driving indigenous supplier firms up the learning curve. Large customers in software-using sectors like finance, pharmaceuticals, and the food industry created customized demand for flexible response from indigenous suppliers. Most founding entrepreneurs had previously worked in indigenous software companies or those in related industries like computing; only a third were previously employed in the foreign-owned sector. Finally, to return briefly to the illustration with which this contribution began, optical networking software in Northern Ireland, distinct but nevertheless linked to the Republic of Ireland, we see human capital upgrading of a different kind stimulating generative development. As noted, Northern Ireland has a path dependent, selective second level education system that al-
58
Philip Cooke
lows a significant share of its youth to specialize in science, engineering, and computer science. In the UK tradition, of which this is a survival, such graduates would move from the province to higher qualification occupations elsewhere in the UK, particularly London and other large city regions. As the old path dependence on shipbuilding and textiles (e.g. specialty linen) has atrophied, a regional breakthrough in telecommunications and internet software, including optical networking has been prioritized in the region’s new growth strategy. Involved in this is a restructuring and thinning-out of the development agencies, stimulation of academic entrepreneurship, completion of eight incubators for software and biotechnology, a second targeted sector, establishment of a regional science park network, and, finally, stimulation of equity investment, including public venture capital. Some one hundred spinouts in varieties of communications software have come into existence during the 1990s. Among these are firms originating in university incubators like QUBIS at Queen’s University in Belfast and the Jordanstown, Belfast, campus of Ulster University. Among the better known of these are Kainos, a sizeable spin-out from QUBIS that specializes in production engineering software with a strong capability for producing solutions for firms engaged in global production networks needing a fast response rate to re-design and order changes. Another outstanding Belfast software firm is MINEit, which in 2000 became only the UK’s second ever winner of the European Information Society Technologies awards first prize. Its Easyminer “worm” or locator software allows e-commerce websites to be tailored according to customer “click stream” behavior. Yet another leading firm is 8over8, a 2000 IPO attracting some £50 million in venture capital, including a majority equity share from an Italian investor (Cooke, Roper, and Wylie, 2001). Both parts of Ireland have moved from path dependence on declining or stagnating sectors, making regional breakthroughs by learning three key things that were also pronounced in the Swedish, Israeli and Indian examples. These are: Recognizing the crucial nature of advanced technological skills and the research and training organizations that produce them in order to have a chance of making a regional breakthrough, and escaping from growth barriers associated with path dependence. Learning a new innovation model that is not a planning or management model associated with redistribution policy designed centrally, but a generative model aimed at stimulating endogenous growth. This takes knowledge from the laboratory bench, and with venture capital or other forms of equity investment (including public), commercializes the results as innovative products and services. Adopting an approach that is open to and understands the fundamental message of globalization and the Silicon Valley profile, which is
REGIONAL TRANSFORMATION AND REGIONAL DISEQUILIBRIUM
59
that it consists in the integration of global value chains and regional clusters of knowledge-based economic activity. These, in brief are common threads, differentially approached, that have yielded breakthrough “generative growth” away from prevailing path dependent trajectories into new knowledge economies.
3.5 CONCLUSIONS In this contribution, an argument has been developed that suggests that the idea of path dependence is a species of technological determinism at heart. Critics may say that it carries with it connotations of socio-cultural and institutional trajectories that are resistant to change and have to be accommodated. However, upon detailed examination these constraints derive ultimately from a particular type and use of technology, broadly defined. The interactive organizational breakthrough of Silicon Valley was examined to identify the deep structures of the prevailing model. These were shown to be laboratory research transformed into innovations by firms through the application of entrepreneurship and equity investment skills in clusters. These clusters or local value chains were then integrated with global value chains, developing competitive advantage. Thereafter, and depending on institutional path dependence of a broad kind, it was then shown that, in different cases, one or other of the three key components of entrepreneurship, equity financing or skills upgrading was deployed in Sweden, Israel, India and both sides if the border in Ireland. This broke regions out of previous trajectories and into regional breakthroughs in knowledge economy sectors. In each new case, again for different reasons, welfare, wage, and employment improvements had occurred, though by no means perfectly, either in local or regional income distribution, or in relation to socio-political conflict caused by other forces in some settings. Spreading such breakthroughs to new socio-political settings was argued to be a sensible policy action line for governance structures to pursue because of its likely moderation of such polarization.
ACKNOWLEDGEMENTS This chapter was prepared for the International Workshop on “Rethinking Regional Innovation and Change: Path Dependency or Regional Breakthrough?” held at the Baden-Württemberg Academy for Technology Assessment, Stuttgart, February 28-March 1, 2002. I have benefited from discussions with many people in developing the views presented in this contribu-
60
Philip Cooke
tion. Chief among them have been Morris Teubal, Franco Malerba, Charles Edquist, and Jane Marceau who, in various ways, commented on a first project paper given at the DRUID summer conference in 2000. I also wish to thank Aidan Gough and Angela MacGowan of the Northern Ireland Economic Council, and Jim Wolstencroft and David Duncan of IRTU in Northern Ireland for confirming my evolving thoughts about the importance of equity investment to regional breakthrough in a difficult regional economic setting. Frédéric Richard of UNIDO influenced me on the importance of breakout from path dependence. Finally, Anders Malmberg kindly read the paper and advised on interpretations of the Swedish cases. I am grateful to all these colleagues, none of whom bears any responsibility for the final outcome.
REFERENCES Arthur, B., Increasing Returns and Path Dependence in the Economy. Ann Arbor: Michigan University Press, 1994. Asheim, B., Clark, E., Creativity and Cost in Urban and Regional Development in the New Economy. European Planning Studies 2001; 9:805-812. Bahrami, H., Evans, S., “Flexible Recycling and High Technology Entrepreneurship.” In Understanding Silicon Valley, M. Kenney, ed. Stanford: Stanford University Press, 2000. Balasubramanyam, V., Balasubramanyam, A., “The software cluster in Bangalore.” In Regions, Globalisation, and the Knowledge-Based Economy, J. Dunning, ed. Oxford: Oxford University Press, 2000. Best, M., The Capabilities and Innovation Perspective: The Way Ahead in Northern Ireland. Belfast: Northern Ireland Economic Council, 2000. Brown-Humes, C., Hiring Winter Sends a Chill to Sweden’s High Tech Heart. Financial Times 2001; May 10, 11. Castells, M., The Rise of the Network Society. Oxford: Blackwell, 1996. Cooke, P., Knowledge Economies: Clusters, Learning and Cooperative Advantage. London: Routledge, 2002. Cooke, P., Roper, S., Wylie, P., Regional Innovation Strategy for Northern Ireland. Belfast: Northern Ireland Economic Council, 2001. Cooke, P., Schienstock, G., Structural Competitiveness and Learning Regions. Enterprise and Innovation Management Studies 2000; 1:265-280. Counsell, A., Human Potential Drives Software Solutions. Financial Times 2001; February 21, xiv.
REGIONAL TRANSFORMATION AND REGIONAL DISEQUILIBRIUM
61
David, P., Technical Choice, Innovation and Economic Growth. Cambridge: Cambridge University Press, 1975. Dertouzos, M., Lester, R., Solow, R., Made in America. New York: Harper, 1989. Edquist, C., Systems of Innovation. London: Pinter, 1997. Florida, R., The Creative Class. New York: Basic Books, 2002. Fuente, A., Vives, X., “The Sources of Irish Growth.” In International Perspectives on the Irish Economy, A. Gray, ed. Dublin: Indecon, 1997. Gardner, D., Dotcommers Hunt for Alternative Jobs. Financial Times 2001; June 27, 14. Isaacs, A., Silicon Valley as a Case of Regional Innovation, presentation to UNIDO International Conference on “The Process of Innovation & Learning in Dynamic City-Regions,” Shenzen, China, 7-9 December, 2002. Jones-Evans, D., Klofsten, M., Universities and Local Economic Development. European Planning Studies 1997; 5:77-94. Jonsson, O., Innovation Processes and Proximity – the Importance of a Regional Milieu, Paper to Seminar “Science and Regions” Mölle, Sweden, June 19-20 (published in European Planning Studies, 10, 705-722), 2001. Latour, B., From the World of Science to the World of Research? Science 1998; 280:208-209. North, D., “Institutions and Economic Performance.” In Rationality, Institutions and Economic Methodology, U. Mäki, B. Gustafsson, C. Knudsen, eds. London: Routledge, 1993. O’Malley, E., O’Gorman, C., Competitive advantage in the Irish indigenous software industry and the role of inward foreign direct investment. European Planning Studies 2001; 9:303-321. Porter, M., On Competition. Boston: Harvard Business School Press, 1998. Porter, M., Takeuchi, H., Sakakibara, M., Can Japan Compete? London: Macmillan, 2000. Porter, M., The Competitive Advantage of Nations. New York: The Free Press, 1990. Saxenian, A., Regional Advantage. Cambridge: Harvard University Press, 1994. Srinivas, S., The Information Technology Industry in Bangalore: a Case of Urban Competitiveness in India? Paper to Asian Urbanization Conference, London, 1997. Sturgeon, T., “How Silicon Valley Came To Be.” In Understanding Silicon Valley, M. Kenney, ed. Stanford: Stanford University Press, 2000. Teubal, M., Avnimelech, G., Gayego, A., The Israeli software industry: analysis of the information security sector, Proceedings DRUID Summer Conference on The Learning Economy – Firms, Regions and Nation Specific Institutions, Aalborg, Danish Research Unit for Industrial Dynamics, 2000. Teubal, M., The systems perspective to innovation and technology policy: theory and selected topics. (mimeo), 2000.
62
Philip Cooke
Teubal, M., von Tunzelmann, N., Varieties of Capitalism and Policy Reform. (mimeo), 2000. VINNOVA, The Swedish Biotechnology Innovation System. Stockholm: VINNOVA, 2001. Zucker, L., Darby, M., Brewer, M., Intellectual Human Capital and the Birth of US Biotechnology Enterprises. American Economic Review 1998; 88:290-306.
Chapter 4 SWITCHING TIES, RECOMBINING TEAMS: AVOIDING LOCK-IN THROUGH PROJECT ORGANIZATION?
Gernot Grabher
4.1 INTRODUCTION 4.1.1 From path dependency to lock-in Ironically, while economists can still embrace the crude Darwinism of Spencer’s “survival of the fittest,” biologists for some time have challenged the received evolutionary model (see, for example, Smith, 1984; Gould and Lewontin, 1984; Dupré, 1987). In their view, evolution cannot simply be regarded as a one-dimensional process of optimization, a beneficent and unilinear journey from the lower to the higher form of organization. Evolution, in this sense, does not proceed along a single grand avenue towards perfection but along multiple paths which do not all lead to optimal change. The fact, that some developmental paths produce ineffective solutions and suboptimal outcomes, is not an indication of evolutionary failure but a precondition for evolutionary selection: no variety, no evolution. The evolutionary process thus necessarily entails development through failure: “imperfections are the primary proofs that evolution has occurred, since optimal designs erase all signposts of history” (Gould, 1987:14). This fundamental critique of the “survival of the fittest” paradigm, challenges (neoclassical) assumptions of “historical efficiency” in which survival implies efficiency and mere existence proves optimality (see, for example, Hodgson, 1993; Foster and Metcalfe, 2001). Competition, as a first lesson drawn from this evolutionary reasoning, does not necessarily favor the fittest and more efficient form of organization: competition, in other words, is not an optimizer. Evolutionary theory, moreover, brings into focus how the future development of an economic system is affected by the path it has traced in the past. In fact, the idea of path-dependency rejects the notion that “from
64
Gernot Grabher
whatever starting point, the system will eventually gravitate to the same equilibrium” (Hodgson, 1993:204). Positive feedback, in this perspective, reinforces negative tendencies. This view on path dependencies has been conceptually elaborated and empirically supported particularly in studies on the “lock-in” syndrome, most convincingly exemplified in the account on the QWERTY keyboard layout (Arthur, 1989 and 1994; David, 1985 and 1992). These studies raise the possibility that free markets might lock-in economic development to a particular path that does not gravitate to the optimum (see also Hodgson, 1993:204). Increasing returns from learning effects and network externalities (Katz and Shapiro, 1985) yield real immediate benefits that can preclude selection in the long run of the most efficient organizational form or technical design (Arthur, 1989 and 1994; David, 1985 and 1992; Carroll and Harrison, 1994). Once an economy is locked into a particular trajectory, the costs of shifting strategies outweigh the benefits of alternatives. This approach to economic history, generally speaking, stresses the possibility that the very mechanisms that foster allocative efficiency might eventually lock-in economic development to a path that is inefficient viewed dynamically. At the same time, the mechanisms that are conducive for the synchronic adaptation of the economy to a specific environment may undermine an economy’s diachronic adaptability. This dialectic between adaptation and adaptability also echoes, more or less clearly, through studies of the (rise and) decline of old industrial areas (see, for example, Tichy, 1984; Grabher, 1993 and 1994; Glasmeier, 1994; Hudson, 1994; Morgan and Nauwelaers, 1999). The concept of “regional lock-in” has been explored more specifically in some detail against the background of the crisis of the Ruhr area (Grabher, 1993 and 1994; see also Hassink, 1997 and 2002) and the dual economic trajectories of BadenWürttemberg (Braczyk, Schienstock and Steffensen, 1995). The lock-in syndrome evolves from a circular amplification of three interrelated processes. First, close inter-firm relations embedded in long-standing personal ties resulted in serious shortcomings in so-called boundary-spanning functions that are essential for making external information relevant for the firm. Second, the high degree of cohesion favored the emergence of common orientations and perceptions that ultimately culminated in a hegemonic worldview that locked the region cognitively in the narrow interpretative frames of “groupthink” (Janis, 1972). Third, regional political intervention kept the region effectively on its path, even when this path became a dead-end. The political “culture of consensus” turned into cohesive coalitions against major changes in which status was privileged over knowledge and power over learning (see also Morgan and Nauwelaers, 1999; Narula, 2002).
SWITCHING TIES, RECOMBINING TEAMS
65
4.1.2 Antidotes against lock-in While this canonical story on how regional trajectories lead into an evolutionary dead end might appear plausible, it seems far less obvious how these positive feedback dynamics can be avoided. In reversing the basic causalities of the positive feedback of lock-in, however, some generic ingredients for regional adaptability, in other words, for keeping regional trajectories open and reversible can be singled out. Diversity. Cognitive coherence and organizational homogeneity increase the liability to lock-in (see also Uzzi, 1997). In general, for evolution to work there must always be a variety of forms from which to select. “Selection is like a fire that consumes its own fuel ... unless variation is renewed periodically, evolution would come to a stop almost at its inception” (Lewontin, 1982:151). Diversity allows evolution to follow, at the same time, different paths that are associated with different sets of organizational forms. When selection starts not from a single trajectory but from a broad and diverse range of evolutionary alternatives, the risk decreases that local maximization results in an evolutionary dead end. The proliferation of a broad spectrum of different organizational forms and diverse practices - as opposed to the diffusion of a single “best practice” - provides a richer “selection environment” for regions, firms and individual actors to co-evolve (Wagner and Altenberg, 1996; Grabher and Stark, 1997:536; see also Rantisi, 2002:591). Redundancy. The reproduction of diversity depends on the ability of different levels of efficiency to co-exist. The co-existence of different levels of efficiency, in fact, is a precondition for evolution since “the rate of increase in fitness of any organism at any time is equal to its genetic variance in fitness at that time” (Fisher, 1930:35). Adaptability thus presupposes a certain tolerance of inefficiency or, phrased in organizational terms, of slack and redundancy (Cyert and March, 1963:38; see also Arthur, 1999). Redundancy of business models, philosophies and organizational forms and practices provides for a richer genetic pool for the evolution of new organizational mutations. Rivalry. In addition to the diversity of organizations, adaptability indeed crucially depends on the organization of diversity (see also Morin, 1974:558). Organizational diversity is most likely to yield its evolutionary potential when different organizational forms and principles coexist in an active rivalry. Such rivalry fuels a sustained engagement that recreates different ways to organize, interpret, and evaluate the same or similar business activities. Rivalry reproduces the diversity and richness of the genetic pool of organizational forms. Rivalry, phrased differently, fosters cross-fertilization and recombination of organizational forms and practices (Jacob, 1977:1163; Holland, 1992:26; Grabher and Stark, 1997; Stark, 1999:74).
66
Gernot Grabher
Loose coupling. Very strong and dense network relationships facilitate the development of uniform subcultures and strong collective identities. A particularly dense and tightly coupled network, however, might promote cohesiveness while hindering, as the emblematic stories of the regional lock-in elucidate, the ability to gain information and mobilize resources from the environment (Weick, 1976; Uzzi, 1997). Network research posits an inverse relationship between the density of coupling of network ties and their openness to the environment. In particular the “weak ties” (Granovetter, 1973) indirectly connecting actors or bridging the “structural holes” (Burt, 1992; Ahuja, 2000) that become obligatory “passage points” (Latour, 1988) between relatively isolated groups of actors are crucial for the adaptability of networks (see also Albert, Jeong and Barabasi, 2000).
4.1.3 Aim of the chapter This chapter seeks to outline a form of social organization that, at first glance, appears to produce at least some of these evolutionary antidotes against lock-in: the project ecology (Grabher, 2002a, b, c). Project ecologies provide the social and organizational fabric for temporary and recurrent collaboration in projects. Despite dense patterns of interaction, project ecologies seem not to evolve into a stable, coherent, and tightly coupled institutional network. Due to the transience of relationships, project ecologies rather resemble a heterarchic form of social organization (see Hedlund, 1986; Hedlund and Rolander, 1990; Hedlund, 1993; Hedlund, 1994; Stark, 1999; Grabher, 2001). To what extent can project ecologies then be described in terms of diversity, rivalry, redundancy, and loose coupling? The chapter aims at answering this question by empirically exploring the project ecology of the British advertising industry and its center of gravity, London, Soho1. Although film production represents the emblematic project-based media industry (see, for example, Faulkner and Anderson, 1987; DeFillippi and Arthur, 1998; Jones, 1996; Sydow and Staber, 2002), a study of the advertising business offers insights from an industry in which project organization only more recently has become part of the canonical repertoire of organizational forms (see also Ekstedt et al., 1999). The Soho advertising business played a particular role by challenging more Fordist “Madison-Avenue” style advertising through a new breed of creative agencies which pioneered momentous innovations in project organization (Lash and Urry, 1994:138-142; Leslie, 1997:1022-23). The chapter ventures into the organizational and social ecology in which advertising projects typically are embedded. It subsequently examines the firms, personal relations, locali-
SWITCHING TIES, RECOMBINING TEAMS
67
ties, and corporate networks around and through which projects typically are built.
4.2 THE FIRM LEVEL: PROFESSIONAL RIVALRIES Basic definitions of projects also capture the essence of organizing production in advertising by stressing, first, the temporal limitation of collaboration and, second, the diversity of skills involved in accomplishing a complex task (for the locus classicus, see Goodman and Goodman, 1976:494; see also Lundin and Söderholm, 1995; Lundin and Midler, 1998). More specifically, a stylized version of the production process in British “second wave” advertising is a collaborative effort between account management, account planning, and “creatives” (creative professionals). The prime function of account management is to liaise with the client from the preparatory pre-project stages to the completion of the project. In fact, as trade jargon goes, the account manager is the client in the agency (Quinn, 1999:30). The world of the account manager is ideally organized around the co-ordinates of briefs, deadlines, budgets, brand share; in short, (s)he incorporates the business ethos of advertising. As much as account management represents the client perspective in the agency, account planning brings the consumer into the production process (Cooper, 1998). The task of account planning is to survey “ways of using” and to articulate what practices of consumption mean to those engaged in them (Leslie, 1999:1447). To the extent that research methods and criteria provide legitimization for the task, account planning incorporates the scientific ethos of advertising. In contrast, the professional ethos of creatives is rather indirectly tied to the business and scientific logics of the industry. The central creative imperative is the “freshness” of ideas (Wells, Burnett and Moriarty, 1998:381). Although it is not “art” that creative teams are or should be doing, it should comply with the conventions, standards, and styles of the community of the creatives (Bullmore, 1999:52). To the extent this community overlaps with the art world, creatives incorporate the artistic ethos of advertising. The high intra-organizational diversity within British agencies provides a broad interface for processes of mutual learning. The ongoing confrontation of difference in ethos in the course of the project, however, also provides causes for power struggles and rivalry. The most notorious sources of rivalry are located in the contested terrain between the realm of the creatives and the account planners, between the artistic and scientific ethos. Creatives are anxious to defend their work as expressions of authentic “emotional” creation against the “rationalistic” imperialism of the account planners who, in the
68
Gernot Grabher
creatives’ view, extract emotions out of creative ideas (Shelbourne and Baskin, 1998:78). A most effective organizational device to limit rivalry between the artistic and scientific logics is the deadlines of projects. That is the point in time when dispute has to be settled, polyphony has to be filtered into a single voice. Deadlines of projects, on the one hand define a temporary organizational arena in which the limited duration of the collaboration in the project preserves the identity of different logics (Lindkvist, Söderlund and Tell, 1998:948). In fact, they meet an organizational imperative according to which creatives, account managers and account planners, despite intense collaboration “should remain strangers to some extent” (MD9)2. Temporary limitation prevents any single perspective from becoming corrupted by a hegemonic view. In other words, deadlines provide antidotes against lock-ins into particular cognitive or aesthetic patterns. On the other hand, the limited duration of projects contains the rivalry between the different logics from gaining too much disintegrative momentum. By temporally limiting the centrifugal drift of the different logics, deadlines, perhaps, define the narrow edge between creative and uncreative conflict.
4.3 THE LEVEL OF PERSONAL RELATIONS: TIGHT AND LOOSE COUPLING The comparatively high diversity of professional profiles involved in British agencies corresponds not only with a respective degree of internal differentiation of the project along the lines of professional ethos, culture, and status. Concurrently, this implies that projects are embedded in a broader range of environments each of which offering different resources and representing different “communities of practice” (Wenger, 1998). In fact, ties and loyalties of creatives, account managers and account planners with their respective communities and project partners appear at least as binding as current agency affiliations: “We are loyal to the people rather than the company (MD12),” goes the business mantra.
4.3.1 Client networks: aiming at lock-in Particularly in larger agencies, accounts are based on two-to-three-year contracts with provisions for renewal on an annual basis. Yet, the notion of the contract implies a degree of formality, binding force and egalitarian status of
SWITCHING TIES, RECOMBINING TEAMS
69
contracting partners that hardly squares with the practice of the advertising business (Ghosh and Taylor, 1999). The major function of contractual arrangements in regulating client-agency relationships indeed appears of limited substance: “By and large it’s a bit like marriages. Having a contract really only provides the mechanism of how to deal to get divorced, how to get paid” (MD1). In consequence, “it’s a strange combination of sort of long term, but also has no security to it” (AP2). The long-term dimension in these relations is anchored in the ties between marketing manager on the client side and account manager on the agency side. The relative robustness of these personal ties vis-à-vis the rather fragile contractual agreement is revealed in the common practice of account managers who take along “their” client when they leave an agency. Conversely, marketing managers tend to stick to their established relation with an account manager in case they change their company. The strength of these mutual loyalties is rooted in the “communicative thickness” of the relationship between account manager and marketing manager, which essentially involves ongoing processes of clarification, modification, re-specification, and deflection of the client’s intentions, unintentionally as well as intentionally. Rather than a process of implementation, this relationship engages in an ongoing process of “translation” (Latour, 1986:267). Once accounts are placed with an agency, the relation with the client thus increasingly evolves into performing projects with clients rather than of realizing projects for clients (see Girard and Stark, 2002). By strengthening personal ties and accumulating client-specific expertise, the agency, in a sense, aims at “locking in” the client.
4.3.2 Creative networks: avoiding lock-in A different social logic lies underneath the project-based collaboration between in-house creative teams and external collaborators. The ethos of creatives also incorporates the professionalism of the execution of the idea for the campaign. Quality of the execution, however, is less an attribute of the personal mastery of graphic crafts but reflects skills and knowledge in selecting most “talented” project partners from an excessively fragmented and specialized pool of professional specialists (Sellers, 1999:62). This selection procedure is shaped by experience and affected by the shadow of potential future collaboration and, over time, fills a pool of preferred project partners, each of which represents a particular specialization. Patterns of collaboration with outside project partners thus reflect preferences of the in-house creative team for certain directors each of which “has its distinct flair” (CR23) or certain
70
Gernot Grabher
photographers renown for their particular style in narrowly understood subgenres (Wells, Burnett and Moriarty, 1998:392-393). The governance of relations in the creative realm of advertising projects resembles practices in jazz improvisation, a “prototype organization” designed to maximize innovation (Barrett, 1998:607-616; Weick, 1998; Hatch, 1999). Improvisation, essentially, implies a deliberate interruption of habit patterns and resistance to the temptation to become locked in to routines of past success, thereby squelching experimentation. One of the most widespread practices in jazz is “taking turns” that is swapping back and forth the roles of soloing and supporting other soloists thereby rotating “leadership” within the band. In advertising projects, in particular the collaboration between the creatives and the film director is characterized by a similar migration of control. Just as jazz bands typically vary their composition from time to time, only to reappear in their original line-up, teams in the creative realm of advertising production re-group from project to project but reappear recurrently as a reconfiguration around a relatively stable set of core relationships. This variance in composition reflects, on the one hand, obvious particularities of a specific project such as the demand for a different skill set for poster, radioor TV-commercials. On the other hand, composition of project teams is deliberately altered by the creative teams, reflecting the core imperative of their professional ethos, as an art director explains: “you work with your favorites ... but you also try new people, because of new ideas, new approaches ... you look for freshness” (CR7; see also Usai, Delmestri and Montanari, 2001). The deliberate alteration of teams, in other words, aims at enhancing “learning by switching” (Dornisch, 2002). Such diffuse “learning by switching” evinces itself through decoupling occurring within particular projects and through the competition motivating movements from project to project. Learning by switching, potentially at least, breaks up collaborative dead-ends and interrupts positive feedback loops.
4.4 THE LOCAL LEVEL: REDUNDANCY OF PRACTICES Spatial proximity in collaborative relations and the degree of specialization, in general, seem inversely related in the trade. Whereas networks in the technical and the lower end of the creative realm are spatially more limited, ties with specialized directors or photographers reach out to international creative centers like New York or Los Angeles (see also Nachum and Keeble, 1999:22-25). In London, however, the entire spectrum of activities associated
SWITCHING TIES, RECOMBINING TEAMS
71
with advertising activities ranging from printing, lithography, graphic design, photography and music to film direction, production and post-production is concentrated in a tiny district of roughly one square mile and the directly adjacent areas.3 Soho, the center of gravity of what is colloquially labeled “adland” or “ad village,” offers a pool of creative talent and potential project partners whose richness is unmatched in Europe and probably only surpassed by New York (Nachum and Keeble, 1999:21-24 and 2000:11; Miller et al., 2001:65).
4.4.1 Accidental interaction Without doubt, this Marshallian-style co-location of specialists and collaborators offers benefits. It reduces different variants of transaction costs (Krugman, 1995); allows for a quintessentially metropolitan 24/7-pace of face-toface interaction between collaborators (Sassen, 1995); and facilitates the strategic monitoring of competitors (Storper, 1997; Malmberg and Maskell, 2002:438; Pratt, 2001). These lines of reasoning strongly focus on various types of efficiency gained through geographical clustering. Adaptability, however, is promoted through social processes that do not follow a narrow economistic logic. “Being there” (Gertler, 1995) also induces interaction that, rather than by strategic purpose, simply is prompted by the principle of neighborhood. Neighborhood acts as a sort of random generator that also brings into contact actors that lie outside the reach of (global) pipelines that are set up for purposes of strategic monitoring (see also Owen-Smith and Powell, 2002). Actors are not only deliberately “scanning” their environment in search of a specific piece of information but are also surrounded by a “dense environmental texture” (DeLillo, 1984:169) of rumors, impressions, recommendations, trade folklore, and strategic misinformation (see also Pratt, 2001; Thrift, 1994). Variations in the level of such “noise” is indexing significance (see Brown and Duguid, 2000a:244) when, for example, rumors grow or die down. The point about co-location, in fact, is not the richness and diversity of the noise as such. Rather, clustering facilitates the emergence of “interpretive communities” (Brown and Duguid, 1991) which filter and process noise into patterns of signals. Face-to-face contacts and “hanging out” together facilitate casual conversation that allows expansion of contacts and knowledge about other actors (see also Sarbaugh-Thompson and Feldman, 1998:693). Casual conversation keeps open communication channels so that they are available for substantive purposes such as the “triangulation” of rumors or gossip, e.g. the availability of a particular project collaborator. Taken together, in addition to the mere richness of noise and information, processes of “negotiating meaning” tie
72
Gernot Grabher
project clusters together (Brown and Duguid, 2000b:22; Grabher, 2002a:209; Christopherson, 2002).
4.4.2 Permeability of boundaries The consistent demand for fresh perspectives opens the door, at least a stretch, for people without a specialist professional training background. As a consequence of the low formal barriers to entry and low segmentation of the labor market, the spectrum of career patterns and biographical paths is extraordinarily broad and diverse. Closely related to this labor market structure is the permeability of the border between trivial culture and “high art” which appears to be unusually high in London. In other words, migration of people, organizational and aesthetic concepts between these worlds is not suppressed by rigid cultural attributions. This in particular holds true for the relation between advertising and film production (Lash and Urry, 1994:138-142). The relative ease of migration between different communities of practice and genres facilitates exchange and catalyses novel combinations of established routines and perceptions. The practice of project-based collaboration in Soho, thus, maximizes recombinatory options between a diverse range of skill sets, biographical backgrounds, and cultural orientations. Yet, the limited duration of collaboration prevents identities from becoming blurred and indifferent, in other words, it preserves those rivalries that ignite processes of deeper mutual (self-) understanding and reflection.
4.4.3 Rivalry between forms Rivalry not only refers to the obvious project team vs. project team-, agency vs. agency-competition. In addition, rivalry comes to the fore in the contested terrain of boundaries between professions, project teams, organizations and, in fact, in the understanding of the sub-sectors of the trade. Reflecting on a recent shift in boundaries in Soho, account planning has not only been introduced, modified, adopted, or rejected in agency-specific ways as a new employment profile within agencies (Rainey, 1998). Moreover, account planners in London have started organizationally independent account planning agencies and loosely associated networks that offer their services to advertising agencies. Instead of understanding the ad village in terms of the classical dramaturgy of invention/diffusion/adoption of the one “best way,” ongoing rivalry between different ownership forms and organizational practices produces a
SWITCHING TIES, RECOMBINING TEAMS
73
multiplicity of best practices. The traditional set of independent owner agencies on the one hand and agencies, which are affiliated with global groups on the other, is complemented by two other forms. First, the London ecology also hosts agencies that are part of rather loose federations of agencies, which, without any cross-ownership, collaborate episodically on a project basis. Second, in 1995 the world’s first employee-owned agency, St. Luke’s, was founded in London. Although St. Luke’s does not have the status of a “model,” its commercial and creative success turned the philosophy of the agency into a serious “what can we learn from” issue. In this way, the “angry young agency” which swept aside conventions of the ad village, ironically, played an important role for Soho, if not by intention though. By turning the viewfinder, from ad village’s point of view, to unknown corporate territories, and unorthodox business practices, the agency pioneered new forms of clientcentered project-organization. More generally, the sustained rivalry between various ownership and organizational forms and the continuous arrival (and disappearance) of new maverick agencies enriches the pool for potential new recombinations of projects across the contested boundaries of projects, agencies and the ad village.
4.5 THE CORPORATE LEVEL: LOOSE INTEGRATION IN GLOBAL NETWORKS The organizational space of projects is nowadays rarely confined to personal relations and firms in a specific locality. Projects increasingly are embedded in the context of international and global corporate networks. In the advertising business, international corporate networks literally are coined “networks.” These networks in turn, at an even accelerating pace, are integrated into global “communication groups.” The three major groups, WPP, Interpublic, and Omnicom, account for some 40 per cent of the world top 100 advertising organizations’ gross income and report growing market shares (Campaign, 23 March 2001; Advertising Age, 29 May 2001; Interpublic, Omnicom, WPP Annual Report, 2000, 2001).
4.5.1 Diverse portfolio In a striking similarity, the three major groups are built around three major advertising networks all of which are performing projects and handling accounts for different clients or for different brands of the same client. As much
74
Gernot Grabher
as the groups aim to keep their own corporate profile as low as possible, as much they are reiterating the brand identities of their networks (see Wells, Burnett and Moriarty, 1998:119). The resulting diversity of brands within the group allows a resolution of the conflict rule of “exclusivity” according to which an agency is unable to serve the business of directly competing clients. In addition, the rather loose integration into groups secures the identity of individual agencies. It thus reduces the risk for agencies that have been taken over to lose key employees who would refuse to work for a “Madison Avenue behemoth.” Due to the strong mutual loyalties of the project teams with clients, as already elaborated, the danger of losing clients with the departure of personnel is particularly high. The structural similarity of the major groups, however, is not limited to their core business that rests on three advertising networks. In parallel moves, they have diversified into non-advertising marketing and communication services, which, in 2000, accounted for 55-58 percent of the three majors’ income (2000 Annual Reports of Interpublic, Omnicom, and WPP). While the groups traditionally have had public relations arms, they more recently moved, with rather different priorities and organizational strategies, into market research, strategic marketing consulting, identity consulting, direct marketing, in-store marketing, health care marketing, design services, new media, and other specialist services. In order to integrate these diverse services into specific packages, groups have to meet the managerial challenge to create and sustain a client-focused project for a given period of time that cuts across the boundaries and identities of the involved teams, agencies, and networks of the group.
4.5.2 Internal rivalry Rivalry appears a more appropriate concept for understanding the internal organizational logic of these diversified groups than competition since the latter takes the boundaries of the firm as well as the boundaries of its internal units as given parameters. As such it might capture some of the more obvious dimensions of rivalry when, for example, the leading networks within the groups directly “pitch” against each other to win new accounts. However, more critical than this classical firm vs. firm competition are those dimensions of rivalry that challenge the boundaries of the individual networks, agencies, and other organizational units involved. With annual growth rates in excess of 50 percent until early 2001 (WPP Annual Report), new media services have triggered unprecedented rivalry within the groups about how to approach new business opportunities. On the one hand, although at decreasing levels, the groups invested in new media
SWITCHING TIES, RECOMBINING TEAMS
75
start-ups. On the other hand, the group’s leading advertising networks pursued the same agenda by spinning off and broadening their own internet operations. This is rivalry within the group: new business opportunities are approached from both, classical advertising agencies who are broadening their business domain as well as from specialized businesses who are bundling their resources in order to defend, of what they regard, as their domain. The result, however, is not a once-for-all “either/or” solution but a sustained engagement that recombines different ways to organize, interpret and evaluate the same or similar business activities. Instead of regarding the restructured organization as the result of the search for the best solution, the result is rather an organizational configuration that is better at search (see also Stark, 1999).
4.6 CONCLUSION: THE AMBIVALENCE OF SWITCHING The interdependencies between a particular project and the firms, personal relations, localities and corporate networks from which these projects mobilize essential sources, unfold the relational space of a particular project ecology. Project ecologies, phrased differently, provide the social and organizational fabric for temporary and recurrent collaboration in projects. This fabric is not fixed to a specific physical place, to a distinct “island of innovation” (see Amin and Cohendet, 2003). Project ecologies rather unfold a relational space in which local ties are linked up with the wider networks of corporate organization, clients, and project collaborators. Project ecologies, in other words, are not a territorial innovation model based on a stable, coherent, and tightly coupled institutional network. Due to the transience of relationships, project ecologies rather resemble a heterarchic form of social organization (see Hedlund, 1986; Hedlund and Rolander, 1990; Hedlund, 1993; Hedlund, 1994; Stark, 1999; Grabher, 2001). However, do the heterarchic features immunize project ecologies against lock-in? Without doubt, heterarchies produce a whole range of critical ingredients for adaptability that prevent regional trajectories from turning into an evolutionary dead end. First, in project ecologies the temporary limitation of collaboration preserves the diversity of particular professional and organizational identities involved. The limited duration of collaboration in a project prevents the perspectives embodied in the various professional ethos’ involved from becoming corrupted by a hegemonic view. The continuous emergence (and disappearance) of novel business philosophies and maverick agencies reproduces diversity of organizational practices and forms in the
76
Gernot Grabher
village. Neighborhood does not lead to a quasi-epidemic spread of a single “best practice” but rather triggers (agency) specific ways of adoption, recombination, or outright rejection. Within the corporate group, diversity within the portfolio of agencies and business activities is strategically fostered to avoid account conflicts and the loss of creative talent (and, with them, their established ties with key clients). Second, rather than drifting towards cognitive homogeneity and organizational coherence, diversity at all levels of the project ecology is reinforced and further differentiated by rivalry. The sustained rivalry between different ways to organize, interpret, and evaluate the same business activity, in this view, enriches the pool of potentially new recombinations of projects. Somewhat in dissonance with the harmonizing tone echoing particularly through the current learning-discourse (see also Hudson, 1999), diversity and rivalry are not reduced to notorious causes of disturbance and interruption in anyhow high-pressured organizational settings. Instead, they are seen as an indispensable element of project-based creativity and organizational evolution. Third, instead of tightly coupled network configurations, project ecologies recombine loose and tight patterns of interaction. In the course of projects, rather close reciprocal ties of collaboration alternate, to some extent, with loose relationships and arm’s length market transactions at project peripheries. The organizational logic of projects, at least in highly creativitydriven contexts implies the deliberate switching of relationships to break up potential collaborative dead ends. In project ecologies, learning processes do not necessarily occur through sustained interaction but through “switching” in which minimal connectivity between projects is present. Such diffuse “learning by switching” (Dornisch, 2002) evinces itself through decoupling occurring within particular projects and through the competition motivating movements from project to project. Learning by switching, potentially at least, breaks up collaborative dead ends and interrupts positive feedback loops. Fourth, the adaptability of project ecologies is enhanced by the redundancy of business models, philosophies, and practices. Redundancy, however, seems confined to the local and corporate level – both of which are characterized by a wealth of resources, skills, and organizational forms. This sharply contrasts with the individual project stripped of its social context. In fact, the project is symptomatically designed to minimize any redundancies and organizational slack: “there is no fat at all in [this] system” (Lash and Urry, 1994:124; see also Hobday, 2000:888). Whereas the social context of projects reproduces redundancy, the project itself thus represents a hyperefficient form of organization that aims at a problem-specific and temporally limited use of resources. The project appears as the ultimate lean organiza-
SWITCHING TIES, RECOMBINING TEAMS
77
tion, unburdened by the costs and organizational slack involved in the reproduction of recombinatory options. This duality between “lean” project and “fat” context, however, exposes project ecologies to the cardinal risk of temporary collaboration: learning and training effects accruing in a particular project might evaporate after project completion (Ekstedt et al., 1999; Hobday, 2000). Whereas “learning by switching,” in other words, potentially breaks up collaborative dead ends, excessive “switching” could undermine the sedimentation of skills and competencies and thus erode an important source of creativity and evolution. Project ecologies, in conclusion, produce evolutionary anti-dotes against lock-in. Notwithstanding this evolutionary strength, however, project ecologies do not represent a “best-of-best” configuration. Long-term adaptability is not ensured in a self-propelling and universal fashion. In particular, slack and redundancies provide organizational buffers for the “incubation” of new projects as well as repositories in which knowledge sediments and is transferred from project to project. To avoid their erosion, they have to be shielded from a narrow orientation on the efficiency of the individual project that neglects the reproduction of the wider project ecology.
ACKNOWLEDGEMENT The empirical account on the British advertising industry is a revised and shortened version of the paper “The Project Ecology of Advertising: Tasks, Talents and Teams,” published in Regional Studies, 36(3):245-262.
NOTES 1.
Methodologically, this chapter is based on 68 semi-structured interviews in advertising agencies and collaborating film- and post-production companies, which, on average lasted 2 hours and have been conducted between Spring 1998 and Summer 2000 in central London. Information was supplemented by a variety of secondary sources including interviews with representatives of the Institute of Practitioners in Advertising (IPG) and the Account Planning Group (APG) and data from industry reports, trade press, business reports, and press releases.
2.
Quotations from agency interviews are marked according to the professional specialization of the interviewee broadly defined with MD (managing director), AM (account management), AP (account planning) and CR (creatives) followed by the consecutive number of the agency in the sample.
78
Gernot Grabher
3.
The area loosely referred to as Soho is defined by the following boundaries: Oxford Street to the north, Regent Street to the west, Charing Cross Road to the east, and the south side of Leicester Square to the south.
REFERENCES Albert, R., Jeong H., Barabasi A., Attack and error tolerance in complex networks. Nature 2000; 406:378-82. Ahuja, G., Collaboration networks, structural holes, and innovation. Administrative Science Quarterly 2000; 45:425-55. Amin, A., Cohendet, P., Knowledge in action: communities and competencies in the learning organization. Oxford: Oxford University Press, 2003 (forthcoming). Arthur, W.B., Competing technologies, increasing returns, and lock-in by historical events. Economic Journal 1989; 99:116-131. Arthur, W.B., Increasing returns and path dependence in the economy. Ann Arbor: University of Michigan Press, 1994. Arthur, W.B., “The end of economic certainty.” In The biology of business: decoding the natural laws of enterprise, J. H. Clippinger, ed. San Francisco: Jossey-Bass, 1999. Barrett, F. J., Creativity and improvisation in jazz and organization: implications for organizational learning. Organization Science 1998; 9:605-622. Braczyk, H.-J., Schienstock G., Steffensen B., “The region of Baden-Württemberg: a post Fordist success story?” In Industrial Transformation in Europe, E. J. Dittrich, G. Schmidt, R. Whitley, eds. London: Sage, 1995. Brown, J.S., Duguid, P., Organizational learning and communities of practise: towards a unified view of working, learning, and innovation. Organizational Science 1991; 2:40-57. Brown, J.S., Duguid, P., The social life of information. Boston: Harvard University Press, 2000a. Brown, J.S., Duguid, P., “Mysteries of the region: knowledge dynamics in Silicon Valley.” In The Silicon Valley edge: a habitat for innovation and entrepreneurship, C.-M. Lee, W.F. Miller, M.G. Hancock, H.S. Rowen, eds. Stanford: Stanford University Press, 2000b. Bullmore, J., “The advertising creative process.” In The advertising business, J. P. Jones, ed. London: Sage, 1999. Burt, R., Structural holes: the social structure of competition. Cambridge: Harvard University Press, 1992. Carroll, G.R., Harrison, J.R., On the historical efficiency of competition between organizational populations. American Journal of Sociology 1994; 100:720-749.
SWITCHING TIES, RECOMBINING TEAMS
79
Christopherson, S., Project work in context: regulatory change and the new geography of media. Environment and Planning A 2002; 34:2003-2015. Cooper, A., How to plan advertising. London: Cassell/The Account Planning Group, 1998. Cyert, R., March, J.G., A behavioral theory of the firm. Englewood Cliffs, NJ: Prentice-Hall, 1963. David, P.A., Clio and the economics of QWERTY. American Economic Review 1985; 75:332337. David, P.A., “Path Dependence and the Predictability in Dynamic Systems with Local Network Externalities: A Paradigm for Historical Economics.” In Technology and the Wealth of Nations, C. Freeman, D. Foray, eds. London: Pinter, 1992. DeFillippi, R.J., Arthur, M.B. Paradox in project-based enterprise: the case of film making. California Management Review 1998; 40:125-138. DeLillo, D., White Noise. New York: Viking, 1984. Dupré, J.A., The latest on the best: essays on evolution and optimality. Cambridge: MIT Press, 1987. Dornisch, D., The evolution of post-socialist projects: trajectory shift and transitional capacity in a Polish region. Regional Studies 2002; 36 (Special Issue):307-321. Ekstedt, E., Lundin, R.A., Söderholm, A., Wirdenius, H., Neo-industrial organising. Renewal by action and knowledge in a project-intensive economy. London: Routledge, 1999. Faulkner, R.R., Anderson, A.B., Short-term projects and emergent careers, evidence from Hollywood. American Journal of Sociology 1987; 4:879-909. Fisher, R.A., The genetic theory of natural selection. Oxford: Clarendon, 1930. Foster, J. Metcalfe, J.S., “Modern evolutionary economic perspectives: an overview.” In Frontiers of evolutionary economics. Competition, self-organization an innovation policy, J. Foster, J.S. Metcalfe, eds. Cheltenham: Edward Elgar, 2000. Gertler, M., “Being there”: proximity, organization, and culture in the development and adoption of advanced manufacturing technologies. Economic Geography 1995; 71:1-26. Ghosh, B.C., Taylor D., Switching advertising agency – a cross-country analysis. Marketing Intelligence and Planning 1999; 17:140-146. Girard, M., Stark, D., Distributing intelligence and organizing diversity in new-media projects. Environment and Planning A 2002; 34:1927-1949. Glasmeier, A., “Flexible districts, flexible regions? The institutional and cultural limits to districts in an era of globalization and technological paradigm shifts.” In Globalization, institutions, and regional development in Europe, A. Amin, N. Thrift, eds. Oxford: Oxford University Press, 1994.
80
Gernot Grabher
Goodman, R.A., Goodman, L.P., Some management issues in temporary systems: A study of professional development and manpower – The theatre case. Administrative Science Quarterly 1976; 21:494-501. Gould, S.J., The panda’s thumb of technology. Natural History 1987; 1:14-23. Gould, S.J., Lewontin, R.C., “The spandrels of San Marco and the Panglossian paradigm: a critique of the adaptionist Programme.” In Conceptual issues in evolutionary biology, E. Sober, ed. Cambridge: MIT Press, 1984. Grabher, G., “The Weakness of Strong Ties: the Lock-in of Regional Development in the Ruhr Area.” In The embedded firm: on the socioeconomics of industrial networks, G. Grabher, ed. London, New York: Routledge, 1993. Grabher, G., In Praise of Waste. Redundancy in Regional Development (Lob der Verschwendung. Redundanz in der Regionalentwicklung). Berlin: edition sigma, 1994. Grabher, G., Ecologies of creativity: the village, the group, and the heterarchic organisation of the British advertising industry. Environment and Planning A 2001; 33:351-374. Grabher, G., Cool projects, boring institutions: temporary collaboration in social context. Regional Studies 2002a; 36 (Special Issue):205-214. Grabher, G., The project ecology of advertising: tasks, talents and teams. Regional Studies 2002b; 36 (Special Issue):245-263. Grabher, G., Fragile sector, robust practice: project ecologies in new media. Environment and Planning A 2002c; 34:1911 -1926. Grabher, G., Stark, D., Organizing diversity: evolutionary theory, network analysis and postsocialism. Regional Studies 1997; 31:533-544. Granovetter, M., The strength of weak ties. American Journal of Sociology 1973; 78:13601380. Hassink, R., What distinguishes “good” from “bad” industrial agglomerations? Erdkunde 1997; 51: 2-11. Hassink, R., The role of political lock-ins in economic geography: a research agenda. Unpublished manuscript, Socio-Economics of Space, University of Bonn, 2002. Hatch, M.J., Exploring the empty spaces of organizing: how improvisational jazz helps redescribe organizational structure. Organization Studies 1999; 20:75-100. Hedlund, G., The hypermodern MNC – a heterarchy? Human Resource Management 1986; 25:9-35. Hedlund, G., “Assumptions of hierarchy and heterarchy, with applications to the management of the multinational corporation.” In Organization theory and the multinational corporation, S. Ghoshal, O.E. Westney, eds. New York: St. Martin’s Press, 1993. Hedlund, G., A model of knowledge management and the N-form corporation. Strategic Management Journal 1994; 15 (Special lssue):73-90.
SWITCHING TIES, RECOMBINING TEAMS
81
Hedlund, G., Rolander, D., “Action in heterarchies – new approaches to managing the MNC.” In Managing the global Firm, C. A.Bartlett, Y. Doz, G. Hedlund, eds. London: Routledge, 1990. Hobday, M., The project-based organisation: an ideal for managing complex products and systems? Research Policy 2000; 29:871-893. Hodgson, G.M., Economics and evolution: bringing life back into economics. Cambridge: Polity Press, 1993. Holland, J., Complex adaptive systems. Daedalus 1992; 121:17-30. Hudson, R., “Institutional change, cultural transformation, and economic regeneration: myths and realities from Europe’s old industrial areas.” In Globalization, institutions, and regional development in Europe, A. Amin, N. Thrift, eds. Oxford: Oxford University Press, 1994. Hudson, R., The learning economy, the learning firm and the learning region: a sympathetic critique of the limits to Learning. European Urban and Regional Studies 1999; 6:59-72. Interpublic, 2000, Annual Report, New York. Jacob, F., Evolution and tinkering. Science 1977; 196:1116-1166. Janis, I.L., Victims of groupthink: a psychological study of foreign-policy decisions and fiascoes. Boston: Houghton, 1972. Jones, C., “Careers in project networks: the case of the film industry.” In The boundaryless career, M.B. Arthur, D.M. Rousseau, eds. New York: Oxford University Press, 1996. Katz, M.L., Shapiro C., Network externalities, competition, and compatibility. American Economic Review 1985; 75:424-440. Krugman, P., Development, geography and economic theory. Cambridge: MIT Press, 1995. Lash, S., Urry, J., Economies of signs and spaces. London: Sage, 1994. Latour, B., “The powers of association.” In Power, action and belief, J. Law, ed. London: Routledge and Kegan Paul, 1986. Latour, B., The pasteurization of France. Cambridge: Harvard University Press, 1988. Leslie D.A., Flexibly specialized agencies? Reflexivity, identity, and the advertising industry. Environment and Planning A 1997; 29:1017-1038. Leslie, D.A., Consumer subjectivity, space, and advertising research. Environment and Planning A 1999; 31:1443-1457. Lewontin, R.C., Human diversity. New York: Scientific American, 1982. Lindkvist L., Söderlund J., Tell F., Managing product development projects: on the significance of fountains and Deadlines. Organization Studies 1998; 19:931-951. Lundin, R. A., Söderholm A., A theory of the temporary organization. Scandinavian Journal of Management 1995; 11:437-455.
82
Gernot Grabher
Lundin, R. A., Midler, C., Projects as arenas for renewal and learning processes. Boston, Dordrecht, London: Kluwer Academic Publishers, 1998. Malmberg, A., Maskell, P., The elusive concept of localization economies: towards a knowledge-based theory of spatial clustering. Environment and Planning A 2002; 34:429-449. Miller, P., Botham, R., Gibson, H., Martin, R., Moore, B., Business clusters in the UK - a first assessment (1, Main Report). London: Department of Trade and Industry, 2001. Morgan, K., Nauwelaers, C., “A regional perspective on innovation: from theory to strategy.” In Regional innovation strategies: the challenge for less-favoured regions, K. Morgan, C. Nauwelaers, eds. London: The stationary office and Regional Studies Association, 1999. Morin, M., Complexity. International Social Science Journal 1974; 26:555-582. Nachum, L., Keeble, D., Neo-Marshallian nodes, global networks and firm competitiveness: the media cluster of Central London. ESRC Centre for Business Research, Working Paper 138. Cambridge: Cambridge University, 1999. Nachum, L., Keeble, D., Foreign and indigenous firms in the media cluster of central London. ESRC Centre for Business Research, Working Paper 154. Cambridge: Cambridge University, 2000. Narula, R., Innovation systems and “inertia” in R&D location: Norwegian firms and the role of systemic lock-in. Research Policy 2002; 31:795-816. Omnicom, 2000, Annual Report, New York. Owen-Smith, J., Powell, W.W., Knowledge networks in the Boston biotechnology community. Paper presented an the conference on science as an institution and the institutions of science, University of Siena, January 25-26, 2002. Quinn, J., “The account executive in an advertising agency.” In The advertising business, J.P. Jones, ed. London: Sage, 1999. Pratt, A., Firm boundaries? The organization of new media production in San Francisco 1996-98. Paper presented at the workshop “Beyond the firm? Social and spatial dynamics of projectorganization,” University of Bonn, 27-28 April 2001. Rainey, M.T., “The planning context.” In How to plan advertising, A. Cooper, ed. London: Cassell/The Account Planning Group, 1998. Rantisi, N.M., The local innovation system as a source of “variety”: openness and adaptability in New Yorks City’s Garment District. Regional Studies 2002; 36:587-602. Sarbaugh-Thompson, M., Feldman M.S., Electronic mail and organizational communication: does saying “hi” really matter? Organization Science 1998; 9:685-698. Sassen S., “On concentration and centrality in the global city.” In World cities in a world system, P.L. Knox, P.J. Taylor, eds. Cambridge: Cambridge University Press, 1995. Sellers, J., “The art director.” In The advertising business, J.P. Jones, ed. London: Sage, 1999.
SWITCHING TIES, RECOMBINING TEAMS
83
Shelbourne, J., Baskin, M., “The requirements for creativity: a director’s perspective.” In How to plan advertising, A. Cooper, ed. London: Cassell/The Account Planning Group, 1998. Smith, J.M., “Optimization theory in evolution.” In Conceptual issues in evolutionary biology, E. Sober, ed. Cambridge: MIT Press, 1984. Stark, D., “Heterarchy: distributed authority and organizing diversity.” In The biology of business: decoding the natural laws of enterprise, J.H. Clippinger, ed. San Francisco: Jossey-Bass, 1999. Storper, M., The regional world. Territorial development in a global economy. New York, London: Guilford Press, 1997. Sydow, J., Staber U., The institutional embeddedness of project networks: the case of content production in German television. Regional Studies 2002; 36 (Special Issue):215-228. Thrift, N., “On the social and cultural determinants of international financial centers: the case of the City of London.” In Money, power and space, S. Corbridge, R. Martin, N. Thrift, eds. Oxford: Blackwell, 1994. Tichy, G., A sketch of a probabilistic modification of the product-cycle hypothesis to explain the problems of old industrial areas. Research Memorandum Nr. 8401, University of Graz, 1984. Usai, A., Delmestri, G., Montanari, F., Human capital, social capital and performance: an empirical test from an entrepreneurial project-based industry. Milan, University Luigi Bocconi: unpublished manuscript, 2001. Uzzi, B., Social structure and competition in interfirm networks: the paradox of embeddedness. Administrative Science Quarterly 1997; 42:35-67. Wagner, G.P., Altenberg, L., Complex adaptations and the evolution of evolvability. Evolution 1996; 50:967-76. Weick, K., Improvisation as a mindset for organizational analysis. Organization Science 1998; 9:543-555. Weick, K., Education organizations as loosely coupled systems. Administrative Science Quarterly 1976; 21:1-19. Wells, W., Burnett, J., Moriarty, S., Advertising principles and practices. Upper Saddle River: Prentice-Hall, 1998. Wenger, E., Communities of practice: learning, meaning, and identity. Cambridge: Cambridge University Press, 1998. WPP, 2000, Annual Report, London.
Chapter 5 KNOWLEDGE-INTENSIVE SERVICES AS A KEY SECTOR FOR PROCESSES OF REGIONAL ECONOMIC INNOVATION: LEAPFROGGING AND PATH DEPENDENCY
Hans Joachim Kujath
5.1 INTRODUCTION If we consider regions as important nodes of the global business networks, then we will also have to bear in mind the new paradigm of industrial production connected to the processes of globalization and Europeanization. For years, this paradigm shift has been associated with regional concepts of production, such as industrial districts. Industrial districts, consisting of small enterprises in the “Third Italy” or the production complexes of BadenWürttemberg in Germany, were regarded both as best practice in flexible production and as models worthy of adoption. Furthermore, flexible consumer orientation and innovation management often involves large enterprises taking on the role as system integrator. Such types of regional production complexes may be more or less widespread. However, in my opinion, they lead us in the wrong direction when dealing with the complex field of regionalization and globalization. Industrial districts and in part the concept of the “hub and spoke district” all allude to the old production system. Here, management, product development, and employment are closely interconnected, i.e. they are spatially closely interconnected. Under the conditions of globalization these relationships, however, dissolve into a worldwide extending network of often formally independent, but centrally controlled and managed enterprises. When we applied network analyses to the German and European rail vehicle sectors, we found that the manufacturing supply chains are organized nationally and increasingly transnational. This is in fact a recent development. From the historical perspective, manufacturing industries like rail vehicles – in the early stages of its development – were organized as “Marshallian Industrial Districts,” where small and medium-sized enterprises were collaborating,
86
Hans Joachim Kujath
embedded in regional contexts. The Berlin region, the Rhine-Ruhr region, but also southern Germany, and here especially the Karlsruhe and Nürnberg regions were industrial districts. Obviously, the manufacturing industries emanated from these regional contexts, not least due to economic restructuring and spatial centralization processes. Consequently, from the formerly regionally interconnected supply chains only fragments remained. This applies to all advanced industrial systems (Dybe and Kujath, 2000). The old manufacturing districts, however, laid the foundation for the emergence of agglomerations with an individual structure and with a particular spatial concentration of technological and other knowledge bases in a variety of branches. Locational network analysis allows us to recognize the specific focal points of competence and knowledge that are concentrated in different German regions: Engineering services, electrical engineering/electronics/information science or centers of high-tech research and development and of market research and consulting as well as the production of prototypes. Regional networks obviously accommodate particular sets of common skills and they stimulate the reconfiguration of the economy by applying these skills and the internally rooted knowledge. A globalized economy hence appears to generate unique centers of excellence in the form of innovation regions primarily in the advanced economies. Simultaneously, mass production is concentrated in so-called standardization regions, which have the skills necessary for using innovation.
5.2 NEW REGIONAL CLUSTERS OF KNOWLEDGE PRODUCTION AND SERVICES 5.2.1 Knowledge services as accelerator of knowledge conversion and diffusion Industrial manufacturing, in the narrow sense, is no longer the source of dynamic growth within innovation regions; it is rather the service sector with its different forms of specialization. Great importance has been attached to the knowledge services regarding innovation and regional competitiveness, because they surmount the narrow topical and spatial limits of knowledge transfer in informal relationships and in the everyday exchange of experiences (learning by doing, learning by using). Knowledge services systematically contribute to the acceleration of information and knowledge. They rationalize knowledge production and distribution and thus speed up the distribution of
KNOWLEDGE-INTENSIVE SERVICES AS A KEY SECTOR
87
new knowledge developed in the region or accumulated from global relations (Dunning, 1993). In a world without such services - like in some economic areas influenced by tradition (handicraft, trade) - the transfer of knowledge is object to narrow topical and spatial limitations. Here, transformation remains a gradual change of processes and products, and it is connected to daily experiences and incremental learning steps. The dynamic of innovation in these systems is constricted, because knowledge is generally related to the learning individuals as implicit knowledge and knowledge spillovers are dependent on rather informal contacts and observations. Knowledge services have the tendency to break up these obstructions and to accelerate the diffusion of knowledge by converting it into explicit knowledge to their customers. They contribute to the articulation of implicit knowledge (externalization) and generate new explicit knowledge through combination with other external knowledge. “It is a process of reduction and conversion which renders especially easy the transmission, verification, storage and reproduction of knowledge.” (Foray and Lundvall, 1996) Finally, new implicit knowledge (internalization) is generated in a recursive process as part of the codification process for the service provider as well as for the customer. According to Nonaka and Takeuchi (1995) the described spiral-form movement is central to both the individual and the organizational learning process (Table 5.1). An important dynamic factor is that knowledge, which is codified, can be transferred over great distances at very low costs and that the existence of information technologies renders such a conversion of knowledge attractive. On the other hand, the systematical conversion of knowledge will increase the demand for high performance information and communication technologies and stimulate their development (Foray and Lundvall, 1996:23).
5.2.2 Knowledge services as seedbed of the new knowledge economy Knowledge services, however, do not only accelerate the explication of knowledge and its diffusion; they are also the starting point for the emanation of a knowledge economy and of knowledge markets. “By codification, knowledge picks up more and more of the properties of a commodity. Market transactions are facilitated by codification: codified knowledge can be more precisely described and specified in terms of content and intellectual properties, and this can reduce uncertainties and information asymmetries in any transaction involving knowledge.” (Foray and Lundvall, 1996:22)
88
Hans Joachim Kujath
In order to understand this function of the advanced services within the process of conversion of tacit knowledge into a commodity, we have to pay attention to the actual service process as part of the relation between service provider and customer. A typical feature of the relation between provider and customer is the service supplier’s need to acquire strategically relevant knowledge from the customer in order to provide his service (e.g. problem analysis of the value generation chain, interpretation and sorting out of the information) and to use this knowledge for subsequent treatment and consultation. Knowledge services link the existing knowledge of the customer to be explicated to the individual knowledge of the service provider (his capacity potential), but also to external knowledge and thus develop customer specific information and knowledge products. These in turn can serve to implement innovation in the production and marketing process, e.g. of a new technology, a new production process or a new marketing strategy. The character of a service can also be effective in the case of application oriented R&D, if the search process of the innovation activities is connected to a demand assessment and a customer specific solution tailoring. From codification and storage during the process of service provision for one customer it is only a small step to disengage from the close relationship to one single customer by developing specific work packages that can easily be standardized, like content-products or products of process management (programming). The service economy is just generating productivity reserves by separating the service process from the customer relation: the typical concept of service as a realm resistant to technical innovation and productivity growth is being broken up and a process of service industrialization is being introduced.
KNOWLEDGE-INTENSIVE SERVICES AS A KEY SECTOR
89
If we distinguish the service process of the service provider by a phase of pre-combination and a phase of final combination, then the following interrelation becomes obvious. The stronger the customer is integrated in the phase of final combination, the more the autonomy of the service provider will be restrained and the knowledge production will be converted into an individual act of customer specific provision of services like in the case of company consultancy or legal services. In other cases the pre-combination takes on a dominant position, i.e. a pre-fabricated part is generated rather autonomously, while the customer integration into the process occurs relatively late. Here, the development of the knowledge product - the codification of knowledge - can be designed as an industrially organized process that may be followed by another service. In connection with the technological push in the information and telecommunication sectors, these types of services have developed rapidly. Consequently, more and more information and knowledge products, originally tailor-made products created through a close exchange between service provider and customer, are transformed into standardized products, which either are incorporated in new material commodities or become tradable via information infrastructures (Rubalcaba-Bermejo, 1999). Thus not only innovation within the service sector itself takes on speed, but access to extended markets can lead to the emergence of a self-sustained (regional) export basis even within the service sector itself. In this context, the information and communication techniques set an important framework condition for the facilitated transfer of information and knowledge products. Simultaneously, they stimulate the codification of more knowledge than ever before. Similar to the development of the manufacturing industry from customer-related handicraft, the service economy is released from local and regional bonds, providing access to global markets and the growth of the knowledge economy.
