European Science and Technology Policy: Towards Integration or Fragmentation? 1848443307, 9781848443303

Table of contents :
Contents......Page 6
Contributors......Page 8
Acknowledgements......Page 11
Abbreviations and acronyms......Page 12
Introduction......Page 18
PART 1 : Historical perspective......Page 22
1. Post-war research, education and......Page 24
2. Intergovernmental cooperation in......Page 45
3. A European Research Area built by......Page 65
4. The 'European Research Area'......Page 85
5. Analysing Community policies......Page 99
PART 2 : Theoretical issues......Page 120
6. From the Lisbon Agenda to the......Page 122
7. The returns to public research......Page 148
8. Scale and scope in research......Page 163
9. ERA and the role of networks......Page 181
10. Transnational collaboration......Page 196
PART 3 : Achievements......Page 212
11. The EU R&D under-investment:......Page 214
12. Does the 'European Paradox' still......Page 235
13. The European Research Area and......Page 258
14. Performance of European science:......Page 277
15. The levelling off of the integration......Page 312
16. The European Research Area as......Page 333
PART 4 : Conclusions......Page 350
17. The future of the European......Page 352
Conclusion and perspectives......Page 374
Index......Page 378

Citation preview

European Science and Technology Policy

European Science and Technology Policy Towards Integration or Fragmentation?

Edited by

Henri Delanghe PhD, Expert, IDEA Consult, Belgium

Ugur Muldur Plz D, European Commission, Belgium

Luc Soete PhD, Director UNU-MERIT and Professor of International Economics, Maastricht University, the Netherlands

Edward Elgar Cheltenham, UK Northampton, MA, USA

% Henri Delanghe, Ugur Muldur and Luc Soete 2009

All rights reserved. No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic. mechanical or photocopying, recording, or otherwise without the prior permission of the publisher. Published by Edward Elgar Publishing Limited The Lypiatts 15 Lansdown Road Cheltenham GIos GL50 2JA UK Edward Elgar Publishing, Inc. William Pratt House 9 Dewey Court Northampton Massachusetts 01060 USA

A catalogue record for this book is available from the British Library Library of Congress Control Number: 2009936761

Mixed Sources

FSC

Produn group from well-managed f o r e m and o t h a conbollcd sovrcer w . f s r o r g CeR no. SA-COC.1565 0 19% Mrcrt stewardshipCouncil

ISBN 978 1 84844 330 3 Printed and bound by MPG Books Group, U K

Contents List of' contributors Acknowledgements List qf abbreviations and acronyms Introduction

vii X

xi xvii

PART 1 HISTORICAL PERSPECTIVE Post-war research, education and innovation policy-making in Europe Peter Tindemans Intergovernmental cooperation in the making of European research Pierre Pupon

A European Research Area built by the Member States? Christian Svunfeldt The 'European Research Area' idea in the history of Community policy-making LUCUGuzzetti Analysing Community policies Jim Drutw~a

3

24 44

64

78

PART 2 THEORETICAL ISSUES From the Lisbon Agenda to the Lisbon Treaty: national research systems in the context of European integration and globalization Robert Boyer

101

The returns to public research funding Kris Aerts and Dirk Czurnitzki

127

Scale and scope in research Nicholas S. Vonortas

142

vi

European science und techno lop^. polic!

9 ERA and the role of networks Stefano Breschi and Franco Malerbu 10 Transnational collaboration in public research funding and publicly supported research in Europe Henri Delanghe, Brian Sloan and Ugur Muldur

PART 3 ACHIEVEMENTS The EU R&D under-investment: patterns in R&D expenditure and financing Vincent Duchene, Elissavet Lykogianni and Arnold Verheek Does the 'European Paradox' still hold? Did it ever? Giovanni Dosi, Patrick Llerena and Mauro Sylos Labini The European Research Area and human resources in science and technology Wendy Hansen Performance of European science: research networks and profiles of EU countries in a global perspective Anthony F.J. van Raan, Thed N. van Leeunm and Clara Calero- Medina The levelling off of the integration of European technology Dominique Guellec and Helgne Dernis The European Research Area as industrial policy tool Luc Soete PART 4 CONCLUSIONS 17 The future of the European Research Area Paraskevas Caracostas, Ugur Muldur and Kristian Orsini Conclusion and perspectives

Index

Contributors Kris Aerts, PhD, is Innovation Manager at K.U. Leuven R&D and is affiliated with the Department of Managerial Economics, Strategy and Innovation of K.U. Leuven, Belgium. Robert Boyer is an Economist at CEPREMAP (CEntre Pour la REcherche en itconoMie et ses Applications) and currently Invited Professor at the International Center for Business and Politics, Copenhagen Business School. Stefano Breschi is Research Fellow at KITeS-CESPRI, and Professor of Industrial Economics at the Department of Economics, Bocconi Unversity, Milan, Italy. Clara Calero-Medina is a Researcher at the Centre for Science and Technology Studies (CWTS), Leiden University, Leiden, the Netherlands. Paraskevas Caracostas works at the European Commission's DirectorateGeneral for Research. Dirk Czarnitzki is Professor of Industrial Organization and Strategy at the Department of Managerial Economics, Strategy and Innovation of K.U. Leuven, Belgium; Research Unit Coordinator Innovation Surveys at the Center for R&D Monitoring (ECOOM) at K.U. Leuven, and is affiliated with the Centre for European Economic Research (ZEW), Mannheim, Germany and the Centre for Industrial Economics at the University of Copenhagen, Denmark. Henri Delanghe, PhD, is an Expert at IDEA Consult, Belgium. Helene Dernis is at the Organisation for Economic Co-operation and Development Directorate for Science, Technology and Industry Economic Analysis and Statistics Division. Giovanni Dosi is Professor of Economics at the Sant'Anna School of Advanced Studies, Pisa, Italy.

Jim Dratwa works at the European Commission's Directorate-General for Research, and is Professor in Political Science at the Facultes Universitaires Saint-Louis, Brussels.

...

VIII

European science und technology p o l i c ~

Vincent DuchCne, PhD, is an Expert at the Competitiveness, Regulatory Management and Innovation Department, IDEA Consult, Brussels. Dominique Guellec is at the Organisation for Economic Co-operation and Development Directorate for Science, Technology and Industry Economic Analysis and Statistics Division Luca Guzzetti is a Professor at the Universita di Genova. Italy. Wendy Hansen is a Researcher affiliated with MERIT (Maastricht Economic and Social Research and Training Center on Innovation and Technology). Patrick Llerena is Professor of Economics at the University of Strasbourg and a Researcher at BETA UdS-CNRS, UMR no. 7522. Strasbourg. Elissavet Lykogianni, PhD, is Senior Consultant in the Competitiveness, Regulatory Management and Innovation Department, IDEA Consult, Brussels. Franco Malerba is Director of KITeS-Cespri, and Professor of Industrial Economics at the Department of Economics, Bocconi University, Milan, Italy. Ugur Muldur, PhD, is at the European Commission, Directorate-General for Research. Kristian Orsini, PhD, is at the European Commission's DirectorateGeneral for Research. Pierre Papon is Emeritus Professor at the Ecole de Physique et Chimie de Paris and Honorary President of the Observatoire des Sciences et des Techniques (OST). Brian Sloan, PhD, is at the European Commission, Directorate-General for Research. Luc Soete is Director of UNU-MERIT and Professor at the University of Maastricht in the Netherlands. Christian Svanfeldt, PhD, is at the European Commission, DirectorateGeneral for Regional Policy. hlauro Sylos Labini is Assistant Professor at the IMT Lucca Institute for Advanced Studies, Lucca, Italy. Peter Tindemans is Director of Global Knowledge Strategies & Partnerships. He previously worked as Director of Research and Science

Contributors

ix

Policy at the Netherlands Ministry of Education, Culture and Science and served as President of the Organisation for Economic Co-operation and Development's Megascience Forum.

Thed N. van Leeuwen, PhD, is a Senior Researcher at the Centre for Science and Technology Studies (CWTS), Leiden University, Leiden, the Netherlands. Anthony F.J. van Raan is Professor of Quantitative Studies of Science and Director of the Centre for Science and Technology Studies (CWTS), Leiden University, Leiden, the Netherlands. Arnold Verbeek is a Senior Expert at the Competitiveness, Regulatory Management and Innovation Department, IDEA Consult, Brussels. Nicholas S. Vonortas is Director of the Center for International Science and Technology Policy and Professor at the Department of Economics of The George Washington University, Washington, DC.

Acknowledgements Editorial assistance by Gaetane Lecocq and Kristian Orsini is gratefully acknowledged. Nevertheless, any errors are to be attributed ro the authors only. None of the chapters in this book represents any official position of the European Commission.

Abbreviations and acronyms AFA ALLEA ALMA BERD BRITE CAP CAs CATRENE CEA CERDA CERN CIP CIPAST CNEXO CNRS COST

CREST CSOs CURA DARPA DEISA DESY DFG DGRST DLR EADS ECB ECMWF

activities of foreign affiliates All European Academies Atacama Large Millimetre/submillimetre Array business expenditure on R&D Basic Research in Industrial Technologies for Europe Common Agricultural Policy coordinated actions Cluster for Application and Technology Research in Europe on Nanoelectronics Commissariat a 1'Energie Atomique European Committee of Research and Development European Organization for Nuclear Research Competitiveness and Innovation Framework Programme Citizen Participation in Science and Technology Centre National pour 1'Exploitation des Oceans Centre National de la Recherche ScientifiqueINational Institute for Scientific Research Cooperation europeenne dans le Domaine de la Science et de la Technologie (European Cooperation in Science and Technology) Scientific and Technical Research Committee Civil Society Organizations Community-University Research Alliances (Canada) Defense Advanced Research Projects Agency Distributed European Infrastructure for Supercomputing Applications Deutsches Elektronen Synchrotron Deutsche Forschungs Gemeinschaft Delegation Generale de la Recherche Scientifique et Technique German Aerospace Center European Aeronautic Defence and Space Company European Central Bank European Centre for Medium-Range Weather Forecasts

xii

ECSC EDC EDCTP EEC EGEE EIB EIRMA EISCAT ELDO ELWW EMBL EMBO EPO ERA ERA-NET ERC ERDA ESA ESF ESFRl ESO ESPRIT ESRF ESRO ETW EUCLID EUI EUMETSAT Euratom EUREC EUREKA FA0 FDI FhG FP FPR

European science and rechnology p o l i q

European Coal and Steel Community European Defence Community European & Developing Countries Clinical Trials partnership European Economic Community Enabling Grids for E-sciencE European Investment Bank European Industrial Research Management Association European Incoherent Scatter Scientific Association European Space Vehicle Launcher Development Organization European Laboratories Without Walls European Molecular Biology Laboratory European Molecular Biology Organization European Patent Office European Research Area European Research Area Network European Research Council European Agency for Research and Development European Space Agency European Science Foundation European Strategy Forum on Research Infrastructures European Southern Observatory/European Organisation for Astronomical Research in the Southern Hemisphere European Strategic Programme for Research and Development in Information Technology European Synchrotron Radiation Facility European Space Research Organization European Transonic Windtunnel European Cooperation for the Long Term in Defence European University Institute European Organisation for the Exploitation of Meteorological Satellites European Atomic Energy Community European Network of Research Ethics Committees Pan-European network for market-oriented, industrial R&D Food and Agriculture Organization foreign direct investments Fraunhofer Gesellschaft Framework Programme Nordic Research Policy Council

Abbreviations and acronyms

GAP GATT

GBAORD GEANT GERD GLOSPERA GMES GPS GPT HCRs HERD HOP! HRST ILL INCO INRA

INSERM IP IPC IPRs IPS IRAM IS1 IST

IST-RTD ITER JRC JTIs KBE KWG LERU LHC M&As MNCs MNEs MPG MR-N MR-U NAMA NASA

... xlll

gender action plan General Agreement on Tariffs and Trade government appropriations or outlays for R&D Europe-wide e-Infrastructure gross expenditure on R&D Global Systems and Policy Design for the ERA Global Monitoring for Environment and Security global positioning system General purpose technology highly cited researchers higher education R&D spending macro-economic impact of high oil price in Europe human resources in science and technology Institut Laue-Langevin specific international scientific cooperation activities Institut National de la Recherche Agronomiquel National Institute for Agricultural Research Institut National de la Sante et de la Recherche Medicale intellectual property International Patent Classification intellectual property rights integrated projects Institut de Radioastronomie Millimetrique Institute for Scientific Information information society technologies information society technologies, research, technological development and demonstration International Thermonuclear Experimental Reactor Joint Research Centre joint technology initiatives knowledge-based economy Kaiser Wilhelm Gesellschaft League of European Research Universities Large Hadron Collider mergers and acquisitions multinational corporations multinational enterprises Max Planck Gesellschaft Nordic Council of Ministers for Trade and Industry Nordic Council of Ministers for Education and Research non-agriculture manufacturing National Aeronautics and Space Administration

xiv

NATO NCoE NECs NEF NF NICe NIfS NIH NOAA NOES NORDITA NORDVULK NorFA NORIA NOVA NRA-NET NFSR

NSF NSI NTBFs OECD OEEC OMC OPEC PCT PPP PREST RACE RCN RFP RJVs RSFF RTD S&E SDI SEA SGP SSAs ST1

European science and technology p o l i c ~

North Atlantic Treaty Organization Nordic Centres of Excellence national ethics councils Nordic Energy Research NordForsk Nordic Innovation Centre Scandinavian Institute of Maritime Law national institutes of health National Oceanic and Atmospheric Administration networks of excellence Nordic Institute for Theoretical Physics Nordic Volcanological Center Nordic Academy for Advanced Study Nordic Research and Innovation Area Nordic Forestry, Veterinary and Agricultural University Network Nordic Research Area Net Belgian National Fund for Scientific Research US National Science Foundation Nordic Sami Institute new technology-based firms Organisation for Economic Co-operation and Development Organisation for European Economic Co-operation open method of coordination Organization of the Petroleum Exporting Countries Patent Cooperation Treaty purchasing power parity Policy Research in Engineering, Science and Technology Research and Development in Advanced Communications in Europe Research Council of Norway research framework programme research joint ventures risk-sharing finance facility research and technological development science and engineering Strategic Defense Initiative Single European Act Stability and Growth Pact specific support actions science, technology and innovation

Abbrrvi~tion~s and acronjms

STRePs Tekes TFRs TNO

TRIPS UFC UNESCO VLT WEU WTO

xv

Specifically Targeted Research Projects Finnish Funding Agency for Technology and Innovation total fertility rates Netherlands Organisation for Applied Scientific Research Trade Related Aspects of Intellectual Property Rights University Funding Council United Nations Educational, Scientific and Cultural Organization Very Large Telescope Western European Union World Trade Organization

Introduction Henri Delanghe, Ugur Muldur and Luc Soete This is a book about the most important concept underpinning current European Union (EU) research policy. It focuses on the notion of the European Research Area (ERA), whose achievement will become the main objective of EU research policy once the Lisbon Treaty enters into force. Simply put, the ERA involves a European 'internal market' for research, where researchers, technology and knowledge circulate freely, effective European-level coordination of national and regional research activities, programmes and policies takes place and new initiatives are implemented and funded at European level. This book describes the history of the concept, analyses some of its underlying assumptions, assesses some of its achievements and takes a brief look at its future. The ERA concept was launched formally in 2000 through the Commission Communication Towards a European Research Area and was provided with a new impetus in 2007 through the Commission Green Paper The European Research Area: New Perspectives. Yet the historical origins of the concept date from much earlier, in fact from at least 1974. This book explains how the ERA concept relates to two competing and alternating approaches to Community research policy and fits into the broader European post-war research landscape, consisting, in addition to Community research policy, of national research policies and intergovernmental research initiatives. In both the 2000 ERA Communication and the 2007 ERA Green Paper, the ERA was discussed mainly in terms of the features it should have. The 2007 ERA Green Paper, for instance, mentioned 'an adequate flow of competent researchers with high levels of mobility between institutions, disciplines, sectors and countries'; 'world-class research infrastructures, integrated, networked and accessible to research teams from across Europe and the world, notably thanks to new generations of electronic communication infrastructures'; 'excellent research institutions engaged in effective public-private cooperation and partnerships, forming the core of research and innovation "clusters" including "virtual research communities", mostly specialised in interdisciplinary areas and attracting a critical mass of human and financial resources'; 'effective knowledge-sharing svii

xviii

Europrun science and teclznologj~poliq

notably between public research and industry, as well as with the public at large'; 'well-coordinated research programmes and priorities, including a significant volume of jointly-programmed public research investment at European level involving common priorities, coordinated implementation and joint evaluation'; and 'a wide opening of the European Research Area to the world with special emphasis on neighbouring countries and a strong commitment to addressing global challenges with Europe's partners (pp. 2-3)'. Yet such an approach towards the definition of the concept does not render explicit the assumptions underlying the concept. For instance. one of the assumptions underlying the ERA concept appears to be that the optimization of the European research and innovation system as a whole generates larger scientific, technological and societal benefits than that of the older but smaller-scale individual national research and innovation systems, thereby glossing over the historical character and path dependency of national research and innovation systems. The ERA concept also appears to favour coordination over competition, whether in terms of research infrastructures or centres of excellence or research programmes, and to put a lot of faith into the existence of advantages of scale and scope in research. At the same time, focusing on features also leaves many questions unanswered concerning, for instance, the geographical borders of the concept (for example, does the ERA coincide with the EU and does it therefore enlarge as the EU enlarges; is the ERA the same in all S&T fields?). the conceptual borders of the concept (for example, does the ERA concern science, or innovation, or both?) and so on. This book analyses some of the assumptions underlying the ERA concept and looks into some of the unanswered questions. So far, progress towards the achievement of the ERA has been assessed mainly in terms of the policy actions taken at different levels of government. Much less attention has been paid to assessing the overall impact of the ERA on Europe's science and technology (S&T) performance and its contributions to the achievement of societal objectives. This book not only looks at policy actions taken but also assesses the ERA'S broader S&T and societal impact. The originality of this book lies in the breadth of its perspective, its evidence-based approach and frank investigation of issues, the quality and mix (academic, practitioner) of its contributors and contributions, and its timeliness given the 2007 re-launching of the ERA debate. The book is structured in four parts. Part 1 provides a historical perspective. In Chapter 1, Peter Tindemans sets the scene through a description of the development of the post-war European research landscape at national,

