The new economy in transatlantic perspective: Spaces of innovation

Hübner, Kurt (Ed.) Book The new economy in transatlantic perspective: Spaces of innovation Routledge studies in governance and change in the global era, No. 4 Provided in Cooperation with: Taylor & Francis Group Suggested Citation: Hübner, Kurt (Ed.) (2005) : The new economy in transatlantic perspective: Spaces of innovation, Routledge studies in governance and change in the global era, No. 4, ISBN 978-0-203-42096-6, Routledge, London, https://doi.org/10.4324/9780203420966 This Version is available at: https://hdl.handle.net/10419/240576 Standard-Nutzungsbedingungen: Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden. Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen. Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Terms of use: Documents in EconStor may be saved and copied for your personal and scholarly purposes. You are not to copy documents for public or commercial purposes, to exhibit the documents publicly, to make them publicly available on the internet, or to distribute or otherwise use the documents in public. If the documents have been made available under an Open Content Licence (especially Creative Commons Licences), you may exercise further usage rights as specified in the indicated licence. https://creativecommons.org/licenses/by-nc-nd/4.0/ The New Economy in Transatlantic Perspective What’s left from the new economy? This book takes an unfashionable perspective and shows that despite all the mistaken ideas and exaggerations, the technological changes of the 1990s still have important effects today. Economic history shows that technological revolutions tend to generate deep economic and social crises before a temporary state of equilibrium is reached. The established modes of accumulation and regimes of regulation of national capitalisms and international capitalism have been undermined by the collapse of the high tech asset bubble. Financial markets are still in disarray. What can be observed, however, is that some national economies are better positioned to tackle the crisis than others. Why is this? This and other important questions are tackled by an international team of contributors including Daniele Archibugi, Harald Hagemann, Bruno Amable, Martin Heidenreich and David Gibbs. This volume should be of great interest to all those working at the intersection of international politics and economics. Kurt Hübner is professor of Political Science and the director of the Canadian Center for German and European Studies, York University, Toronto. Routledge studies in governance and change in the global era 1 New Socialisms Futures beyond globalization Edited by Robert Albritton, John Bell, Shannon Bell and Richard Westra 2 International Crisis Management The approach of European states Marc Houben 3 Citizenship in a Global World European questions and Turkish experiences Edited by E.Fuat Keyman and Ahmet Içduygu 4 The New Economy in Transatlantic Perspective Spaces of innovation Edited by Kurt Hübner Other titles in this series include: States Against M arkets Edited by Robert Boyer and Daniel Drache Political Ecology Edited by David Bell, Lessa Fawcett, Roger Keil and Peter Penz Health Reform Edited by Daniel Drache and Terry Sullivan Democracy, Citizenship and the Global City Edited by Engin F.Isin Between the Market and the Public Domain Redrawing the line Edited by Daniel Drache The New Economy in Transatlantic Perspective Spaces of innovation Edited by Kurt Hubner First published 2005 by Routledge Published 20 17 by Routledge 2 Park Square, Milton Park, Abingdon, Oxon OX14 4RN 711 Third Avenue, New York, NY 10017, USA Routledge is an imprint of the Taylor & Francis Group, an iriforma business Copyright © 2005 editorial matter and selection, Kurt Hubner; individual chapters, the contributors Typeset in Times by Wearset Ltd, Boldon, Tyne and Wear The Open Access version of this book, available at www.tandfebooks.com, has been made available under a Creative Commons Attribution-Non Commercial-No Derivatives 4.0 license. The publisher makes no representation, express or implied, with regard to the accuracy of the information contained in this book and cannot accept any legal responsibility or liability for any errors or omissions that may be made. British Library Cataloguing in Publication Data A catalogue record for this book is available from the British Library Library of Congress Cataloging in Publication Data A catalog record for this book has been requested ISBN13: 978-0-415-33608-6 (hbk) ISBN13: 978-0-415-40692-5 (pbk) Contents List of illustrations vii Notes on contributors x Acknowledgments xii PART I Macroeconomics of innovation 1 1 Spaces of innovation: introductory remarks on thecomparative political economy of the new economy KURT HÜBNER 3 2 The new economy in a growth crisis GEORG ERBER AND HARALD HAGEMANN 21 3 Innovations, economic growth and productivity in the new economy MARTIN ZAGLER 45 PART II Institutional matrixes 63 4 Is there an institutional base of thenew economy? BRUNO AMABLE 65 5 Europe in the innovation race DANIELE ARCHIBUGI AND ALBERTO COCO 87 6 Innovation and social security: an international comparison MARTIN HEIDENREICH 109 7 Transnationalization of European governance in the information age: the role of policy networks DAVID GIBBS 133 PART III Spaces of innovation 153 8 Urban governance, interspatial competition and thepolitical geographies of the new economy : reflections on the western European case NEIL BRENNER 155 9 Do regional systems of innovation matter? MICHAEL FRITSCH 189 10 Internet-based electronic business: a sociology of discontinuities and failures of new companies in the fields of entertainment and technology in California GERHARD KRAUSS 207 11 The new economy assets of the Berlin Metropolitan region: development chances and threats STEFAN KRÄTKE 231 Index 261 vi Illustrations Figures 1.1 Technological innovation and regulation 11 2.1 Moore’s Law 25 3.1a Growth and inflation, USA 1961–1969 48 3.1b Growth and inflation, USA 1982–1990 49 3.1c Growth and inflation, USA 1991–2000 49 3.2a Growth and inflation, EU15 1960–1973 50 3.2b Growth and inflation, EU15 1983–1992 50 3.2c Growth and inflation, EU15 1994–2000 51 3.3 Growth and inflation, UK 1993–2001 51 3.4 Growth and inflation, Austria 1993–2001 52 4.1 The five types of capitalism in two dimensions 78 4.2 Principal components analysis; projection of countries on the first factorial plane 79 6.1 Public social expenditure and gross domestic expenditure for research and development (in percentage of gross domestic product, 24 OECD countries, 1999) 115 8.1 Spatial Keynesianism and the political regulation of uneven development 165 8.2 The changing European urban hierarchy 170 8.3 Brunet’s blue banana 172 8.4 Glocalization strategies and the new politics of uneven development 177 8.5 State strategies and the political regulation of uneven development: a schematic periodization of the western European case, 1960–2000 182 9.1 Main actors in a regional innovation system 194 11.1 Regional economic profile of German metropolitan regions (1996): Berlin compared to Hamburg, Munich and Frankfurt 234 11.2 Local clustering of the media industry in Berlin (2000): multimedia firms (producers and service providers) 243 11.3 Media cities in Germany 247 11.4 Global media cities in Europe: classification based on local concentrations of global media firms’ establishments in European cities 2001 251 2 1 Spaces of innovation Introductory remarks on the comparative analysis of the new economy Kurt Hübner The new economy and beyond Reports on the death of the new economy are widely exaggerated. A couple of years after the failure of the so-called new economy, the drama of the burst of the dot.com bubble of 2001, it takes little courage for a statement like this. A closer inspection of the preas well as the post-bubble literature shows that today’s condemnation of project new economy is as strong and widespread as was its praise during the 1990s. Both attitudes are understandable, at least to some degree, but nevertheless misleading. The expectation that the new information and communication technologies (ICTs)1 would fundamentally change the nature of a capitalist money economy and prepare the way to a nirvana economy were unfounded from the beginning, at least if one takes the historical experiences of previous fundamental technological changes as well as the well-established knowledge of economic theory as measuring rods. New technologies have neither banished the business cycle nor have they abolished inflation and unemployment. However, there is no doubt that ICTs have shaken the trajectories of developed capitalist economies in fundamental ways. New lead sectors have been established and existing sectors have undergone technical and to some degree organizational transformations: skill profiles of employees have been changed and the technical content of physical capital goods has been altered. In a longer perspective, the economic surge in the 1990s and the years since the burst of the financial bubble should be understood as two sides of a deep-seated structural change of the given regime of accumulation and mode of regulation that are driven by the launch of fundamental technological changes. The appellation new economy, therefore, should not be taken literally but as a preliminary name for a growth regime that will or will not ultimately establish a compatible regime of accumulation and mode of regulation with the potential for a renewed golden age of capitalism. When exactly such a regime will evolve and where this will happen only time will tell. From today’s perspective it seems to many observers that there is only one answer to those questions: the new growth regime will be established in the US.2 This proposition carries at least one implicit thesis, namely that Europe will fall back in the global race for innovations. Even though the financial meltdown was area-wide, the financial and technological resources that are still available in the US seem more than sufficient for the resurrection of ICT activities. While good reasons for such a perspective exist, the same perspective holds for some of the European economies. The burst of the bubble may not have changed the race totally but reshuffled the pole positions. What used to be mainly a US-dominated game has turned into a competition between spaces of innovations on both sides of the Atlantic. Economies that have the capacity to generate fundamental technological innovations or have immediate access to them have the advantage of reaping the economic fruits of this fundamental change, usually in above-average growth rates of GDP and productivity. In this respect it seems almost dramatic that the economies of the European Union not only experienced a stop of their catch-up process with the US since 1995 but lost until 2003 one-fifth of their gain in output per hour relative to the US from the period 1950 to 1995 (Gordon 2004). This reversal in growth dynamics has raised not only political concern in European capitals but also started an intense academic debate about the reasons for this relative decline. In contrast to the mainstream analysis of European growth of GDP and employment,3 the contributors of this volume underline the centrality of innovations for economic growth. While the analytical connection between innovations and economic growth is very well established, there is an enduring discussion about the numerical impact of specific technologies on growth dynamics. Usually, three channels of influence are distinguished. (i) Innovations trigger investments, which improve the quality of the capital stock and generate increases in productivity. (ii) Innovations lead to new sectors of production which enhance employment. (iii) The use of product and process innovations shapes the methods of production on the company level and contributes to the improvement of competitiveness.4 As most of the chapters of this book show, such a Schumpeterian approach has to come with a distinct institutional flavor if it wants to overcome the blind spots of growth accounting practices. Technical innovations can be the motor of economic growth, but there is urgent need for good brakes, a navigation system as well as a developed infrastructure to allow for a sustainable drive. It is a wellestablished fact that not all economies have the same capacity to generate or access technological change at the same pace at the same point in time. Sometimes forgotten seems to be the further insight that not all national economies possess the social capabilities (Abramovitz 1986) to create the institutional matrix for successful innovatory activities. There exists not only a huge gap in the distribution of ICTs between the club of the OECD economies and the rest of the world, buzzed as the digital divide, but 4 KURT HÜBNER this gap also exists inside the OECD club. A few forerunners and a host of latecomers divide the pie unevenly.5 It would be a simplification, however, to draw a picture of North American ICT-haves and European ICT-have-nots. The reality is highly nuanced. Although it is true that the US is leading the pack and Europe fell behind decisively during the 1990s, such a picture only gives a rough and cloudy image. Aside from the many practical problems for comparisons that arise out of highly different methods of measurement of ICT-activities in national accounts statistics,6 OECD research teams have been relatively successful in gathering a pool of data that allows a comparative analysis of ICT between those two economic blocs (OECD 2004; van Ark et al. 2003). Those data show that a small group of European economies have been very successful in installing, developing and exploiting ICTs; Nordic countries such as Finland, Denmark and Sweden, also economies such as the British, for some indicators show an even better performance than the forerunner the US. Based on a back-of-the-envelope calculation using data on the share of ICT investment in overall gross capital formation, the share of ICT value-added in overall private business value-added and for the IT capital stock per hour worked,7 the Nordic countries (Sweden, Finland and Denmark) are very close to the US. A second group consisting of the biggest European economies Germany and France on the one side and the two small economies of the Netherlands and Ireland on the other side follow those forerunners. They tried hard to keep up with the US and the European lead economies in ICT but only with mediocre results so far. The third group of the laggard economies consists of Greece, Spain, Portugal and Italy, which lag far behind with respect to every single indicator. It is this olive belt which makes the image of a digital transatlantic divide. While only sketched in charcoal, this picture is pretty much similar to other empirical attempts to measure the technological capabilities of capitalist market economies.8 More concise calculations of the impact of ICTs show that the usage hypothesis which has been brought forward by analyses of the US case, is not very convincing in the case of Europe. First, the US user industries (finance, wholesale and retail trade) that show the highest growth rates of productivity during the 1990s have no counterparts in Europe. Second, improvements in productivity in Europe mainly came from the production side of ICTs. Both findings refer to the fact that diffusion processes seem to be far more developed in the US regime of growth than in its European counterparts (von Ark et al. 2003; Daveri and Silva 2004).9 As Amable argues in this volume, the degree of usage and production of new technologies in a given economy depends very much on the institutional base of a growth regime. If one accepts the argument that the capacity to generate and access cuttingedge technologies is the “single most important force driving the secular process of economic growth” (Bresnahan and Trajtenberg 1995: 1),10 then the uneven spatial distribution of this capacity seems to be one of the keys to understanding the halt of the secular convergence process that occurred during the 1990s. Contrary to the arguments of mainstream economics, it is not technological THE NEW ECONOMY AND BEYOND 5 progress per se but the ingenious coupling of technological and institutional changes that drives economic growth in the medium term.11 Viewed from this perspective, it was the successful combination of technological breakthroughs and social innovations that gave Europe the world economic leadership in the nineteenth century; and it was the transition toward a regime of mass production and mass consumption that evolved jointly with a new economic role for the state in the process of capital accumulation that allowed the US to overtake the European lead economies in the early twentieth century. History amply demonstrates that those periods of leadership do not last forever. While there are serious doubts about the argument that leadership comes and goes in a regular wave-like pattern where the former forerunners turn into losers and the latecomers come to the front, it has been convincingly shown that forerunners have not been able to hold their position over the long term (Arrighi and Silver 1999; Reuven y and Thompson 2001). Dominant nations in the global economy come and they usually go in irregular fashion. This seems to be different in the current situation. Unlike the avant-garde of European economies in the nineteenth century that lost their formerly superior position to the US, this dominant power renewed its ability to lead the innovative race and to safeguard its comparative advantages during the 1990s. Even though the reasons for this resurrection are manifold, it seemed obvious at the time that it was driven by technological innovations, mainly ITCs, generated and made exploitable by the smooth working of the normative principles of a free market economy. The governments of the European Union’s member states acknowledged the challenge of the revival of the US at their Lisbon Summit in 2000 and agreed on the goal to make the EU by 2010 into “the most competitive and dynamic knowledge-based economy in the world, capable of sustainable economic growth with more and better jobs and greater social cohesion” (European Union 2000). The main instrument to make this happen would be an increase in research and development expenditures. The Barcelona Summit in 2002 confirmed this approach by explicitly targeting R&D expenditures up to 3 percent of EU-GDP in 2010 (European Union 2000). Given current expenditures, it would need an increase of about 33 percent up from 2004 to make this happen (Meister and Verspagen 2004). Even though it is a well-established fact that R&D expenditures show positive effects for economic growth, this linkage neither works automatically nor does growth get proportionally stronger with an increase in the share of R&D expenditures in GDP (Rodriguez-Pose 1999). The new economy of the 1990s is the offspring of innovations that date back at least to 1971 when Intel introduced the 4004 computer chip and prepared the ground for the personal computer.12 ICTs as the technological base of this evolving growth regime can best be understood as General Purpose Technologies (GPTs) that generate new products, new ways of producing new and old products and services as well as new ways to think about generating and using new technologies (Bresnahan and Trajtenberg 1995). Such transforming 6 KURT HÜBNER technologies neither fall from heaven nor do they appear very often in history. Steam and electricity are two of the few examples of GPTs that students of economic and technological history agree upon (Jovanovic and Rousseau 2003). Following the work of David (1999), Crafts (2002) and others, ICTs will be conceptualized as the latest version of this kind of technology (see Erber and Hagemann and Zagler in this volume). ICTs can be characterized as an enabling technology that opens possibilities rather than provides final products. Such technologies not only need innovational complementarities (Bresnahan and Trajtenberg 1995:84) in the form of adequate infrastructures, but even more so need complimentary institutional forms to come to grips with their main defining feature, namely uncertainty. As Lundval (1998:407) reminded us, innovation is by definition a process where “all alternative outcomes cannot be known in advance” and carries a degree of uncertainty that cannot be clarified by consideration of previous probabilistic distributions. This area of the unknown opens up the chances for new products and processes that come with extra profits for the successful innovator. There are also increased chances for big failures. The short history of the new economy illustrates this two-headed monster. From boom to bust There are at least two versions of the tale of the new economy; both deal with the nexus of ICT and financial markets. One story talks about the upbeat expectations of the financial markets during the 1990s that helped to start a self-fulfilling technological/financial virtuous circle, nourished by the success stories of a few forerunners in the technological race. The eBays, Amazons and Oracles, SAPS and Microsofts of this world have not only changed the ways we think of companies and their modes of operation. It is foremost the products they and others have marketed and integrated into the daily life of average households of OECD-economies that have changed consumer as well as producer bundles of goods and services. Even though it is obvious that the Internet and the computer have drastically altered ways of communications and economic as well as social forms of transactions, be it the way products and services are produced, or be it how they were delivered, it is less clear what the economic effects of the new products and processes look like. Neither is there any kind of clear-cut vision of how the optimal microeconomic business models and macro modes of regulation will and should look. Extrapolation from individual success stories to the macroeconomic level was, as investors had to learn the hard way, a problematic procedure. In the second half of the 1990s it seemed as if the lead sectors of technological renewal for a few developed economies passed the critical threshold, enabling them to harvest the fruits of technological change in the form of increasing rates of productivity and economic growth. New technology firms listed at the stock markets and enjoyed enormous jumps in their values, which provided them with splendid amounts of working capital to THE NEW ECONOMY AND BEYOND 7 drive their projects forward. When the equities of new economy companies started to tumble in 2000, the expectation that the achievements of the forerunners would more or less automatically trickle down to the latecomers was gravely disappointed. As a result, catching-up turned into a race-to-the-bottom where the free fall of the forerunner economies brought the latecomers along. All of a sudden, the new economy was like the emperor with no clothes and it was precisely this crisis that seemed to demonstrate the nakedness of this technologybased business model. Even though the share of ICT investment in overall private investment was still comparatively small at the height of the ICT boom, the speed of those investments was extraordinary. In only a few years, ICT investment turned into one of the central investment categories, reaching close to 30 percent of nonresidential gross fixed capital formation in the US in 2001. Even the ICT latecomers of the olive belt could mark increases of the ICT share in GDP of 5 to 10 percent. The build-up of this capital stock was spurred by drastic price reductions for computers and chips and increases in speed and range of user opportunities. Driven by highly positive expectations about the future profitability of these technologies, the equity prices first of “new technology firms” and then also of “old technology firms” experienced—even in historical perspective—an enormous rise, which was not covered by an increase in actual profits.13 This discrepancy, expressed in a stellar rise of P/E values, was the sharp point that burst the bubble in 2000.14 This familiar business cycle story has had some peculiar twists. First, the capital accumulation process in respect to income in these new lead sectors was not profit-driven, rather a strongly wealth-driven activity. Due to the enormous rise in equities and overall asset inflation, companies used the increase in their capitalization value as collateral, as a tool to finance further investment and acquisition. The disconnection between actual and future profits and current fictional values produced the now infamous notion of irrational exuberance, a phrase that captures the systemic nature of the restricted rationality of the economic actors quite nicely. Second, the hype over the revolutionary effects of ICT went hand in hand with an enormous increase in income inequality. It was not just that the highly demanded knowledge worker got higher premiums for his or her skills. It was mainly the upper management levels that intro duced winner take all-structures into the enumeration frame, a sign of the gilded income pyramid of a past which seemed to be gone forever (Phillips 2002). The other version of the tale of the new economy follows a slightly different narrative by arguing that the burst of the bubble should not be understood as one more example of the regular crises so familiar in the history of capitalist growth but as a specific type of a crisis, which points to fundamental inconsistencies of the regime of accumulation and mode of regulation. Crises can be seen as decisive moments, like episodes in a heart patient’s recovery. 