5.2.3 Types of the new knowledge economy We were able to delineate three regional specialization patterns for the knowledge economy. These include: A regional cluster of global exchange, i.e. market-oriented knowledge services. A regional cluster of production-related research and development, i.e. the so-called servindustrial economy. A regional cluster of industrialized information and knowledge production, i.e. the new knowledge industry.
90
Hans Joachim Kujath
Cluster of globalized knowledge exchange Globalized economic exchange leads to a rising importance of services, which support transnational enterprises to perform successfully on the respective national and international markets: market research facilities, legal consulting, accountancy consulting, advertising agencies, insurance agencies, real estate brokers etc. These intermediaries, who advice, prepare decisions and control the results, are highly interdependent. Regions of this type develop as global knowledge marketplaces. As service nodes they represent coordinating points between the service and production networks with different geographical extensions and formality (Andersson, 1995:29; Sassen, 1996 and 2001). Transnational businesses as well as national industrial enterprises are attracted to such regions not in order to manufacture, but in order to accumulate locally obtainable knowledge, e.g. about national and international market conditions, competitors, consumer demands, state regulation, cultural particularities, but also technological knowledge. Besides the Frankfurt region, the Munich region has become a gateway to the German market and an information hub for everyone attempting to set foot in the European market. Consequently, regional innovation cannot be perceived as technological innovation in the narrower sense, but as a specific ability of the regional service providers to provide new knowledge about the market environment for the customers. Service providers within these regions operate as mediators between the national/regional and the global codes of the industrial system. For foreign industrial enterprises, the service networks are particularly attractive, because they help to provide access to national markets. The performance of the service networks, however, is dependent on intelligent communication; its success in turn depends on the shared perceptions and ways of interpretation, i.e. on the appropriateness of the regional institutional arrangements and on the opportunities to utilize these arrangements. Service providers ultimately accumulate tacit knowledge from the interaction contexts and codify it, i.e. make it comprehensible/intelligible for their customers. Servindustrial Economy The innovation system of the technopoles with its dominant technological development trajectories has a completely different focus. In Germany, this pattern can be found around Munich and Stuttgart, but also in the Berlin region. Here, information and knowledge providers, universities and research facilities unite in the development of new products in advanced technology, e.g. in the field of biotechnology and in information and communication technology, and even in “older” technologies like automobile manufacturing
KNOWLEDGE-INTENSIVE SERVICES AS A KEY SECTOR
91
(cf. Castells, 1994; Sternberg and Tamáscy, 1999; Dybe and Kujath, 2000). Specific knowledge-intensive and business-related services take on the task to rationalize business organization, to develop and to improve products and to evaluate information about the entrepreneurial environment. A particular and changing division of labor among industry and service sector emerges that modifies the input ratio of the production factors in favor of knowledge, human capital, and information. Consequently, the initial production process is increasingly dominated by services. Since technological services frequently emanated from the existing industrial system, manufacturing and services are often located in spatial proximity. Spatial proximity, however, is not a necessary precondition (Rubalcaba-Bermejo, 1999). This becomes particularly obvious in the highly globalized system of automobile production. In this comparatively mature industry, the old production sites are changing gradually to areas where knowledgeintensive service providers, development offices and other specialists work in close spatial relation with new concepts for automobile products. Manufacturing, however, is only rarely located within the region. Similar developments can be observed in the area of information and communication technologies. Many businesses have maintained their research facilities and design centers as well as manufacturing sites for prototypes in Silicon Valley, while standardized parts are manufactured in other regions, particularly in Asia. Providers of knowledge services in this kind of regions guarantee the subsequent development of the regional technological innovation system, i.e. they produce a stream of information that flows directly as input into industrial manufacturing. These regions, former locations of industrial production, are thus converted into locations of knowledge production for the industry. With regard to the institutional system, educational aspects (human capital production), the transfer of knowledge and basic research gain in importance. Usually, the institutional arrangements are more tightly organized in comparison to our first case, the global market place: Research and development networks or focal networks around a leading industrial enterprise with the objective to develop and attain the latest technology. Path dependency dominates the service systems of the servindustrial economy. R&D activities are connected to research traditions, which in turn depend on former industrial focal points in the region. Additionally, path dependency within a targeted generation of knowledge is determined by the research know-how and the research infrastructure established in the past.
92
Hans Joachim Kujath
Knowledge industry The third type of regional innovation is characterized by tendencies of industrialization within the service sector itself. This process may be called services revolution, since information and communication technologies transform the traditional characteristics of services in a never before experienced extent: knowledge becomes storable, easy to save and easy to communicate. The places from which the systems originate can be described as regions of the knowledge industry: here, for instance, platforms for electronic commerce, logistic systems, digital procurement systems, program applications and the like are produced. This also includes the so-called “content production,” i.e. the media industry and the particularly fast-growing multimedia sector (cf. Scott, 1988; Sträter et al., 1998). Berlin and Munich are both German centers of the media industry. The development of information and communication technologies functions as a basic innovation for the tremendous transition from knowledge services to knowledge commodities. The industrialization of knowledge services establishes an independent sector that produces codified knowledge, which is extremely concentrated spatially. Although knowledge commodities are in general universally available and after a short period often even global collective commodities, the prototype construction, however, is usually the result of personal skills to combine creativity, experiences, scientific knowledge, and organization knowledge. The fact that knowledge becomes tradable will not only lead to shifts in productivity like in manufacturing industries, but it will open up remarkable international sales opportunities - regardless whether the product is stored physically on hardware or conveyed by electronic media. Finally, due to the international opportunities to trade service products, the knowledge industry will presumably rise as leading supporter of economic growth in the future (Freeman and Soete, 1997:404). Regions with this field of competence will reinforce their position as growth poles for non-manufacturing industries and for international trade of knowledge commodities.
5.2.4 Effects on regional economic systems There is controversy about how this changing mode of production and distribution of knowledge really affect regional systems, and which institutional and organizational forms the knowledge-based economy will possibly attain. Foray and Lundvall (1996) point out that the movement of the border between codified and personalized knowledge in favor of the former drives productivity and economic growth. The new information and communication technologies would make it both possible and economically attractive for
KNOWLEDGE-INTENSIVE SERVICES AS A KEY SECTOR
93
information businesses to transform knowledge that has persisted in personalized or tacitly held form into a codified form. Formerly locally and regionally captured knowledge resources thus become spatially mobilized. Knowledge is released from the limitations of the regional social space in a way that the initial region gradually tends to lose its comparative advantage as location of knowledge. As a result of this ubiquitous availability of a growing quantity of knowledge resources, it is assumed that an increased locational competition arises among the regions in the advanced economies, basically in terms of producing new knowledge and by safeguarding knowledge leads and advantages. One could argue that the process of universalizing knowledge – accelerated by the knowledge service providers – would eventually lead to destabilizing even economically powerful regions. Scholars who suggest that the capability to generate new knowledge and to codify existing knowledge is connected to individuals and their learning abilities oppose this point of view. They point out that these capabilities are rooted in specific relationships and personal communication networks. Soete (1996) and other innovation scholars emphasize the accumulation of new knowledge; its codification and its application represent a significant element of the new knowledge services. Generation of targeted knowledge is conceived of as a helical learning process of individuals and organizations (learning curve effect). This means that the learning routines themselves would be bound to personal communication and social patterns of interaction to an extreme extent, which are ruled by shared beliefs, and modes of interpretation (Polanyi, 1958:212). Here, spatial proximity often is supposed to be essential. However, spatial proximity does not automatically imply the emergence of regionally linked systems of knowledge exchange and the regional clustering of service companies. Still, face-to-face contacts are extremely relevant for the generation of highly complex information like scientific and technological knowledge, but the communities of the knowledge economy may not, or only to a limited extent, be embedded in local networks and milieus, rather than they frequently operate in national and international contexts of communication. To sum up the above, two processes seem to be at work here that render formerly incontestable locational advantages ubiquitous and interrupt regional pathways of development: Firstly, the internationalization process of factor and commodity markets, and secondly the more systematical codification and marketing of knowledge by knowledge-intensive services. It is of course correct that the personalized share of knowledge, which is not tradable, is rather difficult to transfer in space. However, the more job markets become international and knowledge gains in mobility, the easier this knowledge - and particularly highly qualified knowledge - will become detachable from the respective regional connections. This process can be primarily ob-
94
Hans Joachim Kujath
served within national institutional frameworks, but increasingly also in the international realm. Given this reduced significance of proximity to sources of knowledge and between service providers and users, several questions arise. Why should companies concentrate spatially and will regional economic pathways stabilize? Or, will this mean the decline of unique and absolute locational advantages rooted in regional and national path dependencies? Comparative advantages like the infrastructure endowments of the region integrating several global networks or advantages of agglomeration, e.g. in providing a critical mass of human capital necessary to develop codified knowledge, may become decisive. Furthermore, a diversified and large market of inputs and a high number of firms of the same service industry may generate external urbanization and localization benefits (Sternberg, 1991).
5.3 EMPIRICAL FINDINGS IN THE REGIONS OF MUNICH AND BERLIN The following section presents first findings from an empirical investigation carried out in the regions of Munich and Berlin in 2002 concerning the interconnectedness of the knowledge economy within the economic space. The investigation collected data regarding the use of ICT, the employment, the product strategies, the character and spatial scope of inter-firm relations and the information strategies. Furthermore, data on the mobility behavior of the enterprises as well as location factors and structural data of the service providing companies were raised. The innovative approach of this empirical work is based on the verification of the earlier defined new types of services (chapter 5.2), in particular with regard to the patterns of interaction, i.e. the interorganizational relationships among the firms (e.g. between service providers, customers, suppliers and information sources), the communication patterns embedded therein, and the forms of mobility induced by this interaction. The empirical research was conducted following a multi-step quotation procedure combined with a random selection that lead to a sample of 3,000 enterprises for each region. Fifteen percent of the enterprises from the Berlin region returned the questionnaire, while the ratio was 20 percent for the Munich region.
KNOWLEDGE-INTENSIVE SERVICES AS A KEY SECTOR
95
5.3.1 The roots of the knowledge economy in both regions We found quite dramatic restructuring processes in the different regions we investigated (Munich, Berlin) that transformed the regional economic cultures into new types of regional service centers. During the past decade, the knowledge economy has pushed the formerly strong manufacturing industries into the background in both regions. The share of jobholders in the manufacturing sector in the city of Berlin decreased to only 30 percent while in Munich their share is still over 40 percent. But the explanations for this transformation in favor of the service industry and the knowledge economy are quite different: In the case of Munich, these tendencies cannot be interpreted as a “breakthrough” yet. In contrast, they have to be recognized as the result of strong regional development pathways that predominantly originated from old manufacturing clusters or service clusters and further developed the respective knowledge resources by accumulation of similarly targeted knowledge. However, the ties between knowledge services and the new knowledge economy on one hand and the still strong local manufacturing complexes of the machine, automotive and aircraft manufacturing and the electrical industry on the other hand are decreasing without disconnecting entirely. The multi-media industry, information technology (IT) services as well as the financial and insurance industry clusters are resulting from strong regional development pathways, too. These services particularly contribute to the expansion of the regional economy and gradually disconnect from the old pathways of economic development by generating new individual pathways of the knowledge economy. The case of Berlin in contrast illustrates a process of economic transformation frequently referred to as “leapfrogging,” since the historical development paths are neither pursued, nor forced to overcome. Due to the process of system transformation, the largely de-industrialized region offers almost no “inherited” economic points of connection neither for the continuation of economic development paths, nor for a breakthrough. Moreover, the knowledge economy in Berlin lacks the opportunity to reconnect to a historically grown regional economic development characteristic for Munich and other European innovation regions. Instead, it has to leapfrog this phase of reconnecting to existing regional economies by generating new path dependencies out of its own capacities. Attempts to revitalize old manufacturing cores in Berlin - e.g. transportation and electrical industries - and to develop a knowledge economy around these manufacturing cores largely failed.
96
Hans Joachim Kujath
5.3.2 Spatial boundaries of the knowledge economy in Berlin und Munich Building upon the theoretically derived types of services it is possible to organize the results of the empirical investigation in a way that the significance of the knowledge services and the knowledge economy becomes discernible in both regions (Table 5.2). Thus, it can be demonstrated that roughly half of the enterprises surveyed belong to the new information services and to the knowledge economy. Within this group, the “global services” represent the majority. The investigation, however, reveals significant regional differences. Thus, in the Berlin region, the type of the classical locally bounded service still dominates, while in the Munich sample the “classical information service providers” and “global services” are quite evenly represented, reaching one third respectively.
To reveal the role of new knowledge services in the economic development of both regions, we examine the following spatial relations: Between service producers and users. Between intermediary suppliers and service producers. Between service producers and sources of knowledge. Spatial relations between service producers and users These relationships inform about the transformation of the service activities from non-basic to basic activities. While traditional services in general supply only the local marketplace, i.e. they engage in spatial proximity to their users to overcome distance barriers, it is to be expected that the information services and the knowledge economy disconnect from their spatial limitations and develop as a new export basis for the regions. These services play an
KNOWLEDGE-INTENSIVE SERVICES AS A KEY SECTOR
97
active role for the development of their regions since, parallel to their expansion, the entire regional economy gains considerable growth incentives. Our investigation confirms this notion only as a tendency, however without establishing a remarkable substitution of face-to-face contacts typical for these services by a general standardization of the information products. The type “information industry” with standardized information products, which, like the products of the manufacturing industry, are traded on global marketplaces, is still the exception. One of the problems that obstruct a selfsustained growth in the production of information lies in the legal protection of ownership rights upon information commodities. Now as before personal customer contacts and the provision of customer-specific information solutions determine the interaction of the metropolitan information service providers. Differentiating the whole process of the services into the three phases (1) “during placement of order,” (2) “during processing,” and (3) “customer support or care,” the dominance of personal contacts and the personal transfer of knowledge become clearer. In all three phases, 90 percent of the enterprises that responded deemed personal customer contact as very important or important. Three quarters of all service enterprises establish personal customer contacts on the spot. Nevertheless, the regional connection to the customers inside the respective region becomes less important. The information service providers noticeably become enterprises acting on a trans-regional level and with partly globalizing market relations. Home market effects and close relationships to a diversified and large local market for the final good lose influence on spatial behavior of the service industry. More than 60 percent of all information service providers surveyed simultaneously have customers inside and outside the region. Therefore, the transport links of the regions are gaining central importance as location criteria for the information service providers. As we expected the information providers in the economically strong high-tech region of Munich have by far a larger and spatially more wideranging national and international clientele than the service providers in the Berlin region. In Munich, more than 30 percent of all information service providers have customers predominantly from outside their own region, in Berlin the proportion is still below 20 percent (Table 5.3). This points out the differences between the regions regarding the structure of services provided as well as to a different level of development and divergent market spreading.
98
Hans Joachim Kujath
Spatial relations between intermediary suppliers and service producers In addition to the customer relations, the spatial connections to intermediary suppliers are equally important for an evaluation of the role of services in regional development. They inform about the dependence of the service providers and the actors of the knowledge economy on regional communities of knowledge production. By the same token it becomes evident whether, and to what extent, a specific pattern in the division of labor evolves in the production of knowledge commodities or if, similar as in the manufacturing industry, the diversified supply chains will extend beyond the limits of the regional boundaries. The survey established for both regions demonstrate that, in contrast to the customer relations, the spatial connections to intermediary suppliers were much more firmly organized inside the regions. Hence, only about 30 percent of the supplier relations possess a national range while an additional five percent extend beyond the national boundaries. Furthermore, it became evident that 80 percent of all firms with supplier relations had at least one supplier inside their respective region. Apparently, suppliers of intermediary products are seeking proximity to their clients inside the large and compact marketplaces of agglomerations. Corresponding to the early stages of industrialization, new value adding clusters of knowledge production emerge inside the agglomeration spaces. In fact, these clusters can generally only prevail in the currently small segment of the knowledge economy comprising standardized commodities as the media or the software industries. The productivity advantages of a spatially embedded system for the production of information can be successfully employed only in the knowledge industry. Spatial relations between service producers and sources of knowledge One of the main reasons, which might explain the concentration of knowledge-based services in the metropolitan regions is the fact that these regions
KNOWLEDGE-INTENSIVE SERVICES AS A KEY SECTOR
99
are especially well-equipped with knowledge resources, which can be accessed outside economic relations and above all by personal contacts and conversations. Sociological theories of knowledge milieu have pointed out the difficulties to communicate mental models and perceptions, experiences and normative judgments as a basis of economic action. This kind of personalized knowledge can only be mobilized by specific socially and spatially embedded interaction and thus would compel enterprises to connect spatially. Surprisingly, and in contrast to the evident quest for proximity to sources of knowledge we expected, the service providers in both urban regions we investigated make only little use of the advantage of local knowledge about technologies, markets, economic practice and regulation outside their functional business relationships. The results of our investigation confirm the overwhelming significance of personal contacts as a means to expand the knowledge of the service providers. Despite the existence of modern communication and information technologies or the tremendous supply of special journals and scientific literature available, face-to-face contacts and personal conversations are indispensable to mobilize those shares of implicit knowledge that are difficult to grasp or only transferable at high expenses. Around 81 percent of the service companies that responded draw their knowledge either entirely or in combination with other means of knowledge from personal conversations. Mainly the development of the respective branches, the general market development, technological information and product information shape the contents of these communications. The hypothesis often advocated by the proponents of the agglomeration theory as well as in public that metropolitan regions are particularly attractive for service companies because of the density of relations among the knowledge holders, however, could not be confirmed. Although a large group of the service providers gains its knowledge from communication links inside the respective metropolitan places, the group obtaining knowledge from outside the region is far from being insignificant. We can further emphasize and elucidate our findings by distinguishing the different locational opportunities to combine the information sources according to the choice of the service providers. One third of the service companies surveyed had all their information sources inside their own region. One quarter obtains all information from outside the region. Consequently, permanent spatial proximity to the knowledge holders concentrated in the regions must be of rather little relevance for those companies that are part of the said quarter. Hence, other factors have to be more important for the decision to stay in the metropolitan region. The largest group, however, consists of firms, which have distributed their knowledge sources over their own region as well as over external regions. This group makes up more than 40 percent of the sample, and the respective firms are either more oriented towards their specific region or towards other
100
Hans Joachim Kujath
regions. Yet also this observation challenges the notion that the region, where the service provider is located, offers particularly favorable conditions for the development of his own knowledge-based capacity.
In interpreting the results rather carefully, we can state that face-to-face contacts are in general indispensable for the service companies aiming at strengthening the entrepreneurial knowledge base. At the same time, a large part of the companies is not directly or only partly connected to the network of knowledge holders in their region of origin. The communication networks of these companies are characterized by supra-regional or trans-regional extensions and depend upon a mix of direct, personal communication and new information and communication technologies. The significant spatial extension of the companies’ communication across regional boundaries is not contradictory to the local “knowledge milieu,” which potentially exists on other levels and which emerges in the first place among individual actors outside the companies’ realm. It shows, however, that the spatial clustering of service businesses in the metropolitan regions does not unanimously correspond to the necessity to seek spatial proximity to the metropolitan knowledge resources as an input for their own knowledge base. Beyond this general result, our regional comparison reveals significant divergences between the Berlin and Munich samples regarding the spatial distribution of the location of information sources. Altogether, the service providers from the Munich region are involved in external contexts of knowledge acquisition to a much greater extent than those from the Berlin region, whereas in Berlin, the local urban environment or, the local information network is more important. This becomes even more evident if we examine the specific combinations of knowledge sources for each region. The proportion of enterprises with exclusively local information networks in Berlin is by far greater than in the Munich region. If we compare both samples, one apparent explanation for this observation is the fact, that in the Munich region the group of “global services” with a spatially wide-stretching network
KNOWLEDGE-INTENSIVE SERVICES AS A KEY SECTOR
101
is represented relatively often, while in Berlin the classic, locally bound service providers with local information networks are more important. Equally remarkable is the connection between company size, measured by the number of employees, and the range of the information network. In the Munich sample, large companies with a trans-regional range are more frequently represented than in the Berlin agglomeration. If, from the perspective of the individual company, the traditional factors of spatial proximity to the sources of knowledge within the given region are less important, then which are the factors responsible for locational relationship? Obviously, the modern and high-capacity transportation and communication infrastructures that support the trans-regional information exchange, to which only the large agglomerations can be connected, represent one decisive factor. These infrastructures will increasingly assume greater importance for the service economy the more they specialize and the more they are oriented towards supra-regional markets both on the national and on the international scale. Certainly, the opportunity to recruit growing numbers of more and more qualified staff provides is an important aspect for the establishment of a business at a given location too. The large urban agglomerations provide for adequately specialized and sufficiently large job markets. Additionally, the living standard of a region and also the prestige or image of a region are most likely to play an important role.
5.4 KNOWLEDGE SERVICES – STRATEGIC OPTION FOR DE-INDUSTRIALIZED REGIONS Our considerations and empirical findings have demonstrated that knowledge services have the function of propelling innovation for industrial production and for the service sector itself (industrialization of the knowledge production). At the same time, they are also the motors of sectoral reconstruction, of the conversion of the locational structures, and of the spatial division of labor. Because the knowledge industry rises as a driving force of the economy in the advanced economies, the hopes of those regions with economic and structural weaknesses like Berlin are directed towards knowledge services. A catalytic function for the reorientation of the regional economy has been ascribed to these services, the function of a nucleus and an assistant initiating the necessary stimulus for new developments. Since, however, the industrial foundation for the development of advanced knowledge services in regions with structural weaknesses has ceased to exist, the new service clusters would have to concentrate on expanding their markets beyond the regional scope from the beginning. Organic growth, based on local demand of industrial
102
Hans Joachim Kujath
firms from which external marketplaces can be opened up is rarely manageable in these regions.
5.4.1 Leapfrogging strategies Two strategies are conceivable for this kind of regions: Strategy (1): industrialized knowledge production: If industry and knowledge-intensive services do not necessarily have to be closely connected spatially, the mutual interdependence between industry and knowledge services over long distances can be realized at least at the national scale. Therefore, reconstruction of the industrial basis might not be a necessary precondition for the emergence of the respective services in de-industrialized regions. Possibly - and there is some evidence provided by the tendencies of the transition from service production to self-reliant information industries delivering standardized inputs for manufacturing and standardized final products at the same time – the regional potential regarding knowledge-intensive services may be transformed directly into an economic export base while the stage of reconstructing manufacturing industries can be abandoned (leapfrogging). Certain regions in Germany like Berlin are highly developed locations for training and basic research in a variety of areas. Our findings from empirical investigation in Berlin confirm that these resources can constitute a (collective) basis from which the economically useful development of knowledge and, hence, also the development of enterprises specialized in production and distribution of knowledge commodities appears possible. Knowledge producing apparently can become a self-sustained export basis for the urban areas, especially if service products are converted into tradable commodities. Strategy (2) – servindustrial economy: Although it appears little plausible that regions affected by de-industrialization could reconnect to industrial traditions that for instance the regions Munich and Stuttgart still possess, the knowledge existing in regions like Berlin and the services based on this knowledge could become the foundation for new industries. In this case, it is not the manufacturing industry that provides the basis for the development of the service economy. In contrast, research and development as well as related services define “centers of excellence” from which the new industries emerge and manufacturers are stimulated to establish research and development facilities. Such centers of excellence are represented e.g. by the technology region Aachen and in the Berlin region too. The starting point for both regions no longer are the old industries, but leading research facilities and out sourced services of the new technologies, like bio-chemistry and biotechnology, medical technologies, opto-electronics. The success of the Aachen case proves that technical universities can constitute the nucleus for
KNOWLEDGE-INTENSIVE SERVICES AS A KEY SECTOR
103
economic reorientation, as long as they offer a profile compatible to the new competence fields. Several hundred young technology firms in microelectronics, medical technologies etc. have emerged from this nucleus. Meanwhile, the region is also attractive for external investors. Research and development departments have settled near the university and the recently established businesses, thus constituting a bridge to the global networks of technological transfer. Thus, a technopole may emerge within in a region that has been characterized by the coal, iron, and steel industries, university initially being also a part of the production system of mining and steel processing. There are similar processes to observe in Berlin although they are less noticeable than in Aachen. The medium-level technologies still strong in Germany like mechanical engineering, car manufacturing, fine mechanics, optical and electrical industry do not constitute a sustainable focus of economic development in regions of this type. This strategy therefore equally abandons the medium-level stadium of industrial development and assumes the lead in technological innovation. Both options are “leapfrogging”-strategies. They do not comprise the model of regional networking in knowledge-intensive services and traditional industrial branches or services, still residual existing in several German regions. Instead, they aim at abandoning this stage of development. This will be possible in all areas where internationally competitive knowledge industries can be successfully developed and thus knowledge can be transformed into tradable, storable, and conveyable commodities – e.g. movie production, software production and engineering services. Yet another strategy that connects to the existing knowledge resources and that attempts to establish regional “centers of excellence,” from which new production-related technological clusters could emerge, appears equally reasonable. However, both strategies can rely upon the fragile regional markets only to a limited extent. Consequently, the strategies cannot be reduced to measures of regional integration, but will have to support the developing service economy in finding access to interregional and international markets.
5.4.2 Policies to promote the knowledge economy Agglomerations with structural weaknesses that are nonetheless equipped with highly developed research and development potentials and with a decent basis of human capital like Berlin may be able to pursue “leapfrogging” strategies. Usually, these attempts comprise the aggregation of knowledge potentials in the prospectively suitable areas of competence and the improvement of the supportive framework conditions. These basic conditions comprise the following specific collective resources:
104
Hans Joachim Kujath
Infrastructures (communication and mobility facilities). Research and development potential (collective knowledge). Educational and training environment (human capital). Pool of relationships (networks). Rules and norms providing trust and reliability (institutions). Regional policy aiming at competitiveness and innovation in the fields of the knowledge economy can primarily embark on these regional resources. The main task of regional policy is to recognize the existing and prospective areas of competence and to sustain these areas with the appropriate collective resources. However, in many regions exposed to the pressures of reconstruction and affected by de-industrialization we can observe that regional policy exclusively oriented towards the fields of competence and sectoral knowledge cluster is far from solving the economic difficulties of restructuring. This is not only due to the different economic points of departure. Of equal importance are the abilities for and the orientation towards action on the part of the political actors, the executive abilities of the regional institutional arrangement and furthermore the shared behavioral codices as well as the joint regional visions. Regions in transition are usually politically and culturally fragmented, with the institutional system deteriorating, too. In these regions often just the collective resources are missing that enable the actors to coordinate their actions while searching for solutions to their problems (Capello, 1999). The case of Berlin and many other agglomerations in transition indicate that it is relatively easy to create new organizations to help stimulating the interaction among the actors. Far more difficult appears the generation of a new and common understanding, of a shared vision and shared rules of cooperation both internally and among these organizations and their actors regarding to the new demands of the knowledge economy (cf. Amin and Thrift, 1995:103). Opposed to traditional functional strategies based on a narrowly defined relation between objectives and measures, approaches emphasizing “contextual management” have been recommended for some time already to complement the operational concepts. Contextual management abandons the regulation of details and in contrast attempts to motivate the regional actors to draw up common visions and to collaborate in jointly translating these visions into action and measures (regarding the socio-economic context cf. Matzner, 1995).
KNOWLEDGE-INTENSIVE SERVICES AS A KEY SECTOR
105
NOTES 1.
By knowledge services not only the areas of telecommunication, multimedia, onlineservices, computer hardware and software are considered, but in a wider sense the acquisitions, production, joining, saving, supervision, interpretation and analysis of information. This includes e.g. also R&D, market research and consulting.
REFERENCES Amin, A., Thrift, N., “Globalisation, Institutional ‘Thickness’ and the Local Economy.” In Managing Cities: The New Urban Context, Healey, P. et al. eds. Chichester: Wiley, 1995. Andersson, A.E., “Creation, Innovation and Diffusion of Knowledge: General and Specific Economic Impacts.” In Technological Change, Economic Development and Space, Bertuglia, C.S., Fischer, M.M, Preto, G. eds. Berlin, Heidelberg: Springer, 1995. Capello, R., Spatial Transfer of Knowledge in High-Technology Milieux: Learning Versus Collective Learning Process. Regional Studies 1999; 33:353-365. Castells, M., Hall, P., Technopoles of the World. London: Routledge, 1994. Dunning, J.H., “Multinational Enterprises and the Growth of Services: Some Conceptual und Theoretical Issues.” In International Corporations in Services, Sauvant, K.P., Mallampally, P. eds. United Nations Library on Transnational Corporations. London, New York 1993; 12: 3474. Dybe, G., Kujath, H.-J., Hoffnungsträger Wirtschaftscluster: Unternehmensnetzwerke und regionale Innovationssysteme: Das Beispiel der deutschen Schienenfahrzeugindustrie. Berlin: Sigma, 2000. Foray, D., Lundvall, B.-A., “The Knowledge-Based Economy: From Economics of Knowledge to the Learning Economy.” In Employment and Growth in the Knowledge-Based Economy, OECD ed. Paris: Proceedings of an OECD, 1996. Freeman, C., Soete, L., The Economics of Industrial Innovation. Third Edition. London: Pinter, 1997. Matzner, E., “Der sozioökonomische Kontext. Argumente für eine neue (wirtschaftspolitische) Denkform.” In Die Reformfähigkeit der Industriegesellschaften, Bentele, K., Reissert, B., Schettkat, R. eds. Frankfurt/Main:Campus, 1995. Nonaka, I., Takeuchi, H., The Knowledge-Creating Company, London: Oxford University Press, 1995. Polanyi, M., Personal Knowledge: Towards a Post-Critical Philosophy. London: Routledge & Kegan Paul, 1958.
106
Hans Joachim Kujath
Rubalcaba-Bermejo, L., Business Services in European Industry: Growth, Employment and Competitiveness. Luxembourg: Office for Official Publications of the European Communities, 1999. Sassen, S., Metropolen des Weltmarkts. Die neue Rolle der Global Cities. Frankfurt: Campus, 1996. Sassen, S., “Global Cities and Global City-Regions.” In Global City-Regions. Trends, Theory, Policy, Scott, A.J. ed. Oxford, New York: Oxford University Press, 2001. Scott, A.J., Metropolis. From the Division of Labor to Urban Form. Berkeley: Univ. of California Pr., 1988. Soete, L., “Globalization, Employment and the Knowledge-Based Economy.” In Employment and Growth in the Knowledge-Based Economy, OECD ed. Paris:OECD, 1996. Sternberg, E., The Sectoral Cluster in Economic Development Policy: Lessons from Rochester and Buffalo. Economic Development Quarterly 11 1991, 5:342. Sternberg, R., Tamascy, C. Munich as Germany’s No. 1 High Technology Region: Empirical Evidence, Theoretical Explanations and the Role of Small Form/Large Firm Relationships. Regional Studies 1999; 33:367-377. Sträter, D. et al., Netzwerkstrukturen und Kooperationsbedingungen von Multimedia in der Region München. Arbeitsbericht 126. Stuttgart: Akademie für Technikfolgenabschätzung, 1998.
Chapter 6 ENTREPRENEURSHIP AS A SOURCE OF PATH DEPENDENCY
Udo Staber
6.1 INTRODUCTION In the arena of management and business strategy, entrepreneurs are often described as heroes and their organizations are considered places of worship and recipes for success. The admiration of entrepreneurs is also evident in much of the literature on regional economic development and in related literatures on industrial districts, business clusters, and innovation systems. Orthodox institutional accounts of regional development interpret entrepreneurship as providing reflexive and creative solutions to economic problems. Newer institutional approaches, while noting the constraining influence of local structures and history, also emphasize the ability of entrepreneurs to shape their environment strategically and with foresight. In general, entrepreneurship is normally considered a “seedbed for innovation” and a source of new opportunities, especially in less favored regions. This chapter explores a competing argument. In addition to being a source of innovation, entrepreneurship can also contribute to inertia in regional economic development processes. In many theoretical camps, it is often assumed that the relative abundance of particular business practices represents the outcome of some process (such as natural selection, strategic planning, or institutional regulation) yielding competitive advantages to those practices. Such thinking is based implicitly on the assumption of historical efficiency. By making this assumption, one expects that the cause-effect relations in the system play themselves out to an equilibrium state quickly and independently of the particulars of the development process (for a critique of this assumption, see Carroll and Harrison, 1994). However, there is plenty of research to indicate that development processes often contain positive feedback loops that can generate inferior outcomes. These processes can be economic, such as increasing returns to scale (Arthur, 1989), socio-political,
108
Udo Staber
such as reputation and legitimation (DiMaggio and Powell, 1983), or cognitive, such as cumulative learning (Nelson and Winter, 1982). Contexts where these processes occur are considered “path dependent,” in the sense that prior choices and actions affect future developments. I suggest that entrepreneurship is such a context for economic development, as it is implicated in the process of cumulative causation. Entrepreneurs are influenced by their own history, and entrepreneurship is a process that involves cycles of adaptation and adjustment among mutually dependent actors. The tendency to impute to entrepreneurs a capacity for novelty and fundamental change is flawed, to the extent that entrepreneurship contains the types of self-reinforcing mechanisms that can lead to path dependency, such as the tendency to build strongly coupled social networks and to learn myopically from success. Path dependency in entrepreneurship deserves attention in regional policy because of its implications for the ability of local actors to depart from existing routines and of regions develop new growth trajectories. Path dependency can reduce the variability of organizational forms and practices necessary for fundamental economic change. The chapter is organized as follows. It first outlines common assumptions about entrepreneurship and argue that the empirical evidence to validate these assumptions is rather thin. I then summarize arguments from a variety of literatures and theoretical perspectives in organization studies, strategic management, regional studies, and entrepreneurship that suggest a number of reasons why entrepreneurs do not always introduce significant variations, but rather reproduce existing organizational forms and business practices. In the final section, I indicate factors that are likely to moderate the extent to which entrepreneurs are locked into evolutionary paths. I discuss the implications of these contingencies for regional policy aimed at stimulating innovative entrepreneurship.
6.2 COMMON ASSUMPTIONS ABOUT ENTREPRENEURSHIP Conventional thinking on the subject of entrepreneurship is represented by the Schumpeterian ideal of innovations that destroy old competencies and enhance new ones (Schumpeter, 1934). Schumpeterian entrepreneurs are seen as persons who are “always on the go,” creating viable firms and introducing new products in markets they help to define. By unsettling current practices, they force a move away from equilibrium, as opposed to moving towards (a new) equilibrium. For regional planners, the value of entrepreneurs is that they introduce new organizational forms in the space of possibilities that
ENTREPRENEURSHIP AS A SOURCE OF PATH DEPENDENCY
109
might otherwise have gone unexplored. Entrepreneurs are said to (re-)vitalize regional economies, variously described as “hot spots” (Pouder and St. John, 1996), “industrial districts” (Brusco, 1982), “learning economies” (Asheim and Isaksen, 1997), or “innovation systems” (Cooke et al., 1997). The idea of entrepreneurs strategically creating useful variations contains several arguments, made not only in the “popular” literature, that appear as myths on closer examination (Harrison, 1994). Chief among these is the argument that the founding of a new business involves a series of relatively unproblematic decision-making and generic organizing processes. Founders are described as having well defined conceptions of what their organization’s strategy is or ought to be and of how to organize for maximum effectiveness. Relationships with discussion partners are sought strategically, and information is gathered mostly based on economic efficiency. However, many studies show that the entrepreneurial process is above all a social, and not always rational process, overlaid with social content, bound by social controls, and subject to constraining contextual influences (Low and Abrahamson, 1997). Personal beliefs and preferences keep organizing efforts minimalist and simple, and not always rational, with implications for path dependency, as I discuss below. The road to entrepreneurial success is long and thorny, rather than direct and smooth, as is popularly assumed. Entrepreneurship involves a long gestation period, in which founders make mistakes, change their strategies, reassemble resources, retrace past steps, and revitalize old connections. Reynolds and White (1997), for example, estimate that the average time required for building a small enterprise is at least one year, which does not include the preparation time spent before the firm is founded and registered. Only a very small minority of all organizing attempts end in a successful organizational founding. Most potential founders abandon their efforts before reaching the point of registering a new firm. Surveys in the United States found that, of the respondents indicating the intention to set up their own business, about two thirds made some preparation effort (e.g., training, identifying potential customers, and seeking financial support), but less than one third founded a business during the following three years (Katz, 1990; Reynolds and White, 1997). Many of the individuals who say they are “giving serious thought to starting a new business” are engaged in no more than a few activities, such as trying to obtain a license or writing a business plan, and most potential founders abandon their plans in the preparation phase altogether. Because available data are imprecise and probably not representative, the extent of potentially viable entrepreneurial innovations that “fail” before they become visible is not known (Katz and Gartner, 1988). The problems and liabilities of entrepreneurship do not end with successful founding. Most startup firms do not survive for long. Newly founded
110
Udo Staber
firms face obstacles that few are able to overcome. Founders need to convince suppliers and investors of their continued creditworthiness, employees of available career building opportunities, and special interest groups of their legitimacy. Studies in many countries reveal similarly high infancy failure rates, indicating that only about half of newly founded firms survive their first five years or so (Storey, 1994). A further argument is that business startups introduce variability in markets and represent significant deviations from current methods of doing business. The high organizational turnover (founding and failure rates) one can observe in some business populations is often viewed as an indication and source of entrepreneurship and innovation. However, the large majority of startups are not innovations in the sense that they open up new niches, create new organizational forms, and depart much from current knowledge. Research repeatedly shows that the vast majority of business founders enter industries in which they have experience (Sorenson and Audia, 2000) and that entrepreneurs do not stray too far from existing routines. Most founders add few new competencies to the industries they enter, but rather emulate existing organizational forms (Aldrich and Martinez, 2001). Most radically new ideas are weeded out in the organizing process, because of serious obstacles in resource acquisition and social legitimation. Despite contrary empirical evidence, the popular assumption of business founders adding a significant element of innovative entrepreneurship has remained remarkably resilient in regional policy circles, where it is often believed that economic growth is led by very small firms. Some commentators have gone as far as to suggest that “the example set by entrepreneurs offers a solution to the institutional, attitudinal, and cultural ills of presentday Western societies,” or that “small businesses are the seedcorn of our future prosperity” (quoted in Rainnie, 1985:146). The conventional policy implication has been to propose that development agencies should do everything they can to promote the start-up of new small businesses to stimulate economic growth. But this strategy is risky, because it imputes to business founders the role of innovators they can rarely fulfill.
6.3 ENTREPRENEURS ON THE ROAD TO PATH DEPENDENCY Evolutionary economists have observed that firms within an industry tend to converge toward certain practices, specifically those practices that were selected and retained by efficient market forces (Alchian, 1950; Hirshleiffer, 1977). By extension, firms conforming to prevailing practices should, on
ENTREPRENEURSHIP AS A SOURCE OF PATH DEPENDENCY
111
average, accrue positive returns. Deviant firms, by contrast, assume the risks of divergent approaches and incur the inefficiencies of experimenting with (potentially) suboptimal approaches. Similarly, one may expect that most founders will shy away from divergent practices and, instead, will reproduce existing organizational forms. To discern situations where entrepreneurial activities are locked into established routines and where entrepreneurs are not able to identify and capitalize on new opportunities, one may look for conditions supporting positive feedback mechanisms. If positive linkages are strong enough, they create positive feedback (Arthur, 1989). Below, I discuss a number of factors, buttressed by empirical research, which can be understood as self-reinforcing mechanisms. They reflect the difficulties entrepreneurs face when trying to introduce innovations, and the resulting tendency to converge on current practices, rather than travel entirely new paths.
6.3.1 Entrepreneurs Face Pressures of Social Conformity The challenges involved in getting a new business on the road to success are extremely difficult to master, even for founders with business experience. In the kinds of uncertain environments in which entrepreneurs are expected to thrive, there are few predictive templates for behavior that will guarantee success. What makes the discovery of useful templates so difficult is that it requires the entrepreneur to make reasoned judgments about the future, a period about which ignorance is bound to remain considerable. This ignorance will frustrate any attempt to select the appropriate strategies and practices. Only through post hoc analysis of what worked in the past can one begin to understand the mechanisms operating in any given locale or context. However, because such analysis is oriented to the past, it can also generate the kinds of self-reinforcing processes that lead to path-dependent developments. Because of obstacles and uncertainties in the founding process, rational founders often rely on existing routines and conform to existing models, rather than deviate far from well-traveled paths. While there are pressures toward conformity inherent in all social domains, one may expect these pressures to be especially strong in entrepreneurial domains because of the uncertainties surrounding the discovery of which products, services, organizational forms, and so forth are likely to be accepted. Entrepreneurs operate in a social space that defines the meaning of business practices. They read specialized magazines, attend specialized workshops and training seminars, and contribute to focused interest associations. Through their participation in social space they come to learn and accept practices and organizational forms considered desirable and legitimate (Aldrich and Fiol, 1994).
112
Udo Staber
Like other individuals, entrepreneurs are embedded in local social structures that induce them to follow the more established routes, rather than challenge the dominant cultural paradigm (Staber and Aldrich, 1995; Johannisson and Monsted, 1997). Social embeddedness is particularly strong in geographically peripheral regions with long traditions of how to conduct business (Kristensen, 1994). In these regions, local institutional structures and processes operate as social controls that inhibit norm-breaking behavior. Local structures may support the kind of valuable and sustainable distinctiveness that can be an important source of the region’s competitive advantage. But they may also stabilize options and narrow the range of perspectives, and thus constrain development paths, rather than open up entirely new possibilities. To the extent that entrepreneurial activities are conducted in a context of beliefs about what entrepreneurs should do when they innovate, the outcomes of entrepreneurship are social products. These are only incompletely specifiable in the beliefs of any one entrepreneur but are part of a larger social context within which personal relationships develop. Entrepreneurs have been described as business persons acting on the brink of insanity and running their business as a symbol of rebellion or as a drama revolving around psychological needs (Kets de Vries, 1996). However, this view downplays the importance of social relationships introducing order in the process of entrepreneurship. It overlooks the fact that entrepreneurship requires a certain level of trust to minimize relational risk with respect to partners, stakeholders, and resource providers. Entrepreneurs face social pressures to conform. They move in social space that requires some degree of familiarity and mutual understanding about the meaning of business activity. Mutual understandings can lead to commitments with self-reinforcing consequences, as research on the formation and development of social networks in business indicates (Seabright et al., 1992; Nooteboom et al., 1997). Repeated interactions encourage the formation of habits and routines, which are not easily deflected by otherwise disruptive events.
6.3.2 Success Constrains Innovation While most entrepreneurs have difficulty finding the right path to success, those who do reach success may become locked into a pattern of perception and behavior that constrains, rather than enables, further innovation. Success in the past is not always the best route to innovation in the future. Much empirical research on entrepreneurship focuses on the presumed causes of success, including personal traits (e.g., achievement orientation, self-esteem), human capital features (e.g., managerial and occupational ex-
ENTREPRENEURSHIP AS A SOURCE OF PATH DEPENDENCY
113
perience), or contextual factors (e.g., market competition, institutional supports). The typical hypothesis is that business owners who have the right traits and operate under the right circumstances are more likely to behave and succeed as entrepreneurs. However, the results of this research have been mixed. In part, this is because most studies suffer from a selection bias, by sampling only successful individuals (those who have successfully founded a business) and not evaluating their attributes against a comparison group. In addition, investigators normally work with cross-sectional data sets that make it difficult to examine cause and effect relations. It is possible that some of the factors that are believed to be causes of entrepreneurial success, such as internal locus of control, focused strategies, and institutional support, are actually its consequences. Success may cause the entrepreneur to focus on and attribute credit to a narrower range of options and practices than is normally associated with the idea of entrepreneurship, focusing on those perceptual and behavioral outcomes that have been valued to begin with. That is, success may have selfperpetuating consequences, keeping individuals from looking beyond the range of factors already known to them. Among the perceptual consequences are the ways in which entrepreneurs learn about why they succeeded. Such learning is not an unbiased activity, but often involves “superstitious learning” and “competency traps” (Levitt and March, 1988), myopia (Levinthal and March, 1993), and attribution errors (Staw et al., 1981). Research on learning in organizations has uncovered a tendency of individuals to attribute outcomes to the events that preceded them. These events are often chosen from a highly limited set of possibilities, a set that is constrained by selective attention and well-tried routines. When outcomes are positive, individuals may attribute them to internal practices and decisions on which they already have focused. In the process of weeding out negative impressions, myopic learning and faulty attribution induce individuals to give increased attention to fewer options and practices that they continue to exploit. They may learn over time to focus their attention on fewer strategic choices and to concentrate on those actions that they believe to be the cause of success. Entrepreneurs may be especially inclined toward such biased learning, to the extent that they are driven by narcissistic needs of control and self-aggrandizement, primitive defense mechanisms, and other irrational motives (Kets de Vries, 1996). Faulty learning is more likely to occur after perceived success than failure. When entrepreneurs mistakenly credit a particular practice or decision for a positive outcome, they obtain little information to tell them they might be wrong. A re-evaluation of past practices, consultation, soul-searching, and, eventually, a reorientation toward new actions are more likely to occur when an outcome is perceived as failure. However, given success, entrepreneurs
114
Udo Staber
face the risk of becoming overconfident in their own capacities, and such overconfidence can lead to complacency. As a result, they may become reluctant to explore new opportunities and, instead, prefer to exploit their current activities. Confident that they are doing the right thing, they fail to monitor closely the needs of customers, the perceptions of partners, or the actions of rivals. Their strategies may become so focused on a narrow range of possibilities that they are likely to overlook important changes in the environment (Miller, 1993). As a result, the possibility of avoiding or breaking out of pathdependent developments is reduced.
6.3.3 Imitation Reduces Variability Under conditions of uncertainty, many business founders take the present as a guide to the future and look for organizational and behavioral models that can be emulated (Brittain and Freeman, 1980). Imitation represents a potentially highly efficient means to enter an industry, by avoiding some of the costs of trial-and-error learning, extending the organization’s repertoire of potentially effective practices, and fulfilling critical legitimation requirements. There is a long history of theory and research on the effects of uncertainty on behavior in organizations, with implications for entrepreneurship. Most of this research focuses on the problems that uncertainty causes for decision makers in their adoption of appropriate structures and processes (March, 1981). To cope with uncertainty, decision makers may take action to buffer their organization from the effects of external contingencies. Alternatively, they may look outside their organizations and incorporate the practices and routines of other organizations operating in their domain. Various organizational and social psychological theories support this idea. Social comparison theory predicts that uncertainty causes individuals to compare their behavior with that of relevant others and adjust accordingly. Neo-institutionalist theorists argue that uncertainty drives mimetic behavior, as individuals adopt the legitimated practices of others. Learning theories predict that individuals look towards others for behaviors that have reinforcing consequences. And decision-making theories propose that uncertainty causes individuals to economize on search and evaluation costs by imitating relevant others. Entrepreneurs may prefer to copy, or make only minor adjustments, to what they perceive are the successful practices and organizational forms in their relevant environment, as they attempt to cope with the uncertainties surrounding the introduction of innovations. To the extent that innovations lack the taken-for-granted character of well-tried products and forms, entrepreneurs will find it difficult to attract external funding, qualified labor, and other vital material and symbolic resources. Hence, they will look towards
ENTREPRENEURSHIP AS A SOURCE OF PATH DEPENDENCY
115
others who have launched businesses in the same or similar industries as cues about the state of the environment (Stuart et al., 1999). In new industries, knowledge about successful practices and organizational forms is often available only to “insiders,” and the essential features of innovations have not been codified yet. In such situations, existing organizations are the only training grounds for knowledgeable entrepreneurs (Brittain and Freeman, 1980). Potential entrepreneurs may look towards existing organizations for indicators of successful innovations. They may interpret an increase in the number of business startups in their domain as indicators of the receptiveness of the environment to additional firm creations. Research in organizational ecology has shown a tendency among business founders to imitate others in expanding business populations. Growing organizational populations broaden the networks that connect persons with the inclination and skills to succeed in creating particular organizational forms (Staber, 1997). The imitation of existing organizational forms may, over the long term, have noxious consequences for entrepreneurship, at the level of the individual founder, as well as the level of industry and regional economy. Copying may limit an entrepreneur’s search for alternative trials. On the assumption that entrepreneurs, like other founders, invest in learning about existing organizational models only if they expect an improvement in performance, they will have overconfidence that the outcomes they achieved from imitation were the best that could be achieved. Their conclusions will then be based not on actual results from multiple attempts of imitation (some of which would have failed), but rather on a comparison with the presumed results of other imitations that were never tried. Gould (1990) explored this problem as a general problem of historical reasoning in evolutionary theory. As long as there is no obvious and immediate failure of imitation, actors may rationalize that the outcome would have been worse if the imitation had not been attempted. If an imitation occurs, and if the result is an improvement over the previous state, actors will tend to conclude that it is the imitation that produced the improvement. The consequence of this reasoning is that the entrepreneur will not attempt alternative trials, including some that have greater potential value and that break new ground, but will conform to the existing routines. At the level of the population of entrepreneurs, or the level of the regional economy, practices will then converge on the routines that satisfy a given level of performance, and one that is assumed to be the best achievable. Entrepreneurs will not test new variations and will be content to engage in those practices that are considered legitimate. Copying, although efficient, thus prevents path-breaking discoveries that are potentially more efficacious. Moreover, if imitability improves over time, if, for example, learning processes become more efficient, path dependency may deepen. The presence of learning routines and the improved
116
Udo Staber
effectiveness of acquiring specialized knowledge may reduce the actor’s capacity to absorb new ideas and to widen the horizon of alternative possibilities.
6.3.4 Entrepreneurial Networks Tend Towards Closure Entrepreneurs are not only constrained to travel along old paths, for reasons explained above, but they also don’t travel them alone. Although entrepreneurs are often depicted in the popular media as individuals pursuing their idiosyncratic ideas single-handedly and without much reliance on others, they normally draw on the input from a variety of individuals with whom they are connected through social networks (Reynolds and White, 1997; Stuart et al., 1999). In the struggle to find a place for themselves in emerging markets, social networks are crucial assets for business founders (Dubini and Aldrich, 1991; Sorenson and Audia, 2000). Networking allows them to enlarge their scope of action, economize on time, and gain access to resources and opportunities otherwise unavailable. The question is, what kinds of networks do they construct, given the environment in which they operate? Do they draw on established connections to familiar persons, with a proven record of reliability and cooperation? Or do they pragmatically build new relations with previously unknown individuals and discard those individuals who are no longer useful? The neo-classical economic approach to entrepreneurship treats business owners as rational actors who pursue clearly economic goals and make decisions based on their assessment of relative risks and opportunities (Knight, 1921). Entrepreneurs are described as individuals who fearlessly enter unfamiliar territory and seek advice and resources from whoever might be willing to provide them. Information is seen as a commodity and networks as the marketplace in which information is traded. This perspective views social networks as a means of obtaining access to resources with exchange value. Founders build social networks pragmatically, selecting and discarding members based on their perceived pecuniary contributions and changing business needs. From this perspective, networks are transitory, pragmatic, and held together by the narrow economic calculus of individual actors. But most studies of the networks of business founders indicate that an economic analysis of entrepreneurship cannot be separated from an understanding of the social and institutional context in which economic relations are embedded (Granovetter, 1993). Research on new venture creation, for example, finds that the market’s assessment of the value of ventures relies heavily on the social structure in which the ventures are embedded, especially in uncertain environments. The greater the level of uncertainty, the greater is
ENTREPRENEURSHIP AS A SOURCE OF PATH DEPENDENCY
117
the value of close social ties (Stuart et al., 1999). Entrepreneurs behave rationally and instrumentally if they conduct their economic activities through personal networks, which include persons on whom they can rely. Entrepreneurial exchanges may begin in an impersonal way, but often develop into social relations that provide access to information and give it meaning. Mutual trust and the force of habit can be powerful mechanisms for blunting opportunistic motives. Close social networks do not automatically lead to path dependency. Entrepreneurs, like other business founders, make mistakes as they acquire and discard network members in their constant search for useful information. And some mistakes can lead to significant innovations with big returns. But the greater risk is that in social networks with strong social ties built on intimate friendship and long-standing relations, people’s visions are clouded and their abilities to make pragmatic business decisions are constrained. One may imagine a close-knit entrepreneurial network that persists because trust is enforced through the ability of network members to confer unique rewards (Portes and Sensenbrenner, 1993). The network’s sanctioning capacity may be used to police what are considered deviant ideas and practices. Or, cohesive ties may amplify the pressure to reciprocate favors and investments. The more cohesive the network, the greater the risk of acquiring a tainted reputation for cutting ties (Raub and Weesie, 1990), and thus the greater the disincentive to create new ties. Network members are locked into endless mutual exchanges, even if they see no further benefits from the exchange. Strong bonds may also serve as a filter for information and perspectives reaching the entrepreneur, thus creating a cognitive lock-in that isolates him/her from other, potentially more valuable sources of information (Uzzi, 1997). The entrepreneurial dilemma is that, while networking provides access to resources and information necessary for innovation, social ties may force the individual to conform to current understandings and practices. To the extent that entrepreneurial networks tend towards social closure, they contribute to path dependency.
6.4
CONTINGENCIES OF PATH DEPENDENCY
None of the above is meant to suggest that entrepreneurship will always produce outcomes that increase the likelihood of path dependency. The extent to which entrepreneurial behavior is path-dependent is moderated by a number of factors. Environmental resource diversity, error and chance, and the imperfections of imitation influence the variety of evolutionary paths that may be travelled in search of entrepreneurial opportunities.
118
Udo Staber
6.4.1 Environmental Diversity Ecological models of industry evolution hold that the structure of interdependencies among existing firms affects the branching of new resource spaces and the ability of existing firms to identify and exploit new resource spaces (Romanelli, 1989). Interdependence has evolutionary value in a system of actors who, to the extent that they follow different institutional logics and practices, contribute to the system’s information richness and innovative capacity. When environmental selection operates from a broad and diverse range of options, the risk decreases that entrepreneurial strategies create path dependencies that lead to an evolutionary dead end. This argument predicts that innovative entrepreneurs are most likely to emerge in regions that are decentralized, contain diverse industrial competencies, and have institutions which nurture a repertoire of diverse, and possibly even incompatible routines and solutions (Grabher and Stark, 1997). Studies in industrial geography have demonstrated the benefits of knowledge spillovers in economically diversified regions, as reflected, for example, in higher adoption rates of new technologies (Harrison et al., 1996) and lower business failure rates (Staber, 2001). Entrepreneurs may benefit from the crossfertilization of ideas from different lines of work. The innovative possibilities of entrepreneurship are thus tempered by the diversity of surrounding institutions and institutional structures. The question remains how regions develop diverse institutions and institutional structures in the first place. If entrepreneurship benefits from environmental diversity, then diverse regions are more likely to stimulate new developments than homogeneous regions. But to the extent that the less favored or declining regions have more homogeneous industrial structures, they are likely to continue on their old development paths.
6.4.2 Uncertainty, Error, and Chance Evolutionary theory considers environmental uncertainty an important source of entrepreneurial variation, because it creates incentives and opportunities for new learning (Campbell, 1965). Uncertainty may raise the awareness of entrepreneurs to changing conditions and the continual need to adapt, and may prevent them from becoming too complacent. It may broaden entrepreneurs’ perspectives by presenting challenges that are impossible to ignore. By contrast, entrepreneurs operating in stable environments are more likely to be constrained by their past behavior and to adopt practices they are loath to abandon, especially if their past performance has been within acceptable lim-
ENTREPRENEURSHIP AS A SOURCE OF PATH DEPENDENCY
119
its. One would expect that environmental uncertainty reduces the likelihood that entrepreneurs become locked in evolutionary paths. Reference to environmental uncertainty alone does not explain the mechanisms by which actors identify and mobilize the available resources and how they imbue the structure of interdependencies with social meaning in the process of resource mobilization. Entrepreneurs are critical agents as they fill new resource space, but they also make mistakes, re-trace old tracks, and end up in dead ends. Entrepreneurial strategies of resource mobilization are not free of error and noise. Several lines of research in business strategy, organization theory, and entrepreneurship suggest that decision processes contain elements of surprise and non-rationality. When individuals’ goals are ambiguous, means-ends relations are unclear, and participation in a system of actors is fluid, decisions tend to have a strong random component and learning is largely haphazard (Starbuck, 1983). Scholars have long noted that errors and chance events can introduce variety into business populations, with potentially adaptive value. Opportunistic tendencies, coupled with bounded rationality, increase the risk that individuals will choose suboptimal solutions. However, currently suboptimal solutions may also carry the seeds of success in different future circumstances. Environmental uncertainty thus creates opportunities for new discoveries. But this effect is tempered by the tendency of entrepreneurs to seek security in stable social networks and to imitate current practices, as explained above.
6.4.3 Imperfect Imitation Neo-institutional theory predicts that uncertainty concerning the context in which imitation takes place plays an important role in modeling processes and outcomes (DiMaggio and Powell, 1983). The precise impact of uncertainty, however, may depend on the individual’s assessment of the value of the practice or organizational form to be copied. Simulation studies of bandwagon adoption, for example, show that greater differences between the adopter’s assessment of returns to a particular innovation can cause lower bandwagon effects (Abrahamson and Rosenkopf, 1993). Uncertainty above a certain threshold level can also prompt the imitator to attempt a variety of imitation modes (Narduzzo and Warglien, 1996), such as copying practices assumed to be successful or emulating the most commonly used practices. Such behaviors may raise the possibility of introducing new and pathbreaking innovations with potentially greater value. Research on decision-making and learning suggests that actors suffer from information processing limitations, attribution errors, escalating com-
120
Udo Staber
mitments, excessive self-confidence, and unwillingness to look beyond their personal perspectives. These limitations may lead to outcomes that are completely disconnected from intentions, especially under conditions of uncertainty, thus introducing error in the imitation process. Imitation may also be imperfect because of uncertainty about an organizational model’s success or an industry’s attractiveness, especially at the early stage in the development of an organization or industry. In addition, the causes of superior performance may not be well understood, even if the level of performance is perceived accurately. The ambiguities surrounding imitation present difficult challenges for the entrepreneur and may lead to the adoption of ineffective practices. Causal ambiguity limits an entrepreneur’s ability to learn and emulate best practices, and thus induces the entrepreneur to depart (unknowingly) from previously traveled paths. Copying existing models may, in the end, only guarantee that there will be enough failures upon which environmental selection can operate. However, imperfect imitation may also lead to the adoption of deviant practices, some of which may turn out to have greater adaptive value, depending on environmental circumstances. Imperfect imitation thus increases the range of possible variations (Campbell, 1965).
6.5 POSSIBILITIES FOR POLICY INTERVENTION The moderating factors identified above have implications for policy interventions that aim at the creation of new development paths. The costs of coping with environmental diversity, guarding against environmental uncertainty, and reducing decision-making errors may be so substantial for the individual entrepreneur that they eliminate the majority of (potential) startups from the range of possibilities. But limitations at the individual level may have higher-level collective value by increasing the range of variation necessary for innovation to occur. Imperfections, mistakes, and chance events ensure that positive feedback loops at the system level are not always instantaneous or closed. In a given business population and at any given time, there will be a mix of adaptive and maladaptive routines, some of which may blossom into entrepreneurial and innovative startups (Gartner et al., 1989), as long as environmental selection forces can operate. It is at the system level where strategic policy intervention may have its greatest impact on entrepreneurship as a source of new variations with potential value. Policy intervention should address the system of interaction, rather than the intentions and skills of individual entrepreneurs, and should orient itself to the interdependencies among relevant actors in the system. The idea is to encourage the formation and development of networks of entrepreneurs and
ENTREPRENEURSHIP AS A SOURCE OF PATH DEPENDENCY
121
actors in the institutional setting who are loosely embedded in a dense, yet open and dynamic web of economic and social relations. Loose embeddedness means, in this context, that relations evolve through a dynamic mix of competition and cooperation, to provide both the incentive and capability to innovate. Only when specialization and idiosyncrasy on the part of individual founders are matched with complementarity, social trust, and a sense of collective purpose is the capacity of the regional production system as a whole enhanced to diversify its products and venture into new markets. The policy goal should be to create a regional production and exchange complex with sufficient levels of interaction to prevent entrepreneurs from working in isolation from one another and ending up in an unproductive competitive squeeze. The policy argument advanced here builds on the evolutionary logic of variation as the raw material for change and innovation (Campbell, 1965). For innovation to be possible, there must be a variety of options from which to select. The greater the range of variation in practices, organizational forms, strategies, and so forth, whatever their source, the greater the opportunities for change, including change with more desirable outcomes. The problem is that whether the available options are useful, in the sense of opening up new development opportunities, is not known ex ante, but can be known only after the fact (Dennett, 1995). The outcomes are essentially indeterminate, involving processes and selection criteria that are difficult to control, certainly at the level of regional economic development. Thus, a useful policy strategy may be to invest in the discovery of possible alternatives, of any kind, rather than seeking to minimize strategic conflicts, encourage convergence towards a particular set of agreed-upon routines, and reduce informational redundancies. Based on this reasoning, policy initiatives should support diversity and tolerate ambiguity. Investments should be made to initiate relationships that encourage exchange among actors with diverse competencies and preferences. Policy should invest in building a reservoir of diverse, but shared knowledge, to ensure the possibility of continuous learning. Investments could be in form of trials and building databases that enable comparisons with the existing routines, to broaden the range of possible solutions to be considered. Institutional actors can also add to variation and new learning by supporting the creation and exchange of tacit knowledge, and not only articulated and codified knowledge. Tacit knowledge has its strongest effect on innovation when individuals with diverse and complementary knowledge can meet and collectively articulate their ideas (Lawson and Lorenz, 1999). By contrast, a policy that is premised on a belief in strategic clarity and prescience runs the risk of cutting off prematurely potential sources of innovation.