Introduction

xix

intergovernmental and European level. In Chapter 2, Pierre Papon zooms in more specifically on the development of intergovernmental initiatives in the field of research. In Chapter 3, Christian Svanfeldt assesses the extent to which Member States have amongst themselves engaged in crossborder research programme collaboration. In Chapter 4, Luca Guzzetti traces the development of the ERA idea at Community level. In Chapter 5 , Jim Dratwa updates the 2007 ERA Green Paper assessment of policy actions taken so far towards the achievement of ERA. Part 2 focuses on theoretical issues. In Chapter 6, Robert Boyer questions some of the assumptions underlying the ERA, such as the focus on continent-wide coordination at the expense of the optimization of national research and innovation systems, the benefits expected from researcher mobility and the geographical scope of ERA. In Chapter 7, Kris Aerts and Dirk Czarnitzki discuss the rationale for public intervention in the field of research. In Chapter 8, Nicholas S. Vonortas assesses the existence of advantages of scale and scope in research. In Chapter 9, Stefano Breschi and Franco Malerba analyse the networks that have emerged as a result of the Community Framework Programmes. And in Chapter 10, Henri Delanghe, Brian Sloan and Ugur Muldur discuss the extent of, rationale for and impact of cross-border programme and project collaboration. Part 3 focuses on achievements. In Chapter 1 1 , Vincent DuchCne, Elissavet Lykogianni and Arnold Verbeek assess whether the R&D investment gap between the EU and the US is decreasing. In Chapter 12, Giovanni Dosi, Patrick Llerena and Mauro Sylos Labini take an evidence-based approach towards the concept of the 'European Paradox'. In Chapter 13, Wendy Hansen discusses the ERA from the perspective of human resources. In Chapter 14, Anthony F.J. van Raan, Thed N . van Leeuwen and Clara Calero-Medina take a closer look at Europe's scientific performance. In Chapter 15, Dominique Guellec and Helene Dernis assess Europe's performance in the field of patents. And Chapter 16 by Luc Soete discusses the emergence of ERA as a form of industrial policy. The final chapter of this book in Part 4, by Paraskevas Caracostas, Ugur Muldur and Kristian Orsini looks towards the future. By the end of the book not all outstanding questions on ERA will necessarily have been answered. In fact, some new questions may have been raised. Yet at least a serious start will have been made with an evidencebased approach towards the issue.

PART 1

Historical perspective

1. Post-war research, education and innovation policy-making in Europe Peter Tindemans 1 VECTORS OF CHANGE AND THE EMERGENCE OF A NEW EUROPE IN RESEARCH AND EDUCATION It is rather daunting to try to cover 60 years of policy-making on research, education and innovation in Europe. It entered the period as a collection of states with neither clearly articulated policies for science and technology (S&T) nor integrating European mechanisms and now displays a full set of well-developed national policies next to a many-faceted Community Framework of the European Union and several intergovernmental cooperation schemes. Main trends will necessarily be covered; conceptual subtleties will therefore be glossed over. Yet one issue has to be dealt with: what policies are we actually dealing with? Names abound: research policy, policies for research and technological development, technology policy, science policy and, increasingly, innovation policy. As this chapter focuses on major changes in the funding for scientific research and technological development since World War I1 and the accompanying institutional changes, nationally and at a European level, no attempt is made to distinguish between research policy, technology policy or science policy. Differences would be of second-order importance. For example, statistically, expenditure on S&T is defined as broader than expenditure on research and development (R&D), one major difference being the inclusion of expenditure on science education. However, the remits of science policy and research policy in this respect can be considered to be essentially the same, as they do not follow this distinction: the scope of science policy hardly ever includes science education, though one ministry may be responsible for both. Innovation policy, which began to come to the fore in the 1970s but really gained currency in the 1990s, is of a different nature. In essence, one might say that no new institutions or major funding mechanisms have been created for innovation policy apart from innovation councils or platforms. These councils or platforms underline

4

European science and technology policy

that innovation is very much about interaction between various stakeholders and more integrated approaches between various policy domains without lumping together responsibilities or funding streams for these domains. This is, in fact, the essence of the key concept that nowadays underlies most thinking about innovation policy: national or regional systems of innovation (see for example, Lundvall, 1992). World War I1 brought to universities and research organizations across continental Europe not only destruction but also the wide recognition that science must be a key pillar of post-war reconstruction. The promises of the civilian use of atomic power were perhaps the major driver but warfare had opened many more windows on science and engineering's cornucopia. Aeronautics and the beginnings of space technology, radar and other electro-magnetic forms of communication and the computer were only some of the developments waiting to be exploited on an unprecedented scale. The prestige of scientists, physicists in the first place, had grown tremendously. Governments came into action to create organizations to fund and carry out especially basic research. Funding research was not alien to governments before the war but it was largely limited to a few areas where governments had direct operational responsibilities such as defence. telecommunications, agricultural extension services, geological surveys, civil works from roads to dikes and health. Those government interventions, moreover, could hardly be called policies: research efforts mostly took place in laboratories within operational branches of government: the postal and telecommunications service, the corps of civil engineers, the air force or the navy.' The overall levels of pre-war R&D efforts were low: of the order of onequarter to two-thirds of a per cent of national product in the US, the UK and Japan (David et al., 2000, p. 498). In addition, government's share was often much smaller than nowadays. Compared with 53 per cent in 1975, the US government funded only 20 per cent of overall R&D in 1940. In Europe, the comparable government shares were probably higher but reliable data is often missing. It took Edgerton and Horrocks (1994) quite some effort to give much more detailed and much higher estimates for pre-war R&D expenditure of British firms than Bernal was able to cite in a path-breaking book (1939, Appendix 111). Of course, there was some indirect support for research through government funding of higher education but only pre-war Germany funded basic science at some considerable level with its public support for the Kaiser Wilhelm Gesellschaft (KWG) and the predecessor of the Deutsche Forschungs Gemeinschaft (DFG). Now. after the war, many governments began to pour money into basic science: universities quickly came to be seen no longer as higher education institutions but as higher education and research institutions.

Post-wur research, education and innovation policymaking

5

Along with increased funding, especially for academically oriented research and the technologies emerging from World War 11, the 1940s and 1950s saw the evolution of the institutional pattern for performing and funding research and development that would largely determine the national styles for science and technology policy for a long time to come. Germany, the UK and France illustrate the different approaches that were taken. A first key characteristic for Germany (Krull and Meyer-Krahmer, 1996) is the way significant portions of its basic research and much of its applied research are organized in independent institutes under a rather powerful umbrella organization. Almost all of them have been established in the 15 years after the war beginning with the Max Planck Gesellschaft (MPG) in 1946, which essentially took over the institutes of the Kaiser Wilhelm Gesellschaft. The Fraunhofer Gesellschaft (FhG) for applied research dates from 1949 (though very small then) and the Helmholtz Gesellschaft, grouping research institutes centred around large research facilities of which the nuclear research centres from the 1950s are a key part, dates from 1959. A single large funding agency for academic research, the DFG, and the Wissenschaftsrat (German Council of Science and Humanities), which dates from 1957, completed the federal institutional backbone of the German science system. The balance between the federal government and the states, another typical feature of Germany's policy style, was strictly maintained after World War I1 and is nowhere more discernable than in the university arena. German universities, several of which were traditionally research strongholds, depend on the states for most of their funding and state ministries went about this in widely different, discretionary and often detailed ways. Combined with the constitutionally defined uniquely strong position of full professors, this left little room for central university management or for country-wide policies for academic research. Only the independent D F G and the Wissenschaftsrat, which in its composition (scientists, a few outsiders from industry or trade unions, but also state and federal politicians), reflects the federal nature of Germany, would be able to exert national influence. Finally, despite all the independent research organizations, Germany came to stand out by developing in a number of stages, involving the Ministry for Nuclear Affairs (1955) and its successor the Ministry for Scientific Research (1962), a powerful federal ministry of research and technology. This became responsible for almost all public expenditure on R&D outside the universities, including the oversight of the large research organizations. In the UK, the various sector ministries kept playing a much stronger role in funding research in their laboratories despite the brief interlude of the powerful Ministry of Technology from 1964 till 1970. Apart from that,

6

European science and technology policy

coordination was either absent or pursued through the Prime Minister's office. Universities were funded both for education and research through an intermediary body, the University Funding Council (UFC), with the exception of Oxford, Cambridge and London, for which the Ministry of Education remained responsible until the Higher Education Funding Councils succeeded the UFC in the 1990s. Some five (the numbers vary over the years), not one single, research councils fund additional university research as well as a rather small number of institutes outside universities, some of which are of the large-facility type. Private foundations such as the Wellcome Trust and other charities have played from right after the war a much larger role in funding research than elsewhere in Europe. France (Papon, 1979) is different again. The focus right after World War I1 was on the Centre National de la Recherche Scientifique (CNRS), which originated from before the war and was now put forward as the organization to carry out basic as well as applied research, as well as on effectively developing science policy. That was not going to last. The establishment of the Commissariat a 1'Energie Atomique (CEA) set nuclear things, the main propellant for science after the war, apart, the Defence Ministry put a fence around defence research and the same happened for aerospace. CNRS retreated on basic science, not as a funding agency, but as an assembly of research centres. In addition, other 'spin-offs' restricted the realm of CNRS, such as INSERM' for medical sciences, INRA' for agronomical research and CNEX04for oceanographic research. CNESSwould become, as mentioned already, a kingdom of its own for space R&D and was soon to be a key player in the French mixture of industry policy, policy in general and science and technology policy. French universities on the whole had little tradition in research and the Grandes ~ c o l e shad almost none at all. Neither would there be an equivalent of the German and UK research funding agencies, though eventually in the 1960s and 1970s. CNRS set up more and more 'unitis mixtes', research units at universities. The centralization on the Paris region would continue for several decades after World War 11. Within the government, administrative structures were set up to coordinate public research efforts linked to the French way of developing five-year plans for all sectors of society. The success of both the scientists pointing out forward-looking ways for investing in science, epitomized by the 'Comite des Sages', and their administrative counterpart drawing up coordinated investment plans for science, the Delegation Generale de la Recherche Scientifique et Technique (DGRST), depended very strongly on the interest of the Prime Minister and the President. As this began to wane, coordination gave way to the special interests linked to special programmes in atomic energy, space, aircraft industry, high-speed trains or the computer industry. In smaller countries, a mixture of these

Post-war research, education and innovation policy-making

7

institutional and policy approaches was to be seen. All of this formed the background for the successive attempts at European integration in the area of research and technology since the early 1950s. While institutional conditions gradually took shape, the actual fast growth in expenditure on R&D was part of a process driven by several factors. After the initial years of hardship, industrial expansion took off in the 1950s and key industries rapidly expanded their research efforts. The need for highly skilled personnel set in motion the expansion of higher education. And while the Cold War in general and the launching of Sputnik in particular were less of a trigger in Europe, they certainly added to the conviction that strong science and technology efforts contributed to the resilience of Western societies. The 1940s and 1950s were, of course, also the years in which the process of European integration took off. It was only natural to incorporate research and development in the first two European Communities, the European Coal and Steel Community (ECSC) of 1951 and Euratom (European Atomic Energy Community) of 1957. But the European Economic Community (EEC) established in one go with Euratom had at that time no provisions for a Community responsibility for science and technology, thereby underlining that apart from two specific sectors, science and technology policies were absent from the debate about European integration in general. That issue, and more generally the limitations of the rationale for European action, is perhaps nicely illustrated by the fate of a proposal to establish a European university. In the discussions leading up to the Euratom Treatyh in 1955, Walter Hallstein, the German Secretary of State, promoted the development of a full-fledged European university in the context of this treaty. Whereas, as will be discussed below, the objective of Euratom, namely the integration of the nuclear industry in Europe, provided a clear logic to establish a joint research centre, this was not easily applicable to the concept of a European university. Collaboration of existing universities within the six Member States was the preferred option.

2 TOWARDS MORE EXPLICIT AND BROADER SCIENCE POLICIES With the institutional landscape for doing research in universities, industrial laboratories, government institutes and not-for-profit organizations taking shape and funding levels on the rise, governments and industry began to think in a much more systematic way about government policies and industrial strategies and management approaches for research.

X

European science and rrchnologj~polic!.

In a number of countries, and also in Europe, as the previous section has shown, policy responsibilities for science had already been defined. But only in 1961, when the Organisation for European Economic Co-operation (OEEC), which was established originally to distribute the Marshall Plan funds, was transformed into the OECD (Organisation for Economic Co-operation and Development), serious and systematic attention started to be given to identifying and strengthening the key components of science policy. Alexander King, who played a major role in the not very successful efforts to focus OEEC's mind on science and in OECD's key contributions to turning science policy into a regular part of 'established' government policies, tellingly pictures the mindset of contemporaneous European governments. He describes the reaction of the Canadian ambassador to the OEEC, who had been invited to report on his visits to the scientifically advanced countries in Europe at the eve of OEEC's transition into OECD: Research activity had grown in a haphazard manner in response to specific needs. . . . and was seldom considered in terms of its economic potential or contributions to human understanding. It was hardly ever seen in policy terms. . . . His formal report . . . stressed the need for each country to develop a national science policy that recognized the importance of science as a basis for technological innovation and economic growth as well as its important intellectual functions. (King, 2001, p. 341) The OECD created a Directorate for Scientific Affairs and an Advisory Group was established under the chairship of the French Delegue General 3 la Recherche Scientifique et Technique, Piganiol. Its 1963 report discussed priorities to help allocate resources within science and how to use research results for various societal goals and put forward two main recommendations (OECD, 1963). Each government should set up a central mechanism to study science policy and review the total national effort in research and the OECD should convene a meeting of ministers responsible for science policy or science. Only somewhat later, the OECD condensed the experience of several countries and the ongoing studies on the place of science in relation to the sector policy domains of governments into three key recommendations: countries should draw up a science budget, which at least should be a compilation of all government expenditure on R&D. they should avail themselves of a coordinating mechanism within government and they should seek to benefit from an independent advisory body drawn from science, industry and other sectors of society. Much effort went into building up a framework for gathering reliable statistical data in order to measure and compare public and private expenditure on R&D. In 1970, agreement was reached between the OECD

Po~t-liarresemrch, educutwn and innovation polrcy-makmg

9

Member States on the Fruscuti Munuul, which has developed into the standard methodology worldwide for collecting R&D data and is now in its 2002 version. Reviews of national science policies, which the OECD started to organize in the 1960s, further contributed to the appreciation of the importance of science policy and the organizational and instrumental variation countries had developed in the meantime to conduct their policies.' Yet it was not only government policies that concerned the OECD. An important strand in the activities of the OEEC was devoted to management and productivity in industry. The US experience was, of course, leading in the period of reconstruction in Europe. OECD now transplanted this to industrial research and technological innovation: the complex process of turning research into successful products or production processes. The result was the establishment in 1966 of the European Industrial Research Management Association (EIRMA)Rin which the OECD's King and Philips' Casimir were the key players. The construction of explicit policies around priority setting in science and applying research to societal goals was accompanied by efforts to try to use the various sophisticated techniques often developed in a US defence context, such as operational research, systems analysis, planning and budgetary systems or forecasting methodologies (OECD, 1972). Interest in these techniques has subsided now and their use is largely limited to, for example, portfolios of large applied research projects. The epoch of the 1960s, during which science policy matured, is conveniently summarized by referring to two very influential OECD publications. The first is the Brooks Report (OECD, 1971); the second is a three-volume series on The Research System (OECD, 1972-74). The Brooks Report put science and technology squarely in the context of societies and governments facing challenges in all domains, including now explicitly the social domain with uncertain and changing goals. Industrial societies had become very complex, economic growth per se was no longer seen as a sufficient overall objective and disenchantment with technology and science had in many instances replaced the optimism of the reconstruction period. For science and technology to respond to new priorities, a wideranging set of recommendations was presented. A few examples illustrate the scope (OECD, 1971, pp. 89-108): 'This implies a much closer relationship between policies for science and technology and all socio-economic concerns and governmental responsibilities than has existed in the past', 'national and international economic agencies [are urged to] take greater account of scientific and technological factors in the implementation of their policies', 'Member States should channel their technological policies into areas capable of producing alternative, socially oriented technologies,