8 KURT HÜBNER Economic systems are not patients, yet crises play a similar role. This is particularly true for capitalist money economies that can intrinsically feature spans of long-term stability but reproduce and transform themselves through periodic crises. Crises act as corrective mechanisms, which punish the economy for overshooting and underachieving the functional requirements of a profitdriven system of markets and so prepare the ground for more stable periods of capital accumulation. Such corrections are not free lunches. On the contrary, economic crises come with enormous costs that are distributed highly unevenly between different classes of economic actors. Not all economic crises are born equal and one should add, not all crises are driven by the same factors. The distinction between two types of crises can help to understand the crisis of the new economy in a more comprehensive way. From a functionalist perspective, small crises are defining elements of the regular business cycle (Juglar) that develop in regular seven to nine year intervals. Although such crises drive marginal producers out of business and by this generate unemployment as well as reduce tax incomes of the fiscal state, they do not necessarily induce fundamental changes in the economic behaviors of actors or in the economic/political structures of the overall system. Instead, they are built-in automatic correctors that bring derailed economies back on their established trajectories. In contrast, great crises not only indicate grave problems in the process of capital accumulation and profitability, they also hint at fundamental structural and institutional restrictions of the once-established institutional setting, which can only be overcome by deep-seated changes. In other words, great crises have a trajectory-changing quality. Such crises have at least a twofold nature. First, there are those elements of great crises where the established institutional settings are exploited and exhausted and no longer provide the adequate incentives and constraints for dynamic accumulation of capital. Crises of this type do not show the same regularity as small crises and even more importantly, only develop over a long period of time, every thirty to forty years or so. The last example of such a crisis was the crisis of Fordism, which occurred in the 1970s and ended the golden age of capitalist accumulation and growth after the Second World War. The second kind of great crisis has a related but different nature. The basic elements of this crisis are not the exploitation of an established regime of accumulation and corresponding mode of regulation but the lack of an adequate mode of regulation for a newly developing regime of accumulation. Unlike the first type of great crisis, this is a comparatively open process where adjustments and changes both in the mode of regulation as well as in the regime of accumulation can occur in an asynchronic way leading to non-complementary developments in both realms. In the best case, this type of crisis leads to a new socio-economic technological paradigm, which is characterized by a complementarity between the mode of regulation and the regime of accumulation. Yet crises of this type do not come with such a guarantee. Great crises carry a trait of “Knightian THE NEW ECONOMY AND BEYOND 9 uncertainty”: actors are aware that they have to act differently according to the changing circumstances but they can neither rely on previous experiences nor are they equipped with a proactive knowledge. The search for adequate practices needs time and involves costs. Figure 1.1 illustrates the complex relationships between technological innovations in form of GPTs, modes of regulation and productivity. Technological paradigms (TPs) induced by GPTs have the potential to generate increases in the growth rates of productivity as well as in the scope of economic activities and by this they can lift given productivity levels of an economy. Such lifts only happen if—and only if—the TP is accompanied by complementary institutions of production, consumption and regulation, indicated by a Social System of Production, Consumption and Regulation (SSPCR). Graphed as an isosceles triangle, A′B′C represents the best of all possible worlds. In contrast, ABC represents an inferior constellation due to its lack of a complementary SSPCR. This imbalance generates a comparatively lower level of productivity than in the best-of-all-world-state. Compared to the forerunner of the new TP, A still may represent a higher level of productivity and thus be an improvement in social welfare but one that is not living up to its potential. As simple as it is, Figure 1.1 allows for some sketchy generalizations from the short history of the new economy. First, the concrete economic potential of ICTs was at no point clear for the actors. As discussed above, innovations are literally innovations because nobody knows exactly what the technological, social and economic implications of technological breakthroughs will be. Once inventions turn into innovations, their economic quality is characterized by a high degree of uncertainty. Economic actors are in a steady research and trial process that leads them well into the zone of risk and beyond. This zone of uncertainty is a general feature of economic activities that reach into the future. In so far as innovations hold promise of future profits plus technological rents, some actors will enter this zone of uncertainty and induce imitators to follow. What started with a nucleus of actors willing to handle the challenges of uncertainty turns into a growth motor that attracts more actors from a variety of sectors. Second, the economic exploitation of innovations is a time-consuming process. Their potential is not only unknowable in the present, but a number of costly trials must occur, many of which will fail. Given the trait of uncertainty, the observation holds that in the trial and error phase new actors dominate the landscape. It comes as no surprise that out of the sample of companies that ranked highest on the NASDAQ index during the 1990s the majority of players only came into existence after 1970. The economic success of first mover companies does not automatically translate into a macroeconomic success. The debate of the so-called Solow Puzzle has highlighted the time gap between the launch of innovations and its showing up in statistical indicators like labor productivity (see Erber and Hagemann and Ziegler in this volume). As David (1999) and Crafts (2002) argued vehemently, GPTs like the electric dynamo and ICTs need time to properly adapt. Not only do companies need the 10 KURT HÜBNER time to develop appropriate business models, but new technologies also need infrastructural support in the broadest sense. Seen in this light, ICTs have not behaved differently than previous GPTs that likewise needed long periods of social learning to deliver the expected productivity rates. GPTs are best understood as moving targets, which develop into a range of new processes and products. Unlike Gordon (2003) who strongly argued that ICTs cannot live up to the “great inventions” of the past, argue that ICTs have not yet used up all of their potential. If David (1999) is right in that it took nearly half a century before the US economy understood the concept of the electric motor, ICTs will demonstrate their productivity and growth effects only in the next decades.15 Even though time is an essential analytical category in understanding the delay between technological innovation and economic effects, it is only the social use of time that is of importance. Moving anywhere close to ABC in the graph requires a process of social learning as well as institution building. Third, innovations can clash with given sets of institutions. The beginning of the twenty-first century has witnessed the turn from the first to the second element of a great crisis in the world of capitalist market economies. Based on the technological breakthroughs of the 1960s and 1970s, these economies have experienced strong processes of capital deepening since the second half of the 1980s due to the changing structures of investments in favor of efficiency improving capital goods, a process that accelerated in particular between 1995 and 2000. However, the promises of the technological breakthroughs were not realized immediately, at least not on the macroeconomic level. It was only in the second half of the 1990s that significant increases in productivity and stronger economic growth could be observed and this was only Figure 1.1 Technological innovation and regulation. THE NEW ECONOMY AND BEYOND 11 Gordon, R.J. (2003) “Why Was Europe Left at the Station When America’s Productivity Locomotive Departed?,” Paper Presented at the DEMPATEM Conference, Seville, October 18. ——(2004) “Five Puzzles in the Behavior of Productivity, Investment, and Innovation.” Available online at http://faculty-web.at.northwestern.edu/economics/ gordon/ FivePuzzles.pdf (May 11, 2004). Henwood, D. (2003) After the New Economy, New York: The New Press. Jovanovic, B. and Rousseau, P.L. (2003) “General Purpose Technologies,” unpublished manuscript. Available online at http:www.econ.nyu.edu/user/jovanovi/ GPT.pdf (May 13, 2004). Krugman, P. (1991) Geography and Trade, Cambridge: MIT Press. ——(1999) “The Role of Geography in Development,” International Regional Science Review 22(2): 142–161. Linz, S. and Eckert, G. (2002) “Zur Einführung Hedonischer Methoden in die Preisstatistik,” Wirtschaft und Statistik 10:857–863. Lipsey, R.G. (2000) “New Growth Theories and Economic Policy for the Knowledge Economy,” in Rubenson, K. and Schuetze, H.G. (eds) Transition to the Knowledge Society: Policies and Strategies for Individual Participation and Learning, Vancouver: Institute for European Studies, University of British Columbia. Lundval, B.A. (1998) “Why Study National Systems and National Styles of Innovation,” Technology Analysis & Strategic Management 19(4): 407–421. Mann, C.L. (2003) “Globalization of It Services and White Collar Jobs: The Next Wave of Productivity Growth,” Policy Brief 03/11, Washington, DC: Institute for International Economics. Meister, C. and Verspagen, B. (2004) “European Productivity Gaps: Is R&D the Solution?,” Merit-Infonomics Research Memorandum Series, 005. Mokyr, J. (2003) “Long-Term Economic Growth and the History of Technology,” Draft Section for the Handbook of Economic Growth, edited by Philippe Aghion and Steven Durlauf. Available online at http://www.faculty.econ.northwestern. edu/ faculty/mokyr/Durlauf.PDF (May 15, 2004). Morgan, K. (2004) ‘The Exaggerated Death of Geography: Learning, Proximity and Territorial Innovation Systems,” Journal of Economic Geography 4(1): 3–21. Nightingale, P. (2003) “If Nelson and Winter Are Only Half Right About Tacit Knowledge, Which Half? An Aearlan Critique of Codification,” Industrial Corporate Change 12(2): 149–182. OECD (2003a) The Sources of Economic Growth in the OECD, Paris: OECD. ——(2003b) ICT and Economic Growth: Evidence from OECD Countries, Industries and Firms, Paris: OECD. ——(2004) The Economic Impact of ICT: Measurement, Evidence and Implications, Paris: OECD. Ofek, E. and Richardson, M. (2002) “The Valuation and Market Rationality of Interest Stock Prices,” Oxford Review of Economic Policy 18(3): 265–287. Oliner, S. and Sichel, D. (2002) “The Resurgence of Growth in the 1990s: Is Information Technology the Story?,” Journal of Economic Perspectives 14(4): 3–22. Patel, P. and Pavitt, K. (1991) “Large Firms in the Production of World’s Technology: An Important Case of ‘Non-Globalisation,’” Journal of International Business Studies 22(2): 1–21. 18 KURT HÜBNER Phillips, K. (2002) Wealth and Democracy: A Political History of the American Rich, New York: Broadway Books. Polanyi, M. (1966) The Tacit Dimension, Garden City: Doubleday. Reuveny, R. and Thompson, W.R. (2001) “Leading Sectors, Lead Economies, and Economic Growth,” Review of International Economy 8(4): 689–719. Rodriguez-Pose, A. (1999) “Innovation Prone and Innovation Averse Societies: Economic Performance in Europe,” Growth and Change 30(1): 75–105. Schettkat, R. (2001) “Sind Arbeitsmarktintegrität die Ursache der Wirtschaftsschwäche in Deutschland? Der niederländische und deutsche Sozialstaat im Vergleich,” WSI Mitteilungen 11:674–684. Shiller, R.J. (2000) Irrational Exuberance, Princeton: Princeton University Press. Stockhammer, E. (2004) “The Rise of European Unemployment: A Synopsis,” Working Paper No. 76, Amherst: Political Economy Research Institute. United Nations Commission on Trade and Development (UNCTAD) (2003) “E-Commerce and Development,” Report 2003, Geneva: UNCTAD. van Ark, B., Inklaar, R. and McGuckin, R. (2003) “The Employment Effects of the ‘New Economy’: A Comparison of the European Union and the United States,” National Institute Economic Review 86–98. Verspagen, B. and Schoenmakers, W. (2004) “The Spatial Dimension of Patenting by Multinational Firms in Europe,” Journal of Economic Geography 4(1): 23–42. Wheale, P.R. and Amin, L.H. (2003) “Bursting the Dot.Com ‘Bubble’: A Case Study in Investment Behaviour,” Technology Analysis and Strategic Management 15(1):117–136. Zeng, J. (2004) “Reexamining the Interactions between Innovation and Capital Accumulation,” Journal of Macroeconomics (forthcoming). THE NEW ECONOMY AND BEYOND 19 20 2 The new economy in a growth crisis* Georg Erber and Harald Hagemann Introduction Until the end of 2000 there seemed to be a broad consensus in the economy that the increasingly intensive use of modern information and communication technologies (ICTs) had brought about a sustainable acceleration of growth and productivity. This was certainly the case in the US, which was the undisputed leader of this development. The economic upswing that the US had been experiencing since the beginning of the 1990s differed significantly from previous ones. Thus, at the Lisbon Summit in March 2000, following the US example, EU leaders decided to set their target at a 3 percent minimum annual real GNP growth rate over the next decade. Since the beginning of 2001, however, a more sober view has come up with regard to such development prognoses.1 Currently it seems quite likely that Germany will end the first half of this decade with an average annual GDP growth rate of 1 percent, well below the already low average growth rate of 1.56 percent for the 1990s. Hopes for the dawning of a continuous period of noncyclical economic growth based on new technologies—a “golden age” of the information society (see Schwartz et al. 1999)—have obviously vanished. However, issues remain to be clarified regarding the changes taking place in long-term economic growth and the new types of cyclical effects the ICTs may have on the economy as a whole. Thus, there is an urgent need for research that investigates these questions and the explanations postulated so far. This will produce a clearer picture of the growth and cyclical effects of ICTs on the economy. Future research should especially focus on changes in endogenous growth cycles that are rooted in the new technologies. The ICT industry’s current crisis The European Information Technology Observatory (EITO) recently issued a forecast for the market development of ICTs (EITO, Update 2002). According to this report, EITO expects 5.7 percent growth of the ICT markets worldwide in 2004, while there was a 4.5 percent growth in 2003 (see Table 2.1). For 2000, however, the annual growth rate was still 12.7 percent. Both the US, which led the new-economy boom, and Japan are currently performing better than Europe. It is expected that these growth differences will continue in the coming year. However, in the US, with regard to ICT expenditure growth, ICT spending is falling behind the rest of the world including Japan and Europe. Compared with the forecasts published in spring 2002, the latest data published by EITO for the IT-industry have been significantly adjusted year after year to meet negative trends. This reflects the delay of a significant recovery in the OECDcountries. After the bust of the UMTS-bubble in the telecommunication industry, Europe experienced a particularly rapid decline from doubledigit growth. However, the slow growth of the IT-industries is even more pronounced. To some extent, this might be a statistical artifact because no revised data for the telecommunication industries have been published yet. The data currently available also enable us to predict a more severe growth slowdown for information technologies than for communication technologies. The latter’s significantly faster growth until 2000 can be attributed to the massive economic boom in the area of mobile phones in Europe, the US and Japan. In terms of world market volumes, the two areas possessed similar market Table 2.1 Annual growth rates of the information and communication technologies,2000–2004 Sources: EITO (2001), EITO Update (October 2002) include only new estimates for IT in 2003 and 2004 plus own calculations. 22 GEORG ERBER AND HARALD HAGEMANN shares in 2001:1.133 billion euro for information technologies and 1.159 billion euro for communication technologies (for 2004 the numbers are 1.060 billion euro for information technologies and 1.318 billion euro for communication technologies). As in 2002 and 2003, the ratio shifted away from information technologies and in favor of communication technologies. With regard to the various regions’ world market shares, the US is the unchallenged world leader in information technology; in 2001 its market share was 42.5 percent. Europe follows at a distant second with 29.4 percent. In communication technologies, however, Europe (29.1 percent) is ahead of the US (25.9 percent). The other countries are just ahead of Europe with a 33.1 percent market share. It is likely that the differing focal points between Europe and the US revealed by these data will continue to exist in the future (see Table 2.2). In particular, the rapid upgrading of mobile phone networks in Europe to the UMTS standard can lead to large differences in market structures between Europe and the US. In light of recent developments, it remains highly uncertain to what extent the investment risks associated with this development will result in a sustainable competitive advantage for Europe. At the same time, numerous jobs in German ICT industries are moving away from the production of hardware—and increasingly software toward ICT services. This confirms observations of a shifting trend in world market shares. This also demonstrates the increasing human capital intensification of the information society itself: a decrease in the demand for labor and resources required for the direct production of ICT equipment goods. According to the German Federal Statistical Office, in 2001 there were 83,000 jobs in communication hardware (10.1 percent of the total ICT volume) and 108, 000 in IT hardware (13.2 percent of the total ICT volume). The large majority of jobs were in the areas of telecommunication services (247,000 or 30.1 percent of the total ICT volume) and software and IT services (382,000 or 46.6 percent of the total ICT volume).2 However, the slump in the ICT industry will significantly reduce the employment in Germany: in 2002, 28,800 jobs were lost in the ICT- industries. In 2002, the number of jobs in ICT-industries rose by 10.1 percent to 820,000. However, by the end of 2002, this number had declined to 791,000. In 2002, Deutsche Telekom, which accounts for 90 percent of all revenues and is still Germany’s biggest communication service provider, announced to cut down its employment by about 50,000 jobs until 2005. Since 2000, to these developments, mobile phone manufacturers (Nokia, Ericsson, Siemens, Alcatel) considered relocation from Europe to Asia, in particular to the PR of China, which is expected to be one of the world’s largest growth areas over the next couple of years. This relocation of production sites from Europe to China led to substantial job losses in Europe since 2001. Even if the mobile phone market in Europe recovers, this will not change companies’ strategic decisions. Thus, the jobs lost during the recession will not be replaced in the future. THE NEW ECONOMY IN A GROWTH CRISIS 23 Table 2.2 Forecasts for 2004 on the world market for information and communication technology Sources: EITO (2001) and EITO Update (October 2002) plus own calculations. Note a including Eastern Europe. Similarly, in 2002 the expectations of a solid recovery in the global semiconductor industry were erased. However, there were expectations for a strong recovery in 2003,3 with annual growth rates of 19.8 percent and 21.7 percent in 2004. However, this short-term boom will settle down to a long-term annual growth of about 8 to 10 percent. Moore’s Law According to many assessments, an important determinant of the growth effects induced by ICTs is the long-term constant increase in computer processing power. Over the last decades, this indicator has doubled approximately every 18 months.4 The following analysis applies Moore’s Law—in a slightly modified form (see Figure 2.1)—in combination with the data published by Kurzweil (2000). A logarithmic linear regression of the relationship between computer performance—measured in MIPS (million instructions per second)—and the costs required for this level of performance in US dollars at 1998 prices produces a value of approximately 31 percent annually for the period of almost one century. Moore’s Law would call for 58 percent. Thus, the fall in economic costs is not as extreme as the increase in physical processing performance. If Moore’s Law were to remain valid for the coming decades, computer performance would increase approximately one-hundred-fold by 2010 and more than ten-thousand-fold (in exact terms, by a factor of 10.231) by 2020 compared to current computer processing power, for example, that of a Pentium IV processor. Even in the second half of the 1990s, the US trend in cost reductions accelerated rapidly: it approximately doubled. Thus, this shows the rapid increase in future technological potential in information processing. However, the fundamental question with regard to economic effects still remains to be 24 GEORG ERBER AND HARALD HAGEMANN answered: is the supply side of a continual increase in information processing performance adequate to induce a sustainable rise in growth and productivity? ICT-induced acceleration of economic growth in the 1990s Since the middle of the 1990s, the question of what ICTs have contributed to macroeconomic expansion and productivity growth has been the subject of heated debate among leading economists in the USA. In his presidential address to the 113th meeting of the American Economic Association, Dale W.Jorgenson Figure 2.1 Moore’s Law. THE NEW ECONOMY IN A GROWTH CRISIS 25 summed up the current state of research on this question (see Jorgensen 2001). First among the findings he presented is the idea that the dramatic acceleration of macroeconomic expansion and productivity growth in the US can be attributed mainly to the new ICTs. Second, as shown in Table 2.3, the long-term trend of a radical decline in computer prices has accelerated significantly since 1995. This has been accompanied by a dramatic increase in computer performance: from a decline of 15.77 percent on a yearly average in the first half of the 1990s to 32.09 percent in the second half. According to Jorgenson (2001), this can be traced back to a shift from a three-year to a two-year product cycle in the computer semiconductor industry. The motto “faster, better, cheaper” has thus become the trademark of the new economy. Three channels of impact The radical decline in prices5 and the simultaneous improvement of performance have led to a major increase of investment in computers, software and telecommunication equipment, as well as a rising demand for information services. In 1999 alone, US businesses invested as much money in ICTs as they did in the entire decade of the 1970s. In 2000, this investment dynamic was even more profound and the result thereof was a dramatic increase in businesses’ ICT- capital intensity. This has also played a central role in increased labor productivity since 1995. In this regard, three channels of ICTs’ impact on the economy can be distinguished: •The first channel of impact is direct and consists of the effects generated by rapid technical progress itself within the ICT-capital goods producing sector. Oliner and Sichel (2000) state that the growth of total factor productivity6 in the US computer sector increased from 11.3 percent in the first half of the 1990s to 16.6 percent in the second half. Meanwhile total factor productivity growth of the semiconductor sector even doubled, rising from 22.3 percent to 45 percent in the same period. This demonstrates the outstanding position of the semiconductor sector among the ICT sectors with regard to productivity trends. Despite the relatively small dimensions of the ICT sector even in the US, the sector only made a total contribution to value added activities of approximately 8.3 percent in 2000—it boosted macroeconomic development and productivity growth significantly. The previously mentioned studies produced similar results when calculating this contribution for the US: an increase in labor productivity of 0.2–0.3 percent. The main explanations for this first channel of impact are the positive effects which are generated by increasing returns to scale that characterize the production of ICT goods. • The most important studies7 state that a further 0.3–0.5 percent increase in macroeconomic labor productivity in the latter half of the 1990s did result from the increasing accumulation and use of ICT goods and services (see 26 GEORG ERBER AND HARALD HAGEMANN Erber et al. 2001:217). These effects of capital intensification upon macroeconomic labor productivity constitute the second channel of impact of ICTs on economic and productivity growth of individual national economies. • ICTs’ positive spillover effects comprise the third channel of impact: they lead to “disembodied” increases in efficiency—that is, independent of investment—in the sense of “learning by doing” in the end-user sectors or accelerate product innovations in other areas of the economy (indirectly induced efficiency increases). The broad use of the rapidly growing potential for information processing and communication stimulates an ICT-specific innovation process throughout the entire economy. These indirect effects are especially emphasized by the proponents of the new economy. They are, however, inherently difficult to measure empirically. Additionally, there are numerous methodological problems regarding their adequate measurement. Thus, the calculations of indirect effects that have been presented in the literature so far are highly contested and constitute the main reason for the divergent results of different studies. Only when the last two channels of impact produce major macroeconomic productivity effects can a sustainable surge in productivity and growth be expected. THE NEW ECONOMY IN A GROWTH CRISIS 27 The current growth crisis in the area of ICTs can thus be seen as the result of an increasing mismatch between purely technological potentials on the one hand and the complementary human resources required to make use of them on the other. At present, the central problem lies in society’s inability to apply available technological potentials to achieve sustainable improvements in social welfare, as product-skill complemen tarities and capital-skill complementarities cannot keep pace with technological development. Thus, now more than ever, resources must be diverted into education and training in order to strengthen forces of endogenous economic growth (see Murphy et al. 1998; Romer 2000). However, the supply side of the state educational system as well as the general and vocational educational systems, lack the necessary flexibility to quickly react to these labor market demands with an expansion of supply. With regard to the skill shortage, the narrow wage spread also provides too little incentives for individuals to expend efforts to learn on their own or obtain training (see Kohnz and Erber 2000). This lies at the root of the muchlamented backlog of ICT applications, which can easily develop into a costintensive bad investment when hardware systems are purchased without planning for the necessary human resources needed to operate them. To sum up, ICTs can only stimulate long-term economic growth when major mismatches in product-skill and capital-skill complementarities are avoided; that is, when structural imbalances do not occur. If structural rigidities are not eliminated quickly, this raises the specter of falling back to a lower growth path. Cyclical effects of ICTs In recent years the discussion on the new economy has focused on the question of the long-term acceleration of growth in the US. Initially, this seemed to make sense in light of the fact that the US had not experienced any significant macroeconomic slumps in the course of the 1990s. However, the current recession in many OECD countries raises the question whether there is a need for more careful analysis, not only of the longterm effects of ICTs but also of their cyclical macroeconomic effects. On the basis of “real business cycle theories” current theory has found that, along with monetary factors, exogenous productivity shocks are the main source of short-term fluctuations (see Lucke 1998). As a result of adjustment processes, the economy transforms these shocks into businesscycle movements. These approaches supplant older theories (Schumpeter 1939), which were based on the overlapping of various business cycles: • an inventory cycle (Kitchin), • an investment in equipment goods cycle (Juglar), • an investment in construction goods cycle (Kuznets), • the long wave cycles caused by basic technological innovations (Kondratieff). 34 GEORG ERBER AND HARALD HAGEMANN For a long time, the theoretical basis of general business cycle discussions was the pattern that is produced when the fixed periodicities attributed to each of these individual cycles are superimposed on one another. However, concurrent with the neoclassical emphasis on the efficiency of an economy with flexible markets, this endogenous view of business trends lost popularity among macroeconomists. In the same way that endogenous growth theory generally did not correspond to the notions of the economic fraternity of this period—whose interest was reawakened only by the work of Romer, Barro, Grossman, Helpman, Aghion and Howitt—the question of an endogenous cyclical trend based on the new technologies has been largely neglected in the current economic debate. In light of the current business trend, this could prove to be a serious flaw and could be used to gain a better understanding of the nature of the modern information society’s cyclical growth. As several recent studies on this topic have shown, growth dynamics changes considerably when a non-neutral, that is, an investment-specific technological change is taken as a basis (see Greenwood et al. 1997). In the framework of a model with endogenous growth, the specific annual investment vintages, together with the structure of the capital stock, can lead to the emergence of endogenous cycles. Oscillations in investment in ICT goods thus have the effect of generating cycles, which in turn trigger an “echo effect” (see Boucekkine et al. 1997) due to replacement needs resulting from the previous investment cycle. Regarding the current development in the world economy, approaches such as these have interesting implications. First, because of the ICT investment boom in the second half of the 1990s, the current slump can be interpreted as a new type of ICT investment cycle. Along with the disproportionately large economies of scale (when compared to the previous year), the non-recurring replacement pressure exerted by the threat of the Y2K bug and the strong incentive it provided to make early ICT investments could have triggered this boom. The latter would explain the drastic reduction in ICT investments after the turn of the millennium, as market saturation has occurred since then, at least temporarily. However, if one postulates a boom in ICT investment goods and uses the period of depreciation as a base, then—using a vintage capital model as a theoretical approach—one would be able to calculate a strong echo effect in replacement needs. Likewise, the question of a changed pattern of market fluctuations still requires further research. Thus, at its core, the classical (Juglar) business cycle is based on a reinvestment need or echo effect. With the sharply increasing share of ICT-capital goods in total equipment investments or in the physical capital stock thereof, there is an increase in the share of investment that implies a need for reinvestment after three to five years—instead of after seven to eleven years— due to their shorter technological and economic lifetime. The echo effect of ICT- capital goods exhibits a temporal period due to rapid technological obsolescence that corresponds most closely to the inventory cycle a la Kitchin and less to the THE NEW ECONOMY IN A GROWTH CRISIS 35 classical time frame of the Juglar cycle. This dramatic increase in the depreciation requirements for ICT-capital goods counteracts the tendency toward capital-saving technical progress. This is connected to Moore’s Law and therefore constitutes one of the essential arguments in the explanation of Solow’s productivity paradox. To empirically test the explanation of the new economy’s current crisis that has been put forward here only in summary form would require sufficiently reliable data, which currently are nowhere available. Nevertheless, a number of extremely interesting research perspectives emerge out of this study. The question of which endogenous cyclical patterns a market economy exhibits ultimately deserves as much attention as the analysis of long-term development which has been the focus of growth theory for a long time. Differences in growth performance between Germany and the US Unlike in the US, Germany’s economy barely shows any sign of the new economy in the macroeconomic data on growth, productivity and employment trends in the 1990s. The contribution of the ICT-capital stock to overall growth of GDP was just 0.05 percent in the first half of the 1990s and doubled to only 0.11 percent in the second half. This has led to the media exaggeratedly portraying the US’s return to a “golden age” of growth while Germany’s economy plods along in a “melancholy state.” Even a high growth of the ICT-capital