122
Udo Staber
6.6 CONCLUSION Entrepreneurs are usually described as individuals with rare characteristics. Some of these characteristics reflect valued traits, such as decisive leadership skills and strong achievement motivation. Others resemble the deficits of people suffering from psychological ailments that make them look foolish. The entrepreneurial desire to try out new things may indeed require practices that would normally be considered foolish: Preparing for the unknown, working within unknown contexts, and developing entirely new resources (Aldrich and Fiol, 1994). Foolish behavior includes experimentation, deviation, and correction, sometimes with clear intentions, but often with unintended consequences. And it is in this capacity that entrepreneurship is often viewed as a means of avoiding or breaking out of path-dependent developments. Entrepreneurship may indeed offer regions a way out of developmental lock-ins, but it is also constrained by existing structures and practices, as discussed above. Specific entrepreneurial practices face serious legitimation hurdles, especially if they contradict taken-for-granted assumptions about how to conduct business. Entrepreneurs face different challenges than business founders who merely reproduce or extend existing organizational forms. Whereas founders who enter existing markets and work with existing organizational models need to understand primarily the constraints that are already in place, innovative entrepreneurs operate in a situation of extreme uncertainty and fluidity. Environmental selection forces may be severe and may not be understood as well as those confronting other founders. Creating and marketing novel products and services may require new organizational forms and practices, which raise legitimization issues and call for novel strategies. The problem is that entrepreneurs may need to portray their new activities as sufficiently familiar to obtain social acceptance, but such portrayal is inconsistent with the need to distinguish their activities as unique and valuable to build competitive advantages. Because it is difficult for entrepreneurs traveling in unchartered waters to attract critical resources in an environment that offers no clear models, they depend heavily on institutional support from government bodies, consultants, funding agencies, interest associations, and so forth, but such support is inherently precarious. The long shadow cast by history over the local institutional environment and social infrastructure sets limits to the possibilities of path-breaking entrepreneurship. The policy implications are not as straightforward as one might hope. The enormous investment in time and effort committed during the last few decades to understand and promote entrepreneurship as a means to develop less favored regions or revitalize ailing regional economies has not led to the realization of its full potential. Many investigators have been content with
ENTREPRENEURSHIP AS A SOURCE OF PATH DEPENDENCY
123
enumerating attributes that are thought necessary for promoting regional entrepreneurship. The list of attributes includes cultural resources such as trust and local identity, as well as infrastructural resources such as venture capital and support networks comprised of universities, research institutes, technology transfer centers, training institutes, and consultants. The lesson of this literature, as interpreted by many policy practitioners, is that regions which combine these attributes will be able “to grow the next Silicon Valley” (Miller and Cote, 1985), and that the right combination of attributes can be “engineered,” with more or less predictable outcomes. However, the various processes by which these attributes can be brought to life and trigger innovation are less well understood. I propose that policy makers should ask if choice, action, and discretion really occur in the sense of guiding and directing entrepreneurs, or whether entrepreneurship is best described as an outcome of “natural selection” or self-organizing processes. While individual agency does matter, it is circumscribed by inherited cultural and institutional structures that constrain future developments. Attempts by policy makers to identify generic elements of innovative entrepreneurship and to proceed on the assumption that the intended outcomes of entrepreneurship can be created strategically are understandable, but flawed. It is impossible in a fluid and uncertain environment, which includes contingencies existing outside the region and beyond political control, to devise the right strategy for success. The best lesson for regional policy may be that the definition of developmental strategies should be left to the business actors themselves, within an institutional framework that supports variation in practices and organizational forms mobilizes a variety of autonomous organizations, and to let entrepreneurial practices unfold “naturally.” While an institutional framework can be built to facilitate entrepreneurship by supporting new knowledge creation, disseminating information, and widening the range of possibilities in general, the specific solutions chosen must be sensitive to local path dependencies, without being wholly governed by them. Unique combinations of competition and cooperation are a discovery process that is the driving force of the regional production system. Thus, different regions will experience different ways of preserving or breaking out of path dependencies.
REFERENCES Abrahamson, E., Rosenkopf, L., Institutional and competitive bandwagons: Using mathematical modeling as a tool to explore diffusion of innovation. Academy of Management Review 1993; 18:487-517.
124
Udo Staber
Alchian, A., Uncertainty, evolution, and economic theory. Journal of Political Economy 1950; 58:211-221. Aldrich, H., Martinez, M., Many are called, but few are chosen: An evolutionary perspective for the study of entrepreneurship. Discussion paper, Chapel Hill: University of North Carolina, 2001. Aldrich, H., Fiol, C., Fools rush in? The institutional context of industry creation. Academy of Management Review 1994; 19:645-670. Arthur, W., Competing technologies, increasing returns and lock-in by historical events. Economic Journal 1989; 99:116-131. Asheim, B., Isaksen, A., Location, agglomeration and innovation. European Planning Studies 1997; 5: 299-330. Brittain, J., Freeman, J., “Organizational proliferation and density dependent selection.” In The Organizational Life Cycle, J. Kimberly, R. Miles eds. San Francisco: Jossey-Bass 1980. Brusco S., The Emilian model: Productive disintegration and social integration. Cambridge Journal of Economics 1982; 6:167-184. Campbell, D., “Variation and selective retention in socio-cultural evolution.” In Social Change in Developing Areas: A Reinterpretation of Evolutionary Theory, H. Barringer, R. Mack eds. Cambridge: Schenkman 1965. Carroll, G., Harrison, J., On the historical efficiency of competition between organizational populations. American Journal of Sociology 1994; 100:720-749. Cooke, P., Uranga, M. Etxebarria G. Regional innovation systems: Institutional and organisational dimensions. Research Policy 1997; 26:475-491. Dennett, D., Darwin’s Dangerous Idea: Evolution and the Meanings of Life. London: Simon and Schuster, 1995. DiMaggio, P., Powell, W., The iron cage revisited: Institutional isomorphism and collective rationality in organizational fields. American Sociological Review 1983; 48:147-160. Dubini, P., Aldrich, H., Personal and extended networks are central to the entrepreneurial process. Journal of Business Venturing 1991; 6:305-313. Gartner, W., Mitchell, T., Vesper, K., A taxonomy of new business ventures. Journal of Business Venturing 1989; 4:169-186. Gould, S., Wonderful Life: The Burgess Shale and the Nature of History. New York: Norton, 1990. Grabher, G., Stark D., Organizing diversity: Evolutionary theory, network analysis and postsocialism. Regional Studies 1997; 31:533-544. Granovetter, M., “The nature of economic relationships.” In Explorations in Economic Sociology, R. Swedberg ed. New York: Russell Sage Foundation 1993.
ENTREPRENEURSHIP AS A SOURCE OF PATH DEPENDENCY
125
Harrison, B., The small firms myth. California Management Review 1994; 36:142-158. Harrison, B., Kelley, M., Grant, J., Specialization versus diversity in local economies: The implications for innovative private-sector behavior. Cityscape 1996; 2:61-93. Hirshleiffer, J., Economics from a biological viewpoint. Journal of Law and Economics 1977, 20:1-52. Johannisson, B., Monsted, M., Contextualizing entrepreneurial networking: The case of Scandinavia. International Studies of Management and Organization 1997; 27:109-136. Katz, J., Longitudinal analysis of self-employment follow-through. Entrepreneurship and Regional Development 1990; 2:15-25. Katz, J., Gartner, W., Properties of emerging organizations. Academy of Management Review 1988; 13:429-441. Kets de Vries, M., The anatomy of the entrepreneur: Clinical observations. Human Relations 1996; 49: 853-883. Knight, F., Risk, Uncertainty and Profit. New York: Houghton Mifflin, 1921. Kristensen, P., Spectator communities and entrepreneurial districts. Entrepreneurship and Regional Development 1994; 6:177-198. Lawson, C., Lorenz E., Collective learning, tacit knowledge and regional innovative capacity. Regional Studies 1999; 33:305-317. Levinthal, D., March, J., The myopia of learning. Strategic Management Journal 1993; 14:95112. Levitt, B., March, J., Organizational learning. Annual Review of Sociology 1988; 14:319-340. Low M., Abrahamson E., Movements, bandwagons, and clones: Industry evolution and the entrepreneurial process. Journal of Business Venturing 1997; 12:435-457. March, J., “Decisions in organizations and theories of choice.” In Perspectives on Organization Design and Behavior, A. Van de Van, W. Joyce eds. New York: Wiley 1981. Miller, D., The architecture of simplicity. Academy of Management Review 1993; 18:116138. Miller, R., Cote, M., Growing the next Silicon Valley. Harvard Business Review 1985; JulyAugust: 114-123. Narduzzo, A., Warglien, M., Learning from the experience of others: An experiment on information contagion. Industrial and Corporate Change 1996; 5:113-126. Nelson, R., Winter, S., An Evolutionary Theory of Economic Change. Cambridge: Harvard University Press, 1982. Nooteboom, B., Berger, H., Noorderhaven N., Effects of trust and governance on relational risk. Academy of Management Journal 1997; 40:308-338.
126
Udo Staber
Portes, A., Sensenbrenner, J., Embeddedness and immigration: Notes on the social determinants of economic action. American Journal of Sociology 1993; 98:1320-1350. Pouder, R., St. John, C., Hot spots and blind spots: Geographical clusters of firms and innovation. Academy of Management Review 1996; 21:1192-1225. Rainnie, A., Small firms, big problems: The political economy of small businesses. Capital and Class 1985; 25:140-168. Raub W., Weesie J., Reputation and efficiency in social interactions: An example of network effects. American Journal of Sociology 1990; 96:626-654. Reynolds, P., White, S., The Entrepreneurial Process: Economic Growth, Men, Women, and Minorities. Westport: Quorum, 1997. Romanelli, E., “Organizational birth and population variety: A community perspective on origins.” In Research in Organizational Behavior vol. 11, B. Staw, L. Cummings eds. Greenwich: JAI Press, 1989. Schumpeter, J., The Theory of Economic Development: An Inquiry into Profits, Capital, Credit, Interest and the Business Cycle. Cambridge: Harvard University Press, 1934. Seabright, M., Levinthal, D., Fichman, M., Role of individual attachments in the dissolution of interorganizational relationships. Academy of Management Journal 1992; 35:122-160. Sorenson O., Audia P., The social structure of entrepreneurial activity: Geographic concentration of footwear production in the United States, 1940-1989. American Journal of Sociology 2000; 106:424-462. Staber, U., An ecological perspective on entrepreneurship in industrial districts. Entrepreneurship and Regional Development 1997; 9:45-64. Staber, U., Spatial proximity and firm survival in a declining industrial district: The case of Baden-Württemberg. Regional Studies 2001; 35:329-341. Staber, U., Aldrich H., Cross-national similarities in the personal ties of small business owners. Canadian Journal of Sociology 1995; 20:441-467. Starbuck, W., Organizations as action generators. American Sociological Review 1983; 48:91102. Staw B., Sandelands, Dutton, J., Threat-rigidity effects in organizational behavior: A multilevel analysis. Administrative Science Quarterly 1981; 26:501-524. Storey, D., Understanding the Small Business Sector. London: Routledge, 1994. Stuart, T., Hoang, H., Hybels, R., Interorganizational endorsements and the performance of entrepreneurial ventures. Administrative Science Quarterly 1999; 44:315-349. Uzzi, B., Social structure and competition in interfirm networks: The paradox of embeddedness. Administrative Science Quarterly 1997; 42:35-67.
Chapter 7 GEOGRAPHICAL PROXIMITY AND THE DIFFUSION OF KNOWLEDGE The case of SMEs in biotechnology
Delphine Gallaud and André Torre
7.1 INTRODUCTION Since the principle that the capacity for innovation is a driving force in the growth of firms or other productive systems has been acknowledged, public policies hold to the view that geographical proximity plays a part in the process of the circulation of technology and knowledge. Fostering the kind of face-to-face relationships needed to establish and maintain a common pool of knowledge about companies or business concerns has become a public policy concern. It is deemed necessary to encourage local interactions by promoting the setting-up of networks or local systems of innovation (e.g. clusters or technopoles) where the circulation of knowledge is reinforced by the opportunity for frequent contacts, thanks to the common location of the actors (Porter, 2000). This chapter aims to question the relevance of these ideas, to reject all idealistic notions on the subject, and to enquire as to whether geographical proximity is really needed for the diffusion and exchange of knowledge (see Rallet and Torre, 2000). In particular, we will ask ourselves about the spatial dimensions of technology interactions. The fundamental question we shall raise concerns the relation between geographical proximity and the different modes of external acquisition of technologies. More precisely, do firms need permanent geographical proximity in order to acquire the technological knowledge they need to innovate? Can they do without this variable? Alternatively, do they only need interactions mobilizing geographical proximity more temporarily – that is only during certain stages of the R&D and innovation process? We shall briefly present the studies that uphold the idea of the importance of permanent geographical proximity in the diffusion of knowledge, regrouping of firms and spatial spillovers. Then, we analyze the process of
128
Delphine Gallaud and André Torre
the transmission of knowledge, first by assessing the absorptive capacity of the firms, then by integrating the spatial dimension of the external acquisition of technology. This will allow us to highlight the possibility of satisfying the needs for geographical proximity through a temporary coming together of the actors of the innovation process. We will end with an evaluation of the localized nature of innovation interactions in the field of biotechnology in France, by comparing pharmaceutical activities to those related to agriculture and the agro-food industry.
7.2 THE DIFFUSION OF KNOWLEDGE ON A LOCAL LEVEL: THE CONTRIBUTION OF CONTEMPORARY LITERATURE Even if the concept of proximity is not always directly referred to, contemporary literature regarding the transmission of knowledge in a spatial framework continually alludes to this notion. Although everyone agrees on the importance of geographical proximity in the process of the diffusion of knowledge, the main aim of the systemic type of analyses is to highlight the real or assumed qualities of groupings of technology firms, and to define the LIPS (Localized Innovation Productive Systems) that hold the promise of a potential local technological development. A more econometric approach, on the other hand, aims to investigate the role of proximity in the process of the transmission of knowledge, based on the modeling of geographical externalities related to innovation and technology. In the 1980s and 1990s, the relation between geography and technology became the focus of attention in regard to the institutional framework of the production of innovations, as differences had appeared between countries or regions that had a comparable level of development but were characterized by unequal innovation rhythms (Lung et al., 1999). Since then, research has been dedicated to various subjects such as innovative milieux (Ratti et al., 1997; Crevoisier, 2001), technological districts (Antonelli, 1986), technopoles or science parks (Monck et al., 1988; Longhi, 1999) and, in general, to localized systems of production and innovation (Lundvall, 1992; Maskell and Malmberg, 1999), so as to highlight the complex connection between spatial concentration and technological advantage, and then to reveal the organisational component underlying this type of local operation. These different approaches have two common characteristics: they postulate the effectiveness of local operations and highlight the importance of the organisational component. So it is, that studies concerning innovative milieux have underscored the importance of connections between the different local actors as
GEOGRAPHICAL PROXIMITY AND THE DIFFUSION OF KNOWLEDGE
129
regards the technological development of a given region or geographical area, particularly when they have technology supplier-user type relations that can help to reduce technology leakages and promote the implementation and development of local learning opportunities. Research concerning technopoles, which is often of a less theoretical nature, systematically attempts to highlight the advantages of grouping local high- tech firms on the same territory, especially in regard to the production of innovations, not only because of the concentration of potential for research or innovation, but also because of the synergetic effects arising from the collaboration between local firms. Most of these characteristics can be found in the analyses of regional innovation systems, that include the setting-up of a local network based on technological complementarities, as well as an institutional dimension illustrated by implementation policies of public authorities in terms of support to innovation or the training of engineers or scientists, and where the relation between science and industry occupies a central position. All in all, and as the latest syntheses on innovation clusters have shown (Porter 2000), the idea that firms and productive systems benefit from the spatial concentration of their research and innovation activities, is widely accepted nowadays: permanent geographical proximity is seen as an essential condition for technological success, particularly in the case of small and medium-sized enterprises (SMEs). Nevertheless, serious doubts are now being voiced, as to the characteristics themselves, or even the merits, of the process of spatial concentration that has been engaged in, particularly concerning the ability to transfer knowledge that is often termed as tacit knowledge, without cost and without any particular effort (Rallet and Torre, 2000). Whereas this research takes it for granted that permanent geographical proximity plays a part in the process of innovation and the transmission of knowledge, studies on geographical externalities attempt to verify the role of this proximity in the transmission of knowledge by calculating the maximum distance that a technological externality could cover. One of the characteristics of innovation is to produce externalities. Due to the peculiar nature of this activity, that is sometimes compared to the production of a (semi) public good, the results cannot be totally appropriated by the innovator, as part of the knowledge is diffused into the economy without the innovator being able to prevent it, or even being aware of it.1 When innovation (or R&D) is likened to information, there is an unlimited leakage of results that concerns the overall economy, but the approach in terms of knowledge leads us to analyze the possibility of diffusing this knowledge, as well as the geographical area it covers. From an empirical point of view, the fact that there is a high concentration of innovative activities contradicts the hypothesis of a complete diffusion of R&D results, which would allow activities to be equally distributed throughout the territory. The over-concentration
130
Delphine Gallaud and André Torre
of innovative activities, which is even greater than the production activities (Audretsch and Feldman, 1996), is then often accounted for by the characteristics of the externalities that are assumed to have a limited geographical extension. Autant-Bernard and Massard (1999) have compiled four types of studies dedicated to calculating the externalities of knowledge (or spillovers) and to their spatial area, respectively based on: The use of patents as markers of externalities (Jaffé et al., 1993). The geographical concentration of innovations (Feldman, 1999; Audretsch and Feldman, 1996). Geographical coincidence (Jaffé, 1986; Anselin et al., 1997). Local interaction (Anselin et al., 1997; Wallsten, 2001), to which one may add (Feldman, 1999). Knowledge incorporated in capital or investment goods. All these approaches conclude that externalities exist and that their geographical extension is limited; this explains the concentration of firms in certain areas and supports the idea that geographical proximity is an important factor in the diffusion of knowledge. However, two issues limit the significance of this research. First, how to measure the geographical extension of externalities is still much debated. Second, the role of proximity appears to be dependent on how the channels of transmission of externalities are analyzed. (Autant-Bernard and Massard, 1999).
7.2.1 The measurement of geographical extension is still much debated Some of the above-quoted studies do not really offer an independent estimation of spatial externalities: the authors use a predefined geographical area, which pre-supposes, but does not prove, the existence of externalities. Thus, the first three methods (patents, concentration, and coincidence) do not offer a true measurement of externalities. There is no calculation of the elasticity of R&D expenditure in relation to the innovation capacity of the company of reference and even less of the distance externalities are supposed to cover. Assuming that externalities exist, such studies model their effects and, in fact, they measure agglomeration phenomena. These methods generally postulate the role of local dimensions by using predefined geographical areas: States (Jaffé, 1989; Feldman, 1999), metropolitan areas (Jaffé et al., 1993) and Counties (Anselin et al., 1997 in their first evaluation). Notions of distance, when they are introduced into the gravity and coverage indicators used by these authors, are pre-defined. For instance, according to Anselin and al.
GEOGRAPHICAL PROXIMITY AND THE DIFFUSION OF KNOWLEDGE
131
(second measurement), R&D may have been carried out within a radius of 50 or 75 miles around the County of reference. Other recent studies make use of Geographical Information Systems (GIS) to model the distance range of technology spillovers. Thus, Wallsten (2001) uses GIS to analyze the probability of a firm benefiting from government support for innovation that is given to an adjacent firm. It locates firms without using a predefined geographical zone and shows that firms receiving financial support are situated close to each other, in a radius of one tenth of a mile, often on the periphery of urban areas. Even if these are strategic externalities linked to information rather than R&D, and although participating in a government program is liable to introduce a different angle, one sees that the distance effect, if it is not predefined, still varies noticeably from one author to another (from one tenth of a mile to 50 miles). This leaves room for many interpretations. After all, it was not until the publication of Orlando’s work (2000) that these methods allowed the simultaneous calculation of externalities and of distance.
7.2.2 The channels of transmission of externalities Zucker et al. (1994) are the first to highlight the role of effective interaction in the diffusion of knowledge, by showing that geographical proximity is not sufficient to enable benefits from the externalities of technology. A similar finding is put forward by Cockburn and Henderson (1998), who consider that externalities can only be received if firms stay in contact with scientists (especially by co-authoring articles). Audretsch and Stephan (1996) continue in line with these studies, showing that 70 percent of the contacts that are maintained between SMEs and scientists are not local but vary according to the main tasks assigned to the scientists (which may include the transfer of knowledge, offering a quality signal to investors, or participation in the scientific committee of a SME). These results should be considered with caution, as it is only the transfer of knowledge that is concerned in the analysis of innovation networks. Zucker et al. (1998) measure the relationships between the productive scientists and innovation output of firms. Productive scientists (or “starscientists”) are defined by the number of articles published. The innovation output of firms is assessed by three indicators: the number of products being developed, the number of products on the market, and the net growth of employment, that all express the stages in the process of innovation, from invention to economic performance (Autant-Bernard, 1999). Zucker et al. conclude that only the influence of researchers linked to firms by research cooperation has a significant impact on innovation performance. It is therefore not enough
132
Delphine Gallaud and André Torre
for firms to be situated near universities – they must also effectively cooperate with the local scientists to capture the externalities and express them in terms of an increase in innovative results. Thus, the notion of interaction shows that the effectiveness of geographical proximity is limited, as it involves a certain organization of the said proximity. Thus, the analysis of the process of transmission of knowledge externalities leads us to reconsider an automatically positive role for geographical proximity, and to consider instead the role of an effective cooperative type of interaction between SMEs and scientists. An organizational constraint is added to the geographical factor, as the idea that firms could enjoy the results of R&D without costs is replaced by the organization of its dissemination. Moreover, the definition of these externalities is problematic. If one considers that knowledge is in the very air, then externalities can be apprehended without cost and geographical proximity is enough. On the other hand, knowledge becomes much less easily transferable when it is built into the human capital. Geographical proximity alone is then an insufficient guarantee, and the (expensive) diffusion of knowledge will have to be organized. This is why it is necessary to investigate the actual interaction between agents, particularly concerning the external acquisition of technology, even if this requires certain changes in the prior definition of externalities and their content, in order to get into the black box of technology interaction.
7.3 GEOGRAPHICAL PROXIMITY AND ABSORPTIVE CAPACITY Any company that wishes to benefit from external technologies must rely on its relations with agents who have that knowledge, whether they are markets for knowledge, or more direct interactions. However, if this operation is to be fruitful, the firm must have sufficient internal ability to assimilate or reproduce this imported knowledge. Studies concerning the absorptive capacity of companies analyze how a receiving firm may pick up a greater or lesser quantity of technology spillovers, the internal organization of the said company enabling it to minimize its R&D investments in order to absorb available knowledge from its environment. Cohen and Levinthal (1989) show that, besides producing new information that can be used as input for innovation, R&D also facilitates the assimilation of external knowledge. Thus, the absorptive capacity of a firm corresponds to the amount of external knowledge that it is able to use. This is a multiplying factor for sums invested in intra industrial R&D, and a stock of extra-industrial knowledge (mostly made up of university research), that does not take into account geo-
GEOGRAPHICAL PROXIMITY AND THE DIFFUSION OF KNOWLEDGE
133
graphical distance; only the technological distance between the transmitter and the receiver of the externality is considered, This work can be continued through research on the connection between the internal characteristics of the firm and the type of external knowledge that is absorbed. Mangematin and Nesta (1999) have shown that the possibility of absorbing external knowledge gives rise to three series of specific problems: The distinction between applied and fundamental knowledge, that leads to the question of what type of knowledge companies import, and highlights two aspects of the absorptive capacity, respectively, research activities and development activities. This scientific skill enables us to discern the advance of knowledge leading to technological opportunities, and to reduce the uncertainty regarding the value of cooperation projects and of a technological component that would contribute to the development of innovation and to its introduction on the market (Arora and Gambardella, 1994). Only a company that has a high absorptive capacity is in a position to absorb both types of knowledge. The distinction between tacit and codified knowledge. Codified knowledge can be expressed through the use of a code and “embedded” in a medium that is not dependent on the person who possesses knowledge, whereas tacit knowledge cannot be expressed in a code or used by a third party independently of the possessor of that knowledge. Even if we distinguish knowledge that is predominantly tacit or predominantly codified, both forms of knowledge are complementary (Nonaka, 1994), which is why firms invest in channels for the diffusion of knowledge that are more expensive than simply reading scientific and/or technical articles. Thus, a high absorptive capacity, due to the diversification of the means of access to knowledge, allows one to absorb codified and tacit knowledge through a greater number of different channels. On the other hand, a reduced absorptive capacity limits opportunities for cooperation. The processes of recontextualization. Knowledge is often generated in a particular context, specific to the company, laboratory or productive system that it originated from (and which aims to prevent this knowledge from leaking out). This specificity gives rise to some difficulty in the appropriation of knowledge, as the lack of certain elements of information that would help to clarify it, makes it difficult to use in a different context. The absorptive capacity of technology users must therefore resort to a process of recontextualization of technology spillovers (Guilhon, 2000), that is expensive and requires once again, specific absorption skills.
134
Delphine Gallaud and André Torre
In view of these considerations, the potential complementary between internal and external R&D (without internal R&D, it is impossible to use knowledge that one has not generated by oneself) brings us to the question of the role of external cooperation, an aspect that Cohen and Levinthal have not studied in any detail. However, although they have not explicitly broached the subject of space, their thesis is in opposition to the arguments promoting localized production and innovation systems (technopoles in particular). Whereas the LIPS approach assumes that geographical proximity can compensate for the shortcomings in the internal organization of firms, particularly SMEs, the analysis in terms of absorptive capacity, recommends that internal organizational abilities within the firm be used for the assimilation of external knowledge, even if the latter originates from close neighboring firms. Thus, studies on the diffusion of knowledge at a local level (whether it is the geographical knowledge externalities or the LIPS) typically consider the geographical proximity of the actors as the most important or a relatively very important factor. However, research on the absorptive capacity of firms gives greater precedence to the internal organizational capabilities of firms, or to the organizational proximity, as referred in the economics of proximity (Gilly and Torre, 2000)2 , and above all it makes it possible to raise the question concerning the stages of the innovation process in which geographical proximity is necessary.
7.4 THE SPATIAL DIMENSION OF THE EXTERNAL ACQUISITION OF TECHNOLOGY Although they explicitly refer to an external acquisition of knowledge, Cohen and Levinthal’s works hardly discuss the way in which this acquisition process takes place, and in particular the types of interaction between firms. However, these interactions are extremely important, as no firm possesses internally all the knowledge it needs for its production process. A firm that wishes to acquire external knowledge can get information made public through conferences, trade fares, publications, symposia, exhibitions... but most knowledge it wishes to acquire is private (or semi public) and can only be acquired from other firms or organizations. These acquisitions range from commercial transactions (the markets of technology) to research cooperation. The latter can be more or less formalized, whether it concerns the relations with public research organizations (contracts between universities and industries) or with other enterprises (vertical cooperation, that corresponds to the relations with clients or suppliers, and horizontal cooperation with the competitors (a rare form which concerns less than 10 per-
GEOGRAPHICAL PROXIMITY AND THE DIFFUSION OF KNOWLEDGE
135
cent of R&D agreements)). In cases where knowledge is public, geographical proximity has no impact because knowledge can be acquired wherever the innovating firm is located in relation to the productive source of knowledge. Things are different when the information is not divulged; it can be beneficial for the firm that seeks to acquire it to be located in the proximity of the productive organization, especially if it does not have a monopoly on this information. The different types of interactions made in order to obtain external knowledge can be classified according to their relation to space and to the need for geographical proximity: Informal relations for the exchange of technological knowledge: these require a permanent and high level of geographical proximity, because of the tacit and contextualized nature of the knowledge in question, and the need to continually validate and enrich this knowledge. Research collaboration: the joint accumulation of knowledge is facilitated by a relation of geographical proximity. Exchange of researchers: geographical proximity is required in the case of an exchange of tacit knowledge, but can be obtained by temporarily importing a skill. Purchase of patents and licenses: it is useful to distinguish between license agreements that do not provide any technical assistance and those that do. The former represent a spot transaction on the knowledge market: geographical proximity does not exist. Where the license agreement does make provision for technical assistance, the patentee must organize the transmission of technical know-how defined within the field of the patent, which implies the transmission of both the codified and tacit knowledge needed for the industrial utilization of the invention. Geographical proximity facilitates the transmission of this knowledge, and also lowers the costs involved, particularly when the agreement provides that the patentee shall personally undertake the start-up of production and training sessions for the licensee’s staff (Gaudin, 1993). Industry-university research contracts: when the content of this research is tacit, it requires geographical proximity, particularly in the urban areas or within specific geographical zones (Grossetti and Nguyen, 2001). However, as soon as the results are acquired and presented in a codified form, the need for geographical proximity disappears. Vertical or horizontal cooperation: although geographical proximity is frequent in client-supplier relations, its role in horizontal coopera-
136
Delphine Gallaud and André Torre
tion is not that obvious. Here too, temporality is significant, as there are temporary moments where proximity is necessary. In the case of relations involving competing companies, it is advisable to try and obtain organizational proximity at a distance in order to retain access to local markets. The process of acquisition of external competencies in its spatial dimension may be described as follows. When a technical object or a technology is introduced, it is essential to implement a decoding and/or recontextualization phase (or even a decontextualisation phase in the case of an imitation procedure) with the help of the absorptive capacity of the firm. The purchase of external technologies on the markets for knowledge is a point in case. However, this also holds true for other types of interactions, and particularly for relations of a cooperative nature. Meetings between people must take place, in order to decode or to decontextualize knowledge and to establish protocols for the distribution of tasks between the participants in the operation, whether they are people inside the firm or outside contributors or newly arrived on the scene. Cooperative relationships often require geographical proximity, even if it is temporary, because of the frequency and the intensity of the interactions that we have just noted. The following phase of internal recodification or recontextualization requires a lesser degree of geographical proximity in the relationships within the network, as it involves a considerable amount of exchanges of codified knowledge within the company and an integration of the absorptive capacity of the latter. The type of learning referred to is mostly internal, and the starting-up of the production process relies extensively on codified knowledge, which becomes the major element in the process, as external support and contributions can be realized by means of long distance communication. It is possible to specify the needs for geographical proximity in the particular case of cooperation that was studied by Léveque et al. (1996) who analyzed the link between the forms of internal organization of R&D and the types of external cooperation. This study proposes three main archetypes that, in the case of firms, correspond to an equal number of R&D forms3. They are respectively: The exploration of ways to implement an enduring technical transformation, by seeking to acquire new knowledge and finding out how to use it in the innovative process. Exploration R&D, which allows new knowledge to be produced and internalized, generates cooperation that in turn increases learning opportunities. The high level of uncertainty attendant on this activity is matched by the large number of agreements entered into with partners (especially scientists) who are in a position to generate new knowledge.
GEOGRAPHICAL PROXIMITY AND THE DIFFUSION OF KNOWLEDGE
137
The exploitation of a relatively well-known innovation technique. R&D exploitation is the expression of a routine way to produce innovation. Cooperation, that includes a higher proportion of agreements between clients and suppliers than exploration R&D, occurs when both parties seek to develop identical skills but the products and the markets are complementary. The imitation of innovative processes initiated by competing firms. Imitation R&D, that requires rapidity and flexibility, is generally completely internalized. The number of external relations drops with the shift from exploration (that can be distributed between several sites) to exploitation (characterized by the use of relatively commonplace sources of external technology) and then to imitation (a situation in which external models are reproduced, often because the firm does not have the internal skills needed to produce innovation). This pattern, which enables us to connect the internal organization of R&D (and hence its absorptive capacity) to a certain type of external relations, applies for the most part to differentiated oligopolies, in other words, to relatively big firms. It has to be amended by taking into consideration the size of the firms. SMEs for instance, trapped between their innovation rationale and the insufficiency of their own skills in regard to the creation of technology, are not in a position to so easily develop certain modalities of internal R&D. Based on these considerations, and taking into account the fact that organizational proximity is substantially necessary for the overall process of acquiring external knowledge, the relation between the different types of R&D and geographical proximity can be defined (see Table 7.1). Here, geographical proximity is defined as the opportunity for frequent face-to-face meetings between agents, according to Lung et al., 1997 and Rallet and Torre, 2000.
In this instance, the term “permanent or temporary geographical proximity” applies to the process, and not be confused with the location of firms, which differs according to their size. It means that there exists a need for geographical proximity that can be met through a temporary coming together of the actors of the cooperation process.
138
Delphine Gallaud and André Torre
Put very simply, big firms, subsidiary groups or universities, can avoid the constraint of a permanent location in the initial phase of exploration by using devices such as sending out research teams or doctors, a solution that is allowed because they have such a large pool of human resources to draw from. In smaller firms, on the other hand, the coincidence of the need for exploration R&D with the need for permanent geographical proximity during this process is often a factor that determines the locality of the firm, as tasks related to the different phases of the R&D process are assigned to the same person. Therefore, these firms have to be located close to other firms or laboratories, even if geographical proximity is only needed during one of the phases of the R&D process. Again, geographical proximity and the location of firms should not be confused, as the existence of permanent geographical proximity during the exploration R&D process does not require the company to be located nearby, except in the case of smaller companies, where permanent geographical proximity does entail this kind of constraint in terms of their location. Finally, it should be noted that the role of geographical proximity, of greater or lesser import according to the phases and types of innovation project involved, decreases over time. Geographical proximity can be either: Permanent and complementary to organizational proximity during the phase of the joint production of fundamental, tacit and contextual knowledge. Temporary and complementary to organizational proximity during the phase of the absorption of knowledge developed during the scientific phase, which means that this knowledge must be recontextualized in order to test it in various situations. Totally dominated by organizational proximity, as in the case of coordination at a distance, which does not require face-to-face interactions. This is often the case during the development phases of prototypes and clinical trials, or when the findings of research are being codified. Thus, the thesis according to which the need for geographical proximity in the process of external acquisition of knowledge leads to a co-localization of firms or organizations in one same area proves erroneous. Indeed, the great number of modes of external acquisition allows firms to obtain techniques or knowledge even if they are not localized in the proximity of the enterprise or laboratory that has produced them. The more commercial interactions are, the less necessary becomes geographical proximity. On the other hand, the closer one gets to cooperation, the more important it becomes. However, the need for geographical proximity can be met in a temporary manner through the
GEOGRAPHICAL PROXIMITY AND THE DIFFUSION OF KNOWLEDGE
139
mobility of qualified staff, in particular by organizations that possess the necessary resources to carry out this type of operations.
7.5 EXTERNAL ACQUISITION OF KNOWLEDGE AND PROXIMITY IN THE FIELD OF BIOTECHNOLOGY The agricultural food-processing industry (AFI) has undergone major changes over the last thirty years, and in particular, there has been a complete reversal of France’s position, which has shifted from being an importer to being an exporter. The AFI now contributes overwhelmingly to the positive trade balance, following the modernization and extension of farms, the rapid increase in the output, the intensification of breeding, etc... these factors generate productivity gains that are reflected in profound institutional modifications. Today, thanks to the boost in France’s international trade, it has become the world’s second largest agricultural exporter after the USA, with particularly strong performance in food grains, sugar, dairy products, wine, and beverages. However, although the growth in these activities has been extremely significant, the AFI still comes last in the field of research. In France, due to the lack of researchers and of funds available for R&D, relatively little research has been carried out in the private sector (1 percent of the GDP); moreover, the level of research in the private sector differs considerably according to the type of company concerned. Thus, big groups in the agricultural food-processing industry enjoy many R&D activities (often organized worldwide), and this is the case for about 300 French SMEs. Most of the small companies, on the other hand, do not have formal access to R&D, and, in the best of cases, they make use of transfer centers or intermediation institutions. Public research plays an important part in this context, and the INRA in particular, has a pivotal role, as contracts passed between firms and public laboratories are often presented as a substitute for the lack of internal R&D. Several explanations can be given in order to resolve this apparent paradox: The dynamics of this sector is not based on innovations, or only incremental ones. This situation arises because of knowledge and know-how that are already present, but have not been fully exploited and that, as soon as they are used, become engines of productivity. Agronomic innovations are not linked to research in this sector ; this explanation substantiates the hypothesis of a diffusion arising from other industries. Joly and Lemarié (2000) note that Pavitt (1984)
140
Delphine Gallaud and André Torre
classifies the AFI in the category of industries where innovation is almost entirely due to technical progress that has been built into the capital goods purchased by them, and very little to their own research. The social return on research is particularly high in the fields of agriculture and the food trade. Huffman and Evenson (1993) show that the impact of research is particularly significant in the AFI, more for agriculture than for breeding and more for scientific research than for applied research. This result justifies the many studies carried out in the seed sector, because genetic engineering and improvements represent an important area of innovation and contribute massively to the increased yield of the large-scale farming industry. Biotechnologies are defined as a number of techniques and knowledge related to the use of the living in the industrial processes of production (Ducos and Joly, 1988). This is not enough to define a sector (Porter, 1990) even though certain firms have specialized in these activities. These techniques are used in the pharmaceutical industry, agriculture, the agro-food industry, and the environment (they are still in the experimental stage in the latter case). In France, an economic environment is emerging that constitutes a production sector composed of biotechnology firms and firms offering complementary activities: manufacturing of specific instruments and equipment, consulting and technical expertise (Lhuillery, 2002). Chemical firms remain the main suppliers of innovation in the agro-food industry. It is the case, in particular, of the production of flavorings - additives produced by a few firms in the agro-food industry, but essentially manufactured by the chemical industry. There are currently about one hundred flavorings obtained thanks to biotechnologies. Thus, this activity is still developing. The market is not much internationalized and SMEs as well as groups are present on the market. Whereas pharmaceutical and chemical firms produce techniques related to biotechnologies, agricultural and agrofood firms are essentially users. The market is internationalized in the chemical and pharmaceutical industries (there is world oligopoly), but a few SMEs have specialized in specific products. However, in the long-run, it remains to be seen if they will be able to maintain these activities. These firms have to innovate constantly in order to create markets on which they obtain a temporary monopoly. Lemarie et al. (2001) have highlighted the specific characteristics of these companies as far as France is concerned: Very few of them operate on final consumer markets; they are mostly suppliers for other enterprises, whether in the field of health or of the AFI. Their main business is to design, develop, and manufacture cus-
GEOGRAPHICAL PROXIMITY AND THE DIFFUSION OF KNOWLEDGE 141
tom-made genetic or biological material. The main business of SMEs of the AFI (a third of the sample) is the perfecting of diagnostic kits in order to determine, for instance, if GMO are present- most of them are situated in the Ile de France, a region that specializes in pharmacology, as well as in five other regions: Alsace, Auvergne, Aquitaine, Brittany (regions that specialize in the AFI) and Rhône-Alpes (a region that specializes in pharmacology). Their development is related to different models of cooperation networks. Companies with a fast growth rate (a minority, mostly involved in pharmacology) have a dense network of cooperation with universities and public research organizations. Geographical proximity has very little relevance for these firms, that are used to recruiting skills through their international scientific network and who therefore favor organizational proximity. Finally, firms with a parent company include mainly SMEs connected to agricultural markets, that have been created by groups with a view to flexibility and decreased risks, and that are situated near the customers. Catherine et al. (2002) also highlight the recourse to local networks, by connecting their development to the founders’ profile. Companies are usually established in the area where its founders had already worked previously, particularly in the case of scientists, which would seem to indicate that this human capital is more specific and less transferable than the capital of the managers. Thus, successful SMEs combine local skills (particularly in the scientific field) and global skills (mostly in the managerial field). Thus, we can distinguish (Catherine et al., 2002) two main types of companies: Firms with a low level of technological complexity and small exploitation and imitation R&D budgets, which lead to incremental innovations, and rely on internal skills that stem from the core business of the firm. These SMEs, that make up 2/3 of the total number of SMEs and whose outlets are mainly in the AFI and agriculture sectors, are often suppliers for big groups and operate on markets with a narrow niche. They have an essentially local network, and rely on funding within a demarcated geographical area. They can either be classified as service providers or as manufacturers of generic products as the case may be, and these activities enable them to rapidly become operational and to make a considerable profit. Firms that have a higher level of technological complexity and a big volume of exploratory R&D, whether they are leaders on their market and make radical innovations, or research providers. These companies, with outlets mainly in the pharmacology sector, take on the
142
Delphine Gallaud and André Torre
risks and uncertainty of programs that big groups consider too risky, and they look for wider markets. They can combine purchases of licenses that imply a very limited geographical proximity, with means such as co-developments in partnership that necessitate greater geographical proximity. Cooperation between SMEs linked with agricultural outlets that tend to practice exploitation R&D is often a customer-supplier type relationship (Léveque et al., 1996) that requires ties of geographical proximity, (Cf. III, below), particularly in the initial phases of research and development. Nevertheless, the modalities of cooperation in R&D are liable to change over time due to the influence of the dynamics of competition. The initial phases of the innovation process require extensive application to external collaboration, or even joint contracts in order to carry out the exploratory R&D, the number and variety of contracts decreasing at a later stage, and resembling suppliercustomer types of cooperation (Léveque et al., 1996). The theory suggesting that networks are more extensive and diversified in pharmacology than for SMEs connected to agriculture and the AFI has to be qualified, because they all generally go through a phase of exploratory R&D when they enter the market. The information (very scarce) about SMEs in this sector, shows that big integrated networks (Depret and Hamdouch, 2000), that include different organizations and are developed by big groups from the pharmaceutical industry, often create a plant biotechnology sector outside of their core business. This is the case for Aventis, for example. It is important not to confuse these networks with the networks formed by SMEs, that have specificities in terms of their objectives and/or constraints (survival through access to the network, for instance) and are more motivated to create an area of expertise (specific skills) for themselves (and thus to develop a strategy of niches). The strategy of the groups leads one to believe that a certain number of techniques, initially perfected in pharmacology, can then be transferred to the AFI. However, the gap between the results of research and the market, is more significant in the case of plant biotechnology than for biotechnology in the pharmacology sector, which is almost continually linked to the advance in fundamental research. Whereas pharmacology is marked by a series of discoveries converted as quickly as possible into innovations, plant biotechnology is often characterized by a lesser volume of scientific discoveries and a slower renewal of skills. Moreover, as these techniques are generally relatively older, the life cycle of the industry can more frequently correspond to the phase of exploitation R&D, where the agreements are less in number, involving extensive sub-contractual relations, the aim of which is to reduce the cost of developing innovations. Consequently, the recourse to scientific
GEOGRAPHICAL PROXIMITY AND THE DIFFUSION OF KNOWLEDGE
143
cooperation could be more selective and be used as a means to complete the data base of the SME on certain specific issues, in order to carry out a particular project instead of maintaining a diversified absorptive capacity. The interactions mobilized within this framework would then only require a temporary geographical proximity, that has however, proved fatal to SMEs in terms of locality.
7.6 CONCLUSION What is the role of geographical proximity in setting-up and operating of external acquisition of technology? Is it fundamental, as public policies that favor the interaction between science and industry at the local level in the hope of obtaining a better circulation of knowledge, seem to imply? This chapter argues that a body of literature (local systems of production and externalities) considers permanent geographical proximity as a necessary condition for the diffusion of knowledge. However, the literature dealing with transmission channels for externalities shows that geographical proximity only influences the innovative performance of firms if there is effective interaction between the agents. In order to benefit from externalities, firms have to make costly efforts of organization. Our analysis is based on the hypothesis that organization is the first modality in the transmission of knowledge, and that geographical proximity can be temporary, particularly in the initial phases of the R&D processes. The smaller firms are then more acutely aware to fulfill the need of geographical proximity. This pattern, applied to plant biotechnology, reveals that SMEs related to the AFI and to agriculture are part of a less diversified and more localized innovation network than pharmaceutical SMEs and are more involved in frequent and repeated contacts with the clients and the suppliers. The localized nature of their technological interactions can be explained by different variables, but not by a need for permanent geographical proximity in the process of knowledge circulation. It stems more from the lack of human abilities and financial means, that forces them to choose a close to companies and laboratories, which are a source of knowledge, in the hope of benefiting from the transmission of knowledge that, however, only concerns short term projects.
NOTES 1.
We only take in account the common sense of the definition of the externalities, often presented in the literature on geography of innovation.
144
Delphine Gallaud and André Torre
2.
We define organizational proximity as the membership to the same group of economic entities, linked with interactive processes.
3.
In reality, few businesses use only one of these methods. Rather, businesses may use different methods over time and according to the types of competition to which they are confronted.
REFERENCES Anselin, L., Varga, A., Acs, Z., Local geographic spillovers between University research and high technology innovations. Journal of Urban Economics 1997; 42:422-448. Antonelli, C., Technological Districts and Regional Innovation Capacity. Revue d’Economie Régionale and Urbaine 1986; 5:695-705. Arora, A., Gambardella, A., Evaluating technological information and utilizing it. Scientific knowledge, technological capability and external linkages in biotechnology. Journal of Economic Behaviour and Organization 1994; 24:91-114. Audretsch, D., Stephan, P., Company scientist locational links: the case of biotechnology. American Economic Review 1996; 86:641-651. Audretsch, D., Feldman, M., R&D spillovers and the geography of innovation and production. American Economic Review 1996; 86:630-640. Autant-Bernard, C., Massard, N., Econométrie des externalités technologiques locales et géographie de l’innovation: une analyse critique. Economie Appliquée 1999; 52:35-68. Catherine, D., Corolleur, F., Coronini, R., Les fondateurs des nouvelles entreprises de biotechnologies et leurs modèles d’entreprise. Une approche par les compétences and les ressources illustrée sur le cas français. Revue Internationale des PME 2002; 15:63-92. Cockburn, I., Henderson, R., Asorptive capacity, coauthorship behaviour and the organization of research in drug discovery. The Journal of Industrial Economics 1998; 46:157-182. Cohen, W., Levinthal, W., Innovation and learning: the two faces of R&D. The Economic Journal 1989; 99:569-596. Crevoisier, O., L’approche par les milieux innovateurs: état des lieux et perspectives. Revue d’Economie Régionale and Urbaine 2001; 1:153-165. Depret, M., Hamdouch, A., Pharmacie et biotech l’ère des réseaux. Biofutur 2000; 203:44-48. Ducos, C., Joly, P.B., Les biotechnologies, Paris: La Découverte, Repères, 1988. Feldman, M., The new economics of innovation, spillovers and agglomeration: a review of empirical studies. Economics of Innovation and New Technology 1999; 8:5-25. Gaudin, J.H., Guide pratique de l’ingénierie des licences et des coopérations industrielles. Paris: LITEC, 1993.
GEOGRAPHICAL PROXIMITY AND THE DIFFUSION OF KNOWLEDGE
145
Gilly, J.P., Torre, A., Dynamiques de proximité, Paris: L’harmattan, 2000. Grossetti, G., Nguyen, D., La structure spatiale des relations science-industrie en France: l’exemple des contrats entre les entreprises and le laboratoires du CNRS. Revue d’Economie Régionale and Urbaine 2001; 2:311-326. Guilhon, B. et al., Technology and markets for knowledge. Boston: Kluwer Academic Publishers, 2000. Huffman, W.E., Evenson, R., Science for Agriculture: a long term perspective, Ames: Iowa State University Press, 1993. Joly, P.B., Lemarié, S., Cinquante ans d’innovation en Agriculture. Economie Rurale 2000; 86-97. Jaffé, A., Technological opportunity and spillovers of R&D: evidence from firms patents, profits and market value. The American Economic Review 1986; 76:984-1001. Jaffé, A., Real effects of academic research. The American Economic Review 1989; 79:957970. Jaffé, A., Trajtenberg M., Henderson R., Geographic localization of knowledge spillovers as evidenced by patents citations. The Quarterly Journal of Economics 1993; 108:577-598. Lemarie, S., Mangematin, V., Torre A., Is the Creation and Development of Biotech SMEs Localized? Conclusions drawn from the French Case. Small Business Economics 2001; 17:6176. Léveque, F., Bonazzi, C., Quental, C., “Dynamics of cooperation and industrial R&D: first insights into the black box 2.” In Technological cooperation, Coombs R., ed. Cheltenham: Edward Elgar, 1996. Lhuillery, S., Panorama des entreprises françaises de biotech, séminaire REPERES, MENRT, 2002. Longhi, Ch., Networks, collective learning and technology development in innovative high technology regions: the case of Sophia-Antipolis. Regional Studies 1999; 33:333-342. Lundvall, B.A. “Relations entre utilisateurs et producteurs, systèmes nationaux d’innovation et internationalisation.” In Technologie et Richesse des Nations, Foray, D. and Freeman, Ch. eds. Paris: Economica, 1992. Lung, Y. et al., Organisation spatiale et coordination des activités d’innovation des entreprises. Rapport pour le Commissariat au Plan. 1997. Lung, Y., Rallet, A., Torre, A., Connaissances et Proximité Géographique dans les processus d’innovation. Géographie, Economie, Société 1999; 1:281-306. Mangematin, V., Nesta, L., What kind of knowledge can a firm absorb? International Journal of Technology Management 1999; 18:149-172. Maskell, P., Malmberg, A., Localised learning and industrial competitiveness. Cambridge Journal of Economics 1999; 23:167-185.
146
Delphine Gallaud and André Torre
Monck, C.S., Porter, S.P., Quintas, P., Storey, D.J., Science Parks and the Growth of High Technology Firms. London: Croom Helm, 1988. Nonaka, I., A dynamic theory of organizational knowledge creation. Organization Science 1994; 5:14-37. Orlando, M., On the importance of geographic and technological proximity for R&D spillovers: an empirical investigation. Kansas City: WP Federal Reserve Bank of Kansas, 2000. Pavitt, K., Sectoral patterns of technical change: towards a taxonomy and a theory. Research Policy 1984; 13:343-373. Porter, M., The competitive advantage of nations. London: Macmillan, 1990. Porter, M., “Locations, clusters and company strategy” In The Oxford handbook of economic geography, Clark and al. eds., Oxford: Oxford U. Press, 2000. Rallet, A., Torre, A., Is Geographical Proximity necessary in the Innovation Networks in the Era of Global Economy? GeoJournal 2000; 373-380. Ratti, R., Bramanti, A., Gordon, R., The Dynamics of Innovative Regions. Aldershot: Ashgate Publishing, 1997. Wallsten, S., An empirical test of geographic knowledge spillovers using geographic information systems and firm-level data. Regional Science and Urban Economics 2001; 31:571-599. Zucker, L., Darby, M., Armstrong, J., Intellectual capital and the firm: the technology of geographically localised knowledge spillovers, WP NBER 4946, 1994. Zucker, L., Darby, M., Brewer, M., Intellectual human capital and the birth of US biotechnology enterprises. The American Economic Review 1998; 88:290-305.
Chapter 8 CONTINUITIES, RUPTURES, AND RE-BUNDLING OF REGIONAL DEVELOPMENT PATHS: LEIPZIG’S METAMORPHOSIS
Harald Bathelt and Jeff Boggs
8.1 INTRODUCTION Leipzig has a long history as a site of cultural production and learning. Its book publishing industry, centered in the city’s Graphisches Viertel (Graphical Quarter), dominated the German book trade for well over two centuries, from the 18th century until the mid-20th century. Today, Leipzig is but a shadow of its former glory as the hub of German-language book publishing. However, where old media once reigned, during the 1990s a new media industry has grown phoenix-like from Leipzig’s proverbial ashes. To explain this sequence of events, we develop a conceptual framework that treats regional development as an evolutionary process, dependent on numerous local industries that vie for dominance, experience crises, and – through the ensuing struggle for stability – shape a region’s development path. This conceptual framework is built around interactive learning. Unlike some literature on evolutionary changes, we incorporate technological and political ruptures into our model of regional development paths and eschew technological determinism. From this, we detail the basic structure of a model in which bundles of overlapping technological trajectories drive regional development (for a more detailed account, see Bathelt and Boggs, 2003). We argue (in section 8.2) that regional development paths consist of multiple, partially overlapping, partially unrelated technological trajectories. Regional development paths have to be treated this way because individual firm competencies cannot be spatially aggregated into a comprehensive, homogeneous regional path. We then build (section 8.3) on the idea of windows of locational opportunity by incorporating theories of crisis from the regulationist approach. This allows us to conceptualize the impact of sectoral and technological as well as political crises on those institutions that enable interactive learning.
148
Harald Bathelt and Jeff Boggs
Such crises, in turn, create growth opportunities for previously marginal technological trajectories housed in a region. We conceptualize this process by introducing the concept of regional re-bundling. Regional re-bundling is the mechanism through which individual agents in a region shift their remaining capital (broadly defined) towards new or different uses in an attempt to eke out a living. At the aggregate level, regions recover from crisis as their actors continually re-bundle local assets at hand and combine them with new assets from outside until they find a configuration that generates growth. As a result of these crises, new growth opportunities arise for novel or previously marginal industries. In turn, these expanding industries can drive the region’s development path. In some cases, newly composed regional bundles are the precursors of new localized clusters of dynamic interrelated networks of economic activities (or “industry clusters” for short) driving regional development. In the empirical discussion contained in section 8.4, Leipzig is viewed through this conceptual lens. Finally, in section 8.5, we draw some conclusions regarding a re-conceptualization of regional development paths.
8.2 INTERACTIVE LEARNING, TECHNOLOGICAL TRAJECTORIES AND REGIONAL DEVELOPMENT This section synthesizes literature on interactive learning, technological trajectories, and windows of locational opportunity. We point out the strengths and faults of these concepts to lay the groundwork for a model that explains regional development as an evolutionary process. In developing the concept of regional development path, we argue that technological trajectories cannot easily be aggregated into a single, comprehensive regional trajectory. Instead, we employ a more complex understanding of regional development paths as bundles of overlapping technological trajectories.
8.2.1 Technological Change and Interactive Learning In contrast to traditional views on technological change, interactive learning assumes that innovation generally results from search activities and experiments in or related to the production process (e.g. Arrow, 1962; Rosenberg, 1982). These search activities are driven by reflexive social practices. They are also contingent; searchers do not know if they will succeed (Malecki, 1991).
CONTINUITIES, RUPTURES, AND RE-BUNDLING
149
Interactive Learning Interactive learning is crucial for innovation processes in complex leadingedge technologies (Lundvall, 1988; Gertler, 1993 and 1996). Such innovations depend on competencies distributed among a variety of different actors. This requires interactive problem-solving and collective action. Storper (1997) argues the goals of complex innovation processes are rarely predetermined. They become defined through actions while processes are underway. Due to reflexive behavior, agents constantly redefine goals according to new information about the success of previous actions. Interactive innovation processes rely on relations between particular people. These relations are mediated through conventions as expressed in accepted norms, rules, and practices. Even in less complex technologies, interactive learning processes and systematic feedback loops play an important role in technological progress. Spatial Implications of Interactive Learning Frequent interaction between specialists in firms promotes the generation of implicit, tacit knowledge that is not easily codified. This knowledge is localized in that it privileges those actors, firms, and places where this interaction takes place (Maskell and Malmberg, 1999). At the same time, explicit knowledge which is codified can also be quite “sticky” (Markusen, 1996). This is particularly the case when codified knowledge is being adjusted in interactions within a specific context or “community” (Wenger, 1998). Contextualized or embedded knowledge constantly enriched by localized knowledge is less prone to “ubiquitification” (Maskell and Malmberg, 1999). Therefore, tacit and even contextualized codified knowledge assets generate competitive advantages to regional actors who share this knowledge (Asheim, 1999; Gertler, 2001; Belussi and Pilotti, 2002). Relational Consequences of Interactive Learning As a result of interactive learning, innovation is a relational process. This has three consequences for regional production systems (Bathelt and Glückler, 2002 and 2003). First, the direction and particular results of innovation processes are context-dependent because they are bound to a particular set of agents and institutions. Firms must be understood as being embedded in their respective socio-economic context of supplier and customer relations (Grabher, 1993). Second, contextuality implies that change is path-dependent since it is based on past, often irreversible decisions regarding a particular technology (Arthur, 1988). Yesterday’s social experience shapes today’s interactive
150
Harald Bathelt and Jeff Boggs
learning. This, in turn, shapes future search processes and innovations. Third, at the same time, future developments are not easily predictable. The outcomes of interactive learning and the way in which this affects future interactions are fundamentally open and undetermined (Sayer, 2000); that is, innovation is a contingent social process.
8.2.2 Technological Trajectories: Privileging Continuity and Lock-in In a relational perspective, the direction of technological change is to some degree pre-structured by existing technologies, albeit not in a deterministic way (Dosi, 1982 and 1988). The history of developing new and adapting existing technologies defines a particular technological trajectory. Along this path, firms accumulate specialized knowledge about specific technologies, their uses and organizing their uses (Nelson, 1995). This knowledge is based on experience gathered from interactive learning and collective action. While the literature recognizes discontinuous technological change, many debates and empirical studies in this field focus on technological continuity and lock-in. However, the focus on continuity can blur the view of change in that it overlooks the potential of unexpected ruptures or discontinuities. This is especially obvious in Arthur’s (1988) model of market-share evolution for two competing technologies. Here, lock-in results from rational individual behavior, implying that path dependency is irreversible; that is, the users’ decisions necessarily lead to total market dominance of one technology over others. Yet, Arthur’s (1988) model of path dependency is problematic (Clark, Tracey and Lawton Smith, 2001) because it ignores that competing technologies can co-exist in the medium or long run (e.g. IBM’s and Apple’s competing operating systems). This quasi-deterministic view results from the assumption that actors are atomistic utility-maximizers. In reality, however, economic actors are embedded in structures of social relations, through which they communicate decisions for a particular technology and within which power asymmetries shape choices made (Bathelt and Glückler, 2002 and 2003).
CONTINUITIES, RUPTURES, AND RE-BUNDLING
151
8.2.3 Regional Development Paths as Bundles of Overlapping Technological Trajectories Regional trajectories are not simple projections of technological trajectories onto a territory. Because it is problematic to aggregate technological trajectories into a single regional trajectory, we conceptualize regional development paths as multiple, partially overlapping, and partially unrelated technological trajectories. A model of regional development paths must incorporate technologically unrelated regional actors. The regional actors might be spread across several different technological paradigms, none of which may have their territorial core within that particular region. Even in leading-edge hightechnology regions, such as Boston’s Route 128, the local economy is not a homogeneous set of industries tied into a single technological trajectory (Bathelt, 2001). However, this is not always noted in empirical work trying to identify the composition and direction of regional development trajectories, as seen in the work of Rigby and Essletzbichler (1997). Aggregating Technological Competencies of Firms into Regional Trajectories? Rigby and Essletzbichler (1997) use spatial aggregations of firms as proxies for regions and project the technological competencies of firms into a comprehensive regional development trajectory. They suggest that regions, just like firms (Cantwell and Fai, 1999), can be characterized by their technological and organizational structures. Thus, regional economic change is viewed as a consequence of variation and selection processes within that particular region. In this view, regions with superior capacities for interactive learning and extended knowledge pools grow faster than other regions. It is suggested that through this process particular regional innovation paths evolve. Regions are not Homogenized Actors We agree with this conceptualization on most accounts. However, this interpretation risks treating regions as actors. Furthermore, regions usually do not house just one technological trajectory. Therefore, it is problematic to assume that coherent firm behavior exists which simply translates into an equally coherent regional path as could be implied from Rigby and Essletzbichler (1997). Figure 8.1 schematizes some possible relationships between a region and one or more technological trajectories, which form the region’s development path. We provide this figure because we wish to geographically ground regional development trajectories, and doing so requires anchoring
152
Harald Bathelt and Jeff Boggs
Figure 8.1: Relationships between regions and technological trajectories.
technological trajectories in real places. Technological trajectories are usually described in aspatial terms, existing in some extra-dimensional space, not clearly anchored to any real location. Technological trajectories, however, do have geography and are based in particular places. This is due to the grounded actions of those agents engaged in interactive learning. It is through localized interactive learning, or the particular spatiality of innovation, that a given technological trajectory progresses. These locations are not necessarily restricted to a single region; they could easily be nodes within a network that spans continents. Nonetheless, in cases where the exchange of tacit knowledge occurs most readily with
CONTINUITIES, RUPTURES, AND RE-BUNDLING
153
close physical proximity of agents, we expect a significant part of this anchoring to attach to a particular location. Figure 8.1 contains three ideal cases for illustrative purposes. Each case takes place within a region of indeterminate size at time t, while each technological trajectory is assumed to be discretely bound or coherent at time t. Case A describes a region housing one single technological trajectory. We expect that in this case the region is dominated by industries applying the same associated technologies. Case B depicts a region which houses two unrelated technological trajectories, each of them being largely located outside the region’s border. The peripheral location of these technological trajectories designates that only some of the interactions of agents engaged in propelling those technological trajectories forward can be found in the region. Case C details a region anchoring two dominant, unrelated technological trajectories, as well as two peripheral technological trajectories, likely contributing to a diversified regional economic base. These cases illustrate that a region can be composed of many different technological trajectories which can be partly or largely located outside the region. The important point is that these trajectories can be fairly independent from one another not only because they rely on different assets (in the widest sense) and capabilities but also because they are not connected through important transaction networks. How Today’s Peripheries Become Tomorrow’s Cores: Geographical Industrialization Storper and Walker (1989) avoid the aggregation problem pointed out above by conceptualizing regional development paths in a different way. They contend that a process of geographical industrialization generates uneven territorial development. Geographical industrialization (defined in a broad sense) revolves around three factors: territorial expansion, territorial differentiation, and territorial instability. In this conceptualization, territorial expansion is linked to the creation of novel industries, based on new technologies. Storper and Walker (1989) argue that novel industries tend to form outside of existing industrial centers because they have few specific locational requirements at this stage (e.g. with respect to the characteristics of local suppliers, workers and infrastructure). This occurs because the necessary inputs do not exist anywhere and instead must be cobbled together from whatever materials are at hand (i.e., many locations offer an appropriate mix of modifiable inputs). Thus, numerous territories exist which could house this industry, opening a window of locational opportunity. Over time, Storper and Walker (1989) assume that a handful of locales come to specialize in this once-novel industry, driving selective clustering (Scott and Angel, 1987; Boschma, 1997).
154
Harald Bathelt and Jeff Boggs
However, we find that the model of geographical industrialization has some missing links and, through this, needs to be expanded to conceptualize regional development paths. First, this model does not explain how regions appear to re-invent themselves. Second, it does not clearly articulate a connection between regional expansion and political crises; that is, regional expansion following on the heels of war or wider societal crises (Bathelt, 1997).
8.3 NATIONAL RUPTURES AND REGIONAL DEVELOPMENT PATHS We attribute both faults in the model of geographical industrialization to a general under-politicization of this model. Like most models in economic geography, it makes too few references to the wider socio-political conditions underpinning capital accumulation. Therefore, we suggest a reconceptualization of regional development in which a multiple-trajectory understanding of development is systematically combined with one of discontinuities through national ruptures.