10

Europeun science und trchnologj,policj,

that is, technologies capable of directly contributing to the solution of present infrastructural problems, of satisfying so far neglected collective needs, and finally of replacing existing environmentally deleterious technologies'. In addition, there were recommendations on technology assessment, on the matching of higher education and the labour market. on the challenges multinational firms pose to national governments, on the balance between fundamental research and research addressing problems in sectors such as health or urban mass transportation, on social sciences, on a mechanism closely related to science policy - to investigate longrange policy issues in all domains and on science and underdevelopment. A tall order, indeed, for the ministers, agencies or other bodies that had been or were at last being set up in the European OECD Member States. The three-volume series, The Research System, deserves mentioning because it put the limelight on the systems nature of research efforts. Originally conceived as a comparative survey of the organization and financing of fundamental research, it became a state-of-the-art analysis of the various components of the research system, their interactions and the way governments attempt to influence these. Universities, enterprises, government and other not-for-profit research institutes or research councils and government policies and the institutional framework for the latter were discussed in their historical evolution and from the point of view of the specific challenges that each of these elements was facing. Rich in detail and comparing the real-life performance of key actors throughout the OECD area, the three volumes were a genuine treasure trove to which policy-makers turned in addition to the higher level and normative proposals of the Brooks Report. At the end of this OECD story a few remarks are needed about the United Nations Educational, Scientific and Cultural Organization (UNESCO). UNESCO has contributed significantly to the analysis and discussions on science policy, not only for developing countries. The UNESCO series, Science Policy Studies and Documents, aimed to 'collect, analyse and disseminate information concerning the organization of scientific research in Member States and the policies of Member States in this respect.19 Since 1965, a steady flow of overviews of national science policy and science organization, also in European countries, and of analytical studies has been published, one example being UNESCO (1970). However, UNESCO's impact on the development of the organization of and policies for science in Europe has never come near the influence of the OECD and virtually disappeared in the 1970s. Before turning to European cooperation and integration, the question of how Europe was doing compared with the US in the area of science and technology by the late 1960s needs revisiting. The situation in the late -

P o ~ t - w rreseurch, education and ~nnovutionpolicy-making

11

1940s and early 1950s was clear: the US had accumulated huge experience with applying science and technology in areas that could begin to pay off immediately in the civilian sector. It had learned how to organize science and technology and the Cold War gave a fresh boost. Continental Europe had to be built up and was lacking capital, which was not differcnt for the UK. So catching up and investing was the game to bc played. Had Europe caught up sufficiently by the late 1960s? The alarm bell was rung by Servan-Schreiber in Le D+ um&iccrin (1967). Europe was lagging dangerously in technology development. The technological gap had if anything only deepened. Upon closer scrutiny of the figures, it turned out to be true but much less so. If one focused on civilian technologies and was not as ardent a believer in a full civilian spin-off of military R&D efforts, the picture was much less bleak (King, 2001, p. 346). But the ball park had changed in other areas and these needed attention. Quite new was the research university. The major US universities were benefiting enormously from the huge investments in the NSF (National Science Foundation) but also from the Department of Defense: huge labs were based on campus at universities such as MIT, Stanford or Berkeley. They were now mounting research efforts at an unprecedented scale; the brain drain from Europe, which began in the 1930s for political reasons, now accelerated as these universities became the Mecca for European scientists (OECD, 1972-74; Smith and Karlesky, 1977). The agency system blossomed. The National Institutes of Health (NIH) was not a newcomer but its pre-war role was rather modest. Now it grew fast and developed a huge grant programme for university science. NIH's overall budget skyrocketed from $8 million in 1947 to more than $1 billion in 1966. The Atomic Energy Commission was at first the driving force behind the development of nuclear technology but its role was greatly expanded after its merger in 1974 with the Office for Coal Research into the Energy Research and Development Agency and later the Department of Energy, which came to run large programmes in basic science. The Department of Defense created DARPA (the Defense Advanced Research Projects Agency) in 1958 in response to the Sputnik launch and became famous as sponsor of very innovative research in basic science or building on basic science. Other organizations were to follow or to increase substantially their efforts in science and technology, such as the Department of Agriculture. How could a divided Europe respond to these immensely powerful sectoral agencies? The 1950s saw a third completely new phenomenon, the research parks and high-tech regions where high-tech firms, universities and their spin-offs and research institutes (often defence-related) developed new forms of symbiosis. Stanford Research Park led the pack and soon Silicon Valley was born. On the East Coast, Route 128 through suburban Boston

12

European science und rechnologj~policj

was quickly gathering worldwide renown, powered in the first place by MIT. And there were the huge federal labs, some originating from the atomic weapon labs from the war, others freshly created for basic science. So new challenges there were indeed.

3 EUROPEAN COOPERATION AND INTEGRATION In hindsight, the outcome in terms of institutional and other forms of cooperation between European countries of the first round of skirmishes in the late 1940s and the 1950s was not unexpected. In two areas, coal and steel, which were crucial for laying to rest the ghosts of war, and nuclear power, which held the promise of a major new industrial development, the political will was strong enough for a few bold people to break new ground. New political and executive institutions were established in ways resembling US-style sectoral approaches. Since the coal and steel industries were technology-based, it was not a big step to incorporate research in the ECSC. The Coal and Steel Research Fund established by the ECSC Treaty effectively continues to exist today as a ring-fenced part of the Framework even if the ECSC Treaty has expired. For Euratom it was imperative to have a strong basis in research; this was about creating a new industry through the mastering of complex technologies and the development of new scientific reactor concepts. Thus the Euratom Treaty defines a major role for the Commission in promoting and coordinating national research efforts and complementing these with a Community research and training programme (Euratom Treaty, Article 4). The Joint Research Centre (JRC), which was pivotal for Euratom's development mission, was established and became operational as of 1959 with new headquarters laboratories in Ispra in Italy and three branches in Karlsruhe in Germany, Petten in the Netherlands and Mol in Belgium. But it ran into an existential crisis ten years after its founding because it soon became clear that the industrial aims of Euratom, that is, to establish a truly European nuclear industry, based on solid European (JRC) nuclear R&D, foundered on national industrial and political interests. Nevertheless. the Euratom Treaty had opened the way for a very successful integrated European research and development effort in an area where industrial interests were few and far between. Thermonuclear fusion was to become the only area where national and community efforts harmoniously went together and where a common strategy and leadership for the European Commission in global negotiations turned out to be feasible. A key reason for this was the availability of substantial amounts of European money. Actors in the later debates of the 1970s and in the last ten years, under the guise of the

Post-wur reseurch, education and innovation policy-muking

13

European Research Area, instead focused on the coordination of national programmes to work on common R&D challenges. It was easier but success has rarely been theirs. Establishing CERN (European Organization for Nuclear Research) was relatively straightforward: a windfall profit from the newly discovered promises of the atom and based on strong international relations between key European scientists and with some US ones. And they were influential with governments. The ESO (European Southern Observatory/European Organisation for Astronomical Research in the Southern Hemisphere), the basis for which arose early 1953, was already in for a bumpier road. It took until 1962 and required substantial financial support from the Ford Foundation, which promised to pay 20 per cent of the investment cost, to pass the hurdle. Space technology was in a too early stage of development in the 1950s and its industrial promises were not yet clear enough to make a compelling case for moving forward in an integrated way as was done through Euratom. Moreover, the political shockwave that Sputnik created in the US was not strong enough in Europe to change that situation. Economic cooperation in general, at the basis of the EEC Treaty, was not seen in conjunction with research. And the same fate that befell Hallstein's attempt to create a European university befell the International Institute of Science and Technology (IIST), which was proposed in 1960 in the context of the North Atlantic Treaty Organization (NATO) as a response to the shortage of personnel in areas considered vital for responding to the Soviet challenge and to MIT and Route 128 (Krige, 2006, pp. 191ff.). The final proposal, which came from a high-level working group chaired by previous MIT president Killian and established by the NATO Science Committee of October 1961, was not going to fly. Resistance among the scientific communities in Germany, France and the U K put an end to the attempt in 1963-64. Notwithstanding the examples of particle physics and Southern Hemisphere astronomy, higher education and basic research were turfs that national governments and most academics and funding bodies keenly guarded. The European University Institute (EUI), which was established in 1972 in Florence and mainly provides PhD training in the fields of economics, social sciences, history, culture and law, proves that progress was possible only in a very special area and on a rather small scale. The second wave of discussions in the area of research took place between 1965 and 1974 and they were about the principles of Community R&D policy (for the history see Guzzetti, 1995). Its eventual outcome was pragmatic as it opened the door to an expanding role of the EC in R&D but buried for a long time a principled discussion on what EC and national responsibilities in the area of science, technology and innovation (STI)

14

European science and technology p o l i q

should be. Such pragmatism would remain the hallmark of the political vicissitudes of subsequent Framework Programmes (FPs) until the first years of the 21st century: expanding budgets and a growing number of goals but burying the substantial question of Community and national responsibilities under the mantra of subsidiarity. The discussion was started off by the PRESTI0 Working Group under Marechal, the successor as DGRST to Piganiol, who also figured in the OECD context. It was established in 1965 within the Medium-Term Economic Policy Committee of the EEC to examine Community policies for scientific and technological research. It did not get very far, however. Eventually PREST turned to coordination as its key subject, which led to the creation of the intergovernmental COST" programme in 1971 and its first seven concerted actions. The real discussion was initiated by Spinelli, who became Commissioner in 1972 and undertook to define a Community policy. He was convinced that the European Communities (after the Fusion Treaty of 1967, the three Communities had the same Council of Ministers and the same Executive, that is, the Commission) should be actively involved in financing research in Europe and in rationalizing its organization, for example, through centres of excellence, that this should not be limited to basic science but should also include applied research and that research and industrial policy should go hand in hand. The Commission's proposals of 1972 included the creation of a European Committee of Research and Development (CERDA) consisting of independent experts from science and industry whose task would be to advise on research programmes to be funded. A funding body, the European Agency for Research and Development (ERDA), was also to be established. It was modelled on the US NSF, but its approach would be more top-down and include, for instance, the selection and funding of centres of excellence or the supervision of national investment in large research facilities. Spinelli, who was responsible for both industry and research and technology, had already merged the staff of the Commission into one Directorate-General to facilitate a more integrated approach to these areas. But his proposals were not to be implemented. A new and larger Commission entered the scene after the UK, Ireland and Denmark acquired EC membership in 1973. Spinelli's portfolio was split up and he himself inherited industry; Dahrendorf, research. Dahrendorf changed tack almost completely. His focus was on basic research and he did not think that the EC should play a key role in the funding of research. Neither should it intervene in the institutional landscape as it had evolved in the EC Member States. Instead it should limit itself to coordination. It was Dahrendorf who was largely responsible for what was finally agreed upon in 1974: the principle of coordination

Po~t-warreseurch, education and innovution policy-making

15

was enshrined, CREST (Scientific and Technical Research Committee) was established and its being composed of Member State representatives underlined the weight attached to coordination, the Commission was authorized to make a beginning with Community R&D programmes though the budget was very small and the first foresight exercise, E+30, which was to catch a glimpse of future developments in science, technology and society, was set up. One cannot underestimate the importance of this discussion. The EC was firmly launched on a coordination track. The FPs budgets have, of course, increased significantly but they still revolve around networking, coordination and indirect action, which is EC jargon for stating that it is the companies, universities and labs in the Member States that carry out the research. The Joint Research Centre would remain a bone of contention. And only in the field of thermonuclear fusion, a truly European effort exists with the Commission in a role as closely resembling that of the US Department of Energy as is possible in EC circumstances. But in this case, the Euratom Treaty provided the rationale and paved the way for making available the money for common central efforts and for effectively coordinating national programmes. It helped that industrial interests were far away in the future. Andre underlines the importance of the events of the early 1970s when he traces the roots of the European Research Area back to Dahrendorf and emphatically not to Spinelli, the villain in Andre's cast (Andre, 2006, p. 133). It all depends, of course, on one's definition or view of the European Research Area (ERA), which for Andre is centred on coordination. It is ironic that the European Research Council (ERC), which was established in 2007, and the EU's much more pronounced role with regard to research infrastructures, which in 2008 are both touted as cornerstones of ERA, are rather more Spinelli's children. It is not surprising that in academic circles, little support was to be found for more radical changes in the European landscape. This was vividly illustrated by the establishment in 1974 of the European Science Foundation (ESF) through which academies of science (with a key role for the Royal Society), funding agencies and organizations such as CNRS and the MPG were quick to take the action away from governments. A true child of its time, the ESF was founded with a very limited budget and a focus on coordination. In areas where everyone, helped by the successes of CERN and ESO, realized that common approaches were necessary, cooperation between European countries did occur. But the Commission was not involved. Instead, the European Space Agency (ESA), the European Molecular Biology Organization (EMBO) and the European Molecular Biology Laboratory (EMBL), the Institut Laue-Langevin (ILL), the European

16

European science and technology policj

Synchrotron Radiation Facility (ESRF) or the European Transonic Windtunnel (ETW) were all established by an ad-hoc group of countries. usually later joined by other countries, also from outside Europe. The details of their gestation are not important here; they are all highly successful global powerhouses in their fields and there are more examples close to the field of research, such as the European Centre for MediumRange Weather Forecasts (ECMWF) or the European Organisation for the Exploitation of Meteorological Satellites (EUMETSAT). However, while to an observer comparing Europe's policy performance over the past decades with that of the US in this important area of large facilities for science, the US decision process with its budget uncertainty from year to year may seem peculiar, Europe's decision machinery has virtually ground to a halt. Only recently, policy circles at the national and European levels have realized that Europe's disposition for science policy may not be totally appropriate. The successive EC FPs have become another pillar of cooperation in Europe. Whereas in the 1970s, EC activity was largely limited to a small but successful programme (without a pre-defined programme) justly called 'Science' and a few scattered activities in, for example, health research, the FPs as they evolved during the 1980s and 1990s had a three-fold origin. The first was the ESPRIT (European Strategic Programme for Research and Development in Information Technology). Second were the traditional Euratom programmes, including the programme for the JRC. And third was the continued pressure to spend more EC money on research and link research to an ever larger number of policy goals. ESPRIT was an initiative very much along the lines of Spinelli. Commissioner Davignon's proposal to the 12 main European manufacturing companies in information technology to engage in cooperation in what became known as pre-competitive research under a programme with substantial European funding was initially met with some suspicion but became highly successful. Its success may have lured the Commission into believing too easily that the programmatic approach of ESPRIT could be copied to most other areas, thereby disregarding two key reasons why ESPRIT succeeded. The number of industrial players was small and they were all large companies with significant R&D laboratories with mostly very good relations with universities and research institutes. And ESPRIT'S budget was substantial indeed. Relative to national budgets, it was far larger than what was allocated to any other area in any FP. It was substantial even in countries with large R&D investments (much more than 10 per cent). Programmes like RACE (telecommunications'") and BRITE-EURAM (materials") were initially set up like ESPRIT as independent programmes separately put to the Council of Ministers and the European Parliament

Post-war research, education and innovation policymaking

17

for decision-making. The first FP was largely born out of the desire to simplify decision-making. Of course, there could hardly be long-term planning, except in the Euratom area, because a new Commissioner had the free hand to change whatever he or she liked or disliked, an opportunity not lost on most. And as more and more scientists and companies got involved in consultations, projects and evaluations, an almost unstoppable dynamic of pressure arose in national capitals and in the DirectoratesGeneral responsible for the FPs to increase the budget. The role of science in international relations created another dynamic. Relations with non-European and accession countries had to be established. Budgets continuously went up, but so did the number of objectives and areas. Moreover, during the 1980s and part of the 1990s, the bias towards ICT and energy was very significant, leaving many applicants disappointed. The growth of the number of areas was the result of another imbalance, which had to do with the goals of the Community. The Euratom Treaty and the ECSC Treaty had very specific objectives for their R&D component. When research and technological development (RTD) was incorporated into the EEC Treaty through the Single European Act of 1986, reference was made to industrial competitiveness only. As Community responsibilities and activities grew formally and in practice, a sometimes heated discussion flared up as to where the boundaries lay for the FPs, which resulted in the explicit recognition in the Maastricht Treaty of 1992 that RTD could also target all other Community responsibilities and objectives. The Constitutional Treaty set yet another step by explicitly mentioning space and security. The last boundary was crossed with the European Research Council: basic research, or frontier research as the new buzzword has it, has been accepted as a category. It seems unavoidable that the discussions on the European Research Area involve a fundamental rethinking of the objectives and the position of the FP.