stock of 7.35 percent and 11.65 percent in the first and second half of the 1990s (see Table 2.4) did not have a similarly strong impact on aggregate growth as it did in the US.13 To some extent, this is due to the fact that Germany has not built up the same amount of ICT-capital stock in the previous decade. Therefore, its overall share of total capital just rose from 2.5 percent to 5 percent in the 1990s— much below the level in the US. A further reason for this difference in performance between the US and Germany is due to a dramatic decline in the productivity of ICT-capital investment in Germany. German companies seem to be constrained by organizational rigidities and value chains that limit the effective use of ICTs. The capital-skill complementarity between ICT-equipment and the workforce also hampers an efficient use of ICT compared to the United States. This has triggered a debate on increasing immigration of high-skilled foreigners via greencards to Germany.14 Furthermore, a smaller sector of producers of ICT equipment, software and services diminishes the impact of ICT compared to the US. This is particularly the case as the embodiments of ICT into traditional manufacturing products like machinery and electrical equipment or automobiles are not well calculated in the National Accounting Systems which favor the highly visible information technology investments in office computers but less so the embedded ICT in traditional products. Looking at input-output-table data for Germany, it becomes clear that nominal growth in gross production values in the 36 GEORG ERBER AND HARALD HAGEMANN computer industry is mainly due to a tripling of imports and a stagnant domestic production level. Thus, growth of this industry took place elsewhere in the globalized economy, such as the US, Japanese or Asian NICs and in European production locations such as Ireland. The fact that the long economic upswing in the US led to both an acceleration of productivity growth and a sharp increase in employment is particularly striking. The result was an unemployment rate of 4 percent—distinctly below the level (6 percent) long considered as “natural” or the “non-accelerating inflation rate of unemployment” (NAIRU). The surge in the labor productivity growth rate caused such a striking reduction in the inflation pressures that American monetary policy was able to follow an expansionary course without endangering its objective of monetary stability. In contrast, after the short boom of 1990/91 that was the immediate result of reunification, Germany’s economy suffered a deep recession in 1992/93 and recovered only very slowly thereafter. However, until today, economic growth rates remain significantly below that of the other EU member states. Additionally, the difference between labor productivity growth per employee and per working hour denotes that a decreasing or stagnant employment pattern was highly unfavorable to facilitate a high productivity growth rate. The restructuring of the German economy, in particular of the East German part, could not be accomplished by a straightforward increase in labor productivity. Incapability in utilizing the costly ICT-capital stock by flexible working hours, with the exception of some large scale plants like semiconductor producers or a tiny group of start-up companies, further decreased the efficiency by lower operating hours for ICT-capital stock compared to other countries. A waste of economic resources in building up overcapacities in the residential real estate market and weak privatized former state companies, which did not manage a transition to highly effective and international competitive enterprises, slowed down the economic transition of the East German states. Subsidies by the German federal and state governments which made the payment of a high level of wages in the East possible—e.g. the unit labor costs are still about 20 percent higher in the Eastern states than in the Western states with no sign of closing this gap in the near future—have decreased investment incentives in many areas of the East German economy since then. Investments could only be attracted at high social costs, that is by subsidizing investments to compensate for the higher wage costs associated with the East German productivity-wage gap. Many of these investment subsidy programs were cut down in the late 1990s because of their distorting effects and increasing budget problems. As a result, the investment boom of the first half of the 1990s came to an end. THE NEW ECONOMY IN A GROWTH CRISIS 37 Table 2.4 Decomposition of the German GDP 1992–2000 Notes a In constant prices (1995=100). Deviations between aggregate and the sum of components is due to rounding errors. b Labor income share=ratio of labor income to GDP. c Capital stock including residential real estate. 38 GEORG ERBER AND HARALD HAGEMANN Furthermore, the financing of the above productivity wage-level in East Germany was and still is accompanied by an increase of the social security payments for West German workers. Rising wage costs in Germany caused by high social transfers to the East put extra pressure on the West German labor market. This increased the incentive for German and foreign investors to relocate the demand for low-skilled labor abroad to other locations like Asia, in particular to the PR of China, or more nearby to the Central and Eastern European states, such as the Czech Republic, Hungary and Poland, where wages for low-skilled labor are just a fraction of those in West Germany. This lowered the investment and employment incentives in Germany in general and also contributed to the unbundling of the national value chains and changed the allocation structure from an efficiency-oriented toward a more social-justice based one. A short surge of foreign direct investment into Germany associated with the deregulation of the telecommunication industry and the auctioning of the UMTS licenses in 1998 until 2000 could not be maintained after the high expectations linked to these developments ended up being overoptimistic. Higher expected returns on investment in the US, in particular in the new economy industries, encouraged high foreign direct investments to turn to the US. Megamergers like those of Deutsche Bank/Bankers Trust, Daimler-Benz/Chrysler, Deutsche Telekom/Voicestream and numerous investments into companies of the NASDAQ led to a significant shift of investment flows. Although the annual labor productivity growth rate of Germany remained above that of the US until 1998, it would be false to assume that the stronger economic dynamics in Germany had been the result of higher speed of diffusion of new technologies. For decades, the catching-up process vis-à-vis the US was characterized by efforts to close the productivity gap accompanied by very limited long-term employment gains (see Erber et al. 1998). However, in the 1990s this resulted mainly from intensified employment reductions in lessproductive sectors and the significantly lower capacity of the German labor market to absorb workers with lower qualifications. As a result, Germany’s strong lead was gradually reduced throughout the 1990s until it disappeared altogether in 1998. Ever since, Germany has not been able to catch up with the US. The macroeconomic consequences of German reunification provide significant explanations for this development. The acute need for infrastructure investment in the new Bundeslän-der of former East Germany limited the possibilities for investment in new technologies compared to other OECD countries. These developments were further intensified by a flawed tax policy, which led to excessive expansion of the construction industry and an associated massive misallocation of capital. Additionally, German monetary policy was more restrictive compared to that of the US, which prevented stronger demand growth and therefore higher productivity growth that— in accordance with Verdoorn’s Law—goes hand in hand with economic growth. THE NEW ECONOMY IN A GROWTH CRISIS 39 Thus, Germany’s hopes of duplicating the US development of the late 1990s are thoroughly understandable. However, the cyclical downswing that began as early as the end of 2000, has silenced naive hopes that with the new economy, the US has seen the end of the business cycle or that at least any further recessions will be shorter or milder. After re-unification, German investment activities had a rather different focus than in the US. This was the case because of the way in which the integration of East Germany was implemented caused inefficient allocation of resources. Additionally, it was partly due to wrong subsidies in the construction sector, a predisposition with investment into traditional industries, such as steel plants and basic chemical plants and significant waste due to corruption in the distribution process of these subsidies as well as wages that were higher than productivity. Furthermore, many infrastructure projects of local governments encouraged gold plating of infrastructure in areas where it was not needed, or where it only had the capability to create short-term employment effects. Because these misallocations continued over the entire decade, the traditional pattern of high productivity-driven growth was lowered in Germany. As incentives for workers and investors were relocated away from the restructuring of the economy toward new high growth areas, investments did not generate a sufficiently high return. The newly created group of small- and medium-sized companies in East Germany also suffered from undercapitalization and a lack of experience to operate and sell their products in West German or even international markets, thus not creating a significant customer base. When Germany ran into a slowdown or mild recession, the West German economy caught a cold while the East German economy easily got pneumonia. This still makes the German economy vulnerable for the predictable future. It emerged during the 1990s as a laggard in growth performance among EU countries and it could become a sick man of Europe if major corrections continue to emphasize on social redistribution rather than efficiency. Summary In summary, with regard to macroeconomic growth and business cycle effects, the ICT research landscape is dotted with “uncharted territories” that render it more difficult to make exact assessments that are quantitatively reliable and to derive predictions for future developments. Further research needs to focus on achieving the following aims. First, it should attempt to close the gap in the data by creating a dependable statistical basis on the specific investment behavior regarding ICT products and the analysis of human capital creation, especially concerning ICT-specific qualifications. Second, it should develop new concepts for the theory of growth cycles that make it possible to specify respective hypotheses and to empirically test a theory of endogenous growth cycles against the background of technology-specific diffusion, by relying, for example, on the basis of the work of Murphy et al. (1998). 40 GEORG ERBER AND HARALD HAGEMANN The initial question of this section, whether the new economy is currently in a crisis, can be answered by stating that on the one hand, the cyclical character of ICTs has become particularly clear in the current phase of the business cycle. As a result, we are indeed experiencing a growth crisis of the new economy. However, even in the midst of the crisis, the economic growth of the ICT markets still remains at 4.0 percent, which is significantly higher than the worldwide average of 2.6 percent. For 2003 and 2004, the comparison of growth forecasts between the US and Germany yields a continuing positive difference of 2.5 percent and 3 percent for the “new” economy in the US against 0.6 percent and 1.0 percent for the “old” economy in Germany (see DIW 2003:2). Consequently, there is a continuous trend toward an information and network society. Notes * We thank Markus Schreyer and Stephan Seiter for valuable comments on an earlier draft. 1 See the results of the current “Six Institutes Forecast” of Germany’s economic research institutes in DIW 2002. 2 See BITKOM’s publication on the Internet on jobs in information technology and communication (in German), 2000. Available online at http://www. BITKOM.org. 3 See 2002–2005 Forecast: SIA Projects Double-Digit Growth for Global Semiconductor Sales, Press Release, Semiconductor Industry Association, San José, November 6, 2002. 4 Moore’s Law refers to “the observation made in 1965 by Gordon Moore, cofounder of Intel, that the number of transistors per square inch on integrated circuits had doubled every year since the integrated circuit was invented. Moore predicted that this trend would continue for the foreseeable future. In subsequent years, the pace slowed down a bit, but data density has doubled approximately every 18 months. This is the current definition of Moore’s Law, which Moore himself has blessed. Most experts, including Moore himself, expect Moore’s Law to hold for at least another two decades” (webopedia 1998). 5 However, the rapid decline of semiconductor prices may be attributable not only to cost reductions based on technological efficiency increases in production but is also a result of increased competition between industry leaders like Intel and new market entrants. Price wars in the semiconductor market may have played a role in the acceleration of the price decrease observed in the second half of the 1990s. Aizcorbe (2002) provides evidence that the mark-up price for microprocessors of Intel has declined significantly from 1993 to 1999 as a market leader in this field. Usually such price wars are not sustainable in the long term and therefore cannot be expected to contribute to future productivity growth. 6 Total factor productivity, along with its share of costs, is the weighted sum of the partial factor productivities, for example labor and capital productivity. To measure the influence of investment in ICT equipment goods, ICT-specific goods are separated from other equipment goods. THE NEW ECONOMY IN A GROWTH CRISIS 41 7 See Council of Economic Advisors 2000 and 2001; Gordon 2000; Jorgenson and Stiroh 2000; Oliner and Sichel 2000; Whelan 2000. 8 See Evans et al. 1998. The modular character of numerous ICT products enables them to be recombined into new products with new characteristics ad infinitum. Thus, depending on what it is equipped with, a PC can be used for a variety of things: playing computer games, taking measurements, producing CAD drawings, word processing, making music, accessing the Internet or making videos. The possible product combinations thus exceed the imagination. 9 For the most recent estimates, see EITO Update 2002. 10 As anecdotal evidence one should realize that on the standard PC bought from a supermarket chain like Aldi in Germany with Windows XP and a couple of application software packages pre-installed one has about 80,000 files stored on the hard disk without any personal data files store on it. If they are not extremely large and need high computing capacities because of complex calculations and high volume of these calculations, text or numeric data files are hardly capable to exceed the needs of ordinary customers at the same speed as in the past. 11 ASPs provide comprehensive services in the remote maintenance of hardware and software, offering services that are highly time-consuming for users such as data backups, software updates, etc. at a reasonable price. Because of the bundling of services at computing centers and the availability of qualified specialists there, the end-user can avoid a process that would otherwise be cost intensive. By offering special software packages at “pay-per-use” fees, companies can tap into new customer segments that would otherwise never purchase the licenses to these products due to the high price. Nevertheless, it remains to be seen whether or not the expectations for ASPs will be fulfilled on the market. At present, the ASP market is still in its infancy. 12 As we know from production theory, a Leontief-type limitational consumption function as the dual form on the consumption side compared to the supply side with its limitational production function just increases the utility of a user for a particular type of equipment as a durable consumption good or its respective services only when the user has the necessary capabilities to unlock its services. 13 Calculations are based on the most recent national data for Germany published by the national statistical office in August 2002 and the capital stock data for the break down into different types of capital stock are based on van Ark et al. (2002). 14 However, the current recession in the ICT and other industries in Germany reversed this tendency so that many of the new migrant ICT-workers were fired first when the companies shed labor during the recession. Bibliography Aizcorbe, A. (2002) Why Are Semiconductor Prices Falling So Fast? Industry Estimates and Implication for Productivity Measurement, Washington, DC: Federal Reserve Board. BITKOM (2001) Europäischer Markt Für Informationstechnik Und Telekommunikation Wächst Um 6.8%, EITO. Available online at http://www.eito.com/ Press-rel/pressrel-4.html (December 4, 2003). 42 GEORG ERBER AND HARALD HAGEMANN Boucekkine, R., Germain, M. and Licandro, O. (1997) “Replacement Echoes in the Vintage Capital Growth Model,” Journal of Economic Theory, 74(2): 333–348. Christensen, C.M. (1997) The Innovator’s Dilemma, Boston: Harvard Business School Press. Council of Economic Advisors (2000) Annual Report, Washington, DC: Council of Economic Advisors. ——(2001) Annual Report, Washington, DC: Council of Economic Advisors. DIW (2002) “Die Lage der Weltwirtschaft und der deutschen Wirtschaft im Herbst 2002,” DIW-Wochenbericht, 43. ——(2003) “Grundlinien der Wirtschaftsentwicklung 2003/2004,” DIW-Wochenbericht, 1–2. Erber, G., Hagemann, H. and Seiter, S. (1998) Zukunftsperspektiven Deutschlands im Internationalen Wettbewerb: Industriepolitische Implikationen der neuen Wachstumstheorie, Heidelberg: Physica. Erber, G., Hagemann, H., Schreyer, M. and Seiter, S. (2001) “Produktivitätswachstum in der ‘New Economy:’ Übergangsphänomen oder Strukturbruch?,” in Heise, A. (ed.) USA—Modellfall der New Economy?, Marburg: Metropolis. European Information Technology Observatory (EITO) (2001) “ICT Skills in Western Europe,” European Information Technology Observatory 2001, Frankfurt: EITO. ——(2002) “Update 2002,” European Information Technology Observatory, Frankfurt: EITO. Evans, G.W., Honkapohja, S. and Romer, P. (1998) “Growth Cycles,” American Economic Review 88(3): 495–515. Gordon, R.J. (2000) “Does the ‘New Economy’ Measure up to the Great Inventions of the Past?,” Journal of Economic Perspectives 14(4): 49–74. Greenwood, J., Hercowitz, Z. and Krusell, P. (1997) “Long-Run Implications of Investment-Specific Technological Change,” American Economic Review 87(3): 342–362. Griliches, Z. (1969) “Capital-Skill Complementarity,” Review of Economics and Statistics 51(4): 465–468. Helpman, E. (1998) General Purpose Technologies and Economic Growth, Cambridge, MA: MIT Press. Jorgenson, D.W. (2001) “Information Technology and the U.S. Economy,” American Economic Review 91(1): 1–31. Jorgenson, D.W. and Stiroh, K. (2000) “Raising the Speed Limit: U.S. Economic Growth in the Information Age,” Brookings Papers on Economic Activity 1: 125–211. Kohnz, S. and Erber, G. (2000) Lohnspreizung und Arbeitslosigkeit: Theoretische Erklärungsansätze und Stand empirischer Forschung, Berlin: DIW. Kurzweil, R. (2000) The Age of Spiritual Machines: When Computers Exceed Human Intelligence, London: Penguin. Lucke, B. (1998) Theorie und Empirie realer Konjunkturzyklen, Heidelberg, Berlin, New York: Springer. Murphy, K.M., Riddell, W.C. and Romer, P.M. (1998) Wages, Skill, and Technology in the United States and Canada, Cambridge, MA: National Bureau of Economic Research. Oliner, S.D. and Sichel, D.E. (2000) “The Resurgence of Growth in the Late 1990s: Is Information Technology the Story?,” Journal of Economic Perspectives 14(4): 3–22. THE NEW ECONOMY IN A GROWTH CRISIS 43 GDP growth rates would start to decline after six years, whereas inflation rates again overtook growth rates after a six-year period. In the last expansion we can only identify seven years due to data limits—a different image is presented. First, growth rates accelerate from year to year, reaching their highest level at the end of the recorded period. Second, inflation rates again decline during the expansion. Given the relatively small growth rates of the last European expansion—in particular compared to the US—we find inflation rates that are initially above GDP growth rates, but then end well below. In that respect, the same novel pattern that has emerged in US macroeconomic time series can also be identified in Europe. However, this Figure 3.2a Growth and inflation, EU15 1960–1973. Figure 3.2b Growth and inflation, EU15 1983–1992. 50 MARTIN ZAGLER pattern is not homogenous across Europe. Whilst the UK follows the US and the general European pattern in the last expansion, Austria, for instance, does not. This is shown in Figures 3.3 and 3.4. In that respect, the new economy has installed itself in some, but not all, economies. The question that remains is what prevents economies from developing a “new economy.” The last stylized phenomenon of the new economy are stock markets. In the medium range, they have been more volatile than they were in the second half of the last century. From December 1997 to March 2000, stock markets roared. The New York Stock Exchange experienced a 26.6 percent increase. This development spread internationally, with London gaining 37 percent, Amsterdam 45.4 percent and Frankfurt even 77.7 percent. Most remarkably, however, was the increase of the technology-prone NASDAQ—it almost tripled its stock market capitalization and exhibited an annual increase of 76.6 percent. From March 2000 onwards, stock prices plummeted, with Amsterdam dropping 6 percent, Frankfurt 13.4 percent, London 9 percent—only New York slightly improved at 1.7 percent. Most startling was the performance of the technology values on the NASDAQ, which almost halved its value, dropping by 42.8 percent in only nine months (Helmenstein and Zagler 2001). The chapter addresses these three stylized facts: the productivity puzzle, the seemingly unexplainable macroeconomic performance and the stock market boom which simultaneously accompanied the expansion of the ICBT sectors. The following section briefly sketches a model that specifically focuses on the product characteristics of the ICBT sector. The other sections will provide an explanation of these empirical phenomena. A simple model Let us assume that firms produce output using labor, physical capital and ICBT capital according to the technology given in equation (1), which we simplify to a Cobb-Douglas specification, (1′) According to Table 3.1 and Table 3.2, the deepening of non-ICBT capital has not been a major issue over the period of interest. We therefore assume that the capital to labor ratio is constant and equal to k, further simplifying technology to, (1″) where Bt collects all exogenous variables, It turns out to be useful to think of this production function (equation (2)) to contain two separate technologies, a INNOVATIONS, ECONOMIC GROWTH AND PRODUCTIVITY 51 traditional technology Zt=BtLt, which uses labor and physical capital and a new technology Xt, which uses ICBT capital.1 Then we can identify the optimal transfer price pt, which the firm would be willing to pay for an additional unit of ICBT capital by its marginal product, (4) Similarly, we can identify the optimal transfer price for the old technology qt, Figure 3.2c Growth and inflation, EU15 1994–2000 (sources: OECD National Accounts, own calculations). Figure 3.3 Growth and inflation, UK 1993–2001. 52 MARTIN ZAGLER (5) As the total factor productivity Bt is exogenous, all revenues from within the old technology go to the factor labor. As one unit of labor produces Zt/Lt=Bt units of the old technology production factor, the wage paid to labor is equal to the transfer price times productivity, wt=Btqt. We find that the ratio of nominal spending for ICBT capital to nominal spending for other physical capital is constant and depends only on a, (6) Whilst this appears appropriate for a given state of ICBT technology, firms will increase their nominal spending for ICBT if new applications arrive. These are valuable on their own but also enhance the value of all existing ICBT applications. We will capture this feature by assuming that a depends positively on the number of available ICBT applications n, a=α(n). Different ICBT applications are certainly no perfect substitutes. In order to further simplify matters, we will assume that the elasticity of output with respect to aggregate ICBT capital is identical to the elasticity of output with respect to a particular ICBT capital good i, (7) Figure 3.4 Growth and inflation, Austria 1993–2001 (sources: OECD National Accounts, own calculations). INNOVATIONS, ECONOMIC GROWTH AND PRODUCTIVITY 53 The optimal transfer price pi,t, which the firm would be willing to pay for an additional unit of a particular ICBT application again equals its marginal product, (4′) Providers of ICBT are innovators, who will be the single provider (directly or via licenses) for ICBT applications demanded by final goods producers (Bresnahan and Trajtenberg 1995). Assuming that these ICBT providers can turn one unit of the final product into one unit of the ICBT application2 xi,t=yi,t, they maximize profits by setting marginal revenues equal to marginal costs αPi,t=1. The mark up turns out to equal 1/α, (8) The mark up declines as the nominal share of the ICBT sector increases. This profit squeeze is due to the fact that competition within the ICBT sector gets fiercer as new applications partially drive out older applications. Given that all ICBT firms will ask an identical price, the demand will be identical for all ICBT firms and equal to, (9) As technology in the ICBT sector (equation (7)) exhibits constant returns to scale, we can identify a price index for ICBT applications pt without referring to purchasing shares. Dividing equation (4′) by equation (4), taking everything to the power of α/(α−1), summing over all i and eliminating quantities from equation (7), we find that the price index for ICBT applications equals, (10) where we have made use of mark-up pricing (equation (7)) to eliminate the summation. The productivity paradox revisited Productivity differs between sectors. First, productivity in the old technology Zt equals output per worker, which equals Zt/Lt=Bt. Note that the productivity of 54 MARTIN ZAGLER the old technology increases continuously at the rate of total factor productivity. However, as the share of the old technology in production (1−α) declines continuously, we should expect a continuous decrease in total factor productivity. In a particular ICBT firm, one unit of the final product produces one unit of an ICBT application, hence productivity equals one by definition, xi,t/yi,t=1. However, the ICBT sector as a whole does exhibit productivity gains due to increases in variety. We obtain a measure of productivity of the ICBT technology straightforward by dividing output as defined in equation (7) by all ICBT inputs that is the sum of all yi,t final product inputs in the ICBT sector. In order to better understand the empirical results, we will pursue a different track. First, note that aggregate nominal spending for ICBT applications ptXt must equal the sum over all expenditures for all particular applications pi,txi,t. Making use of the mark-up equation (8) and substituting final product inputs for ICBT application outputs, after some rearrangement, we obtain, (11) Productivity in the ICBT sector is growing proportionally to the rate of innovation. The reason is that the expenditure based price index for ICBT inputs (equation (10)) corrects for productivity gains at the sectoral level. If, by contrast, the ICBT price index would be simply a weighted average of all particular ICBT input prices pi,t, the aggregate price index would equal 1/α and we would observe no productivity gains in the ICBT sector. This was common practice until the revision of prices by the BEA. After the revision, we should expect the productivity increase to show up in the capital stock, which remains the dominant source of increases in labor productivity according to Tables 3.1 and 3.2. There is evidence that BEA adjustments of quality are incomplete (Pakko 2002). This implies that an even greater share of labor productivity growth should be attributed to ICBT capital. Economic growth Equation (11) allows us to interpret new technology as a product of the inputs into the new technology (xi,t) and productivity growth in the new technology, which depends on the number of ICBT innovations n. Substituting this into the production function (1″), we can rewrite per capita output as a function of raw inputs and technological change, (2′) INNOVATIONS, ECONOMIC GROWTH AND PRODUCTIVITY 55 We can now separate growth in output into an extensive component—the growth in factor inputs—and an intensive component—the change in technology. There is a natural limit to reinvest output into the new technology. It does not make any sense to reinvest all output yi,t into the produc tion process, as nothing would be left for consumption. Noting from equation (4) that the nominal spending share for ICBT capital equals a and eliminating ICBT factor inputs from equation (11), we find that the share of ICBT inputs in production equals α2, (11′) Given that the price for the factor inputs yi,t is equal to the price of output Yt, which are both normalized to unity, this implies that the nominal share of raw inputs is α2, whereas the share for ICBT capital Xt was a. The difference, α(1−α), is the profit share for monopoly suppliers of ICBT inputs, which they use to finance costs of innovation. Substituting (equation (11′)) into labor productivity (equation (2′)), we find, (2″) Ignoring the impact of structural change (changes in α), we find that per capita output growth