8.3.1 Conceptualization of Crisis in the Regulationist Approach The regulationist approach furnishes a conceptual framework describing how cybernetically related, national institutional compromises reproduce society, forestall economic crises, and mediate economic restructuring (Dunford, 1990; Robles, 1994). The central question of the regulationist approach is to understand what causes the variability of economic and social dynamics in time and space. Among others, this means analyzing how and why an economic formation moves from a pattern of stable, robust growth to one characterized by stagnation or contraction. It also involves the search for explanations for variations in growth and crisis between nations (Lipietz, 1987; Boyer, 1988 and 1990). The institutions of a particular mode of development do not always ensure expanded reproduction. Within a dynamic economy, ensembles of institutions that once stabilized capital accumulation can later fail. This results in changing economic, demographic, environmental, technological, social, and geo-political conditions and undermines the ability of interacting individuals and organizations to maintain social compromises. When this “misregulation” (Boggs, 1997) persists, expanded reproduction stalls, and a crisis results (Boggs and Knudsen, 1996; Barnes, 1997).
CONTINUITIES, RUPTURES, AND RE-BUNDLING
155
Sources of Crises Two forms of national crises have shaped Leipzig’s economic development path during the 20th century. Economic reproduction was blocked due to (i) war and (ii) radical socio-political change. However, these are crises at the national, not the regional scale. We have yet to clarify how national crises affect regional structures and development patterns and what kind of ruptures take place within the institutional compromises at a regional level. From this, we develop the idea that crises can lead to a recombination of local assets that, in turn, stimulate the creation of new capabilities. This eventually leads to local re-development and even to the formation of new industry clusters, sometimes of the Marshallian variety. The Movement from Crisis to Stability What happens after a crisis? Regulation scholars contend that balanced reproduction does not magically follow a crisis (Aglietta, 1979; Lipietz, 1987; Boyer, 1990 and 2000; Hirsch, 1990; Jessop, 1992). Instead, they contend that crises are only overcome through institutional and organizational experimentation. Through these experiments, actors within the economy attempt to restore conditions that are favorable for further capital accumulation. New ensembles are generated piece by piece through political struggles and provisional compromises. Hirsch (1990) states that at any time in a society, secondary models of development co-exist with the established model. During crises, some of these secondary models, previously seen as untenable by most actors within the system, are viewed, and implemented as possible means for securing stability. From these experiments and struggles, new compromises may eventually result. They open up new windows of locational opportunity (Storper and Walker, 1989; Bathelt, 1997; Boggs, 2001).
8.3.2 Ruptures in Interactive Learning and Regional Development Paths Political and sectoral crises strain a regional economy’s social cohesion. Crises disrupt existing transactional networks, releasing resources for alternative uses. Suppliers and service firms that previously focused on the needs of the dominant sector are now open to new ventures and technologies developed in other sectors and/or other regions. Crises lead to two kinds of adjustments: (i) geographical shifts, as assets leave the region; and (ii) sectoral/technological shifts, as assets are re-deployed to the region’s other sectors and technolo-
156
Harald Bathelt and Jeff Boggs
gies. Through sectoral/technological shifts, innovations and talents developed in one sector filter into another.
8.3.3 Re-bundling the Regional Development Path Regions begin to recover from a crisis when agents re-bundle the capital at hand (i.e., financial, physical, human, social capital) for a new round of accumulation. Thus, interactive learning through systematic and trivial observation, information exchange, feedback, imitation, experimentation, and other reflexive social practices enables the region’s actors to re-bundle technological trajectories. Sometimes this becomes the basis of a new regional industry core promoting a novel development path. However, not every new bundle of technologies and related economic activities will have a significant positive impact on a region’s development. In order to shape a region’s development path so that the local effects of a crisis are overcome, a new ensemble of competencies or specialized cluster of interrelated economic activities must develop. This cluster must be anchored into the local economy by nonubiquitous resources found within a specialized social division of labor; it must have a certain minimum size; it must develop particular institutions to reproduce itself; and it must access markets and internalize novel inputs, partially located outside the region. The concept of re-bundling thus does not imply that regional crises are overcome through regional assets alone. In fact, re-bundling processes are often initiated or supported by external agents and their particular assets. However, in order to anchor theses initiatives within a region and to trigger regional development, local actors will also have to be mobilized. Interaction between internal and external actors will have to be established to enable a re-combination of assets and spur new rounds of capital accumulation towards a renewed development path. In this conceptualization, regional bundles are complementary to clusters in that they emphasize a particular perspective, focusing on regional assets. Clusters refer to the interrelated economic activities within a value chain and, thus, emphasize a sectoraltechnological perspective on production systems. Complementary to this, the concept of regional bundles draws attention to the actors and assets that are situated within a particular region or available to that region. Therefore, the bundle perspective focuses on those agents and assets which can be mobilized to anchor a particular technological trajectory within a particular region. Regional bundles and clusters are only the same if all corresponding action and interaction is restricted to that region.
CONTINUITIES, RUPTURES, AND RE-BUNDLING
157
Figure 8.2: Sequence of events leading from crisis to re-bundling.
Figure 8.2 clarifies the relationship between crises, technological trajectories, re-bundling and regional development paths. The sequence shown in this diagram begins at time t at which the regional development path is dominated by a single technological trajectory. Note that this could just as easily be a region housing two or more technological trajectories, with any number
158
Harald Bathelt and Jeff Boggs
of peripheral technological trajectories athwart the regional boundary. In the next stage at time t+1 a crisis – political or sectoral/technological – has shattered the technological trajectory. This rupture creates an opportunity motivating agents remaining in the region to re-bundle capital in attempts (many in vain) to eke out a living. At time t+2, this re-bundling might only lead to a weakly specialized regional economy, with a regional development path characterized by stagnation or dominated by a branch plant economy (case a). This outcome will result if re-bundling does not generate marketable novelty (e.g. in the form of radical innovations) or competitive efficiency (e.g. in the form of incremental innovations). However, re-bundling of regional assets, as we see in case b, can instead lead to the formation of a cluster at time t+2 which drives regional growth. This is often considered to be the optimal motor of regional economic development (e.g. Scott, 1996 and 1998). Maskell (2001), Malmberg and Maskell (2002) and others have contributed to a more nuanced understanding of the processes driving the evolution of a regional cluster. Based on this literature, clusters can be divided clusters along their vertical, horizontal, power, institutional and external dimensions (Bathelt, 2002). Vertical Dimension The vertical cluster dimension consists of complementary firms, such as a network of supplier, service and customer relations. Once a specialized cluster of activities exists, this creates incentives for other suppliers and customers to move to this cluster (Krugman, 1991 and 2000; Scott, 1988). This process forms a localized value chain or parts of it. New firms starting up in this cluster gain access to economies of scale, minimize transaction costs, and benefit from information sharing and inter-firm communication (Lundvall, 1988; Storper, 1995 and 1997; Maskell and Malmberg, 1999; Cooke, 1999; Gertler, 2001). Horizontal Dimension While often overlooked in studies of regional networks, the horizontal dimension of a cluster is as important as the vertical (Maskell, 2001; Malmberg and Maskell, 2002). The horizontal dimension refers to relations between a firm and its direct competitors. Although rivalry might prevent them from closely cooperating or directly sharing knowledge, it does stimulate product differentiation and increased efficiency. Strong inter-firm rivalry thus serves to accelerate innovation and increase competitiveness (Porter, 1990 and 1998; Grabher, 2002).
CONTINUITIES, RUPTURES, AND RE-BUNDLING
159
Power Dimension A cluster’s horizontal and vertical dimensions are shaped by existing power asymmetries that affect the agents’ ability to react to changes in their regulatory environment and external markets (Clegg, 1989; Taylor, 2000; Bathelt and Taylor, 2002). Through power relations that are fluid but temporarily fixed and accepted by the cluster actors, information can be exchanged and translated quite efficiently within the cluster. This, in turn, strengthens the cluster’s internal coherence (Murdoch, 1995; Allen, 1997). Institutional Dimension In order to support the re-bundling process, new institutions must be created. These are created either from the ground up, or by salvaging practices from the wreckage of past institutions. Long-term growth depends on institutions that encourage and enable inter-firm cooperation and collaboration towards coherent projects (Hodgson, 1988; Maskell, 2001). External Dimension Interactive learning does not happen in isolation. It depends on information about external markets and technologies (Bathelt, 2002; Bathelt and Taylor, 2002). If internal linkages are too exclusive and too rigid, such social relations threaten the competitiveness of a firm or a whole group of firms (Uzzi, 1997). Agents must establish linkages with external information sources or else they know nothing about relevant market trends and new technologies. The local innovative successes of agents depend not only on internal or local assets, but also on external assets. In the following section, we understand the rise and decline of Leipzig’s book publishing industry and the subsequent growth of television and film production through the lens provided by our conceptual framework. We hope to show the utility of this conceptual framework for understanding regional development as a process which is driven by periods of growth due to rebundling and punctuated by decline due to external crises.
8.4
CONTINUITY AND RUPTURE IN LEIPZIG’S DEVELOPMENT PATH
Since the mid and late 1990s, local politicians and planners have highlighted the importance of Leipzig as “the city of media.” They frequently draw direct
160
Harald Bathelt and Jeff Boggs
parallels between the city’s pre-War tradition in book trade and publishing and the now growing cluster of activities in TV/film production and new digital media. The following examination calls into question this claim of continuity (for a more detailed analysis, see Bathelt and Boggs, 2003).
8.4.1 Centuries of Continuity: The Evolution of Leipzig’s Book Publishing Industry Until 1945 Leipzig, which had developed into an important trading and population center in the late Middle Ages, became one of Central Europe’s earliest book publishing sites (v. Schroeder, 1904; Schulz, 1989; Wittmann, 1999). For this, its cultural proximity to the Protestant Reformation was also important. By the late 18th century, Leipzig housed numerous firms operating in narrow niche markets. These publishers and subsidiary trades further specialized and concentrated in Leipzig’s Graphisches Viertel. This geographical concentration developed into a “true” Marshallian industrial district of interrelated industrial activities (Boggs, 2001). Numerous institutional innovations further shaped and strengthened the industry: (i) the Börsenverein des Deutschen Buchhandels, a member-run industry organization representing the interests of the various firms involved in book publishing, (ii) the Leipziger Bestellanstalt, a clearinghouse for book orders (through which orders were processed faster, with fewer mistakes and delivered in bulk at a lower cost). These institutions allowed the industry to improve in the first case its ability to shape markets, and in the second case its productivity. In 1938, Leipzig housed over 300 book publishers and 500 allied firms, as well as the Deutsche Bücherei (the German equivalent of the U.S. Library of Congress). Overall, a regional bundle of book publishers and wholesalers developed into an important cluster that marked Leipzig’s overall regional development path until the Second World War.
8.4.2 Regional Rupture I and Its Consequences: Leipzig’s Book Publishing Industry in the G.D.R. The Second World War dramatically ruptured the evolution of the local media sector, as well as other industries in Leipzig. While the Graphisches Viertel was rebuilt following the Second World War, it was integrated into the international socialist division of labor of which the German Democratic
CONTINUITIES, RUPTURES, AND RE-BUNDLING
161
Figure 8.3: Firms of Leipzig’s book publishing industry, 1939 (computed from Adressbücher Redaktion des Börsenvereins des Deutschen Buchhandels zu Leipzig, 1938).
Republic (G.D.R.) and interacted little with West Germany. Its former western market was now split amongst publishers in West Germany. This re-bundling included the migration of some of the industry’s institutions and assets to West Germany. By 1989, the once diverse Leipzig’s publishing industry had been vertically integrated into just 36 VEB (Volkseigene Betriebe; or people’s own enterprises). The industry no longer formed a dynamic hub able to induce on-going growth (Boggs, 2001).
162
Harald Bathelt and Jeff Boggs
8.4.3 Regional Rupture II and Its Consequences: Initial Attempts of Re-constructing Leipzig as a Media City after German Reunification At the time of German Reunification in 1990, Leipzig’s book publishing industry, like most industries in the former G.D.R., was not well positioned for market-driven competition. In 1990, the Förderverein Medienstadt Leipzig (Development Association of the Media City Leipzig), a loosely organized public-private partnership, began agitating for re-development of the Graphisches Viertel as a site for book publishing and affiliated industries (Baier, 1992; Denzer and Grundmann, 1999; Schubert, 2000). The traditional location of print media in the Graphisches Viertel was supposed to develop into the center of Leipzig’s re-invigorated media sector. However, this did not work (Figure 8.3). Leipzig’s book fair did not re-gain its international importance. Furthermore, Leipzig’s publishing houses did not flourish and instead scaled back employment and output. Book publishing thus underwent massive de-industrialization with German Reunification. This second crisis also had strong effects on unemployment and social welfare in the region throughout the 1990s.
8.4.4 Re-bundling a New Regional Development Path: Leipzig’s New Media Industries since the mid-1990s The recent growth of the media industry is not embedded in Leipzig’s traditional industrial structures and institutional settings. Instead, local agents began re-bundling local assets left in the aftermath of drastic ruptures in the region’s development path. Leipzig’s cluster of firms in TV/film production and new digital media, primarily dependent on the newly created publicbroadcasting facilities of MDR (Mitteldeutscher Rundfunk; or Middle German Broadcasting), has yet to develop into a dynamic regional bundle. It does not yet have the same importance for the regional economy as did the former book publishing industry. But why Leipzig at all? In the area of electronic media, such as the television and film industry, Leipzig had never played an important role (Sagurna, 2000). The G.D.R. television and film sector, characterized by a rigid hierarchical organization, was concentrated in East Berlin and nearby Potsdam. The decision to establish MDR in Leipzig was due, in part, to Leipzig’s physical location in the center of southeastern Germany, historically known as Middle Germany. This decision was a political compromise between the three Länder (German states) involved: Saxony, Saxony-Anhalt, and
CONTINUITIES, RUPTURES, AND RE-BUNDLING
163
Figure 8.4: Firms and institutions of Leipzig’s “new” media industry, 2000 (computed from Stadt Leipzig, 2000).
Thuringia. Due to its central location, Leipzig was the only city in Saxony that would be readily accepted by politicians from Saxony-Anhalt and Thuringia, since it is located close to both Länder. Despite Leipzig’s lack of television and film resources, a new media cluster developed in the city during the 1990s around the activities of MDR. While much smaller than that Potsdam-Babelsberg, it is still significant to the local economy. According to a survey of Bentele, Polifke, and Liebert (1998), this TV/film production and new digital media cluster had about
164
Harald Bathelt and Jeff Boggs
8,000 permanent employees in several hundred firms in 1998 (not including the thousands of freelance consultants). With its establishment, MDR rapidly became the most important anchor organization in supporting the establishment of television- and film-related media branches in Leipzig (Sagurna, 2000; Schubert, 2000; v. d. Berg, Braun and v. Winden, 2001). Spurred by MDR’s need for local suppliers, thanks in part to localcontent self-regulation, we observe a re-bundling of Leipzig’s local assets through interactive learning. Given that most of the G.D.R.’s formal institutions did not survive Reunification, new firm formation is the site of rebundling. Bentele, Liebert, and Polifke (2000) found that 84 percent of the firms included in their survey were founded within the Leipzig region. According to interviews conducted by Bathelt (2002), these firms were established by people who either were born in and never left Leipzig or had studied or worked in Leipzig for many years. Many of these entrepreneurs felt an attachment to Leipzig that went beyond the purely economic. MDR-related establishments and privatization of former VEB also spurred local growth in the media industry. Employees from the former G.D.R. television and film industry reported skepticism about working for people from West Germany and, instead, started up their own businesses. Another group of firms in the West German television and film businesses established branches in Leipzig or relocated part of their activities to the region to acquire contracts from and offer services to MDR. In some cases, MDR asked professional film teams, technicians, cutters, reporters, news agencies and other media specialists from other regions to establish a branch in Leipzig and offered future contracts to attract specialized firms. These “transplants” played an important role in the development of Leipzig’s media industry since they brought professional expertise and specialized experience into the region that did not exist previously. Overall, a network of specialized activities and competencies developed which became an incentive for euphoric local start-ups and university spin-offs (Bathelt, 2002; Figure 8.4). The new institutional structure supports this growth. Graduates from special training programs and institutions of higher learning, such as the Hochschule für Technik, Wirtschaft, und Kultur (College for Technical, Economic, and Cultural Studies), establish a basis for further firm formation which is, in turn, supported by start-up consulting and financial funds providing start-up money. Overall, a regional network of institutional support is beginning to develop which provides a distinct institutional thickness (Amin and Thrift, 1995). This is likely to drive future start-up activities in related media industries. At this point, however, the growth process is still in an infant stage and the cluster of media activities has not yet fully unfolded its vertical and hori-
CONTINUITIES, RUPTURES, AND RE-BUNDLING
165
zontal structure, supporting institutions and linkages with external growth markets.
8.5 CONCLUSION Leipzig’s book trade up to the Second World War was deeply rooted in the local economy, with its book publishing industry characterized by outstanding overall capabilities, clearly shaping the regional economy. This can be viewed as a situation where a particular technological trajectory arrayed around book publishing was geographically centered in Leipzig, triggering a dominant regional bundle of interrelated industrial activities. The region had excellent potential for further agglomerated growth barring any sectoral/technological or political crises. However, due to post-Second World War economic and societal ruptures as well as German Reunification in 1990, Leipzig’s book publishing industry was locked out of major German markets. Despite this, a new TV/film and digital media cluster began to develop because of an administrative choice to locate MDR in Leipzig. This was a poorly equipped regional environment for these media branches, populated with inexperienced and technologically limited firms. We argue that this development can be viewed as a re-bundling process where agents redirect and re-combine local and non-local assets towards different or novel bundles of interrelated industrial activities in order to open up new growth opportunities. With this case study, we illustrated the utility of our conceptual framework, and especially the idea of re-bundling in the wake of crisis as an important factor in understanding a region’s development path. Regional development paths consist of bundles of technological trajectories, some dominant, some operating on the region’s periphery. These regional development paths are not continuous, but instead are punctuated by sectoral/technological crises and political ruptures. Crises provide agents chances for interactive learning geared towards re-bundling local capital, spurring growth and change in formerly peripheral or entirely novel industries. The chapter developed a conceptual framework incorporating these claims, and then used this framework to analyze the demise of Leipzig’s book publishing industry and the unrelated growth of its broadcasting, film, and digital media industry. This explained how Leipzig’s pre-War and postReunification media industries are not extensions of the same regional development path or a continuous history as “the city of media.” Instead, they are manifestations of different regional development paths.
166
Harald Bathelt and Jeff Boggs
ACKNOWLEDGEMENTS We would like to thank Gerhard Fuchs and Philip Shapira for organizing the Workshop on “Rethinking Regional Innovation and Change: Path Dependency or Regional Breakthrough” at the Akademie für Technikfolgenabschatzung in Stuttgart. Thanks are also due to Andreas Koch and Sandra Wassermann for their excellent support. We are also grateful for constructive comments which we received from Ron Boschma, Phil Cooke, Meric Gertler, Gernot Grabher, Anders Malmberg, Dominic Power, and Lena Tsipouri.
REFERENCES Adressbücher Redaktion des Börsenvereins des Deutschen Buchhandels zu Leipzig, Adressbuch des Deutschen Buchhandels 1939 (Directory of the German book trade 1939). Leipzig: Verlag des Börsenvereins des Deutschen Buchhandels zu Leipzig, 1938. Aglietta M., A theory of capitalist regulation: The US experience. London, New York: Verso, 1979. Allen, J., “Economies of power and space.” In Geographies of economies, R. Lee, J. Wills, eds. London: Arnold, 1997. Amin A., Thrift N., “Living in the global.” In Globalization, institutions, and regional development in Europe, A. Amin, N. Thrift, eds. Oxford, New York: Oxford University Press 1995. Arrow K. J., The economic implications of learning by doing. Review of Economic Studies 1962; 29:155-173. Arthur, W. B., “Competing technologies: An overview.” In Technical change and economictheory, G. Dosi, C. Freeman, R. R. Nelson, G. Silverberg, L. L. G. Soete, eds. London, New York: Pinter Publishers, 1988. Asheim, B., Interactive learning and localised knowledge in globalising learning economies. GeoJournal 1999; 49:345-352. Baier, H., ed., Medienstadt Leipzig: Tradition & Perspektiven (Leipzig as a media city: Traditions and perspectives). Berlin: Vistas, 1992. Barnes, T. J., “Theories of accumulation and regulation: Bringing life back into economic geography.” In Geographies of economies, R. Lee, J. Wills, eds. London: Arnold, 1997. Bathelt, H., Chemiestandort Deutschland. Technologischer Wandel, Arbeitsteilung und geographische Strukturen in der Chemischen Industrie (German chemical industry. Technological change, the division of labor and geographical structure). Berlin: Edition Sigma Bohn, 1997. Bathelt, H., Regional competence and economic recovery: Divergent growth paths in Boston’s high technology economy. Entrepreneurship and Regional Development 2001; 13:287-314.
CONTINUITIES, RUPTURES, AND RE-BUNDLING
167
Bathelt, H., “The re-emergence of a media industry cluster in Leipzig,” In European Planning Studies 2002; 10:583-611. Bathelt, H., Boggs, J., Towards a re-conceptualization of regional development paths: Is Leipzig’s media cluster a continuation of or a rupture with the past? Economic Geography 2003; 79 (forthcoming). Bathelt, H., Glückler, J., Wirtschaftsgeographie: Ökonomische Beziehungen in räumlicher Perspektive (Economic geography: Economic relations in spatial perspective). Stuttgart: UTB - Ulmer, 2002. Bathelt, H., Glückler, J., Towards relational economic geography. Journal of Economic Geography 2003; 3: forthcoming. Bathelt, H., Taylor, M., Clusters, power and place: Inequality and local growth in time-space. Geografiska Annaler 2002; 84B:93-109. Belussi, F., Pilotti, L., Knowledge creation, learning and innovation in Italian industrial districts. Geografiska Annaler 2002; 84B:125-139. Bentele, G., Liebert, T., Polifke M., Medienstandort Leipzig III: Eine Studie zur Leipziger Medienwirtschaft 2000 (Leipzig as a location of the media industry III: A study of Leipzig’s media economy 2000), Leipzig: Stadt Leipzig, 2000. Bentele, G., Polifke, M., Liebert, T., Medienstandort Leipzig II: Eine Studie zur Leipziger Medienwirtschaft 1998 (Leipzig as a location of the media industry II: A study of Leipzig’s media economy 1998). Leipzig: Stadt Leipzig 1998. Boggs, J., A regulationist analysis of the conditions preceding and following German Reunification. Unpublished Masters thesis. Bloomington (IN): Department of Geography, Indiana University, 1997. Boggs, J., Path dependency and agglomeration in the German book publishing industry. Paper presented at the Annual Meeting of the Association of American Geographers, New York, 2001. Boggs, J., Knudsen, D., “The current transition in industrial capitalism.” In The transition to flexibility, D. C. Knudsen, ed. Boston, London, Dordrecht: Kluwer, 1996. Boschma, R., A. New industries and windows of locational opportunity. A long-term analysis of Belgium. Erdkunde 1997; 51:1-19. Boyer, R., “Technical change and the theory of ‘régulation.’” In Technical change and economic theory, G. Dosi, C. Freeman, R. R. Nelson, G. Silverberg, L. L. G. Soete, eds. London, New York: Pinter, 1988. Boyer, R., The political in the era of globalization and finance: Focus on some regulation school research. International Journal of Urban and Regional Research 2000; 24:274-322. Boyer, R., The regulation school: A critical introduction. New York: Columbia University Press, 1990.
168
Harald Bathelt and Jeff Boggs
Cantwell, J., Fai, F., Firms as the source of innovation and growth: The evolution of technological competence. Journal of Evolutionary Economics 1999; 9:331-366. Clark, G.L., Tracey, P., Lawton Smith, H., Agents, endowments, and path-dependence: A model of multi-jurisdictional regional development. Geographische Zeitschrift 2001; 89:166181. Clegg, S., Frameworks of power. London: Sage, 1989. Cooke, P., “The co-operative advantage of regions.” In The new industrial geography: Regions, regulation and institutions, T. J. Barnes, M. S. Gertler, eds. London: Routledge, 1999. Denzer, V., Grundmann, L. Das Graphische Viertel - ein citynahes Mischgebiet der Stadt Leipzig im Transformationsprozeß: Vom Druckgewerbe- zum Bürostandort (The transformation of Leipzig’s Graphical Quarter: From printing and publishing to modern office functions). Europa Regional 1999; 7:37-50. Dosi, G., Technological paradigms and technological trajectories: A suggested reinterpretation of the determinants and directions of technical change. Research Policy 1982; 2:147-162. Dosi, G., “The nature of the innovative process.” In Technical change and economic theory, G. Dosi, C. Freeman, R. R. Nelson, G. Silverberg, L. L. G. Soete, eds. London, New York: Pinter, 1988. Dunford, M., Theories of regulation. Environment and Planning D: Society and Space 1990; 8:297-321. Gertler, M.S., Implementing advanced manufacturing technologies in mature industrial regions: Towards a social model of technology production. Regional Studies 1993; 27:665-680. Gertler, M.S., Local knowledge: Tacitness and the geography of context. Paper presented at the Annual Meeting of the Association of American Geographers, New York, 2001. Gertler, M.S., Worlds apart: The changing market geography of the German machinery industry? Small Business Economics 1996; 8:87-106. Grabher, G., “Rediscovering the social in the economics of interfirm relations.” In The embedded firm. On the socioeconomics of industrial networks, G. Grabher, ed. London, New York: Routledge, 1993. Grabher, G., The project ecology of advertising: Tasks, talents and teams. Regional Studies 2002; 36:245-262. Hirsch, J., Kapitalismus ohne Alternative? (Capitalism without alternative?) Hamburg: VSA, 1990. Hodgson, G.M., Economics and institutions: A manifesto for a modern institutional economics. Cambridge: Polity Press, 1988. Jessop B., “Fordism and post-Fordism: A critical reformulation.” In Pathways to industrialization and regional development, M. Storper, A.J. Scott, eds. London, New York: Routledge, 1992.
CONTINUITIES, RUPTURES, AND RE-BUNDLING
169
Krugman, P., Geography and trade. Leuven: Leuven University Press; Cambridge, MA, London: MIT Press, 1991. Krugman, P., “Where in the world is the ‘new economic geography’?” In The Oxford handbook of economic geography, G.L. Clark, M.P. Feldman, M.S. Gertler, eds. Oxford: Oxford University Press, 2000. Lipietz, A., Mirages and miracles: The crises of global Fordism. London: Verso, 1987. Lundvall, B.-Å., “Innovation as an interactive process: From user-producer interaction to the. national system of innovation.” In Technical change and economic theory, G. Dosi, C. Freeman, R.R. Nelson, G. Silverberg, L.L.G. Soete, eds. London: Pinter, 1988. Malecki, E.J., Technology and economic development: The dynamics of local, regional, and national change. Burnt Mill: Longman; New York: Wiley. 1991. Malmberg, A., Maskell P. The elusive concept of localization economies. Towards a knowledge-based theory of spatial clustering. Environment and Planning A 2002; 34:429-449. Markusen, A., Sticky places in slippery space: A typology of industrial districts. Economic Geography 1996; 72:293-313. Maskell, P., Towards a knowledge-based theory of the geographical cluster. Industrial and Corporate Change 2001; 10:921-943. Maskell, P., Malmberg, A. Localised learning and industrial competitiveness. Cambridge Journal of Economics 1999; 23:167-185. Murdoch, J., Actor-networks and the evolution of economic forms: Combining description and explanation in theories of regulation, flexible specialization, and networks. Environment and Planning A 1995; 27:731-757. Nelson R. R., Evolutionary theorizing about economic change. Journal of Economic Literature 1995; 23:48-90. Porter, M.E., The competitive advantage of nations. New York: Free Press, 1990. Porter, M.E., Clusters and the new economics of competition. Harvard Business Review 1998; November-December:77-90. Rigby, D.L., Essletzbichler J., Evolution, process variety, and regional trajectories of technological change in U.S. manufacturing. Economic Geography 1997; 73:269-284. Robles, D.F., French theories of regulation and conceptions of the international division of labor. New York: St. Martin’s, 1994. Rosenberg, N., Inside the black box: Technology and economics. Cambridge, New York: Cambridge University Press, 1982. Sagurna, M., “Der Medienstandort Leipzig im Freistaat Sachsen (Leipzig’s role as a media location in Saxony).” In Medienstadt Leipzig: Vom Anspruch zur Wirlichkeit, H. Grunau, W. Kleinwächter, H.-J. Stiehler, eds. Leipzig: Monade, 2000.
170
Harald Bathelt and Jeff Boggs
Sayer, A., Realism and social science. London: Sage, 2000. Schubert, D., “Die Stadt Leipzig und die Medien als Wirtschaftsfaktor (The city of Leipzig and the economic importance of media).” In Medienstadt Leipzig: Vom Anspruch zur Wirklichkeit, H. Grunau, W. Kleinwächter, H.-J. Stiehler, eds. Leipzig: Monade, 2000. Schulz, G., Buchhandels-Ploetz (Ploetz book trade directory). Freiburg: Ploetz, 1989. Scott, A.J., New industrial spaces. Flexible production organization and regional development in North America and Western Europe. London: Pion, 1988. Scott, A.J., Regional motors of the global economy. Futures 1996; 28:391-411. Scott, A.J., Regions and the world economy: The coming shape of global production, competition, and political order. Oxford, New York: Oxford University Press, 1998. Scott, A.J., Angel D.P., The US semiconductor industry: A locational analysis. Environment and Planning A 1987; 19:875-912. Stadt Leipzig, ed., Branchenhandbuch Medien: Region Leipzig 2000 (Firm directory of the media industry in Leipzig 2000). Leipzig, 2000. Storper, M., The resurgence of regional economics, ten years later. European Urban and Regional Studies 1995; 2:191-221. Storper, M., The regional world. Territorial development in a global economy. New York, London: Guilford, 1997. Storper, M., Walker R., The capitalist imperative. Territory, technology, and industrial growth. New York, Oxford: Basil Blackwell, 1989. Taylor, M., “Enterprise, power and embeddedness: An empirical exploration.” In The networked firm in a global world: Small firms in new environments, E. Vatne, M. Taylor, eds. Aldershot, Burlington: Ashgate, 2000. Uzzi, B., Social structure and competition in interfirm networks: The paradox of embeddedness. Administrative Science Quarterly 1997; 42:35-67. v. d. Berg, L., Braun, L., v. Winden W., Growth clusters in European metropolitan cities. A comparative analysis of cluster dynamics in the cities of Amsterdam, Eindhoven, Helsinki, Leipzig, Lyon, Manchester, Munich, Rotterdam and Vienna. Aldershot, Burlington, USA: Ashgate, 2001. v. Schroeder, F., Die Verlegung der Büchermesse von Frankfurt am Main nach Leipzig (The relocation of the book fair from Frankfurt/Main to Leipzig). Leipzig: v. Jäh & Schunke, 1904 (Reprinted by Zentralantiquarit der DDR, 1970). Wenger, E., Communities of practice: Learning, meaning, and identity. Cambridge: Cambridge University Press, 1998. Wittmann, R., Geschichte des deutschen Buchhandels (History of the German book trade). Munich: Beck, 1999.
Chapter 9 CAN LESS FAVORED REGIONS CHANGE THEIR DESTINY? LESSONS FROM EUROPE
Lena J. Tsipouri
9.1 INTRODUCTION The goal of economic convergence among member countries and their constituent regions is one of the pillars of the European Union (EU), particularly after the adoption of the Single European Act in 1986. Yet, neither academics nor politicians are in agreement among themselves as to how convergence can be achieved. One perspective is that European integration will automatically lead to economic convergence through the liberalization of capital and labor markets. The contrary view predicts continued economic divergence between Europe’s core and peripheral regions. The trouble is that long-term empirical evidence confirms neither position. Economic convergence or divergence in the Europe Union can be measured at both the national and regional levels. Available policy instruments, interpretations of determining factors, and empirical findings in terms of convergence or divergence vary according to which level is examined. The “national” level refers to the distinction between advanced countries and cohesion countries. In the latter, Gross Domestic Product (GDP) per capita is lower than 90 percent of the EU average. These countries (Spain, Portugal, Ireland, and Greece) are eligible for support from the EU’s Cohesion Fund in the areas of transport and environmental protection. National convergence is a crucial political issue, as decision making in the Union is made at the Council level, which is composed of national ministers. The “regional level” is sub-national, leaving the decision of the administrative division to the member states. Thus, European regions differ in terms of size, performance, and political structure, plus they lack powerful political representation.1 The distinction between advanced and “less-favored regions” relies on GDP per head. There are approximately sixty less-favored regions (LFRs, called also Objec-
172
Lena J. Tsipouri
tive 1 regions in EU jargon) in the European Union where GDP per capita is lower than 75 percent of the EU average. Reasons behind the low performance of these regions vary, as do their productive structures. Yet, they share certain characteristics such as insufficient accumulated physical infrastructure, inadequacies in human and social capital, a low share of high-tech production in either manufacturing or services, and low intangible investments. This differentiated performance among European regions and the study of the destiny of the less prosperous among them is of generalized interest, because there are certain unique characteristics in their recent development process. These offer new insights for the academic literature and for policy: Financial resources used to be scarce, but they are not anymore. Regional development funds redistributed via the EU budget have increased considerably since the mid-1980s; the lack of availability of physical capital can no longer be used as an explanation for low growth rates. There is an interesting mix of regional development tools and mechanisms. Regional development plans follow a mainstream logic, funding mainly physical infrastructure, but at the same time, there is an increasing effort to direct more and more support towards technology, information society enhancement, and innovative actions. The development planning and delivery mechanisms are unique and evolving over time. The planning process is a shared responsibility between an outsider (the European Commission acting as a supervising authority for European taxpayers’ money), and national and regional authorities. Shifts in responsibilities and power can be observed over the years and constitute by themselves an interesting research topic in regional aid effectiveness. The “subsidiarity” principle, suggesting that decision-making should be taken at the lowest possible level, would give regions the ability to plan, while national authorities decide on the distribution of funding. Performance within these regions varies greatly. Some regions clearly have closed the gap, represented best by the Ireland’s spectacular recent growth; other regions have maintained their relative positions; while still others suffer from a further widening of the development gap, despite ample financial support for 15 years. This chapter rapidly reviews the existing literature on convergence and divergence. It then examines individual less-favored regions in Europe and tries to identify which ones have succeeded in “changing their destiny.” After discussing reasons for the presence or lack of success, the chapter offers con-
CAN LESS FAVORED REGIONS CHANGE THEIR DESTINY?
173
cluding insights into the development processes underway in these different categories of regions.
9.2 OVERVIEW OF THE LITERATURE The regional development process in Europe has been extensively studied in the academic literature and in a variety of policy studies. The latter contain relevant data and valuable insights into regional diversity. The neoclassical growth model accounts evidence on convergence. It is even suggested that open-economy versions of the theory predict higher rates of convergence than those observed empirically (Barro et al., 1995). These models have been criticized from various points of view, (as by Levine and Renelt, 1992) suggesting that if alternative methods are used evidence is provided against the convergence hypothesis (Quah, 1993). Other approaches suggest that these models may measure only the convergence by certain “clubs” of countries (Baumol, 1986) and not state the general rule (Maurseth, 2001). More recent criticism refers to the need to consider interactions between the regions and countries in the process of their growth, be it technology spillovers, trade, or other linkages indicating that geographical proximity may be more relevant than common administrative borders. Endogenous growth theories, through the adoption of deficient excludability and the cumulative aspects of knowledge, imply spillovers leading to increasing returns that are consistent with competitive markets (Romer, 1986). Subsequent academic models then deny automatic convergence and R&D and proxies for knowledge become determining factors on whether convergence will occur or not. Technology becomes a key variable: Innovations in the advanced countries/regions increase divergence. Technological diffusion in less competitive countries or less favored regions determines whether they will converge or diverge. Yet diffusion is all but self-evident and automatic. It depends on the ability of the weaker countries/regions to be successful recipients of transborder technology transfer, which depends not only on their economic attractiveness for foreign direct investment, but also on their domestic absorption capacities (Abramovitz, 1993; Cohen and Levinthal, 1989). Economic geography further complicates the model when exploring the interplay between increasing returns, market size, and geographical distance (Krugman, 1991). With increasing returns and transportation costs, firms will tend to establish in larger markets (Maurseth, 2001). In the context of this vibrant theoretical debate, the application of the convergence-divergence dilemma in the European regions is of general interest. The rich and competitive regions, known as “core regions” (i.e. the advanced regions in central Europe composed mainly of the South East of the
174
Lena J. Tsipouri
UK, Northern France and Germany, Northern Italy. Belgium and the Netherlands) systematically have a much higher GDP per head. Research has focused both on their development in relation to peripheral regions, and on the implications of trade liberalization and wider market integration on cohesion. The results of empirical research are inconclusive. They differ considerably depending on the number of regions used in the samples, the periods studied, and the methodologies applied. Sala-I-Martin (1996a) suggests convergence at a steady rate of about 2 percent for 1950-1990, while with more recent data Neven and Gouyette (1995) find that from the early 1980s convergence stagnates. More empirical research in the mid-1990s demonstrates that European regions were heading for divergence. After a slow but steady reduction of differences in GDP per capita across European regions during most of the post-war period, there were signs of reversal in this trend. Based on data from 143 European regions, it is suggested again that the traditional convergence effect is present: poorer regions grow on average faster than richer regions. The halt of convergence observed in the 1980s should be seen as temporary. While poorer regions seem to grow faster than richer regions, splitting the pace of convergence into periods shows substantially different convergence rates. In addition, which denotes that the standard deviation of GDP per capita decreases over time and as such should be seen as a stricter measure, confirms the tendencies to converge. It also confirms that geographical and technological factors seem to be important. In the period 1980-1994, innovation and technology spillovers contributed to economic growth, and peripheral regions caught up with central regions (Maurseth, 2001). In a recent survey on regional disparities in Europe (Cuadrado-Roura, 2001: 335) it is argued that “on the one hand most of the neo-classical analyses tend to show a certain global predominance of economic convergence (Barro and Sala-I-Martin, 1991; Sala-I-Martin, 1994, 1996a, 1996b; Dunford, 1994) on the other, this trend is widely debated (Cuadrado et al., 1998; Figleton, 1999; Magrini, 1999). It is even suggested that the European regions are evolving towards a polarization model or the so-called ‘twin peak’ model (Chaterji, 1993; Quah, 1996a, 1996b). The existence of such disparities is due to various causes.” In the same article, European regional disparities are classified into three periods: From 1960 to the mid-1970s, the prevailing process is convergence. From the mid-1970s, economic convergence by regions and countries comes to a halt. From the mid-1980s to 1996, disparities existing in the global level are clearly stabilized ... and at the same time, there is both convergence and divergence.
CAN LESS FAVORED REGIONS CHANGE THEIR DESTINY?
175
Policy reports argue that disparities between Member States of the European Union continue to exist even though they have been substantially reduced since 1988. The main change affects the countries benefiting from the Cohesion Fund: these noticeably moved closer to the Community average in terms of their per capita GDP, but within the Member States these disparities have sometimes even been aggravated. In 2003, the evidence suggested by the European Commission (European Commission, 2003) is that economic convergence is continuing in the Member States yet not at all levels: At the national level, the cohesion countries are continuing to make up lost ground; with the exception of Greece, they continued to narrow the gap. Ireland’s per capita GDP increased from 64 percent to 119 percent of the EU average between 1988 and 2000. At the regional level, the EU disparities are narrowing. Within Member States, by contrast, disparities have worsened. There is thus a certain consensus in academic and political circles that regional cohesion is not automatic and further research is necessary to identify determining factors that make a difference to the pace of regional development. Part of this research tries to group converging regions and effects of financial support. In an analysis trying to quantify the impact of the completion of the Single Market, measuring pre and post 1987 convergence tendencies, Cambridge Econometrics suggest that there was a faster rate of regional convergence in Europe due to a significant performance in problem regions, with the exception of Greece and Southern Italy. However, this was not reflected within the countries. In other words, the indication that the performance of poorer regions improved after 1987 appears to be mainly due to a better performance by poorer countries, linked to the Community Support Framework (CSF) spending. The Single Market does not appear as relevant. The equilibrium implied by regional growth rates in 1987-1993 is characterized by greater economic convergence than that implied by growth rates in the pre-1987 period. Additionally, the rate at which equilibrium is being approached appears to have increased, but it is still slow (Cambridge Econometrics, 1997). Other studies refer to the distribution and effectiveness of financial support. Taking increasing regional development support into consideration in an ex ante assessment of the effects that the second CSF was likely to have on the economy of Greece, Christodoulakis and Kalyvitis conclude on very positive outcomes if positive externalities are taken into account. These authors stress the importance of hard infrastructure and indicate that soft infrastructure should take into account the demand that is likely to develop for such services (e.g. R&D, culture, or health). The concentration on top-down ac-
176
Lena J. Tsipouri
tivities that cannot attract substantial demand from the production and service sector should be avoided (Christodoulakis and Kalyvitis, 1998). Adapting to the international debate many authors have started investigating new aspects such as the relevance of technology, networking effects and governance levels in Europe. In their book on Globalization, Institutions, and Regional Development in Europe, Amin and Thrift argue that regional economic prosperity in Europe will depend on the degree to which regions are able to mobilize flexible institutional strategies (Amin and Thrift, 1996). The conclusions are that the supranational-national-sub national levels shape roles and strategies (cf. Ansell et al., 1997) ... and that regional labor markets in Europe are not flexible enough to absorb shocks. Region-specific changes in the demand for labor indicate that, in the long run, workers migrate into regions with booming labor markets to look for employment (Decressin and Fatas, 1995). The centralized wage setting in European Member States works against regional convergence (Faini, 1999). Checking for important variables, notably R&D effort, investment support from the EU, the structure of GDP and differences in unemployment, suggest that they have had a diverging impact. Generally, the inclusion of R&D efforts and directed credit leads to a significant increase in the explanatory power of a model tested by Fagerberg and Verespagen in 70 regions. Three different “growth clubs” are identified, each with its own dynamics (Fagerberg and Verspagen, 1996). Further analysis confirmed these findings and opened up the discussion for employment and migration (Fagerberg, Verspagen and Caniels, 1997). Finally, although not referring directly to convergence, Robert Putnam (1993) enriches the regional growth debate with his emphasis on the role of “social capital.” This refers to features of social organization, such as networks, norms, and trust that facilitate coordination and cooperation for mutual benefit. Social capital enhances the benefits of investment in physical and human capital. Some regions of Italy, such as Emilia-Romagna and Tuscany, have many active community organizations. Citizens in these regions are engaged by public issues, not by patronage. They trust one another to act fairly and obey the law. Leaders in these communities are relatively honest and committed to equality. Social and political networks are organized horizontally, not hierarchically. These “civic communities” value solidarity, civic participation, and integrity. Democracy works here. At the other pole are “uncivic” regions, like Calabria and Sicily, aptly characterized by the French term incivisme. The very concept of citizenship is stunted there. Engagement in social and cultural associations is meager. From the point of view of the inhabitants, “public affairs” is somebody else’s business – for example, i notabili, “the bosses,” or “the politicians” – but not theirs. Laws, almost everyone agrees, are made to be broken, but fearing others’ lawlessness, every-
CAN LESS FAVORED REGIONS CHANGE THEIR DESTINY?
177
one demands sterner discipline. Trapped in these interlocking vicious circles, nearly everyone feels powerless, exploited, and unhappy. It is hardly surprising that representative government here is less effective than in communities that are more civic. The conclusion from this brief survey is that there is no uniform pattern of convergence at the European regional scene. Time-horizons, types of regions, institutions, and technology may affect the relative growth of regions. There are indications of convergence and indications of divergence in different periods, different countries, and different regions. There are clubs of converging regions and others falling behind. There has been leapfrogging in some cases. Most importantly, there are explanatory variables like innovation, regional development aid, and social capital, but none of them have been tested on such a universal validity that can be mathematically confirmed to be a way for regions to change their destiny. Thus, what needs to be done is a detailed discussion of policy and performance in the less favored regions, to identify potential lessons.
9.3 THE WINNING REGIONS This section attempts to identify winning regions that succeeded in growing rapidly over the period 1988 to 1998. The former year marks the beginning of a coherent approach in regional policy in the EU. Based on objective measures of GDP and population, each region with an LFR status was entitled to receive support from all European Structural Funds3 in an effort to catch up with its neighbors and be in a position to compete in the Single European Market. Prior to this, regional support was not only much smaller but was distributed without a coherent, EU-national-regional long-term agreement. This agreement known as the “Community Support Frameworks” (CSFs) marks a turning point in European regional planning. The analysis in this section ends in 1998, due to data availability. The total population of European less-favored regions trying to change their destiny through transfer of capital in a CSF context is composed of approximately 60 regions. To date, the planning process has encompassed three distinct periods: 1988-1993, 1994-1999, and 2000-2006. A negotiation on the status and the amounts per region and period starts approximately two years before each new planning period starts. In 1988, 45 regions were given LFR status. Beginning in 1994, in the second programming period, 12 new regions entered the eligibility list for regional support: five of them from the EU advanced countries, as their GDP grew less than the EU average and they fell below the 75 percent mark, and seven because of new political events. Thus, the reunification of Germany added 6 new Bundesländer, while the accession
178
Lena J. Tsipouri
of Austria brought a poor region with it. No region lost the LFR status, although it was clear from the beginning that the Italian region of Abruzzi had reached the 75 percent frontier and its aid diminished considerably. Support during the 1994-1999 period was higher and better managed, so the results were quite encouraging. In planning for the 2000-2006 period, ten regions performed so well that they passed over the 75 percent mark and became ineligible for further support. Ireland, which was previously one of these regions, was split into two regions: one-half remained eligible for support, while the other half – now one of Europe’s most prosperous regions – entered a process of “phasing out” which provided symbolic transitional support to help sustain growth. Less favored regions originate from both advanced and cohesion countries. During this period, nine new regions (11 if the 2 Irish sub-regions are added) entered the population, three LFR regions from the UK, and three regions from each Finland and Sweden. These changed their status from “sparsely populated” into an LFR in an effort to streamline support eligible categories of the EU regional development policy. Thus, through entries and exits, the number of regions eligible for support is now 58, yet twenty percent of them are different from the previous period.
The regions overcoming the 75 percent threshold are the most successful ones at first sight: Belgian Hainaut, French Corsica, French Hainaut, German East Berlin, Italian Molise, Irish South and East, Dutch Fevoland, Portuguese Lisbon Tagus Valley, Spanish Cantabria, and Northern Ireland and the Highlands and Islands in the UK. By the same definition the regions, which performed so badly in relative terms that they joined the LFR league in a previous period, were the Highlands and Islands, Merseyside, Cantabria, South Yorkshire, West Wales and Valleys, Cornwall and the Isles of Scilly. Other regions that joined for political reasons (late accession or change of definition, are self-evidently not included). This means that Cantabria and the Highlands and Islands are at the
CAN LESS FAVORED REGIONS CHANGE THEIR DESTINY?
179
border of entering and exiting the less favored status, leaving just Corsica, Molise, Ireland and Northern Ireland as the major winners. There are still both theoretical and practical problems concerning why the simple entry to or exit from the “LFR” league is not sufficient to draw conclusions on the medium and long-term development trends. The most important practical problem relates to changes in the statistical methodology in data selection, where some regions present such abnormal tendencies that they have to be eliminated as outliers. Suggesting that the legal status alone is a sufficient criterion to determine winning regions, or even more, regions that have changed their destiny would be naïve. From a theoretical point of view, there is no academic argument as to why “75 percent” should be the best threshold. It constitutes a political compromise rather than any aspiration suggesting the rise of endogenous forces, when the specific number is reached. There is also no single argument at to why it is better or worse than 70 percent or 80 percent. However, even if 75 percent could be proven as clear-cut threashold, in a closer discussion it is difficult to suggest that all the exit regions have changed their destiny. French and Belgian Hainaut were already very close to the 75 percent mark. The EU administration itself has contested their nomination as Objective 1 in 1994 and nobody was surprised to see them losing their LFR status in 1999. When very close to the threshold there are two additional reservations. First, despite an effort to compile accurate statistics, when a region comes close to the 75 percent mark, there is a political effort to remain below the target, in order to maintain fund transfers. Second, it is in general assumed that the informal sector is higher in LFRs and in the Southern countries (see Le Monde, November 13, 2001, on the incorporation of the Italian informal sector into the taxation system), so Southern regions very close to 75 percent may also be considered as having crossed the barrier. OECD statistics4 demonstrate a very large divergence of the informal economy as a GDP share. With 1998 data (and all the reservations for their validity) as presented on Table 9.2, one can easily assume that the bad performance of the Greek and Southern Italian regions, observed both on the EU statistics and mentioned in the academic literature, can be attributed to the above average size of the informal economy in these two member states. Ideally, one should rely on regional data on the informal economy, yet there is no evidence available. Assumptions about higher informal revenue in rural regions, and regions geographically further from national capitals are too vague to be of any value. A quantitatively more elegant method is to study regions by the level of percentage points they could improve their GDP per head in purchase power standards (PPS). Table 9.3 presents regions considered as LFRs since the beginning of the period in descending order of GDP growth. When trying to
180
Lena J. Tsipouri
conclude on persistent winners there are several reasons why in this case we need to concentrate on the original 48 regions: The three regions from the rich countries, notably France, Belgium and the Netherlands entering in 1994 should be excluded, not only because of the controversy about their status but also because of their immediate exit afterwards The German LFRs, which have a clear Objective 1 status, have received a totally different treatment,5 compared to other European LFRs and they merit a distinct study The six Scandinavian regions were incorporated for different reasons than poverty or low productivity. Table 9.3 helps identify the regions that grew most rapidly. Again, there are serious considerations that relate to the effects of diminishing returns. Some of the most rapidly growing regions in Europe started so low that one could only expect them to start a catching-up procedure rapidly: their levels of infrastructure, technology and skills can be assumed to have been so low in 1988 that any additional infrastructures would have very high returns. Based on the work of Abramovitz (1993), the catching-up process is easier for laggards, in particular when they succeed in developing new institutional instruments. In that sense, the regions that grew rapidly but still have a low position cannot be considered as having really changed their destiny. The counterargument is that this is not an automatic process and some of the particularly lagging regions further deteriorated their low ranks. Assessing this would require different research addressing reasons for falling behind rather than abilities to change destiny.
CAN LESS FAVORED REGIONS CHANGE THEIR DESTINY?
181
182
Lena J. Tsipouri
Combining the two classifications, one can argue that there are three types of winning regions in Europe that were, to a certain degree, in a position to change their destiny in the 10 years studied (see Table 9.4). Firstly, those that demonstrate a very rapid growth crossed the 75 percent EU average threshold, despite the fact that they already started high. These regions do not owe their success to their dramatically low starting position, could grow despite diminishing returns, and are now in a position to compete with the EU average and potentially, core regions. These should be called the golden winners. (Attica, actually not included by official statistics, is added due to a very high informal economy and a long negotiation for its acceptance by the Commission). Regions, which grew rapidly but remain below the 75 percent threshold, are considered as partial winners (silver). Finally, regions that grew rapidly but this growth is partly6 attributed to their low start also indicate that they are in a process of changing their destiny. Nevertheless, one
CAN LESS FAVORED REGIONS CHANGE THEIR DESTINY?
183
should not forget that this growth was performed in a period of continuous international growth of both wealth and trade. It is now, with relatively higher GDP/head levels, that they will prove whether they did indeed transform their economic structures. Needless to say that even this classification suffers from the inadequacy of GDP per head as a development indicator.7 The regions that seem to have changed their destiny can be distinguished into: The capitals of the three cohesion countries8 that have been doing best, and East Berlin. The growth of the metropolitan areas is at least partly explained by the urban agglomeration argumentation. Corsica, Cantabria, and Molise among the less developed non-urban areas in the cohesion countries, plus Northern Ireland and Highlands and Islands in the UK.
9.4 SEARCHING FOR EXPLANATIONS OF SUCCESS What differentiated the performance of the two categories of winning regions compared to the others? Did regional support make the difference or can other explanatory variables be used? Moreover, if the latter is the case, do these other explanatory variables indicate path-critical institutions or ways to influence incremental changes? Finally, is whatever we learn from those successful regions enduring? If yes, can lessons be transferred to other, less successful regions? Table 9.5 show that market size, measured in terms of total population, is not significant, with the exception of the capital cities.
184
Lena J. Tsipouri
Physical capital was not found to play a crucial role either. There is no correlation whatsoever between per capita transfers and success. Transfers are negotiated on the basis of the initial position of the regions, but a large share of these funds is managed at the national level. Often, major physical infrastructure is concentrated in the most advanced regions, while many other investments are distributed via national channels (health, security, intangibles) and thus can be assumed to reinforce current set-ups. Thus, if physical capital is insufficient, one has to turn to other variables that try to capture human capital or technology. A series of correlations between growth rates on the one hand and the potential explanatory variables on the other were tested: Regional development funding, as a proxy to physical capital since most of the LFRs use all national funds as a leverage for mobilizing EU resources, and as a consequence (with a major exception in Germany) there are no major funds available except the CSF. Educational levels, as a proxy for human capital, measured in shares of the population in the primary, secondary and tertiary education
CAN LESS FAVORED REGIONS CHANGE THEIR DESTINY?
185
Technology, measured in terms of gross expenditure on research and development to gross domestic product (GERD/GDP). This particular indicator may be misleading, as the component of GERD that increases productivity and GDP in the long run is the part of GERD that comes from the business sector (OECD, 2000). As data for business expenditure on R&D (BERD) is not available at the regional level, the use of GERD data includes a top-down pump of funds into the public system, usually insufficiently linked with the productive sector.
9.5 THE ROLE OF TECHNOLOGY AND INSTITUTIONS Technology gaps are by far, wider than gaps in wealth or revenue. This has triggered the idea of their relevance for development. New indicators on the state of technological advance confirm that technological innovation and the growth of the knowledge-based economy are lagging behind in the countries of Southern Europe. The number for patents applied for per million of the population in Finland, Sweden and Germany is at least twice the European average compared with a figure of less than half in Ireland, Italy, Spain, Greece and Portugal (European Commission, 2003). Technology as a driving force of development is now broadly recognized not only in the academic literature (Fagerberg, 1988, Fagerberg and Verspagen, 1996) but also by policy makers (European Commission, 2001a). Regions with a strong knowledge-based technological infrastructure realize greater numbers of product innovations (Feldman, 1994). Feldman and Florida (Feldman and Florida, 1994) undertook survey research that did show a positive connection between the propensity to innovate and a region’s supply of university and industry R&D. Other studies are less positive. Overall, achievements and regional development efforts in the 1970s and 1980s have been modest in terms of both stimulating sustained improvements in the economic competitiveness of the less favored regions, and in terms of encouraging self-sustaining growth based on the mobilization of local resources and interdependencies. Partly in response to these failings, more progressive policy communities have begun to explore a new alternative, designed to secure economic competitiveness by mobilizing the endogenous potential of the less favored regions through efforts to upgrade the local supply-side infrastructure – to unlock the “wealth of regions” (Amin, 1998). As successful policy schemes demonstrated rapid catching-up effects, and the Third Italy as well as certain originally poor regions in Spain and
186
Lena J. Tsipouri
Ireland, started demonstrating above average or even spectacular growth rates, their good practices spread and policy makers started imitating them. The European Commission analyzed the social returns on investment of R&D and strongly encouraged member states to increase R&D resources. Initially, the share of CSF funds absorbed for R&D, and later the share devoted to innovation and the enhancement of the information society, became benchmarks for efficient regional development spending. During the process towards developing a European Research Area, the Commission adopted a special communication for synergies between its research and regional policy9. Despite this top-down enforcement of increasing R&D spending, Oughton and colleagues claim the existence of a regional innovation paradox. This involves the apparent contradiction between, on the one hand, the comparatively greater need to spend on innovation in lagging regions, and, on the other hand, their relatively lower capacity to absorb public funds, earmarked for the promotion of innovation, and invest in innovation-related activities, compared to more advanced regions. They suggest as a remedy the integration of technology and innovation policy (Oughton et al., 2000). This should not be seen as a paradox but as a rule. There are two main reasons why less favored regions do not invest in technology: first, when resources are scarce and physical infrastructure lacking, it is difficult to dedicate those resources to higher uncertainty and longer-term intangible investments. Politically at least the physical investments receive priority. Second, the economic argument is that in the vicious circles of underdevelopment, R&D is largely driven by demand and this demand does not exist in less favored regions. Many top-down initiatives for technology investments, in particular those trying to create bigger intermediaries in the public sector, have been negatively evaluated (Higgins et al., 1999). The capacity thus created, remained idle because the initiatives did not build trust relationships with the demand side. Data and analyses indicate that the technology gap between the less favored regions and those in the Member States where research and innovation expenditure is highest (Germany, France, Sweden, and Finland) has widened rather than narrowed (with the notable exception of Ireland). The technology gap is reflected at the level of the regions (European Commission, 2001b). Yet, if one looks at the successful regions GERD/GDP seems to be higher than in the average LFRs, with the exception of the low-start regions that clearly based their success in tourism (see Table 9.6). GERD and business expenditure on R&D (BERD), as proxies for innovative behavior, need a much larger discussion. Twelve billion Euros for RTDI in the last decade were given to LFRs through the CSFs. This constitutes public funding, for which economic exploitation is by far not always assured. Absorptive capabilities in these regions could hardly be created
CAN LESS FAVORED REGIONS CHANGE THEIR DESTINY?
187
within the time frame studied. It was mainly the university system, that proved more resilient to change, that absorbed the CSF-R&D and Framework Program (the centrally managed competitive R&D funds of the European Union) funding. Yet the best academic research teams network globally not locally. Consequently, increasing publicly funded R&D in less favored regions may theoretically even have a negative effect on growth: the best regional talent is given the opportunity to work in international consortia whose intellectual property rights are outside the region. GERD and business expenditure on R&D (BERD), as proxies for innovative behavior, need a much larger discussion. Twelve billion Euros for RTDI in the last decade were given to LFRs through the CSFs. This constitutes public funding, for which economic exploitation is by far not always assured. Absorptive capabilities in these regions could hardly be created within the period studied. It was mainly the university system, that proved
188
Lena J. Tsipouri
more resilient to change, that absorbed the CSF-R&D and Framework Program (the centrally managed competitive R&D funds of the European Union) funding. Yet the best academic research teams network globally not locally. Consequently, increasing publicly funded R&D in less favored regions may theoretically even have a negative effect on growth: the best regional talent is given the opportunity to work in international consortia whose intellectual property rights are outside the region. Using BERD as an indicator is much more reliable. Empirical work by the OECD confirms that it is business expenditure for R&D, which has the highest contribution to total factor productivity. BERD, and in particular business-financed expenditure, is extremely low in the LFRs. One may argue that public spending in technology enhancement can be beneficial for regional development only when it leads to increasing business spending. Only Ireland among the winning regions shows a high and systematically increasing share of BERD/GDP. The next effort is to find proxies for absorptive capabilities. From the data available at the regional level, educational attainment was the most appropriate proxy. With the exception of capital cities, the share of the labor force with high education is rather low, while in Portugal in particular, the share of the population having only primary education is extremely high (see Table 9.6). Correlations were again not significant. These remarks leave us with one alternative: to try to test the influence of institutional set-ups in regional development. One major characteristic of the LFRs is their long-term perception of uncertainty that leads them to individualistic behaviors and a low propensity to network. In an effort to fight against this attitude the European Commission has launched a series of incentives that aim to help regional administrations network and improve their planning and implementation capabilities. As indicated on Table 9.6 substantial proportion of successful regions applied for funding to become a RIS (regional innovation system) or RISI (regional information society initiative) region. These initiatives offered funding for the design and pavement of the implementation of long terms strategies in innovation policy and the information society respectively. This approach is used to identify whether such schemes can be used as a remedy for drawbacks in LFRs as identified by the extensive evaluation of innovation support in the second programming period of the CSFs (Higgins et al., 1999, p. 9). This study suggested as major problems: A lack of coordination between the bodies responsible for public research and those responsible for private research. A lack of coordination between universities and businesses.
CAN LESS FAVORED REGIONS CHANGE THEIR DESTINY?
189
In many regions there seems to be no coordination at all on science and technology policy, or between departments of industry and departments of education. In some regions, there is overlap and inadequate coordination between national and regional measures. There is little involvement of those involved in RTD in the region, particularly those in the private sector, in policy planning. An effort was made to investigate the relationship between institutional “thickness” (Amin, 1998), using as a proxy the number of R&D organizations and intermediaries active in a region and regional growth. Measuring institutional thickness is difficult and inevitably subjective. In an attempt to quantify institutional thickness, the three evaluators of the R&D and innovation efforts in the EU less favored regions ranked institutional thickness with values 1-3, to represent: estimated research done locally in 1999; the existence of higher education institutions, R&D organizations, and shared technology support facilities as a measure for institutional thickness; and the propensity to network among firms. The typology presented in Table 9.7 symbolizes correlations between institutional thickness and relative growth. Some differences in the performance with data from Table 9.3 are attributed to differences in the statistical sources and the periods studied. Conclusions from Table 9.7 are both interesting and controversial: the dataset used suggests low growth in two of the winning regions (Corsica and Molise) possibly due to differences in the data used (both time horizon and datasets). In order to confirm theory, the diagonal should be much more densely populated than the other boxes, which is only partially true. Actually the conclusions that can be drawn are: 1. Four of the winning regions are located in the high thickness, high growth category. Thus, to some extent, institutional thickness does offer an explanation to changing destinies. 2. Regions with medium or low institutional thickness may grow very rapidly, because they are either starting very low (Western Macedonia, Thessaly) or because they are endowed with important natural resources (mainly tourism orientation as in Algarve, Southern Aegean, Ionian Islands) 3. Regions that are institutionally thick may fall behind, although it is unlikely. When this is the case, it applies more to regions that are industrially declining than original less favored ones.
190
Lena J. Tsipouri
Overall this whole, exercise leaves important questions open and suggests that more factors should be investigated, such as interregional technology transfer (which is low for most parts of the European Union), entrepreneurship and leadership.
9.6 CONCLUSIONS The increasing regional development funding by the European Union in the last 15 years is a case of particular interest for regions trying to change their destiny. Looking at less-favored regions, trends suggest that from 1960 to the mid-1970s the prevailing process was convergence. From the mid-1970s, economic convergence by regions and countries comes to a halt, and from the mid-1980s to 1996 (when transnational resources increase), disparities existing at the global level are stabilizing; there is at the same time convergence and divergence. Countries and regions on a European scale converge, while regions within countries diverge. This leads to the assumption that when re-
CAN LESS FAVORED REGIONS CHANGE THEIR DESTINY?