4 REFORMS IN HIGHER EDUCATION, RESEARCH INSTITUTES AND INNOVATION AS THE NEW PARADIGM From the 1970s onwards, two developments stand out at national level. They were in fact much related. The Brooks Report was not the only report but certainly an influential one putting innovation centre stage. In the first place, industrial innovation is involved but innovation in all sectors of society and government is involved as well. The emergence of 'enabling technologies', no longer characteristic of one industry only but pervading many, did much to focus attention on how governments and also

18

European science and technology polic!

universities and research institutes could and should support innovation in industry and in other sectors. The spotlight was put on the performance of these institutions and it became clear that reform was necessary to enable them to meet the growing expectations. Reform of universities, research organizations and funding agencies and of the overall science and technology system has therefore been a major trend starting in the 1970s and continuing throughout the next decades. A few examples of such reform will be given. Universities in Europe reflect national traditions. Their organization. curricula, degrees, autonomy (or lack thereof). links to (or even position within) government, personnel status, funding and so on differed and still differ in many respects from country to country. Mobility between countries was low. The 1960s and 1970s confronted universities and governments with crucial decisions to be made as they had to cope with vastly expanded enrolment and rising costs to keep up with the requirements of new and emerging fields of science and technology. There were also the new expectations from industry and the need to adapt so as to work increasingly across the boundaries between the traditional disciplines. In many ways, UK universities resembled the US ones and it comes as no surprise that the first Science Park was most likely created at Cambridge University in 1970. Europe has, of course, its tradition of technological universities, which in most cases have been working in close connection with their industrial environment. But their focus was usually on the larger established companies with sizeable research laboratories. Organizing for entrepreneurship was a new challenge and this required a more entrepreneurial approach to running a university as well. Countries in North-Western Europe were the first ones to undertake serious reform in higher education. The Nordic countries, the Netherlands and the UK were among the countries where new funding models were introduced, based, to a much larger extent, on performance. Curricula were shortened. More professional internal management models were adopted. Quality assurance was institutionalized. Autonomy was increased, though in many cases not yet sufficiently. The Bruegel Policy Brief 'Why Reform Europe's Universities?' (Aghion et al., 2007) convincingly argues that up to this day more autonomy, in particular with regard to budgets, hiring, remuneration, course design and student selection, is crucial for improved performance. The agreement to introduce readable and comparable degrees under a three-cycle model and national qualifications frameworks, which forms the core of the Bologna process initiated in 1999 by European governments, was an important step in increasing efficiency and transparency and thus in encouraging mobility. Applied research organizations have felt the winds of change as well.

Post-war research, education and innovation policymaking

19

Until the late 1970s, TNO, the Netherlands Organisation for Applied Scientific Research, was an organization heavily linked to and controlled by several government ministries. Reform then overhauled its internal organization to put management in charge, organize supervision by societal stakeholders directly (instead of indirectly through government departments) and at arm's length, base its financing model on responsiveness to clients, and assist management in introducing new personnel policies. System-wide reform was also necessary. One reform concerned the strengthening of links between independent institutes for basic and sometimes applied research and universities. A good example is the Max Planck Gesellschaft in Germany, where it was increasingly felt that the distance to the universities, with their constantly evolving student populations, had become too large. In France, this kind of reform was part of the considerable turnaround of the French system as it had developed after World War I 1 (Mustar and Laredo, 2002). CNRS is still responsible for large parts of French basic research efforts but much of its money goes to unit& miuttes, joint research groups of CNRS and a university, which are based at the university. The traditional French focus on large programmes, often built around large firms, such as the 'Plan Calcul' for the computer industry. has largely disappeared as well and like everywhere else, small and medium enterprises (SMEs) now play a considerably larger role in innovation and research policy. Indeed, because of the erosion of the concept of national champions, let alone nationalized companies, interventionist French industrial policy has largely disappeared. The Gvandes ~ c o l e snowadays perform high-quality research and in addition to the Paris region, several regional concentrations of research and innovation have developed, wellknown examples being Grenoble and Sophia Antipolis. System-wide reform was, of course, necessary in the Central and Eastern European countries after 1989 as their systems, which were based on Academies of Science running large parts of basic research and on state companies accounting for most industrial production in a plannedeconomy mode, had to adapt to a world where universities and private innovation-focused companies would be the dominant players. Many of these reforms were prompted or at least influenced by innovation moving centre stage. The key point here is that during the 1960s, innovation policies - to the extent that the word was used in the first place - still very much focused on a few sectors such as aircraft, space, nuclear energy and electronics (including computers) (Pavitt and Walker, 1976). Civilian government R&D, government procurement, industrial mergers and attempts at European cooperation were the key instruments mentioned by Pavitt and Walker. In the 1970s, the concept was

20

European science and techno log^ p o l i q

broadened and the innovation processes in firms in general and the ways in which governments could stimulate innovation in companies, began to be studied systematically. The Six Countries Programme epitomizes these efforts (Sweeney, 1985). The independent examination of and discussion on instruments and policies to stimulate innovation, their effectiveness and the industrial and technological environment in which they had to operate was the goal of the two-yearly meetings. Organizations from four countries (France, Germany, the Netherlands and the UK) held their first meeting in 1974 and were soon joined by organizations from Canada and Ireland and later from other countries. The innovation performance of companies is determined by many factors and quite a few policy domains affect these determinants. Support for R&D in firms, stimulating interaction between firms and universities, government procurement policies, environmental regulations or promoting well-functioning capital markets are examples. Governments began to try to capture this wide array of possible instruments in broadly conceived innovation policies. In Europe, the Netherlands was one of the first countries to do so (Ministries of Education and Science et al., 1979). In these new forms of innovation policies, two related concepts were central: interaction between various actors (firms, universities and research institutes, capital providers, governments and so on) and a systemic approach. The latter meant (Gaudin, 1985) an analysis of the logic of these actors, of institutions and of conditions conducive to innovation. The approach has been elaborated and empirically underpinned and two notions summarizing this body of knowledge are now very often the basis for innovation policies: national (or regional) systems of innovation (see for example, Lundvall, 1992) and clusters (Porter, 1998). A crucial point, which has a bearing on the discussion about the European Research Area, is that these concepts - and the newer nicely fitting concept of open innovation - strongly emphasize that even in an age of globalization, the local aspect of innovation is still very important. Innovation policy has to recognize this dimension and much of it will remain the responsibility of national and/or regional governments.

5

EPILOGUE: RELEVANCE FOR THE EUROPEAN RESEARCH AREA

A major area in which Europe cannot be complacent is higher education. Reform has been and is ongoing. But among European universities, there should be more differentiation. Too many of them have the ambition to be research universities. A comparison with the US, where about 200 universities carry out almost all research at universities. makes this clear.

Post-war reseurch, education und innovation policy-making

21

More autonomy is needed as well and will help to increase differentiation. Obviously this is in the first place a challenge for national governments and institute leadership. A European assessment mechanism building on national evaluation schemes and adopted by national governments would, however, be an important step at European level. It would increase transparency and give institutions a much better view of their position in Europe and this could be a strong incentive for their strategic positioning. National governments are also the ones that need to address the issue of mobility as the key stumbling blocks in this regard have to do with nonportable and widely differing social security and pension schemes. But there are things that need to be done at European level and the balance struck in the 1970s between national and community responsibilities needs reconsideration. Coordination between national policies and funding agencies does not seem to be able to substitute for the effectiveness and efficiency of European-level instruments to support research and technological development in a few areas, as the US example suggests. Having a European-level funding mechanism for basic research is definitely such an area. This will also contribute to greater differentiation among universities, as the US, where a few federal agencies are responsible for funding about 60 per cent of university research, demonstrates. The creation of the ERC is, therefore, a major step and should be an opportunity to fundamentally discuss the FP, as it has developed in a somewhat haphazard way. Independence and larger budgets are key issues requiring attention and the balance between European-level funding and the money from national funding agencies needs revisiting. This is another major challenge for national governments. Intergovernmental collaboration so far has resulted in excellent large European research facilities. But the difficulties European countries experience in deciding about new facilities, apart from new generation instruments at existing laboratories, illustrate that a genuine decision-mechanism and a European-level budget are essential for equipping Europe with essential research infrastructures. Europeanlevel funding for basic research through an independent ERC and funds to contribute to research infrastructures should be the two mainstays of an EU budget for research and technological development. Another one, once more suggested by the federal mission-oriented agencies in the US, is funding for large strategic R&D programmes in areas of an evidently European nature. Global change issues, energy issues, many environmental issues, transportation issues or, for example, certain health issues, are such European missions. The newly developed instrument of Joint Technology Initiatives, the first of which have been launched in early 2008, could possibly evolve into a mechanism of some relevance in this regard. Setting them up, however, has been very slow and their organization is

22

European science und technology policj

rather cumbersome, raising doubts as to whether they can be the sustainable mechanism that Europe needs. Finally, innovation has found its way not only into national policies of national governments. At European level too, innovation is the talk of the town. And whereas in the areas mentioned above, there seem to be compelling reasons t o reconsider the balance between national responsibilities and those of the Community, such arguments are largely lacking for supporting innovation. Of course, a genuine European patent would be a n important step, but innovation and applied research. with the exception of large European mission-driven areas, continue to have strong local elements.

NOTES The establishment by law of the Netherlands Organisation for Applied Scientific Research (TNO) in 1930 to help enterprises and society benefit from science and technology was an exception. The creation of the Belgian National Fund for Scientific Research (NFSR) in 1928 was not: true enough King Albert I was the instigator. but its financial means came from industrialists, banks and private benefactors. not the government. Institut National de la Sante et de la Recherche Medicale. Institut National de la Recherche Agronomique. Centre National pour I'Exploitation des Oceans. Centre National d'Etudes Spatiales. Largely based on information on website of the European University Institute: http:l/ www.iue.it/About (accessed 15 May 2009). For example, France (1966) and the United Kingdom and Germany (1967). all published by OECD, Paris. http:Nwww.eirma.org/f3/cmps-index.php?page=home (accessed 15 May 2009). The quote is from a resolution adopted by the 1I th General Conference of UNESCO in 1960 (Resolution 2.11 316). Policy Research in Engineering, Science and Technology. European Cooperation in Science and Technology. Research and Development in Advanced Communications in Europe. Basic Research in Industrial Technologies for Europe.

REFERENCES Aghion, P., M. Dewatripont, C. Hoxby, A. Mas-Colell and A. Sapir (2007). 'Why reform Europe's universities?, Bruegrl Policy Brief, 4, Bruegel Foundation, Brussels. Andre, M. (2006), 'L'espace europeen de la recherche: histoire d'une idee', Revue d'histoire de I'intPgration europienne, 12 (2), 13 1-50. Bernal, J.D. (1939), The Social Function of Science. London: Routledge and Kegan Paul. David, P.A., B.H. Hall and A.A. Toole (2000), 'Is public R&D a complement or

Post-war research, education and innovation policymaking

23

a substitute for private R&D? A review of the econometric evidence'. Research Policy, 29 (&5), 497-529. Edgerton, D.E.H. and S.M. Horrocks (1994). 'British industrial research and development before 1945', Economic History Review, 47 (2), 21 3--38. Gaudin, T. (1985), 'Definition of innovation policies', in: G. Sweeney (ed.), Innovation Policies: An International Perspective, London: Francis Pinter. Guzzetti, L. (1995), A Brief' History of European Union Resmrch Policy, Luxembourg: Office for Official Publications of the European Communities. King, A. (2001), 'Scientific concerns in an economic environment: science in OEEC-OECD', Technology in Society, 23 (3), 33743. Krige, J. (2006), American Hegemony and the Postwar Reconstruction of'Scirncr in Europe, Cambridge, MA: The MIT Press. Krull, W. and F. Meyer-Krahmer (eds) (1996), Science and Teclznology in Germany, London: Cartermill Publishing. Lundvall, B.-A. (ed.) (1992), National Systems of'lnnovation: To~vardsa Theory of Innovation and Interactive Learning, London: Francis Pinter. Ministries of Education and Science, Economic Affairs, Agriculture, Nature Management and Fisheries (I979), Innovatie (in Dutch), The Hague: Staatsuitgeverij. Mustar, P. and P. Laredo (2002), 'Innovation and research policy in France (1980-2000) or the disappearance of the Colbertist state', Research Policy, 31 (I), 55-72. OECD (1963), Science and the Policies of Governmmts, Paris: OECD (also known as the Piganiol Report). OECD (1971), Science, Growth and Society, Paris: OECD (also known as the Brooks Report) OECD (1972), Analytical Methods in Government Science Policy, Paris: OECD. OECD (1972-74), The Research System, 3 volumes, Paris: OECD. Papon, P. (1979), 'Centres of decision in French science policy: the contrasting influences of scientific experts and administrators', Rclsearch Polio., 8 (4), 384-98. Pavitt, K. and W. Walker (1976), 'Government policies towards industrial innovation', Research Policy, 5 (I), 11-97. Porter, M. (1998), 'Clusters and the new economics of competition', Harvard Business Review, 76 (6), 77-90. Servan-Schreiber, J.-J. (1967), Le DClfi am6ricain, Paris: Denoel. Smith, B.L.R. and J.J. Karlesky (1977), The Universities in the Nation's Research Effort, New Rochelle, N Y : Change Magazine Press. Sweeney, G. (ed.) (1 985), Innovation Policies: An International Perspective, London: Francis Pinter. UNESCO (1970), Manual for Surveying National Scientific und Teclznological Potential. Paris: UNESCO.

2. Intergovernmental cooperation in the making of European research Pierre Papon Until the mid-20th century, scientific research in Europe was a national undertaking, although cooperation between scientists had been an old tradition, even during war periods (Rossi, 1999; Blay and NicolaTdis, 2001). At the end of the 1940s, scientific cooperation on a large scale was considered a means for establishing durable peace and Europe-wide cooperation took shape in the 1950s. It was intended to tackle efficiently the reconstruction of Europe's scientific potential and to face new and major scientific challenges like those represented by nuclear physics. This reconstruction effort thus opened a new era for European science.

1

THE CHALLENGES O F 'MEGASCIENCE' AND O F MODERNIZATION: THE CERN AND ECSC MODELS

The emergence of atomic physics in the 1930s began to transform the organization of scientific research since for the first time since the building of astronomical observatories, large facilities such as particle accelerators became necessary for the advancement of science. During the war, the building of the atomic bomb presaged the advent of a new era of 'megascience' dominated by large machines conceived and operated by hundreds of scientists and technicians. Nuclear physics. an area in which European researchers had achieved great breakthroughs, was considered a starting point for European cooperation and in December 1949, the Conference Europeenne de la Culture meeting in Lausanne proposed multinational cooperation in this domain. It was followed in June 1950 by a meeting of physicists organized by Pierre Auger and sponsored by the United Nations Educational, Scientific and Cultural Organization (UNESCO). Then an intergovernmental meeting also convened by UNESCO launched the foundation process of the European Organization for Nuclear Research (CERN), which was created by an intergovernmental treaty signed in

Intergovernmental cooperution

25

June 1953 by 12 Western European countries (but including Yugoslavia). Switzerland, which had played an important role in the creation of CERN, attracted the new organization to Geneva by donating the land on which it was established in 1954 (Strasser and Joye, 2005). Its first accelerator began operation in 1959 (Hermann et al., 1987). CERN was the first of a series of European laboratories created through an international treaty, with its own facilities and a permanent staff possessing a specific international status. European physicists gathered at CERN to perform experiments with large machines and large particle detectors operated by multinational teams in partnership with national laboratories. Over time, the organization became the world's focal point for high energy physics, attracting researchers from countries outside Europe like the US, India and China. Today, 20 European states have joined CERN (eight other states being observers), while China and the US participated in the construction of the Large Hadron Collider (LHC), which in 2008 became the world's largest accelerator.' One should emphasize that CERN was created with the objective of initiating strong European cooperation in an emerging field, where Europe was in danger of being outpaced: assembling scientific competence and infrastructures was an obvious advantage for Europe and the only way to remain in the scientific competition in high energy physics. The road to success was eased by the fact that there were few obstacles to European cooperation in this domain, which is rather far removed from civilian or military applications. CERN has been without doubt one of the most successful undertakings of European cooperation in science. It is governed by a council in which each Member State is represented and that fixes the policy of the organization and designates its director (it meets periodically at ministerial level). It has succeeded in launching a policy for the construction of new machines that gave a leading edge to Europe in particle physics. Although several countries have kept in operation important national facilities in nuclear physics (France and Germany, for example), CERN has become the unique European laboratory for experimental high energy physics. This situation is explained by the necessity of concentrating financial and technical means to construct machines that need to be operated with increased energy and thus with increasing costs. Building such facilities requires advanced planning involving scientists and engineers through ad hoc panels: this was the case with the LHC.' The track on which the CERN scientific venture was launched was completely different from the one that the European countries followed to initiate the political and economic cooperation leading to the European Union. Providing Europe with a modern capacity for the production of coal and steel and thus ensuring the basis of its independence was the

26

European science and technologj poliq.

motivation behind the May 1950 proposal of Jean Monnet and Robert Schuman to create the European Coal and Steel Community (ECSC). The 'pooling' of France's and Germany's steel production capacities was also a political means to contribute to closing a long series of conflicts between those countries, which had been disastrous for Europe. The six founding Member States of the ECSC (Belgium, France. Germany. Italy, Luxembourg and the Netherlands) signed in 1951 the Paris Treaty establishing the ECSC and launching the great venture of European integration. Under the ECSC 'High Authority', a very successful European research programme on coal and steel was established but, in contrast with CERN. no common facilities were constructed (Bossuat, 2006). After the rapid success of the ECSC, its six Member States founded together in March 1957 through the treaties of Rome, the European Economic Community (EEC) and the European Atomic Energy Community (Euratom). Euratom's goal was to lay the basis for a common policy for the development of nuclear energy and thus for Europe's energy independence (an objective that is still a matter of concern 50 years later). it has been managed as a scientific and technological tool of the European Commission to implement a 'common' nuclear energy policy with the support of the Joint Research Centre's (JRC) laboratories (which has never really materialized) and a fusion programme. The EEC Council of Ministers determined the orientations of common policies, but the EC, the ECSC and Euratom played a kind of 'federal' role in their implementation and they were thus not governed like intergovernmental organizations such as CERN. Intergovernmental cooperation in the field of research and the EEC were thus, since the beginning, evolving on two different tracks.