equals the sum of growth is exogenous technical change Bt and the rate of innovation n. Note once again that we will measure the latter effect if —and only if—we correctly adjust ICBT prices for productivity changes. If the ICBT sector is bigger, then we should be able to devote more resources to the innovation of novel ICBT applications (Zagler 2002). Hence, an emerging new economy may actually exhibit higher growth rates of potential output, thus easing the inflationary pressure from monetary expansions. But there is another important aspect that can explain why, during the last expansion, prices have not been accelerating despite higher growth rates, as there is an intrinsic inertia in ICBT prices. In order to understand this, we have to turn to pricing mechanisms of our simple model. Prices and inflation Final product providers pay implicit transfer prices to the old and new technology. Unit costs will equal the sum of costs for the old technology and for the new technology, divided by output, (12) 56 MARTIN ZAGLER where we have made use of technology (equation (1″)) to express unit costs as a function relative to real sector shares. We can use nominal sector shares (equation (6)) to eliminate all quantities, (12′) which we can separate into a unit cost component c depending solely on α and an implicit expenditure based producer price index Pt, which is a Cobb-Douglas weighted index of the two transfer prices pt and qt. Pressure for changing producer prices will depend on changes in unit costs or equivalently changes in the producer price index Pt,3 (13) We find that increasing productivity-adjusted wages are a first source for increasing pressure to adjust prices upwards. As the size of the new technology increases, this cost-push element of inflation declines. This is the first bias in the inflation rate and already leads to a decline of the inflationary pressure. Over the past decade, this has been exploited by monetary policy in the US. If ICBT transfer prices are simple weighted averages of individual ICBT prices, the ICBT price index would be constant and the story would stop here. However, if individual ICBT prices are corrected downward to adjust for the added benefit for one application from the adoption of another,4 it turns out that inflationary pressure further declines. Evidently, the faster new applications get adopted, that is the faster the rate of innovation, the more inert prices will be. The regional divide In the previous chapter, we assumed that when firms make their pricing decisions, they consider the state of the ICB technologies as given. The result is certainly a good approximation in the short term. In the long term, however, firms can realize cost advantages by adopting new ICBT technologies at a greater extent. In the long term, perfect competition drives unit costs to the price of the final product, which we have normalized to unity, a result which we can obtain by substituting equation (6) into equation (12). But this implies that unit costs cPt as defined in equation (12′) will equal unity, or (12″) INNOVATIONS, ECONOMIC GROWTH AND PRODUCTIVITY 57 This gives us an expression for relative prices, which we can use to identify relative real shares of the old to the new technology, (6′) An increase in the number of innovations will exhibit an unambiguously positive impact on the nominal share of the ICBT sector, α. Hence the first term will always be increasing over time. A sufficient condition for an increasing relative real share of the new technology would be the second term increased over time, which, taking logs and derivatives, will be the case if, and only if, (6″) where is the elasticity of the ICBT sector share with respect to ICBT innovation. A large sector share will typically imply a small elasticity, as there will be no more room for an expansion of the ICBT sector. Hence an excessively large number of ICBT applications will help an economy to overcome this technological threshold. In this respect, only technology prone countries will manage to develop ICBT production techniques, whilst others will stay behind. This explains the regional divide in the use of ICB technologies and the emergence of the new economy in some parts of the world, but not in others. Stock market volatility5 Over the medium term, stock markets have been more volatile over the last few years than they were in the second half of the last century. The adoption of new technologies is one reason for the considerably higher volatility. We will analyze this difference within the model framework by investigating the different reactions of firms predominantly using the old technology (with a close to zero) and firms predominantly using the new technology (with a close to unity) to a sudden small λ-fold increase of intermediate inputs, driven, for instance, by an increase in productivity in the respective sector. With the old technology, an increase in supply, dyt/yt=λ, will not change profits, as perfect competition ensures zero profits. In an efficient stock market, the price of a stock reflects the discounted stream of future profits. Hence stock prices should not react to the productivity shock. In order to sell the additional output, firms will have to cut prices by a proportional amount, dPt/Pt=λ. Unless productivity gains permanently render every unit of labor more efficient, this will lead to losses and eventually a withdrawal of the original increase in supply. With the new technology, an increase in the supply of each ICBT application, dxi,t/xi,t=λ, will force every ICBT provider to cut prices. Given a price elasticity 58 MARTIN ZAGLER of demand equal to 1/(1−α), firms will cut prices by (1−α) λ. As profits in the provision of ICBT applications are proportional to revenues, this will lead to an increase in profits by αλ. Profits will therefore increase less than proportionally, but given a discount rate of r, the price of the company stock will increase more than proportionally if α >r, or if the share of ICBT use in production exceeds the real interest rate. Hence it comes as no surprise that the stock market boom set in only after the ICBT share was exceeding roughly 6 percent. By itself, this can explain the information technology stock market boom. As is evident, an increase in quantity of all ICBT applications will induce a proportional increase in aggregate factor supply (equation (7)), dXt/Xt=dxi,t/xi,t=λ and a decrease in every price will lead to a proportional decline in the price index (equation (10)), dpt/pt=dpi,t/pi,t=−(1−α)λ. An increase in profits increases firms’ incentives to engage in innovation. Free entry ensures that the number of firms in the service sector will drive discounted running profits down to their previous level, which is just enough to finance the cost of innovation. There are three possible ways in which this process comes to an end. First, the initial effect—in our example an increase in productivity—is reversed. Second, the expansion in quantity raises the demand for labor. If the service sector is large enough, this will raise wages, inducing an increase in prices. Not surprisingly, on the doomsday of the information technology stock market boom, warehouse workers at the Internet book seller Amazon were threatening to strike for higher wages. Third, once the first new innovators succeed to enter the product market, they will start to have an impact on aggregate prices and quantities. Aggregate ICBT provision will increase, whilst the ICBT price index (equation (10)) will fall. This implies that the increase in quantity of every ICBT provider falls and the bubble begins to implode. Conclusions The aim of this chapter was to explain several distinguished stylized facts of the new economy. We found that these facts are low rates of total factor productivity despite high rates of innovation in information-, communication- and biotechnologies (ICBT), but high rates of capital deepening due to investment in ICBT, an inflation inertia despite high rates of economic growth and full capacity utilization, regional differences in the emergence of macroeconomic phenomena associated with the new economy and an increasing volatility in stock markets. A simple model of production, which embeds both an old technology and a new technology in a single production function, can mimic these facts surprisingly well. If ICBT innovations increase productivity only through their impact on existing or parallel innovations, we should expect the productivity gains to show up in the capital deepening term in a conventional growth accounting equation and not in total factor productivity, and even then only if one correctly estimates the impact of innovations on ICBT prices. INNOVATIONS, ECONOMIC GROWTH AND PRODUCTIVITY 59 in areas such as labor markets, finance, scientific systems and education. Since the weightless economy seems to be based on ICT and since ICT favors relocation and reorganization of the value chain, institutions that promote change and flexibility are best suited to the full exploitation of the benefits of the new technologies. An extreme version of this thesis is that there exists one best way to achieve the most satisfying economic performance regardless of the historical period, the dominant technological paradigm or the international regime. The new economy thesis referred to above does not require such a starting point. On the one hand, the institutional features of Continental European countries might have been instrumental in achieving a high degree of economic stability and growth in the post-war period, but they might no longer be suited to the new growth trajectory for a variety of reasons. On the other hand, the institutions of the Anglo-Saxon economies appear to favor entrepreneurship and investment in ITC and hence the emergence of the “new” economy. Therefore, if there is an ideal economic model at the present time, it would appear to be the US system. This statement questions the existence of institutional diversity among developed economies. Diversity is a fact that can be easily observed, but the interpretation of this diversity is a matter of debate. Should diversity be interpreted as heterogeneity of countries with respect to the process of adaptation to the ideal model? However, several researchers argue that diversity should not be understood in such a way (see Hall and Soskice 2001). In fact, diversity reflects the fact that there is no single one-best-way for a modern capitalist economy. Moreover, several varieties of capitalism may coexist. Each one of these may possess its own strong and weak points, reflected among other things in its pattern of trade specialization2 or its innovation system.3 The purpose of this contribution is to briefly analyze the diversity of capitalism and attempt to assess whether one particular type of capitalism is particularly suited for the new economy trajectory. The first section of the chapter analyses the diversity of capitalism with the help of the concept of institutional complementarity. The second section proposes a typology of modern capitalism based on identified complementarities. The third section suggests an empirical analysis and a typology of countries with respect to the production and diffusion of ICT and compares the results with the “varieties of capitalism.” The fourth section offers some tests of ICTs’ effects on economic growth. The fifth section provides a conclusion. Institutional complementarity Globalization and institutional diversity The current debate about transformations affecting contemporary economies is often centered on globalization and the consequences of new ICTs. As a result of 66 BRUNO AMABLE the deregulation of international trade and increasing competition on the product markets, globalization brings about an intensification of international economic relationships. The liberalization of financial activities facilitates international investment flows and tends to generalize the principles of market-based corporate governance worldwide. These include the provision of attractive incentives to high-ranking executives and the necessity of takeovers in order to achieve an efficient market for corporate governance. Capital has become more mobile, leading to an extension of the sphere within which private companies can act. As a result, the possibility for public intervention in the economy decreases. Firms have centered their strategies on the global market rather than national or “regional” spheres. This has increased the firms’ outside options and more generally their bargaining position vis-à-vis labor and the state. A consequence of this trend is increased competition between national territories that define their competitiveness in terms of their “factor endowment,” infrastructure, but also (and above all) local economic institutions. These regulations affect the functioning of factor markets (labor and capital) as well as the efficiency of the educational and training systems. The implications of this increased competition can be understood intuitively. If countries do not wish to fall behind in terms of their international competitiveness, countries experiencing economic difficulties need to align themselves with current best practices. But how can a country recognize these best practices? A rapid comparison of growth and unemployment performance over the 1990s would lead to the conclusion that the institutions of the USA and more generally “Anglo-Saxon” economies come very close to best practices. The USA was able to explore the new growth trajectory commonly understood as the new economy. The notion of a new economy is composite (see OECD 2000). The effect of this intensification of technical progress has been to raise productivity gains wherever they had been stagnating (in the United States). Technical progress has been concentrated in certain technologies and sectors, especially those involved in information and communication technologies. Innovation and technological dynamism played a greater role in the definition of competitiveness, both for firms and for countries as a whole. If ICTs define a new technological paradigm, it is tremendously important to adopt and perhaps produce these technologies in order to benefit from the productivity advances that they will make possible. Besides, technological dynamism at the onset of a technological cycle is seen as being dependent on small innovative firms. In order to develop efficiently, these firms need a favorable environment that includes flexible labor markets, facilitation of company forming and splitting, the availability of qualified personnel and easy access to venture capital. Market-based institutions, which are supposed to favor flexibility and entrepreneurship, play an important part in this story. As a result of the combination of globalization and the new economy, one might expect to observe a disappearance of national institutional specificities. Such specificities are the main preoccupation of a more or less homogenous IS THERE AN INSTITUTIONAL BASE OF THE NEW ECONOMY? 67 school of thought that focuses on the varieties of capitalism. This diversity is not seen as something that is accidental or temporary, but rather as the consequence of mechanisms that can be grouped under the generic title of institutional complementarities (see Aoki and Dore 1994; Aoki 2000, 2001; Amable 2000; Amable et al. 2000a). In this view, economies basically diverge in terms of the institutions that characterize them, depending on the particular aspect of the economy that is being studied. Thus, the labor market can be more or less regulated too, wage bargaining more or less centralized and the financial systems more or less reliant upon the banks or on the freedoms they have extended to the financial markets. In general, education is organized quite differently from one country to the next, with more or less close ties to the industrial sector, varying levels of university independence and different intensity of competition between private companies. Institutions’ influence on the economy should not be considered independently from one another. Rather, they exert a joint influence. Institutions affecting one area of the economy (e.g. the labor market) will have consequences beyond that particular area, if only because of general equilibrium effects. A rather simple example is that of wage bargaining: the outcome depends on each party’s outside options. These outside options are in turn dependent on the institutions affecting other areas than the labor market. These include the alternative job for the worker, which may depend on its skill level and hence on the institutions concerning the education and training system. Meanwhile, the alternative option for the firm may depend on its relocation possibilities, i.e. on the regulatory environment or the liquidity of the financial market. Amable et al. (2002) provide several definitions for institutional complementarity: • The differential definition requires continuity and is derived from the standard definition of complementarity in economics. The marginal efficiency of a certain institution is positively related to the presence/intensity of another institution in another area. For instance, if there is a complementarity between deregulated labor markets and deregulated product markets, less regulation in the labor market increases the marginal gain to deregulation in product markets. • The same logic as above applies to the case of comparative perform ance. However, a comparison between several situations is made. Adhering to product and labor markets, which may be regulated or not, there are four possibilities. Which is the best institutional combination? One considers discrete changes in the institutional environment instead of an institutional continuum. The conclusion may differ from that obtained with the differential method.4 • In the case of the dynamic definition, the presence of one institutional form in one area leads to the adoption of an institutional form in another area. The foundations of the institutional dynamics are left unspecified. 68 BRUNO AMABLE • The principle of conformity to a general logic maintains that institutions are said to be complementary when they have the same operating principles. Thus, the liberal or the coordinated logic leads to the adoption of deregulated or coordinated market features respectively. The first three definitions are very clearly related to one another. They make implicit reference to a “performance” criterion, which is necessary for appreciating complementarity. One may also rely on local criteria such as the level of unemployment for labor markets, the level of investment or the cost of capital for the financial system and the rate of innovation for the innovation system. However, the logic of institutional complementarity seems to imply the consideration of an aggregate indicator: welfare, GDP level or growth rate. Of course, this is not a matter of statistics, it rather concerns the way in which specific institutional forms are chosen. Agents with conflicting objectives will have different criteria for appreciating the performance of institutions: wages for workers or profits for firms. Several questions arise from such an assumption. How are each agent’s criteria related to a global performance index? Do agents internalize the interrelations between criteria? Workers may be aware that future employment depends on investment and hence on profits. This internalization is itself dependent upon institutions. Whether agents’ time horizons are long term or short term may depend on the capacity of institutions to stabilize the forecasting perspectives. The fourth definition is a little more problematic. The institutional complementarity approach considers economic models or systems as a set of complementary institutions. Whether one can sum up this pattern of complementarities into one general logic is debatable. For instance, if we define Anglo-Saxon economies through a set of complementary institutional forms such as a deregulated labor market and a markets-based financial system, can we then identify a general principle which would ensure that the whole liberal market economies (LME)5 model is coherent and viable? This remains an open question. The challenge to the institutional complementarity approach is accountable for the existence of diverse institutional configurations (economic models). Their stability and viability is long term but not eternal as changes need to be possible. These changes may result from different factors (see Amable et al. 2002): • A changing environment: technology is often mentioned in this respect. The existence of technological regimes suggest that inherent technological requirements may make industries perform differently in countries with distinct institutional systems. However, technology may evolve over time. Consequently, a change in the technological requirements of well-established industries may put a country’s institutional system under stress, eventually leading to a break-up of the historic compromise that underlies the system. Moreover, technologies may benefit differently from distinct institutional IS THERE AN INSTITUTIONAL BASE OF THE NEW ECONOMY? 69 systems during their evolution from nascent to mature stages. Hence, an exogenous dynamic pattern of technological evolution may, from time to time, put pressure on any institutional setting. For instance, the new technological regime centered on ICTs is supposed to have endangered the compromise between labor and capital on which economies, different from the Anglo- Saxon model, relied. • Unintended consequences: agents may take decisions that have unintended affects on the institutional structure. This can be caused by the presence of externalities, or because the consequences of their decisions are effective well beyond their time horizon. For instance, workers eager to obtain the highest income from their savings may favor the emergence of financial markets, which may destabilize a bank-based financial system, which was important in the system of institutional complementarities. If this bank-based system was complementary with a stable employment relation (see Amable et al. 2000b), its weakening may worsen the relative position of workers. • Conscious attempts at institutional design: institutions are the result of deliberate decisions. In times of crisis, agents may have their own motives for altering the institutional structure. Whether change actually takes place depends on the polity. The current debates over the necessity of “reform” in Continental Europe refer to these kinds of attempts to orient economies toward a more market-based system. Diverse types of capitalism Many studies on the diversity of capitalism have been published (see Jackson 2002). These studies either focus on specific countries or analyze more general categories such as liberal market economies versus coordinated market economies. In a theoretical analysis and an exercise in international comparisons, Amable et al. (1997) highlight four main types of capitalism or Social Systems of Innovation and Production (SSIP): • the market-based SSIP, encompassing the USA, the UK, Australia and Canada; • the social-democratic SSIP in the Scandinavian countries; • the meso-corporatist SSIP in Japan; • the “European” SSIP of France, Germany, Italy and the Netherlands. Rather than more or less stylized descriptions of the main characteristics of a given country, SSIP should be seen as ideal-types. Table 4.1 provides a summary of the main characteristics of each SSIP. The classification of SSIP into four groups resulted from a difficulty in classifying European countries. A homogeneous European group, which lumped together France and Germany with Italy and possibly other southern European countries raised some questions about the precise definition of European SSIP. 70 BRUNO AMABLE Limiting oneself to the realm of innovation, Italy and other Mediterranean countries seemed very different from Continental Europe, most notably in terms of R&D intensity and the importance of high-tech industries. Amable and Petit (2001) offer new analyses, which provide for a refined typology of SSIP. Extending the empirical analysis to 21 countries6 and updating it to include the end of the 1990s, they distinguish between six SSIP: • market-based SSIP; • social-democratic SSIP and meso-corporatist SSIP, with Korea joining Japan; • the “European integration” (or “public”) SSIP. This includes the countries that already belonged to this SSIP (France, Germany, the Netherlands). However, Italy is now categorized as a “Mediterranean” variant while Belgium and Ireland are now included in this category; • an “Alpine” variant of the preceding SSIP now comprises Austria and Switzerland; • the “Mediterranean” variant of the European SSIP is made up of Spain, Italy, Greece and Portugal. Besides the differentiation between the European SSIP into three variants, the main results reported in Amable and Petit (2001) were that there did not seem to be any general pattern of convergence toward the marketbased SSIP. Rather, there had been an advance of some market mechanisms in specific areas (mostly the financial system), while globalization affected SSIP in a differentiated manner. The market-based and “Mediterranean” SSIP were not affected as a group of countries by the empirical analyses of data that related to the new pattern of internationalization.7 Other groupings of countries were reshuffled, indicating a process of recomposition within the SSIP. IS THERE AN INSTITUTIONAL BASE OF THE NEW ECONOMY? 71 Table 4.1 Four types of social systems of innovation and production Centralization of wage bargaining 72 73 Diversity of capitalism The previous characterizations of SSIP were based on institutional variables as well as variables reflecting the economic structure. For instance, these include scientific, technological and industrial specialization. As a close connection between institutions and economic structures was assumed, this argument seems logical (Amable 2000). Furthermore, data on economic and technological structures are readily available whereas variables on institutional elements related to these elements are much more difficult to find. The analysis proposed in Amable (2003) distinguishes, as much as possible, institutional from economic variables. Five institutional areas are identified: product market competition, the labor market, the financial sector, social protection and the education sector. Based on previous results and other contributions to the literature (Aoki 2001; Hall and Soskice 2001), five different varieties of capitalism are distinguished: • market-based, or liberal market economies or the Anglo-Saxon model: the USA, the UK, Australia and Canada; • social-democratic economies: Sweden, Finland and Denmark; • Asian capitalism: Japan and Korea; • Continental European capitalism: France, Germany, Belgium, Austria, Ireland and Norway. • south European capitalism: Italy, Spain, Portugal and Greece. The institutional characteristics of these five varieties of capitalism are briefly summarized in Table 4.2. The existence and relative stability of differentiated models of capitalism is based on complementarities between the institutional features specific to each type of capitalism. The complete description of the institutional complementarities is outlined in Amable (2003). It is worth emphasizing the difference between this representation of the diversity of capitalism and the “variety of capitalism” approach of Hall and Soskice (2001). Hall and Soskice distinguish between two basic types of capitalism: liberal market economies (LME) and coordinated market economies (CME). The coordination dimension is crucial for understanding the pattern of differentiation among countries. In LME, actors primarily coordinate among each other through market signals whereas non-market relationships dominate in CME. Thus, countries can be arrayed along a continuum reflecting the extent of non-market coordination. This dimension is present in many areas of the economy such as processes of wage bargaining and the financial system. Due to its one-dimensionality, Hall and Soskice’s partition of countries cannot distinguish more than two models of capitalism. Also, even if the distinction 74 BRUNO AMABLE between some blatant LME such as the USA or the UK and some typical CME such as Sweden or Germany is clear, the position of some other countries is more ambiguous. For instance, Hall and Soskice (2001) cannot clearly identify France and Italy as either CME or LME. Being intermediate cases, they are also assumed to be less competitive than any of the pure types and thus condemned to join one or the other club sooner or later. It is possible to consider more than two types of capitalism, if one allows for more than one dimension for differentiation. The institutional complementarities between the various institutional forms of Table 4.2 allow these kinds of dimensions. In order to check the empirical relevance of the 5-type partition, Amable (2003) performs an empirical analysis, using variables that characterize the institutional areas taken into account previously. A cluster analysis broadly confirms the existence of five to six types of capitalism, depending on whether one distinguishes Switzerland and the Netherlands as a separate group apart from the rest of Continental Europe. The cluster analysis’ results are presented in Figure 4.1: the two dimensions describe the most fundamental oppositions between the different models. A first basic dimension is the regulation of markets. The one end of the spectrum shows deregulated markets, the other represents few obstacles to competition and regulated markets and limited competition. This dimension does not so much oppose LME and CME as found in Hall and Soskice (2001), but rather LME (or market-based economies, or “North Atlantic” capitalism) to south European capitalism. The second dimension is that of social protection and separates social-democratic countries from Asian capitalism. With regard to the competition and regulation of markets, continental capitalism might be considered as an intermediate form between market-based capitalism and South- European capitalism. Continental capitalism also possesses some of the socialdemocratic model’s features, particularly a less extensive social protection system, as is shown in Figure 4.1 (p. 76). ICT and the diversity of capitalism The spread of the new economy can, in part, be assessed through the relative importance of ICTs in production and consumption. As mentioned earlier, the new economy in general and ICTs in particular are often associated with marketbased capitalism. How accurate is this judgement? In order to answer this question, there is a need for an empirical analysis of the ICT sector. The sample of countries is the same that was used for the analysis of the diversity of capitalism. The OECD’s database on ICTs provides various statistical indicators on the production and diffusion of ICTs. These indicators are available for most OECD countries. A cluster analysis of these indicators checks whether ICTs exhibit the typology of varieties of capitalism introduced in the previous section. IS THERE AN INSTITUTIONAL BASE OF THE NEW ECONOMY? 