191
gional aid increases, only regions able to manage this support well will converge. The significant downturn in economic growth after the year 2000 may trigger a particularly negative impact in the most vulnerable regions in the Union. Despite limitations in the data and political interference, some European LFRs seem to have been able to change their destiny in the last decade. However, despite ample funding, only few regions were persistently successful over the years, mainly the capital cities of the cohesion countries, the second most important regions, and some tourist areas. The explanations of success are difficult, and statistical methods did not lead to significant results. Physical capital transfer, market size, educational levels, and even GERD were not found critical. Soft factors, such as institutional thickness, and the willingness of the regional administration to change routines, are observed in most of the successful regions but they are only partial indicators of success and cannot be reliably and objectively measured. Besides, even when such factors are identified as crucially important, their political relevance depends on the extent to which they are subject to change, the extent to which social capital can be created top-down or the potential of technology intermediaries to stimulate demand.
NOTES 1.
The Committee of the Regions has a consultative rather than decision-making character.
2.
The European formal rules adopted and practiced in the past.
3.
European Regional Development Fund (ERDF), European Social Fund (ESF), FEOGA and the European Investment Bank.
4.
OECD statistics seem to be rather conservative. For comparison a detailed recent study by the Institute of Industrial Research suggests that the Greek informal sector represents 36.7 percent of GDP (excluding criminal activities, which would further increase the share) compared to 29 percent estimated by the OECD.
5.
This includes the very high transfer of physical resources from the German Bund, including the “Solidaritätsbeitrag” paid by the Western German employees, but also the institutional framework that allowed direct investment from the West.
6.
Partly only, because other low starters did not grow as rapidly.
7.
This analysis refers to the 15 member states and their regions in the year 2002. Things will change considerably in the future through the “eviction effect,” notably a
Lena J. Tsipouri
192
statistical lowering of the threshold of eligibility because of the accession of 10 new, poor member states. 8.
Madrid would have been the fourth to complete the picture but was not eligible for regional support, having already at the decision-making period reached a higher GDP per head than suggested by the threshold.
REFERENCES Abramovitz, M., The Search for the Sources of Economic Growth. Journal of Economic History 1993; 53:217-43. Amin, A., “An Institutionalist Perspective on Regional Economic Development,” Expert Document, OECD Conference on “New Challenges in Regional Policies,” Borås, Sweden, 28-29 January 1998. Amin, A., Thrift N., Globalization, Institutions, and Regional Development in Europe. Oxford: Oxford University Press, 1996. Ansell, C.K., Parsons C.A., Darden K.A., Dual Networks in European Regional Development Policy. Journal of Common Market Studies 1997; 35:375-403. Barro, R.J., Sala-i-Martin, X., Convergence Across States and Regions. Brookings Papers on Econ. Activity 1991; 1:107-182. Barro, R., Mankiw, G., Sala-I-Martin, X. Capital Mobility in Neoclassical Models of Growth. The American Economic Review 1995; 85:103-115. Baumol, W.J., Productivity Growth, Convergence and Welfare: What the Long Run Data Show. American Economic Review 1986; 76:1072-1085. Cambridge Econometrics, Regional Growth and Convergence, Europe, The Single Market Review. Luxembourg: Office for Official Publications of the European Communities, 1997. Chatterji, M., Convergence clubs and endogenous growth. Oxford Review of Economic Policy 1992; 8:57-69. Christodoulakis, N., Kalyvitis, S., The Second CSF (Delors Package II) for Greece and its Impact on the Greek Economy, An ex-ante assessment using a macroeconomic model. Economics of Planning 1998; 31:57-79. Cohen, W., Levinthal R., Innovation and Learning the two Faces of R&D. Economic Journal 1989; 99:569-96. Cuadrado-Roura, J.R., Regional Convergence in the European Union: From Hypothesis to the Actual Trends. The Annals of Regional Science 2001; 35:333-356. Cuadrado-Roura, J.R., Mancha, T., Garrido, R., Convergencia regional en Espana. Hechos, tendencias y perspectivas. F. Argentaria & Edit. Visor, Madrit, 1998.
CAN LESS FAVORED REGIONS CHANGE THEIR DESTINY?
193
Decressin, J., Fatas, A., Regional Labor Market Dynamics in Europe. European Economic Review 1995; 39:1627-1655. Dunford, M., Winners and Losers: The New Map of Economic Inequality in the European Union. European Urban and Regional Studies 1994; 1:95-114. European Commission, Second Progress Report on Economic and Social Cohesion, Unity, solidarity, diversity for Europe, its people and its territory, 2003. European Commission, Unity, solidarity, diversity for Europe, its people and its territory, Second Report on Economic and Social Cohesion. Luxembourg: Office for Official Publications of the European Communities, 2001a. European Commission, The Regional Dimension of the European Research Area (549 final), Brussels 2001b. Fagerberg, J., International Competitiveness. The Economic Journal 1988; 98:355-374. Fagerberg, J. Verspagen, B., Heading for Divergence. Regional Growth in Europe Reconsidered. Journal of Common Market Studies 1996; 34:431-448. Fagerberg, J., Verspagen, B., Caniels, M., Technology, Growth and Unemployment across European Regions. Regional Studies 1997; 31:457-466. Faini, R., Trade Unions and Regional Development. European Economic Review 1999; 43:457-474. Feldman, M.P., Florida, R., The Geographic Sources of Innovation: Technological Infrastructure and Product Innovation in the United States. Annals of the Association of American Geographers 1994; 84:210-229. Feldman, M.P., The Geography of Innovation. Boston: Kluwer Academic Publishers, 1994. Fingleton, B., Estimates of Time to Economic Convergence: An Analysis of Regions of the European Union. International Regional Science Review 1999; 22:5-34. Higgins, T., Tsipouri, L., van der Lande, R., Thematic Evaluation: Impact of Structural Funds (1994-1999) on Research, Technology Development and Innovation (RTDI) in Objective 1 and 6 Regions. European Commission, 1999. Krugmann, P., Geography and Trade. Cambridge: Mass MIT Press, 1991. Levine, R., Renelt, D., A Sensitivity Analysis of Cross-Country Growth Regressions. The American Economic Review 1992; 82:942-963. Magrini, S., The Evolution of Income Disparities among the Regions of the European Union. Regional Science and Urban Economics 1999; 29:257-281. Maurseth, B., Convergence, Geography and Technology. Structural Change and Economic Dynamics 2001; 12:247-276. Neven, D., Gouyette, C., Regional Convergence in the European Community. Journal of Common Market Studies 1995; 33:47-65.
194
Lena J. Tsipouri
Organisation for Economic Cooperation and Development, OECD Science, Technology, and Industry Outlook 2000. Paris, France, 2000. Oughton, C., Landabaso, M., Morgan, K., The Regional Innovation Paradox: Innovation Policy and Industrial Policy. Working Paper 00/04. Birkbeck - University of London: School of Management and Organizational Psychology, 2000. Putnam, R.D., The Prosperous Community. The American Prospect 1993; 13:36-42. Quah, D., Galton’s Fallacy and Tests of the Convergence Hypothesis. Scandinavian Journal of Economics 1993; 95:427-443. Quah, D., Empires for Economic Growth and Convergence. European Economic Review 1996a; 40:1353-1375. Quah, D., Twin Peaks: Growth and Convergence in Models of Distribution Dynamics. Working Paper n.280, Centre for Economic Performance, 1996b. Romer, P., Increasing returns and long run growth. Journal of Political Economy 1986; 94:1002-37. Sala-I-Martin, X., Regional Cohesion: Evidence and Theories of Regional Growth and Convergence. European Economic Review 1996a; 40:1325-1352. Sala-I-Martin, X., The Classical Approach to Convergence Analysis, Economic Journal 1996b; 106:1019-1036. Sala-I-Martin, X., Cross Sectional Regressions and the Empirics of Economic Growth, European Economic Review 1994; 38:739-47. Tsipouri, L., “Typologies and policy menus.” In Thematic Evaluation: Impact of Structural Funds (1994-1999) on Research, Technology Development and Innovation (RTDI) in Objective 1 and 6 Regions, Higgins, T., Tsipouri, L., van der Lande, R. eds. Dublin: report to the European Commission, 1999.
Chapter 10 INNOVATION CHALLENGES AND STRATEGIES IN CATCH-UP REGIONS Developmental Growth and Disparities in Georgia, USA
Philip Shapira
10.1 INTRODUCTION This chapter examines innovation challenges and strategies in catch-up regions, drawing on the experience of the state of Georgia in the southeastern United States. The chapter begins by discussing the broader context of regional convergence in the U.S. and recent trends towards divergence. Current U.S. development debates and models are briefly discussed. The situation of Georgia is then examined, with an assessment of the state’s developmental strategies.
10.2 TRENDS IN REGIONAL CONVERGENCE AND DIVERGENCE IN THE UNITED STATES Over the past eight decades, there has been a significant convergence of the economic fortunes of U.S. regions, at least as measured by per capita income. Previously high per capita income regions (Mideast, New England, Far West and Great Lakes) have converged towards the U.S. average. Conversely, previously low per capita income regions (Southeast, Southwest, Plains, and Rocky Mountain) have been raised towards the U.S. average (see Figure 10.1). Similarly, at the state level, there has been a broad decline in per capita income disparities between the states, as shown by the long-run declining standard deviation in state personal income per capita (Figure 10.2). For example, in 1950, the highest per capita income state – Alaska – had a per capita income almost three times that of Mississippi, the lowest. In 1999, per capita income in Connecticut, the highest per capita income state, was about 1.9 times that of Mississippi, which remained the lowest (Bernat, 2001).
196
Philip Shapira
Figure 10.1: Regional relative to U.S. per capita income, BEA regions, 1929-1999.
Figure 10.2: Dispersion in state per capita income in the U.S., annual standard deviation, 1929-1999. Source: BEA, State Personal Income, 1929-1999, U.S. Department of Commerce, 2000. Cited in Wisconsin Department of Revenue, http://www.dor.state.wi.us/ra/perc0201.html.
INNOVATION CHALLENGES AND STRATEGIES
197
Despite the consistency in the poorest state, there has in fact been considerable change in the geographical distribution of per capita income among the states. Five of the top 10 states were in the west in 1950, while the bottom 10 states were all in the Southeast. By 1999, states west of the Mississippi River comprise only two of the top 10 states, while in the Southeast four states (Georgia, North Carolina, South Carolina and Tennessee) had moved out of the bottom ten (Bernat, 2001). There are varied explanations for this long-run convergence in regional and state per capita incomes in the U.S. Neo-classical economists highlight the mobility of labor, capital, and technology in fostering long-run convergence towards equilibrium. From an evolutionary perspective, it has been argued that systematic learning and knowledge transfer has occurred, for example as states benchmark and replicate developmental strategies pioneered in other regions (Eisinger, 1988). Institutional and political theorists point to the role of public policy in reshaping regional economic fortunes in recent decades. This includes the contribution not only of explicit federal regional policies and inter-state regional groups (e.g. the Southern Growth Policies Board) but also more significant “implicit” federal regional policies such as the national highway system or the allocation of defense and federal R&D spending (Markusen et al., 1992) and the development of more inclusive, growth-oriented political leadership in previously lagging regions, especially in the “New South” (Allen, 1996). However, there are substantial limits to these arguments. In particular, it should be noted that most of the convergence in regional and state economic fortunes took place between the 1950s and 1979. Since then, convergence has ceased and new disparities have emerged. In the 1980s, a new round of defense spending may have disproportionately aided high-income regions, such as New England. In the 1990s, growth in new information and other high technology industries appears to have favored a subset of metropolitan areas in selected U.S. regions. In 1999, the 10 metropolitan areas with the highest levels of per capita income were located either on the east or west coast. Similarly, of the 10 fastest growing metropolitan regions by per capita income between 1997 and 1999, all had strong growth in new economy industries such as software development and advanced business services (Bureau of Economic Analysis, 2001). This group included San Jose (California) Austin (Texas) and Seattle (Washington). Over this period, per capita income gaps between urban and rural areas widened, with metropolitan per capita income averaging 1.4 times that of non-metropolitan areas in the U.S. Within metropolitan areas, central city and suburban disparities (often demarcated along racial and ethnic lines) in income, employment and opportunity continue to be of concern.
198
Philip Shapira
10.3 CURRENT U.S. REGIONAL ECONOMIC DEVELOPMENT POLICY DEBATES AND STRATEGIES Although extreme regional and state differences have been reduced, the cessation of convergence and the rise of new regional disparities have given rise to renewed analytical and policy debate about regional economic development in the U.S. For example, new growth theorists suggest that information, knowledge and human capital flow most readily among firms that are proximate rather than far apart (Krugman, 1991); this regional “stickiness” will continue to foster regional divergence in economic fortunes. Such arguments have been used to explain the development of high technology and other spatially concentrated industrial agglomerations in the United States. Similarly, social capital and industrial organization perspectives have been employed to account not only for the emergence of high technology clusters, but also for differences in performance between clusters. An example: Saxenian’s (1996) comparison of the success of the more flexibly-organized industries, innovative institutions and networked relationships of Silicon Valley in California compared with the vertically-structured and apparently less successful industries of the Route 128 complex focused around Boston, Massachusetts. Conversely, path dependency arguments – focusing on the cumulative effects of regional history and limitations in institutions – have been reengaged to suggest why lagging regions continue to lag (Salstron, 1997). These conceptual insights have, in turn, prompted a series of revised and new regional development strategies in the U.S. States, local governments, and institutions such as state universities and experiment stations have long sought to promote regional technology transfer or to develop new technology parks (one of the most notable being North Carolina’s Research Triangle, begun in the mid-1950s). More recently, states have considerably increased their investments in regional innovation policies, including applied university research centers, industry-university partnerships, technology incubators, the commercialization of research, grants, venture capital, financial awards, and other programs using research and technology for economic development (Coburn and Berglund, 1995). By the mid-1990s, states were spending at least $2.7 billion of their own funds (excluding federal and industry dollars) on applied research and technology programs, mostly with regional economic development aims (Berglund, 1998). Although there is much variation in these strategies and many important nuances of implementation by local agencies, there are some broad characteristics of note. First, current strategies tend to emphasize innovation and technology-based approaches to regional economic development. Such strategies
INNOVATION CHALLENGES AND STRATEGIES
199
tend to emphasize regional innovation systems and the development of local innovation demand and capabilities, rather than the technology-push methods of previous U.S. technology transfer programs. Second, strategies are highly decentralized, often with state and local organizations taking leadership roles. The federal role is typically one of offering frameworks and matching funding. In certain cases, federal agencies will provide additional funds to build capacity in states that severely lag in science and technology.1 Third, implementation through collaborative arrangements is common, for example through public-private partnerships or regional networks. A “layered” approach is typically employed, involving multiple organizations but often with lead institutions (such as a university or non-profit organization). In many cases, state regional innovation strategies are added to more traditional approaches of infrastructure development and industrial attraction (with these traditional approaches updated through such efforts as building high-capacity data networks and marketing to advanced service or cluster-supporting industries). A fourth characteristic of recent regional innovation strategies is what might be called the “rediscovery of the foundation.” To date, states have focused a great deal of effort and resources on developing new high-technology industries in their regions. This continues to be their greatest emphasis, albeit with some variations in the technologies of choice (for example, rather than semiconductors or computers, today the targets include biotechnology, ecommerce, nanotechnology, and various fusion technologies such as bioinformatics, bioengineering, or communication interfaces). However, this emphasis on the latest technologies notwithstanding, states have also increasingly recognized the importance of existing, mature industrial bases, often comprised of many small and medium-sized firms in industries such as carpets, food processing, plastics, metals, machining and engineering. To stimulate these industries to be more efficient, effective, and innovative, states (in concert with the federal government) have established industrial modernization programs. These programs, mostly organized under the umbrella of the U.S. Manufacturing Extension Partnership (MEP), engage public and private technology and business service providers to upgrade existing manufacturing enterprises. Nearly 70 MEP centers with more than 400 offices and 3,100 affiliates and serving about 30,000 firms a year are now in operation in all 50 U.S. states (Shapira, 1998; National Institute of Standards and Technology, 2002). A fifth characteristic is that, arguably, current state strategies seek to be more systemic than earlier efforts, sometimes known as a shift from retailing to wholesaling state innovation services. For example, whereas previously individual freestanding programs were developed in the states to promote high technology enterprises, today there is greater emphasis on the strategic
200
Philip Shapira
planning and development of regional innovation capabilities. For example, in 2002, a business-led New Economy Transition Team in Northern Kentucky developed a detailed strategy to position the region as a center for life sciences and information technology, advanced manufacturing and financial services. Initiatives are planned to create a new public-private organization to invest in companies, products, technologies, and research commercialization in pharmaceuticals; to establish a technology and commercialization corridor and increase technology training; to develop advanced manufacturing and recruit information management intensive companies; and to develop seed, angel and venture capital funding for the region. The Northern Kentucky initiative will form part of a statewide strategy for the new economy, comprised of multiple regional strategies developed through bottom-up planning. Throughout the U.S., comparable strategies to coordinate new research, technology, and economic development initiatives are being put into place.2 Similarly, there has been a substantial growth in region-wide initiatives to foster industrial clusters. Indeed, promoting regional industry clusters has emerged as one of the most prevalent concepts in local and regional development in the U.S. Debates about new growth theory, networks and regional social capital, along with comparisons with European industrial districts, have boosted cluster-led initiatives. States that have developed industry cluster concepts include Arizona, California, Connecticut, Florida, Minnesota, North Carolina, Ohio, Oregon, and Washington, while hundreds of U.S. cities are reported to have developed cluster strategies (Bergman and Feser, 1999). Typically, clusters involve planning and analysis of existing and potential industrial linkages, assessments of hard and soft regional assets, strategy development, investments in key institutions and programs, social capital investment through the thickening of associations, and cluster promotion and marketing. Whereas in the past, many regional innovation efforts sought to mirror well-publicized successes (for example, Silicon Valley), today there seems to be a greater emphasis on “distinctiveness” – identifying and developing indigenous elements that will allow a regional cluster to assume a more distinctive (and up-market) competitive posture. A final feature of current U.S. regional innovation measures is an emphasis on learning, benchmarking, and discursive comparison. One of the features of federalism in the U.S. is the role that states play as “laboratories” for new policies and programs (Osborne, 1988). Policy and fiscal autonomy allows states and localities to pioneer new programs and concepts. While many fail, those efforts that look successful are rapidly borrowed by other states and often by the federal government itself. While this iterative experimentation and learning process is not new, it has been given new impetus by recent efforts to promote benchmarking and performance measurement, thereby providing a greater (albeit not perfect) analytical basis for compari-
INNOVATION CHALLENGES AND STRATEGIES
201
son. Current and new organizations, such as the National Governor’s Association, the National Association of State Land Grant Universities, the Southern Technology Council, the Modernization Forum, and the State Science and Technology Institute, play critical roles in mediating and promoting exchange and processes of discursive comparison.
10.4 APPLICATION OF THE DEVELOPMENT MODEL: THE CASE OF GEORGIA To probe in detail the extent to which new U.S. economic development models are being applied and the effects of such models, the chapter now turns to the case of a U.S. state – the southeastern state of Georgia (see Figure 10.3). The state of Georgia has undergone a substantial long-term transformation of its developmental position. Following the Civil War in the midnineteenth century, the state was a poor agricultural region, not so fundamentally reconstructed from the pre-war plantation slave-based economy, with little industry and many of its towns in ruins (as indelibly portrayed through the words and images of Margaret Mitchell’s “Gone with the Wind”). Yet, beginning around the same period, business (led by such figures as Henry Grady), political and (at times) community movements got under way to modernize the state through technology and the development of industry, using the agency of government to promote change (McMath, 1991; Combes and Todd, 1995). These movements, albeit with turns and changes, have continued through to the present, as the state has moved through successive phases of infrastructure development, social modernization (especially in the fight for civil rights and the removal of racial barriers in the 1960s), and now to the promotion of innovation. In recent years, Georgia has aggressively pursued most aspects of the new development model. The state has been one of the fastest growing in the U.S. in terms of population growth and can now point to Atlanta (just a small railway junction in the nineteenth century) as an internationally-recognized city and business head-office location (Porter, 2002). Yet, despite significant success, the state still suffers from social, economic, and spatial inequities in part inherited from past patterns of development and, arguably, in part a function of how the new development models have been applied in the state. These trends and development issues are discussed in the following sections of the chapter.
202
Philip Shapira
Figure 10.3: State of Georgia: Major cities
10.4.1 Context Georgia is the tenth largest U.S. states in terms of population and in recent years has been one of the fastest growing states in the nation in population and employment growth. Although state per capita income is still slightly below the national average, over the past half century Georgia (along with much of the larger U.S. southeast) has gone through a significant era of catch-up. A benchmark with New England (and the state of Massachusetts) is particularly interesting, given that in the early twentieth century Georgia competed aggressively with this northern region on the basis of lower wages so as to attract textile and other manufacturing. In 1929, Georgia’s per-capita income was just 38 percent of that in Massachusetts, but by 1978 it had risen to 82 percent. However, more recently, although Georgia has continued to move slowly towards the U.S. average, the state was unable to match a renewed growth spurt by Massachusetts, such that Georgia’s per-capita income
INNOVATION CHALLENGES AND STRATEGIES
203
Figure 10.4: Per capita income trends – Georgia and other reference areas, 1929-1999. Source: Analysis of data from U.S. Bureau of Economic Analysis.
had dropped to 74 percent of that in Massachusetts by the year 2000. (See Figure 10.4.) There are perhaps two parallel stories here, the second of which being one that this chapter will explore more deeply. The first story is, of course, how Massachusetts recovered in the 1980s and maintained its per capita income position in the 1990s. This story is not examined here, although others have written about it.3 The second story is about what happened to Georgia. After growing and catching up for many decades with a developmental model of industrial attraction and modernization, has Georgia more recently become constricted by the legacy of this model? Or, has the state – through massive investment in new innovation-based development models – laid the seeds for significant transformation of its economic base towards a higher value-added research and knowledge-based economy? Today, roughly 17 percent of the total employment base in Georgia is in manufacturing (of the rest, 26 percent is in services and 25 percent in wholesale and retail trade). The manufacturing sector has grown steadily over recent decades in employment and establishments (which now total about 12,000, most of which are small and medium-sized enterprises with have fewer than 500 employees). Much of Georgia’s manufacturing has involved traditional resource-intensive sectors such as pulp and paper and food processing, or modern but routine branch plants. Throughout the state’s manufacturing sector, the effects of the historic domination of industrial attraction
204
Philip Shapira
strategies based on low costs can be seen. For example, the 1999 Georgia Manufacturing Survey suggests that Georgia manufacturers use rather standard technology strategies or low-cost strategies to compete in the market for customer sales. The 1999 survey asked manufacturers to rank six strategies from 1 (highest importance) to 6 (lowest importance): low price, high quality, innovation/new technology, quick delivery, adapting product to customer needs, value-added customer and product services. Nearly half of the manufacturers compete primarily through standard quality-related technologies and techniques. Only 8 percent of Georgia firms report competing primarily through innovation or new technology. The bubble chart (Figure 10.5) shows that manufacturers competing primarily through innovation strategies have both higher returns on sales and higher employee wages. However, most Georgia manufacturers use strategies associated with low wages and lower returns on sales. More broadly, industrial research and development (R&D) spending remains low, Georgia ranks in the bottom half of states in patent generation, and public R&D spending has been dominated by defense procurement. For example, the Rand Corporation estimates that the U.S. Department of Defense accounts for nearly 90 percent of federal R&D spent in Georgia.
Figure 10.5: Corporate strategies and associated returns, Georgia manufacturers 1999. Source: Georgia Manufacturing Survey, 1999. N=727.
INNOVATION CHALLENGES AND STRATEGIES
205
10.4.2 State Innovation Strategy For more than one hundred years, Georgia has pursued state development projects aimed at using technology to promote industrial and economic development. The earliest efforts included the founding of an engineering institution (Georgia Tech) in 1885 and the subsequent development of a state university system and technical colleges. In the 1960s, the state established an industrial extension service to transfer technology and know-how to existing industry through field offices of Georgia Tech. One of the first technology incubators in the U.S. – the Advanced Technology Development Center – was established by Georgia Tech in 1980 in Atlanta. Efforts in the 1980s and 1990s to strengthen the state’s research and scientific base were embodied in the establishment of the Georgia Research Alliance (GRA), a private non-profit organization that coordinates university research efforts to generate technology-based economic development (Combes and Todd, 1995; Lambright, 2000). The efforts evolved into a targeted technological development strategy focused on chip and broadband areas (Yamacraw). The state invested roughly $40 million in fiscal year 2000 in these initiatives, with nearly all of this investment going to GRA and Yamacraw. These programs have been supported by physical and human capital initiatives. Public and private investments have been made in high-speed telecommunications (e.g., the Georgia Statewide Academic and Medical System). The state has invested in intellectual capital through the HOPE scholarship program, QuickStart training of technical and service-level workers through the Georgia Department of Technical and Adult Education, and the University System’s Intellectual Capital Partnership Program (ICAPP), which provides expedited educational programs for companies with university-level workforce needs. No single body manages or coordinates these technology development activities. Some key programs are managed by units of Georgia Tech (e.g., the Economic Development Institute (EDI)) and the GRA, while other state agencies, educational institutions, and non-profit organizations manage others. Partnerships and multiple interlocking relationships enable these organizations to work with other state and local organizations (as there is considerable private-sector technology-based association activity in the Atlanta area). Private-sector technology associations such as the umbrella Technology Association of Georgia have offices on the Georgia Tech campus. GRA board members and Georgia Tech administrators also serve on the boards of directors of state and local economic development organizations. Interlocking boards are part of the state’s strategy to establish complementary relationships among Georgia’s technology development organizations programs.
206
Philip Shapira
10.4.3 Innovation Programs There are many publicly-sponsored innovation programs in Georgia. Indeed, a recent study uncovered more than 100 programs with a technology outreach component (Youtie et al., 2000). An analysis of these programs indicates the following profile (see also Table 10.1): Most technology-related programs are located in Atlanta or its suburbs. Less than 30 percent can be found outside the Atlanta Metropolitan Statistical Area. Seventy-five percent were established in the 1990s, with the median year of establishment being 1994. The majority are run by academic institutions. The most common capabilities these programs emphasize are business association, training, expert assistance, applied research and financing. Some programs support inter-firm collaboration assistance, while a small number aim at promoting technology leadership. To illustrate the range and depth of Georgia’s larger innovation-related programs, a summary of four key programs follows: Georgia Manufacturing Extension Program Georgia was one of the first states in the country, along with North Carolina, to set up a state-funded industrial extension program. The Georgia General Assembly created the program at Georgia Tech in 1960, and the first field office opened in Rome, Georgia, in 1961. The program, located in Georgia Tech’s Economic Development Institute (EDI), currently operates 17 field offices concentrated in rural Georgia and has partnerships with regional economic development groups, state labor offices, private-sector consultants, trade associations, and other contacts. Its statewide service activities include various assessments, feasibility analyses, problem-solving guidance, direct technical assistance, referrals and resource matching, implementation assistance, information dissemination, training workshops and seminars, and demonstrations. The field offices provide a gateway to EDI’s more specialized services such as information technology, ISO 9000/quality, lean manufacturing, industrial marketing, environmental technology, and specialized sponsor programs such as Trade Adjustment Assistance (U.S. Department of Commerce). In addition, EDI calls on researchers and faculty at Georgia Tech and regional national laboratories (e.g., Oak Ridge National Laboratory, NASA) with which Georgia Tech has formal agreements to serve as the designated technology transfer agent for the state. Full-time personnel with engineering,
INNOVATION CHALLENGES AND STRATEGIES
207
technology transfer, management, economic development, and industrial experience staff EDI’s regional offices and technical centers. In 1994, Georgia joined the national Manufacturing Extension Partnership (MEP), administered by the National Institute of Standards and Technology (NIST) of the U.S. Department of Commerce. Georgia now operates an $8.5 million manufacturing extension program (about $3 million of which
208
Philip Shapira
comes from the state). Today, the Georgia program typically serves 700 to 900 manufacturers a year, which puts the program in the top quintile of all MEP programs in the U.S. Customer survey information since 1994 indicates that more than 80 percent of Georgia MEP clients take action as a result of the assistance. The most common benefits are in owner and employee skills. Evaluation studies have found that Georgia MEP clients tended to have higher productivity growth than did non-clients (Shapira and Youtie, 1998). Advanced Technology Development Center (ATDC) Established in 1980, ATDC was one of the nation’s first technology incubators. It offers entrepreneurial services including space, guidance, and support for early-stage new technology companies. ATDC also supports corporate R&D units through a “landing party” program. The Faculty Commercialization Program (small grants to support faculty research with commercial potential) is also run by ATDC. In 1997, ATDC started Netcelerate, a virtual incubator that provides Web-based membership, electronically provided services, mentoring, and linkages with venture capital sources to some 200 entrepreneurial members. ATDC is a unit of Georgia Tech/EDI. It has nine staff associates and receives about $1.5 million a year in state funding. Atlanta has been the main focus of ATDC activities (with incubators at Georgia Tech and in alliance with Emory University). But there have been efforts to establish incubators in other areas. An incubator at Warner Robins has been in operation for several years, and there are now new initiatives to create incubators in Savannah and Augusta. Several external studies of ATDC’s impacts have been conducted (Culp and Shapira, 1997; Phillips, 2003). However, ATDC mostly relies on client surveys to gather impact information. Key measures are amount of venture capital raised, sales generated, and number of jobs created. Since 1986, more than 120 firms have been admitted. Fifty member firms have raised $38 million in venture capital and employ 370 people. As of 1998, ATDC graduates generated a reported $350 million in sales and more than 4,000 jobs. Georgia Research Alliance The Georgia Research Alliance (GRA) was formed in 1990 as a collaborative research initiative among six major research universities in Georgia (Lambright, 2000). GRA was charged to invest in building the state’s research infrastructure in targeted areas. The investments were designed to generate economic development results—new company start-ups as well as hightechnology firm relocations and retention of existing industry.
INNOVATION CHALLENGES AND STRATEGIES
209
GRA has several programmatic elements. Eminent scholars, of whom there are 32 as of April 2000, are recruited in targeted areas based in part on a GRA supplementary endowment to be used for facilities, equipment, and other non-salary expenses. GRA also invests in facilities and specialized equipment such as the 150,000-square-foot Georgia Center for Advanced Telecommunications Technology (GCATT) building. GRA’s Technology Development Partnership program funds industry-university collaborative research with significant commercial potential. A private non-profit organization consisting of a small staff (president, two program managers, and administrative support) forms GRA management. This group functions as a virtual holding company for the program, develops strategy, finds financial resources, and interacts with state and local economic development organizations. GRA has a board of trustees composed of presidents of the member universities and executives from technology and other businesses who have served in state and local leadership positions. Besides its direct programmatic elements, GRA has made investments in key aspects of Georgia’s technology development infrastructure. In 1994, GRA encouraged the formation of Alliance Technology Ventures, the first public initiative to establish an early-stage venture capital fund in the state. GRA has supported investments to expand ATDC to convert GRA’s research investments at member universities into commercial applications. GRA employs indicator monitoring and benchmarking to assess program impacts. The program uses three main sources for monitoring and benchmarking: the National Science Foundation’s Academic Institutional Profiles, the State Science and Technology Institute’s state-by-state program descriptions, and PricewaterhouseCoopers Money Tree Surveys. In addition, the program has gathered some 50 research and commercialization success stories attributed to its support. Through to fiscal year 2000, the state of Georgia invested $276 million via the GRA in research and development programs at its six member universities, matched by $65 million in private funds. In turn, GRA reports that this investment has helped to attract over $600 million in additional sponsored research. Yamacraw Initiative Yamacraw is a public-private partnership in broadband, digital signal processing, and systems-on-a-chip technologies. The venture includes (1) marketing and promotion of a high-bandwidth communications business cluster by the Georgia Department of Industry, Trade, and Tourism, (2) expansion and retention of the pool of qualified electrical engineers and scientists in the state by the University System of Georgia, and (3) a GRA/GCATTestablished electronic design center and research consortium.
210
Philip Shapira
Yamacraw received $14 million for fiscal year 2000. Specific goals over a five-to-seven-year period include (1) creation of up to 2,000 electronic design and software engineer jobs, (2) attraction of four times the venture capital currently available to companies, and (3) establishment of 10 highbandwidth communications companies/divisions in Georgia. Assessment and Outcomes Georgia’s investments suggest that the state places substantially more emphasis on research infrastructure investment rather than on technology transfer/infusion activities. Of the state’s technology budget, only about 10 percent goes directly to technology transfer activities, with the remainder allocated to research (Youtie, Shapira, and Mohapatra, 2000). The most concentrated, formal set of technology transfer programs is at EDI - the manufacturing extension program and the ATDC. Georgia’s technology development and transfer programs are for the most part located in or near Atlanta. Universities, particularly Georgia Tech, are the dominant organizing mechanism for technology development and transfer. The GRA also plays an important role, although this private nonprofit organization again works primarily through universities to encourage collaborative research initiatives. Examining the effect of these programs by linking investments to technology and economic development outcome indicators is difficult. A mixed picture of good performance in some areas but lags in others is presented by measures such as venture capital attraction, patent output, the percentage of employees working in technology-based jobs, or levels of wages and educational attainment (see Table 10.2). The amount of venture capital in Georgia has grown such that the state ranks among the top in the nation. A PricewaterhouseCoopers MoneyTree Survey placed Georgia among the top five states in terms of venture-backed investments. The American Electronics Association has ranked Georgia among the five fastest growing states in high-tech employment growth. The state also ranks relatively high for new spin-offs. Three research universities, (Georgia Tech, University of Georgia, and Emory University) are ranked among the top 50 in total R&D expenditures in the U.S. Public and private technology-based incubator programs and services have supported an expanding number of start-ups (mostly in the Atlanta area). Other indicators are not as strong, however. Georgia ranks in the middle or near the bottom of states in terms of patent output per capita, percentage of employees working in technology-based jobs, and private research and development. Perhaps the greatest success of Georgia’s state economic development efforts has been the growth in and around metro Atlanta of a significant high
INNOVATION CHALLENGES AND STRATEGIES
211
technology research and business complex – all the more remarkable bearing in mind that this was a location that two-to-three decades ago had very little innovation capability or employment. Yet, there is another less successful side of the picture. The Corporation for Enterprise Development (2003) in its Development Report Card noted well below average performance in human resource development for Georgia. K-12 public school systems in many parts of the state continue to perform weakly, with low high school graduation rates (Table 10.2). There is a relatively high proportion of working poor households and poor performance on a series of quality of life rankings. Other reports have noted the emergence of “multiple Georgias” – going beyond the contrast between the “two Georgia’s” of booming metropolitan Atlanta and the stagnant rest of the state to the “five Georgias” identified by the state Rural Development Council. Based on an economic vitality index, these five Georgias comprised 8 rapidly developing counties (19 percent of the state’s population), 42 developing counties (49 percent of the state’s population), 48 existing and emerging growth center counties (23 percent of the state’s population), 43 lagging rural counties (8 percent of the state’s
212
Philip Shapira
population), and 8 declining rural counties (0.5 percent of the state’s population) (Georgia Rural Development Council, 2000).
10.4.4 Unequal Growth and Sub-Regional Disparities Notwithstanding precisely how the sub-regions of the state are divided, the clear point is that a structural divergence in growth paths has emerged in Georgia. The Atlanta metropolitan region has clearly changed its position, adding more than 2.1 million new residents from 1969 through 1999, to reach its current size of about 3.9 million people (or 50 percent of the state’s population). In many ways, Atlanta has built its growth on the attraction and immigration of human capital and skills from throughout the United States and internationally. Every expectation is that metropolitan Atlanta will continue to grow, raising significant concerns about transportation congestion, environmental problems, and urban management. Much of Atlanta’s growth has occurred in the northern and ever expanding peripheral rings of the city, with much less development in older black neighborhoods (although parts of the central city are now undergoing revitalization and gentrification). While significant problems of poverty remain in Atlanta and many residents have not seen improvements in income, the overall metropolitan-wide picture is one of growth. However, interestingly, the story outside of Atlanta is not one primarily of decline. Seven medium-sized Georgia cities grew by 33 percent in population between 1969 and 1999 and now house a total of 2 million people (or about 26 percent of the state’s population).4 Each of these medium-sized cities added people. In aggregate, the state’s rural areas have also added population. In short, the challenge facing Georgia is not so much population (or even employment) decline, but one of poor quality of growth and low quality employment in many parts of the state. This challenge is illustrated by disaggregating per capita income trends within the state (Figure 10.6). The Atlanta metropolitan area held steady at the national average per capita income between 1969 and 1981, then grew sharply in the 1980s and 1990s, reaching a per capita income of 114 percent of the U.S. average in 1999. Conversely, mid-sized Georgia metropolitan cities saw relative per capita income growth in the 1970s and early 1980s, but since 1984 their per capita income level has been stagnant when compared with the U.S., and remains at 85 percent of the national average.5 The state’s non-metropolitan areas saw little change in their relative per capita position in the 1970s and 1980s, some relative improvement during the economic boom of the early and mid 1990s, with a tail-off more recently to 74 percent of national per capita income. Of course, these averages mask considerable variations, from the wealthiest county in Georgia (Fulton County in the
INNOVATION CHALLENGES AND STRATEGIES
213
Figure 10.6: Georgia per capita income trends, 1969-1999. Source: Analysis of data from U.S. Bureau of Economic Analysis.
Atlanta metro region) with a 1999 per capita income of $45.5 thousand to $13.8 thousand in Long County in the southeastern part of the state. The recognition of these regional inequities has prompted a parallel set of developmental measures in Georgia. Again, there are many programs, but to simplify these efforts tend to focus at two targets. The first target comprises the poorest, mostly rural areas of the state, where efforts are geared towards upgrading infrastructure, education, and social services and to attracting incoming employers. There is little capability in many of these areas to sustain innovation. The second target comprises the mid-sized cities of the state, where it could be said that modernization has already occurred, there are branches of major international companies, university and college infrastructures exist, and where the quality of development needs to be upgraded. Here, in most of these locations, there is some capability for innovationbased development, although it is often weak. Among the programs that focus on these mid-size metro and rural target regions are the following: Regional Advisory Councils /Regional Development Councils These are the traditional mechanisms of regional economic development coordination used in the state. Currently, Georgia is structured into twelve multi-county regions. Each region has an advisory council of public and private sector representatives. These councils advise the State in the development and implementation of regionally significant community and economic
214
Philip Shapira
development initiatives. Appointments are coordinated by state development agencies. The councils have long been conduits to channel federal economic development funding. Recently, the functions of these councils have expanded. They now include coordinating strategic economic development planning; maximizing the effectiveness of development resources through the collaboration and advice of local private and public sector leaders; promoting public and private sector partnerships at the regional level; seeking local input; identifying regional solutions for natural resource and other management issues; and aiding regional leadership programs, workforce development strategies, downtown revitalization, rural development initiatives, and existing industry support programs. Rural Development Council The Rural Development Council was established by the Governor in 1999 to serve as an advocacy and leadership organization to reduce the inequities facing many rural areas of the state. The Council has a membership appointed by the Governor from rural communities, business, and state agencies. Its responsibilities include advocacy for rural Georgia, the promotion of economic diversification, the avoidance of “one size fits all” strategies, promotion of equal opportunity, fostering of innovative solutions and economic, social, and environmental balance, support for local initiative, and promotion of collaboration and cooperation. One of the most useful functions of the Council has been to prepare a series of information and benchmarking studies that highlight spatial contrasts and inequities in the state and argue for stronger state vision, coordination, and investment to aid lagging rural areas. OneGeorgia Authority This is the largest and most ambitious “flag-ship” initiative to reduce regional, economic, and social inequalities in Georgia. OneGeorgia was created by the Governor and state legislature in 2001 to assist the state’s “most economically challenged areas.”6 The OneGeorgia Authority will use one-third of Georgia’s tobacco settlement funds. Over $60 million per year is anticipated for the Authority, with total funding of $1.6 billion projected over 25 years. To allocate these resources, a 4-tier ranking of Georgia counties has been developed, based on average rates of unemployment, per capita income, and percentage of residents in poverty. OneGeorgia will focus its efforts on the disadvantaged counties in Tier 1 (72 counties) and Tier 2 (34 counties), with some projects eligible in Tier 3 (35 counties) and none in Tier 4 ( 1 8 counties). This scheme includes economically lagging rural areas and the central districts of most medium-sized Georgia cities. At present, two types
INNOVATION CHALLENGES AND STRATEGIES
215
of project funds are available. An Equity fund provides matching awards and grants to eligible communities to help build the necessary infrastructure for economic development. The Equity fund is broad and can be used for a variety of activities to help eligible communities to assist economic development preparation. Eligible projects include traditional projects such as water and sewer projects, road, rail, and airport improvements, and industrial parks as well as workforce development projects, technology development or tourism development proposals. A second fund, known as the EDGE (Economic Development, Growth and Expansion) fund is used to assist Georgia communities compete for a business with another community from outside the state. It is of course too early to assess OneGeorgia, and it is likely that this ambitious program will evolve. Interesting, while the mission statement of OneGeorgia emphasizes the values of collaboration, innovation, cooperation, adaptation, and regionalism, the major funding programs are rather traditional in orientation (infrastructure, business development, and business attraction), although the organization does have the flexibility to accept a variety of project applications. Greater Georgia At the other end of the funding scale (i.e. with very limited funding) is Greater Georgia, an initiative to develop cluster strategies in mid-sized cities in the middle part of Georgia.7 The Greater Georgia project aims to use a regional industry cluster focus to enhance local organizational, leadership, planning, and project management capabilities for technology-based economic development; strengthen strategies and opportunities to grow and attract technology-based enterprises; and ensure that participating cities are deploying world-class best practice approaches to fostering technology-based development. Four cities are currently involved in the project: Augusta, Columbus, Macon, and Savannah. The project is implemented by the Economic Development Institute of Georgia Institute of Technology (Georgia Tech) in conjunction with the Georgia Tech School of Public Policy. The Georgia Department of Industry, Trade, and Tourism is the project sponsor. The development of International Partnerships is an integral element of the Greate Georgia program. International benchmarking is being undertaken to provide matched city comparisons and to establish linkages for technologydevelopment capacity building. Four European cities are being invited to participate as International Partners: Cork, (Ireland), Dundee (Scotland), Heilbronn (Germany), and Pisa (Italy). These European cities share similar characteristics with their Georgia counterparts, including similarities in population size, relationships to larger metropolitan areas, industrial structure,
216
Philip Shapira
higher educational facilities, and technology strategies. In addition to benchmarking and sharing of information, representatives from International Partner cities will be invited to participate in exchange activities with Georgia cities. This will provide opportunities to learn about each other’s technologybased economic development strategies, to discuss issues, and to consider opportunities for enhancement and further collaboration. Greater Georgia began activities in Summer 2001 and has continued through to 2004. While small, Greater Georgia is an explicit effort to strengthen the use of benchmarking, reflective, and learning strategies to Georgia’s medium sized cities – to try to move them beyond traditional business attraction strategies and competition with one another. If successful, Greater Georgia should result in strengthened capabilities and leadership in the four cities and a collaborative and internationally-aware approach to identifying new innovation-based opportunities.8 Among the promising outcomes of the project has been a commitment by the Governor to establish innovation centers in midsize cities which can serve as nuclei for new public and private applied research, development, and commercialization initiatives in targeted technologies. Assessment The Georgia case confirms that the changing of regional development trajectories is not an easy task to accomplish: where there is progress, it takes considerable time and effort, and a sustained public role. Moreover, the benefits are not necessarily equally shared. The state’s innovation and technology developmental initiatives have had discernable and positive effects in boosting research, innovation, and high technology employment in Georgia. But the benefits from these policies have been secured by a segment of the population, with strong benefits going to talented scientists and engineers, research universities, high technology enterprises, and high technology employees (whose wages are significantly above the Georgia average) particularly in the Atlanta region and including a large proportion of new-comers. Black residents in Atlanta and all social groups in regions outside of the Atlanta metropolis have gained less and “trickle down” impact has been limited: indeed the success of Atlanta has made more visible the static or lagging economic conditions found elsewhere. Yet, it is notable that the state government has responded to these inequities and developed a parallel set of policy mechanisms to address regional disparity. Are the state’s efforts to foster greater regional equity in the state likely to be effective? Time will tell, but already some serious issues can be raised. In part, Georgia’s efforts meet earlier calls for “dual agenda” state programs, to broaden the scope and targets of innovation policies (Bozeman, 2000) and
INNOVATION CHALLENGES AND STRATEGIES
217
the state has gone further than many in developing compensating programs. However, while the state’s innovation programs seek very much to “break the mold” and establish Georgia as a global location for leading-edge technology and business development, the major parallel programs for disadvantaged regions (excluding small efforts such as Greater Georgia) are very much consistent with traditional strategies of infrastructure development and business attraction. Even if successful, such strategies may reinforce the modern but routinized nature of work in these regions, but not significantly raise the value-added quality of development (and associated living standards). On the other hand, it can reasonably be argued that, due to an accumulation of historical and institutional factors, the basis for alternate development strategies remains lacking in much of peripheral Georgia. While that may be true, the most hopeful exception to this could be in the state’s medium-sized cities where there is some foundation, albeit at present weakly developed and supported, to leverage innovation assets and capabilities for higher valueadded growth.
10.5 CONCLUDING POINTS Georgia has adopted many of the features of the new U.S. regional economic development models discussed earlier in this chapter. The state has certainly focused on innovation and technology in its approach to economic development. There is a strong emphasis on implementation through public-private partnerships. There is some attention to the technological and business foundation of small and mid-sized enterprises, particularly through the state’s extensive industrial extension system. Benchmarking and comparison has also been prevalent. At the same time, other features of the new models have been more faintly applied. While the state has emphasized innovation and technology, this has tended to be on a sector basis (i.e. information technology or life sciences) anchored by major research universities, rather than working more broadly on the strengthening of regional innovation systems (especially outside of Atlanta). Moreover, while there is attention to technology transfer for existing and mature small and medium-sized firms, the level of investment is far lower than that accorded to high technology research and new venture development. It can also be added that while the state has promoted publicprivate partnerships, often the most successful technology-based efforts have been those where the state has made a sustained and long-term public commitment (as for example with the development of its research universities). Given Georgia’s historical context of a weakly developed private innovation infrastructure, the ability of the state to pursue sustained goals is a strength,
Philip Shapira
218
although at times the goals are perhaps narrowly constructed (for example, again for historical and constitutional reasons, the state has not been able to take such a direct role in the strengthening of K-12 public education systems where control is largely in the hands of local school districts). Finally, while there has been benchmarking, this has rarely been extended to explicit evaluation of state innovation and technology programs. What insights are there from the Georgia case with regard to the debate about path dependency versus regional breakthrough? One hates to have to say this, but there is a fair measure of “it depends” in assessing the state’s position on this axis. Viewed from a long-term perspective, relative to where the region started, Georgia has very substantially changed its regional position, aided by state (as well as federal policies not yet fully discussed in this chapter). On the other hand, with some exceptions (e.g. research services, information technology, multimedia, soft drinks, transport and logistics), the state is still not a recognized innovation leader. Yet, it could fairly be argued that the Atlanta metropolitan region has broken through the urban hierarchy and has emerged as a regional, national, and international center. In understanding where there has been significant change, it has frequently required the creation of new institutions and organizations, not necessarily strongly connected with the old (for example, in the development of new technology research centers not connected with old-line industries or in the rise of CNN in Atlanta which pioneered fresh approaches to the broadcasting and marketing of news). Additionally, new human capital resources have been drawn into the state (more than 3 million people, leading to almost a doubling of the state’s population over three decades), which has been a vital element in bringing in new skills and ideas. Unfortunately, many existing communities, both in Atlanta and outside, have been constrained by their own past developmental experiences and inequities and have not fully participated in the state’s growth. In sum, Georgia does not present a clear-cut case where path dependent structure beats out policy planning and change agency or vice versa. Rather, it is surely an example of the continued tension between these two dynamics, with both forces being reflected in the developmental results obtained. That might be said to be the inevitable state of a catch-up region that is closing some (but not all) gaps.
NOTES 1.
An example is EPSCoR – the Experimental Program to Stimulate Competitive Research – sponsored by the U.S. National Science Foundation in conjunction with several U.S. states to promote sustainable science and technology infrastructure and
INNOVATION CHALLENGES AND STRATEGIES
219
research improvements, so that EPSCoR researchers can compete for Federal and private sector R&D funding. See http://www.ehr.nsf.gov/epscor/. 2.
For updated information on U.S. state and regional innovation initiatives, see the State Science and Technology Institute, http://www.ssti.org.
3.
For example, Lampe’s (1988) volume highlights the ability of Massachusetts to draw on long-established strengths in research and new product development in engineering an economic turnaround in the 1980s.
4.
The mid-sized Georgia metropolitan areas are Albany, Athens, Augusta-Aiken, Chattanooga-Dalton, Columbus, Macon, and Savannah.
5.
Based on an adjusted analysis, not including parts of these cities that extend across state lines.
6.
See: http://www.onegeorgia.org
7.
Note: the author of this chapter is a principal in the Greater Georgia Project. The project was initially named CyberGeorgia, with a rebranding to its current name in 2003. For more information, see http://www.cherry.gatech.edu/mid
8.
The Greater Georgia project has counterparts in earlier regional networking efforts such as the European “Four-Motors Project” and its spin-offs in Canada (Wolfe and Gertler, 2001). However, one difference is that Greater Georgia focuses on lessfavored, traditionally non-innovative areas.
REFERENCES Allen, F., Atlanta Rising: The Invention of an International City 1946-1996. Longstreet Press, 1996. Berglund, D., State Funding for Cooperative Technology Programs. Columbus, OH: State Science and Technology Institute, 1998. Bergman, E.M., Feser, E.J., Industrial and Regional Clusters: Concepts and Comparative Applications, Web Book of Regional Science, Morgantown, WV: Regional Research Institute, 1999. http://www.rri.wvu.edu/WebBook/Bergman-Feser/contents.htm. Bernat, G.A., Jr., Convergence in state per capita income, 1959-99. Survey of Current Business, 2001: June. Bozeman, B., Technology and economic development for whom? The prospect for “dual agenda” state programs. AAAS, Washington, DC: AAAS Science and Technology Policy Year Book, 2000. Bureau of Economic Analysis, Metropolitan Area Personal Income and Per Capita Personal Income: 1999. BEA 01-12, Washington, DC, May 3, 2001.
220
Philip Shapira
Coburn, C., Berglund, D., Partnerships: A Compendium of State and Federal Cooperative Technology Programs. Columbus, OH: Battelle Memorial Institute, 1995. Combes, R. Todd, W., From Henry Grady to the Georgia Research Alliance: A case study of science-based development in Georgia. Annals of the New York Academy of Science 1995 (vol. 798). Corporation for Enterprise Development, 2003 Development Report Card for the States. Washington, DC: CfED, 2003. http://drc.cfed.org. Culp, R., Shapira, P., “Georgia’s Advanced Technology Development Center: An Assessment.” In Technology Incubators: Nurturing Small Firms. Paris: Organisation for Economic Cooperation and Development, OECD/GD (97) 202, 1997:63-74. Eisinger, P., The Rise of The Entrepreneurial State: State and Local Economic Development Policy in the United States. Madison, WI: University of Wisconsin Press, 1988. Georgia Rural Development Council, State of Rural Georgia. Atlanta, GA, 2000. Krugman, P., Geography and Trade. Cambridge, MA: MIT Press, 1991. Lambright, W.H., Catalyzing research competitiveness: The Georgia Research Alliance. Prometheus 2000, 18(4). Lampe, D. ed., The Massachusetts Miracle: High Technology and Economic Revitalization. Cambridge, MA: MIT Press, 1988. Markusen, A., Hall, P., Deitrich, S., Campbell, S., The Rise of the Gun Belt. New York: Oxford University Press, 1992. McMath, R., “Variations on a Theme by Henry Grady: Technology, Modernization, and Social Change.” In The Future South: A Historical Perspective for the Twenty-first Century, J.P. Dunn and H.L. Preston, eds. Urbana and Chicago, IL: University of Illinois Press, 1991. National Institute of Standards and Technology, Manufacturing Extension Program, Gaithersburg. MD: U.S. Department of Commerce. Worldwide Web site http://www.mep.nist.gov , 2002. Osborne, D., Laboratories of Democracy. Boston: Harvard Business School Press, 1988. Phillips, R.G., Evaluating Technology-Based Economic Development: Gauging the Impact of Publicly-Funded Programs. Lewiston, NY: Edwin Mellen Press, 2003. Porter, M., Atlanta-Columbus - Clusters of Innovation Initiative: Regional Foundations of U.S. Competitiveness. Washington, DC: Council on Competitiveness, 2002. Salstron, P., Appalachia’s Path to Dependency. Lexington, KY: University Press of Kentucky, 1997. Saxenian, A., Regional Advantage: Culture and Competition in Silicon Valley and Route 128. Boston, MA: Harvard University Press, 1996.
INNOVATION CHALLENGES AND STRATEGIES
221
Shapira, P., Youtie, J., Evaluating Industrial Modernization: Methods, Results and Insights from the Georgia Manufacturing Extension Alliance. Journal of Technology Transfer 1998; 23 (1): 17-28. Shapira, P., “Manufacturing Extension: Performance, Challenges and Policy Issues.” In Investing in Innovation, L.M. Branscomb, J. Keller, eds. Cambridge, MA: MIT Press, 1998. Wolfe, D.A., Gertler, M.S., Globalization and economic restructuring in Ontario: From industrial heartland to learning region? European Planning Studies 2001; 9:5. Youtie, J., Shapira, P., Mohapatra, S., Technology Infusion: Assessing Current and Best Practice Programs. Atlanta, GA: Georgia Tech Economic Development Institute and the Georgia Tech School of Public Policy, September 2000.
Chapter 11 PATH DEPENDENCY IN BADEN-WÜRTTEMBERG: LOCK-IN OR BREAKTHROUGH?
Gerhard Fuchs and Sandra Wassermann
11.1 INTRODUCTION In 1995, the Centre of Technology Assessment in Stuttgart, Germany, presented a paper on the innovation system of its home region, the state of Baden-Württemberg in southwest Germany. The paper predicted a dire economic future for Baden-Württemberg – which, up to that point, had been regarded as an economic success story and a model region. Close to ten years later, we are able to look again at the economic and political situation of the region and can compare the analysis of the mid-1990s with current conditions. The central thesis of the mid-nineties’ paper was that BadenWürttemberg was affected by a process of path dependent development which, although very successful in the past, would lead the region’s economy into a potential dead end street. A break away from the established paths of economic development was needed to guarantee future sustainable economic growth. Now (in 2003) we should be able to ask ourselves whether this analysis was correct. Has Baden-Württemberg in fact lost its leading position due to its concentration on established strengths or have significant changes taken place that altered the contours of the Baden-Württemberg model and made it more adaptive to changes in the world economy? Additionally, what role did regional policy play in this context? At first glance, Baden-Württemberg seems to be on the winning side again. In 1995 (especially with the most difficult year of 1993 in full view) all major economic indicators looked depressing. The performance of BadenWürttemberg was below the German average. In 2003, Baden-Württemberg is – in spite of a new economic crisis - along with neighboring Bavaria the economically most successful region within Germany. To understand, whether the case of Baden-Württemberg is an example of the rise of new approaches, a (successful) continuation of old practices, or a restructuring of
224
Gerhard Fuchs and Sandra Wassermann
these old practices, we do not seek simply to rekindle debate about whether regional policy can fundamentally change regional development paths, but to go beyond this by questioning and hopefully advancing underlying conceptual frameworks. In this chapter, we will first introduce the concept of path dependency, briefly review the main characteristics of the regional innovation system of Baden-Württemberg, outline the lock-ins that have been identified in the mid-nineties, look to what extent these lock-in effects have in fact been roadblocks for further developments and whether they have been removed. The main thesis of this chapter is that Baden-Württemberg’s success in the last years is the result of its staying “on–course.” The region has demonstrated its ability to introduce incremental reforms that combine old with new structural elements. This has been accomplished with the aid of crucial actors in the economic and political sectors who helped to change the guiding model of how to do things and who can be called “norm entrepreneurs.”
11.2 THE CONCEPT OF PATH DEPENDENCY To understand the changes that have taken place in Baden-Württemberg’s economy, we refer first to the notion of path-dependency, as this is the foundation of our analytical framework. Path-dependent phenomena are caused by positive feedbacks. This implies that history matters in the sense that historical paths once chosen, even by coincidence, then rigidify and dominate future developments. A famous example of technological path-dependency is the often-quoted QWERTY phenomenon of the typewriter keyboard. The typewriter keyboard arrangement solved a temporary mechanical problem on one of the first typewriters. But soon it became the standard for generations to come, even though it has been considered by many to be inefficient and other, potentially better, solutions were available.
11.2.1 Regional Path-dependency In considering how to frame and understand contemporary regional development, we begin with the assumption that in addition to technological trajectories (see Dosi, 1982), there are also regional trajectories. Technological knowledge is not only organized in large-scale technical systems (Hughes, 1987), in branches or professions, but also in regional economic areas. It can be expected that regions are also affected by path-dependency, as institutions play an important role in concepts of regional development. Moreover, they
PATH DEPENDENCY IN BADEN-WÜRTTEMBERG
225
are often used as the explaining factor in these concepts. Regional trajectories consist of economic structures that have developed in relation to pathdependent technologies and embedding institutions. Existing regional structures, institutional thickness, existing relations of cooperation as well as approved policy strategies are path-dependent – which implies, that they can be hard to change. Regional economies can also be understood as spaces of collective (technological) learning. The crucial question is how such learning can be institutionalized. Learning itself can be seen as an evolutionary process, following again path-dependent courses. Technological skills reflect local, regional, and national contexts and environments. Regions that are extraordinarily successful are those where the institutional context and regional networks closely complement the dominant industrial clusters of the region. Regional paths of development thus refer to accumulated competencies, methods, and technologies. This regional stock of knowledge is further developed within regional networks and can be closed towards inputs from outside. These processes of closure, however, might also lead to the development of lock-in situations that prevent regional institutions from adapting to changes in the environment and from learning new knowledge. Path-dependency can be maintained or fractured in particular ways in different locations. However, to make our position clearer, we will differentiate between three types: lock-in, incremental change and breakthrough. Breakthrough means revolutionary changes of crucial economic structures and relating institutions. Incremental changes mean evolutionary changes that do not set up new paths but try to lead old paths into a new direction (see Ackermann, 1999). Lock-ins, in contrast, are characterized by institutional and technological structures which are inefficient but cannot be changed. Grabher differentiates between three kinds of lock-ins: functional lock-in, cognitive lock-in, and political lock-in (see Grabher, 1993). Functional lock-in means that close cooperation within strongly-tied networks impedes contacts with other regions. Future trends that occur outside these networks may be overlooked, with external resources poorly identified and used. Cognitive lock-in means that personal relations lead to shared common ideas, feelings and beliefs that prevent the adoption of new ideas. Political lock-in describes situations where historical trajectories of economic development is maintained by cooperative relations between regional actors which are then unable to see and adopt new ways of thinking and policymaking. Generally, it is assumed that breakthroughs are difficult to achieve in the case of existing institutional thickness and established, distinct economic structures within a certain region. To the extent this is the case in BadenWürttemberg, we have to focus on the chances for incremental change.
226
Gerhard Fuchs and Sandra Wassermann
11.2.2 Guiding models Ideas and informal rules, despite not being written down formally, are interpreted as social institutions and a huge part of governance can be attributed to them. In analyzing path-dependency in regional development, one of the crucial questions is what role is played in regional development by informal institutions such as guiding ideas or patterns of interpretation of new economic and social phenomena. The sociology of technology works with socalled “guiding models” in order to explain the genesis and management of new technologies (see Barben, 1997: 133). We assume that the concept of guiding models can also be used in order to explain changes and developments in regional innovation systems. The notion of a guiding model encompasses two perspectives: on the one hand, we might be able to understand the genesis of new forms of economic behavior and aims; on the other hand we might be able to analyze political, economic and social interactions in managing these new forms (see Barben, 1997: 134). The guiding model concept looks for the role of ideas about given or future possibilities on how to do things. Such ideas condense into concepts which look ahead and which act as frames for orientating interpretations, thinking, decisions, and actions for individual and collective actors within regional innovation systems (see Dierkes et al., 1992: 11). Neither markets, nor hierarchies are able to solve specific co-ordination problems in social networks or in regional innovation systems respectively. Instead, the existence of regional guiding models renders the available frames of orientation (see Abel, 1997: 71). In this way, guiding models manage to connect different logics of action, as this is the case in regional innovation systems, when political, economic and scientific actors meet and develop ideas on how to stimulate innovation. Thus, guiding models reflect goals, towards which discourses are directed, or lines of orientation that structure them (see Dierkes and Marz, 1992: 16). In answering the question of how such guiding models come into existence, we have to leave the structural level and focus on the individual level, as it is usually crucial actors with the “power of definition” who shape and reshape guiding models. These actors take up ideas that are suitable to become guiding models and place them in discussions as long as they are established. However, even for very powerful actors it is not possible to just implant guiding models within a region. Instead negotiation processes between different powerful actors are needed to get relevant social groups to accept the new guiding model - and it is not before this happens, that we can consider a mere idea as a guiding model (see Abel, 1997: 74). In this perspective, guiding models can only be shaped and reshaped but never created out of nothing. This assumption brings us back to the concept of path dependency.
PATH DEPENDENCY IN BADEN-WÜRTTEMBERG
227
In the social sciences “...the role of ideas in particular has gained greater attention. Social constructivism has developed as a school of thought committed to explicating the way ideas are developed, disseminated, and implemented as public policies” (Cox, 2001: 495). Following the idea that social norms can be changed by norm-entrepreneurs, some political scientists have argued that political leaders are able to overcome the path-dependent constraints of existing policy institutions. This is consistent with our view that the relations between institutions are complementary: change in informal institutions may end up in changing formal institutions and the outlook of a regional innovation system. As guiding models can be changed by crucial actors within a society, these actors might be able to overcome the pathdependent constraints of existing policy and social institutions. By initializing new discourses, framing issues and redefining old models and ideas of how to do things, regional development can be realized by integrating old and new institutions.