2 NEW EUROPEAN INSTITUTIONS FOR NEW CHALLENGES CERN's success was an incentive for the creation of other European organizations when it appeared that Europe had to face new scientific challenges in several domains, molecular biology, astronomy and space research being among the most important of them at that time. The life sciences had gone through a deep transformation during the 1950s with the emergence of molecular biology, but they had less political appeal than nuclear physics and with the exception of UK, the development of molecular biology lagged somewhat behind in many European countries. At the suggestion of Victor Weisskopf, the CERN Director-General, European biologists led by John Kendrew decided to found in 1964 the European Molecular Biology Organization (EMBO) in order to convince

Intergovernmentul cooperation

27

European governments to create a new intergovernmental organization to support the development of molecular biology in Europe. Their campaign succeeded with the creation in 1969 of the European Molecular Biology Conference and with the foundation in 1974 of a European laboratory in Heidelberg, the European Molecular Biology Laboratory (EMBL), conceived to operate like CERN although at that time there was no need for heavy infrastructures. The conference is an intergovernmental organization with 27 European Member States, which is presently funding a generic programme in biology, the EMBO fellowships and workshops, . ~ local EMBL antenand that has the EMBL as a central l a b ~ r a t o r yTwo nas in Hamburg and Grenoble (the Deutsches Elektronen Synchrotron [DESY] and the European Synchrotron Radiation Facility [ESRF]) give access to major synchrotron radiation facilities, which are important tools for structural biology research. Although EMBL has established a reputation over the years of being among the world leaders in its field, it did not, unlike CERN, become the 'focal' point for its discipline in Europe and the world since there remain several large national laboratories in Europe and the world and since there is no need for a complete European pooling of funding. Astronomy has several similarities with particle physics: research activities require access to large infrastructures with an increasing cost, and scientific communities share a long cooperation tradition. The necessity of building new large observatories in remote areas, where atmospheric conditions are ideal for astronomy observation, has been a powerful incentive for promoting European scientific cooperation in astronomy. Once more the CERN model inspired the creation of a new European organization with the objective of building joint research infrastructures: the European Organisation for Astronomical Research in the Southern Hemisphere (ESO - European Southern Observatory), which was created in 1962 through an intergovernmental agreement signed by five founding countries. Today, 13 countries are ESO members. It has built and operates state-of-the-art observatories in the Chilean Atacama desert, in particular the Very Large Telescope (VLT) located atop the Paranal, a 2600-metrehigh mountain. ESO is presently building in collaboration with North America, East Asia and Chile, a giant array of submillimetre antennas, the Atacama Large Millimetre/submillimetre Array (ALMA), which will be operational in 2013. ESO has succeeded in establishing itself as the European focal point for optical and infrared ground-based astronomy, although many countries have kept national observatories in activity. ESO is the de facto European partner in the framework of international cooperation in astronomy; this is the case, for example, with the ALMA project. It is also planning the equivalent of its future LHC with the next

generation of giant telescopes beyond the VLT: the Extremely Large OpticaYInfrared Telescope, for which ESO is developing a new design with a primary 30-60-metre-diameter m i r r ~ r . ~ In the post-war period, particle accelerators, nuclear research reactors and optical and radio-telescopes have opened the way to 'megascience' with research infrastructures run as industrial facilities: experiments and observations have to be planned far in advance and executed according to strict planning, and data have to be stored and often distributed all over the world.5 During the 1960s, the picture changed again as scientists studying the properties of condensed matter used nuclear reactors to explore with fluxes of neutrons the structure of solids, liquids and surfaces. Biologists followed suit when they engaged in the study of biomolecular structures and soon the possibilities offered by electromagnetic radiation emitted by particle accelerators were seized by almost all disciplines. Large infrastructures were built in the US, Europe and Japan to serve broad scientific communities. 'Megascience' thus became a widespread phenomenon requiring huge investments, as capital costs amounted to several hundreds of millions of euros for every new facility. The list of large European facilities (a dozen in total) has thus grown under the increasing pressure of scientific needs. By the close of the 1980s. it included the Institut Laue-Langevin (ILL, a high flux neutron reactor in Grenoble), the European Centre for Medium-Range Weather Forecasts (ECMWF) located in Reading (UK) and some smaller facilities such as the European Incoherent Scatter Scientific Association (EISCAT) and the Institut de Radioastronomie Millimetrique (IRAM, run by France, Germany and Spain). The ESRF, created in 1988 by a 12-country intergovernmental agreement, is one of the most recent facilities. It is located in Grenoble nearby the ILL. The political decision to build a European windtunnel for aerospace research was taken when France and Germany agreed on establishing the ESRF.6 The European Transonic Windtunnel (ETW) was built in Cologne (France, Germany, the Netherlands and UK jointly operate the facility) in the vicinity of the German Aerospace Center (DLR). It is the only European infrastructure fully dedicated to industrial research.

3 FACING THE US-USSR DUOPOLY: EUROPEAN COOPERATION IN SPACE Space exploration had remained a dream until the mid-20th century, although the first realization of rockets had been undertaken before that in the Soviet Union, Germany and the US. Nazi Germany had achieved technical breakthroughs with the work of Walter Dornberg and Wernher

Intergovernmental cooperation

29

von Braun in the field of rockets; the V2 was thus built in Peenemiinde and fired on UK cities at the end of the war (Sebesta, 2006).' Although potential applications of space activities appeared rather remote at the beginning of the 1950s, except in the military area, scientists had perceived the interest of rockets for atmospheric research and the detection of cosmic rays. The 1957-58 International Geophysical Year (IGY) catalysed international cooperation, in particular in atmospheric science, for which the use of balloons and rockets was a necessity, but the launching of the first satellite 'Sputnik-1' by the USSR in 1957 with a military rocket created a political shock throughout the world. The US responded by launching the satellite 'Explorer-1' with the military rocket 'Saturn-C', which had been developed under the supervision of Wernher von Braun, and with the creation in 1958 of the National Aeronautics and Space Administration (NASA). Discoveries by the satellite 'Explorer-1' had illustrated the scientific . ~ it was clear that the need for rockets possibilities of space r e s e a r ~ hYet would place space research amidst serious political debates as most efforts for rocket building were in the hands of the military establishment. Meanwhile, the two physicists Pierre Auger and Edoardo Amaldi, who had an interest in cosmic rays and had taken part in the creation of CERN, met in Paris in January 1959 to pave the way towards the organization of space research in Europe along the CERN model (Sebesta, 2006). As a follow up, Amaldi wrote a brief note on 'Space Research in Europe' (Tinjod and Battrick, 1998) in order to mobilize his colleagues. Although the North Atlantic Treaty- Organization (NATO) had been thought of as a possible channel for organizing a Western response to the Soviet challenge in space, there was widespread hostility among European scientists towards linking space science and the military establishment. The UK proposed the idea of combining into a single institution cooperative efforts in the field of satellites and for the development of a launcher on the basis of its 'Blue Streak' military rocket, which it proposed to convert to civilian uses. After a conference convened by Switzerland in 1960 in Geneva at CERN, the British proposal was abandoned and in 1962 an intergovernmental agreement was signed creating two separate organizations: the European Space Research Organization (ESRO) in charge of satellites and research and the European Space Vehicle Launcher Development Organization (ELDO) responsible for building launchers. ESRO was supported by ten countries; among them, as neutral countries, Switzerland, Sweden and Austria did not join ELDO. ESRO succeeded in establishing a European programme for space research with satellite projects, but ELDO was plagued with technological difficulties, cost overruns and political disputes. The European rocket 'Europa' was

30

European science and technolog>.p o l i q ~

a failure and ultimately countries decided to merge ESRO and ELDO into a single organization by creating in 1975 the European Space Agency (ESA). In 1980, European stakeholders established the French company Arianespace with the mission of building the series of 'Ariane' rockets. Today, ESA is an intergovernmental organization with a membership of 17 states (15 EU countries, Norway and Switzerland). which has the mission of carrying out a multifaceted European space programme: research on the Earth's environment, the solar system and the universe. the development of satellite-based technologies and services and the promotion of European industry in those domains. The ESA Council is the agency's governing body. Each Member State is represented there, it adopts the policy of the agency and its programmes and it meets periodically at ministerial level.9 ESA is also a platform for cooperation with space organizations such as NASA in the US. It has created five centres with different responsibilities, the most important one in Noordwijk in the Netherlands. ESA has become the focal point of European space activities, although several countries have kept important activities under national agencies, for instance, Germany with the DLR and France with the Centre National dlEtudes Spatiales (CNES). Since its creation in 1975, ESA has succeeded in creating a space community in Europe with several great achievements (for instance, the 'Mars Express' mission) and with 'Ariane' being considered a great technical and industrial success. Arianespace is launching most ESA satellites and has captured a large share of the market for launchers of commercial satellites using the French space centre in Guiana as a launch base. ESA has managed its Industrial contracts for space programmes according to a geographical distribution key or 'loidu juste retour': their execution by national companies in terms of funding must be roughly proportional to the contribution of Member States to the agency's compulsory and optional programmes. This constitutes a serious constraint as the high-technology input of contracts has to be preserved while technical competences are unevenly distributed among Member States. Artificial satellites rapidly became an operational tool for various public and commercial activities. Meteorology was a case in point and in 1986, European governments decided to create a second European space organization, EUMETSAT (European Organisation for the Exploitation of Meteorological Satellites), which operates the 'Meteosat' meteorological satellites fleet providing meteorological data.I0 EUMETSAT also cooperates with CNES, NASA and the National Ocean and Atmospheric Administration (NOAA) through the Jason programme for ocean obser\ation supporting an important oceanographic activity. Although ESA is a space organization established outside the framework

Intergovernmentul cooperation

31

of European treaties, the EU has become progressively more conscious of the increasing importance of space activities and several priorities of the Research Framework Programmes (climate and environment for example) have directly related to space issues. Furthermore, the success of the US Global Positioning System (GPS) and the development of navigation technique spin-off activities were strong incentives for greater EU involvement in space. In 2003, the European Commission prepared together with ESA a White Paper on European space policy, which was the basis for the framework agreement between the EC and ESA that came into force in 2004 and established a platform for cooperation between the two institutions and the building up of a comprehensive European space policy. They agreed to organize joint meetings at ministerial level, which constitute the 'Space Council'. Space is thus a good example of a field in which the E U succeeded in joining forces with a European intergovernmental organization (ESA) to establish a mechanism for the elaboration of a common policy piloted by a 'high-level space policy group' including a division of tasks: the EU will focus on space-based applications related to its policies (agriculture and transport, for example), while ESA prioritizes access to space (through 'Ariane'), space science and the development of space technologies. The GMES (Global Monitoring for Environment and Security) programme is a joint initiative and the EU is funding with Community funds (in particular the agriculture budget) the Galileo navigation satellite project with the technical support of ESA.

4 TO BE OR NOT TO BE: EUREKA AND EUROPEAN TECHNOLOGICAL COOPERATION Debates about Europe's technological competitiveness have been waged in political circles almost since the first steps taken towards European integration in the 1950s. Although the existence of a so-called 'technological gap' between Western Europe and the US was a matter of discussion during the 1960s, the main motivation behind the EEC decision to launch its first research programme was clearly the necessity of closing at least a 'funding gap' between US and European industrial research." The programme launched in 1974 was mainly devoted to energy, environment and health issues and was indeed modest (300 million ecus), but it was expanded in the 1980s after the adoption of the Single European Act (Grande and Peschke, 1999). In France, President Franqois Mitterrand and his government were, during the 1980s, engaged in the reorganization of the public research sector and of industrial sectors that had been nationalized in 1981, and

32

European science and rechnologj polic)

concerns were growing about the technological competitiveness of French industry in sectors as different as steel and electronics. Similar concerns were expressed regarding European industry as a whole in a French memorandum on the 'Relance europienne', which was published in October 1981 and called for strong cooperation between European industrial partners (Karsenty, 1987; Saunier, 2006). Similar views were expressed in the report Technologie, emploi et croissance (G8 research group, 1982), which the French government prepared for the June 1982 Versailles industrialized countries summit, and in the September 1983 document Une nouvelle etape pour /'Europe: un espace de I'industrie et de la recherche (Editions du Moniteur, 1984). It was at that time that the EC, through the efforts of Commissioner Davignon, promoted the very innovative research programme ESPRIT (European Strategic Programme for Research and Development in Information Technology), which was adopted in 1984 as part of the first Research Framework Programme and intended to develop cooperation between industry and public laboratories. It was feared. however, that the juste retour logic and the difficulty of rapidly mobilizing funding through EC procedures would jeopardize the initiatives to promote rapidly the European technological cooperation. This motivated France to propose in its 1983 memorandum the creation of several European agencies to organize research cooperation between industrial partners. This proposal first met with resistance. But it was paradoxically the US that provided Europe with the opportunity to reach agreement on new modes of technological cooperation when President Ronald Reagan launched the 'Strategic Defense Initiative' (SDI) in 1983 and created the SDI Office in March 1984 to fund the huge technological efforts required for building a shield against Soviet missile attacks with high flux lasers, optical detectors, space platforms, and so on. The US also took initiatives to involve European companies with SDI, which was perceived by political circles - in particular by Mitterrand - as a real technological challenge, which Europe had to face rapidly through a new initiative. After some hesitation - due to US pressure to enrol European countries into SDI within the framework of NATO1? the European Council adopted in June 1985 in Milan the principles of the EUREKA programme, which basically intended to catalyse 'joint ventures' between European partners (companies and public laboratories) to develop new market-oriented technologies. Seventeen countries (several of them, like Switzerland, not EEC members) and the EC agreed to support EUREKA, whose charter and first batch of projects were adopted in November 1985 by a ministerial conference in Hannover. EUREKA was thus an intergovernmental initiative taken in the margin of the EEC but with the clear -

Intergovernmental cooperation

33

purpose of pushing forward European integration through technological cooperation within a framework larger than the EEC, which at that time counted ten members only. No 'agency' was created. Instead a permanent secretariat was established in Brussels to supervise and coordinate the activities launched with the EUREKA label and funded by both partners (industry and public laboratories) and Member States. EUREKA quickly became a success, being well considered by companies and particularly small and medium enterprises (SMEs). Over time, a great number of projects were adopted and new countries joined EUREKA (37 Member States and the EU today).13 EUREKA'S political body is a bi-annual Ministerial Conference (chairship rotates every year among members), which sets the political course for the organization and its development. Key decisions are in the hands of a High-level Group, in which each member country has a representative, which labels and endorses new projects on which partners have agreed. An executive group acts as the executive arm of the conference for implementing decisions and preparing future decisions. EUREKA is functioning according to a 'bottom-up' mode: projects are initiated by partners, evaluated by the national EUREKA offices and after an assessment by the Brussels secretariat endorsed by the High-level Group. In spite of this emphasis on 'bottom-up' initiatives, some kind of sectorial organization has emerged within EUREKA. Thematic networks, which are called 'umbrellas', are focused on specific technology areas to facilitate the generation of new projects. 'Clusters' develop generic technologies of key importance for European competitiveness through large projects, primarily in the area of information and communication technologies at large and more recently in energy and biotechnology. A large cluster like 'Medea +', which was launched in 2001, has thus played a catalytic role in helping European microelectronics companies to stay in the group of world leaders. It built on the achievements of its forerunnersthe JESSI programme in microelectronics and the first 'Medea' - and was succeeded in 2008 by the cluster 'CATRENE' (Cluster for Application and Technology Research in Europe on Nanoelectronics), which is a fouryear programme with a prospect of €6 billion funding. EUREKA has probably had an important impact on the technological competitiveness of European industry, although - with the exception of clusters playing the role of 'virtual' agencies - it has lost part of its initial impetus over time. In recent years, France and the Netherlands have been the main players in the programme, while the UK is almost absent in terms of funding and German participation does not reflect the country's technological and industrial weight.14 The EU 7th Research Framework Programme (FP7) has introduced Joint Technology Initiatives (JTIs),

34

European science and technology policy

new instruments associating industrial partners and public laboratories for well-targeted objectives responding to industrial needs in key areas. JTIs have similarities with EUREKA clusters so that a full merger between EUREKA and the EC programmes is still a pending issue.