75 The clustering presented in Table 4.3 only partly confirms the close relation between the market-based model and the new economy, at least when measured by ICT diffusion. On the one hand, Australia, although a marketbased economy, does not exhibit a clear superiority either in terms of diffusion or production of ICTs. Ireland and the Netherlands on the other hand, are two countries that are close to the Continental European model, belong to the same cluster as the UK, Japan and Korea. Finally, Sweden and Finland, which represent the socialdemocratic model, are in the same cluster as the US. Therefore, if countries are market-based economies, this may indeed facilitate them to explore the technological trajectory of the new economy, however, it does not seem to be a necessary condition. The differentiation of economic performance The analysis now turns to the impact of ITCs on macroeconomic performance. The new economy is held to lead to renewal of economic growth because the nature of technical change has been modified. ITC industries play a double role in this perspective. They enjoy the highest level of technical progress, with productivity gains that are significantly above those in other sectors. They also feature a very high rate of product innovation. Besides, their diffusion throughout the whole economy enables other sectors to improve their own productivity and to modify their offer in a way that encourages innovation. As such, information technologies play a role in their production as well as in their diffusion. As a result, we can expect that countries that have experienced a rapid diffusion of ICTs, or who have been involved in a significant production of ICTs, will be ahead of other countries, at least during the ascending phase of a long-term cycle. It will be easier to understand this ascending phase insofar as its institutional conditions facilitate innovative dynamism. In particular, these conditions relate to the flexibility of the labor market and above all to the dynamism of the financial markets: the availability of venture capital, the possibility that corporate executives can be motivated through the attribution of stock options, the liquidity of the financial markets, etc. Market-based finance can be synthesized in one single indicator—the price index of listed securities. In this view, the stock markets’ dynamism can be crystallized as the changes in this indicator. Another argument is closer to traditional macroeconomics and stresses that the developed countries did not all experience the same macroeconomic conditions over the course of the decade—if only because of the diversity of the macroeconomic policies that were being pursued (monetary policies in particular). The European recession of the early 1990s is not in fact unrelated to the restrictive monetary policies that were being carried out in light of the upcoming economic and monetary union. 82 BRUNO AMABLE Table 4.4 Regression results Note Significance levels: *10%, **5%, ***1%. Is it possible to explain international variations in macroeconomic performance on the basis of the aforementioned elements? The influence of ICT- related variables on growth was tested for a sample of the following ten OECD countries over the period 1991–2000: Canada, Denmark, Finland, France, Germany, Italy, Japan, the Netherlands, the UK and the USA. The test considered two separate ICT variables: the proportion of total GDP accounted for by the ICT-producing sectors and the same figures for sectors employing ICTs. Data on ICT producing and using sectors is drawn from the ICT database of the Groningen Growth and Development Center (University of Groningen, the Netherlands). The ICT-using sectors’ proportion of total GDP replaces a variable that relates to the diffusion of such technologies as it is difficult to obtain capital goods diffusion variables for an entire economy over a period of several years. Additionally, these kinds of variables are not particularly reliable. The real rate of interest is introduced in order to control macroeconomic policy. The test considers two macroeconomic performance indicators: the rate of growth of GDP and the rate of growth of GDP per hour. The findings are summarized in Table 4.4. One single ICT variable is significantly associated with growth rates: the proportion accounted for by the ICT producing sectors. The utilization of these technologies, as far as it can be measured by the employed variable, does not seem to have had any significant influence; in fact, the estimated coefficient is negative. Restrictive monetary policy, recognized with the level of the real rate of interest, seems to have played a negative role over the period; at least for the GDP growth rate regression. The negative coefficient of the lagged productivity level variable indicates a catch-up effect. IS THERE AN INSTITUTIONAL BASE OF THE NEW ECONOMY? 83 Conclusion The diversity of modern capitalism seems to be a durable fact, not because of the existence of rigidity in the process of adapting to the “one-best-way,” but because there are systemic reasons for countries to adopt different institutional structures. Just as the post-war growth regime was characterized by the diversity of capitalism, there is no reason to think that the new economy regime, if it is stable, will generate more institutional convergence. The superiority of the market-based model to other models of capitalism during the 1990s is not that evident. A satisfying growth performance was attributable to “sound” macroeconomic policies (particularly monetary policy) and the ability to produce technologies related to the new paradigm, i.e. ICTs. In this respect, the adequation of market-based capitalism and ICTs is not perfect: not all marketbased economies are particularly competitive with respect to ICT production while some European countries, whose economies are very distant from the market-based model, seem to be performing well. Thus, there is no strong evidence in favor of the impossibility for a non-liberal market economy to engage in the new technological trajectory. As a result, there is room to argue for a diversity of trajectories developing toward a “knowledge-based economy.” Diversity, however, does not mean absence of change. The different models will change by adapting to the new technological regime and new patterns of internationalization. In fact, some types of capitalism are likely to be more affected than others: the Continental European model is certainly under more stress than the market-based model. Some transformations are under way in the fields of product markets competition, labor markets and financial systems, which question the somewhat fragile stability of this model. But this stability is not a simple question of relative economic efficiency; more fundamentally, it is a matter of the stability of the socio-political compromises that underlie the institutions of the European model. Notes 1 The weightless economy is generally understood as comprising four main elements: (i) information and communications technology (ICT), the Internet; intellectual assets; (ii) electronic libraries and databases; (iii) biotechnology; (iv) carbon-based libraries and databases, pharmaceuticals. 2 See the chapter on institutional comparative advantage in Hall and Soskice (2001). 3 See the analysis of social systems of innovation and production in Amable et al. (1997). 4 If certain conditions are met, negotiations between social partners in a given labor market can create the sort of stable compromises that could help the workforce to receive a high level of training. In addition, physical investment is facilitated by the existence of close relationships between banks and firms. In this scenario, the existence of durable relationships and proximity between banks and firms enhances the implementation of long-term investment projects. In return this facilitates the 84 BRUNO AMABLE establishment of stable compromises in the labor market. Conversely, a flexible labor market, one that facilitates employee mobility, is seen here as complementing a financial system that facilitates the reversibility of commitments and the liquidation of investments. This means that the range of potential complementarities can be extended to cover areas such as innovation, professional training systems, etc. 5 Hall and Soskice define LMEs as being unified by market principles. Meanwhile, coordinated market economies (CME) such as Germany are unified by a principle of “strategic coordination.” 6 The United States, Japan, the United Kingdom, Canada, Australia, Italy, Spain, Portugal, Greece, France, Belgium, Denmark, Norway, Sweden, Finland, Germany, Austria, Switzerland, the Netherlands, Ireland and Korea. The empir ical data related to scientific and technological fields, economic structure, the educational system and the labor market As such, this is an extended conception of the innovation system. 7 The data included figures on foreign direct investment, international trade and the evolution of the financial systems. 8 Illustrative variables do not contribute to the definition of the factorial axes. Bibliography Amable, B. (2000) “Institutional Complementarity and Diversity of Social Systems of Innovation and Production,” Review of International Political Economy 7(4): 645–687. ——(2003) The Diversity of Modern Capitalism, Oxford: Oxford University Press. Amable, B. and Petit, P. (2001) The Diversity of Social Systems of Innovation and Production During the 1990s, Document de Travail CEPREMAP no. 15. Amable, B., Askenazy, P., Goldstein, A. and O’Connor, D. (1997) Les systèmes d’innovation à l’ère de la globalisation, Paris: Economica. ——(2000a) “Institutional Complementarities: An Overview of the Main Issues,” Presented at the seminar on Institutional Complementarity and the Dynamics of Economic Systems, Paris, April 5–6. ——(2000b) Institutional Complementarity: Labour Markets and Finance, Paris: CEPREMAP. ——(co-ordinated by Cohen, D.) (2002) “Internet: The Elusive Quest of a Frictionless Economy,” Report presented at Fondazione R.Benedetti “Gli aumenti di produttività e la disuguaglianza digitale nella nuova economia,” Catania, June 15. Aoki, M. (2000) Information, Corporate Governance, and Institutional Diversity, Oxford: Oxford University Press. ——(2001) Towards a Comparative Institutional Analysis, Cambridge, MA: MIT Press. Aoki, M. and Dore, R.P. (eds) (1994) The Japanese Firm: The Sources of Competit-ive Strength, Oxford and New York: Oxford University Press. Beffa, J.L., Boyer, R. and Touffut, J.-Ph. (1999) Les Relations Salariales En France, Notes de la Fondation Saint Simon no. 107. Boyer, R. (2001) “La ‘Nouvelle economie’ Au future antérieur: histoire, theories, géographie,” Document de Travail CEPREMAP no. 13. IS THERE AN INSTITUTIONAL BASE OF THE NEW ECONOMY? 85 Edwards, J. and Schanz, J. (2001 a) “Faster, Higher, Stronger: An International Comparison of Structural Policies,” Lehman Brothers Structural Economics Research Papers no. 3. ——(2001b) “Lehman’s Structural Database. Sources and Methods,” Lehman Brothers Structural Economics Research Papers no. 4, Freeman, C. and Soete, L. (1997) The Economics of Industrial Innovation, Cambridge, MA: MIT University Press. Freeman, R. (2000) “Single-Peaked Vs. Diversified Capitalism: The Relation between Economic Institutions and Outcomes,” NBER Working Paper. Hall, P. and Soskice, D. (eds) (2001) Varieties of Capitalism: The Institutional Foundations of Comparative Advantage, Oxford: Oxford University Press. Jackson, G. (2002) Varieties of Capitalism: A Review, Draft Rieti. OECD (2000) A New Economy? The Changing Role of Innovation and Information Technology in Growth, Paris: OECD. 86 BRUNO AMABLE 5 Europe in the innovation race Daniele Archibugi and Alberto Coco Introduction The aim of this chapter is to comparatively discuss the future trajectory of the socalled new economy in Europe, North America and East Asia. The idea that there is a “new” economy is certainly fascinating and it is not surprising that it has taken so much ground in the business world, the political community and the press. John Maynard Keynes knew very well that expectations play a fundamental role in fostering the business cycle. The hope that we might be experiencing the development of something as intriguing as a new economy has helped some corporations to support their stock market performance, some politicians to be elected or reelected and the press to increase their sales. The academic community is certainly not immune from these tendencies, although its function is to critically examine ideas that might have spread out too quickly. The scientific community should remember that new terms that are boldly introduced are often forgotten rather quickly and are replaced by others that seem more appealing. A certain extent of skepticism does not imply that one does not share the assumption that there is nothing new under the sun: now and then something new does occurs in economic and social life. Major changes have taken place in the last decade and some key components can be singled out. In particular: 1 The exploitation of knowledge has become more and more systematic, with an increasing propensity of business companies now searching for profit and growth opportunities in the exploitation of know-how. 2 The transfer in space of commodities, financial resources, expertise and information has become much easier. While the technical feasibility has increased exponentially, the economic costs have been dramatically reduced. 3 The number of players able to enter into old and new fields has also increased, leading to an accelerated pace of economic competition. Perhaps if one combines these three aspects, they can be labeled a new economy. Students of technological change, however, have preferred to use other terms such as knowledge-based economy, which emphasizes the role played by knowhow and competences in the economic sphere. We prefer to use the term “globalizing learning economy” (Lundvall and Borrás 1998; Archibugi and Lundvall 2001) since this seems to better capture the key role played by human learning in the economic and social landscape. The term globalizing (rather than global or even globalized) should help to remind us that the vast majority of the population of the world still does not have access to know-how that has already become obsolete in other parts of the globe (UNDP 2001). There is a widespread belief that economic growth, employment and welfare in the old continent will be more and more associated to its capability to generate, acquire and diffuse new knowledge. It is therefore not surprising that there is a major policy concern within governments, business and trade unions about the ways to promote scientific and technological activities, to foster innovation in firms and to upgrade the competencies of human resources. In order to develop a proper innovation strategy, Europe has to face the fact that it is composed of a number of states that do retain a substantial autonomy. What the old continent is gaining in terms of variety and diversity, it is losing in terms of lack of cohesion and central policy decision-making. Not surprisingly, Europe can better be described as an agglomeration of different innovation systems. While some regions of the EU are strongly integrated in knowledge transmission, others continue to be peripheral and are excluded by the major technology transfer flows. Therefore, one of the core issues that should be addressed both at the national and at the European policy level is how to integrate the different local and national components into a single innovative system comparable to the systems in the US and Japan. 1 The transformations toward a knowledge-based economy regard Europe as being engaged in a major institutional change. European integration has been driven by a variety of common policies such as the creation of a custom union, a common agricultural policy and, more recently, even a single currency. In spite of the efforts undertaken with the various multi-annual Framework Programs since the early 1980s, European integration is not yet driven by science and technology policy. Not more than 4.6 percent of the European Commission’s total budget is devoted to Research and Technological Development (RTD)— this accounts for less than 6 percent of the total amount spent by EU governments for RTD (Sharp 2001). In spite of the growing amount of resources that the EU has dedicated to RTD, this is still a minor part of the budget. However, it is significant that European governments continue to indicate very ambitious targets for the “old continent.” At the Lisbon Summit in March 2000, the European Council declared its willingness to make the European Research Area the world’s largest knowledge economy. At the Barcelona Summit in March 2002 the goal was formulated that Europe should reach a R&D/GDP ratio 88 DANIELE ARCHIBUGI AND ALBERTO COCO equal to 3 percent by 2010. 2 How realistic are these targets? Additionally, how is Europe performing in the technological race now? In the next section we present a broad set of data describing the technological status of Europe, as regards both investments—by means of the expenditure on R&D—and performance in innovative activities—by means of other well-known technological indicators. We compare Europe with the US and Japan and we highlight the recent evolution. Particular attention is devoted to ICTs since these are more strictly linked to the concept of new economy. In a following section we analyze the phenomenon of scientific and technological collaborations as we assume they reveal a lot about the “attractiveness” of the various regions of the world. Finally, in the last section we discuss the strategies Europe is applying to achieve a more prominent role in the globalizing learning economy. A new European technology gap? Like the US and Japan, Europe is a leading player in the generation of scientific and technological competencies. The combined R&D budget of the EU member countries is almost two thirds that of the US and more than one and an half that of Japan. In terms of scientific articles, the output of the EU is slightly higher than that of the US. But this strongly reflects the size of the EU, which has a population much larger than the US and Japan (see Table 5.1). If we look at the intensities, there are increasing signals that Europe is losing ground in the most dynamic and technologically advanced part of the economy. The concern about an increasing technological gap is certainly not new: as early as the 1960s we heard about “the American challenge” (Servan-Schreiber 1968) Table 5.1 Some indicators of size in the Triad, latest available year Sources: OECD (2001) for gross R&D expenditure; Worldbank (2000) for the other indicators. Note PPP GDP is gross domestic product converted to international dollars using purchasing power parity rates. An international dollar has the same purchasing power over GDP as the US dollar has in the United States. Population and GDP refer to 2001, scientific articles to 1999, GERD and Internet users to 2000. and similar concerns were reiterated in the 1980s and in the 1990s (see, for example, Patel and Pavitt 1987; Archibugi and Pianta 1992). EUROPE IN THE INNOVATION RACE 89 Surely, Europe is not the only region to be concerned about losing its leadership. Similar concerns were echoed in the US (Kennedy 1987; Pianta 1988; Nelson 1990) and if we had sufficient insight into Japan’s discourse, we would certainly find comparable statements. But by simply saying that the neighbor’s grass is always greener, we cannot dismiss the issue about European economies’ poor performance in key aspects of knowledgebased production. Table 5.2 reports some data about the R&D intensity (Gross R&D expenditure, GERD, as a percentage of GDP). The EU’s intensity is equal to 1.93 percent, substantially lower than that of the US (2.69) and Japan (2.98). In the second half of the 1990s Japan grew more than the US, while the EU’s growth remained stagnant. Within the EU, a clear regional divide between the North and the South emerged. The country with the highest intensity, Sweden, has a value that is almost six times higher than that of Greece. Table 5.2 Gross R&D expenditure as a percentage of GDP by country, 2000 and 1995 Sources: EC (2002), data taken from Eurostat and OECD for USA and Japan. Note Data for Greece, Ireland, Italy, Belgium, The Netherlands, Denmark, Sweden, Canada, Norway, Switzerland refer to 1999; Japan data refers to 1996–2000. A similar pattern emerges in terms of business R&D (BERD) as a percentage of the Domestic Product of Industry (DPI), reported in Table 5.3. In this case, the 90 DANIELE ARCHIBUGI AND ALBERTO COCO difference between the first and the last EU country is even higher: Sweden has a BERD intensity 17 times higher than Portugal. The dispersion among European countries has further increased. It will be difficult to find other aspects of economic and social life where the differences between European countries are so profound. Table 5.4 shows the number of patents granted per million people for various countries at the US Patent and Trademark Office (USPTO). The high ratio for the US reflects the fact that inventors and firms are seeking patents in their domestic market. But Japan and the EU are on an equal footing since for both of them the American market is crucial. The figures show that Japan has a ratio that is almost four times higher than the EU’s average. Not even the European countries with the highest propensity to patent, that is, Switzerland and Sweden, Table 5.3 Business R&D expenditure as a percentage of DPI by country, 1999 and 1995 Sources: EC (2002), data taken from Eurostat and OECD for USA and Japan. Note Denmark and Finland have last data for 2000. EUROPE IN THE INNOVATION RACE 91 Hagedoorn 1996). As much as 60 percent of the total strategic technology alliances recorded are international in scope. This form of generating technological knowledge has considerably increased its significance. As a result, the number of recorded agreements has nearly tripled between 1980 and 1982, and 1998 and 2000. The largest and most increasing portion of alliances takes place within the US: 45.8 percent of all the strategic technological alliances recorded between 1998 and 2000 occurred among American firms only; from 1980–1982 it was only 24. 6 percent (NSF 2002). Moreover, US firms have strong ties on both the Atlantic and the Pacific shores: from 1998 to 2000, US companies have participated in as much as 84.7 percent of the recorded technology alliances. On the contrary, the share of intra-European strategic technological alliances has substantially declined: they accounted for 18.2 percent from 1980 to 1982 and less than 10 percent from 1998 to 2000. They have even decreased in absolute terms in the last decade (from 74 percent in 1989–1991 to 53 percent in 1998–2000). European policymakers should be concerned with the strong propensity of European firms to seek American, rather than European, partnerships. Policies carried out at the European level, especially at the European Commission’s level, to foster cooperation in R&D and innovation in the continent have not been able to reverse the propensity of European firms to engage in partnerships with American firms. The first possible explanation would be that the absolute amount of resources devoted to science and technology is much greater in US firms—firms engage in technology alliances with partners who have the adequate expertise. The shift of alliances to the US can thus be seen as the result of the amount of US companies’ investment in knowledge. In order to account for this phenomenon, we divided the number of European alliances undertaken by the total amount of European, US and Japanese business enterprises’ R&D expenditure (BERD). This provides an indicator of the propensity of European companies toward collaboration in each of these regions. The results are reported in Table 5.10. 98 DANIELE ARCHIBUGI AND ALBERTO COCO Table 5.9 Distribution of strategic technology alliances between and within economic blocs, 1980–2000 Source: Our elaboration from NSF (2002). 99 Table 5.10 Propensities for strategic technical partnerships, 1980–2000 Sources: Authors’ elaboration from NSF (2002), data from MERIT database and from OECD (2001). Note Methodology: The number of strategic technological agreements recorded by the MERIT database have been divided by the Business Expenditure on R&D of the region expressed in constant 1992 purchasing power parity US billion dollar. It reads for example that in 1980–1982 there have been 0.8 strategic technology agreements involving European firms for each US dollar billion of European BERD. Although the attractiveness of the US economy emerges to be a bit smaller in relative terms, the results in Table 5.10 confirm that European companies have a greater propensity for American partnership. There are 1.07 European-US partnerships for each billion US dollar BERD, while the equivalent figure for intra-European partnership is just 0.62. Moreover, the European business community has considerably changed its propensity for partnership over the last ten years: for the periods from 1980 to 1982 and from 1989 to 1991, European companies had a larger propensity for European rather than American partners. The figures were, respectively, 0.80 and 0.61 agreements for each billion US dollar BERD from 1980 to 1982 and 1.03 and 0.86 from 1989 to 1991. The lower part of Table 5.10 reports the propensity of American companies to undertake alliances. US companies are now keener to undertake joint ventures with European partners and this is a result of the overall increase of their engagement in collaborations. However, Table 5.10 illustrates a great propensity for internal partnerships, which has grown about three times in the last 20 years. If the new economy is represented—among other things—by strategic 100 DANIELE ARCHIBUGI AND ALBERTO COCO technology partnership, the evidence suggests that this strongly leans out toward the US rather than toward Europe or Japan. Partnerships and collaborations promoted by public research institutions and universities equally play a crucial role in the international dissemination of knowledge. They can take a variety of forms: joint research centers, exchange of students and of academic staff or sharing of scientific information. One of the ways to measure it is by looking at internationally co-authored scientific papers. A dramatic increase in the internationally co-authored papers—facilitated by the diffusion of the Internet and e-mail —is evident in all countries (Table 5.11). From 1986 to 1999, the percentage of internationally co-authored papers has nearly doubled and this represents a clear signal of globalization in the generation of knowledge. On an individual level, European countries are keener to collaborate than the US and Japan. Given the smaller size of each country’s scientific community, this fact should not come as a surprise. From a dynamic viewpoint, the rate of increase has been higher in the US and Japan than in European countries, but this is due to the fact that the US has lowered the growth of national scientific articles (see NSF 2002: Table 5.41). These data clearly show that the academic community in Europe is a valuable asset for the acquisition of knowledge and expertise beyond the borders of their countries. Does the academic community also share the same preference of European firms for American rather than for European partners? Table 5.12 reports the distribution of internationally co-authored collaborations in the triad: Europe is by far the greatest collaborator for the American academic community. From 1995 to 1997 as much as 60.3 percent of internationally co-authored papers in the US involved a partner in the EU. Additionally, Europeans have a strong propensity to collaborate among each other. This fact could be misleading if we think that a paper co-authored by a Dutch and a Belgian scientist is classified as “international,” while a paper co-authored by a Californian and a New Yorker is classified as national. Still, the US remains the single most important nation for Europeans to collaborate with. But the data’s significance lies in the time evolution (and this is not affected by the different size of the countries): by comparing the first period (1986–1988) to the last one (1995–1997), it becomes evident that intra-EU collaborations are increasing (from 56.6 to 69.4 percent of all internationally co-authored papers), while EU-US collaborations are decreasing for the EU as a whole (from 31.9 to 29.0 percent) as well as for each EU member country. Looking at the data from an American perspective, the above tendency is enhanced: the share of intra-US articles in all US co-authored articles declines from 78 to 68 percent, while the coauthorship with authors based in the EU grows from 11 to 19 percent (NSF 2000: Table 6.51). EUROPE IN THE INNOVATION RACE 101 Table 5.11 Percentage of internationally coauthored scientific papers in selected countries in all scientific papers, 1986 and 1999 Notes National rates are based on total counts: each collaborating country is assigned one paper (a paper with three international coauthors may contribute to the international coauthorship of three countries). We could not calculate the EU total, as it would contain multiple counting. We therefore note an inverse tendency: the European business community has an increasing propensity for technological alliances with US firms, while the European academic community has an increasing propensity for intra-European partnership. One of the main policies used by the European Commission in the last decade has been to promote collaborations among European institutions and firms through the instrument of the Framework Programs. The data reported suggest that these policies have been much more successful in academia than in business. It seems as if the limited resources the European Commission disposes (about 4,000 million euro a year in the last approved Sixth Framework Program) have not been enough to face the needs of European industry, while they have revealed to be more effective in the training and mobility of researchers. 