11.3 THE REGIONAL INNOVATION SYSTEM OF BADEN-WÜRTTEMBERG 11.3.1 The success story The federal state of Baden-Württemberg in the south-west of Germany has been looked upon by many as a regional model economy up to the early nineties (Cooke and Morgan, 1990; Gabriel, 1990; Hassink, 1992; Maier, 1989; Sabel, 1989; Schmitz, 1992; Semlinger, 1993). With low levels of unemployment, high rates of industrial investment and export, a reputation for high quality, and well-engineered products, it seemed to have overcome major problems of many regional or national economies: namely how to establish and maintain competitive advantage. A high share of specialized and exportoriented capital goods producers, a dominance of flexible small and mediumsized enterprises (SMEs), and a high capacity for technological innovation within these firms were seen as decisive characteristics. The region enjoyed decades of economic prosperity because of its skilled labor, cooperative industrial relations, well-developed research structures, state and national industrial policies, and close and long-term relations between banks and companies. Social scientists interested in explaining the region’s economic success laid emphasis on institutional factors. They characterized the region as coming close to the model of an industrial district (see Sabel, 1989; Schmitz, 1992; Semlinger, 1994; Rehfeld, 1995) and endowed with a relatively self-
228
Gerhard Fuchs and Sandra Wassermann
contained regional economy and system of governance (Amin and Thrift, 1995: 7). They attributed the economic success of the region to a successful partnership among state government, dominant industries, financial institutions, research institutions, and universities (Wallace, 1994: 68). Baden-Württemberg is one of the “Four Motors of Europe” regions – a group of European regions regarded as being particularly successful in generating economic growth. It has earned this “nomination” thanks to its outstanding economic successes. Baden-Württemberg’s research intensity of 3.9 percent of its GDP (Germany: 2.3%; European Union: 1.8%) is way above average. Its employment rate (percentage of population aged 15-64) is well above the German and European level (1999: 69.5% in comparison with 65.4% and 62.8%); its unemployment rate well below these levels (1999: 5.1% in comparison with 8.9% and 9.4%); and its rate of European patent applications per million people (average for the period 1997-1999) is the highest of all European NUTS2-regions (416.4 in comparison with 227.3 in Germany and 119.4 in Europe).1
11.3.2 Revisionism Baden-Württemberg’s economic statistics are largely undisputed. However, starting in the early 1990s economic problems began to strike this model region. In these years, it became clear that we had been talking about a highly stylized success story. This story had been retold many times after the major contributions by Piore and Sabel (1984), Herrigel (1993) and others. What followed were revisionist interpretations, for example by Heidenreich and Krauss (1998), which demonstrated that some of the traits that had been considered to be the cornerstones of the Baden-Württemberg model were hard to find in empirical reality and seemed more to be properties deducted from the theory of industrial districts and flexible specialization. This section reviews several of the key features previously identified as important in the BadenWürttemberg model in the light of more recent evidence and reflection. Cooperation. Baden-Württemberg is certainly a region with considerable institutional thickness. Intensive cooperation and communication networks within a region are supposed to give rise to synergies that, according to the concept of flexible specialization, represent an important precondition for the success of industrial districts. In our case, however, one must clearly differentiate between vertical and horizontal cooperation patterns. In BadenWürttemberg, the significance of vertical relations between suppliers and buyers is very high. This reflects the extent to which companies in BadenWürttemberg’s three core branches of automotive, electrical and mechanical engineering are tied in regional clusters. However, cooperation between po-
PATH DEPENDENCY IN BADEN-WÜRTTEMBERG
229
tentially competing companies on a horizontal level seems rather the exception than the rule. In the industrial core sectors, relations between potential competitors are far less significant than the concepts of flexible specialization would suggest (Schmitz, 1992; Cooke et al., 1993). Based on a representative survey of West German mechanical engineering companies, Kerst and Steffensen (1995) were able to demonstrate that the number of cooperating companies in Baden-Württemberg is by no means above average. The share of cooperating mechanical engineering companies in Baden-Württemberg (1993: 37%) is on par with the West German average (36%). Therefore, it is doubtful that there is a higher incidence of cooperation activities between competing companies in Baden-Württemberg than in other German regions. Small and Medium Sized Enterprises. The assumption that the structure of Baden-Württemberg’s economy is more strongly determined by small and mid-sized “Mittelstand” companies than the rest of the German economy is also a misconception. Based on the average size of the regional companies, there is no difference between Germany and Baden-Württemberg. On the contrary, it can be argued that the prosperity of Baden-Württemberg’s regional economy is based on the success of a large number of big companies (such as DaimlerChrysler, Porsche, Bosch, and IBM). Technology Transfer. In the conventional Baden-Württemberg model, it was suggested that the region’s major research universities and its universities for applied sciences were closely geared towards the demands of local industry. Large firms were believed to have close cooperative relationships with the research universities, while SMEs cooperated with the universities of applied sciences. However, our research indicates that linkages of technology and knowledge transfer differ from this model. Bigger companies (more than 500 employees) are cooperating with the Steinbeis technology transfer centers (a system that was designed for small firms). Although most very small companies make almost no use of the institutional knowledge available in the regions, about 5 percent of the small companies (up to ten employees) in the region make use of the Steinbeis system. The importance of Steinbeis is also mirrored in data that show that 17.5 percent of the companies that are looking for support for the start of an innovation process are directing their enquiries there. At the same time, some 55 percent of the respondents, named their bank as a significant source of assistance, with 47 for chambers of industry and commerce, 23 percent for chambers of artisans, and 21 percent for trade associations (IHK, 2002:32). Core Industries as Core Problem. Drawing on the general discussion about economic development and industrial policy in Baden-Württemberg, the critics held that Baden-Württemberg’s past success might have become the very reason for the region’s comparatively slow orientation towards new industries. Grabher (1993) has aptly described how the Ruhr region became
230
Gerhard Fuchs and Sandra Wassermann
locked in on a once successful path of development, leading the region into deep economic crisis. According to Morgan (1994: 11), it would be “surprising” if some of the problems described by Grabher were not applicable to the Baden-Württemberg case. Indeed, Braczyk et al. (1995) identify several of these problems. They point to the slow adaptation of Baden-Württemberg’s industry to the “Japanese challenge” as a case of “cognitive lock-in” and to the bias of Baden-Württemberg’s technology policy towards the core industries as a case of “political lock-in.” They also detect a “functional lock-in” in Baden-Württemberg’s high level of economic integration: the fact that the three industrial core sectors are so closely interwoven makes the whole region particularly vulnerable to economic crises. In a similar vein, Heidenreich and Krauss (1998: 223) point out that the major companies in Baden-Württemberg’s core sectors are quite reluctant to source services from external providers. As a result, opportunities for communication and cooperation outside the established trajectories are missed and economic restructuring is hampered by barriers to learning. Heidenreich and Krauss (1998: 229; see also ifo, 1995) furthermore show that BadenWürttemberg’ s system of R&D and technology transfer strongly concentrates on the three industrial core sectors and on advanced technology rather than on the most advanced areas of technology. The symptoms of the limitations of Baden-Württemberg’s regional production system became increasingly apparent in the first half of the 1990s. During the economic crisis between 1992 and 1994, the economy of BadenWürttemberg suffered more than the other states of former West Germany. The GDP declined by 4.7 percent (1993), while the unemployment rate rose from 3.7 percent (1991) to 8.7 percent (1997). This crisis primarily affected the economic core region of Baden-Württemberg, the region of Stuttgart (see Krauss, 1999: 359). The key sectors of Stuttgart’s economy were greatly affected by this worsening situation. The automotive, mechanical engineering, and electronics and electrical engineering industries all underwent processes of personnel retrenchment. Between 1980 and 1995, more than 15 percent of the employment in mechanical engineering was lost; in the electronics and electrical engineering industry, the decline amounted to almost 14 percent of the jobs. Employment related to the production of electronic data processing systems declined by 10.5 percent between 1980 and 1996 (see Statistisches Landesamt, 1997). Long-term employment prognoses for these sectors were quite dim, too. In 1994, one research institute predicted employment to drop between 1991 and 2010 in Baden-Württemberg’s electronics and electrical engineering industry by 9 percent, in the mechanical engineering industry by 15 percent and in the automotive industry by 26 percent (Saebetzki, 1994: 92).
PATH DEPENDENCY IN BADEN-WÜRTTEMBERG
231
The question was asked, given Stuttgart’s exceptional success in capital goods industries, why its economy did not succeed in staging a stronger reorientation to sectors promising greater growth potential. It seemed that Stuttgart’s economy was failing to stake out greater shares in new areas such as information and communications technology, new materials, biotechnology, environmental and power technology, microsystem technology, and production oriented services (Faust et al., 1995), although the industrial, structural, and institutional preconditions in each of the cited areas were certainly not unfavorable. These questions lead invariably to the inertia of established production structures. Let us therefore look at the dominant features of the innovation system of Stuttgart/Mittlerer Neckar.
11.4 STUTTGART AS THE CORE REGION It is apparent that when Baden-Württemberg is referred to, many people in fact are talking about its core region of Stuttgart and Mittlerer Neckar. We note that Baden-Württemberg features several economic regions with very distinct characteristics and problems. However, to be consistent with past practice, in this chapter we too will focus on the core Stuttgart-Mittlerer Neckar region. Data for Baden-Württemberg will only be used if relevant data for the core region are missing and the data can be considered as adequate indicators for the core region as well. The region of Stuttgart comprises, besides the city of Stuttgart itself, five adjacent administrative districts and is home to more than 2.5 million inhabitants. Over one million people, more than a quarter of all employees in Baden-Württemberg, are employed in that area.2 Although there are other economic centers in Baden-Württemberg – above all the regions of Karlsruhe and Mannheim-Heidelberg – the Stuttgart region is the most important of them. The region of Stuttgart had a gross domestic product of 55,400 per employee in 1998, which is 8 percent above the Baden-Württemberg average and 15 percent above the German average. This is impressive but has also to be seen in comparison with other German agglomerations. The comparable data for Munich is 64,600 per employee. The lower rate for Stuttgart still reflects the major importance of the industrial sector in Baden-Württemberg. Other agglomerations have accumulated production-oriented services which feature a high gross product per employee. In Stuttgart, more than a quarter of all employees are still within the manufacturing sector. In Hamburg or Cologne, the rate is around 10 percent. The export orientation of a region reflects its ability to stand firm in the international competition. With an export ratio of 47 percent in the manufacturing sector, Stuttgart is the most export-intensive region in Germany. Only
232
Gerhard Fuchs and Sandra Wassermann
the Munich region achieves similar success. Stuttgart’s unemployment rate is among the lowest in all of Germany. Only Munich again, as a comparable location, has a similarly low rate (4.1 percent in October 2001). Stuttgart-Mittlerer Neckar core area has about 45 percent of all R&D capacities of Baden-Württemberg. The research intensity of private companies in the region is 4.6 higher than the German average. It is still 2.2 times the Baden-Württemberg average. It features two universities, various universities for applied sciences, and vocational colleges. Furthermore there is a high number of institutions of extra-university research like institutes of the Fraunhofer and the Max-Planck associations, institutes of the industrial joint research group, contract research institutes at the university and a number of organizations for technology transfer most notably the Steinbeis Foundation. The Steinbeis-Stiftung für Wirtschaftsförderung (Steinbeis Foundation for Economic Development) is a statewide umbrella organization. In 1998, the technology transfer related activities became a private business activity and therefore were outsourced to a private firm called Steinbeis GmbH & Co. für Technologietransfer. This company maintains a network of about 500 centers3 for technology transfer, in most cases close to a Fachhochschule (university of applied sciences). The Steinbeis-foundation over the 1990s has developed itself into a more effective, market-oriented organization. Only “successful” professors can now lead a Steinbeis center. If a center fails to win contracts in a satisfying manner for a certain period of time then the center will be closed. This is one of the mechanisms to link professors to the changing demands of industry.
11.5 THE PRESENT STATE OF THE REGIONAL ECONOMY The assumption behind the questions asked by the critics of the regional innovation system of Stuttgart is obvious. Innovativeness is considered the result of communication and cooperation opportunities outside of historically evolved and institutionally and organizationally reinforced trajectories. Heidenreich and Krauss assumed that such communication and cooperation opportunities did not exist in Stuttgart. The implication was that barriers to learning and not the maturity of Stuttgart’s product range formed a major obstacle on the road to new innovation-promoting company strategies. As noted in the previous section, when looking at the situation today, there is little doubt that Stuttgart is again performing very well economically. In view of this, we have to ask whether the critical developments, which had been identified in the mid-1990s have been changed and the rule of path de-
PATH DEPENDENCY IN BADEN-WÜRTTEMBERG
233
pendency has been broken up. The question is: what is the reason for the good economic performance? We have to answer the question whether developments can be interpreted as an example of breakthrough or incremental change: The changes we will find in our analysis will be interpreted in the context of “lock-in” or “breakthrough.” As it is difficult to indicate the exact demarcation between incremental and revolutionary changes, we assume that the variety of changes can be localized on a scale where one pole denotes a phenomenon which explicitly indicates lock-in and the other pole denotes a phenomenon which indicates an absolute breakthrough. While the economic data seem to be comforting, some of the disturbing trends recognized in the mid-1990s have kept stable. In comparison to Germany, the Baden-Württemberg regional R&D system continues to be characterized by a specialization on technologies relevant for the construction of machines and cars. The dominance of the manufacturing sector has not been severely challenged. In Stuttgart-Mittlerer Neckar, three sectors remain particularly important measured in terms of both employment and sales: the automotive industry, the mechanical engineering and the electronics and electrical engineering industry (see Table 11.1). Among these three, the car industry stands out more than ever before. Stuttgart maintains and, indeed, has further developed “the largest, thickest, and the most powerful auto cluster in Europe” (Morgan, 1994: 37). As this cluster has strongly influenced not only the region’s economy but also its institutions and culture, it is no exaggeration to still portray Stuttgart as a “car city.” Let us look more closely at the three sectors and recent developments there.
11.5.1 Electrical Engineering Electrical engineering comprises a wide variety of activities: the production of office equipment, computers, electronics, precision engineering and optics. This broad definition has to be kept in mind when looking at the data. Electrical engineering is – in terms of turnover – the third biggest manufacturing
234
Gerhard Fuchs and Sandra Wassermann
sector in the region of Stuttgart. With more than 11.9 billion sales turnover, electrical engineering comprised 15.4 percent of the region’s manufacturing sector. The growth rates of this sector are, however, below the German and the Baden-Württemberg average in spite of the dramatic increases during the boom years of the New Economy. Nonetheless, 56 percent of the sales are from exports, which is well above the Baden-Württemberg average. Growth in this sector is mainly due to export activities. Domestic demand is lagging behind considerably. The sector of electrical engineering employed 75,800 people (2000). The sector is thus the second largest industrial employer. Large companies like Alcatel, Bosch, IBM, and Hewlett Packard are responsible for these figures. Part of the electrical engineering sector is strongly affiliated with the automotive sector. Besides Bosch, this is mainly due for a number of SMEs. In spite of the fact that electrical engineering is statistically part of the manufacturing sector, growth is mostly due to growing service-oriented activities. Direct production-oriented activities continue to decline quite significantly. In the Stuttgart region, 39 percent of the people engaged in R&D activities are working in the electrical engineering sector. This fact signifies that the sector is very research-intensive. 17.3 percent of the employees are working in R&D. In the automotive sector, the respective number is 10.6 percent and in mechanical engineering 9.2 percent. More than the mechanical engineering and the automotive sector, the field of electrical engineering has become the object of immense transformations. As is the case with the other sectors, the traditional strengths of Stuttgart were in the manufacturing realm. However, the major companies of the region (Bosch, Alcatel, IBM, Hewlett Packard) have downscaled or completely closed down their production facilities. This loss was not compensated by increasing employment in services or content. On the other hand, this sector is also much more open to new entries. Fifty-one percent of all technology-oriented/technology-based new firm creations are IT related. The machine tool industry accounts only for 12 percent. Car manufacturing does not appear among the leading sectors for new firms. It must be mentioned, however, that a significant proportion of new firm creations nevertheless are effected by firms providing input for the car manufacturing industry, and thus are not counted as automotive but are nevertheless directly linked to it. Car manufacturing and electrical engineering are interlocked to a significant degree. Mechanical engineering and automotive also show a sizeable degree of interconnections, while this is hardly visible for mechanical engineering and electrical engineering.
PATH DEPENDENCY IN BADEN-WÜRTTEMBERG
235
11.5.2 Mechanical Engineering Mechanical engineering is the second biggest industrial sector in the Stuttgart region. In 2000, about 453 companies were active in this sector with an average employment of 164 people. This illustrates the SME character of this sector. The sector is very heterogeneous, but it also features, as indicated above, a number of significant companies geared towards the automotive sector. Contrary to electrical engineering, the mechanical engineering sector is experiencing a growth rate that is well above the Baden-Württemberg and the German average. The share of the Stuttgart region in the mechanical engineering industry in Baden-Württemberg is above one third. Similar to the electrical engineering sector, export demand is maintaining growth in the face of consistently weak domestic demand. The export rate of the mechanical engineering industry is 54 percent (2000), which is above the BadenWürttemberg average. With respect to employment, the mechanical engineering industry is the third largest industrial sector in the Stuttgart region. In spite of the overall favorable figures, employment is shrinking. The loss of employment is significantly higher than in the rest of Germany. The continuing economic success of this sector, which does not find its imprint in the employment data, is due to a continuing high tech orientation and (resulting) significant productivity increases. Fifty-five percent of the companies in the mechanical engineering sector spend more than 3.5 percent of their turnover on R&D.
11.5.3 Automotive Industry The automotive industry represents by far the most important industrial activity in the Stuttgart region. The global automotive sector is undergoing a severe process of restructuring, which has important impacts on primary manufacturers as well as their suppliers. This reorganization is taking place on a worldwide scale and along the entire value chain. However, this restructuring has helped the Stuttgart manufacturers to stay on top and strengthen their position. There are approximately 100 companies in Stuttgart’s automotive sector. They are responsible for 43.6 percent of the industrial output in the region. This signifies an important growth trend. In 1980, car manufacturing accounted for 28.7 percent of the total output. In 1992, its share has dwindled to 27.3 percent, but since then has steadily risen up to 43.6 percent in 2000. 25.9 percent of all industrial employees are to be found there. It is therefore also the biggest employer in the region. Over one hundred thousand workers
236
Gerhard Fuchs and Sandra Wassermann
were employed in 2000, with employment growing faster than in the rest of Baden-Württemberg as well as in Germany. This development is in some way special to the Stuttgart region. Not all manufacturers in Germany and especially the core manufacturing regions were as “lucky” as Stuttgart with DaimlerChrysler, Porsche, and Bosch (the world’s second largest automotive supplier). Stuttgart did better than the rest of Baden-Württemberg as well as the rest of Germany. A special feature is again the export orientation. In 2000, exports amounted to 65 percent of all production, up from 57 percent in 1996. Almost one-half of the employees working in the automotive sector are fulfilling manufacturing duties. However, as in other sectors, direct manufacturing tasks are declining, however, and services and R&D- related activities are solely responsible for the recent job growth. Car manufacturing is also very research-intensive. Close to 70 percent of companies spend more than 3.5 percent of their turnover on R&D. As mentioned above car manufacturing has been subject to a global restructuring for some years. This global restructuring has found its imprint on the regional level. Stuttgart has benefited from both its strong export orientation in the sector as well as the decision of the major players to turn Stuttgart into a worldwide hub for car manufacturing. All the important functions for car manufacturing are present in the region. Stuttgart could well have fallen victim to the merger negotiations between Daimler and Chrysler. Stuttgart, however, succeeded to get out of these negotiations as a strengthened partner. This put the region also in contrast to the developments in Germany or worldwide, where several regions had to suffer severe losses.
11.5.4 Patent statistics Stuttgart is the leading region in Germany with respect to patent activities. This holds both true for the number of patents per 100,000 inhabitants as well as employees (1998). The ratio is double the German average and 18 percent above the Baden-Württemberg average. The relative position of the region, however, has changed. Other locations like Munich have shown a higher growth rate in recent years and were able to close the gap with Stuttgart. Patent statistics also demonstrate clearly that the Stuttgart economy develops along stable trajectories. Patents are concentrated in car manufacturing and mechanical engineering, with s somewhat lower specialization in electrical engineering. The data also show clearly, that over the last decade new patent specializations were not being established, in spite of all the political efforts to the contrary. The degree of specialization has in fact even increased for the case of car manufacturing and to a more limited degree in mechanical engineering. The most important single companies with respect to patenting ac-
PATH DEPENDENCY IN BADEN-WÜRTTEMBERG
237
tivities remained Bosch, Porsche, DaimlerChrysler, and Alcatel (although the latter has reduced significance). The technological strengths of the region undoubtedly are with car manufacturing and technologies of special relevance for car manufacturing, including auto-related IT. The combination of excellence in these fields remains an essential comparative advantage for the region.
11.5.5 “Old” v. “New” economy trends The data presented in previous sections show that industry sectors that were identified as in “critical” condition in the 1990s continue to dominate the Stuttgart regional economy. This also holds true for the following aspects. The share of the Baden-Württemberg economy in the worldwide trade of R&D-intensive goods has declined constantly (see Tables 11.2 and Figure 11.1). The role of Baden-Württemberg in (non-automotive) high-technology sectors has never been that strong and it continues to decline rather sharply. The most radical downturn is to be observed in the electronics sector (see Figure 11.1). Here Baden-Württemberg has fallen victim to the abovementioned trend of a radical decrease in the employment in the hardware part of the electronics sector. The major companies (IBM, Alcatel-SEL, Bosch, Siemens, Hewlett Packard, and Sony) are still reducing their manufacturing capacities in the region (in the case of IBM none is left) and the service oriented employment in the same sector could not be built up in comparable quantities (see Seufert, 2000).
238
Gerhard Fuchs and Sandra Wassermann
Figure 11.1: Products with high technology content and Baden-Württemberg’s share of world exports in this category (by sector) Source: OECD – International Trade by Commodity Statistics, Rev. 3, 2000.
The concentration on the three core sectors mirrors another potential weakness of the regional innovation system: The economy is concentrated on markets which feature an under average growth on the world market. Goods, which experience the greatest growth potential worldwide, are only represented in a below-average degree. The even increasing concentration on core competences has led to the fact that Baden-Württemberg has lost its lead in more areas than it has achieved new leads (see Lay, Eggers, and Rainfurth, 1999). For example, the regional economy has not secured a strong position in advanced services. Of course, there has been some growth in services, but this increase mostly reflects the general German trend. Data do not reveal that, in Stuttgart or in Baden-Württemberg for that matter, services have any special, pronounced growth dynamic. Overall, this presents us with a conflicting picture. On the one hand we see that most of the traits of Stuttgart’s economy that have been considered to be problematic are still there, in certain ways even extenuated. But we also see that the Stuttgart region remains one of the leading regions of Germany and has done better in recent years than the rest of (less industrialized and more service oriented) the country. The strong concentration on specific technological fields was in the second half of the 1990s a major reason for the above average performance. While some regions declined and others moved upward, Stuttgart kept its leading position. In this respect, the crisis interpretation of the mid-1990s has to be reassessed. One of the leading justi-
PATH DEPENDENCY IN BADEN-WÜRTTEMBERG
239
fications for the assessment of the mid-1990s was that the “old” economy was on the decline. Yet, the subsequent experiences of Stuttgart demonstrated that the “old” economy was well able to stay alive and prosper – under specific circumstances. The discussion after the demise of the “New Economy” bubble has shown that the essential component for economic well-being is not a single-minded promotion of the new economy, but a blending of new and old economy elements. Here, Stuttgart’s industrial sector has done quite well. It has kept its R&D expenditures high, has absorbed new knowledge for developing products that are more competitive, introduced organizational reforms, and has slowly increased its services component. Moreover, there have been other important changes that are not adequately mirrored in the statistics, which will be discussed in the next section.
11.6 THE NORM ENTREPRENEURS At an early stage, the state government of Baden-Württemberg was aware of a potential lock-in situation in the region’s economic development. The person symbolizing the search for a new role for Baden-Württemberg and its core region of Stuttgart was the then minister president Lothar Späth. Already in the late-1980s, he tasked expert commissions to consider the situation and make recommendations. This resulted in a number of activities to tackle lockin phenomena and to develop new industry and service sectors and to build up of a number of new institutions in the field of science. Most prominently among the sectors targeted were biotechnology, multimedia and produceroriented services. Späth was aware of the potential difficulties, but reinterpreted these as opportunities worth taking a risk. In this sense, he viewed globalization as an option to be taken up proactively. In addition, many actors in the region in the early 1990s began to share the assumption that Stuttgart faced not a temporary crisis but profound structural changes, most importantly with the trend towards the globalization of markets, towards the decline of mature industries and towards the rise of highly innovative, knowledge-intensive, and service-oriented industries (New Economy). A broad agreement gradually came into existence on the view that Stuttgart could not successfully cope with these changes merely by concentrating on its traditional key industries. Instead, the call for a fundamental restructuring of Stuttgart’s regional economy became louder. In the industrial sector, Jürgen Schrempp – the most outspoken representative of treating globalization as a positive option – became the head of DaimlerChrysler. While the activities of the politicians were mainly in the realm of image building, changing the priorities of funding, looking for new policy instruments, constructing new institutions, Schrempp transformed the regionally
240
Gerhard Fuchs and Sandra Wassermann
based company Daimler-Benz into a globally oriented and globally based company DaimlerChrysler. Let us look at these developments in more detail.
11.6.1 The automotive sector Contrary to many prophecies, Stuttgart’s automotive sector has not declined. Rather, it has grown in importance in recent years. Underlying this development was the strategic decision to further develop the Stuttgart region into one of the few worldwide important centers of competence for this industry. Stuttgart features a nearly unique combination of managerial, production, research, and infrastructure components. The merger of DaimlerChrysler has further strengthened the position of Stuttgart. Even a worldwide decline in the automotive sector will most probably hit Stuttgart less hard than other less well-positioned regions. Furthermore, Stuttgart has strengthened its position within Germany with respect to both turnover and employment. Compared to the recession year of 1993 in which every seventeenth job was dependent on the automobile industry, this relationship has jumped today to every fourteenth job. Considering only the manufacturing sector, dependence on automobiles has increased from every sixth to every fifth working place. It has to be said that the adaptive capacities of the industrial core are remarkable. This can be especially demonstrated for the case of the car industry by its successful turn to global markets over the 1990s. Exports rose from 43 percent (1991) to 61 percent (2000). While in 1995, 22 percent of all German exports came from Baden-Württemberg, in 2000 this figure climbed to 27 percent. Between 1995 and 2000, more than 50 percent of the increase in the region’s valued-added output can be contributed to the car industry, and its share of regional industrial output has increased from 16 to 21 percent. Successful organizational reforms have been implemented, and an adaptive system of input suppliers was further developed, which expanded even more strongly than the core manufacturers. While in 1993 191,000 people were employed in the core producing companies, 84,000 (30 percent) were with the suppliers. In 1998, 193,000 were with the core producing companies and 142,000 (40 percent) with the suppliers. In the process of the merger between Daimler and Chrysler, a number of U.S. suppliers set up shop in the region of Stuttgart. This did not imply manufacturing capacities, but new marketing, service and research and development facilities. The implication of these developments surely is that the region is very dependent on the strategies of the main companies. This is not only DaimlerChrysler but also Porsche, which staged a broadly admired turn around in the 1990s, along with three further car-producing companies. Next to these big manufacturing firms is a cluster of worldwide leading supplier firms, engi-
PATH DEPENDENCY IN BADEN-WÜRTTEMBERG
241
neering and research companies, service companies, fuel cell cooperation networks, regional competence centers, institutes of the University of Stuttgart, traffic telematics companies, and regional pilot projects dealing with the future of mobility (e.g. STORM, Mobilist).
11.6.2 Parallel institution building One means of increasing regional innovative capacity is by creating institutions. This often proves an inadequate solution, however (see Amin and Thrift, 1995). Harder to achieve is the development of a new regional identity and the generation of synergy effects between institutional and technological development paths. This is the task facing especially those regions which have been very successful up to now, and which have achieved a high level of technological competence in the so-called mature industrial sectors. For these regions, the hitherto established institutions and the institutional thickness thus achieved can even become an additional problem, since training, research and funding facilities tend to stabilize the traditional patterns of industrial development. Against a background of intensified global competition, this problem of institutional inertia and restrictions deserves special attention. The state government of Baden-Württemberg over the last decade was fully aware of the problems of institutional inertia. This institutional inertia has been analyzed in various studies that did show that main pillars of the Baden-Württemberg model were concentrating on the core industries and had difficulties adapting to the new industries. The Steinbeis foundation, for example, did not play any significant role in the early development of the multimedia or the biotechnology industry. The response of the state government was a strategy that combined parallel institution building with a reform of old institutions. Parallel institution building means that for the support and promotion of new industries, new institutions were created. For instance, in the multimedia sector the Medien- und Filmgesellschaft Baden-Württemberg (Media and Movie Association of Baden-Württemberg) was created. In the case of biotechnology, a state agency was established and various regional forms of institution building have been supported. It is noteworthy that these attempts – at least in the beginning - totally bypassed the established institutional structures and had few connections with these structures. The further development of existing capacities and the search for new product and production concepts is thus being supported by many new institutions as well as by expert commissions, a dense network of regional technology transfer institutions and industry policy initiatives. The Future Commission Economy 2000, appointed in 1992, gave the starting signal for the
242
Gerhard Fuchs and Sandra Wassermann
search for “ways out of the crisis.” This was followed by the Innovationsoffensive with a fund of DM1 billion ( 0.52 billion) committed to new technical faculties, data highways, biotechnology parks, software centers. In 1994, an Innovation Council was set up and this council developed, among other things, proposals for creating a biotechnology agency, a microsystem forum and better computer facilities in schools. Within the context of a “Future Campaign” (Zukunftsoffensive) the state government has realized these proposals since 1996. An additional DM1 billion was invested in upgrading technical colleges, colleges of advanced vocational studies, universities, clinics, schools and libraries. Since 1997, the state government has supported the establishment and expansion of five biotech parks at Freiburg, Heidelberg, Esslingen and Reutlingen/Tübingen with a total of 12.5 million. Support has also been given to business start-ups, networked research projects, and regional trade fairs. Start-up companies for example have been supported by subsidy programs, investment shares, and a venture capital fund that was set up in 1998. Within the framework of the programs “Young Innovators” and “Start-up Founders on Campus,” the government supported entrepreneurs from universities and research institutes. From the regional perspective, the most important element has been the foundation of the Association Region of Stuttgart in 1994. This was done in the form of a decision by the state government and is a unique development. No other region became equipped in a similar manner. The most important task of the association is the furthering of economic development. It was clear from the beginning that this association should be primarily concerned with new sectors and new types of activities to support structural change in the region. The association is the product of a top-down development, represents the core of a new regional order, and is in the center of other institutional reforms as well. The association was active in the development and implementation of major projects. It organized regional capacities and resources in order to be able for the region to participate in national as well as European competitions for money. Complex partnerships were built up, organized, and moderated. The operationalization of the aims of the association was done gradually via concrete projects. In an exemplary manner one can name projects for site development, attracting new investors, network projects like the Regional Dialogue Car Production, MediaRegio Stuttgart, PUSH!, Mobilist, BioRegio, regional competence centers, the support program environmental technology and a host of marketing and image campaigns. These are still examples that show how within a short time frame the agency has developed its tasks. It also shows that in various fields there was nothing much in existence before. For example, previously site marketing not considered necessary by the city
PATH DEPENDENCY IN BADEN-WÜRTTEMBERG
243
of Stuttgart, because the local politicians believed that Stuttgart as a strong economic area would not need any further marketing. Central fields of action are the promotion of innovations (not technologies) and the support of cooperation (see Table 11.3). On the one hand, this is the support for the development and use of new products and production processes; on the other hand, it is the instigation and support of cooperation in order to make use of potential synergetic effects in the region. In this sense, the association pursues a structural policy strategy, aiming at the support and stabilization of regional clusters, which should be accomplished by a network and innovation oriented regional policy.
The analysis of structures of regional networking has become an important topic of research over the last couple of years. Beneficial structures of regional networking might offer options for regions, for endogenous economic growth and can make available the resources for cross-regional and cross-national activities. Regional networking can become a restriction if available resources are mainly or exclusively used for a mere reproduction of existing structures and are thus literally “consumed.” This will lead to dangerous lock-in effects. In this case, closely knit relationships and established structures of intertwinement facilitate a strong orientation of actors towards established paths of development and well-proven problem solutions as a result. Regional networking structures, however, may also work as a springboard for the opening up of new cross-regional markets and resources. Thus, a bridge building between regional and global economic trends can be achieved. The success of Stuttgart was linked to the impression that there is a large degree of networking between economic actors going on. Research in the 1990s has repeatedly shown that this is a misconception. Nevertheless an objective demand developed among corporations for an increased amount of cooperation both within existing value chains (automotive sector) as well as beyond (e.g. publishing, multimedia). This holds true both for the level of inter organizational division of labor (between companies) as well as with respect to the innovative capabilities of companies, sectors, and clusters. The Association aims at filling this gap.
244
Gerhard Fuchs and Sandra Wassermann
Another exemplary development to be mentioned here with respect to the reform of institutions refers to the whole system of universities of applied sciences, which has undergone a process of structural change in the second half of the nineties. This type of universities was originally thought to fulfill the demands of the local industry. The severe economic crisis especially in the first half of the nineties and a rapidly decreasing demand for university graduates by regionally based companies, led to a dramatic situation (e.g. few new students) in many a university for applied sciences. The answer was twofold: on the one hand, traditional curricula have been sometimes even dramatically reduced, on the other hand new curricula have been developed, which aim at decreasing the dependency on specific regional or local demand structures. Even smaller universities are now faced with the demand to develop “world class” specializations. The regional orientation thus has been reduced and the importance of the openness to supra regional developments and actors strengthened.
11.7 CONCLUSION Grabher (1993) differentiated between three kinds of lock-ins that were considered to be at work in the Stuttgart region: functional lock-in, cognitive lock-in and political lock-in. As previously discussed, functional lock-in implies that close cooperation within strong ties networks impedes contacts with other regions: future trends that occur outside are overlooked and external resources are poorly identified and used. With respect to functional lock-in, we have analyzed the situation especially for the case of the automotive sector, which seemed to have successfully further developed its strategic position by developing itself into a truly global player and extending regional networks into a global arena. The merger of DaimlerChrysler is a symbol for this as well as the assertiveness of the growing supplier industry. Functional lock-ins are intensified through cognitive lock-ins. Personal relations lead to shared common ideas, feelings and believes. The fact, that entrepreneurs in Stuttgart did not notice Japanese competitors at an early stage and that they neglected to introduce countermeasures in time, was interpreted as a cognitive lock-in. Collective misinterpretations were the reasons why Stuttgart did not evaluated its industrial structure in the context of global division of labor as well as division of technological competences. With respect to cognitive lock-ins it can also be asserted that starting in the mid nineties Stuttgart began to reverse its image of being an old industry region. All major actors were looking towards new possibilities and options. Attempts to build up competencies beyond the core industries (as described in this chapter) are an impressive sign for this.
PATH DEPENDENCY IN BADEN-WÜRTTEMBERG
245
Political lock-ins describe situations where economic development proceeds along historical trajectories that are supported by cooperative relations between political actors, trade unions, employers associations and other embedding institutions. As a political lock-in, we identify the fact that in Stuttgart the economic and technology policy of the regional government was limited to modernization efforts within the core branches of the old economy. These branches received political support in order to stay competitive in globalized highly competitive markets. We have seen, however, that increasingly policy strategies have been introduced, to develop economic and structural alternatives for the region in order to loosen regional dependence on established industrial clusters. In this case it must be said, that the official policy on state and regional level has clearly changed focus and is attempting to support both the modernization of the old sectors as well as the promotion of new sectors. Our analysis has showed that important aspects in this respect have been the policy of parallel institution building. Finally, it has to be said that the analysis of the mid-1990s might have been hampered by a generally held over-estimation of the developments in the new economy. The recent demise of this sector has shown that the mature industries do not necessarily need to look “old.” There is also a promising future for parts of the old industry. Nevertheless, the picture is not only bright. The development of strategies of institutional learning remains a challenge to be mastered. While many “institutionally poor” economic regions seek to adopt the seemingly exemplary institutions of other countries, Stuttgart is faced with the challenge of restructuring and transforming an exceedingly rich institutional landscape. These challenges put the well-considered (and concerted) modernization of regional institutions at the top of the agenda. The expedient further development of communication and cooperation promoting institutions is therefore a central precondition for the design and rejuvenation of innovation-friendly environments. But it has also been demonstrated that institutional reforms and a transformation of existing productive structures is a difficult and extremely risky endeavor. Each of the three core sectors of the Stuttgart economy is following its own path of “learning” and features quite distinct paths of development. Among the three core sectors, only the automotive sector seems to be on a secure future oriented path. Furthermore, the institutional reforms, the reworking of the image of the region of Stuttgart and the establishment of new networking and cooperation structures has so far not found its imprint in the overall data on regional economic performance. The success of this planned reorientation on the Baden-Württemberg model level can only be judged in the future and will be dependent on continuing efforts.
Gerhard Fuchs and Sandra Wassermann
246
NOTES 1.
European Commission, Second report on economic and social cohesion, Luxembourg 2001. NUTS is an acronym for “Nomenclature of Territorial Units for Statistics” – based on territorial definitions developed by the Statistical Office of the European Communities (Eurostat) for producing regional statistics in the European Union. NUTS2 regions are sub-national units with between 0.8 million and 3.0 million inhabitants.
2.
The exact figures (2000) are 1,056,363 people employed in the Stuttgart region, 3,802,494 in Baden-Württemberg. Stuttgart’s share in Baden-Württemberg’s employment is thus 27.8 percent (the numbers refer to persons officially employed and registered in social security schemes.
3.
In 2003, Baden-Württemberg hosted 329 Steinbeis transfer centers. While Steinbeis increasingly creates new centers in other German and foreign regions, Baden- Württemberg remains the main location of this organization (there are 121 Steinbeis centers in other parts of Germany and 14 centers abroad).
REFERENCES Abel, J., Von der Vision zum Serienzug: Technikgenese im schienengebundenen Hochgeschwindigkeitsverkehr. Berlin: edition sigma, 1997. Ackermann, R., Pfadabhängigkeit, Institutionen und Regelreform. Diss. U of Freiburg, 1999. Amin, A., Thrift, N., “Living in the Global.” In Globalization, Institutions, and Regional Development in Europe, A. Amin, N. Thrift, eds. London/Oxford: University Press, 1995. Barben, D., “Genese, Enkulturation und Antizipation des Neuen. Über Schwierigkeiten und Nutzung, Leitbilder der Biotechnologie zu rekonstruieren.” In Technikgenese, M. Dierkes, ed. Berlin: edition sigma, 1997. Braczyk, H.-J., Schienstock, G., Steffensen, B., “The Region of Baden-Württemberg: a Post Fordist Success Story?” In Industrial Transformation in Europe. Process and Contexts, E. J. Dittrich, G. Schmidt, R. Whitley, eds. London and others: Sage, 1995. Cooke, P., Morgan, K., Learning through networking: regional innovation and the lessons of Baden-Württemberg. University of Wales, College of Cardiff. Regional Industrial Research Report, No. 5, 1990. Cooke, P., Morgan, K., Price, A., The Future of the Mitrtelstand. Collaboration versus Competition. Regional Industrial Research Report 13. Cardiff: Department of City and Regional Planning, 1993. Cooke, P., “Introduction: Origins of the Concept.” In Regional Innovation Systems, H.-J. Braczyk, P. Cooke, M. Heidenreich, eds. London: UCL Press, 1998.
PATH DEPENDENCY IN BADEN-WÜRTTEMBERG
247
Cox, R.H., The Social Construction of an Imperative. Why Welfare Reform Happened in Denmark and the Netherlands but not in Germany. World Politics 2001; 53: 463-498. Dierkes, M., Antal, A., Hähner, K., Business in Society, the role of perceptions and principles in organisational learning. Berlin: Wissenschaftszentrum Berlin für Sozialforschung, 1992. Dierkes, M., Marz, L., Leitbildprägung und Leitbildgestaltung: zum Beitrag der Technikgenese-Forschung für eine prospektive Technikfolgen-Regulierung. Berlin: Wissenschaftszentrum Berlin für Sozialforschung, 1992. Dosi, G., Technological Paradigms and Technological Trajectories: A Suggested Interpretation of the Determinants and Directions of Technological Change. Research Policy 1982; 11:147162. Eisinger, P.K., The Rise of the Entrepreneurial State. Madison: University of Wisconsin Press, 1988. Faust, M., Griupp, H., Hummel, H., Klee, G., Laube, T., Münzenmaier, W., Saul, C., Schmoch, U., Waldkircher-Heyne, C., Der Wirtschafts- und Forschungsstandort BadenWürttemberg. Potentiale und Perspektiven. Studien zur Strukturforschung 19, Ifo Institut für Wirtschaftsforschung, Munich, 1995. Gabriel, J., “Innovation-oriented policy in regions with growth dynamics: three winners in the process of structural change – a comparison of Baden-Württemberg, Massachussetts and Emilia-Romagna.” In Innovation and regional development strategies. Instruments and policy coordination, H.J. Evers, J. Allesch, eds. Berlin, New York: Springer, 1990. Grabher, G., “The Weakness of Strong Ties. The Lock-in of Regional Development in the Ruhr Area.” In The Embedded Firm. On the Socioeconomics of Industrial Networks, G. Grabher ed. London, New York: Routledge, 1993. Hassink, R., Regional innovation policy: case studies from the Ruhr area, Baden-Württemberg and the North East of England. Nederlandse Geografische Studies 145, Utrecht, 1992. Heidenreich, M., Krauss, G., “The Baden-Württemberg production and innovation regime. Past successes and new challenges.” In Regional Innovation Systems, H.-J. Braczyk, Ph. Cooke, M. Heidenreich, eds. London: UCL Press, 1998. Hughes, T.P., “The Evolution of Large Technical Systems.” In The Social Construction of Technological Systems, W. E. Bijker, ed. Cambridge: MIT Press, 1987. ifo Institut für Wirtschaftsforschung, Der Wirtschafts- und Forschungsstandort BadenWürttemberg. München (ifo), 1995. IHK Industrie und Handelskammer Region Stuttgart, Potenziale und Hemmnisse in der Technologieregion Stuttgart. Analyse und Handlungsempfehlungen. Stuttgart, 2002. Kerst, Ch., Steffensen, B., Die Krise des baden-württembergischen Maschinenbaus im Spiegel des NIFA-Panels. Arbeitsbericht 49 der Akademie für Technikfolgenabschätzung in BadenWürttemberg, Stuttgart, 1995. Krauss, G., Les problèmes d’adaptation d’une économie régionlae forte: changement et inerties en Bade-Wurtemberg. In: Revue d’Economie Régionale et Urbaine 2; 1999: 353-376.
248
Gerhard Fuchs and Sandra Wassermann
Lay, G., Eggers, T., Rainfurth, C., Industrie in Baden-Württemberg im Wandel von der Produktion zur Dienstleistung? Fraunhofer Institut für Systemtechnik und Innovationsforschung. Karlsruhe; ISI, 2000. Maier, H.E., “Industrieentwicklung und Industriepolitik in Baden-Württemberg. Überlegungen zu den institutionellen Voraussetzungen differenzierter Qualitätsproduktion.” In Dezentrale Technologiepolitik? Technikförderung durch Bundesländer und Kommunen. Stadtforschung aktuelle 20., J. Hucke, H. Wollmann, eds. Basel, Boston, Berlin: Birkfellner 1989. Morgan, K. Reversing Attrition? The Auto Cluster in Baden-Wuerttemberg. Arbeitsbericht Nr. 37, Stuttgart (Akademie für Technikfolgenabschätzung in Baden-Württemberg), 1994. Piore, M. J., Sabel, Ch. F., The Second Industrial Divide: Possibilities for Prosperity. New York: Basic Books, 1984. Rehfeld, D., “Disintegration and Reintegration of Production Clusters in the Ruhr Area.” In: The Rise of the Rustbelt, Ph. Cooke, ed. New York: St. Martin’s, 1995. Sabel, C.F., “Flexible specialisation and the re-emergence of regional economies.” In Reversing industrial decline? Industrial structure and policy in Britain and her competitors. P. Hirst, J. Zeitlin, eds. Oxford: Oxford UP, 1989. Saebetzki, A., Perspektiven der Beschäftigung im Maschinenbau, in der Elektrotechnik und im Straßenfahrzeugbau bis zum Jahr 2010. ZEW-Wirtschaftsanalysen 1, 1994; 2:78-95. Schmitz, H., “Industrial districts: model and reality in Baden-Württemberg.” In Industrial districts and local economic regeneration, F. Pyke, W. Sengenberger, eds. Geneva: ILO, 1992. Schon, D.A., The Reflective Practitioner: How Professionals Think in Action. New York: Basic Books, 1983. Semlinger, K., Economic development and industrial policy in Baden-Württemberg: small firms in a benevolent environment. In: European Planning Studies 1, 1993; 435-463. Semlinger, K., Industrial-district-Politik in Baden-Württemberg. Zwischen Neubesinnung und Neuanfang. Arbeitsbericht Nr. 39, Stuttgart (Akademie für Technikfolgenabschätzung in Baden-Württemberg), 1994. Seufert, W., Informations- und Kommunikationsiwrtschaft räumlich stark konzentriert. DIWWochenbericht 32. Berlin, 2000. Statistisches Landesamt Baden-Württemberg, Region Stuttgart: Seit 1992 gingen 110000 Arbeitsplätze verloren. Eildienst Statistisches Landesamt Baden-Württemberg, Nr. 265/1997, Stuttgart (Statistisches Landesamt), 1997. Wallace, W., Rescue or Retreat? The Nation State in Europe, 1945 - 1993. Political Studies 1994; 42: 52-76.
Chapter 12 RETHINKING REGIONAL INNOVATION POLICY The making and breaking of regional history
Ron Boschma
12.1 INTRODUCTION One of the main issues regional science is dealing with is regional change. There are so many questions involved in this topic, each of which is hard to answer and deserves a paper of its own. It involves questions like: what is meant by change, what kinds of forms of change exist, do regions change (in other words, should we treat them as the appropriate unit of analysis, despite the fact that they are not actors like individuals or firms), how should we explain change, in what way do regions itself play a role in this (and at what level, be it local, regional, national, international), what is the role of public policy (is there any major role to play in a market economy?), is it possible to determine the rate and direction of change by policy makers, and if so, how? All these questions are related in one way or another to the major theme that is addressed in this edited volume: to what extent can regions diverge from established paths of economic development? The literature often suggests there exists some kind of tension between path dependence on the one hand (implying change along established trajectories), and radical breakthroughs on the other hand (implying new trajectories). Much emphasis is then laid on inertial structures that explain why radical change hardly comes about (and why incremental change is the rule rather than the exception). Nevertheless, new (regional) trajectories like Silicon Valley emerge now and then. How to explain them in a world where path dependence is paramount? What about public policy? Is there a role for public policy to play in making or breaking regional history, and if so, how? Yet, when regional policy is itself part of a lock-in process, how can it be capable of developing new trajectories (Boschma and Lambooy, 1999b)? This chapter has two objectives. The first objective is to clarify in detail the nature of regional change. It builds on previous work that sets out how
250
Ron Boschma
notions of evolutionary economics can be related to spatial issues (Boschma and Lambooy, 1999a). In essence, evolutionary economics has two sides. On the one hand, it seeks to explain why change is unlikely to occur due to inertial forces (lock-in) and, when it occurs, why change is of a minor, incremental, and gradual nature. Here we make use of a more unusual interpretation of path dependence to underline that new variety is largely directed and channeled by the existing environment. On the other hand, evolutionary thinking stresses that change may be radical and unpredictable. Selection mechanisms are of less relevance, while human agency, increasing returns, and institutional change play a much larger role in explaining the emergence of novelties. Path dependence here gets the usual meaning of an evolutionary process in which small events become magnified by positive feedbacks (Arthur, 1994; Garrouste and Ioannides, 2001). Elaborating on these aspects of evolutionary thinking in which history matters in different ways, a distinction is made between two ideal-types of evolutionary change, that is, localized and structural change. The second objective is to determine how regional policymaking can stimulate regional trajectories, given the role of path dependence. In a recent publication, we explored how evolutionary thinking can provide new insights to regional policy matters (Lambooy and Boschma, 2001). What are the objectives of such a policy based on evolutionary thinking, and what could be their main characteristics? It is also important to discuss the degrees of freedom policy makers may have to determine the future development of regions. Dalum et al. (1992) state that “implicit in evolutionary thinking there are hidden arguments in favour of non-intervention” (p. 298). So, is there room for effective regional innovation policy-making? Are policy makers largely constrained in stimulating new trajectories, or do they have the power to steer regional development in new directions? The distinction between the two ideal-types of regional change is considered essential for a better understanding of these policy issues. The structure of the chapter is as follows. Section 12.2 deals with the notion of localized change. It explores why regional change is often of a localized nature, that is, incremental, gradual, and cumulative. In Section 12.3, we set out the notion of structural change. Based on evolutionary thinking, it explains regional change in terms of chance, creativity, and structural adjustment. Firms and other organizations perform strategies that, deliberately or not, transform their environment. There is more chance involved in the emergence of new trajectories, which makes it rather unpredictable where and when these will develop. In Section 12.4, we explore the possibilities of making and breaking history. Section 12.4.1 sets out what a regional policy based on localized change may look like, and what are the real options for policy makers in this respect. Section 12.4.2 describes the implications when poli-
RETHINKING REGIONAL INNOVATION POLICY
251
cymaking is focused on new trajectories and structural change. Section 12.5 makes a few conclusive remarks.
12.2 LOCALIZED CHANGE Here, we attempt to define what is meant by localized change. This concept of incremental, cumulative change has become pervasive in evolutionary economics. We briefly explore why regional change is often of a localized nature by referring to notions like routines, path dependence, and selection. In short, localized change reproduces and strengthens existing structures. It takes place in a place-specific and historical context: it is shaped and constrained by its past (structural inertia may cause lock-in), it is systemic and interdependent, it involves (interactive) learning, it is of an incremental, cumulative and localized nature, and it tends to follow certain trajectories (defined in technical, economic and spatial terms) (Dosi et al., 1988). However, this does not imply some sort of historical determinism in which trajectories would move towards some pre-set ends: various, often unpredictable, kinds of reaction may still occur. Nevertheless, although new technologies are often unexpected, stochastic results of search activities undertaken by firms, the resulting changes are localized due to inertial forces. Why is regional change mainly incremental, cumulative, and localized? One reason is that local actors and organizations enjoy the opportunities and potential benefits offered by established trajectories. There is a widespread confidence that there are still many potential benefits to reap from improving existing technology in the near future, as reflected in growing markets, technological potentials, etc. (Freeman, Clark and Soete, 1982). This is reinforced by the fact that the surrounding local system (e.g. the social and institutional framework, the infrastructure) has committed itself to provide the necessary support. In this respect, the established trajectory acts as a stimulus for localized change. Another reason has to do with path dependence in terms of structural inertia. Notice that this interpretation of path dependence is somewhat different from the one given by David (2000), which refers “... to the idea of history as an irreversible branching process” (p. 8). Localized change is largely determined by its past which sets the constraints for future development. It provides an explanation for why it is hard for local systems to induce major changes, and why individuals and organizations in regions face many problems of adjustment when they want to develop or adopt something completely new (see Section 12.3). This is because actors and organizations are constrained by their internal structures, which are rule-bound. Moreover, they are part of a wider selection environment that is to some extent specific to
252
Ron Boschma
localities. In this respect, regions are treated as collective units because of common characteristics (such as a shared culture, a common economic history, etc.), that favor, or not, the dynamic capabilities of the local organizations. According to evolutionary economists, a major inertial force concerns the cognitive framework. At the level of the organization, learning is embedded in the memory of the organization, which is organization-specific, due to tacit knowledge (Nelson and Winter, 1982). It is because of evading uncertainty that organizations tend to undertake routinized behavior. Selection of information is rule-dependent. Searches for new technology are directed to technologies and markets with which firms have become familiar in the past (Heiner, 1983). Local problems and opportunities within existing routines trigger the search process. Certain elements of new information will be selectively left out, whereas other (more familiar) elements that strengthen own vested interests and safeguard feelings of security will be accentuated. Firms tend to ignore promising (or even superior) alternatives that deviate from their routines because high adjustment costs and risks are involved. It is for these reasons that new firms generally play such a pivotal role in the generation of new trajectories. At the level of the region, learning is regarded as a collective process of a localized nature (Camagni, 1991). Regional change is considered to proceed along paths, which form a framework of thought that sets constraints on the technological opportunities and guides the exploratory activities of local firms. The cognitive dimension is reflected in the techno-industrial specialization of an area (such as steel making) and embodied in the capabilities of local firms, other organizations, and skilled labor. They act as a filter for interpreting and assessing new opportunities. Due to its tacit nature, knowledge is shared only between local actors who have common experience: outsiders that lack the experience cannot effectively acquire it. In this respect, regional specialization may act as a constraint on the ability of local insiders to react to changing market signals or technologies. This is true for all the organizations that are part of the local system. It acts as a selection mechanism that guides and constraints creative and adaptive behavior. New variety that does not fit into the environment is bound to disappear. For example, local banks know how to assess the technical and market potentials of projects in the local activities where they have acquired experience. This is even truer when it concerns local firms with which they have built trust-based relationships for a long time. However, the same riskaverse behavior makes that new projects are not given a chance in such a local environment, because there is no knowledge and experience (and trust) to be built on. In other words, the local environment acts as a sort of selection
RETHINKING REGIONAL INNOVATION POL ICY
253
mechanism that may, or may not provide conditions favorable to meet the new requirements of change. There is also a systemic dimension involved. The collective learning process arises because of strong interaction patterns and cooperation networks between local organizations. These may consist of inter-firm relationships (either between competitors, or supplier-buyer linkages), or between firms and other organizations (such as knowledge institutions). According to Hakansson and Lundgren (1997), “there is a path dependency in the development of relationships and networks” (p. 122). The mutual interdependence of the parts of the local system may cause local inertia. When a region consists of a complex system in which each element has a structural position, change brings in instability because positions between elements are disturbed (Hannan and Freeman, 1977; Frenken, 2001). Only localized change which does not upset the functioning of the whole local system is therefore bound to happen. Grabher (1993) has mentioned the problem of political lock-in. This has not only to do with the fact that new developments, even within the very region in which these happen, are not always being perceived as a new way of organizing or producing. People (such as politicians and business executives) notice much later that fundamental changes occur. When they react to new changes, they will also do so in a very routinized (that is, continuous, incremental, and localized) manner (Herrigel, 1993). This is because they might fear losing control, or they have obligations towards other actors in the network. In old industrial regions, established corporations, labor unions, and public authorities tend to oppose new developments in order to secure their own vested interests. Economists like Arthur and David refer to the term “lock-in” as “... the entry of a system into a trapping region – the basin of attraction that surrounds a locally (or globally) stable equilibrium” (David, 2000:10). Lock-in is reinforced by the fact that investments in plants, equipments, labor, organizations, etc. constitute assets that are not easily transferable to other tasks or functions (Hannan and Freeman, 1977; Antonelli, 1995). In other words, high sunk costs prevent organizations, networks, and regions to abandon a previously chosen path, even when they know higher returns may be realized in other unrelated activities. Consequently, the world consists of many specialized islands with very specific and idiosyncratic features (Storper, 1992). When local advantages are cumulative (e.g. localized learning processes with a tacit dimension) and systemic (requiring a specific institutional context), entry barriers for other places are too high (Perez and Soete, 1988). In these circumstances, it is impossible to imitate and copy these successful areas by other regions that lag behind. Since selection does not destroy all variety, there exists a plurality of development paths, or multiple equilibria: economic systems are unlikely to move into a state of best-practice level. Therefore, the
254
Ron Boschma
competitive advantage of successful regions is less likely to be threatened by the rise of potential rivals competing on similar technologies. The greatest threat to their survival may be an internal one: the more specific, the more successful, and the greater the risk of lock-in in the longer run. In sum, localized change is the outcome of a strong selection process: it is therefore local, gradual, and cumulative. It favors a tendency of dynamic stability due to inertial forces. Localized change tends to reproduce existing structures, because it is focused on optimizing the “fit” into the environment (Grabher, 1993). It is typical for a lock-in situation: only change is allowed and awarded that closely fits into the local environment. Adaptation may be perfect as long as the market circumstances do not require any fundamental change. When they do, however, structural problems of adjustment come into being, and the lock-in situation becomes a negative one. We turn to this topic now.
12.3 STRUCTURAL CHANGE Structural change is different from localized change because it requires structural adjustment at all levels (technical, economic, institutional). Gersick (1991) has remarked that it is “like the difference between changing the game of basketball by moving the hoops and changing it by taking the hoops away” (p. 19). In short, structural change is associated with a new type of development that reflects fundamental change and thus, a break from the past. Contrary to the case of localized change, the selection environment is unable to act as a localized stimulus for structural change. Due to strong negative selection pressures, structural change depends heavily on the process of creativity in order to develop: it requires human agency to shape and adapt the local environment to the needs of the new breakthroughs. This also leaves room for chance events to occur: it is rather uncertain and unpredictable where structural change will take place. Structural change has always something to do with dramatic change, something that did not exist before. Due to its novel nature, it is accompanied by problems of adjustment, not only in technical but also in organizational, economic, and institutional terms. This type of change may start and develop in the region itself, or it may be brought in from elsewhere through various channels. In the former case, local actors are involved, and their actions become fully embedded in the region after the required structural changes have been accomplished. In the latter case, it may mean imitation of a basic technology that has been widely used elsewhere, but which gives many problems of implementation because the technology has been taken out of its local context. A fine example is provided by Gertler (1997) who described in detail
RETHINKING REGIONAL INNOVATION POLICY
255
the many difficulties Canadian companies were confronted with when implementing and operating complex German technology in their machinery, because of different national cultures. Only after structural change has been achieved, the basic technology becomes integrated in the region and is able to contribute to its competitive advantage. As noticed before, the selection environment is generally hostile to new breakthroughs. Nevertheless, they take place now and then. Nobody had ever heard of computers before the Second World War, it is now one of the leading industries in the world, with many applications (including software and Internet applications) providing employment to millions of people. There are numerous historical examples of this kind (Boschma, 1997). Before new trajectories can develop, they have to overcome many problems, such as shortages of knowledge, skills, capital, clients, inputs and institutions. This means either establishing new organizations (e.g. new firms, new knowledge institutions, new educational facilities, new suppliers of capital), or restructuring existing organizations. Population ecology studies demonstrate that established organizations have a poor capacity to change in a sufficient manner when confronted with external changes (Hannan and Freeman, 1989; Staber, 1998). In either way, there is a strong need for structural change in the surrounding environment (Freeman and Perez, 1988). Since new high-tech firms require new resources (such as knowledge, skills, institutions) that are unlikely to be available, firms have to create their own mechanisms to satisfy their needs. They do so by creativity: individuals and firms are very active in creating or mobilizing their necessary but missing resources themselves. In order to finance their capital needs, they cannot rely on established financial institutions, because these show routinized (that is, conservative and risk-averse) behavior. Therefore, they perform alternative strategies, such as drawing on family capital, or reinvesting own profits as a major source of investment capital (Oakey, 1993; Hayter, 1997). In the case of (technical) labor skills, for example, firms set up their own training facilities and research departments, or attract skilled labor from outside (Boschma and Van der Knaap, 1999). Thus, new basic variety brings about structural change, because firms actively shape and transform the environment in accordance to their own needs. If this would not occur, new variety that strongly deviates from the surrounding environment would not emerge. In other words, firms co-determine the selection process (Metcalfe, 1994b). However, the local environment is not only upgraded by “matching” strategies through human learning and strategic behavior of entrepreneurs, organizations, and governments. It will also become more favorable as a (non-intended) result of the local growth process itself, bringing about external economies of scale and scope in the area involved. This leads to increasing variation, resulting in increasing returns for individual firms or sectors and other
256
Ron Boschma
parts of complementary, associated networks. This may be associated with market-size external economies (deep division of labor, specialization) and the local accumulation of (tacit) knowledge (or technological spillovers). For example, the local labor market will be upgraded and become more diverse, because the new demand will induce migration into the area and stimulate (public or private) educational institutes to set up new courses (Boschma and Van der Knaap, 1999). Moreover, when the new high-tech firms grow and make profits, specialized local suppliers of venture capital may mushroom in the area because an interesting and expanding market has emerged (Smith and Florida, 2000). This also applies to the institutional framework. Several misunderstandings should be avoided when we talk about structural change. First, it is not a question of pace of change. Structural change is not abrupt or sudden, but evolves slowly. In general, new trajectories emerge and develop gradually: it takes time to build complementary conditions and a favorable environment. Economic historians know this all too well: they have objected to the notion of Industrial Revolution, a term that did not describe well what was going on in the eighteenth and nineteenth century in Britain. For example, the Watt steam engine did not emerge as a kind of “manna” from heaven. It was the result of previous generations of steam engines (like the Savery engine) that were developed and improved. Moreover, it was slowly implemented in sectors like cotton, coal mining, and railways (von Tunzelmann, 1978; Mokyr, 1990). It required many complementary new technologies and organizational changes before it could be effectively implemented. Therefore, new breakthroughs (with accompanying structural changes) are often the result of a long and slow evolutionary process. A second misunderstanding is related to the fact that structural change is often believed to start from scratch. We have stated above that the environment is rather hostile to new breakthroughs, because there is no experience to build on. However, new trajectories may respond to local demands and draw on existing structures and conditions (such as labor, capital, markets), and put these together with new elements in new configurations. Breshanan et al. (2001) have observed this in new ICT-related clusters during their start-up phase in many parts of the world. At first sight, this seems strange because we would believe there is a large mismatch (or low degree of fit) between on the one hand the requirements of radical breakthroughs, and on the other hand, the available structures and conditions. Although this is true in a sense, we have to be aware of the fact that the requirements of radical breakthroughs are not well specified from the beginning. This implies there is not a big gap involved which just needs to be closed. This would also mean a too static approach for a very dynamic process. New trajectories build on generic instead of specific conditions. In this respect, the process of structural adjustment concerns the transformation of
RETHINKING REGIONAL INNOVATION POLICY
257
generic conditions (e.g. basic knowledge, skills, inputs, etc.) into specific ones (that is, e.g. specialized knowledge and skills). Many examples in economic history can be given, in which new trajectories were founded on generic conditions. This seems especially true for engineering complexes, which have provided generic knowledge and skills to many new trajectories in the past (with low entry barriers in terms of technological knowledge, skills, and capital) (Rosenberg, 1982). For example, in the late nineteenth/early twentieth century, new industries like automobiles and electrical engineering could build on generic conditions like a tradition of metalworking, engineering and instrument trades, from which were drawn experienced labor, entrepreneurs, parts and components needed in the new industries (Boschma, 1997). There is more chance involved in structural change than in localized change, for three reasons. First, as noticed before, there is a mismatch between the requirements of new trajectories and the regional environment. This makes it difficult to predict where these will emerge: the local environment has to be transformed before it can really contribute to the competitive advantage of the firms involved. That is, structural change is required. When the requirements of new trajectories are also not well specified during their initial stage of development, it becomes even more uncertain to predict. Second, following Arthur (1994) and Klepper (2002), the spatial formation of new industries depends on historical accidents (such as an early spin-off process), in combination with increasing returns. During the initial stage of development, it is therefore hard to tell beforehand which region will take the lead. However, once one or a few regions take this lead, agglomerations economies (most likely localization economies) will come into being, which makes the regional growth process self-reinforcing. This refers to the many cumulative, self-reinforcing advantages that local firms may enjoy. For example, the more growth, the more complementary, supportive firms (such as specialized suppliers), and other organizations (like R&D-facilities) are generated. Third, the generic resources mentioned above, on which new trajectories may build, are expected to be widely available in space because they do not (yet) have any specific features (Boschma, 1997). Following this line of thought, the actual places of new industries, which shape the evolution of a spatial system for a considerable period, are rather unpredictable. The idea of many potential outcomes with a high degree of uncertainty stands in marked contrast to the relative predictability of the places where localized change evolves. In sum, structural change is an evolutionary process, which goes along with problems of adjustment. In Table 12.1, the differences with localized change are summarized. Selection tends to be rather weak because there are no specific resources in the local environment to benefit from. At most, new
258
Ron Boschma
trajectories can draw on generic resources. Thus, contrary to localized change, the selection environment is unlikely to act as a localized stimulus for structural change. Consequently, structural change depends heavily on the process of creativity in order to develop: the local environment has to be transformed before it can contribute to the further development of new trajectories. The environment is gradually transformed and adjusted, due to deliberate, intended actions of local organizations on the one hand, and the spontaneous growth process itself on the other hand. Thus, structural change concerns the transformation of generic resources into specific ones. As there are more spatial options in the case of structural change (due to chance events, increasing returns and the importance of generic conditions), it is unpredictable where the new development paths become manifest. Finally, both idealtypes of change allow for the potential inefficiency of market outcomes: localized change may develop into a state of lock-in, while early lock-ins of inferior regions may occur in the case of structural change, due to chance events and increasing returns. In the foregoing, we have stressed the features of an ideal-type of structural change. However, the degree of structural change required may vary between sectors/technologies, which implies entry barriers for firms and regions may be different. For biotechnology based on genetic engineering, there will be fewer potential candidates, because advanced scientific knowledge is required. In contrast, other sectors with lower entry barriers, such as the software sector, may have more potential candidates, that is, regions that are well endowed with the required generic conditions (such as people with basic knowledge of computers). The same applies to the catching-up process. For example, complex technologies that are imitated or imported require more absorptive capacity of regions and more restructuring of local organizations than other basic technologies. One of the challenges of future research is to assess for each sector/technology whether structural change may take place everywhere. This means determining which regions are likely candidates and, at the same time, clarifying why it is impossible to predict which of the likely candidates will be the successful one. This is also relevant for
RETHINKING REGIONAL INNOVATION POLICY
259
regional policy-makers, because it determines to a considerable degree their options to intervene successfully in regional development. To this topic, we turn now.