5

EUROPE IN SEARCH OF A EUROPEAN DARPA: THE DILEMMA OF COOPERATION IN DEFENCE

Research and technology have been used in military undertakings throughout Europe's history (Galileo, for example, was an 'advisor' to the Venice arsenal) (Salomon, 2007). The difficulties encountered in the 1970s and 1980s, when ESA and EUREKA were created, stemmed at least partially from the interference between industrial and political issues regarding in particular the role of military technology in relationship with missile forces. Going beyond the intergovernmental agreements that made possible the creation of ESA and EUREKA to establish European cooperation in military technology supposed a minimum of political agreement on the objectives of a European defence policy.15By changing completely the international situation, the end of the Cold War certainly opened the gate to new opportunities for Europe in the defence area, while the implosion of Yugoslavia caused Europe to get engaged militarily in the Balkans. Reinforcing peace and security in Europe was thus the central issue in the Petersberg declaration adopted by the Western European Union (WEU) countries in 1992, while the Maastricht Treaty introduced the concept of a 'Common Foreign and Security Policy'. This 'pillar' was supposed to support policy through intergovernmental cooperation within the EU and be the incubator for a future European defence policy.16 Further progress was rendered possible when France and the UK agreed in 1998 in Saint Malo in France to establish within the EU, an autonomous defence capacity based on reliable forces to face international crises. Although the necessity of enhancing European technological competitiveness in military technology was not yet an explicit objective, the issue was raised by several governments and the cooperative military research programme EUCLID (European Cooperation for the Long Term in Defence) was launched in the 1990s under the WEU umbrella. EUCLID intended to build industrial consortia for developing new military technologies with a philosophy akin to the EUREKA clusters but its yearly budgets were never higher than € 100 million. The US model was often referred to in political reflections about the European competitiveness issue. The Pentagon's research and procurements contracts and the Defense Advanced Research Projects Agency

Intrrgovernmentul cooperation

35

(DARPA) created in 1958 have, through long-term cooperation with academic and industrial laboratories, spearheaded the development of new technologies in several sectors (from lasers to computers) (Weinberger, 2008). For lack of political agreement on a defence policy, Europe - which is not a federal state - was unable to go in a similar direction, although the military industry sector had gone through an important reorganization at the beginning of the 2000s (the merger of several aerospace companies to create the European Aeronautic Defence and Space Company (EADS) being one example). The concept of a European Defence Agency (EDA) was thus in the air and the principle of its creation was adopted at the European Council in Thessaloniki in Greece in June 2003 while the agency itself was created in July 2004." One of EDA's four objectives is 'promoting research in the domain of defence and security with a view to strengthening Europe's industrial and technological potential' (European Union, 2004). All EU Member States except Denmark have joined the agency. In November 2007, EDA adopted its first plan for a 'European defence research and technology strategy' while in 2006, it had already approved its first investment programme with the still modest funding of €55 million. Twenty years after the end of the Cold War, Europe has - through the meanders mixing intergovernmental cooperation and initiatives from the Commission at last laid the basis for building a military research policy but its blueprint has yet to be implemented. -

6 A EUROPEAN PUZZLE: INTERGOVERNMENTAL COOPERATION IN THE BUILDING-UP O F THE EUROPEAN RESEARCH AREA Since the creation of CERN, intergovernmental cooperation has widened the spectrum for research cooperation in Europe. The creation of COST (Cooperation europeenne dans le Domaine de la Science et de la Technologie) in 1971 through an intergovernmental agreement was another initiative intended to establish a forum for European cooperation in the field of research on a broad front at a time when there was not yet an EEC research programme. COST initially played a useful role through workshops, networks and concerted actions but its influence has weakened over the years. The need to share costs and technical competence has been prominent in the history of European cooperation for building particle accelerators, research reactors, launchers like 'Ariane', satellites, telescopes and other sophisticated equipment. The necessity to pool efforts in order to undertake 'joint ventures' in the domain of science was certainly the prime motivation behind this advance of science policy in Europe.

36

European science and techno log^ policj.

cooperation was a positive signal for the building up of a new Europe and that strengthening Europe's scientific base in 'strategic' areas like particle and nuclear physics and later in space-related activities was an investment in the future (Andre, 2006). Looking at the network of the dozen existing European research organizations, can their strengths and weaknesses be identified (Papon, 2001,2004)? Most organizations are dedicated to building and operating large infrastructures and, sometimes, to the realization of experiments on a collaborative basis. This is clearly the case with CERN while ESA, EMBL and ECMWF also have dedicated programmes. One may observe that infrastructures shared by several partners have been successful either when there was a highly focused research community with well-defined scientific needs (which was the case, for example, with CERN and ESO) or else when it was able to offer a permanent outstanding service to a large and very diverse community (which was the case with ILL and ESRF). All large facilities built in Europe since 1954 share these characteristics and their performance is in most cases unchallenged in Europe and sometimes in the world (this is typically the case with CERN). The infrastructure system works according to a 'variable geometry' principle, which has one great advantage: the shareholders are countries volunteering to build an infrastructure to meet the needs expressed by communities in cooperation with national agencies, which will bear the financial consequences. Concerns about quality of service and performance of instruments are thus discussed at the highest level. Each operation requires the initial agreement of 'founding' countries (ten for CERN).I8 One must also stress that intergovernmental cooperation has not followed a unique 'model': CERN, ESA and ESO actually are international organizations operating according to international rules with permanent international staff, while facilities like ESRF, ILL and the European windtunnel are private entities operating as companies according to the legislation of the host country (French and German laws for those organizations) and with national labour law constraints.19 In all cases, decisions are made by boards in which Member States are represented by national delegates (who represent the ministries or national agencies financially supporting the organization) and in some occasions at ministerial level. The variable geometry principle gives political flexibility to the decisionmaking process, although in CERN and ESA, for instance, the political process has been difficult on those several occasions when member countries have put pressure to limit or downsize the budget. In spite of many difficulties and through a back and forth movement, Europe has succeeded through intergovernmental cooperation

I

in expanding the limits of European scientific cooperation on a scale without parallel in the world and CERN, ESA and ESO have become platforms for establishing cooperation with the US, Japan and China. Thus the LHC, the new particle accelerator at CERN, will by the end of the decade be the world facility for high energy physics. This policy, which has been followed on an ad hoc basis for half a century, has largely contributed to transforming research into a truly international venture in several domains and to building in parallel with the EC - the backbone of the European Research Area. One must not forget either that the vision of far-sighted science 'statesmen' has contributed to the success of this policy: Pierre Auger, Hubert Curien, Edoardo Amaldi, Reimar Lust and John Kendrew among others had the cleverness to mobilize politicians to launch the processes."' Despite those successes, there are some areas where intergovernmental cooperation has been weak or non-existent. Oceanography, atmospheric research and engineering sciences, to mention just three of them, are areas where as yet no European infrastructures exist (the cryogenic windtunnel in Cologne being an exception): oceanographic vessels, atmospheric research planes and large engineering facilities remain national. Technological research is, of course, a specific dimension of the European Research Area. ESA has certainly contributed to the emergence of a broad European partnership with regard to space applications thanks in particular to a continuous and productive partnership with Arianespace. In other areas, it is more difficult to assess the balance between success and shortcomings. EUREKA was created in the 1980s with the objective of enhancing European competitiveness in key areas like Information technologies and electronics. One must observe that the EC Research Framework Programme was launched with the similar objective of supporting so-called 'pre-competitive research'. Those programmes and clusters, with their different working modes ('bottom-up' versus 'topdown'), have doubtlessly played a catalytic role for European industry in sectors like information technology. The actual impact of EUREKA on other industrial sectors is less visible and probably smaller. SMEs were certainly attracted by EUREKA'S flexibility and the possibility of transnational cooperation with large companies and public laboratories, but over time, EUREKA'S attractive power has certainly decreased. In addition, the increasing influence of regional governments in research and innovation policies has certainly caused SMEs to seek local funding for their technological development. Lastly, one can actually doubt that support to SME innovation should today be a matter of political concern for intergovernmental cooperation. One can also be briefly reminded that until the creation of EDA in 2004, military research was a blind spot in -

38

European science and techno log^,policx

intergovernmental cooperation, immersed as it was in the maelstrom of European defence policy. A rather large dispersion can be observed in the size of the annual budgets of European research organizations. One can roughly estimate that about €5 billion are spent every year in research and technology activities through intergovernmental cooperation (including the EUREKA funding, which is mostly private), an amount that is approximately of the same order of magnitude as the yearly expenses of the Community Research Framework Programme. Generally speaking, intergovernmental cooperation has contributed to a large extent to putting in place important pieces of the European research and technology puzzle.

7 INTERGOVERNMENTAL COOPERATION AND THE FUTURE O F THE ERA By declaring that Europe should mobilize its talent to develop a 'knowledge-based economy', the Lisbon Strategy adopted in March 2000 has placed the European Research Area (ERA) at the very heart of this ambitious enterprise. However, when reflecting upon the prospects of ERA, one should not forget the lessons from the past: governments have accepted supporting initiatives in favour of European cooperation in the field of research as long as they were convinced that these were contributing to the modernization of Europe (as was clearly the case with the ECSC, the first Research Framework Programme and EUREKA) and to its presence in 'strategic' areas like nuclear energy and space. Intergovernmental cooperation and the various integrated processes launched at the initiative of the EC were the two tracks along which this policy has evolved over the years. If ERA is not able to demonstrate with convincing arguments that European-wide research efforts contribute to addressing important European issues, it will probably face a difficult future. One can draw a tentative list of these issues: energy and climate change are certainly top priorities, cities and transport in tomorrow's Europe are two other joint challenges, public health being a fifth one. One may also consider that enlarging the scientific and technical basis of a competitive agriculture in order to nourish a world with 9 billion inhabitants is an important objective for Europe. In several of these domains, the necessity of pooling resources and talent to define 'common policies' will probably become a necessity. The Treaty of Lisbon has thus identified research, space and energy as domains in which policies would become 'shared competences' between Member States and the EU." A permanent plea should also be undertaken in order to highlight that excellence on a large research front

Intergovernmental coo!,eration

39

is a necessary basis of European policies, this mission being entrusted to the ERC. One should also bear in mind that European integration has walked on two legs by evolving according to two different modes. Schematically speaking, the Rome Treaty has inspired a policy-making process based on an unofficial 'intergovernmental federalism' mode: common or integrated policies were defined and implemented (agriculture, trade, money with the euro, partially research with the Framework Programmes) without applying the law of juste retour. In parallel, governments have launched cooperation on a case-by-case basis with variable geometry schemes, in particular in the domain of science, a process that has given birth to 'intergovernmental organizations' like CERN and ESA. This process has rendered possible the participation in a European science policy of European countries that were not EEC (later EU) Member States (like Norway and Switzerland). These two visions of European cooperation have remained present until now, although one may consider that ESA is becoming more and more a 'federal' agency. Shared competences in science policy can be understood and exercised according to several schemes. They can be exercised through 'concerted actions' between the EU research Framework Programme and European research organizations or agencies like CERN, ESA, ESO and so on in order to use their know-how. Significant progress has recently been achieved in this regard through the cooperation agreement signed between the EU and ESA, which constitutes a framework for a comprehensive European space policy: the satellite navigation system Galileo provides Europeans with the first major opportunity for implementing this policy through cooperation between the EC and ESA (ESA being the technical architect of the project). Looking forward, one can envisage the transformation of ESA into a space agency of the EU having both scientific and technical missions and operating through a mix of funding from member countries and EU programmes (as the Research Framework Programme).** Another area of convergence is research infrastructure policy. The ITER project (International Thermonuclear Experimental Reactor), which aims at building an international machine for testing the experimental feasibility of fusion power, is a recent achievement of scientific policy, whatever one thinks of the real possibility of harnessing the energy of hydrogen atom fusion. The EU has been the European platform or 'voice' for promoting the project on behalf of European governments and an actual substitute to intergovernmental cooperation.?' There are compelling arguments for implementing a comprehensive research infrastructure policy in Europe, which was inaugurated by the founding of CERN. The European

40

European science and rechnology p o l i c ~

Strategy Forum on Research Infrastructures (ESFRI), which was created in 2002 at European Commission level to address those issues, has established a roadmap for future research infrastructures with a European dimension and the forum should play the role of incubator for future initiatives. These may result from direct cooperation between the EC and several governments through joint funding from the Research Framework Programme and national budgets. Another area of potential convergence is the policy to support high technology and, in particular, the mechanisms that have been set up by both the EC and EUREKA. There are obvious similarities between the EUREKA clusters and the Joint Technology Initiatives (JTIs). For example, the JTI 'Nanoelectronics Technology 2020' will be launched in parallel with the EUREKA cluster CATRENE. JTI and EUREKA clusters have clearly the same objectives: mobilizing industrial partners and public laboratories for enhancing European technological competitiveness. Concentrating the means of both the Framework Programme and EUREKA would make sense. This would probably imply EUREKA being transformed into a European agency supporting future JTIs or clusters according to a variable geometry me~hanism,'~ while support to SME innovation would no longer be an objective of European policy. Last but not least, while support to military research is a mission of the Defence Agency, research for 'security' issues may also be funded through the Research Framework Programme. One can probably consider that in the future, the two European tracks - intergovernmental cooperation and EU policy - will progressively converge in order to move forward the ERA train.?' This would suppose an Aggiornamento of the policy supported by the Research Framework Programme. The latter would be transformed into a political and budgetary platform ensuring the global coherence of European science and technology policy and mobilizing European funds to implement it. This would also suppose the existence of a governing body working closely aith the EC and the Council of Ministers and representing at the highest level the Member States. This would imply the complete transformation of CREST (Scientific and Technical Research Committee), which is now a purely advisory body. The construction of the ERA is a long-term enterprise, which assumes a permanent effort of coordination between national governments and the European political focal point that is the Commission representing the common interest of the EU. This construction also assumes an alliance between scientists and the interests of European citizens and their governments which, 50 years ago, rendered possible great ventures like CERN and the first European research programme launched by the ECSC.

Intergovernmental cooperution

NOTES The LHC has been built in an existing 27-km-circumference tunnel (dedicated formerly to the Large Electron-Positron Collider) across the French and Swiss border. In 2007, CERN's budget was about C H F l billion (€600 million). Four thousand five hundred European physicists and 2500 non-European researchers are, by and large, associated with experiments. The budgetary contributions of Member States are proportional to their GDP. The decision to build the LHC, a 7-TeV accelerator, was taken in December 1997 (see Llewellyn-Smith, 2007). EMBO was established as a non-profit organization under Swiss law. Its members arc distinguished molecular biologists. The EMBO and EMBL budgets amount to about El00 million. ESO is an international organization with headquarters in Garching (Germany). Its budget is around €120 million and it has a staff of 600 persons. The ALMA capital cost is about €900 million (twice the VLT cost). This is the origin of the Web, which was conceived by CERN physicists in order to exchange experimental data through a network of computers and telecommunication lines. Presently, I8 countries are members of ESRF. Its operation budget was about €80 million in 2007. Six hundred persons are working on the site, while several thousands of researchers come every year to perform experiments with the light beams of the machine. The US and the USSR benefited from German technical experience in rocketry. which was put into practice to start their missile programmes during the Cold War. The astrophysicist van Allen had embarked Geiger particle detectors on the satellites 'Explorer-I' and 'Explorer-.?', which detected a belt of radiation circling the Earth since called the van Allen belt. ESA's budget is about €3 billion (national contributions are proportmnal to G D P ) and it has a staff of 1900 scientists, engineers, technicians and administrative personnel. Its headquarters are in Paris. In order to facilitate the political compron~isefor the creation of ESA, it was decided that it would work with two kinds of programmes: the compulsory programme financed by every Member State (the scientific programme is compulsory) and the optional programmes funded only on a voluntary basis (or 'a la carte'), one of them, the 'Ariane' programme, being funded for 67 per cent by France, for 20 per cent by Germany and for the remainder by a few other partners. EUMETSAT has 21 Member States and a budget between €200 and 300 million. The OECD had pinpointed the weakness of industry research and development (R&D) funding in Europe and also the role played indirectly by the federal funding of technology in the US. Meanwhile, experts like the late Keith Pavitt from Sussex University had serious doubts about the existence of a technological gap. By the beginning of 1985. the US had issued a kind of ultimatum to their allies: they should respond to their offer of cooperation with SDI within two months. This prompted France to make concrete proposals to its European partners. Hubert Curien, French minister in charge of research, and J. Attali, Mitterrand's adviser, played an important role in the implementation of EUREKA. Ten projects were adopted in 1985, 99 in 1986 and between 100 and 200 every year since. The annual budget allocated by partners (states, companies and public laboratories) was about €1.3 billion in 2005 supporting 318 projects: France's participation amounted to €503 million, that of the Netherlands to €177 million, Germany's to €124 million and that of the U K to only €9 million. European integration had stumbled in the 1950s upon a major obstacle as the treaty creating the European Defence Community (EDC) was not ratified by the French parliament in 1954.