102 DANIELE ARCHIBUGI AND ALBERTO COCO Table 5.12 Distribution of internationally coauthored papers across collaborating countries, 1986–1988 and 1995–1997 Sources: Our elaboration from NSF (2000), data from Institute for Scientific Information. Notes Row percentages may add to more than 100 because articles are counted in each contributing country and some may have authors in 3 or more countries. With regard to European Union, internationally coauthored articles also include those among members countries. Rows report the percentage of the total number of international coauthorships of the country. Columns indicate the relative prominence of a country in the portfolio of internationally coauthored articles of every country. A single Europe for science and technology? The evidence provided confirms that in some vital areas of knowledge and competence building Europe is lagging behind. Contrary to what happened for many periods after the end of the Second World War, the gap between Europe and the US has increased in the 1990s. It is therefore understandable that a major policy concern in Europe is to identify strategies that would allow for the catching up and upgrading of its scientific and technological competence. Europe is dominated by vast regional disparities and they are much wider in terms of scientific and technological competencies than in income, production or consumption. Germany, which for long has been the technological engine of Europe, has to face a major regional problem: the integration of the former German Democratic Republic. The UK, soul of many centers of scientific excellence, has under-funded its universities for more than twenty years. The EU’s enlargement to 15 countries has integrated some small and highly dynamic countries such as Sweden and Finland, but future enlargements will not bring nations with this kind of sophisticated dowry of scientific and technological infrastructures. The candidates to join the EU are nations in which the scientific community suffered hardship for many years while the business world is still far from the competitive proficiency of the EU’s more advanced countries. Future enlargements will lead to a EU with a larger population and an expanded market, but with a somewhat reduced intensity of scientific and technological capabilities. At the Council Summit held in Lisbon in March 2000, European governments set themselves a very ambitious goal: to make Europe the most competitive and dynamic knowledge-based economy in the world. As it often happens with political statements, there is a certain divergence between the target announced EUROPE IN THE INNOVATION RACE 103 and the instruments made available. Too little commitment has been expressed to reach this target. More recently, at the summit of March 2002 held in Barcelona, the EU has set the more ambitious target to raise R&D expenditure to 3 percent of the GDP by 2010. Further, it has set itself the goal to increase the presence of the business sector, which, by 2010, should finance two-thirds of total R&D expenditure. As we know, R&D intensity has a positive relation with economic growth, but such an objective appears very difficult to reach if we simply consider its evolution in the last decade (Schibany and Streicher 2003). One of the results is that it would imply a strongly needed, but hard to realize, increase in research personnel of about 100,000 units per year. An ambitious goal like this requires serious policy measures: an import ant one is the direct financing by governments. This is limited by the Maastricht criteria on public balance on the one hand and, on the other, by the aim to increase the private presence at the expense of the public. Indirect measures include tax incentives for the industry, but there are doubts about the leverage effects they will display. Surely a greater effort by the private sector is necessary, as it is also testified by the low level of venture capital financing in Europe with respect to US and Japan (EC 2002). This form of financing is particularly significant for the promotion of innovative activities by small firms (the so-called start-ups). Thus, what are the possible actions to pursue in order to increase Europe’s level of innovation? It is difficult to give a unique answer: we can take for granted the fact that institutions play a fundamental role in fostering and creating the premise for technical change. Recently, the concept of Innovation Systems (IS) is being replaced by the more complex one of Social System of Innovation and Production (SSIP). This means that not only direct scientific institutions must be considered for the technological upgrading of an area, but rather the economic system as a whole, including the relations in the markets of production factors (such as labor market connections), the educational and the financial institutions which are all linked by mechanisms of reciprocal complementarities (Amable and Petit 2001). A second consideration to keep in mind is that ICT must be considered as a strategic sector of long-term growth. As a result, the initiatives of the EU, expressed in the various Framework Programs, are indeed welcomed and well-posed, but they are certainly not sufficient to achieve the targets set by the European Council. This frequentlycited instrument is promoted every four years and allows the European Commission to progressively revise and enlarge the areas of intervention. In this, it follows the leading principle to enhance intraEuropean co-operation in the socalled pre-competitive research fields. In the last approved Sixth Framework Program, the greatest bulk of resources has been dedicated to Informative Society Technology and Nano-technologies (4,925 million euro over the next four years). Thus, the strategic importance of this sector has indeed been recognized, not only for the new jobs and business that the 3G wireless communication systems, the software architectures and the opto-electronics 104 DANIELE ARCHIBUGI AND ALBERTO COCO networks can create, but also because it perfectly meets the request of “ambient intelligence,” that is the target to link economic growth with welfare purposes. The other research areas are devoted to the same aim: biotechnology, environment and energy. However, if these are the actions that the EC wants to engage in, its budget is too limited. The ambitious targets will require a much larger commitment of national resources in terms of funding of the existing centers of excellence (especially when they have been kept under severe financial restriction), to generate the human resources needed for both the public and the business institutions and to start up new problem-oriented institutions. In addition to the (limited) financial instruments, regulations, standards, procurement, competition, real services and large-scale cooperative civilian projects are essential instruments to create a European Research Area (Lundvall 2002). Parallel to the European commitment, each country makes its own attempts to upgrade its scientific and technological potential. A small country such as Ireland, for example, has managed to improve its technological potential by making the country attractive for multinational corporations. This is not the first time that European governments have preferred to follow an autonomous route. But in an era increasingly dominated by social and economic globalization, the linkages between European nations are so strong that many science and technology policies must be bound to a common European faith. The Commission’s schemes have the advantage to reward excellence and to foster the crucial fields where Europe is lagging behind and to involve Europe’s least developed regions. The difficulty to implement these policies in an EU of 15— soon to be 27—member countries, will continuously increase. The enormous differences between European member countries show that in order to fully exploit the advantages of knowledge, it is crucial to develop strategies for the transmission and diffusion of competencies across areas. Only by reducing regional disparities it will be possible to obtain a European scientific and technological competence comparable to that of the US and Japan. In this regard, the only risk of ICT development is that it could enlarge the already existing gap between northern and southern European countries. A stronger integration between the academic and the business community is needed, which in turn requires major changes in the institutional setting and in the incentives of the existing publicly funded research centers. The evidence reported in this chapter has clearly indicated that a small club of European countries has a scientific and technological intensity on par with, and often superior to, the US. The Scandinavian countries have followed a distinctive approach to competence building based on a highly competent and qualified labor force, generated through massive investments in education and training. This model should serve as the model for European policy-making rather than the American model based on firms competing for market shares and R&D public investment concentrated in national priorities such as defense and space. In this regard, the last Framework Program embarked on the right path. EUROPE IN THE INNOVATION RACE 105 It is evident that successful management of the learning economy will require a much higher political commitment, which should be comparable to the efforts European governments have devoted to creating a single currency. Lundvall (2002) has suggested the creation of a European High Level Council on Innovation and Competence Building chaired by the president of the EU. Its political weight should not be less than that of the European Central Bank. This would be a clear sign that there is a Euro pean political commitment to become “the most competitive and dynamic knowledge-based economy in the world” within the next decade. But words without actions will only allow us to observe that at the end of the decade the aim to increase R&D to 3 percent of the GDP has not been achieved and the European technology gap has further widened. Notes 1 Amable and Petit (2001), Maurseth and Verspagen (1999), Garcia-Fontes and Geuna (1999) and, more broadly, the chapters collected in Archibugi and Lundvall (2001) present some evidence and considerations relating to the lack of a proper European Innovation System. 2 For an assessment of these targets, see Soete 2002; Schibany and Streicher 2003. Bibliography Amable, B. and Petit, P. (2001) The Diversity of Social Systems of Innovation and Production During the 1990s, Paris: La Defense. Archibugi, D. (1992) The Technological Specialisation of Advanced Countries: A Report to the EEC on International Science and Technology Activities, Boston: Kluwer. Archibugi, D. and Lundvall, B.-Å. (eds) (2001) The Globalising Learning Economy, Oxford: Oxford University Press. Archibugi, D. and Pianta, M. (1992) The Technological Specialisation of Advanced Countries: A Report to the EEC on International Science and Technology Activities, Boston: Kluwer. Archibugi, D., Howells, J. and Michie, J. (eds) (1999) Innovation Systems in the Global Economy, Cambridge: Cambridge University Press. Bank, W. (2001) World Development Indicators 2001, Washington, DC: World Bank. Cantwell, J.A. and Iammarino, S. (2001) Multinational Enterprises and Regional Systems of Innovation in Europe, London: Routledge. Chesnais, F., Ietto-Gillies, G. and Simonetti, R. (eds) (2000) European Integration and Global Corporate Strategies, London: Routledge. Community Research and Development Information Service (CORDIS) (2002) Sixth Framework Program, Luxembourg: CORDIS. Daveri, F. (2002) The New Economy in Europe (1992–2001), Milan: IGIER. David, P. and Foray, D. (1995) “Accessing and Expanding the Science and Technology Knowledge-Base,” Science, Technology Industry Review 16:13–68 . European Commission (2002) Key Figures: Towards a European Research Area, Luxembourg: European Commission. 106 DANIELE ARCHIBUGI AND ALBERTO COCO European Informative Telecommunications Observatory (EITO) (2001) Yearbook 2001, Frankfurt a.M.: EITO. European Patent Office (EPO) (1995–2001) Annual Report, Munich: EPO. Fagerberg, J., Guerrieri, P. and Verspagen, B. (eds) (1999) The Economic Challenge for Europe: Adapting to Innovation-Based Growth, Aldershot: Edward Elgar. Freeman, C. (1987a) Technology Policy and Economic Performance: Lessons from Japan, London: Pinter. ——(ed.) (1987b) Output Measurement in Science and Technology , Amsterdam: North Holland. Gambardella, A. and Malerba, F. (eds) (1999) The Organization of Economic Innovation in Europe, Cambridge: Cambridge University Press. Garcia-Fontes, W. and Geuna, A. (1999) “The Dynamics of Research Networks in Europe,” in Gambardella, A. and Malerba, F. (eds) The Organization of Economic Innovation in Europe, Cambridge: Cambridge University Press. Hagedoorn, J. (1996) “Trends and Patterns in Strategic Technology Partnering since the Early Seventies,” Review of Industrial Organization 11(5): 601–616 . Held, D., McGrew, A., Goldblatt, D. and Perraton, J. (1999) Global Transformations: Politics, Economics and Culture, Cambridge: Polity Press. Kennedy, P. (1987) The Rise and Fall of the Great Powers, New York: Random House. Lundvall, B.-Å. (2002) “Innovation Policy in the Globalising Learning Economy,” in Archibugi, D. and Lundvall, B.-Å. (eds) The Globalising Learning Economy, Oxford: Oxford University Press. Lundvall, B.-Å. and Borrás, S. (1998) The Globalising Learning Economy: Implications for Innovation Policy, Brussels: European Commission, D.G. XII. Mathews, J. (2000) Tiger Technology: The Creation of a Semiconductor Industry in East Asia, Cambridge: Cambridge University Press. Maurseth, P.B. and Verspagen, B. (1999) “Europe: One or Several Systems of Innovation? An Analysis Based on Patent Citations,” in Fagerberg, J., Guerrieri, P. and Verspagen, B. (eds) The Economic Challenge to Europe. Adapting to Innovation- Based Growth, Aldershot: Edward Elgar Publishing. Molero, J. (ed.) (1995) Technological Innovation, Multinational Corporations and the New International Competitiveness, Reading: Hardwood Academic Publishers. Mytelka, L.K. (1991) Strategic Partnership. States, Firms and International Competition, London: Pinter Publishers. ——(2001) “Mergers, Acquisitions, and Inter-Firm Technology Agreements in the Global Learning Economy,” in Archibugi, D. and Lundvall, B.-Å. (eds) The Globalizing Learning Economy, Oxford: Oxford University Press. National Science Foundation (NSF) (2000) Science and Engineering Indicators 2000, Washington, DC: US Government Printing Office. ——(2002) Science and Engineering Indicators 2002 , Washington, DC: U.S. Government Printing Office. Nelson, R. (1990) “Technological Leadership. Where Did It Come From and Where Did It Go?,” Research Policy 19(2): 117–132 . Niosi, J. (1999) “The Internationalization of Industrial R&D,” Research Policy Special Issue, Guest Editor J.Niosi, 28(2–3): 107–336 . OECD (2001) Main Science and Technology Indicators 2001–2002, Paris: OECD. EUROPE IN THE INNOVATION RACE 107 on the basis of the compensation hypothesis a reciprocal increase in the relationship between welfare state social security payments and public innovation expenditure can be expected. For each of these two hypotheses, empirical evidence can be found: while the USA and Japan are characterized by a high share of research and development (R&D) expenditures and a low level of social security expenditures, some Scandinavian countries—especially Finland and Sweden—combine high research and development expenditures and a high proportion of researchintensive industries with high social expenditures. The controversy outlined above, therefore, cannot be decided on the basis of single case studies. It is necessary to include a larger group of countries. The method and the data The connection between innovations and social security provisions will be analyzed on the basis of internationally comparative data. On the basis of the data6 collected by the Organization for Economic Co-operation and Development (OECD), this task could simply be solved by bivariate analyses— for example by a regression or a scatter diagram (cf. Figure 6.1). It turns out that the relationship between social security and research and development expenditure exhibits a very strong correlation coefficient of r2=0.22. This can be understood as a confirmation of the compensation hypothesis. However, such an interpretation would not take into account that the strong correlation is perhaps only the result of a third variable not being taken into account—for example the result of the economic prosperity or the integration into the world markets: richer countries, who are more integrated in the world market invest more in research and social security. Due to the small number of developed countries, the control of such intervening variables is normally not possible, because the number of advanced industrial countries is too low: the required data will hardly be available for all 30 OECD countries. Nevertheless, multivariate analyses can be carried out if the necessary data are available for several years. If, for example, data for 20 countries are available for 20 years (in our case: for two decades, the 1980s and 1990s 7), then the number of observations can be increased to a maximum of 400. Classic linear regressions could be carried out through this “pooling” of data, if the observations for one variable for one country at different times were independent from each other. However, this assumption (technically speaking: the assumption of no unobserved heterogeneity) is extremely implausible; it implies, for example, that there is no connection between the income inequalities in a country in the years 1980 and 1981. This assumption can be checked with the Breusch-Pagan Test (Schwarze and Härpfer 2003; Breusch and Pagan 1980). Normally, this assumption will be rejected. It is not only implausible, but also statistically refutable. In this case, different panel regressions can be used—above all models of fixed or random effects. 114 INNOVATION AND SOCIAL SECURITY Models of fixed and random effects differ regarding the assumptions concerning the error terms: models with fixed effects assume countryspecific constant error terms over a period of time; in models with random effects, the country-specific effects are considered as random variables. Whether one or the other assumption applies can be checked with the Hausman Test. The zero hypothesis of the Hausman Test is: the country specific errors and the explanatory variables are not correlated. If this hypothesis cannot be rejected, because the differences between estimated coefficients of the random-effects model and the fixed-effects model are not systematically different from zero, a model with fixed effects as well as a model with random effects can be used. If the zero hypothesis is rejected, then a model with fixed effects is preferred (Greene 2000: chapter 14; Baltagi 2001). An advantage of using exclusively the model with fixed effects is that the estimations are unbiased even if the fixed effects and the explanatory variables are correlated; a disadvantage is that influences of time-constant variables (for example the culture of a specific Figure 6.1 Public social expenditure and gross domestic expenditure for research and development (in percentage of gross domestic product, 24 OECD countries, 1999) (sources: OECD Statistical Compendium own calculations). MARTIN HEIDENREICH 115 country) cannot be identified, because it cannot be distinguished from the fixed error terms. After the Breusch-Pagan Test and the Hausman Test, a third test must be carried out to determine whether the error parameters of the variables are timedependent. A conventional procedure for this is the application of a modified Durbin-Watson Test (cf. Baltagi 2001:95). If the corresponding value clearly deviates from 2, this can be interpreted as a correlation of the residuals over time. The rho value also shown in the following tables indicates the strength of these auto correlations (zero: no auto correlation). The squared correlation coefficient (r2) indicates the amount of variance explained by the entire model (including the 19 dummy variables for the years not listed in Tables 6.3–6.5). With regard to these tests, Tables 6.3–6.5 use fixed-effects models with firstorder autoregressive disturbance terms (AR1). The estimates of these models are carried out using the procedure xtregar of the program STATA 8. The median values of the variables used in the following tables are portrayed in Table 6.2. Innovations and social security: the empirical results In order to analyze the relationship between social security and innovations, two groups of dependent variables have to be distinguished. On the one hand, there are the indicators measuring the input of innovation processes, for example the expenditures for research, development and education. On the other hand there are the output indicators, for example the number of patents or the share of advanced technologies. In this chapter, the connection between social security systems and innovations is discussed on the basis of three different groups of dependent variables. First, the relationship between social security and innovation expenditure, then the connection between social security and patents and finally the connection between public social security expenditure and the relative weight of knowledge-based industries and services. 116 INNOVATION AND SOCIAL SECURITY Table 6.2 Innovations and social security: mean values and standard deviations (20 OECD countries; 1980–1999) Note a The proportion of real net output of high-technology industries (aircraft and spacecraft, pharmaceuticals, office, accounting and computing machinery, radio, television and communications equipment, medical, precision and optical instruments), of medium-high- technology industries (electrical machinery and apparatus, motor vehicles, trailers and semi-trailers, chemicals excluding pharmaceuticals, railroad equipment and transport equipment, machinery and equipment) and from market-related knowledge-based services (post and telecommunications, finance and insurance, business services) can be calculated on the basis of the OECD STAN database (cf. also OECD 2003). MARTIN HEIDENREICH 117 Social security an innovation expenditure Research and development expenditure in percentage of the gross domestic product is a central indicator for the input to innovation processes. However, a weakness of this indicator is that it focuses only on systematic innovation activities especially in larger businesses. Therefore, the inclusion of additional indicators, which measure the role of technical knowledge based on investments in machinery and equipment, software or higher education would be highly desirable. In a recent study, the OECD (2003) proposed a more comprehensive concept of innovation expenditure taking into account all these expenditures. It has been estimated that in the year 2000 in the 30 OECD countries, 2.3 percent of the gross domestic product (2000) was spent on research and development, 1.3 percent on software, 1.3 percent on higher education and 8.4 percent on machinery and equipment. However, the last three types of expenditure can only be estimated for some recent years. Therefore, it is necessary to use other input indicators: the indicator “knowledge investments” will be made up of the expenditure for research and 118 INNOVATION AND SOCIAL SECURITY development and for education (subtracting the share of research and development expenditure in higher education, since this expenditure is included in both positions). On the one hand, these two indicators overestimate the investment in knowledge, since the expenditure for primary school education can hardly be counted as innovation expenditure. However, since the educational expenditure and the expenditure for higher education in the second half of the 1990s are closely correlated (r=0.63), the changes in educational expenditure can be taken as indicators for the size of innovation-relevant investments in human capital. On the other hand, the innovation expenditure is underestimated since the investments in software and into new machinery and equipment are blanked out. The OECD (2003:74) submitted estimations of the investment in software for the last years. However, these data are not available for all the eighties and nineties. In the following, we will therefore take the educational and research expenditure as input indicators. In Table 6.3, different models are reported, in which the connection between innovation expenditure and public social security expenditure is captured in different ways. In each case, the purchasing-power-adjusted gross domestic product (GDP), the integration into the world market (the sum of the import and export quotas), and the dummy variables for the different years (with the exception of the reference year 1980) are included as control variables. The coefficients of the constant and these dummy variables are not reported in the following tables. In the first column of Table 6.3, a model is introduced, which—with the exception of the additional control variables—essentially corresponds to the model portrayed in Figure 6.1. The results are surprising: in contrast to Figure 6.1, there is no significant correlation between research expenditure and social security expenditure. When the economic performance and the integration into the world market are taken into consideration, there is no connection between the two variables. This points to the previously indicated limitations of bivariate analyses. In international comparisons, bivariate analyses can be used only with the utmost caution. However, only 29 percent of the research and development expenditure in OECD countries is financed by governments (OECD 2003:21). 64 percent (2000) of R&D expenditure is carried by the private economy. If only the government-financed R&D expenditure (column 2) are included, a significant connection emerges between R&D and social security expenditure. The connection between public social security expenditure and public and private educational expenditure is also significant (column 2): an active state invests in education as well as in research and in the social security of its citizens. MARTIN HEIDENREICH 119 Table 6.3 The relationship between educational and research expenditure and social expenditure (20 OECD countries, 1980–1999) Sources: OECD: Statistical Compendium and own calculations. Notes Absolute value of t statistics in parentheses; balanced fixed-effects model when the disturbance term is first-order autoregressive; data for 20 OECD coun-tries from 1980–1999. Time dummies and constants included but not portrayed in the table. * significant at 10%; † significant at 5%; ‡ significant at 1%. Breusch-Pagan Test on the existence of country-specific effects (chi-square distributed with a degree of freedom). Hausman Test on failure specifications using a model with random effects (chi-square distributed; number of degrees of freedom shown in the table). 