12.4 RETHINKING REGIONAL POLICY Regional policy presupposes the possibility to intervene in the process of regional development. More in particular, public intervention is focused on a self-defined path of development, different from what is being perceived as the current one. Evolutionary economics could be helpful to formulate regional innovation policy making, since it deals with the long-term processes of structural change, with the increasing variety of technology and organizations, and with the strategies of economic actors to adapt to changing structures. In the remaining part, we go more into detail what regional innovation policy based on evolutionary thinking could mean. This is not an easy task, because there is hardly any literature on this topic (Lambooy and Boschma, 2001). Since evolutionary economics acknowledges the importance of routinized behavior and rigid institutions, the main policy objectives concentrate on stimulating economic change and on enhancing the capability of all organizations to contribute to the innovation process. It is helpful in this respect to make use of the two ideal-types of regional change set out in the previous sections. A preliminary attempt is made to sketch for both types of change the main objectives and features of regional policy making on the one hand, and the degrees of freedom policy makers may have to determine the future development of regions on the other hand. First, we discuss the regional policy implications of localized change. Then, we set out the policy options for stimulating structural change. As such, we link the two types of regional change to two ideal-types of regional innovation policy.
12.4.1 Localized change and regional policy When dealing with localized change, regional policymaking may have the following features. In essence, it is location-specific (accounting for the specific context and the particular needs), it is merely fine-tuning, it focuses on the strengthening of dynamic connectivity between the elements of the regional system, and it puts efforts to reduce the risk of negative lock-in (e.g. through loosening tightly coupled networks). In these circumstances, local policy makers have few degrees of freedom, but are more likely to be suc-
260
Ron Boschma
cessful as long as their actions are localized, that is, focused on reproducing and strengthening the existing structures. In other words, the local environment determines to a large extent available options and probable outcomes of regional policy. As described in Section 12.2, evolutionary economics has a specific interest in how historical paths of development affect change. Therefore, regional policy cannot start from scratch. It has to be sensitive to local trajectories, accounting for the particular technological, economic, and institutional context. Evolutionary policy should take regional variety as a point of departure, that is, it should be based on a bottom-up strategy attuned to the needs and resources of regions (Morgan, 1997; Cooke, 2001). Such an approach accounts for the fact that the development potential of regions differ. This implies that a diversity of policies that allows for a variety of development paths is more likely to generate satisfactory and effective results. This view is fundamentally different from standardized top-down solutions common in Keynesian and neoliberal thinking (Amin, 1998). This does not necessarily mean that regional policy should only focus on the local level. As the literature on national innovation systems has taught us (Nelson, 1993), many elements of the institutional context (such as labor market policies, cultural norms and values, etc.) relevant for particular trajectories are determined at the national, or even international level (Gertler, 1997). Moreover, lagging regions often lack the resources to implement location-specific policy. For this reason, Cooke and Morgan (1998) have claimed the need to complement the region-specific strategy by a top-down approach (at the national or supranational level) that focuses on redistribution mechanisms. However, what should such a regional policy based on localized change look like? Three main objectives of such a policy may be identified: (1) to stimulate the efficiency of the selection mechanism; (2) to realize dynamic connectivity between local actors; and (3) to reduce the risk of lock-in by stimulating openness and diversity in the local economy. First, a major objective for evolutionary policy makers is to ensure that the selection process takes place efficiently. The development and diffusion of new variety in an economic system may be achieved through efficient selection mechanisms, which include both market and non-market factors. Policy measures are directed to minimizing problems of adjusting to change (Metcalfe, 1994b). Here again, history matters. The degrees of freedom to act are related to the argument that the specific local, regional, and national patterns of institutions may offer opportunities for regional authorities to influence the direction of their development paths. In fact, the institutional configuration of countries and regions is often regarded as quite durable, embedded in a specific culture that is, in general, slow to adapt (Perez, 1983).
RETHINKING REGIONAL INNOVATION POLICY
261
In such circumstances, policy becomes more of a “fine-tuning exercise” (Carlsson and Jacobsson, 1997). It is focussed on reproducing and strengthening the local system. Policymaking is directed at upgrading the surrounding environment where found necessary. It may focus on strengthening weak elements in the local context, it may respond to particular local problems, or it may reinforce local interdependencies (enhancing a collective learning capacity). In this respect, policy makers not only influence the structure of the organizations, such as raising the quality of the education system, but also affect the interaction patterns of organizations. For instance, policy efforts may be geared towards stimulating close industry-university links in the field of research and education. This brings us to the second objective of evolutionary policymaking. It concerns stimulating connectivity in (regional) innovation systems, in order to enhance learning and innovation. This view builds on ideas of Lundvall (1988) on the importance of interactive learning as a source of innovation, which may be realized by stimulating the coupling between firms and other organizations (Metcalfe, 1994a). It acknowledges the fact that the production and diffusion of innovations do not occur in isolation, but require complementary organizations (such as venture capitalists, research institutes, technical schools, etc.) whose actions need to be coordinated. Here, policymaking should ensure that all organizations that make up the system span all the necessary range of activities (that is, none are missing or underdeveloped), and that these organizations interact intensively. It is far from sure whether policy makers should actually play a role in such innovation systems. Private actors could fulfill many of these roles (such as supplying venture capital, educating skilled labor, and acting as intermediaries providing information), especially when the system is expanding. Nevertheless, Metcalfe (1994b) stresses the necessity of the policy maker to realize connectivity between the constituent parts of the innovation systems because the objectives and interests of the actors are diverse. The government may then play the role as “broker” and network facilitator. The third objective of regional policy should focus on reducing the risk of regional lock-in. As explained earlier, localized change and path dependent processes may cause lock-in situations that result in sub-optimal outcomes. This comes down to opposing “a situation where an agent is aware of a ‘better’ solution but avoids it” (Magnusson et al., 1997:3). In highly specialized regions, “localization economies” may stimulate innovative behavior along existing trajectories, but tend to hinder regional adaptability, and thus the development of something completely new. This may be attributed to a long-standing commitment of the local environment towards “old” technologies in terms of production, investments, skills, industrial networks, and organizations like R&D centers.
262
Ron Boschma
In order to avoid such lock-in, regional policy should focus on stimulating openness and diversity in local systems. This would make industrial regions more equipped to introduce and adopt qualitative change as a source of growth (Saviotti, 1996). Policy makers should ensure a local power of balance, which prevents powerful organizations to take control of the local system and use it only for their own reproduction (Herrigel, 1993). Grabher and Stark (1997) suggest that loosely coupled networks may provide a solution: “... a loosely coupled network is a good system for localized adaptation. If the elements in a system are loosely coupled, then any one element can adjust to and modify a local contingency without affecting the whole system. A second advantage is that loosely coupled networks preserve many independent sensing elements and therefore “know” their environment better. Third, in loosely coupled networks where the identity and separateness of elements is preserved, the network can potentially retain a great number of mutations and novel solutions than would be the case with a tightly coupled system” (p. 538). What are the options for public policy to determine the future development of regions in a spatial system in which selection forces are strong mechanisms? In an evolutionary world, policy makers have to deal with a great deal of uncertainty. Therefore, they do not optimize. They pursue a policy of trial-and-error, and they learn and adopt in the light of experience (Lambooy and Boschma, 2001). As a consequence, they prefer a policy that is focused on localized change. There is less risk involved, local support will be much stronger, and the guarantee of success may be higher. It implies that policy makers may be successful as long as their actions are localized. This means that the potential impact of regional policy may be large when their policy objectives are strongly embedded in the surrounding environment, building on regional competencies. As firms, policy makers are more likely to fail when their local strategies deviate considerably from the local context. Consequently, the options for policy makers to change fundamentally the course of regional development are expected to be rather limited. This is shown by the many regions that wanted to become a “New Silicon Valley,” but failed to do so. As noted above, it has to be sensitive to local trajectories. However, the prospects for public policy with respect to the third objective, that is, avoiding lock-in, may be not that large. This is because public authorities are often sensitive to the interests of the representatives of the “old” technologies. In such circumstances, the local government is expected not to be much interested in pursuing a policy of change and solving the problem of lock-in. Moreover, to enhance interactive inter-organizational learning, peripheral regions have to build on their local network capacity in order to make their renewal policy successful (Storper, 1995). This requires a culture of trust and strong public bodies, but these are exactly the weakest
RETHINKING REGIONAL INNOVATION POLICY
263
features of many backward regions (Boschma and Lambooy, 1999b). It remains hard to tackle these structural problems, especially when the local public authorities themselves are part of the problem. We come back to this topic in the next section.
12.4.2 Structural change and regional policy This section discusses the possibilities for regions to develop successful strategies that focus on structural change. When dealing with this type of change in which chance and creativity play a role, regional policy may have the following features. In essence, it is generic (accounting for diversity, as potential sources of new variety, as found in diversified urban regions), its main goal is the restructuring of the institutional framework, it aims on building new (regional) systems (stimulating new connections), and it is focused on avoiding early lock-ins of sub-optimal technologies. In these circumstances, policy makers have more degrees of freedom (more policy options) than in a world of localized change, but there is more uncertainty involved in the actual outcomes of regional policy. Chance events and experimentation, among other things, may be responsible for this. In other words, the local environment provides many policy options but gives few clues about the real outcomes of regional policy. In Table 12.2, we have summarized the main differences between the two ideal-types of regional innovation policy that focus either on localized change or on structural change.
Since the selection environment is unlikely to provide specific stimuli in the case of structural change, it is less meaningful to account for the locationspecific context as a starting point for regional policy. As noticed in Section 12.3, new trajectories can at most build on generic conditions. The aim of regional policy here is to ensure diversity in order to increase regional
264
Ron Boschma
adaptability, to keep open a multiplicity of potential resources for new trajectories, and to allow for new, unexpected combinations. Especially so-called diversified urban regions with products and services in almost all sectors are likely candidates for structural change, because they have the advantages of flexibility based on urbanization externalities associated with the proximity of actors from many diverse industries and organizations (Lambooy and Boschma, 2001). They are the centers of communication, they host a large diversity of competences in individuals, firms and other organizations, and they offer attractive living environments that retain or attract the innovators and the “brainworkers.” Specialized regions with a large exposure to the outside world may also (but are less likely to) fulfill the requirement of a diversified environment. In this respect, public policy focuses on stimulating “redundant capacities” in agglomerations, that is, “... the availability of unspecific and uncommitted capacities that can be put to a variety of unforeseeable uses” (Grabher, 1993:265). In short, regional policy that focuses on structural change may have the following features: (1) restructuring the institutional framework; (2) stimulating new connections between (new or restructured) organizations; (3) avoiding early lock-ins of sub-optimal technologies. Since new trajectories require structural change (as described in Section 12.3), the intervention of regional policy should be oriented to transforming the local environment. This implies the creation of new organizations and institutions (as well as the restructuring of old ones) that provide new knowledge, capital, technical education, rules, laws, etc. (Freeman and Perez, 1988). During the very early stages of the emergence of new trajectories, a key task of public policy may be to identify new technological opportunities as soon as possible, and to anticipate the institutional changes required. Such a pro-active and flexible attitude of policy is often called for since organizations and institutions are unlikely to adapt spontaneously. However, as stated in section 12.3, non-public organizations, like firms (such as suppliers of venture capital, or providers of technical and market information), may fulfill these roles as well. In sum, the fortunes of regions depend largely on the ability of their (public and private) organizations to restructure their activities in accordance to the needs of the new trajectories. In Section 12.3, we explained that such a restructuring of organizations and institutions is far from easy to establish. We need to be more precise here (Edquist and Johnson, 1997). With respect to organizations, it may be easier to create new organizations (such as new firms, new schools, or new universities) than reforming established organizations (due to their routines). With respect to institutions, formal institutions like laws and regulations (e.g. enabling genetic engineering) may be more easily modified. However, required changes in informal (hard to observe) institutions (such as norms and values)
RETHINKING REGIONAL INNOVATION POLICY
265
are more difficult to realize, such as promoting entrepreneurship or building stocks of social capital (in order to enhance inter-organizational learning) in backward regions. However, such a restructuring process is a necessary but not a sufficient condition. Public policy has not only to make sure that the complementary elements of the system are fully developed, it should also take care of establishing the connections between them. In other words, a main task of policy makers is to build-up a new system in which the main (local and non-local) organizations are well connected. Once again, the government is only part of a whole set of actors (such as firms, banks, universities, schools, industry associations, etc.) that is involved in the formation of such a new system. As Carlsson and Jacobsson (1997) have put it, “a new technological system may not evolve from the prevailing ones without substantial intervention, although not necessarily by the government, or the government only” (p. 285). As noticed above, it is essential to create or preserve diversity in regional economies, in order to keep open a wide range of opportunities for regional development. A task of policy makers here is to avoid early lock-ins of inferior technologies or development paths that are produced by historical accidents and increasing returns (David). According to Cohendet and Llerena (1997), this may be achieved by launching different research programs, by encouraging experimental competition, and by protecting some weaker, but potential options (in which the government may act as the main user of competing technologies, such as alternative energy systems). In this latter case, the selection process is retarded deliberately, in order to prevent the early exclusion of options that may have a large potential (Herrigel, 1993). Remark that avoiding the risk of lock-in was also part of a policy based on localized change. In that case, however, it focuses on all kinds of inertial forces that have emerged alongside the development of the dominant trajectory in the local environment, at the detriment of regional adaptability and fundamental change. Here, avoiding lock-in as a policy tool has a different meaning: it aims at preventing the early exclusion of new trajectories with major growth potentials. However, avoiding early lock-ins may go at the expense of economic efficiency: it may be extremely wasteful to maintain too many alternatives for some time. A problematic task of public policy is to find a balance here. On the one hand, it should focus on maintaining diversity, favoring potential sources of new variety, but which goes at the expense of static efficiency (Saviotti, 1996). On the other hand, it should eventually select one of the alternatives, preventing economic waste, stimulating increasing returns, but risking the early selection of a sub-optimal outcome (Arthur, 1994). In other words, as Edquist (1997) puts it, “it may be fatal to exclude options too early and wasteful to do so too late. Doing it too early may result in closing what
266
Ron Boschma
may potentially be the most fruitful trajectories and working instead along those that are inferior. Doing it too late may absorb excessive amounts of resources” (p. 27). Having discussed the main objectives of a regional policy based on structural change, we end our discussion considering the degrees of freedom policy makers may have in stimulating new trajectories. There are contradictory options for policy makers in this respect (Lambooy and Boschma, 2001). On the one hand, one could argue that it is beyond the control of planners. Due to chance events and increasing returns, there are no certainties where new breakthroughs have their main locational impacts (Boschma and Lambooy, 1999a). In many cases, new trajectories developed first in unexpected locations, often in non-core regions (Boschma, 1997). Moreover, small changes can result in large deviations from the expected outcomes of a planning process (Kemp, 1997). As a result, policy makers have to deal with fundamental uncertainties, even more so than in the case of localized change. New development paths are almost impossible to be planned: they are almost without exception the result of unplanned human action. Even if they could be planned, regional authorities have to deal with the problem how to scan and evaluate new basic technologies as the potential driving forces for their regional economies. On the other hand, given the scope for human action, we believe policy makers may have an important role to play. Since there is a strong need for restructuring the institutional framework (laws, regulations, etc.), public authorities have a considerable power to provide the basic requirements for new trajectories. Moreover, since space is only of little influence (at most, generic conditions may be a stimulus), there is room for policy makers to act and build-up a favorable environment for new trajectories. This may be especially true for local strategies that are directed to the development of new industries that display increasing returns, with low entry barriers (Arthur, 1994). These highly dynamic and knowledge-intensive sectors build up their own increasing returns in situ as their development proceeds. Although urbanization economies may offer advantages of flexibility secured by a diversity of activities, even inferior places may be likely candidates, because creative behavior of organizations may compensate for the lack of a favorable environment. Thus, regional policy makers may certainly play a role, but it is uncertain whether they succeed in developing new paths: even with optimal strategies, there is no guarantee for being the region in which the new development paths will be located.
RETHINKING REGIONAL INNOVATION POLICY
267
12.5 CONCLUSION Evolutionary regional policy is preoccupied with making and breaking regional history. From an evolutionary perspective, it may be argued that regional policy makers have a considerable role to play. For example, regional policy is more likely to be successful when its political objectives are strongly embedded in the surrounding environment, that is, attuned to the resources and needs of specific regions. On the other hand, regional innovation policy is often confronted with many problems and uncertain outcomes. Options for regional policy are quite limited. This is especially true when policymakers aim at creating new regional trajectories, and when they attempt to abandon or escape the development paths of the past. Uncertainty is then paramount. In general, there is no certainty about where new trajectories will have their main locational impacts, given the importance of chance events, creativity, and increasing returns. Moreover, regional development is often a spontaneous process in which many actors and organizations (including public ones) are involved. Here, a first attempt was made to sketch the contours of a regional innovation policy that is based on two different ideal-types of regional change. In general, evolutionary policy aims at stimulating the responsiveness of local actors to change (given the importance of routinized behavior and inertial forces), bringing together public and private stakeholders, and avoiding risks of lock-in (stimulating diversity). However, it makes a big difference when policymaking focuses on localized rather than structural change. It has not only implications for the main objectives of regional innovation policy; it has also impact on the degree of freedom policy makers may have to induce regional change. However, there are several issues associated with such a distinction between two ideal-types of regional innovation policy that need to be worked out. First, the impact of the institutional context has been treated in a rather simple way. To put it shortly, we argued that regional policy might be more successful in stimulating localized change when it accounts for the institutional context (at the local and national level). In the case of structural change, we claimed that regional policy should focus on restructuring the institutional framework. However, it ignores the question whether particular institutional environments may be more responsive to structural change rather than localized change. For example, Hall and Soskice (2001) argue that so-called liberal market economies (as found in the USA) provide more institutional support for radical innovations because of the presence of more switchable assets. Labor markets are not only more fluid because of fewer restrictions on layoffs, they also encourage the formation of pools of general skills (that is, highly portable skills that are not tied to a particular firm or
268
Ron Boschma
industry) because this is less risky for people in the case of high job insecurity (Estevez-Abe et al., 2001). In such economies, financial markets (with extensive equity markets and many venture capitalists) are also more responsive to the needs of firms that develop or implement radical innovations. This suggests that structural change is more easily to realize in liberal market economies, because radical innovations require less problems of adjustment. Second, much effort has been put in explaining why it is so difficult to pursue an effective regional innovation policy. The main challenge is, however, to explore how such problems may be tackled by intelligent policy making in lagging and backward regions. In the case of peripheral regions, a policy based on structural change may be directed to basic conditions like restructuring the local school system, building networking capacity, or stimulating the entrepreneurial climate. This is different from a policy based on localized change that builds on specific strengths, like existing specializations or specific skills. However, there is a strong need for elaborating on these policy matters in more concrete terms, before the results can become useful for policymakers. For example, what should a location-specific policy look like, and what kind of policy measures should be recommended? These are obvious questions, but far from easy to answer.
REFERENCES Amin, A., An institutionalist perspective on regional economic development. Paper presented at the Department of Geography UCL. London, 1998. Antonelli, C., The economics of localized technological change and industrial dynamics. Dordrecht/Boston/London: Kluwer Academic Publishers, 1995. Arthur, B., Increasing returns and path dependence in the economy. Ann Arbor: University of Michigan Press, 1994. Boschma, R.A., New industries and windows of locational opportunity. A long-term analysis of Belgium. Erdkunde 1997; 1:1-19. Boschma, R.A., van der Knaap G.A., New high-tech industries and windows of locational opportunity: the role of labour markets and knowledge institutions during the industrial era. Geografiska Annaler 1999; 2:73-89. Boschma, R.A., Lambooy J.G., Evolutionary economics and economic geography. Journal of evolutionary economics 1999a; 9:411-429. Boschma, R.A., Lambooy J.G., The prospects of an adjustment policy based on collective learning in old industrial regions. GeoJournal 1999b; 49:391-399.
RETHINKING REGIONAL INNOVATION POLICY
269
Bresnahan, T., Gambardella A., Saxenian A., “Old economy” inputs for “new economy” outcomes: cluster formation in the new Silicon Valleys. Industrial and corporate change 2001; 4:835-860. Camagni, R., ed., Innovation networks. Spatial perspectives. London/New York: Bellhaven Press, 1991. Carlsson, B., Jacobsson, S., “Diversity creation and technological systems: a technology policy perspective,” In Systems of innovation. Technologies, institutions and organizations, C. Edquist, ed. London/Washington: Pinter, 1997. Cohendet, P., Llerena P., “Learning, technical change and public policy. How to create and exploit diversity.” In Systems of innovation. Technologies, institutions and organizations, C. Edquist, ed. London/Washington: Pinter, 1997. Cooke, P., Regional innovation and learning systems, clusters, and local and global value chains. Paper presented at the International Workshop “Innovation clusters and interregional competition”; 2001, November 12-13; Kiel Institute of World Economics. Cooke, P., Morgan K., The associational economy. Firms, regions, and innovation. Oxford: Oxford University Press, 1998. Dalum, B., Johnson B., Lundvall, B.A., “Public policy in the learning society.” In National systems of innovation. Towards a theory of innovation and interactive learning, B.A. Lundvall, ed. London: Pinter, 1992. David, P.A., Path dependence, its critics and the quest for “historical economics.” Working paper. Oxford & Stanford University, All Souls College, June 2000. Dosi, G., Freeman C., Nelson R., Silverberg G., Soete L., eds., Technical change and economic theory. London: Pinter, 1988. Edquist, C., “Systems of innovation approaches. Their emergence and characteristics.” In Systems of innovation. Technologies, institutions and organizations, C. Edquist, ed. London/Washington: Pinter, 1997. Edquist, C., Johnson, B., “Institutions and organizations in systems of innovation.” In Systems of innovation. Technologies, institutions and organizations, C. Edquist, ed. London /Washington: Pinter, 1997. Estevez-Abe, M., Iversen, T., Soskice, D., “Social protection and the formation of skills. A reinterpretation of the welfare state.” In Varieties of capitalism. The institutional foundations of comparative advantage, P.A. Hall, D. Soskice, eds., Oxford: Oxford University Press, 2001. Freeman, C., Clark J., Soete L., Unemployment and technical innovation. A study of long waves and economic development. London: Frances Pinter, 1982. Freeman, C., Perez C., “Structural crisis of adjustment: business cycles and investment behaviour.” In Technical change and economic theory, G. Dosi, C. Freeman, R. Nelson, G. Silverberg, L. Soete, eds. London: Pinter, 1988. Frenken, K., Understanding Product Innovation using Complex Systems Theory, Ph.D. Dissertation, University of Amsterdam and University of Grenoble, 2001.
270
Ron Boschma
Garrouste, P., Ioannides, S., “Evolution and path dependence in economic ideas: past and present.” In Evolution and path dependence in economic ideas. Past and present, P. Garrouste, S. Ioannides, eds. Cheltenham: Edward Elgar, 2001. Gersick, C.J.G., Revolutionary change theories: a multilevel exploration of the punctuated equilibrium paradigm. Academy of Management Review 1991; 1:10-36. Gertler, M.S., “The invention of regional culture.” In Geographies of economies, R. Lee & J. Wills, eds. London: Arnold, 1997. Grabher, G., ed., The embedded firm. On the socioeconomics of industrial networks. London/New York: Routledge, 1993. Grabher, G., Stark, D., Organizing diversity: evolutionary theory, network analysis and postsocialism. Regional studies 1997; 5:533-544. Hakansson, H., Lundgren A., “Paths in time and space – path dependence in industrial networks,” In Evolutionary economics and path dependence, L. Magnusson, J. Ottosson, eds. Cheltenham: Edward Elgar, 1997. Hall P.A., Soskice, D., “An introduction to varieties of capitalism.” In Varieties of capitalism. The institutional foundations of comparative advantage, P.A. Hall, D. Soskice, eds. Oxford: Oxford University Press, 2001. Hannan, M., Freeman J., The population ecology of organizations. American journal of sociology 1977; 5:929-964. Hannan, M., Freeman J., Organizational ecology. Cambridge: Harvard University Press, 1989. Heiner, R.A., The origin of predictable behaviour. The American economic review 1983; 73:560595. Hayter, R., The dynamics of industrial location. The factory, the firms and the production system. Manchester: John Wiley, 1997. Herrigel, G.B., “Power and the redefinition of industrial districts. The case of BadenWurttemberg.” In The embedded firm. On the socioeconomics of industrial networks, G. Grabher, ed. London/New York: Routledge, 1993. Kemp, J., New methods and understanding in economic dynamics. An introductory guide to chaos and economics. Economic Issues 1997; 1:1-27. Klepper, S., The evolution of the U.S. automobile industry and Detroit as its capital. Paper presented at Congress of the International Joseph A. Schumpeter Society; 2002 March; Gainesville, Florida. Lambooy, J.G., Boschma, R.A., Evolutionary economics and regional policy. The Annals of Regional Science 2001; 35:113-131. Lundvall, B.A., “Innovation as an interactive process: from user-producer interaction to the national system of innovation.” In Technical change and economic theory, G. Dosi, C. Freeman, R. Nelson, G. Silverberg, L. Soete, eds. London: Pinter, 1988.
RETHINKING REGIONAL INNOVATION POLICY
271
Magnusson, L., Ottoson, J., eds., Evolutionary economics and path dependence, Cheltenham: Edward Elgar, 1997. Metcalfe, S., “The economic foundations of technology policy: equilibrium and evolutionary perspectives.” In The handbook of industrial innovation, M. Dodgson., R. Rothwell, eds. Cheltenham: Edward Elgar, 1994a. Metcalfe, S., Evolutionary economics and technology policy. The economic journal 1994b; 104:931-944. Mokyr, J., The lever of riches. Technological creativity and economic progress. Oxford: Oxford University Press, 1990. Morgan, K., The learning region: institutions, innovation and regional renewal. Regional Studies 1997; 5:491-503. Nelson, R.R., ed., National innovation systems. A comparative analysis. Oxford/New York: Oxford University Press, 1993. Nelson, R.R., Winter S.G., An evolutionary theory of economic change. Cambridge, Mass: Harvard University Press, 1982. Oakey, R.P., “High technology small firms: a more realistic evaluation of their growth potential.” In Small business dynamics, C. Karlsson, B. Johannison, D. Storey, eds. London: Routledge, 1993. Perez, C., Structural change and the assimilation of new technologies in the economic and social systems. Futures 1983; 5:357-375. Perez, C., Soete L., “Catching up in technology: entry barriers and windows of opportunity.” In Technical change and economic theory, G. Dosi, C. Freeman, R. Nelson, G. Silverberg, L. Soete, eds. London: Pinter, 1988. Rosenberg, N., Inside the black box: technology and economics. Cambridge: Cambridge University Press, 1982. Saviotti, P.P., Technological evolution, variety and the economy. London: Edward Elgar, 1996. Smith, D.F., Florida R., “Venture capital’s role in regional innovation systems: historical perspective and recent evidence.” In Regional innovation, knowledge and global change, Z.J. Acs, ed. London/New York: Pinter, 2000. Staber, U., Inter-firm co-operation and competition in industrial districts. Organization Studies 1998; 4:701-724. Storper, M., The limits to globalization: technology districts and international trade. Economic Geography 1992; 1:60-93. Storper, M., Regional technology coalitions. An essential dimension of national technology policy. Research policy 1995; 24:895-911. Tunzelmann, G.N. von, Steam power and British industrialization to 1860. Oxford: Clarendon Press, 1978.
Chapter 13 ON THE ROLE OF GLOBAL DEMAND IN LOCAL INNOVATION PROCESSES
Anders Malmberg and Dominic Power
13.1 INTRODUCTION This chapter takes its starting point in what we see as an implicit paradox in contemporary writing on regional systems of innovation and regional clusters. On the one hand, the last decade has seen the proliferation of a number of theoretical propositions arguing that learning and innovation are indeed the result of localized processes, and that academic inquiry as well as industrial policy should be geared towards enhancing the understanding (and workings) of regional systems of innovations and regionally defined clusters of similar and related economic activities. On the other hand, the received result of empirical research on industrial innovation and learning processes indicate that they are indeed far from exclusively local. On the contrary, one of the core aspects of globalization is the increasing magnitude of long-distance flows, not just of goods and money but also of market signals and information. What, then, is localized and what is not in processes of learning and innovation? Are notions like regional clusters and regional systems of innovation the most fruitful starting point for analysis and policy aiming at coming to terms with the issue of national and regional growth and development? These questions are the focus of this chapter. The chapter starts by examining the idea that learning and innovation are essentially localized processes and that as such the local characteristics of the regional industrial system explain and determine firms’ and agglomerations’ success in innovation. Innovation processes are on this type of account thought to work best when spatially localized and the more strongly integrated the local system the better the chances of hitting on successful innovations. In contrast to this, we argue that innovation need not work like this and that the openness of an agglomeration and its firms may in fact be more important to innovative capacity than local integration. In order to flesh out this argument, we use the example of the role of customer relations in innovation.
274
Anders Malmberg and Dominic Power
We provide an analysis of several types of customer relations to show that in an increasingly globalized marketplace it seems strange to think that key customers should be close to the firms that serve them or indeed spatially localized at all. The chapter finishes by further exploring the above noted paradox and suggesting that – while there are indeed quite strong factors making for the creation and reproduction of regional clustering and regional systems of innovation – policies for, and studies of, innovation processes could fruitfully start from the idea that global rather than local linkages and outlook may be the most useful method of boosting regional innovativeness.
13.2 LEARNING AND INNOVATION AS LOCALIZED PHENOMENA In much recent work on the geography of innovation, innovation is seen to be rooted in the region and its attributes. This approach supposes that innovation is essentially a localized phenomenon with the implication that policy interventions aimed at strengthening a region’s supply of assets and strengths is the key route to innovation: supply the tools and methods and applications will grow. In this “model” regional innovation and change is the result of two main factors: the supply of regional assets (in particular “soft” ones rooted in the cultural and knowledge base of the region); and innovation strategy and policy directed at encouraging interactions between firms and R&D institutions in the local milieu. Regional innovation and change then is all about using localized innovation strategy and policy and the strengthening of regional assets to react positively to structural change. A stylized version of the general narrative behind this line of thinking can be summarized in four broad assertions. First, in a knowledge-based economy, the ability to innovate is more important than cost efficiency in determining the long-term ability of firms to prosper. Innovation is here defined in a broad sense, as the ability to come up with new and better ways of organizing the production and marketing of new and better products (Porter, 1990; Lundvall, 1992; Nelson, 1993; Nonaka, 1994; Grant, 1996). This does not mean that cost considerations are unimportant, but simply that the combined forces of the globalization of markets and deepening divisions of labor make it increasingly difficult to base a long-term competitive position on cost-advantage alone. Second, empirical research on innovation processes has quite convincingly demonstrated that innovations predominantly occur as a result of interactions between various actors, rather than because of the solitary genius (von Hippel, 1988; Håkansson, 1987). This fits well view the overall Schum-
ON THE ROLE OF GLOBAL DEMAND
275
peterian view of innovations as new combinations of already existing knowledge, ideas and artifacts. Most innovations are based on some form of problem solving. One actor identifies a problem and turns for help to someone else who might contribute to its solution. In an industrial context, these interactions often follow the value chain. The behavior to be expected from a firm facing a particular problem is to turn to a supplier, a customer, a competitor, or some other related actor, in an attempt to get help in specifying the problem and define the terms for its solution. From this, economic geographers and other students of processes of industrial innovation and change have concluded that innovation processes can be best understood using some notion of an industrial system defined upon similar and related economic activities. Third, and this is where “geography” enters the picture: there are a number of reasons why we should expect that interactive learning and innovation processes are not space-less or global, but on the contrary that they unfold in a manner in which geographical space plays an active role. Proximity carries with it, among other things, the potential for intensified face-toface interaction, short cognitive distances, common languages, trustful relations between various actors, easy observation, and immediate comparison (see Malmberg and Maskell, 2002 for further elaborations). In short, spatial proximity seems to play an important role in these processes, and therefore industrial systems should be assumed to have a distinctly localized component. Fourthly and finally, an implication of the above line of argument is that there are reasons to believe that the knowledge structures of a given geographical territory are more important than other characteristics (general factor supply, production costs, etc.) when it comes to determining where we should expect economic growth and prosperity in today’s world economy. It is the region’s institutions, structures, and characteristics – such as its boundaries, governance and government systems, assets, skills, and infrastructure – which are thought to hold the key to economic innovation and its attendant competitiveness and profitability. Arguably, the two most influential concepts in this context are those of the cluster and the innovation system. The cluster concept, of course, was introduced in the early 1990s by Porter (1990; 1994; and 1998) and has had a tremendous impact not only on research economic geography and related disciplines, but also in policy circles (see Malmberg and Maskell, 2002 or Martin and Sunley, 2001 for recent reviews). The cluster concept was originally formulated to describe and analyze how international competitiveness is developed, not within a stand-alone firm or an individual industry, but within bundles of related firms and industries. A country is often successful not only in the production of a certain type of good, but also in industries providing specialized inputs, technology and associated services. Porter’s original ob-
276
Anders Malmberg and Dominic Power
servation was that such clusters of industries are not global. On the contrary, it was possible to discern distinct differences between the dominant clusters industries in different countries. Thus, the cluster was originally seen as a nationally organized system of functionally related firms and industries. At the same time, the observation was made that such clusters are often not evenly distributed across a country but instead agglomerated to a certain region within the country. Therefore, the concept has gradually taken on the double meaning of functionally defined bundles of industries and spatially defined agglomerations of similar and related industries. As we will argue below, and as has been further developed elsewhere (Malmberg, 2003), this is presumably an unfortunate turn of events. During the same period, the innovation system concept has been introduced to describe how the industrial and institutional structure in a country contributes to channeling industrial and technological development along a certain path (Lundvall, 1992; Nelson, 1993). Here, there is a stronger emphasis upon how the broader organization of knowledge and R&D-systems affect industrial development. An innovation system can thus be seen as a network of organizations, people, and institutions within which creation, diffusion and commercial exploitation of new technologies and other types of knowledge takes place. As with the cluster concept, here too, there has been a gradual shift such that an original focus on national systems has gradually given way to an increased focus on regional systems of innovations (Braczyk et al., 1998). While there are, of course, some differences between the cluster and the innovation system concept there are some similarities as well. Both take their starting point in the view that innovation, renewal and development occur as the result of interactions across a spectrum of different actors. Both also have a spatial point of departure in that this interaction is seen to take place in territorially defined milieux. In addition, both have equally had a great impact on the formulation of industrial, innovation, and regional policies in the OECD countries and beyond. This developing research field has come up with a new and fresh approach to the analysis of industrial and regional change. It has reformulated some of the core research questions in economic geography and related disciplines in a fruitful way, and it has resulted in a number of propositions that do indeed challenge some of the pre-existing core assumptions. Examples of such new propositions in this literature include the following: Firms that meet sophisticated demand from customers located nearby will be more innovative. Rivalry between similar firms located in the same regional milieu will trigger learning and innovation.
ON THE ROLE OF GLOBAL DEMAND
277
Knowledge and innovation will diffuse faster between firms in a local milieu than across longer distances. The more local is collaboration between firms, research institutes, and universities, the more innovative the system. Appropriate institutional frameworks and concerted policy action at the local and regional level can stimulate the type of localized interaction that lead to the development of dynamic regional clusters and innovation systems. These are indeed exciting as points of departure, both for further theoretical development and for policy. However, we would argue that for the moment they are best seen as promising hypotheses, rather than as established facts since several of them have proved extremely difficult to support empirically (despite some heroic attempts, see Malmberg and Maskell, 2002, or Basant, 2002 for reviews). In being exposed to empirical analysis, however, the above hypotheses have been somewhat twisted. The typical outline of a study of these issues is that it starts from a notion of a spatially defined cluster – i.e. a regional economy where one has identified the existence of a number of successful firms that seem to be active within one or a few interlinked industries – with the underlying assumption that if this is indeed a “true cluster” one should expect most key relations between firms to take place within this same spatially bounded system. The results from such studies give, almost without exception, rather disappointing results. A bold summary of these results could be presented in the following way. We would thus claim that a stylized summary of present empirical knowledge suggests that: There is generally relatively little local business going on between firms in regional clusters. When asked questions about where the most important suppliers or customers, or competitors are located, most firms tend to report spatially extended networks. Global connections clearly tend to dominate over local ones (see for instance MacKinnon et al., 2001; Markgren, 2001) There is limited formal collaboration across firms (or between firms and other organizations) going on locally. These collaborations tend rather to follow the value chain and therefore, their spatial configuration tends also to be fairly global (e.g. Larsson, 1998) There is sometimes intense local rivalry, in the sense that the firm down the road is often seen as the “prime enemy”: a bit like the rivalry between neighboring football clubs (Arsenal vs. Tottenham, AC Milan vs. Inter Milan). At the same time, many firms report having extremely few direct competitors globally and none locally.
278
Anders Malmberg and Dominic Power
Firms in a regional cluster seem to have rather good knowledge of other local firms and actors, even though they have some difficulties explaining precisely where this knowledge comes from and whether it is valuable in the context of business development. Therefore, one can often see a good deal of disappointment when empirical analyses show little manifest local interaction between firms in regional clusters. These results have led many scholars in the field to gradually shift focus away from researching transaction links (traded interdependencies, to use the phrase introduced by Storper, 1995 and 1997), and other forms of collaboration (e.g. joint projects in technological development, also with mainly “disappointing” results in the sense that it often turns out that the most innovative firms are indeed globally well-connected, see Larsson and Malmberg, 1999). These scholars have shifted their attention to entirely informal, subtle, and often almost unintended interactions taking place as a result of the predominantly localized nature of everyday life, such as untraded interdependencies, unplanned meetings in bars and restaurants, gossip and rumor, spontaneous monitoring of competitors nearby, etc. (e.g. Pinch and Henry, 1999). Work of the latter type is certainly very interesting and the issue of learning by “being there” (cf. Gertler, 1995) clearly deserves further attention. Still, we believe that the disappointing result of most attempts to validate empirically five assertions made outlined earlier is largely a result of a misconception at the outset. We argue that the relative lack of empirical support for some of these key assertions is not just a result of the empirical difficulty of studying complex relations, but also a result of some basic theoretical and conceptual flaws. There is a general lack of studies in economic geography that focus explicitly on how firms go about securing, and remaining open to new ideas and information originating outside the local region. Few studies today seem to have as part of the research design the idea that it is actually a strength to remain open to distant influences that is the key to the prolonged success of a local milieu (see however: MacKinnon et al., 2001). The thesis we are advancing in this chapter is that regional clusters – or regional systems of innovation – should not be expected to be primarily locally integrated, and that there would be much to gain from dropping the underlying assumption that “the more localized interaction, the better.” Interestingly, some clear pointers towards the role of extra-local relations in cluster development are to be found in Porter’s original account. In a section on “insular clusters,” Porter (1990) writes about how an overly inward oriented focus often explains why nations lose competitive advantage. This happens if firms stop employing global strategies that are aimed at offsetting local factor
ON THE ROLE OF GLOBAL DEMAND
279
disadvantages or at selectively tapping into advantages available in other places: The cluster itself, particularly if it is geographically concentrated, may contain the seeds of its own demise. If rivalry ebbs and homebuyers become pliant or lose sophistication, there is a tendency for the local cluster to become insular, a closed and inward-locking system. The problem is exacerbated if most firms lack significant international activities and their primary commercial relationships are with each other (for example suppliers sell almost exclusively to a single domestic industry). Firms, customers, and suppliers talk only to each other. None brings fresh perspectives. (Porter, 1990: 171)
Such clusters are seen to be vulnerable to structural change. Moreover, Porter points to the inherent risk of lock-in that follows from strong local specialization. When skills, assets, and strategies are highly specialized to a particular industry structure, firms can adapt incrementally but may have difficulties dealing with radical innovation (cf. Grabher, 1993; Asheim, 1996). So in a sense, the idea that a dynamic regional cluster is characterized by intense local linkage and interaction is indeed not really present in Porter’s analysis. In order to better illustrate our argument that local linkages and interactions should not be overplayed we present below an argument that rests on the role of customer relations in learning and innovation processes. We suggest that by taking the role of customers and demand into better account a clearer picture emerges of the motivations and driving forces behind innovation. Furthermore such a focus helps us backup the claims made above that in an increasingly internationalized, even globalized, economic sphere it would be a mistake to always privilege the local in innovation processes.
13.3 THE ROLE OF CUSTOMER DEMAND IN THE INNOVATION PROCESS In contrast to the type of theoretical arguments characterized above, in which innovation is considered to be an essentially spatially localized phenomenon, one may argue instead that innovation is better understood in terms of an innovation chain that, due to the interactive multi-stage and multidimensional aspects of innovation, closely mirrors that of the value chain. This implies that we must not assume that innovation is a localized process or indeed that it is necessarily spatial in character. In this section, we argue that innovation can be considered a process or a “continuum with multiple dimen-
280
Anders Malmberg and Dominic Power
sions” (Danneels and Kleinschmidt, 2001; Green et al., 1995; Harmsen et al., 2000) but that it is a continuum that is very often demand-driven. We conclude by arguing that the spatiality of demand may be very different to that of the firm or the regional cluster and that this observation poses a set of interesting problems for economic geography. Many authors note that it is useful to place innovation processes within the wider context of industrial systems. Porter (1990) for instance makes use of the “value chain” to alert us to the role of different actors and agents in the process of bringing new products or technologies to light and to market. A focus on the value chain helps us to focus on where added value is created. This seems logical as for the greatest share of profits possible in order to survive. This implies that businesses will most likely compete and innovate at points where potential added value is highest. Thus, the value chain and chains of innovative action are likely to be closely related to one another. However, all too often value and innovation chains are in someway equated with supply, delivery, and logistical chains. This perhaps reflects the underestimation of the role of the customer, and indeed consumption (cf. Crewe, 2001; Wrigley and Lowe, 1996) and market systems (cf. Metcalfe and Miles, 2000), in economic geographical analysis. In short, customers are far too often relegated to anonymous recipients of products that are in turn seen as the result of relatively closed innovation processes highly reliant on the dynamics of production (and not the chains linking producer to customer). In contrast, in many business areas, demand, consumption processes, and customers have a determinant effect on the innovation process. As Gereffi notes commodity chains can equally be buyer-driven as much as they can be producer-driven; he suggests products such as footwear and garments are more likely to be buyer-driven than others such as automobiles and aircraft which are more likely to be producer-driven (Gereffi, 1994; Raikes et al., 2000; Dicken and Malmberg, 2001). Such a division is useful in making our point but we argue it may downplay the demand side and demand factors in innovation processes in that it principally draws attention to power relations implicit in a commodity chain rather than motivational aspects underlying the impulse to innovate (though of course it is often hard to distinguish motivation from power). In the context of innovation, the value chain concept is more useful than the commodity chain as it is not principally focused on power relations or on a strictly defined commodity. Many products have associated value chains well beyond the core commodity, e.g. the extent to which servicing and financing have come to be core sources of profitability for automotive “manufacturers” (Raikes et al., 2000:398). Focusing on value and its spatiality draws us immediately to acknowledge that the motivation (rather than the means) to innovate and compete may be more important to
ON THE ROLE OF GLOBAL DEMAND
281
understanding innovation. Innovation is a largely cognitive and interactive phenomenon aimed at creating added value. Therefore, the role of motivations and motivational contexts underlying chains of innovative actions, which are closely related to value chains, must be of analytical primacy. We suggest here that there exist many motivations behind innovation processes: such as rivalry, cooperation, survival, and profit. However, one motivation that particularly concerns us here is the motivation to reply to and profit from demand. In responding to demand, businesses are driven to try to understand, interpret, and manipulate the customers’ choice/motivation. This does not imply that innovation always flows from a “customer is always right” or “customer participation” approach. Rather it implies that analytically in order to understand the initiation and life cycle of innovative actions we must first understand how the desire to innovate ultimately is most commonly related to understandings of the demand side. Current and perceived demand conditions most often drive firms to innovate in certain ways and pursue certain courses of action. As Sutton notes the primary activity of most companies is not innovation but the routine work of making money from existing reliable products and services (2001:96). Attention to current demand conditions, and therein revenue, is primary to business operations. It is not surprising that innovative activity that incorporates a direct link to customer preferences and demand conditions characterizes many successful businesses (Cristiano et al., 2000; Ettlie and Johnson, 1994; Goodstein and Butz, 1998; Govers, 1996; Tidd et al., 1997; Whiteley, 1991). Businesses will often most actively pursue innovation in the parts of the value chain nearest the customer. This is due not only to the fact that better incorporation of customer demand and preferences tends to make for more successful products but also because the closer in the value chain to the enduser one’s products/services are the more profitable/value-adding they tend to be. Whilst not true for every product or area of commerce it is generally true to say that in the contemporary world economy the greater proportion of value is added, and profit made, in the stages nearest the customer or end user. For instance, it is not at the stage of cultivating and selling coffee beans that the greatest profit is made but at the point where the beans are turned into, for example, a coffeehouse espresso (cf. Pine and Gilmore, 1999). Furthermore, innovation in areas such as customer service and relations are important for both profitability and business survival. Whiteley (1991), for example, notes that significantly higher profit margins are achieved by firms that are rated highly by their customers – irrespective of whether their product may be inferior to that of competitors. In addition, strategies in sorting and identifying the dynamics of customer bases are now viewed as central to the operation of leading firms. This has begun to involve concerted efforts to
282
Anders Malmberg and Dominic Power
identify which customers are more profitable than others (cf. Cristiano et al., 2000; Stalk and Webber, 1993) and aggressive efforts, such as mergers and acquisitions (see “The many-sided company,” Financial Times 26-11-2001), to obtain strategic customer and demand data. Lastly, recent research on geographies of consumption (e.g. Crewe and Beaverstock, 1998; Gregson and Crewe, 1997) alerts us to the fact that “value” itself can be a highly contested concept and that the spatialities of consumption and consumers are crucial to the constructions of value and desirability. If the highest levels of value added are likely to be found in parts of the value chain nearest the customer it is also likely that the customer will be the focus and motivation behind most innovative and entrepreneurial activity. Innovation processes most often depend on the role of customers and the dynamics of demand. Given these propositions, how do we understand the role of customers and demand in determining motivational backgrounds to innovation processes or the customers’ often-direct role in innovation? The rather lame answer is that no universal rules govern such things but what we can say is that the characteristics of customers and demand (or the perceived characteristics of them) will play a crucial role in the impetus behind initiating many types of business innovation and the direction they take. None of this is especially new and few economic geographers would question that the customer has an important place in the innovation process and indeed that innovation is often customer-driven. However, once one recognizes that the spatiality of demand and customers is often very different to that of the firm and is often “located” beyond the regional cluster – increasingly so with internationalization and globalization – the validity of policy approaches based on the suggestion that innovation systems function best when the interactive processes of innovation are highly localized becomes questionable. In contrast to the assumption underlying approaches such as the Porterian “diamond” or ongoing cluster policy initiatives it is not always true to say that sophisticated and participatory customers will be proximate to regional clusters (cf. Lundequist and Power, 2002). On the contrary, for many firms key customers are located far away. Indeed, in the context of globalization, economic innovation in advanced economies is increasingly aimed at capturing or taking advantage of demand and customers wherever located. A cursory treatment of three hypothetical types of customer may illustrate better our point. First, the long distance customer. Increasing distances between firms and the customers and markets they serve can be considered to have various effects. In particular we may expect customer relationships to be harder to maintain and less spontaneous with resulting effects on the ability to tailor products correctly or predict future demands best. The problems of distance to customers and points of sale are also problematic for firms with
ON THE ROLE OF GLOBAL DEMAND
283
multiple foreign subsidiaries; companies whose geographical map is constantly being redrawn. Serving customers and winning new ones is especially problematic for multi-site and multi-national organizations as innovative and responsive products can come from anywhere in the organization’s network. It is not necessarily true that greater front line autonomy creates a more responsive and innovative subsidiary and case studies show that higher levels of local autonomy can restrict the transfer of innovations across borders (Birkinshaw and Hood, 2001; Tidd et al., 1997). With customers located outside your immediate market area or a customer base splintered across space and served by multiple business units it is clear that the spatiality of demand can be radically different to that of the production process and organizational foci, creating a series of problems of scale and distance that affect firms’ innovation efforts. Secondly, rather than analyzing globalization and new technologies increasing the footlooseness of firms, perhaps we should be considering that the footloose customer is a common and important component of the world economy. Increased inter-market communication, comparability, exchange rate volatility, and lessening trade restrictions have meant the customer (at least the OECD-based one) has an unprecedented range of choices. The reality of global competition in many sectors means, of course, that buyers have access to suppliers from, in theory, around the world which in turn makes firms’ efforts to retain and acquire customers even more important, and problematic, than before. The customer may constantly be moving through and engaging in different markets and supply chains with limited loyalty to these. Thirdly, the growth of mediated forms of buying and consumption spaces (such as the Internet or electronic trading platforms used by stock and derivatives markets: see Power, 2001 and 2002) allows for the essentially aspatial customer. Such customers can appear rather abstract to firms in that limited information about their preferences or characteristics can be derived. Lack of information on the part of “suppliers” and the difficulty in targeting customers using highly mediated markets can mean that power is almost entirely concentrated in the hands of the consumer. The more highly mediated the market structure, the more likely that it will be demand-driven, and that the success of innovative products and services will be entirely reliant on the adaptiveness and innovativeness of customers and the consumption spaces they use. We have argued thus far that innovation processes can be argued to be increasingly demand-driven and that innovation is perhaps best understood as a process or chain characterized by demand motivating innovative activity. Such an understanding places some degree of analytical precedence on the analysis of demand functions, the customer and consumption spaces. Furthermore, whilst the main producers of a certain type of good may be highly
284
Anders Malmberg and Dominic Power
localized in regional clusters, innovation – in everything from R&D to marketing – should be seen not as a rooted entity or capability but rather as a process formed through customer relationships that need not be significantly fettered by spatial factors or boundaries. If we believe that demand, customer relationships and the like, are constitutive and often initiatory of business innovation – i.e. that innovation processes are demand-driven and occur through interactions between suppliers and customers – then it would appear that we must start to look beyond the region or the “local” in order to understand and promote innovation processes. Taking into account the demand side into explanations of innovation thus presents economic geographers with a difficult set of questions to do with the spatiality of the process. These questions are not difficult in the sense of being impossible to answer rather they are difficult because it makes policy support and advice aimed at supporting regional economic growth supported by business innovation even more difficult than before.
13.4 REVISED HYPOTHESES We started this chapter arguing that much of the literature on innovation valorizes the role of localized conditions in successful innovation processes. The message it seems is that innovation processes best occur within a strong local system and furthermore the implication of this is that the more localized and internalized the innovation process the better it will be. In contrast, we have argued that we need to understand that in contemporary business life this may indeed not be the case. This can be clearly seen when one considers the role of demand and customers in structuring and determining innovation processes and chains. Thus, the central contention of the chapter emerges from a basic a question. Could it be that the set of factors that explain the existence of regional clusters of similar and related economic activity are not the same as the set of factors which determine whether firms in such spatial settings are innovative? Agglomeration theorists, by and large, believe that agglomerations exist because they enhance firm performance. It is said that a number of factors work to enhance the performance of firms based in agglomerations. To brutally paraphrase, it seems to us that three such enhancing factors can be seen to be common to most accounts (with old Marshall still echoing in the background): the benefits of deepened local divisions of labor (“When an industry has chosen a locality for itself [...] subsidiary trades grow up in the neighborhood, supplying it with implements and materials …”); labor market and skills pooling effects (“a localized industry gains a great advantage from the fact that it offers a constant market for skill”); and localized knowledge
ON THE ROLE OF GLOBAL DEMAND
285
spill-over effects (“if one man starts a new idea, it is taken up by others and combined with suggestions of their own”) (Quotations form Marshall, 1920/1960:271). The lesson taken from this is generally that the more each of these local factors the better for the performance (and innovativeness) of firms. What is interesting, then, about this is that analytically these factors are said to do two things: first, explain why a number of firms agglomerate in a certain area; and second, why these firms should achieve global competitiveness from being there. Instead we propose that a starting point in answering the question put forward in the paragraph above is that perhaps we should regard regional clusters and regional systems of innovation not primarily as systems of interacting firms but rather as spatial agglomerations of people with skills and competences. In short one should analytically distinguish why a large number of firms and people (firms are after all in one sense the sum of their personnel) agglomerate, and stay, in a certain place before one goes on to thinking about how firms benefit from such agglomeration. With this in mind, the explanation of the existence of agglomerations should, perhaps, be primarily sought in the labor market. One may imagine that a person with a specialized profession will be better off in a region where there are many potential employers. Firms carrying out a specific type of economic activity will correspondingly be better off in a region where there is a large supply of people with the relevant skills and competences. In addition further mechanisms through which a given pattern of regional clustering will be reproduced over time will come into play. For example, a critical mass of people with specific skills and competences will tend to attract investments from the outside in this particular line of business. In turn, people interested in this type of activity, the “wannabees” of the industry in question, will be attracted to, or trained within, the region in a cumulative process. Furthermore, industrial renewal through spin-offs and new firm formation will also tend to be localized processes: we know from numerous studies that the overwhelming majority of new firms start in the region where the founder resides. Thus, there is a whole series of mechanisms that explain why regional clustering of similar and related economic activity should exist. However, it is the next step in the argument that we believe may be erroneous: i.e. to assume that such a localized system of skills and competences works better if the firms there are strongly integrated locally. Another way of saying this is that Marshall’s argument of a localized division of labor argument and localized knowledge spillovers have been taken too seriously. Maybe it is precisely the other way around. Given that there are already some powerful mechanisms leading to the creation of regional specialization, perhaps the most successful region is the one that can upgrade and innovate by linking up globally. In other words, regional clusters come into existence and
286
Anders Malmberg and Dominic Power
develop over time through a set of mechanisms related to labor market specialization, knowledge accumulation, and spin-offs. The interactions and information flows leading to innovation, however, need not, do not and should not be expected to be confined to the regional cluster. We believe that our brief analysis of customer demand in the innovation process points in this direction. So do the negative results from several studies designed to capture local inter-linkage in regional clusters. Therefore, by way of conclusion, we believe that some of the hypotheses presented earlier in the chapter should be rather radically reformulated, to read instead as: Firms that manage to sense sophisticated demand from and building relations with key customers globally as well as locally will be more innovative. Knowledge and innovation will diffuse, more or less automatically between firms in a local cluster through local labor market dynamics and various community ties. The challenge for firms is to link up with global flows of knowledge. Local collaboration between firms, research institutes, and universities might be good, but it is most crucial for firms to try link up with the very best partners in the world and, in most cases, these will not be found nearby. Appropriate institutional frameworks and concerted policy action at the local and regional level can contribute to the development of dynamic regional clusters and innovation systems, if geared towards opening up the regional system to the outside world. We are fully aware that these are just hypotheses, and as such of limited value. We do think, however, that our analysis of the role of demand and customer relations does lead in this direction and we would, for the sake of argument at least, propose that a fruitful point of departure for future work on regional clusters and systems of innovation would gain from dropping the problematic assumption that “the local is always beautiful.”
ACKNOWLEDGEMENTS The authors would like to thank all those involved in the Stuttgart “International Workshop Rethinking Regional Innovation and Change – Path Dependency or Regional Breakthrough?” for their help in the preparation of this chapter. Dominic Power would like to acknowledge the support of the European Community through their Marie Curie Fellowship program (HPMF-CT2000-00668).
ON THE ROLE OF GLOBAL DEMAND
287
REFERENCES Asheim, B., Industrial districts as “learning regions”: a condition for prosperity. European Planning Studies 1996; 4:379-400. Basant, R., Knowledge flows and industrial clusters. An analytical review of literature. Mimeo. Ahmedabad, India: Indian Management, 2002. Birkinshaw, J., Hood, N., Unleash innovation in foreign subsidiaries. Harvard Business Review 2001; March: 131 -137. Braczyk, H.-J., Cooke, P., Heidenreich, M., eds., Regional Innovation Systems. London: UCL Press, 1998. Crewe, L., The besieged body: geographies of retailing and consumption. Progress in Human Geography 2001; 25:629–640. Crewe, L., Beaverstock, J., Fashioning the City: Cultures of consumption in contemporary urban spaces. Geoforum 1998; 29:287-308. Cristiano, J., Liker, J., White, C., Customer-driven product development through quality function deployment in the US and Japan. Journal of Product Innovation Management 2000; 17:286-308. Danneels, E., Kleinschmidt, E., Product innovativeness from the firm’s perspective: its dimensions and their relation with product selection and performance. The Journal of Product Innovation Management 2001; 18:357-373. Dicken, P., Malmberg, A., Firms in territories: a relational perspective. Economic Geography 2001; 77:345-363. Ettlie, J., Johnson, M., Product development benchmarking versus customer focus in applications of quality function deployment. Marketing Letters 1994; 5:107-116. Gereffi, G., “The organization of buyer-driven global commodity chains: how US retailers shape overseas production networks.” In Commodity Chains and Global Capitalism, G. Gereffi, M. Korzeniewicz, eds. Westport, Conn.: Praeger, 1994. Gertler, M., Being there: proximity, organization and culture in the development and adoption of advanced manufacturing technologies. Economic Geography 1995; 70:1-26. Goodstein, L., Butz, H., Customer value: the linchpin of organizational change. Organizational Dynamics 1998; 27:21-34. Govers, C., What and how about Quality Function Deployment (QFD). International Journal of Production Economics 1996; 46-47:575-585. Grabher, G., “The weakness of strong ties. The lock-in of regional development in the Ruhr area.” In The embedded firm: on the socio-economics of industrial networks, G. Grabher, ed. London: Routledge, 1993.
288
Anders Malmberg and Dominic Power
Grant, R., Prospering in dynamically-competitive environments: organizational capability as knowledge integration. Organization Science 1996; 7:375-387. Green, S., Gavin, M., Aiman-Smith, L. Assessing a multidimensional measure of radical technological innovation. IEEE Transactions in Engineering Management 1995; 25:186-192. Gregson, N., Crewe, L., The bargain, the knowledge and the spectacle: making sense of consumption in the space of the car boot sale. Environment and Planning D: Society and Space 1997; 15:87-112. Harmsen, H., Grunert, K., Declerck, F., Why did we make that cheese? An empirically based framework for understanding what drives innovation activity. R&D Management 2000; 30:151-166. Hippel, E.V., The Sources of Innovation. Oxford: Oxford University Press, 1988. Håkansson, H., Corporate technological behaviour: Co-operation and networks. London: Routledge, 1987. Larsson, S., Lokal förankring och global räckvidd: en studie av teknikutveckling i svensk maskinindustri [Local embeddedness and global reach: a study of technological development in the Swedish machinery industry]. Uppsala: Geografiska Regionstudier nr. 35. Uppsala Universitet, 1998. Larsson, S., Malmberg, A., Innovations, competitiveness and local embeddedness. a study of machinery producers in Sweden. Geografiska Annaler Series B Human Geography 1999; 81B:1-18. Lundequist, P., Power, D., Putting porter into practice? Practices of regional cluster building: Evidence from Sweden. European Planning Studies 2002; 10:685-704. Lundvall, B.-Å., National Systems of Innovation: towards a theory of innovation and interactive learning. London: Pinter, 1992. MacKinnon, D., Chapman, K., Cumbers, A., Networks, learning and embeddedness amongst SMEs in the Aberdeen oil complex. Working Paper 3. Aberdeen: ESRC Research Project: Innovation, Networks and Learning, Department of Geography, University of Aberdeen, 2001. Malmberg, A., “Beyond the cluster. Local milieux and global connections.” In Remaking the Global Economy, J. Peck, H. Yeung, eds. London: Sage, 2003. Malmberg, A., Maskell, P., The elusive concept of localization economies: towards a knowledge-based theory of spatial clustering. Environment and Planning A 2002; 34:429-449. Markgren, B., Är närhet en geografisk fråga? Företags affärsverksamhet och geografi - en studie av beroenden mellan företag och lokaliserings betydelse. [Is proximity a geographical question in business relationships]. Uppsala: Uppsala University, Department of Business Studies, 2001. Marshall, A., “Industrial Organization, Continued. The Concentration of Specialized Industries in particular Localities.” In Principles of Economics, Book IV, Chapter X, A. Marshall, ed. London: Macmillan, 1920/1960.