European science uiid technology policy In 1997, the Amsterdam Treaty restated the necessity of gradually establishing a common defence policy through intergo\emmental cooperation and not according to the traditional mediation of the EC. The Greek presidency made strong efforts for this adoption on the basis of a report on A Secure Europe in a Better World presented by Javier Solana (Solana. 2003). Security issues were then addressed by a group of experts. which met in 2004 at the initiative of Solana and presented its conclusions in the report Reseurd~for u Secure Europe (European Commission, 2004). As far as ILL is concerned, for example. the initial decision to build the reactor was taken by France and Germany but the U K later joined the Grenoble institute: the UK did the same with ESO. ESRF and ILL are both societes ciriles in France. They are managed according to private law (but without taxes). The influence and role of 'Americans' (European scientists who had emigrated before the war to the US) like Peter Weisskopf and Leo Szilard should not be forgotten. Article 2C of the Lisbon Treaty defines research as an area of shared competence but with a rather tortuous phrasing: 'In the areas of research. technological development and space, the Union shall have competence to carry out activities, in particular to define and implement programmes; however, the exercise of that competence shall not result in Member States being prevented from exercising them.' After several years of discussion, a political agreement was reached in April 2008 for the funding of Galileo from the EU budget: €3.4 billion partly with unspent funds from the agriculture common policy budget. Keeping in existence several national agencies with dedicated programmes would probably be necessary as. for example, in Guiana. CNES is operating the space base with its launching facilities and its security and safety requirements (see Blamont, 2008). ITER is being built in Cadarache (France). Its objectives and governance have been specified in an international treaty signed in 2006 (see Haignere and Bigot, 2008). An international fusion energy organization to build and operate ITER and a 'European common enterprise', located in Barcelona. to coordinate the European involvement in the project (Switzerland being a member of this organization) have been established. Article 169 of the European Treaty might provide a mechanism for such concerted activities between States and the Commission. JTIs will be autonomous 'common enterprises' with a variable geometry. The Eurostars programme, dedicated to the support of R&D in SMEs, is a recent and first example of such joint Eureka-EU programme (€90 million per year); it was adopted by the E U in June 2008. 'Reinforced cooperation' involving several Member States might be a way to initiate specific programmes and build new research infrastructures. This concept was introduced by the Amsterdam Treaty and included in the Lisbon Treaty.

REFERENCES Andre, M. (2006), 'L'espace europeen de la recherche: histoire d'une idee', Revue d'histoire de I'intPgration europienne, 12 (2), 1 3 1-50, Blamont, J. (2008), 'L'espace d u XXIesiecle sera prive', La Recherche, 416, 58-61. Blay, M . a n d E. Nicola'idis (eds) (2001), L'Europe des sciences, Paris: L e Seuil. Bossuat. G. (2006). 'Les c o o p e r a t i o n s europeennes p o u r la recherche scientifique e t technique', Journal of European Integration History, 12 (2), 5-10. Editions d u M o n i t e u r (1984), ' U n e nouvelle e t a p e p o u r 1'Europe: u n espace d e I'industrie e t d e la recherche', L'annPe politique, c!conomique et sociale en France 1983,220. E u r o p e a n C o m m i s s i o n (2004), Research for a Secure Europe - Report of the Group

Intergovernmentul coopemtion

43

uf'Personalitit~.sin the Field qf'Security Resmrch, Luxembourg: Ofice for Oficial Publications of the European Communities. European Union (2004), Officiul Journul oj'the Europeun Union, 17 July, p. 245. G8 Research Group (1982), Sommet de Versuilles, 4-6 juin 1982, nlpporf cir Monsieur le Prbsident de Iu Rbpuhlique fran@e uu Sommet dr puys industriulisi..~, Technologie, etnploi et croissunce, Toronto, www.g8.utoronto.ca/ summit/l982versailles (accessed 15 May 2009). Grande, E. and A. Peschke (1999), 'Transnational cooperation and policy networks in European science policy-making', Research Policy, 28 (1). 43-61. Haignerk, C. and B. Bigot (2008), 'Le programme ITER', Futurihks, 339, 29-43. Hermann, A , , J. Krige, U. Mersits and D. Pestre (1987), Historj! uf CERN: Launching the Europeun OrganizutionJbr Europeun Research, vol. I , Amsterdam: North Holland Publishers. Karsenty, J.-P. (1987), 'Analyse socio-economique de la cooperation scientifique et technologique europeenne-gknese et ambitions d'Eureka', PhD thesis: Universite Paris 1, Paris. Llewellyn-Smith, C. (2007), 'How the LHC came to be', Nature, 448 (7151) 2814. Papon, P. (2001), L'Europe de lu science c2t de lu technologic, Grenoble: Presses Universitaires de Grenoble. Papon, P. (2004), 'Research infrastructures in Europe', Minerw, 42 ( I), 6 1-76. Rossi, P. (1999), La nuscitu d e l b scienzu modernu in Europu, Rome: Laterza. Salomon, J.-J. (2007), Lrs scientijiques, Paris: Albin Michel. Saunier G. (2006), 'Eureka: un projet industriel pour l'Europe, une reponse a un defi strategique', Journal of Europeun Integration History, 12 (2), 57-74. Sebesta, L. (2006) 'Choosing its own way: European cooperation in space; Europe as a third way between science's universalism and US hegemony'?', Journul of Europeun Integrution History, 12 (2), 27-56. Solana, J. (2003) 'A secure Europe in a better world', European Council, Thessaloniki, http/eu.int/uedocs/cmsUpload/78397.pdf(accessed 16 May 2009). Strasser, B. and F. Joye (2005) 'L'atome, l'espace et les molecules: la cooperation scientifique internationale comme outil de la diplomatie helvetique (195 1-1983)', Relutions internationules, No. 121, 59-72. Tinjod, N. and B. Battrick (1998), The Archives of'tlze Europeun Space Agency, E. Amaldi Memorandum, Noordwijk: ESA, pp. 19-24. Weinberger, S. (2008), 'Still in the lead?', Nuture, 451 (7177), 390-93.

3. A European Research Area built by the Member States? Christian Svanfeldt This chapter centres on publicly funded intra-European cross-border research programme cooperation and its relation to the development of the European Research Area (ERA).' The focus is on research programme cooperation between Member States taking place outside the frame of intergovernmental research cooperation schemes including programmes like COST2and E U R E K A b r infrastructures like the European Organization for Nuclear Research (CERN) and the European Molecular Biology Laboratory (EMBL). Although cross-border research programme cooperation has been rather insignificant so far in terms of frequency and volume of funding, there are signs that the situation may be changing. There is an explicit recognition of the need for scale and scope in order for research to be effective and for national research efforts to be well connected to European or worldwide research activities. The key questions addressed in this chapter are: what are the rationale for, drivers of and barriers to public research programme cooperation between EU Member States? And can such cooperation be a viable Member State-driven strategy to construct ERA? The ERA-NETS are exploited as sources of information to understand better the current situation and recent trends. Nordic research cooperation and in particular the Nordic Research and Innovation Area (NORIA) are looked at as cases of non-EU-led multilateral research cooperation with objectives similar to those of ERA.

1 THE RATIONALE FOR RESEARCH PROGRAMME

COOPERATION BETWEEN MEMBER STATES IN RELATION TO THE ERA OBJECTIVES The rationale for public investment in knowledge in general and science and technology in particular is multiple and relates to different socioeconomic objectives. Research investment is recognized as a key element

A European Reseurch Area built by Member Srutes?

45

for economic development and competitiveness and the chief response to the challenges of globalization. Although there is a great diversity in how Member States' research policies are formulated, most deal with similar issues: how to achieve critical mass of effort in currently important areas that require it . . .; how to develop flexible inter-disciplinary structures and the 'critical diversity' they require; how to sustain expensive infrastructure; how to develop creative interactions with business; and how to embed competitive structures that stimulate excellence." Cross-border cooperation is increasingly seen as a n instrument necessary for adequately dealing with these issues, and although the European dimension has been rather weak in most Member States' research and innovation strategies, there is a growing awareness amongst national policy makers a n d research funding bodies that national efforts are often subcritical in terms of scale, scope and depth. Policy and strategy documents speak of the need t o better exploit global scientific resources by, for instance, stimulating research cooperation with foreign centres of excellence, promoting the outward mobility of young researchers and the inward mobility of senior researchers, influencing the direction and content of European research policy t o become better suited to national needs and so on. The creation of centres of excellence to concentrate resources and achieve higher impact is part of these strategies to make national resources more visible and attractive a t the international level.

1.1

The European Research Area

ERA can in some ways be conceived of as a n internal market for knowledge. According to the 2007 ERA Green Paper: 'the ERA concept combines: a European 'internal market' for research. where researchers, technology and knowledge freely circulate; effective Europeanlevel coordination of national and regional research activities, programmes and policies; and initiatives implemented and funded at European level . . . The European Research Area that the scientific community, business and citizens need should have the following features: an adequate flow of competent researchers with high levels of mobility between institutions, disciplines, sectors and countries; world-class research infrastructures, integrated, networked and accessible to research teams from across Europe and the world . . .; cxcellent research institutions engaged in effective public-private cooperation and partnerships, forming the core of research and innovation 'clusters' including 'virtual research communities', mostly specialised in interdisciplinary areas and attracting a critical mass of human and financial resources; effective knowledge-sharing notably between public research and industry, as well as with the

46

Europeun science und techno log^. policj

public at large; well-coordinated research programmes and priorities. including a significant volume of jointly-programmed public research investment at European level involving common priorities, coordinated implementation and joint evaluation; and a wide opening of the European Research Area to the world with special emphasis on neighbouring countries and a strong conimitment to addressing global challenges with Europe's partners'.' The ambition of ERA hence goes further than promoting excellence and the free flow of knowledge. Research is also increasingly being seen as a key element in addressing European challenges ('Grand Challenges') and the coordination of and priority setting among research investments and efforts at European level are thus given a higher importance. 1.2

Bilateral and Multilateral Research Cooperation Policies

R&D cooperation policies serve different purposes, from fulfilling strategic research-related objectives to supporting other external relations p ~ l i c i e s .Historically, ~ Member State bilateral or multilateral research programme cooperation has been rather insignificant in terms of frequency and volume of funding. A mapping of the bilateral research and development (R&D) agreements of the EU-15 in 2001 showed that most agreements were goodwill oriented and that in general, no funding was associated with these agreements. The goodwill agreements mainly served purposes - such as forging stronger cultural or economic ties - other than the direct advancement of science. The goodwill agreements also mainly aimed at facilitating R&D collaboration in general in broadly defined areas whereas the strategic agreements had more narrow scientific objectives of more strategic importance, for instance, joint facilities and joint research centres. Most cooperation agreements were directed at researcher mobility.' Research cooperation agreements do not necessarily involve programme cooperation. A study from 1999 indicated a very low level of cross-border cooperation taking place in national R&D programmes within the EU;8 a 2005 update, which surveyed 127 national programmes within Europe, showed a very similar situation. Although 80 per cent of programmes spent money across borders, this was mainly to fund national participation in international programmes. The benefits of cooperation were increased research capacity and higher-quality results as well as, to a lesser extent, the opening up of new markets and the faster exploitation of results. Lower costs were perceived as a benefit in 10 per cent of the cases. Almost none of the programmes funded foreign participants. The main reasons for funding foreign participants were to make up for shortfalls in the quantity or quality of national research capacities; to bring in

A European Research Area built by Member Stutes?

47

foreigners to foster the international competitiveness of the research community and to match aspects of the internationalization of business." One of the few examples of 'common pots'1° with no 'justt' retour'" is the Nordic cooperation centred on the Nordic Council of Ministers. Yet the Nordic countries have also been relatively active in terms of research cooperation outside the frame of the Nordic Council. One example is NORDITE. l 2 This is a continuation of previous Finnish-Swedish cooperation in ICT, which was extended to also include Norway and operates through joint calls with virtual pot fundingL3Other examples include the Finnish-Swedish co-funded Wood Material Science Research Programme (2003--06);L4 and the coordination of Nordic food research programmes through a Nordic Research Area Net (NRA-NET)."

2 DRIVERS OF AND BARRIERS TO RESEARCH PROGRAMME COOPERATION BETWEEN MEMBER STATES - EVIDENCE FROM THE ERANETS

The ERA-NET scheme was conceived under the 6th Framework Programme (FP6) as a bottom-up mechanism for the coordination of national (and regional) research programmes. The scheme supports the networking of research programme owners and managers; activities range from information exchange to joint calls. ERA-NETS have been welcomed widely by the research community, which appreciates their bottom-up nature and 'variable geometry' character. An evaluation of the ERA-NET scheme by the end of FP6 found that they fulfilled a real need and helped to overcome barriers to c ~ o r d i n a t i o n .By ' ~ the end of 2007, the 71 ERA-NETS launched under FP6 had resulted in more than 80 joint calls totalling over €800 m i l l i ~ n .As ' ~ part of the deliverables, most ERANETS have mapped their research area in terms of funding, coordination and fragmentation. Many have also looked into the barriers to increased cooperation and coordination including barriers to joint calls and joint programming. The following sections build on evidence gathered from the ERA-NETS.

2.2. Complementarities, Duplication and Critical Mass Science and technology fields differ significantly in terms of funding, degree of coordination (or fragmentation) and performance (see Figure

48

European science and techno1og.v policy

3.1). No straightforward linear relation exists between these characteristics since each science and technology (S&T) field is unique and the benefits and costs associated with S&T cooperation vary considerably across field^.'^ While the presence of complementarities and commonalities among research funding programmes may constitute a good starting point for multilateral cooperation, the dividing line between duplication and overlap on the one hand and similarity and commonality on the other hand is not easy t o draw." The existence of very similar issues within R&D programmes may simply reflect the importance and priority aspect of those issues for the research community or public authorities hut it may also indicate that there is a critical lack of scope.

Noter: The X-axis represents a qualitative assessment of the degree of coordination among Member State (MS) research programmes and of funding and institutional fragmentation. The Y-axis presents the logarithmic ratio of public R&D investment in Europe fMS+European Commission (EC)) compared with the US. The size of each field is directly proportional to the amount of European public funding (MS + EC). Main sources are evaluations, scientific reports. etc. of different S&T fields. data from New Cronos (for example. government appropriations or outlays for R&D [GBAORD]) and US government information.

Source: European Commission (2008b).iU

Figure 3.1 A diverse European research landscape: degree of European coordination and European funding volume (absolute and relative to the US) by S&T area

A European Research Areu built by Member States?

2.3.

49

Main drivers of cooperation

Avoiding duplication, achieving critical mass in research projects or programmes by pooling limited resources, exchanging good practices and accessing outside expertise are generally widely accepted as benefits of cooperation (see Table 3. I).?' Although there is no clear-cut evidence that national research programmes lack critical mass or that the European research landscape is fragmented, specific research fields from medicine to energy provide examples that illustrate the potential benefits of cooperation for the achievement of critical mass." This is not least the case in the field of shared infrastructure." There are several examples of areas for which the ERA-NETS have made a difference in providing the necessary scale and scope, for example critical mass of R&D funds and resource^,'^ researcher^^^ or actors2%r in terms of impact.

2.4 Barriers to cooperation The barriers to cooperation are many. They range from softer barriers such as lack of experience with cooperation and reluctance to cooperate with strangers to barriers of a technical nature such as incompatible programming routines and cycles and to legal barriers to cross-border funding (Table 3.2).

2.4.1 Cultural, linguistic and mental barriers At the project level, differences in culture and linguistic obstacles may make finding project partners and writing proposals more difficult.'%t the programme level, the same kinds of obstacles also make it more difficult to define common objectives and design and implement programmes. At the political level, stakeholders need to be convinced that win-win situations exist. Fears of loss of influence and control make decision-makers unwilling to open up national programmes, as do fears of 'unjust returns"" and the difficulty of defining the national added-value of project results." Beside a general reluctance to engage in cooperation due to a 'strangerdanger' attitude," there can also be a real need for building trust in very sensitive fields, such as security research, which may take a very long time. 2.4.2 Different objectives, different timing, different procedures Research programme cooperation adds additional complexity to the design, management and funding of programmes. It may also add additional complexity for the researchers targeted by the programmes. Defining the content and timing of a joint calls1may be a challenging task as national

50

Table 3. I

European science und tec/znolog>~ polic~.