120 If the public and private educational and research expenditure is added and the overlapping expenditure for research in higher education is deducted (column 4), the correlation between knowledge investments and social security expenditure is no longer significant. If the public and the private research and educational expenditure is taken as an indicator for the input of innovation processes, neither the efficiency nor the compensation hypothesis can be confirmed. However, this is not valid if only public educational and research expenditure (column 5) is included: the correlation between public social security and educational and research expenditure is highly significant. The compensation hypothesis therefore applies if it is understood as a statement about political actors: states that invest a great deal in social security also invest a great deal in education and research. However, a general correlation between innovation expenditure and social security expenditure—as predicted by the compensation hypothesis—cannot be proved.8 In the last three columns of Table 6.3, the connection between social security expenditure and public educational and research expenditure is examined more closely. At first, the public social security expenditure is divided into transfer payments and into expenditure for the public provision of services (column 6). Both indicators are significantly correlated with research and educational expenditure. A state which invests in research and education does not have any preference for transfer payments or for a developed public service. In the seventh column, the hypothesis is tested that “future-oriented” social expenditure (expenditure for families, for active labor market policies and for health services) in comparison to ex post measures (unemployment benefit, pensions, etc.) are more closely associated with research and educational expenditure. After the inclusion of three different “ex ante” types of social expenditure it can be seen that only the expenditure of publicly-financed health services is correlated significantly with the amount of the research and educational expenditure. This indicates the importance of “life sciences,” which will be one of the future growth markets in view of the older population of most OECD-countries. In the last column, the hypothesis is examined that the relative weight of social security contributions and of taxes on income and goods and services influences the state commitment to education and research. It could be supposed that a high share of social security contributions limits the possibilities of the state to finance research and education investments. This hypothesis cannot be confirmed. In conclusion: neither the negative correlation between social security and innovation expenditure predicted by the efficiency thesis nor the positive correlation predicted by the compensation thesis can be adequately proven. However, a positive correlation between public social security and educational and research expenditure can be proved. This connection cannot be interpreted— in the sense of the compensation hypothesis—as expression of functional MARTIN HEIDENREICH 121 requirements, but it reflects the fact that welfare states are engaged in education and research as well. Social security and international patents In the next stage, the relationship between social security expenditure and the results of innovation processes will be analyzed, considering the patent activities of the respective national economies. These patent activities can be detected by using three different patent categories: firstly, by the number of applications for patents (per million residents) to the European Patent Office (EPO); secondly, by the number of the patents approved by the United States Patent and Trademark Office (USPTO); and finally through the number of the patents approved in Japan, the European Union and the USA (“triadic patents”). The number of the European patents is negatively correlated with the proportion of public social security expenditure (Table 6.4, column 3). However, this is not true for triadic and US patents (columns 1 and 2). Contrary to the European experience, in the USA and the AmericanEuropean-Japanese triad, high social security expenditure does not have any negative effects on the patent applications. Even if, additionally, the amount of the educational and research investments is included, the correlation between social security expenditure and European patent activities remains significant (column 6). Such a relationship cannot be proved with the triad and US patents (columns 4 and 5). This points to a European particularity—a negative impact of welfare state activities on patent activities. In the next stage, the educational and research expenditure is divided into three components (columns 7–9): public and private educational expenditure; industry-financed R&D expenditure; and governmentfinanced R&D expenditure. In all cases, the R&D expenditure of businesses has a clear positive influence on the patent intensity in the triad, in the USA and in Europe. Public research, however, is negatively correlated with patent intensity. In the USA, the corresponding coefficient is even significant at the 5 percent level (column 8). The influence of social security expenditure on the patent level is no more significant. At first sight these results are calming: on the basis of a central output indicator, the patent applications, the expected negative effects of the efficiency hypothesis on the innovativeness of the respective national economies could not be confirmed. However, this is not applicable to the patent registrations with the EPO. If it is assumed that there is a “home bias,” and therefore a propensity to apply for patents in one’s own economic area, then this can be interpreted as a minor innovativeness of the relatively strongly developed European welfare states. 122 INNOVATION AND SOCIAL SECURITY Table 6.4 The relationship between patent applications and social security expenditure (20 OECD countries, 1980–1999) Notes See Table 6.3. 123 Genschel, P. (2003) “Globalisierung als Problem, als Lösung und als Staffage,” in Hellmann, G., Wolf, K.D. and Zürn, M. (eds) Die Neuen Internationalen Beziehungen—Forschungsstand und Perspektiven in Deutschland, Baden-Baden: Nomos. Greene, W.H. (2000) Econometric Analysis, 4th edn, London: Prentice Hall International. Hall, P.A. and Soskice, D.W. (2001) “An Introduction to Varieties of Capitalism,” in Hall, P.A. and Soskice, D.W. (eds) Varieties of Capitalism: The Institutional Foundations of Comparative Advantage, Oxford: Oxford University Press. Heidenreich, M. (1999) “Gibt es einen europäischen Weg in die Wissensgesellschaft?,” in Schmidt, G. and Trinczek, R. (eds) Globalisierung: Ökonomische und soziale Herausforderungen am Ende des Zwanzigsten Jahrhunderts, Sonderband 13 der ‘Sozialen Welt,’ Baden-Baden: Nomos. Hirst, P. and Thompson, G. (1996) Globalization in Question: The International Economy and the Possibility of Governance, Cambridge: Polity Press. Kanter, R.M. (1995) World Class: Thriving Locally in the Global Economy, New York: Simon & Schuster. Kaufmann, F.-X. (1999) “Sozialstaatlichkeit unter den Bedingungen moderner Wirtschaft,” in Korff, W. (ed.) Handbuch der Wirtschaftsethik, Gütersloh: Gütersloher Verlagshaus. Kittel, B. and Obinger, H. (2003) “Political Parties, Institutions, and the Dynamics of Social Expenditure in Times of Austerity,” Journal of European Public Policy 10(1): 20–45. Marshall, T.H. (1977) “Citizenship and Social Class,” in Marshall, T.H. (ed.) Class, Citizenship, and Social Development (with an Introduction by Seymour Martin Lipset), Chicago: University of Chicago Press. Münch, R. (2001) Offene Räume. Soziale Integration diesseits und jenseits des Nationalstaats, Frankfurt a.M.: Suhrkamp. OECD (1997) The Measurement of Scientific and Technical Activities: Proposed Guidelines for Collecting and Interpreting Technological Innovation Data (Oslo Manual), Paris: DSTI, OECD. ——(2003) OECD Science, Technology and Industry Scoreboard, Paris: OECD. Rammert, W. (1997) “Innovation im Netz: Neue Zeiten für technische Innovationen: Heterogen verteilt und interaktiv vernetzt,” Soziale Welt 48(4): 397–416. ——(2000) “Ritardando and Accelerando in Reflexive Innovation, or How Networks Synchronise the Tempi of Technological Innovation,” Technical University Technology Studies Working Papers No. WP-7–2000. Rieger, E. and Leibfried, St. (2003) Limits to Globalization, Cambridge: Polity. Rodrik, D. (2000) “Has Globalization Gone Too Far?” in Held, D. and McGrew, A. (eds) The Global Transformations Reader: An Introduction to the Globalization Debate, Cambridge: Polity Press. Schmidt, M.G. (1988) Sozialpolitik: Historische Entwicklung und internationaler Vergleich, Opladen: Leske und Budrich. Schumpeter, J. (1935) Theorie der Wirtschaftlichen Entwicklung, 4th edn, Leipzig: Duncker and Humblot. ——(1976) Capitalism, Socialism and Democracy, 5th edn, London: Allen & Unwin. Schwarze, J. and Härpfer, M. (2003) Are People Inequality Averse, and Do They Prefer Redistribution by the State? Evidence from German Longitudinal Data on Life Satisfaction, Bamberg: University of Bamberg. 130 INNOVATION AND SOCIAL SECURITY Siebert, H. (1997) “Labor Market Rigidities: At the Root of Unemployment in Europe,” Journal of Economic Perspectives 11(3): 37–54. Sinn, H.-W. (1995) “A Theory of the Welfare State,” Scandinavian Journal of Economics 97(4): 495–526. Soskice, D. (1999) “Divergent Production Regimes: Coordinated and Uncoordinated Market Economies in the 1980s and 1990s,” in Kitschelt, H., Lange, P., Marks, G. and Stephens, J.D. (eds) Continuity and Change in Contemporary Capitalism, Cambridge: Cambridge University Press. Therborn, G. (1995) European Modernity and Beyond: The Trajectory of European Societies 1945–2000, London: Sage. MARTIN HEIDENREICH 131 132 7 Transnationalization of European governance in the information age The role of policy networks David Gibbs Introduction This chapter is concerned with the role of policy networks in shifting power relationships in local, regional, national and supra-national governance, together with the impact of such networks upon the promotion of innovative policies on information and communication technologies (ICTs). At the policy-making level, the perceived importance of ICTs to the future development of localities has engendered a range of responses to try and harness the new economy for the benefit of local populations and the local economy. Thus several cities, towns and regions have developed their own policies on ICT to try and foster the development of new sectors and the take-up of ICTs by existing sectors. One response has been to encourage the development of institutional capacity within areas to develop ICT initiatives, as part of broader arguments about the potential of “associationist” development and “learning regions” (Amin and Thrift 1995; OECD 2001). It has been argued that ICTs as a technology have the power to coalesce groups, in this case as groups of actors seeking to shape the “space of flows” (Castells 1989). Increasingly urban and regional partnerships that synthesize ICTs and governance have emerged. These partnerships are attempting to shape the “space of flows” through trying to exert some form of control over exogenous and endogenous features of the informational age (Southern 2000). While much of this falls outside local control, for example in the construction of infrastructure, there are still attempts to enforce territorial control over the process (see Gillespie et al. 2001). How such control is exerted and the success or otherwise in doing so, may vary substantially from place to place. In the European Union, such developments have been encouraged and assisted by EU policy makers in their attempts to promote an EU “Information Society” (Gibbs 2001). These developments can be seen in the context of broader debates about changing governance structures. These include shifts from government to governance, the “hollowing out” of the nation state and the rise of multi-level governance. The structure of the chapter is as follows. The next section provides an overview of the development of ICT policies within the European Union and outlines the specific measures taken to try and encourage the development of a European Information Society. Following these sections, the focus shifts to an examination of debates on changing governance structures in an effort to contextualize EU policies within such debates. Evidence from the two main EU Information Society policy networks is then presented and a final section provides some conclusions from the analysis. European Union policies on ICTs: developing the information society From the 1990s onwards, the European Commission (EC) has developed a range of policies and initiatives which form a strong political drive or “policy push” to develop a European Information Society (Dai 2000a). These are embodied in a series of reports and white papers, including the Delors’ White Paper Growth, Competitiveness and Employment (CEC 1993), the Bangemann Report Europe and the Global Information Society (CEC 1994) and First Reflections of the High Level Group of Experts (CEC 1996d). They deal with the social and societal aspects of the information society and propose a range of regulatory, technological and structural initiatives to stimulate its development. Such initiatives have encouraged experimentation with ICT uptake and have focused on such developments as providing infrastructure for peripheral areas, enabling them to be more pro-active in working with core areas, as well as regional initiatives aimed at promoting an integrated approach to policies for an Information Society. The European Commission argues that new jobs and more efficient services will arise from the plethora of new technologies now coming to maturity. A European strategy on the Information Society is essential, it is argued, given estimates that half of all jobs will be in industries that are either major producers or intensive users of information technology products and services (McQuaid 2002). The European Union’s approach to the Information Society is illustrated in the following quotation: The information society is on its way. A “digital revolution” is triggering structural changes comparable to last century’s industrial revolution with the corresponding high economic stakes. The process cannot be stopped and will lead eventually to a knowledge-based economy. (CEC 1994) In consequence, the EU proposes that a crucial policy goal is to transform Europe into an Information Society. The importance attached to this is also associated with its perceived relevance in achieving other EU objectives, notably greater integration and cohesion, as well as being related to the demands of economic competitiveness—especially vis-à-vis the USA (Dai 2000b). Official EU literature leaves one in little doubt as to the inevitability of such a process: “the challenge is to manage the transition to the information society in order to 134 DAVID GIBBS optimize the number of jobs retained and created. It is not a question of whether the information society is achieved, but how, when and where it is achieved” (CEC 1995a, emphasis added). This goal and the implications for local and regional development are outlined in a number of policy documents. The Delors’ White Paper viewed the information society as a powerful engine for employment growth that would bring benefits to all regions and all parts of society. The highly influential Bangemann Report similarly viewed the development of an information society as a positive development for regions, particularly peripheral regions. Exactly how these benefits were to be achieved is not evident in the report, although considerable stress is placed upon the capacity of the free market and liberalization to deliver economic growth. The report was presented to a European Council meeting in Corfu (1994) which recommended that Bangemann’s proposals should be taken forward through designating a responsible ministerial level appointment in each member state, thus creating a Council of Ministers to direct the information society proposals.1 The next step was the publication of the EC’s action plan (1994) entitled Europe’s Way to the Information Society, which summarized information society initiatives already under way, or planned, by the Commission. In this action plan the major emphasis was placed upon developing a new EU-wide regulatory framework, with limited mention of regional imbalances. The regional consequences of the information society were largely left to the operation of the EU’s regional development funds, with a greater emphasis upon the information society in the Community Support Frameworks and in the Single Programming Documents agreed for Objective 1 regions for 1994–1999.2 The Commission launched a study of the impacts and benefits of the information society for regional, economic and social cohesion. At this stage, only pilot projects and specific networks were envisaged in co-operation with cities and regions. In 1994, EU policy was further developed at a European Council meeting in Essen where it was proposed that the information society would bring new jobs, stronger regional cohesion and facilitate greater solidarity between rich and poor regions. One concrete development from the meeting was the inauguration of a Forum on the Information Society and the installation of a group of High Level Experts to deal with social and societal issues. This latter group produced the Working Document on the Social and Societal Aspects of the Information Society in 1995. Six key themes were identified as the major agenda items for the Commission in the short to medium term: the quantity and quality of work; the way work is organized; the efficiency of the labor market; education and training; upgrading systems of healthcare; and improving regional and urban cohe sion. The document stated that economic cohesion could be enhanced through the use of electronic networks by: • Improving access to markets and sources of information; • Bringing work to areas of high unemployment; • Increasing the competitiveness of businesses at a regional level; THE ROLE OF POLICY NETWORKS 135 • Reinforcing positive externalities through shared infrastructures; • Enhancing social cohesion through ICT usage; • Reducing exclusion by making health, education, the arts and cultural services more accessible; • Using the power of ICTs to provide on-line forms of these services to all groups in society. This group pointed out that ICT usage may not automatically lead to enhanced cohesion and that they could lead to greater centralization of jobs and services and to a “fast” and “slow track” information society. The questions asked by the group in relation to spatial development were: • What are the regional and urban development consequences of disparities in infrastructure provision and service levels? What can be done to assist less wellequipped regions and areas? • How do different levels of provision affect social cohesion at both regional and urban levels? What can be done to increase the accessibility of these new essential services? • How can regions and towns be helped to optimize their use of the new possibilities of ICT services? • How can ICTs be used to improve the quality of life in the towns and regions of Europe, by allowing innovation in urban government and greater accessibility of services? Preliminary answers to these questions were given in the Interim Report of the High Level Group of Experts Building the European Information Society for Us All: First Reflections of the High Level Group of Experts (CEC 1996d). This outlined both the threats and opportunities that ICTs present for peripheral regions and suggested a number of policy measures. They stressed the need for an integrated EU approach to the information society in order to increase social cohesion, as well as the need for a more focused approach to infrastructural support and to secure access to a reasonable service level at a reasonable price for different social groups. The group proposed that social policy merited at least equal consideration with economic policy in formulating a European approach to the information society. A subsequent development was the establishment of the Information Society Forum by the EU which produced its first report in 1996 which encouraged the Commission to launch EU-wide awareness raising initiatives, promote ICT usage and access and explore the social and economic implications of the Information Society (CEC 1996c). In relation to the latter point, in 1996 the Commission adopted a green paper on Living and Working in the Information Society: People First. The paper’s purpose was to deepen the dialogue on such social and societal aspects, with an aim to implementing specific proposals in 1997. 136 DAVID GIBBS Further measures “to place Europe at the forefront of the Information Society” are outlined in CEC (1996b): • Liberalization of Europe’s telecommunications market, which was intended to lead to the development of new services and promote the uptake of ICTs; • Supporting regional Information Society initiatives to enhance cohesion and to take social and societal aspects of the Information Society into account; • Recognition of the role of education and training through the development of a Europe-wide learning action plan (CEC 1996e); • Supporting the development of a strong EU multimedia content industry, which was expected to create 1 million new jobs in the next 10 years3; • Incorporating a specific ICT element into the EU’s Fourth Framework Program for Research and Technological Development, which has been continued within the Fifth and the new Sixth Framework Program; • Playing a role in developing and shaping the international context of the information society. Finally, the European Commission launched an eEurope initiative intended to ensure that European economies exploit the opportunities offered by new technologies and do not lose out to competition, particularly from the USA (CEC 2000). This has three main objectives, with a number of sub-measures within each (see Table 7.1). EU policy initiatives have begun to directly emphasize the spatial consequences of the Information Society program. As the CEC (1996a:8) states: “the relevant disadvantage of peripheral regions […] can be reduced through direct connection […] However, the participation in the information society of all European regions will not happen automatically, but requires active policy measures” (emphasis in the original). These policy measures are envisaged as involving local, regional and national initiatives, not only to attract infrastructure investments but also to stimulate the demand for new information services. Emphasis is placed on the transfer of experiences and best practice already developed by advanced regions (e.g. in strategy development, in the use of “telecottages” and in teleworking). Meanwhile, rural areas and less-favored regions are encouraged to facilitate private-public or cross-regional partnerships, elaborate regional strategies and launch pilot applications (CEC 1996a). The document proposes that future policies will build on those local initiatives that have already been supported and to integrate policy initiatives on the Information Society with Structural Funds policies. The more recent eEurope Action Plan, An Information Society for All, similarly stresses the Structural Funds’ role in improving ICT infrastructure (CEC 2000). THE ROLE OF POLICY NETWORKS 137 Governance, policy networks and EU policy A particular feature of the EU’s Information Society policy has been the emergence of networks of cities and regions to link together locally based partnerships. Such policy networks have become an increasingly important approach to the development of EU policy (Bennington and Harvey 1998). These policy networks can be defined as “a (more or less) structured cluster of public and private actors who are stakeholders in a specific sector of policy and possess resources which allow them to affect policy outcomes” (Peterson 1997:7). The growth of policy networks is part of a broader shift from government to governance or from a more linear, statedominated political system to one which involves non-hierarchical, multilevel governance by a mix of public and private actors (Jordan 2001). Thus, “political power and institutional capability is less and less derived from formal constitutional powers accorded the state but more from a capacity to wield and co-ordinate resources from public and private actors and interests” (Peters and Pierre 2001:131). This shift has led to a much greater recognition of the roles played by supra-national and sub-national state and nonstate actors and the complex interactions between them. Institutional relationships need not be vertical and “do not have to operate through intermediary levels but can take place between, say, transnational and regional levels, thus bypassing the state level” (Peters and Pierre 2001:132). These changes have been recognized as reflecting new “geographies of governance” as state functions are redistributed vertically, both upwards to supranational organizations and downwards to cities and regions, as well as horizontally to non-state actors (MacLeod and Goodwin 1999). This does not, however, necessarily reflect the decline of the nation state, rather that “the continuous reshuffling and reorganizations of spatial scales are an integral part of social strategies and struggles for control and empowerment” (Swyngedouw Table 7.1 Objectives of the eEurope Initiative Action Plan Source: CEC 2000. 138 DAVID GIBBS 2000:70), where the nation state continues to play a key role. However, the tendency for national governments to become less interventionist in some spheres has encouraged sub-national authorities to develop their own strategies to deal with adjustment problems consequent upon new economic circumstances (OECD 2001). There is thus a complex interrelationship between changing governance structures and the development of new economic forms based upon new technological developments. Theoretically, “the workings of policy networks are critical for promoting the diffusion of policy innovations” (Mintrom and Vergari 1998:128). Such networks may act to promote policy learning in the regions, though this depends on the structure of the network and the patterns of interaction (Benz and Fürst 2002). Such networks are a key means of exercising power by the EU (through the European Commission), but where policy making is achieved through sharing power with a range of actors from European, national and local levels. As the recent EU White Paper on governance states “the Union cannot develop and deliver policy in the same way as a national government; it must build partnerships and rely on a wide variety of actors” (CEC 2001b:32). Much EU decision making now takes place within policy networks. In describing patterns of multi-level governance in the Europeanization of regional policies, Benz and Eberlein (1999) argue that EU policy making is achieved through power sharing between different levels of government and that policy networks are formed for collaboration based on various combinations of government authorities. Such relationships are not simply at the level of local-European alliances, signifying the “hollowing out” of the nation state, but rather reflect a shifting mix of national-sub-national alliances, as well as EU-local alliances (Jessop 1994). The EC promotes such networking activity as a means of building coalitions in favor of its own renditions of policy, thus increasing its leverage with the Council of Ministers and the European Parliament (Richardson 1996). The EC has recently argued that networking is also an essential component of legitimating European governance, incorporating local and regional knowledge and conditions and showing “policies in action” (CEC 2001b:18). The EC expects such policy networks to develop policy proposals and exchange experience, thus mobi lizing the regions as potential allies of the Commission in that they seek further funding and/or an extension of European policies (Tömmel 1998). Such developments though are far from a one way process. Regions and urban areas (or at least those fractions that claim to represent or speak for local areas) use such networks as a means of seeking greater autonomy and a voice in EU policy making, sometimes as a means of circumventing nation state control. While Lloyd and Meegan (1996:78) point out that “the Commission and its advisers are a key source of new scripts about the regional development process —both in the sense of how to and in terms of their views on the most appropriate institutional arrangements,” regional authorities are also able to rewrite such scripts for their own purposes. Thus, the rise of multi-level governance may THE ROLE OF POLICY NETWORKS 139 interpreted as an outcome of the new economy crisis. The collapse in 2000 also hit the Berlin multimedia firms, particularly the Internet business. The failures of firms and employees’ dismissive attitude in this sub-sector were driving forces for a large number of discharged multimedia specialists to immediately establish new start-up firms. They hoped to remain in the sector and to gain success with new business concepts or to get an acquisition offer from the strong established companies in the new media sector. However, the Berlin metropolitan region remains a strong and dynamic locational center of the media industry. In 2000, the regional chamber of commerce published a survey among the resident firms in the sub-sectors of the media and information and communication industries to evaluate the city’s particular strengths and weaknesses (IHK 2000). According to this, the media industry firms regard the availability of highly qualified creative employees, the local media culture in terms of the local clustering of many specialist firms and institutions and the city’s cultural attractiveness as the most important strength of the region. Meanwhile, the urban government’s political support for the media sector got the poorest judgment. The resident ICT firms emphasize Berlin’s R&D capacities as the region’s most important strength, as well as the availability of highly qualified specialist personnel, the local clustering of many specialist firms and service providers and the high quality of the telecommunication infrastructure. The Berlin metropolitan region (including Potsdam/Babelsberg) holds a leading position in the particular field of post-production activities of the film industry. In the year 2000, of the recorded 211 specialist firms for digital image processing and computer animation in the German film industry 25 percent were located in the Berlin metropolitan region, whereas Munich and Hamburg held a share of 13 percent respectively, Frankfurt-Main 9 percent and Cologne 7 percent (Krätke and Scheuplein 2001). The digital post-production and computer animation activities represent a field of intersection between the media and information technology industries (including software development). These activities reveal a particularly strong dynamic in those locational centers which have a clustering in both sub-sectors (which applies not only to Berlin, but also to Los Angeles etc.). Regarding the spatial organization of Berlin’s media industry, the most important characteristic is the formation of local enterprise clusters within the inner urban area of the metropolis (Krätke 2002a). This is particularly well illustrated in the multimedia sector. The Berlin media industry contains a cluster of multimedia firms of considerable size (with 424 establishments in 2000 and more than 800 recorded in 2001). Today, many well known multimedia agencies like Pixelpark, Aperto, MetaDesign and others are located in Berlin and the startup activities are continuously strong. Against this background, Berlin is now the second largest location of Internet firms in Germany after Munich. The knowledge- and technology-intensive