ON THE ROLE OF GLOBAL DEMAND
289
Martin, R., Sunley, P., Deconstructing clusters: chaotic concept or policy panacea? Presented at Regional Studies Association conference on Regionalising the Knowledge Economy, 2001 November 21, London, 2001. Metcalfe, J., Miles, I., eds., Innovation Systems in the Service Economy: measurement and case study analysis. London: Kluwer Academic, 2000. Nelson, R., ed., National Innovation Systems: a comparative analysis. Oxford: Oxford University Press, 1993. Nonaka, I., A dynamic theory of organizational knowledge creation. Organization Science 1994; 5:14-37. Pinch, S., Henry, N., Discursive aspects of technological innovation: the case of the British motor-sport industry. Environment and Planning A 1999; 31:665-682. Pine, B., Gilmore, J., The Experience Economy: work is theatre and every business a stage. Boston MA: Harvard Business School Press, 1999. Porter, M., The Competitive Advantage of Nations. New York: The Free Press, 1990. Porter, M., The role of location in competition. Journal of the Economics of Business 1994; 1:35-39. Porter, M., Clusters and the new economics of competition. Harvard Business Review 1998; November-December: 77-90. Power, D., “Information and Communications Technology and the Integration of European Derivatives Markets.” In Worlds of E-Commerce: economic, geographical and social dimensions, T. Leinbach, S. Brunn, eds. Chichester: Wiley, 2001. Power, D., IT and institutions in the structuring of European finance: urban impacts. Economic and Industrial Democracy 2002; 23:335-356. Raikes, P., Jensen, M., Ponte, S., Global commodity chain analysis and the French filiere approach: comparison and critique. Economy and Society 2000; 29:390-417. Stalk, G., Webber, A., Japan’s dark side of time. Harvard Business Review 1993; July-August: 93-102. Storper, M., The resurgence of regional economies, 10 years later: the region as a nexus of untraded interdependencies. European Urban and Regional Studies 1995; 2:191-221. Storper, M., The Regional World. New York: Guildford, 1997. Sutton, R., The weird rules of creativity. Harvard Business Review 2001; September:94-103. Tidd, J., Bessant, J., Pavitt, K., Managing innovation: integrating technological, market and organizational change. New York: Wiley, 1997. Whiteley, R., The Customer-Driven Company: moving from talk to action. Reading, MA: Addison-Wesley, 1991.
290
Anders Malmberg and Dominic Power
Wrigley, N., Lowe, M., eds., Retailing, Consumption and Capital: towards the new retail geography. Harlow: Longman, 1996.
Chapter 14 THE REGIONALIZATION OF INNOVATION POLICY: NEW OPTIONS FOR REGIONAL CHANGE?
Knut Koschatzky
14.1 INTRODUCTION Since the beginning of the 1990s, but specifically during the last few years, the “region” (i.e. sub-national spatial entities) has not only gained importance in theoretical discussions, but because of convincing new theories, also in national technology and innovation policy. This can be observed at the supranational level of the European Union and at the level of different European countries (Koschatzky and Sternberg, 2000:494-499). The basis of this development was the “re-discovery” of space in economic theory. This was exemplified by the emergence of “new growth” and “new trade” theory and the “new economic geography” based thereon (see Krugman, 1991; 1995; 1998), as well as the multi-faceted analyses of national and regional innovation systems and the political implications of their being influenced (cf. among others Cooke, 1992; Nelson, 1993). The decisive impulse proceeded from the cluster concept, developed and actively diffused by Michael Porter (1990; 1998). This approach, along with network economics on which the cluster concept is based, went so far that regionally focused and networkbased development concepts are hailed as the new panacea of a (technology) policy searching for innovative starting points. Until the first half of the 1990s, innovation and technology policy was primarily intended to assure national technological competitiveness and was thus oriented towards the target of national growth. Spatial implications existed only implicitly in the geographical distribution of recipients of public promotional funding. Now a conscious change of direction towards the regional level, its structures, and regional economic and innovative potentials can be observed. Regions are being regarded in this context as starting points for national policy measures, whereby not only growth, but also balancepolicy targets are pursued. Regions are also seen as a spatial action frame-
292
Knut Koschatzky
work within which self-organization processes to develop and strengthen the regional knowledge basis are initiated through external impetus. Innovation as the foundation of economic activity is no longer limited only to technological innovations in the present knowledge and learning society. With a view to the challenges for regions resulting from internationalization and globalization, new regional arrangements and organization forms, as well as the generation of “regional innovations” are required. At present, an increasing tendency can be observed in regional policy to reject the classical infrastructure promotion and a change of direction towards the development of “soft” intangible factors and human resources promotion. This applies especially to European policy, which was already characterized by a growing innovation orientation even before the formulation of the concept of a European Research Area (ERA), for example within the framework of the Regional Innovation Strategies (RIS) (cf. Landabaso and Youds, 1999; Landabaso et al., 2001). Innovation-oriented regional policy, which was already postulated at the end of the 1970s (cf. Ewers and Wettmann, 1980), is now being filled with new content. The regional orientation in innovation policy presents fundamental preconditions for institutional re-organization and the formation of new policyeconomic and institutional arrangements. To what extent such processes will be effective within individual regions depends on the type of policy and the measures implemented, but also on whether impetus for change comes from within or without the region. It also depends on the absorptive capacity of regional policy actors and on the impact of regional interest coalitions between policymakers and industry on regional structural change. This chapter draws on the following thesis: the regionalization of innovation policy offers the chance, by means of more efficient allocation of funds and the delegation of political responsibility to regional coalitions of interest, to promote regional structural change, and to break up routines and habitual development paths. This is, however, only possible if regional institutions can be successfully brought to renew themselves. While most regions in a country have the potential for endogenous structural changes, only a few regions by contrast are suited to set up new technological development paths with national or international impacts. The chapter will thus discuss possible policy contributions to regional change and takes the need for policy intervention in an imperfect market for granted. It is structured as follows: First, Section 14.2 deals with the discussion of the theoretical cornerstones of a regionalization of innovation and technology policy, while Section 14.3 outlines the tasks and characteristics of regional innovation and technology policy. These explanations form the foundation for the development of a regional typology in Section 14.4 and for the discussion of questions about the scope for action and development op-
THE REGIONALIZATION OF INNOVATION POLICY
293
tions that are available at the regional level. In the concluding Section 14.5, policy conclusions are derived and opportunities that could evolve for regional structural change from the increasing regional orientation of innovation policy are discussed.
14.2 THEORETICAL CORNERSTONES OF A REGIONALIZATION OF INNOVATION AND TECHNOLOGY POLICY Regional economics has been concerned with the spatial distribution of economic activity for several decades. However, it was the work of Lucas, Romer, Krugman and other economists who, by introducing external effects and focussing on the regional accumulation of knowledge, re-discovered spatial aspects in economic theory (see for example Lucas, 1988; Romer, 1986 and 1990; Grossman and Helpman, 1990; Krugman, 1979 and 1991). Specifically, the work of Krugman about geography and trade laid the foundation of a revival of economic geography (Krugman, 1998). This “new economic geography” is not only fuelled by models of the new growth theory and the new trade theory, but by many other theoretical concepts dealing with the regional distribution of technological development and innovative activity. Among the most popular are Porter’s reflections about the factors influencing the competitive advantage of nations (Porter, 1990), the different contributions to the cluster concept (Porter, 1998) and the theory-based empirical studies about knowledge spillovers and their spatial reach (e.g. Anselin et al., 1997).1 The more recent regional-economic concepts and theories are at least partly influenced by findings from innovation economics. These stress that innovation is not a linear, but an evolutionary, cumulative and feedback process, which can only be realized in the cooperation and in the economic and social interaction of different actors, and as a result produces technological, organizational and social innovations (Koschatzky, 2001:62). Despite the differences between the individual concepts, there are several common aspects in the explanation of innovation and regional development. Many concepts emphasize the importance of spatial proximity between the protagonists of a production or value-added chain as well as for innovation. Proximity advantages are explained by positive external effects and thus productivity and cost advantages, which can be realized by a flexible and specialized division of labor between small firms, operating in spatial proximity. Although all of the approaches use the term “region,” there is a different understanding about what defines the spatial context and which spatial entities make up a region. Spatial proximity, understood as metric distance
294
Knut Koschatzky
measure, is more closely defined in the concept of industrial districts than in the concept of regional innovation systems, which are mainly regarded as provinces, regions, or Bundesländer (cf. Cooke and Morgan, 1994; Braczyk et al., 1995). According to the concept of industrial districts and the cluster concept, external effects predominantly develop from localization advantages, i.e. from regional specialization (cf. Pyke and Sengenberger, 1992; Baptista and Swann, 1998; Porter, 1998). In this context, new growth theory and empirical spillover studies stress the necessity for a spatial concentration of knowledge providers, research institutes, enterprises, and physical infrastructure for the realization of knowledge externalities (cf. Carrincazeaux et al., 2001 for the localization of R&D activities in France). Network and milieu-oriented approaches also allow the possibility for the development of decentralized production and innovation clusters (cf. Coombs et al., 1996; DeBresson and Amesse, 1991; Staber, 1996; Tijssen, 1998; and Tödtling, 1999 for different aspects of innovation networking). Regional innovation differences are thereby no longer explained only by locational parameters, but by the ability of economic actors in a region to establish intra- and interregional information and production networks, to participate in network integration and to profit from these networks by collective learning processes (e.g. as stated in the concept of learning regions; cf. Hassink, 2001 for a recent overview). In contrast to regional specialization, the innovative milieu approach argues with diversified economic structures in a region for stimulating innovative activity (cf. Maillat et al., 1993 and 1995; Ratti et al., 1997). The mechanism for stimulating co-operation and innovation is the existence of a regional culture and identity, which creates the basis for trustful cooperation and from which informal, hierarchy-poor, and horizontal networks between the regional actors develop. Culture and identity might be a source for change, but could as well inhibit economic change in a way that they generate lock-in effects by a strong regional economic and social embeddedness of firms and other institutions (Oinas, 1997:30)2. On the other hand, according to the concepts of learning regions and of regional innovation systems, spatial proximity and thus the embeddedness in a regional social system supports the generation of collective learning processes and the exchange of information and knowledge, especially when the knowledge is tacit and therefore spatially immobile. These are those localized routines, knowledge and learning processes which Storper (1995) termed “untraded interdependencies.”3 Especially significant for the explanation of the regional structural change is the concept of governance which is a central element in regional innovation systems. If political actions can steer regional development processes, which is mainly the case in public regional innovation systems and less in new economy innovation systems (Cooke, 2001), then “multi-level governance (MLG)
THE REGIONALIZATION OF INNOVATION POLICY
295
relationships” play a special role. This governance system creates the preconditions for regional openness, the docking into supra-regional, national and supra-national policy levels and the integration of regional innovation systems in globally operating technological and enterprise systems (cf. Cooke, 2002:136-137).4 MLG relationships however can only lead to regional structural change if the learning capability and absorptive capacity of the regional policy and promotional institutions, as well as the political networks existing between them are sufficiently developed (Koschatzky, 2001:334; Marin and Mayntz, 1991:18). Regarding their policy implications, the concepts allow the following conclusions about success factors of regional innovation promotion. These are related to: An innovation-oriented local or regional institutional system with flexible policy networks, a regional capital market and a governance system with appropriate financial authority. An institutional structure that is rich in learning, knowledge transfer, and qualification aligned. Intensive local and regional networking, enhanced by national and international cooperative linkages between regional actors, which facilitates mutual knowledge exchange and enables collective learning processes. A creative and entrepreneurial-oriented human capital that contributes to a continuous renewal of the regional enterprise stock. The emphasis on open, flexible and competitive networks as a key element for structural change and the utilization of so far underdeveloped knowledge potentials, the possibility to support network building processes by public policy intervention and the importance of spatial proximity at least in early phases of technological development and innovation processes made the “region” an interesting political action field. It is assumed by policy makers that within a limited spatial entity with a limited number of actors, public funds for initiating and supporting network building can be allocated much more precisely and perhaps also more efficiently than in measures without a regional focus. These aspects in particular contributed to the popularity of regionally-oriented policy supporting measures.
14.3 REGIONAL INNOVATION AND TECHNOLOGY POLICY The terms “innovation policy” and “technological policy” are often synonymously used, although innovation policy represents the intersection of re-
296
Knut Koschatzky
search and technology policy (Meyer-Krahmer, 1989:1). With regard to a broad definition of innovation, innovation policy aims at the support of science and economy from the first generation of an idea up to its introduction onto the market. In this way, scientific, technological, economic, organizational, and social aspects of the socio-economic change are raised for discussion. Technology policy is defined more closely and is understood as “... policy concentrated on scientific-technical areas” (Meyer-Krahmer, 1997:1) Its main objective is the promotion of application-oriented research and development as well as the use of R&D results in the form of new technology in industry. 5 As innovations are principally generated and determined by demand, which is greatly influenced by the various policy arenas (e.g. innovations in environment technology brought about by new legal regulations and standards), the linking of the different policy arenas achieves a special significance (Meyer-Krahmer, 1997:2). This increasingly also applies to regional policy, which should create the economic, industrial and structural foundations for an innovation and technology promotion at a regional level. As can be seen from theoretical concepts and underlying empirical studies, important framework conditions of technology development still have a national or even regional component, such as the R&D infrastructure, human capital and the general innovation environment (e.g. defined via the level of regulation, the standards system, the availability of investment and venture capital). However, it cannot be assumed that each nation or region has the same chances and starting conditions in innovation competition, as the development of globally distributed competence centers shows. An answer to these processes on the European level lies in the creation of a European Research Area (ERA), in order to bundle and strengthen the research potentials available in the individual states and regions (cf. European Commission, 2000). The EU Commission is assuming that industrial and scientific research has to be embedded in specific regional contexts (European Commission, 2001). It is expected that through the interaction of the different factors determining regional innovation and by changes in institutional and organizational framework conditions the adaptive flexibility of economic and scientific actors can be increased, and thus the regional knowledge base broadened. Regional institutions will be enabled, by means of regionally organized learning processes, to participate in and enhance international knowledge generation. With regard to the theoretical foundation of a regionally-oriented innovation policy, its major purposes should be the promotion of regional clusters, the improvement of the efficiency of regional innovation systems and the stimulation of competition between regions (cf. ISI et al., 2000:539-543): A promotion of clusters seems appropriate for industries and technologies that are in an early phase of their life cycle. In order to structure critical masses, the localizing effect of “tacit knowledge”
THE REGIONALIZATION OF INNOVATION POLICY
297
and spatially limited spillovers demand a high degree of regional concentration during that early development phase. The cluster approach seems to be less suitable for mature industries or technologies. In cluster promotion, region-oriented innovation policy can have positive effects not only for the private sector, but also for regional development. Nevertheless, miracles should not be expected: At least with a sectoral or technologically non- discriminating promotion a kick off towards a permanent change of the promoted regions on a higher growth path cannot be assumed. At best, catching up processes of less innovative regions can be accelerated by a constant allocation of public money over time. Regarding technology or sector-specific promotion, however, long-term and sustainable growth impulses can be given, especially if the technology or industry emerges as “key technology,” and if the external effects are localized over a longer period. Silicon Valley is the most well known example. In the long run, the strong specialization of a region might lead to lock-in situations (Grabher, 1993), causing obsolescence and resulting in monostructures as was the case in coal mining and steel producing regions. This danger can be effectively avoided by a mix of technologies and industries of different maturity stages. For national measures for improvement of the efficiency of regional innovation systems, three starting points exist: (i) the improvement of the integration of regional innovation systems into the national innovation system, (ii) the strengthening of the constituent elements of regional innovation systems and (iii) the better networking of the elements of regional innovation systems. In these cases, national innovation and technology promotion should be limited to supporting and stimulating functions. The central, regional or local governments have the function of creating suitable basic conditions for enabling firms to innovate and – where this is efficient – to allow for spatial concentration. Incentive systems by public financial assistance also seem to be possible, but only to the extent that they strengthen the regional self-initiative and the motivation for the development of endogenous objectives and supporting measures. This concerns both regions with technological potentials and spatial units, in which innovation conditions are still underdeveloped. The third element of a regionally-oriented innovation policy is the stimulation of the competition between regions. Competition between regions and their institutions represents an experimental procedure for uncovering superior institutional arrangements because, without the competition of alternative solutions, it is not known which consti-
298
Knut Koschatzky
tutional arrangements or political orders are better suited to serve the interests of the population. With regard to policy intervention at the regional level, the question of the appropriate policy level is usually raised. According to the subsidiarity principle, the next higher (or a next, respectively) level is only then required to intervene, if the lower level is not capable of a commitment. In this manner, supranational or national state policy actors must only become active on the regional level if for example the lack of financial resources inhibits a regional self-organization process, or if interregional aspects such as the balance-orientation of regional policy are to be considered. Superior levels fulfill an important function in regions if they create incentives through public partial funding, which strengthen the regional initiative and the motivation to develop endogenous targets and promotional measures. This applies not only to regions with technological potential, but also to spatial units in which the pre-conditions for innovation are still underdeveloped. However, the shift of responsibilities to lower political levels also means that institutions are entrusted with political management and controlling tasks for which they are not qualified. Although the European Commission attempts, for example, with its RIS programs to strengthen the politicaladministrative competence on the regional level, those regions, however, are still not capable of applying for funding for innovation-promoting measures and spending funds efficiently, which would need funding most. The necessary absorptive capacity, which is a pre-condition for efficient and effective political action, is still missing there. Landabaso et al. (2001:248) and Oughton et al. (2002) describe this fact as “regional innovation paradox.” Before innovation-promoting measures can be successfully implemented in regions, political implementation competence must be improved in these regions. To implement regionally-oriented innovation policy measures, the following tasks must be mastered (based on Koschatzky and Gundrum, 1997:212): 1. Activation and targeted promotion of the regional innovation resources to strengthen the collective learning capability and to develop and apply new technologies and services: in order to achieve this it is first necessary to ensure and develop the competence in formulating, implementing, and administering policy measures in political institutions. In further steps, the needs and deficits of regional innovation actors must be ascertained, the offer to and the demand for available resources must be identified and the activation of relevant resources must be organized. The main task in this context is the creation of framework conditions for regional structural change and for regional growth.
THE REGIONALIZATION OF INNOVATION POLICY
299
2. Coordination and coupling of these resources in regional innovation networks in order to generate regional system innovations. This involves integrating all process steps from research and development to production and marketing by combining all relevant actors from industry, science, politics, and society. For this it is necessary to identify individual actors, possible promoters, and the existing informal and formal networks, to mediate the establishment of networks and to provide financial support as well as to accompany the development of the networks in the course of time. 3. Integration of these regional networks in national and international knowledge and technology networks by creating active interfaces and promotion of supra-regional cooperation to ensure and increase regional competitiveness. Regional openness for new, problem-solving approaches are necessary even if they lie outside existing routines, as well as the willingness to participate as a region in regional competition (benchmarking).
A look at regional innovation initiatives suffices to see that the business community is often only very slightly involved (ISI et al., 2000:498-499). In order to improve this situation, two basic principles can be followed: the demand orientation of the activity and its bottom-up approach. As private-sector enterprises realize innovations thus contributing to their own and to regional competitiveness, especially SMEs should be the focus of innovation promotion. The industrial sector must be integrated in all phases of the development process, in order to increase social capital.6 The creation of trust in combination with common values and norms support the development of a regional culture and contribute to a positive regional development through regional learning. When it comes to the elaboration of a regional innovation strategy and the definition of concepts and measures, the following process has proved to be successful (cf. Figure 14.1): a sound organizational basis (left side) and a mixture of different approaches (right side) by which sufficient information about regional competencies, regional innovation potentials and development prospects can be collected and transferred into the formulation of a regional innovation strategy. In such a strategy, the particular focus should be on a further development of industry- and application-oriented research. R&Dand innovation-related services should assist enterprises in their innovation activities, but also bring new stimuli and new knowledge into the privateindustry sector. Strategies to create or improve the exchange between the single elements of the innovation system also belong to this category. The (partial) strategies often contain programs to create networks between industry and the research sector and/or between industry and educational institu-
300
Knut Koschatzky
tions. Further strategic elements are the promotion of existing technological capacities in the region, for example by improving the access to external knowledge in certain technology fields. A further important factor is the allocation of adequate financial resources for innovation projects. Finally, specific measures for certain groups of enterprises could be envisaged, for example, the promotion of young enterprises or company start-ups, the introduction of new technologies in certain industry branches or the support for the internationalization of economic activities. In addition, support for the market success of the innovation projects – for example by consulting on commercialization or assistance in marketing and export – also belong to these strategic elements.
Figure 14.1: Elaboration of a regional innovation strategy.
The involvement of the business sectors by public-private partnerships is a basic pre-condition for creating new regional development options, as the enterprises, their inter-company networks, and their interactions with scientific institutions are the central innovation actors. If enterprises cannot be persuaded to change existing behavioral patterns (i.e. routines), then it will also not be possible to abandon traditional paths of development.
THE REGIONALIZATION OF INNOVATION POLICY
301
14.4 REGIONAL TYPOLOGY AND PATH DEPENDENCE It has already been mentioned that regions are not identical functional or political-administrative spatial units, but vary in size, political structure, and economic strength. According to the regional framework conditions, different factors influencing innovations affect and influence the innovation behavior of the enterprises (Keeble, 1997:284). Firms can act and cooperate in a different way in an environment with a multi-layered, high-quality innovation infrastructure as in regions which have only few educational and research institutions. Therefore, innovation support measures must be regionally specific and oriented towards the special starting situation. In so-called new economy innovation systems, the status of policy measures is lower, because of the dominant influence of economic factors, than in innovation systems that are strongly characterized by public promotional measures and innovation infrastructure (Cooke, 2001:970). According to the different theoretical concepts and their empirical analysis, three major types of regions can be distinguished (Koschatzky, 2001:362-366): 1. Globally interlinked centers of national and international technological and scientific excellence. Examples are Silicon Valley, the Greater Boston Area, Île de France, Tokyo, and Singapore. In the theoretical and empirical literature (cf. Amin and Thrift, 1994; Andersson and Andersson, 2000; Sassen, 1996), these regions are named as “global cities,” “global hubs,” “gateway regions,” or “technological clusters,” and “competence regions.” Their common characteristics are: Globally acting transnational companies. A specialized, scientifically excellent research infrastructure with international co-operative linkages. A great deal of localized knowledge and learning. A pronounced entrepreneurial climate supported by national and international venture capital. High innovation and R&D expenditure and advanced technological competencies of the firms. An excellent innovation, transportation and communication infrastructure defining the reference for many other regions. 2. Regions which are intensively integrated in national and international innovation networks. Examples are Baden-Württemberg, Rhônes-Alpes, Lombardy, and Catalonia. The regions are important national locations of technological development and the homebase of many large national and international enterprises (cf. Braczyk et al., 1998 for a description of some of these regions). Their major characteristics are: Complex production and innovation networks.
302
3.
Knut Koschatzky
Learning processes mainly organized within the production environment. A well-developed innovation, communication, and transportation infrastructure meeting national and international standards. Close linkages between industry, science, and administration. At least partial political and financial autonomy. Regions with underdeveloped innovation potentials. Examples are old industrial districts and regions in economic transformation. Regions of this type are characterized by heterogeneity. They may be regarded as traditional industrial clusters, industrial districts (to some extent), and peripheral-rural regions. A common feature is the industrial basis, which can be partly highly specialized, consisting of predominantly small and medium-sized and few large-scale enterprises. These are supported by a standard supply of technical and advisory services. The political and financial autonomy is lower than in the type of region (2) and permits the endogenous management of regional economic development in only a limited way. The regional knowledge base consists of a high proportion of codified knowledge, which is complemented by production and market experiences. External and spillover effects can only be realized to a small extent.
While in many, but not in all type-1 regions development is mainly market-driven and policy intervention is confined to the creation of innovation and investment-friendly framework conditions, type-2 regions can be described by a balanced relationship between market forces and public intervention. A good example for this type of region is Baden-Württemberg, where the state government supports market forces in an intelligent way, but is also engaged in defining own technological priorities as incentive for regional firms to develop and adopt new technologies (e.g. in fuel cell technology). Development in type-3 regions is mainly policy-driven and therefore heavily dependent on public intervention. These different regional starting conditions have to be taken into account when it comes to the question to what extent and with which measures a regionally-oriented innovation and technology policy is able to foster a structural change in the regions. This regional typology leads directly to the question of the path dependency of regional innovation systems. When innovation-oriented support measures are implemented in the region, their objective should be, against the background of the national framework conditions and the regional context, to intensify the innovation activities in the region and in addition to promote development and competitiveness. It is difficult, however, to define the optimal development path and to set targets for regional development. In principle, recognizing existing potentials and deriving possibilities to support them
THE REGIONALIZATION OF INNOVATION POLICY
303
creatively (“new combination of the resources”) take center stage. In each regional type, the particular factors influencing development are present to a varying degree. Thus depending on the technological potentials of a region, the use of not only technology policy but also innovation policy promotion instruments seems justified. In regions with high technological excellence, the creation of framework conditions for the generation of new technologies and technology diffusion is the principal focus of policy intervention. Therefore, the promotion of technology development and the increase in technological performance can be a crucial element of regionally-based development concepts, especially if the region is suited to fulfilling the interests of a national (and increasingly also supranational) technology policy, directed towards technological visions and priorities. In many other regions, the emphasis lies on the improvement of the general innovation capabilities of regional institutions, and on jointly generating innovative ideas and concepts. These can also comprise social or organizational innovations besides the technological ones. This kind of innovation policy is mainly maintained by regional policy actors and regional institutions. The question of how many regions are necessary to form the backbone of a country’s technological strength can only be answered by looking at the already existing spatial structures and at the political systems. While in centralized economies, market-oriented industrial R&D activities are mainly confined to a few regions (as it is the case in France and the United Kingdom), other countries are characterized by a more decentralized distribution of R&D activities (e.g. Germany or the United States; cf. Table 14.1). The resulting question of whether a highly or a less concentrated distribution of R&D potentials is more efficient seems to be a rhetoric one, at least in the short run, because countries with both spatial structural characteristics have gained technological competitiveness in certain fields, paying the price of still existing regional inequalities. Nevertheless, this question is important, especially when major centers of R&D activity are unable to fulfill the needs for technological change and for establishing new localized technological paradigms. Regarding the breaking-up of path dependencies, the question is not whether centralized or decentralized systems are more efficient, but whether economies succeed in flexibly adjusting their spatial distribution of R&D and innovation activities to the challenges of global technological competition.
304
Knut Koschatzky
Policy measures, however, if oriented towards innovation and technology promotion or regional development, are only able to establish new, fundamental development paths in exceptional cases. This can be attributed among other things - to the fact that the development of new technoeconomic paradigms (cf. Freeman and Perez, 1988:45-47) is beyond the reach of political action. Mostly through historical coincidence, exactly those ingredients can be found in a region which forms the basis of new technologies. Technology policy can strengthen such processes, but their setting up is rather the exception than the rule. A technology policy of this type would not be oriented towards regional development and regional structural change, but will follow national efficiency and growth criteria (such as e.g. the BioRegio Contest in Germany; cf. Dohse, 2000). The spatial concentration of public funds and thus the intensification of special technological-economic development processes in single regions of an economy becomes an instrument for improving national competitiveness in the international technology competition. Regions which help to reach this target (i.e. those belonging to type 1 or 2) must be characterized by scientific competence, by innovative enterprises, a close interaction between science and industry, by the link of both sectors to international know-how, i.e. by regional openness, by a highly qualified labor force, capable of learning, as well as by competent institutions, which administer the national policy guidelines in the region.7
THE REGIONALIZATION OF INNOVATION POLICY
305
The growth-oriented technology policy of a regional character is faced by the regional innovation policy which is more balance-oriented. It can also contribute to establishing new regional development paths, but only if it succeeds in removing traditional behavioral patterns in enterprises and other institutions and in persuading them to reform. Corresponding to the balanceoriented target of the policy measures, these paths serve to reduce regional disparities and to align regional development levels, for example by initiating learning processes which broaden the regional knowledge base and contribute to an increase in regional competencies. They perform important functions within individual regions and, if achieving the balance objective, also within the economy as a whole, as public funds can then be utilized for other tasks.
14.5 IMPLICATIONS OF AN INCREASING REGIONAL ORIENTATION OF INNOVATION POLICY Regional technology and innovation policy is always in conflict about targets. If the view is predominantly directed towards the conflict between spatial balance and overall economic efficiency of a regionally-oriented innovation and technology policy, it has to be questioned whether a preference is to be given to the development of specialized regions (e.g. competence centers, clusters), with the consequence of a possible increase in regional disparities, or to the broad innovation promotion in a multiplicity of regions with the possible consequence of decreasing national technological competitiveness. This conflict makes clear that regional innovation policy finds itself in the border area to regional structural and balance policy. This is particularly the case when measures are not implemented exogenously (i.e. “from above”), but are formulated on the region’s own responsibility (i.e. endogenously) and own initiative and coordinated with the next higher policy level, thus placing the interests of the individual region (and not of all the regions of a country) in the center of political action. In the last few years, regionally-oriented innovation policy and (innovation-oriented) regional policy have converged in various aspects (cf. Landabaso and Youds, 1999 for the Regional Innovation Strategy approach of the European Commission). Characteristics of the convergence process are: The focusing of political activity on the region. The emphasis of the importance of innovation and technological development to ensure regional and national competitiveness. Whereas this is a central element of innovation policy, which increasingly uses the regional level as a starting-point, regional policy coming from the
306
Knut Koschatzky
regional level increasingly includes innovation-oriented targets in its measures. The recognition that one policy alone cannot provide a comprehensive approach to solving complex regional development problems. The promotion of regional and inter-regional networks which is practiced in both policy fields. The experimental character of the policies. The introduction of contests as an element to increase the efficiency of allocating public funds. Public-private partnerships as a possibility to achieve additional leverage effects of the promotional measures with the use of limited funds. Therefore, different effects are to be expected from an increasing convergence of aims and instruments between innovation and regional political, which can vary according to the type of region and the specific policy targets. On the one hand, the use of regional policy know-how improves the knowledge of innovation policy makers about the specific conditions of the regional action level. Measures could be formulated in a more targeted manner and special regional impacts better achieved, for example in the currently popular area of regional cluster promotion (cf. Steiner, 1998) or in building up regional competence networks. Regional policy makers can learn from the experiences of innovation policy makers and improve the regional education and competence level by investments aimed at research and educational institutions, as well as at the creation of regional learning networks. Regarding the conflict of aims between balance and growth policy targets, the question has to be raised whether a strict separation between growth and efficiencyoriented technology policies and balance-oriented regional policies can still be held. An efficiency-oriented, regionally implemented “picking the winner” strategy – although economically deductible from neo-classical equilibrium models – does not provide an answer to the question about the perspectives of regions without prerequisites for high-tech developments and cannot alleviate existing disparities between regions. Thus, from the regional viewpoint, the conflict between balance and growth orientation defuses itself, since each region should have its own interest in pursuing a growth-oriented strategy, which refers to the endogenous strengths and potentials. However, these strengths do not necessarily correspond to leading-edge potentials in the interregional comparison, which can oppose the (national) efficiency target. On the other hand, some theoretical concepts (cf. section 14.2) favor a decentralized competence structure and a stronger distribution of potentials within a state. A regional orientation in policy measures appears to be very promising to exploit additional development and innovation potentials, above all if with
THE REGIONALIZATION OF INNOVATION POLICY
307
comparatively few public funds leverage effects by means of public-private partnerships and the qualification for EU promotional programs can be achieved (cf. ISI et al., 2000:537-538). However, despite these increasing similarities there are still deficits in coordination between both policy arenas, namely, Where the balance orientation in regional policy and the growth orientation of innovation policy (and especially technology policy) come into conflict in regions. Where different measures of individual policy actors compete against each other. Where promoted regions covered by different programs and development plans do not coincide. Where free-rider effects appear, which could have been avoided by better coordination of the measures. Positive impacts on regional growth and regional structural change are then to be expected if it is possible, by a skillful combination of both policy areas, to arrive at a more efficient utilization of public promotional funds (which so far is not always the case). In this way, development processes in individual regions could be initiated in a more targeted manner, which strengthen the innovation and technology competence of enterprises, broaden the regional knowledge base, and give impetus for continual learning process. As a result, the chances could be improved to create regional competence centers within single nations, but above all within the European Research Area, which would contribute to the growth of the economy as a whole and to a reinforcement of the technological and economic competitiveness of Europe.
NOTES 1.
The fuzziness of spillover studies regarding the kind of knowledge externalities is discussed by Breschi and Lissoni (2001).
2.
For different aspects of “embeddedness” (i.e. structural, cognitive, cultural, political, institutional and regional) see also Baum and Oliver (1992), Gulati and Gargiulo (1999), Oerlemans et al. (2001), Staber (2001), Yli-Renko and Autio (1998), Zukin and DiMaggio (1990).
3.
Regional examples of localized and local learning are given by Asheim and Cooke (1999).
4.
According to Hassink (2001:227), a learning region strategy, which he defines as regional innovation strategy, “...will not be successful if it ignores the impact of na-
Knut Koschatzky
308
tional or even international innovation systems on inter-firm co-operation and innovative behaviour.” 5.
Cf. Cantner and Pyka (2001) for a classification of technology policy.
6.
Social capital is defined according to Putnam (1993) as “... features of social organisations such as networks, norms, and social trust that facilitate coordination and cooperation for mutual benefit.”
7.
This does not mean that similar criteria must apply for market-induced developments driven forward by individual enterprises. Due to the complexity of technological options, market developments and historical coincidence, it also applies here that the exception proves the rule.
REFERENCES Amin, A., Thrift, N., “Neo-Marshallian nodes in global networks.” In Ökonomische und politische Netzwerke in der Region. Beiträge aus der internationalen Debatte, W. Krumbein, ed. Münster: LIT Verlag, 1994. Andersson, Å.E., Andersson, D.E., Gateways to the Global Economy. Cheltenham: Edward Elgar, 2000. Anselin, L., Varga, A., Acs, Z.J., Local Geographic Spillovers between University Research and High Technology Innovations. Journal of Urban Economics 1997; 42:422-448. Asheim, B.T., Cooke, P., “Local Learning and Interactive Innovation Networks in a Global Economy.” In Making Connections. Technological learning and regional economic change, E.J. Malecki, P. Oinas, ed. Aldershot: Ashgate, 1999. Baptista, R., Swann, P., Do firms in clusters innovate more? Research Policy 1998; 27:525540. Baum, J.A.C., Oliver, C., Institutional embeddedness and the dynamics of organizational populations. American Sociological Review 1992; 57:540-559. Braczyk, H.-J., Cooke, P., Heidenreich, M., Krauss, G., eds., Regional Innovation Systems: The Role of Governance in a Globalized World. London: UCL Press, 1998. Braczyk, H.-J., Schienstock, G., Steffensen, B., “The Region of Baden-Wuerttemberg: a Post Fordist Success Story?” In Industrial Transformation in Europe. Process and Contexts, E.J. Dittrich, G. Schmidt, R. Whitley, eds. London: SAGE Publications, 1995. Breschi, S., Lissoni, F., Knowledge Spillovers and Local Innovation Systems: A Critical Survey. Industrial and Corporate Change 2001; 10:975-1005. Cantner, U., Pyka, A., Classifying technology policy from an evolutionary perspective. Research Policy 2001; 30:759-775.
THE REGIONALIZATION OF INNOVATION POLICY
309
Carrincazeaux, C., Lung, Y., Rallet, A., Proximity and localisation of corporate R&D activities. Research Policy 2001; 30:777-789. Cooke, P., Regional Innovation Systems: Competitive Regulation in the New Europe. Geoforum 1992; 23:365-382. Cooke, P., Regional Innovation Systems, Clusters, and the Knowledge Economy. Industrial and Corporate Change 2001; 10:945-974. Cooke, P., Regional Innovation Systems: General Findings and Some New Evidence from Biotechnology Clusters. Journal of Technology Transfer 2002; 27:133-145. Cooke, P., Morgan, K., The regional innovation system in Baden-Württemberg. International Journal of Technology Management 1994; 9:394-429. Coombs, R., Richards, A., Paolo, P., Walsh, V., “Introduction: technological collaboration and networks of alliance in the innovation process.” In Technological Collaboration. The Dynamics of Cooperation in Industrial Innovation, R. Coombs, A. Richards, P. Paolo, V. Walsh, eds. Cheltenham: Edward Elgar, 1996. DeBresson, C., Amesse, F., Networks of innovators: A review and introduction to the issue. Research Policy 1991; 20:363-379. Dohse, D., Technology Policy and the Regions: The Case of the BioRegio Contest. Research Policy 2000; 29:1111-1133. European Commission [Kommission der Europäischen Gemeinschaften], Mitteilung der Kommission an den Rat, das Europäische Parlament, den Wirtschafts- und Sozialausschuss und den Ausschuss der Regionen. Hin zu einem europäischen Forschungsraum. Brussels: Kommission der Europäischen Gemeinschaften (KOM (2000) 6), 2000. European Commission [Commission of the European Communities], Communication from the Commission. The Regional Dimension of the European Research Area. Brussels: Commission of the European Communities (COM (2001) 549 final), 2001. Ewers, H.-J., Wettmann, R.W., Innovation-oriented Regional Policy. Regional Studies 1980; 14:161-179. Freeman, C., Perez, C., “Structural crises of adjustment: Business cycles and investment behaviour.” In Technical Change and Economic Theory, G. Dosi, C. Freeman, R. Nelson, G. Silverberg, L. Soete, eds. London: Pinter Publishers, 1988. Grabher, G., “The weakness of strong ties: the lock-in of regional development in the Ruhr area.” In The embedded firm. On the socioeconomics of industrial networks, G. Grabher, ed. London: Routledge, 1993. Grossman, G.M., Helpman, E., Comparative Advantage and Long-Run Growth. The American Economic Review 1990; 80:796-815. Gulati, R., Gargiulo, M., Where do interorganizational networks come from? American Journal of Sociology 1999; 104:1439-1493.
310
Knut Koschatzky
Hassink, R., The Learning Region: A Fuzzy Concept or a Sound Theoretical Basis tor Modern Regional Innovation Policies? Zeitschrift für Wirtschaftsgeographie 2001; 45:219-230. ISI [Fraunhofer Institute for Systems and Innovation Research], Niedersächsisches I n s t i t u t für Wirtschaftsforschung [NIW], Institut für Weltwirtschaft [IfW], Deutsches Institut für Wirtschaftsforschung [DIW], Regionale Verteilung von Innovations- und Technologiepotentiaten in Deutschland und Europa. Final report. Karlsruhe: ISI, 2000. Keeble, D., Small firms, innovation and regional development in Britain in the 1990s. Regional Studies 1997; 31:319-332. Koschatzky, K., Räumliche Aspekte im Innovationsprozess. Ein Beitrag zur neuen Wirtschaftsgeographie aus Sicht der regionalen Innovationsforschung. Münster: LIT Verlag, 2001. Koschatzky, K., Gundrum, U., “Innovation Networks for Small Enterprises.” In TechnologyBased Firms in the Innovation Process. Management, Financing and Regional Networks, K. Koschatzky, ed. Heidelberg: Physica-Verlag, 1997. Koschatzky, K., Sternberg, R., R&D cooperation in innovation systems – Some lessons from the European Regional Innovation Survey (ERIS), European Planning Studies 2000; 8:487501. Krugman, P., A Model of Innovation, Technology, Transfer, and the World Distribution of Income. Journal of Political Economy 1979; 8:253-266. Krugman, P., Geography and Trade. Leuven: Leuven University Press, 1991. Krugman, P., Development, Geography, and Economic Theory. Cambridge: MIT Press, 1995. Krugman, P., What’s new about the new economic geography? Oxford Review of Economic Policy 1998; 14:7-17. Landabaso, M., Youds, R., Regional Innovation Strategies (RIS): the development of a regional innovation capacity. SIR-Mitteilungen und Berichte 1999; 17:1-14. Landabaso, M., Oughton, C., Morgan, K., “Innovation Networks and Regional Policy in Europe.” In Innovation Networks. Concepts and Challenges in the European Perspective, K. Koschatzky, M. Kulicke, A. Zenker, eds. Heidelberg: Physica-Verlag, 2001. Lucas, R. E., On the Mechanics of Economic Development. Journal of Monetary Economics 1988; 22:3-42. Maillat, D., Lecoq, B., Nemeti, F., Technology District and Innovation: The Case of the Swiss Jura Arc. Regional Studies 1995; 29:251-263. Maillat, D., Quévit, M., Senn, L., eds., Réseaux d’Innovation et Milieux Innovateurs: Un Pari pour le Développement Regional. Neuchâtel: Institut de Recherches Economiquees et Régionales, 1993. Marin, B., Mayntz, R., “Introduction: Studying Policy Networks.” In Policy Networks. Empirical Evidence and Theoretical Considerations, B. Marin, R. Mayntz, eds. Frankfurt a.M.: Campus-Verlag, 1991.
THE REGIONALIZATION OF INNOVATION POLICY
311
Meyer-Krahmer, F., Der Einfluss staatlicher Technologiepolitik auf industrielle Innovationen. Baden-Baden: Nomos Verlagsgesellschaft, 1989. Meyer-Krahmer, F., “Technologiepolitik.” In Handbuch zur Arbeitslehre, D. Ropohl, A. Schmid, eds. München: Oldenbourg, 1997. Nelson, R.R., National Innovation Systems. A Comparative Analysis. New York: Oxford University Press, 1993. Oerlemans, L., Meeus, M., Boekema, F., On the spatial embeddedness of innovation networks: an exploration of the proximity effect. Tijdschrift voor Economische en Sociale Geographie 2001; 92:60-75. Oinas, P., On the Socio-Spatial Embeddedness of Business Firms. Erdkunde 1997; 51:23-32. Oughton, C., Landabaso, M., Morgan, K., The Regional Innovation Paradox: Innovation Policy and Industrial Policy. Journal of Technology Transfer 2002; 27:97-110. Porter, M.E., The Competitive Advantage of Nations and Their Firms. London: Macmillan, 1990. Porter, M.E., Clusters and the new economics of competition. Harvard Business Review 1998; November-December:77-90. Putnam, R, Making Democracy Work – Civic Traditions in Modern Italy. New Jersey: Princton University Press, 1993. Pyke, F., Sengenberger, W., eds., Industrial districts and local economic regeneration. Genf: International Institute for Labour Studies, 1992. Ratti, R., Bramanti, A., Gordon, R., eds., The Dynamics of Innovative Regions. The GREMI Approach. Aldershot: Ashgate, 1997. Romer, P.M., Increasing Returns and Long-Run Growth. Journal of Political Economy 1986; 94:1002–1037. Romer, P.M., Endogenous Technological Change. Journal of Political Economy 1990; 98:71102. Sassen, S., Metropolen des Weltmarkts, Die neue Rolle der Global Cities. Frankfurt a.M.: Campus Verlag, 1996. Staber, U., “Networks and Regional Development: Perspectives and Unresolved Issues.” In Business Networks. Prospects für Regional Development, U.H. Staber, N.V. Schaefer, B. Sharma, eds. Berlin: Walter de Gruyter, 1996. Staber, U., Spatial Proximity and Firm Survival in a Declining Industrial District: The Case of Knitwear Firms in Baden-Württemberg. Regional Studies 2001; 35:329-341. Steiner, M., “The Discreet Charm of Clusters: An Introduction.” In Clusters and Regional Specialisation, M. Steiner, ed. London: Pion, 1-17 (European Research in Regional Science 8), 1998.
312
Knut Koschatzky
Storper, M., The Resurgence of Regional Economies, Ten Years Later: The Region as a Nexus of Untraded Interdependencies. European Urban and Regional Studies 1995; 2:191-221. Tijssen, R.J.W., Quantitative assessment of large heterogenous R&D networks: the case of process engineering in the Netherlands. Research Policy 1998; 26:791-809. Tödtling, F., Innovation Networks, Collective Learning, and Industrial Policy in Regions of Europe. European Planning Studies 1999; 7:693-697. Yli-Renko, H., Autio, E., The network embeddedness of new, technology-based firms: developing a systemic evolution model. Small Business Economics 1998; 11:253-267. Zukin, S., DiMaggio, P., eds., Structures of Capital: The Social Organization of the Economy. Cambridge: Cambridge University Press, 1990.
Index absorptive capacity, 128, 132–4, 136–7, 143, 258, 292, 295, 298 adaptability, 64–6, 71, 75–7, 261, 264–5 agglomeration, 54, 94, 98–9, 101, 130, 183, 257, 264, 273, 284–5 diseconomies, 54 industrial, 198, 276 agriculture, see industry Belgium Hainaut, 178 best practice, 1, 4, 9, 11, 14–15, 26, 65, 73, 76, 85, 120, 216, biotechnology, 49–50, 58, 90, 128, 139, 140, 142–3, 199, 258, book publishing, 147, 159, 160, 162, 165 business foundings, see new firm formation causal ambiguity, 120 chain commodity, 280 value, 45–6, 48, 50, 52, 59, 156, 158, 275, 277, 279–82 change localized, 250–1, 25–4, 257–63, 265–8 structural, 250–1, 254–9, 263–4, 266–7, 274, 279, 292–5, 298, 302, 304, 307 technological, 148, 150, 303 cluster, 46, 48–50, 52–4, 59, 72, 127, 129, 148, 156, 158–60, 198, 200, 216, 273, 275–9, 286, 291, 293–4, 296–7, 305 business, 107, 210 dimension, 158–9 high technology, 198 industrial, 302 industry, 155, 200, 216, 276 insular, 278 media, 162–5 policy, 282 service, 199 social capital, 45 software, 51, 53–5, 57, 60 spinout firms, 55 sub-cluster, 52 technological, 301 see also regional cluster
314
collaboration, 68–70, 72, 75–6, 129, 159, 214–6, 277–8, 286 external, 69, 142 inter-firm, 206 networks, 78 recurrent, 75 research, 135 temporary, 66–8, 75, 77, 80 collective process, 26, 252 communities of practice, 68, 72, 83 consumption, 67, 280, 282–3 co-operation, 294 core assumptions, 276 creativity, 76–7, 83, 250, 254–5, 258, 263, 267 crisis, 11–2, 64, 147, 154–6, 158, 162, 165 customer aspatial, 282 choice, 281 demand, 46, 279, 281–2, 286 footloose, 283 long-distance, 282 preference, 281, 283 discourse analysis, 2 ecology organizational, 66, 115 project, 66, 75–7 economic geography, 275–6, 278, 280, 293 new economic geography, 291, 293 embeddedness, 83, 307 institutional, 83 loose, 121 social, 112, 294 entrepreneurship, 108–117 epistemological position, 8 European Research Area, 186, 292, 296 European Union Cohesion Fund, 171, 175 Community Support Framework (CSF), 175, 177, 186 Less-Favored (Objective 1) regions, 171–172 regional convergence-divergence, 173–174 regional disparities, 174–177 Regional Innovation Strategies (RIS), 292, 298
INDEX
INDEX
Structural Funds, 177 technology, institutions, and regional development, 185–186, 296 evolution industry, 118, 125 local media sector, 160 organizational, 76 social, 7 evolutionary economics, 250–1, 259, 260 exploitation, 44, 47, 51–2, 137, 141–2, 276 externalities, 129–32, 143 geographical, 128–9 knowledge, 130, 134, 294, 307 network, 5, 64, 81 technology, 131 transmission channels, 143 transmission of knowledge, 132 urbanization, 264 film production, see industry Fordism, 13 France agricultural food processing industry, 139–143 Alsace, 141 Aquitaine, 141 Auvergne, 141 biotechnology, 140 Brittany, 141 Corsica, 178–179, 189 Hainaut, 178 Ile de France, 141, 301 Rhône-Alpes, 141, 301 small and medium-size enterprises (SMEs), 139–143 General Agreement on Tariffs and Trade (GATT), 12 generative growth, 56, 59 Germany Aachen, 102–103 adoption of American models, 4, 13 Baden-Württemberg, 85, 223–224, 228–231, 232–244, 301 Berlin, 86, 92, 94–104, 162, 178 Center for Technology Assessment, Stuttgart, xix, 223 East Germany (German Democratic Republic), 1, 160–162 Leipzig, 159–165 Munich, 92, 94–102, 236
315
316
INDEX
Potsdam, 162, 163 reunification and new länder, 177 Ruhr, 26, 86 Saxony, 162–163 Saxony-Anhalt, 162–163 Steinbeis Foundation, 232, 246 Stuttgart, 102, 230, 231–244 Thuringia, 163 geographical industrialization, 153–4 global services, 96, 100 globalization, 3, 5, 273–4, 282–3, 292 governance, 50, 55, 59, 70, 275, 294–5 Greece regions, 179–180 Thessaly, 189 Western Macedonia, 189 heterarchy, 83 history, 5–7, 44, 55, 63–4, 107–8, 114, 122, 147, 150, 165, 198, 249–52, 257, 260, 267 imitation, 2, 10, 114–5, 117, 119–20, 136–7, 141, 156,254 increasing returns, 6, 64, 107, 250, 255, 257–8, 265–7 India Bangalore, 53–55 equity investment models, 53–55 Hyderabad, 54 industrial districts, 107, 109, 160, 200, 294, 302 policy, 273 industry advertising, 66, 70–75, 77, 81 agricultural food-processing (AFI), 139–43 film production, 72, 159–60, 162–3 media, 66, 147, 162, 164–5 innovation, xvi, 4, 7–8, 13, 43, 44–6, 48–50, 52, 55, 58–9, 66, 70, 75, 107– 12, 114–5, 117, 119–21, 127–34, 137–43, 148–9, 151–2, 156, 158, 195, 198–02, 204, 206, 212, 214–8, 250, 259, 261, 267, 273–7, 279, 280–6, 292–307 institutional, 14, 160 networks, 131, 143, 299, 301 organizational, 46 policy, 217, 249, 274, 276, 291–3, 295–8, 302–7 regional, see regional innovation
INDEX
317
social process, 150 strategy, 8, 205, 274, 300, 305, 307 system(s), 43, 48, 50, 52, 107, 109, 127, 134, 260–1, 273–8, 282, 286, 294, 297, 299, 301 technological, organizational and social, 293 institutional compatibility, 12 learning, 1 legacy, 7, 14 transfer, 1, 2, 7–11, 13–5 interaction face-to-face, 71, 138, 275 interdependencies, 49, 75, 118–20, 261, 278, 294 Ireland, Republic of Dublin, 57 East (region), 178 skills upgrading, 55–59 regional development, 178–179, 186 South (region), 178 Israel equity investment models, 51–53 Herzliah Corridor (Tel Aviv–Haifa), 51 Yozma, 51–52 Italy Abruzzi, 178 Calabria, 176 Emilia-Romagna, 176 Lombardy, 301 Molise, 178–179, 189 Sicily, 176 Third Italy, 85, 185 Tuscany,176 knowledge codified, 133 economy, 43, 46–7, 59, 87, 89, 93–6, 98, 103–4 external, 132, 134–5, 137, 300 industry, 89, 92, 96, 98, 101 market, 87, 90, 135 procedural, 12–13 tacit, 24–37, 49, 121, 129, 133, 135, 149, 152, 252, 296 labor market, 1, 72, 284–6
318
INDEX
leapfrogging, 85, 95, 102 strategies, 102–3 learning by switching, 70, 76–7 institutional, see institutional learning interactive, 147–52, 155–6, 159, 164–5, 261,275 region(s), 27-30, 37, 294, 307 life cycle, 44, 142, 281, 296 lock-in, 63–9, 75, 77, 117, 122, 150, 223, 249, 250–1, 253–4, 258–5, 279, 294, 297 loose coupling, 66, 68, 262 Marshall(ian), 71, 155, 160, 284–5 network(s), 7, 46, 48–50, 58, 66, 68–70, 72–6, 83, 115–7, 120, 123, 141–2, 148, 153, 155, 158, 164, 199–200, 253, 259, 261–2, 277, 283, 291, 294–5, 299–300, 306, 308 corporate, 67, 73, 75 local, 141 personal, 66, 73, 75, 112, 117 social, 108, 112, 116–7, 119 see also collaboration, externalities, innovation new firm formation, 50, 57, 110, 115, 164, 285 Netherlands, The Fevoland, 178 New Institutional Economics, 2, 5–6 organizational forms, 65–6, 73, 76, 108, 110–1, 114–5, 121–3 path dependency, xi–xiii, 1–2, 5–8, 12, 14, 24–5, 63, 85, 91, 107–10, 117, 150, 198, 218, 223, 253, 286, 302 policy, see regional policy, innovation policy Portugal Algarve, 189 Lisbon and Tagus Valley, 178 positive feedback, 64–5, 70, 76, 107, 111, 120, 250 production culture, 25 proximity, 46, 49, 70, 135–6, 138–9, 153, 160, 264, 275, 293–5 advantages, 293 economics, 134 geographical, 127–32, 134–8, 141–3 information / knowledge, 128–9 organizational, 134, 136–8, 141, 144 Public Choice, 3, 20 QWERTY, see typewriter keyboard re-bundling, 148, 156–7, 159, 161–2, 164–5
267,
127, 276,
115,
INDEX
319
redundancy, 65–6, 70, 76, 80 regional assets, 156, 158, 200, 274 bundle, 148, 156, 160, 162, 165 cluster, 59, 158, 200, 273–4, 277–80, 282–6, 296, 306 development, 26, 95, 98, 107, 147–8, 151, 153–7, 159, 165, 185–6, 191, 198, 200, 214, 217, 250, 259, 262, 265, 267, 294, 297, 299–300, 302, 304–5, 306 development path, 147–8, 151, 154–8, 160, 162, 165, 305 development policy, 198 innovation, 47, 50, 55, 59, 151, 199–200, 219, 274, 286, 292, 295–9, 307 innovation policy, 198, 250, 259, 263, 267–8, 295, 305 innovation systems, 129, 199, 218, 291, 294–7, 302 policy, 46, 108, 110, 123, 249–50, 259, 260–4, 266–7, 292, 295–6, 298, 303, 305–7 rupture, 147, 154–5, 158, 160–2, 165 regulationist approach, 147, 154 relational process, 149 research and development (R&D), 46–7, 49, 51, 127, 129–39, 142–3, 197, 204, 208, 211, 219, 257, 261, 274, 276, 283, 294, 296, 299, 301, 303–4 rivalry, 65–68, 72–76, 158, 276–278, 281 internal, 74 inter-firm, 158 local, 277 selection, 1, 9, 63–5, 69, 107, 113, 121, 123, 151, 250–3, 255, 257, 260, 262, 265 environment, 65, 118, 120, 122, 251, 254–5, 258, 263 strong, 254 servindustrial economy, 89, 91, 96, 102 Singapore, 301 small and medium-sized enterprises (SMEs), 48, 139–140, 299 social capital, 45 social conformity, 111 Spain Cantabria, 178 Catalonia, 301 spillovers, 118, 127, 130–3, 256, 285, 293, 297 spin-offs, 164, 211, 285 subsidiarity principle, 298 survival of the fittest, 63
320
Sweden academic entrepreneurship, 47–51 Kista, 47–48 Linköping, 47–49 Lund University, 49–50 Taylorism, 13 technological complexity, 141 Toyota system, 14 theory evolutionary, 63, 115, 118 Gramscian power, 10 new growth, 200, 293–4 new trade, 291, 293 of comparative institutional advantages, 4, 14 trajectory, 55, 64–5, 148, 150–4, 156–8, 165, 251, 265 transaction cost, 71, 158 transnational diffusion, 7 typewriter keyboard, 5, 64, 69 United Kingdom Belfast, 58 Cornwall and Isles of Scilly, 178 Highlands and Islands, 178 Merseyside, 178 Northern Ireland, 178–179 Soho (London), 66, 70–75, 78 South Yorkshire, 178 South-East (England), 173–174 West Wales and Valleys, 178 United States California effect, 7 Georgia, 197, 201–19 Atlanta, 201, 212 Advanced Technology Development Center, 208 Georgia Institute of Technology (Georgia Tech), xix, 206 Georgia Tech Economic Development Institute, 206–207 Georgia Research Alliance, 208–209 Yamacraw Initiative, 209–210 industrial cluster policies, 200 Manufacturing Extension Partnership, 199, 208 New South, 197 Northern Kentucky, 200 regional economic development policies, 198–201
INDEX
INDEX
321
regional per capita income trends, 195–197 Research Triangle (North Carolina), 198 Route 128 (Boston, Massachusetts), 26, 44, 198, 301 San Francisco, 43–44 Silicon Valley (California), 31, 37, 43–46, 198, 249, 296, 301 state technology policies, 198 university, 47–9, 53, 56–8, 123, 132, 138, 198–9, 201, 206, 209–11, 214, 264–5 industry, 134–5, 198, 209, 261 network, 141 research, 46–7, 52, 132, 198, 205, 209, 211, 217–8, 277, 286, spin-offs, 164 spinout firms, 47–9 system, 205, 206, 210 value added, 282 venture capital, 4, 44–6, 48–51, 54–5, 57–8, 123, 198, 200, 208–10, 256, 261, 264, 268, 296, 301 windows of locational opportunity, 147–8, 155
1. A. Phillips, A.P. Phillips and T.R. Phillips: Biz Jets. Technology and Market Structure in the Corporate Jet Aircraft Industry. 1994 ISBN 0-7923-2660-1 2. M.P. Feldman: The Geography of Innovation. 1994 ISBN 0-7923-2698-9 3. C. Antonelli: The Economics of Localized Technological Change and Industrial Dynamics. 1995 ISBN 0-7923-2910-4 4. G. Becher and S. Kuhlmann (eds.): Evaluation of Technology Policy Programmes in Germany. 1995 ISBN 0-7923-3115-X 5. B. Carlsson (ed.): Technological Systems and Economic Performance: The Case of Factory Automation. 1995 ISBN 0-7923-3512-0 6. G.E. Flueckiger: Control, Information, and Technological Change. 1995 ISBN 0-7923-3667-4 7. M. Teubal, D. Foray, M. Justman and E. Zuscovitch (eds.): Technological Infrastructure Policy. An International Perspective. 1996 ISBN 0-7923-3835-9 8. G. Eliasson: Firm Objectives, Controls and Organization. The Use of Information and the Transfer of Knowledge within the Firm. 1996 ISBN 0-7923-3870-7 9. X. Vence-Deza and J.S. Metcalfe (eds.): Wealth from Diversity. Innovation, Structural Change and Finance for Regional Development in Europe. 1996 ISBN 0-7923-4115-5 10. B. Carlsson (ed.): Technological Systems and Industrial Dynamics. 1997 ISBN 0-7923-9940-4 11. N.S. Vonortas: Cooperation in Research and Development. 1997 ISBN 0-7923-8042-8 12. P. Braunerhjelm and K. Ekholm (eds.): The Geography of Multinational Firms. 1998 ISBN 0-7923-8133-5 13. A. Varga: University Research and Regional Innovation: A Spatial Econometric Analysis of Academic Technology Transfers. 1998 ISBN 0-7923-8248-X 14. J. de la Mothe and G. Paquet (eds.): Local and Regional Systems of Innovation, 1998 ISBN 0-7923-8287-0 15. D. Gerbarg (ed.): The Economics, Technology and Content of Digital T V, 1999 ISBN 0-7923-8325-7 16. C. Edquist, L. Hommen and L. Tsipouri Public Technology Procurement and Innovation, 1999 ISBN 0-7923-8685-X 17. J. de la Mothe and G. Paquet (eds.): Information, Innovation and Impacts, 1999 ISBN 0-7923-8692-2
18. J. S. Metcalfe and I. Miles (eds.): Innovation Systems in the Service Economy: Measurement and Case Study Analysis, 2000 19. R. Svensson: Success Strategies and Knowledge Transfer in Cross-Border Consulting Operations, 2000 20. P. Braunerhjelm: Knowledge Capital and the “New Economy”: Firm Size, Performance and Network Production , 2000 21. J. de la Mothe and J. Niosi (eds.): The Economic and Social Dynamics of Biotechnology, 2000 22. B. Guilhon, (ed.): Technology and Markets for Knowledge: Knowledge Creation, Diffusion and Exchange within a Growing Economy, 2000 23. M. Feldman and A. Link (eds.): Innovation Policy in the Knowledge-Based Economy, 2001 24. J. de la Mothe and D. Foray (eds.): Knowledge Management in the Innovation Process 2001 25. M. Feldman and N. Massard (eds.): Institutions and Systems in the Geography of Innovation, 2002 26. B. Carlsson (ed.): Tehnological Systems in the Bio Industries: An International Study, 2002 27. D. Audretsch, R. Thurik, I. Verheul, and S. Wennekers (eds.): Entrepreneurship: Determinants and Policy in a European-US Comparison, 2002 28. J. de la Mothe and A. Link (eds.): Networks, Alliances and Partnerships in the Innovation Process 28. F. Belussi, G. Gottardi, and E. Rullani (eds.): The Technological Evolution of Industrial Districts 30. G. Fuchs and P. Shapira (eds.): Rethinking Regional Innovation and Change: Path Dependency or Regional Breakthrough
SPRINGER
ISBN 0-7923-7730-3
ISBN 0-7923-7776-1
ISBN 0-7923-7801-6
ISBN 0-7923-7922-5
ISBN 0-7923-7202-6
ISBN 0-7923-7296-4 ISBN 0-7923-7464-9
ISBN: 0-7923-7614-5
ISBN: 0-7923-7633-1
ISBN: 0-7923-7685-4
ISBN: 1-4020-7172-8 ISBN: 1-4020-7555-3
ISBN: 0-387-23001-7