multimedia firms are regarded as an important sub-sector of the new economy. The recent crisis has particularly hit 242 STEFAN KRÄTKE the multimedia sector and the Internet firms. Thus, the former growth euphoria has vanished and is being replaced by the expectation that a certain number of the currently existing firms might not be able to maintain their hold on the market. As a result, there is still a high fluctuation of start-ups and failures. Within the urban spatial fabric of Berlin, the multimedia firms are concentrated in a number of local clusters, most of which are located in the densely built and mixed use inner urban area. The largest concentration can be found in the eastern city center and extends to the central part of the adjacent inner urban district of Prenzlauer Berg as well as to the newly built Potsdamer Platz area. Within the “City East,” multimedia firms are clustered for example in the ChausseeStreet (which is popularly referred to as “silicon alley”), where the European New Media Center is being built. In the outer urban area we find a local clustering of multimedia firms within the Adlershof “technology and research district.” The so-called Media City Adlershof contains 126 media firms (in 1998) which can set up links with Figure 11.2 Local clustering of the media industry in Berlin (2000): multimedia firms (producers and service providers). DEVELOPMENT CHANCES AND THREATS 243 the many specialized firms of the ICT sector located in Adlershof. Another local concentration of digital media firms outside the inner urban area of Berlin is the Media City Potsdam/Babelsberg, located on the fringes of the metropolis. It is important to note that most local clusters of the multimedia branch in Berlin coincide with local concentration areas of other media industry subsectors, particularly the film industry. The fact that Berlin has a number of common local concentration areas of firms from different media industry sub-sectors (Krätke 2002a) might indicate that the actors involved share common locational preferences. These include the high demand for a dense local environment of differently specialized media industry firms, which is being appreciated in terms of a creativity-boosting atmosphere regardless of the actual existence of interfirm transactional relations. The spatial concentration of the media industry within the urban area of Berlin is related to several influencing factors. Most of the newly established media firms prefer to locate themselves in the inner urban districts in which living and working environments merge with leisure-time culture. For corporate operators and employees in the media industry the local connection between working, living and leisure time activities is an attraction factor that is in harmony with their lifestyle. These people deliberately seek out locations in a sub-cultural urban district that they can use as an extended stage for self-portrayal during working hours and in their leisure time. In the local media clusters there is thus a direct link between certain lifestyle forms and urban organization forms of production space and thus a clear overlapping of the geographies of production and consumption. Particularly in the inner urban districts (like Prenzlauer Berg) the actors of the media scene are contributing to a process of revaluation and gentrification of inner urban localities (as pioneers or real gentrifiers). However, the new economy and media industry establishments might take a location anywhere in the urban spatial fabric, wherever they can find premises for rent equipped with electricity, phone and Internet connections. Nonetheless, the majority of these firms prefer “sexy” inner city locations. First, the actors of the media scene regard these areas as the best urban environment for a particular lifestyle which consciously combines working and leisure time activities locally within culturally attractive districts. Second, there is no scarcity of premises for rent in Berlin’s inner urban districts. Third, in the inner urban districts the firms can find a supply of specialized producer services within short distance. Fourth, the close proximity of other enterprises active in the same industry is a strong pull factor of the existent local media clusters. In some local areas the actors might find the whole value chain of a particular new economy branch being represented by firms within one single building. Additionally, the local cluster formations might contribute to an intensification of communication links between the firms (without a need for formal cooperation agreements) and thus create a space of opportunities which is welcomed particularly by the start-up firms that are facing many uncertainties. 244 STEFAN KRÄTKE Thus, the striving for risk reduction is also a quite important factor of local clustering. Despite these inner urban location factors the overall concentration of culture and media industry firms in Berlin is an attracting factor with regard to the business opportunities being offered by a metropolitan cluster formation. The institutional infrastructures and the knowledge pool of a metropolitan cluster have to be mentioned as well. All these points underline that the media firms of different sub-sectors in Berlin can profit from socially produced locational advantages in the same way as in other large centers of this industry (Krätke 2002a). Moreover, the city as a whole can become an attracting factor in that the symbolic quality of the specific location is either incorporated into the products of the culture and media industry or the origin of these products itself becomes a mark of quality (Scott 1997). Hence, production locations such as New York, Paris and Berlin are perceived in the sphere of culture and the media as being brand names that draw attention to the attractive social and cultural qualities of the cities concerned. This includes, in particular, the perception of Berlin as a social space in which there is a special variety of different social and cultural milieus. With regard to the content and design of their products, cultural production and media industry companies have to contend with rapidly changing trends. For that reason there is a desire to be close to the source of new trends, i.e. the sub-cultures that develop in certain metropolises, such as New York, Paris and Berlin. Cities of this kind are perceived as constituting a social and cultural potential marked by great openness. This in turn enhances their attractiveness for creative talents and makes them a source of inspiration for cultural producers. A marked social and cultural variety and openness, therefore, represents a specific cultural capital for a city, which is highly attractive for certain economic activities and players. This form of a city’s cultural capital is concentrated to a special extent in large metropolises and it acts as an additional agglomerative factor in the (selective) concentration of the culture and media industries in the urban system. Berlin’s position among Germany’s media cities The German regional system contains a number of competing media cities in terms of large regional production clusters in the culture and media industries which are selectively concentrated in the large metropolitan regions (Krätke 2002a). Thus, the urban media clusters are being involved in the locational competition of metropolitan regions on the national as well as on the international scale. However, the media centers reveal a differing mix of their most important media industry sub-sectors. The leading German media cities are characterized by an absolute concentration of firms and employees in all subsectors of the media industry, i.e. they can make use of cross-sectoral impulses. The most important media cities in Germany might be identified by selecting the DEVELOPMENT CHANCES AND THREATS 245 regions belonging to the ten largest centers by the number of employees in different sub-sectors of the culture and media industry. According to this criterion, the German regional system’s culture and media industry has five prime centers which are among the absolutely largest centers in all different sub-sectors. The prime centers are Berlin, Munich, Hamburg, Cologne and Frankfurt-Main. There are also some media cities which belong to the largest centers in only two media industry sub-sectors such as Düsseldorf and Stuttgart with a strong position in the sub-sectors of advertising and publishing. Thus, in a comparative perspective, Berlin is a prime center of the German media industry. However, it has strong competitors not only on the international scale, but also within the German regional system. Note that the large firms in the media industry frequently have located their establishments in various important regional media centers. This allows them to make use of all the particular regional production and innovation capacities and to profit from the media cities’ locational competition. In this way the various regional clusters of the German media industry are being interlinked in terms of the organizational networks of large media firms which strive to be present in all of the large media centers and take part in the respective local media clusters’ inter-firm transaction and communication networks. Berlin’s position as a “global media city” Like other sub-sectors of the new economy, the media industry is a prime mover for globalization processes in the urban system, in which media industry clusters act as local nodes in the global networks of the large media groups. New communication technologies and the emergence of large multinational groups within the culture and media industries contribute to a global flow of cultural forms and products (Held et al. 1999). Since the 1970s, the liberalization and deregulation of the telecommunications and media industries have led to the globalization of the corporate media sector. However, this situation cannot be taken as indicative of a trend toward the global homogenization of cultural consumption. The global companies in the culture and media industries are obliged to take account of specific tastes and cultural preferences in other countries and regions. The producers and market strategists employed by global media firms are well aware of the cultural variety and differentiation of their global audiences, customers and consumers and have long given their products and programs a regional touch with a view to stabilize or enhance their global market success. In other words, they have adapted their products and programs to specific regional or national tastes and cultural preferences. A good example of this is the successful global TV music broadcaster, MTV, which does not simply reproduce the content of the MTV program in the USA for consumption in all the other regions of the world, but has set up a series of regional MTV channels (including several for the European region alone), which produce and vary the program in 246 STEFAN KRÄTKE accordance with regional tastes and preferences. At the same time, this trend toward cultural market differentiation is a driving force for the organization of global production networks in the culture and media industries with local anchoring points in different regions and nation-states. Figure 11.3 Media cities in Germany. DEVELOPMENT CHANCES AND THREATS 247 The media industry is not only characterized by a growing concentration on a world scale and the geographical concentration and formation of clusters in a series of metropolitan regions and global cities, but also by a marked trend toward the globalization of corporate organization (Robins 1995; Pratt 2000). The huge media groups not only occupy a prominent position in the culture industry of individual countries, but are also creating an increasingly global network of branch offices and subsidiaries. This global network of establishments linked under the roof of a media group has its local anchoring points in the leading culture and media industry centers of the worldwide urban system. First, it should be emphasized, however, that the globalization strategy pursued by media companies is primarily geared toward market development and the extension of market penetration through the establishment of a presence in the major international centers of the media industry. Second, the globalization strategy of media companies reveals a strong orientation toward enhancing their innovative capacity for the simple reason that a presence in the leading centers of cultural production gives the global media groups the chance to incorporate the cultural industry’s latest fashion trends as quickly as possible through their integration into the respective local clusters of the regional economy. At the same time, it allows them to exploit the latest technological developments in the media sector (e.g. in digital image processing, special effects, transmission technologies and Internet applications). The global players in the culture industry network at the local level corporate with the small specialized producers and service providers. They thereby establish a global network of their branch offices and subsidiary companies, which links the internationally distributed urban centers of cultural production with one another. The global media groups organize the worldwide marketing of their respective cultural products and thus contribute to the global spread of media content and formats, which are generated in the production centers of the global culture industry, in particular in Los Angeles, New York, Paris, London, Munich and Berlin. Within processes of globalization, the globally operating media firms are at least as influential as the global providers of corporate services, because they create a cultural market of global dimensions on the basis of which the specialized global service providers can ensure the practical management of global production and market networks. Most of the studies on global cities and the international urban system reveal a tendency to reduce the high-ranking world cities to their function as financial centers and centers providing specialized corporate services (Sassen 1991). In other words, they underestimate the role these cities play as locations for industrial production and as centers of the generation of knowledge and industrial innovation. The debate on new industrial spaces and technology districts, in particular, contains numerous references to the major role that global city regions continue to play in industrial development (Scott 1998; Schamp 2000). In many cases they provide central locations for new knowledge-based production chains and for highly innovative production clusters in the fields of ICT, medical 248 STEFAN KRÄTKE engineering, biotechnology, the media industry, etc. These urban innovation centers for new and old industries are often characterized by extensive networking within metropolitan areas and by strong supra-regional connections with innovation centers in other countries’ metropolitan regions. Thus, the relations between innovative companies in the high-technology clusters of Munich and San Francisco (Silicon Valley) contribute to the worldwide networking of industrial innovation processes. The economic and functional changes in position of the global cities are today determined by processes of selective concentration of global service capacities as well as innovative industrial capacities. In their recent joint publication on “Global City Regions,” Scott et al. emphasize the diversity of economic sectors which tend toward agglomeration in the global city regions and which together transform these centers into “regional motors of the world economy.” Thus, it is important to emphasize globalization’s geographical complexity which cannot be reduced to a simple hierarchy of global cities with global service providers. An analysis of the location networks of global media groups makes it possible to trace the global structure of media cities as an interlocking network of internationally operating media firms in the same way as global city research has done with regard to advanced corporate services (Beaverstock et al. 1999; Taylor and Hoyler 2000; Taylor and Walker 2001). Global media firms have established a network of branch offices, subsidiaries and holding companies that are distributed worldwide across a series of cities. The creation of branch offices or subsidiary companies around the world allows for quicker and more direct linkages with potential clients in the respective region, particularly in those places where new media services are entering a phase of expansion. The production cluster of the film industry in Potsdam/Babelsberg on the outskirts of Berlin represents an example that illustrates that cluster firms are not only closely networked on a local scale, but are also integrated into the supraregional location networks of global media companies. In the case of Babelsberg, the local cluster firms are directly linked with the resident establishments of global media firms in Paris, London and New York and indirectly linked with other global players in the film and TV industry thanks to their business relations with the media firms that are resident in Berlin. In other words, the global players in the culture industry network locally with the small specialist producers and service providers, while at the same time they run a global network of branch offices and subsidiaries that permits global linking of the urban centers of cultural production. Starting from this basic concept of a multi-scale interlocking network, an empirical study was conducted on the business location networks of 33 global media firms with a total of 2,766 enterprise units (Krätke 2002a). To qualify as global, a media company had to have a presence in at least three different national economic areas and at least two continents or world regions (USA/ Canada/Latin America; Europe; Asia/Australia; Africa). The geographical organization of global media firms should not be interpreted as a star-shaped DEVELOPMENT CHANCES AND THREATS 249 hierarchical structure, but rather as a network of business units (acting on their own responsibility), between which there are many different information and communication flows that enable special regional or local impulses and customer requirements to be picked up and processed in a flexible manner on a global scale. In a global media city there is a partial overlapping between the location networks of several global media firms in the urban economic space. Here, the local and the global firms in the media industry are linked in a joint development context that can foster the formation of an urban media cluster, whose international business relations are handled primarily via the present global media firms. The result of this analysis is a ranking of cities based on the number of enterprise units (establishments) of global media firms that are located in the respective metropolitan region. The media industry’s world cities were distinguished by means of two criteria and divided into three groups: alpha, beta and gamma world media cities.3 The analysis illustrates the markedly unequal distribution of the global media firms’ establishments over a small number of cities: over 50 percent of the branch offices and subsidiary companies of global media groups are concentrated in just 22 centers of the worldwide urban system. The organizational units of the globalized media industry reveal a highly selective locational concentration on a global scale. The network of global media cities as a whole is a reflection of the locational system run by the westernstyle media industry, which is concentrated mainly in North America and Europe. Prominent among the alpha world media cities are New York, London, Paris and Los Angeles, which are ranked as genuine global cities in virtually every analysis of the global urban system. They are also designated here as world cities in the media industry. However, one needs to be aware that there are interesting deviations from the widely employed global city system. Based on this analysis, the alpha group of global media cities also includes Munich, Berlin and Amsterdam. In global city research, which focuses on corporate services, these cities are ranked as (third-rate) gamma world cities (Beaverstock et al. 1999). In the system of global media cities, by contrast, Munich, Berlin and Amsterdam are included in the top group. Of the 33 global media groups included in the analysis between 18 and 29 are to be found in the alpha group cities, the most located in London (with a presence of 29 global firms and a total of 180 establishments), followed by Paris (26/129) and Los Angeles (25/111). Of course, New York (22/185) also assumes an outstanding position. Additionally, Munich (with 20 global firms and a total of 96 establishments), Berlin (19/70) and Amsterdam (18/64) have achieved a degree of integration into the location networks of global media firms that qualifies them as internationally outstanding centers in the contemporary media industry. Berlin, in particular, could perceive the message spelled out by this analysis as a reason to make the media industry core area of its economic development policy and to further enhance the local creative potential of the culture and media 250 STEFAN KRÄTKE sector as a world media city’s important factor of attraction. The analysis indicates that Berlin is a city that is now (once again) gaining worldwide importance, especially in cultural production and the media industry. Berlin’s international reputation as a cultural metropolis was an important urban development factor in the 1920s. This reputation can now be restored on the basis of current economic developments. However, while Berlin still does not rank as a global city defined as an economic center with global control capacities and as a center of strategic corporate services (Krätke and Borst 2000), it is a first-rank global media city with a worldwide significance and impact. Figure 11.4 Global media cities in Europe: classification based on local concentrations of global media firms’ establishments in European cities 2001. DEVELOPMENT CHANCES AND THREATS 251 ——(ed.) (2001b) “Globalization and the Rise of City-Regions,” European Planning Studies 9(7): 813–826. Scott, A.J., Agnew, J., Soja, E. and Storper, M. (2001) “Global City-Regions,” in Scott, A.J. (ed.) Global City Regions: Trends, Theory, Policy, Oxford: Oxford University Press. Senatsverwaltung für Stadtentwicklung (1997) Gewerbeflächenentwicklung Berlin, Berlin: Senatsverwaltung für Stadtentwicklung. ——(1999) Innovations- und Technologiekonzept, Berlin: Senatsverwaltung für Stadtentwicklung. Senatsverwaltung für Wirtschaft und Betriebe (1997) Wirtschaftsbericht Berlin 1997, Berlin: Senatsverwaltung für Wirtschaft und Betriebe. ——(1998) Wirtschaftsbericht Berlin 1998, Berlin: Senatsverwaltung für Wirtschaft und Betriebe. Storper, M. (1997) The Regional World: Territorial Development in a Global Economy, New York: Guilford Press. Storper, M. and Scott, A.J. (1995) “The Wealth of Regions: Market Forces and Policy Imperatives in Local and Global Context,” Futures 27(5): 505–526. Taylor, P.J. and Hoyler, M. (2000) “The Spatial Order of European Cities under Conditions of Contemporary Globalization,” Tijdschrift voor Economische en Sociale Geografie 91(2): 176–189. Taylor, P.J. and Walker, D.R. (2001) “World Cities: A First Multivariate Analysis of Their Service Complexes,” Urban Studies 38(1): 23–47. 258 STEFAN KRÄTKE 259 Index absorptive capacity 190 agglomeration economies 189 alpha cities 250–9 Alpha World Media Cities 250 alpine SSIP 70 anarchy of the market place 144 Anglo-Saxon economies 65–4 application service providers (ASPs) 31 archipelago economy 169, 183 Archipelago Europe 171, 175 Berlin Technology Transfer Agency 236 beta cities 250 blue banana 171–9; Brunet’s blue banana 172 Breusch-Pagan Test 114–13 bubble 3–4, 7, 11, 21, 59, 217–16, 226, 240 business R&D (BERD) 89, 95 business-oriented services 193–3 Californian economy 209 capital deepening 45–6, 59 capital-skill complementarity 32, 36 catch-up 4 central processing units 30 cluster analysis 78–8 clustering 13, 78, 157, 176, 188, 197, 231, 235, 241–42, 255 Cobb-Douglas specification 52 competition state 176 coordinated market economies (CME) 74, 84, 217–16 cross-sectoral impulses 245 culturalization of the economy 238 Delors’ White Paper 134 digital divide 4 diversity of capitalism 65, 69, 74–3, 83 division of innovative labor 188–91, 195, 198, 200 division of innovative labor 189–9, 192 Durbin-Watson Test 115 EC’s action plan 134 e-commerce companies 207 eEurope Action Plan 137, 146 eEurope initiative 136 efficiency and compensation hypotheses 111 electronic commerce 148, 205–7, 216, 220, 223, 238–7 employability-centered strategy 110 entrepreneurship 199 European Cities’ Dialogue Initiative 251 European Information Technology Observatory (EITO) 19–1 European integration SSIP 70 European urban regions 171 European urban systems 169 European welfare states 108 externalities 69 Fordist accumulation 161–9 Fordist cities 171 Fordist-Keynesian system 176–4 gamma cities 250 General Purpose Technologies (GPT) 6, 9, 28–8 global cities 171 260 Global City Regions 249 global media groups 248–8 global network 129, 246–7 globalizing learning economy 88–6 glocalization strategies 169–7, 176–8 glocalizing competition state regimes (GCSRs) 157, 176–5 governance of cyberspace 145 great crises 8–9 Green Paper on Living and Working in the Information Society: People First 136 greencard in Germany 33, 36 Hausman Test 114–13 High Level Group of Experts 133–31 high-price-high-performance 29 high-technology districts 211 ICBT capital 45–6, 52–4 ICBT capital elasticity 45 ICT expenditure 95–3 image production activities 238–7 information and network society 40 information, communication and biotechnology sectors (ICBT) 44–60 innovation culture 198–9 innovation losers 111 innovational complementarities 6 innovative areas of technology 235 innovator’s dilemma 29 institutional capacity 131 institutional complementarity 65–7 inter-firm network 237 international patents 109, 121 internationally co-authored paper 101–9 internet-based retailer 219 internet-related business models 214 interscalar coordination 183 intersectoral coordination 182 inventory cycle 34–4 investment in construction goods cycle 34 investment in equipment goods cycle 34 IPO 220–19 irrational exuberance 8 Juglar cycle 34 Keynesian welfare state 157 knowledge capital 223 knowledge investments 118–18 knowledge spillovers 195–8 knowledge-based production chains 248 knowledge-based services 109, 115, 124, 126 knowledge-intensive industries 95, 126, 212, 232–2, 238–7 labor productivity 45–6, 52–2 lagging regions 163 landscapes of regulation 180–9 learning economy 88–6, 105, 194 learning regions 131 liberal market economies (LME) 68–8, 74 life sciences 121 life-long employment guarantees 109 linear models of innovation 205 Lisbon Summit 5, 19, 88–6, local governments 40, 144, 147, 164–3, 173–1 local nodes 246 local state 162–60, 174–4, 182 lock-in 198 long wave cycles 34 low-price-low-performance 29 manufacturing establishments 193 Marginalized cities 171 market-based finance 82 media city 234, 242, 245, 250, 251; global media city 246 media industry cluster 234–4, 245, 251–2 Mediterranean SSIP 70 melancholy state 36 metropolitan region 229–31, 241–40, 245, 248–7 metropolitanization 169 mismatch 29, 33–3 mode of regulation xii, 8 Moore’s Law 23, 29, 36 multi-level governance 131–9, 137 multi-media firms 241 NASDAQ 52 national spatial planning systems 174 261INDEX National urban centers 171 neo-corporatist alliance 167 network technologies 215 new economy accumulation strategies 156– 4, 161, 177–6, 182–1 new growth model 155 new regionalism 178–7 non-accelerating inflation rate of unemployment (NAIRU) 39 non-linear models of innovation 207 North Atlantic Fordism 169 one-man-firms 213 patents 91–92 physical capital 45–6, 52–2 physical capital elasticity 45 policy networks 13, 131–9, 137–5, 145–4 political-economic governance 177–6 positive externalities 135 post-Fordist cities 171 process of creative destruction 113 productivity paradox 47, 54 product-skill complementarity 32 public SSIP 70 R&D expenditures 89–8, 97, 102–1 redistribution of resources 111 regime of accumulation 8–11 regional endowment 198–8 regional Information Society Initiative 136, 138; regional Information Society Initiative/ eris@ 139 Regional Information Society Unit (RISU) 139–6 regional innovation systems 188–99 regional production clusters 229, 245 research-intensive industries 113 self-employed freelancer 253 semiconductor sector 23–5, 29, 39; computer semiconductor industry 25; semiconductor and computer sectors 29 Silicon Valley 211 Sixth Framework Program 102–1 small crises 8 social capabilities 4 social cohesion 135 social ecology 212 social openness 210 social rights 110 social security 108–16,121–24 social spatial exclusion 183 Social System of Innovation and Production (SSIP) 69–9, 74; market based SSIP 70; social-democratic SSIP 70 Social System of Production, Consumption and Regulation 9–10 Solow productivity paradox 10,46 space of flows 131 spaces of innovation 11–12 spatial division of labor 158 spatial Keynesianism 161–60 spillover effects 28 start-up firms 205,213,216 start-ups 216 stock market volatility 58 STRATA 7 strategic technology agreements 96–5 super cluster 231 supra-regional linkages 236 tacit knowledge 190 technological obsolescence 30, 35 technological paradigm 66–6 Technology Foundation Innovation Centre Berlin 236 technology gap 88–7 TeleCities Network 139–8, 144–3 telecottages 136–3 territorial barriers 108 territorial inequality 183 territorial restructuring 158 theory of growth cycles 40 total factor productivity 25, 40, 44–6, 51–3, 59 transnational policy networks 148 UMTS 21–2, 39 uncertainty 6; zone of uncertainty 9–10 INDEX 262 uneven geographical development 159 urban location factors 244 usage hypothesis 4 venture capital 215 venture capital firms 212–13, 225 video-on-demand 214, 222; internet based video-on-demand 222 weightless economy 65, 84 Working Document on the Social and Societal Aspects of the Information Society 134 Y2K bug 35 263INDEX