Drivers of cooperation iduntiJied by ERA-NETS

Drivers

Examples from ERA-NETS

Restructuring of research at European level

Creation of lasting networks not only of researchers but also of ministries and agencies (e.g., E-RARE. EUROTRANSBIO. CORE ORGANIC) Bundle scattered national funding into durable 'networks' of resources in contrast with direct but temporary EU research funding (e.g., CORE ORGANIC) Improvement of scientific/technological quality (e.g.. AMPERA) Access to a larger pool of researchers (e.g., INNER) Avoid duplication; encourage patterns of specialization ( e g . PATHOGENOMICS?', MARINERA, CORE ORGANIC, EUPHRESCO) Expensive research facilities not available in all countries (e.g.. CORE ORGANIC) Efficient allocation of resources (e.g.. HY-CO, MARINERA) Share risks, costs and skills related to innovation (e.g.. EUROTRANSBIO, CORNET) Complementarities, fill thematic gaps and possibility to set up a cohesive programme of research (e.g.. NEW OSH ERA, EUPHRESCO, HERA) Development of transnational multidisciplinary approaches (e.g., HERA, EUROTRANSBIO) Achievement of critical mass to make progress in large and diverse research areas (e.g., BIODIVERSA, E-RARE) Increased dissemination of research and of knowledge (e.g.. NORFACE, SAFEFOODERA, HERA) Raise public awareness (e.g., AMPERA, SAFEFOODERA) Previous cooperation and joint activities among partners (e.g.. NORFACE, MARINERA) Gaining a better position in the international market (e.g.. CORNET) Speed up internal cooperation within a country (e.g.. EUROTRANSBIO)

Benefiting from foreign expertise Better allocation of finite resources

Achieving scale and scope

Dissemination of results Other drivers

Sources: Table based on a review of available ERA-NET reports and deliverables. The reports from which the examples are drawn can be found in the bibli~graphy.'~

A European Reseurch Area built hj' Member Stutes? Tuhle 3.2

51

Overvie\t, o f harriers to cooperution identified hy E R A - N E E

Cultural

Barrier Examples

Illustrations

Differences between new and old Member States, smaller and bigger countries Reluctance to cooperate with unknown partners, difficulty of building trust (especially in sensitive fields) and persuading stakeholders of mutual benefits

'While in the U K and in the Netherlands research is almost entirely performed in the universities . . . and funding and strategic priorities are planned and pursued through funding agencies with the contributions of scientists and experts from the universities, in France, Germany, Italy and Spain, the systems are more complex. Besides universities, scientists working in research institutes provide a strong contribution to astrophysical research""

Lack of cooperation experience at international level (especially among funders and programme managers) Linguistic obstacles

PoliticalAdministrative

'There are only few personal contacts, sometimes funders even have not heard each other's names. In some cases funders of one nation were strangers to each other beforc the SNOWMAN Think Tank workshop"' Lack of political will to open 'Various authorities and up national programmes by organisations are responsible decision makers fearing loss for flood risk management ranging from the local, of control, 'unjust return' and too much EU-influence through the regional up to the national level on national funds (vertical distribution of Different time horizons of responsibility). . . . Usually, politics and policies the more broadly spread Difference in research the responsibilities the more priorities, due to ditferent complex are the consultation national interests processes for preparing Diverse and country and implenienting research specific rules and legislation programmes"' (employment conditions, taxes and so on)

52

European science and technology poliq

Table 3.2 (continued) Barrier Examples

Programmelevel

International cooperation is dealt with by national government departments or agencies other than the national research funding agencylies Administrative burden in national organizations Absence of mechanism for joint calls Diverse research programme priorities Management level fragmentation within the national boundaries Diversity of partners (ministries, agencies, research councils and so on) Differences in skills, scientific competences and research potential available in partner countries Different timing of national programme life cycles Differences in funding principles; difficulty concerning funding international partners and setting up a real common Pot Funding fragmentation at country-level Different intellectual property rights procedures (and reticence to share previously performed research) Differences in peer-review

Illustrations

'Slightly more than half of the programmes (53.3%) are potentially open for international participation in projects .. . . But in reality this is often restricted in that way. that the call for proposals, the programme information, applicant forms, etc. are only available in the national language. Consequently mainly national researchers submit proposals for these programme^'^‘ 'Funding is provided through several channels. A fraction is directly managed by research institutes; while another fraction is provided by the government through a nationwide competition, often open to all scientific disciplines. This is done either by specific agencies . .. or directly by the relevant ministers'46 'Different research funding agencies have different rules for the safeguarding of and support for Intellectual Property Rights. . .. This is an issue of greatest relevance

A Europeun Reseurrh Areu built by Member StrrtesY

Tublt. 3.2

53

(continued) Barrier Examples

Illustrations

Difficult to preassess added-value of cooperation projects Difficulties in project management (linguistic barriers, lack of experience of participation in international consortia, difficulty in finding partners)

for the researchers, research

institutions and research funders participating in cross-border activities and must be addressed with great urgency'47

Soutws: Left column: the list of barriers draws upon a report by the SURPRISE ERANET4X;right column: illustrations from various ERA-NET reports (see bibliography).

programmes differ across Member States in their set-up and organization: programmes have different life cycles and their objectives are set in relation to national strengths and priority areas ranging from the development of science to support for industry and increased competitiveness. Different time horizons also exist at the political level, which makes it more complicated to set up programmes needing additional funding.j4 In addition, international cooperation is often dealt with by dedicated departments at agency or ministerial level and these may not always have the same objectives and priorities as the departments dealing with thematic areas. National programmes may also suffer from internal inefficiencies, such as slow decision-making, administrative complexity, unclear evaluation criteria and limited funds, reducing their scope and the possibilities to include more innovative research.35Such problems obviously have adverse effects on the possibilities for international cooperation. The diversity of national funding scheme^,'^ actors, rules and procedures for budgets, peer review, programme management, intellectual property rights (IPRs), etc., makes programme cooperation more d i f f i ~ u l tJoint . ~ ~ programmes thus need to accommodate a diverse set of national rules, practices and legislation, including social security, tax, employment, and fiscal rules.

2.4.3 Different resources available Differences between (potential) partner countries in terms of scientific competence and research potential and differences in terms of the scale of available resources (both financial and human) make international cooperation difficult.3RThe reciprocal opening up of national programmes may

54

European science und teclitiolog~~ polic!,

demand the creation of a 'common pot' for funding and a new set of rules to allow for the funding of foreign participants, which might in turn also demand a change in national legislation. Funding principles and decisionmaking procedures have to be agreed upon and international evaluation teams have to be set up and managed.j9 In addition, often time-consuming new administrative and reporting requirements have to be put in place.-'O The increased administrative burden cannot always be justified and countries with scarce resources in terms of available funding and/or personnel - especially smaller countries from the new Member States - may not be able to engage fully in joint programmes. The EU Structural Funds are an important source of research funding in these countries and the bulk of it cannot be used for international cooperation under current legislation. Small countries may also benefit less from common pots," and too far-reaching international coordination may have a negative impact on national coordination and competition.

3 NORDIC COOPERATION AS A MODEL? The Nordic countries have a long history of cooperation in a range of different fields including research and innovation. An important part of this cooperation is administered centrally at the Nordic level by the Nordic Council of Ministers but there is also a range of other initiatives that sometimes involve just two or three countries. Finland's Tekes and Sweden's VINNOVA have, for instance, a long tradition of cooperation in technology related programmes like Wood Wisdom in wood technology and INWITE, EXITE and NORDITE in ICT. Often only specific areas or aspects of national research programmes are concerned and funding varies between nationally based and 'virtual pot'. There are also some pan-Nordic networks, such as the Nordic Forestry, Veterinary and Agricultural University Network (NOVA),4yand the Nordic Cancer Union,'' which has a research budget of about f 1 million per year, and which uses a real 'common pot'." Cooperation in the field of research and innovation under the Nordic Council of Ministers became an object of debate during the second half of the 1990s and at the beginning of the 2000s. The debate was influenced by developments at EU level and the fact that Nordic cooperation was being scrutinized in the light of Finland and Sweden joining the EU. Efforts were made to define the added-value and rationale of Nordic research cooperation. A milestone was the White Paper on a Nordic Research and Innovation Area (NORIA),52which was presented in late 2003 and led the Nordic Ministers for Education and Research to agree on the overriding goals for Nordic cooperation:

A European Research Area built b~ Member Srur~s?

0

55

to give the Nordic countries a leading position in competence development and research, with an emphasis on the development of human resources; to stimulate the exchange of experience and development of research and innovative Nordic initiatives in this field; to strengthen and develop Nordic co-operation to promote common Nordic interests at the international level, with particular emphasis on the adjacent areas.

Nordic research cooperation involves Nordic research institutions, fixedterm research programmes, Nordic Centres of Excellence (NCoE), grant schemes and the coordination and planning of major infrastructure investments. The overall objective is to promote research of the highest possible international quality. The Ministers of Education and Research are responsible and are aided by a Committee of Senior Officials, which prepares work and meetings. Central to these activities is the Nordic Research Board NordForsk (NF).13 N F was set up in 2005 as a direct result of proposals included in the NORIA White Paper.54Its overriding primary task is to promote efficient cooperation between the Nordic countries in terms of research and researcher education. It thus replaces and incorporates the functions of the Nordic Research Policy Council (FPR)55and the Nordic Academy for ~ took over the Centre of Excellence Advanced Study ( N O ~ F A )N. F~ also programme initiated by the three joint committees of the Nordic research councils. NF is one of the three pillars of NORIA (also referred to as the triple helix), the other two being the Nordic Innovation Centre (NICe) on the one hand, and the Nordic Council of Ministers for Education and Research (MR-U) and the Nordic Council of Ministers for Trade and Industry (MR-N) on the other hand. NICe is responsible for innovation support at the Nordic level, which In addition to NICe and NF, Nordic also includes applied re~earch.~' Energy Research (NEF)'"unds research projects that demonstrate Nordic utility in the areas of: integration of energy markets; renewable forms of energy; energy efficiency; the hydrogen society and the consequences of climate change for the energy sector.59The ambition of NEF is to create a NORIA in energy ."I Although Nordic cooperation under the Nordic Council of Ministers is relatively limited in terms of direct funding from the Nordic Council €30 million for research and education cooperation and €16 million for innovation, energy and regional policy cooperation in 2007 it has been effective in some areas, notably energy, where Nordic positions in Europe have been strengthened. During an informal meeting of Prime Ministers in -

56

European science und trchtzolog~polic).

Punkaharju in Finland in 2007, it was acknowledged that Nordic cooperation needs to be better equipped to meet the challenges of globalization and that a Nordic globalization initiative should be taken with research and innovation featuring amongst the prioritized areas. It was decided that €8 million should be set aside for this purpose in 2008.61As a result. 14 initiatives were taken in various policy areas, one of which was an initiative for Nordic Excellence in Research. A very ambitious proposal aimed at addressing energy, environment and climate issues, from causes and effects to solutions, was put forward in the spring of 2008. The proposal went as far as proposing an integrated approach in which NordForsk, NICe and NER would act together and draw heavily on additional private and public funding, including E U - f ~ n d i n g Further .~~ discussions led to a revised and reduced proposal being adopted in late October 2008. It contained five pilot areas and total direct funding would amount to around €47 million, of which €12.5 million would come from the three institutions NF, NICe and NEF and an additional €20 million of direct funding from the Nordic Council of minister^.^^ As such it can be seen as an instance of pioneering multilateral research and innovation programme cooperation, coordinating multiple types of actors and instruments. including also the EU leveLM It should be noted that many of the barriers to cooperation identified by the ERA-NETS are also visible within the context of Nordic cooperation, as many initiatives rely on both Nordic and national funding

4 CAN MEMBER STATES' BILATERAL AND

MULTILATERAL RESEARCH COOPERATION BE DEVELOPED FURTHER? There is a growing awareness that national resources alone do not always suffice and that there is a need to be well connected to European or worldwide scientific resources. A result is that international research cooperation is considered of more strategic importance. This is clearly the case for Nordic cooperation, for which the shift has already been translated into concrete actions. At a more general level, the need to strengthen scientific ties with the outside world is expressed in most countries' policy papers, although the emphasis is put mainly on exploiting European and international scientific resources for the purpose of strengthening national systems. An essential question is whether the advantages outweigh the inconveniences and additional costs associated with cooperation for the actors involved. The benefits may be less direct and more difficult to appropriate for national actors and specific national objectives may not be met.

A European Research Area built by Member States?

57

It is obvious that some areas are more open to cooperation than others. Societal problems (such as road safety, quality of education, child care, employment and so on) often present possibilities for mutual learning from best practice or field trials in different communities. There are few competing or conflicting interests, results can be transferred and transposed easily and significant gains can be made for cooperating partners. Cooperation becomes more complicated in areas with stronger national or industrial interests. It is also obvious that the complexity of cooperation increases with the number of countries involved. More far-reaching cooperation will be heavily reliant on a non-negligible harmonization effort of procedures and regulation related to research programming, calls and evaluation. It will also depend on successfully building trust and learning how to work together towards shared objectives, something that may take considerable time. The European Commission's recent joint programming initiative6hay become a decisive stepping-stone in the development of more strategic bilateral and multilateral cooperation. Its importance may well go beyond the areas officially singled out for joint programming as national consultation processes and negotiations between Member States may well lead to spin-offs in terms of cooperation outside of the official joint programming initiative, possibly with fewer countries involved but not necessarily less ambitious. If successful, joint programming may present a wider model for the organization of European research. Nordic cooperation presents another interesting, more integrated model, in which fewer countries are involved but with objectives that are wider and go further. Alternatively the organization of ERA could be based on several smaller ERAS, NORIA being one amongst many. The Nordic model implies a certain degree of homogeneity among the participating countries in terms of culture, values, priorities and resources, which could be more easily emulated by similar smaller groupings of countries (for example. the Visegrad grouph7or at regional level, the GrossIGrande Regionh8)rather than at EU-27 level. Such intergovernmental cooperation neither would nor could replace EU-level action but would rather provide platforms for more effective and efficient interaction between the national, intergovernmental and EU levels. This approach might be able to reconcile ambitions of excellence with more balanced regional development and better embed research policy into a wider policy context. It may also be more viable in a heterogeneous and possibly further expanding Europe. The disadvantage compared with the lighter ad hoc model of joint programming could be reduced dynamism and unsuitability to address particular issues due to a more fixed geometry. However, these two models could very well co-exist and be mutually reinforcing.

58

European science and technology policy

CONCLUSIONS - AN ERA BUILT BOTTOM-UP BY THE MEMBER STATES? Could ERA be built on a bottom-up, voluntary basis by the Member States? Looking at the current state of cooperation between Member States, an answer is not that easily given. Member States' research programme cooperation has historically been low. Comparing the barriers to cooperation with the drivers, the barriers outnumber the drivers by far. The Nordic countries have been in the forefront but it is only recently that this cooperation has become strategic and funding volumes have until now been relatively low. There are few signs that a voluntary approach would be as far-reaching as the ERA ambition in terms of specialization and the creation of European centres of excellence. However, there are clear signs that cooperation and to some extent coordination are considered more important for strengthening national research and innovation systems. The ERA-NET scheme has provided a platform for learning and cooperation between research programme owners and managers. Its success has meant that the basis for cooperation is larger now and since cooperation to a large extent is a learning and adaptation process, the balance between barriers and drivers is likely to shift in a positive direction. The combination of a more fertile ground for cooperation with the recognition of the increasing need to rely on external scientific and technological resources will lead to-a much higher degree of cooperation in the future in terms of both volume and intensity and depth. In addition, the European Strategy Forum on Research Infrastructures (ESFRI) is playing an important role in the development of new pan-European research infrastructures. It is difficult to assess what the situation would have been in Europe without those initiatives hut given the barriers to cooperation that ERA-NET partners have encountered and dealt with and the perceived benefits stemming from this cooperation, the process towards more cooperation has probably gained speed and momentum. The answer to the question whether ERA could be built by Member States' bilateral and multilateral research cooperation efforts is probably a reserved yes; yes, but not without the helping hand of EU research policy as initiator and facilitator of cooperation. But as the intergovernmental and EU levels are interdependent and complement and reinforce each other - especially so in an increasingly globalized setting - a reliance solely on Member States' efforts would not make sense.

A European Reseurch Area built by Member S r ~ t e s : ~

NOTES This chapter has benefited from input from Claudia Brugia. COST European Cooperation in Science and Technology is a European intergovernmental network for the coordination of nationally funded research activities (www. cost.esf.org). EUREKA is an intergovernmental initiative supporting industrial R&D (www.eureka. be). LERU (2007). p. 14. European Commission (2007). p. 2. GLOSPERA (2004). Technopolis Group (2001). Technopolis Ltd. (1999). Optimal Ltd.. VDIIVDE Innovation and Technik GmbH (2005). Funding from several countries put together into a single pot. Ju.stc. retour. refers to the expectation of a particular country to receive out of a programme for its own rescarchers, institutions, ctc. an amount of subsidies proportional to its contribution to the programme's budget. VINNOVA, the Research Council of Norway and Tekes. CJnder virtual pot funding, national funds are used to fund own nationals only. For instance, in a joint Finno-Swedish project with virtual pot funding, the Fmnish researchers would get their funding from Finland and the Swedish ones from Sweden. Run by VINNOVA (Sweden) and Tekes (Finland), see Nordic Forest Research Cooperation Committee SNS (2002). In 2002, a NRA-NET on Processing for Food Safety was established through a specific support action funding a Nordic ad koc group of programme managers and programme makers. In 2003, this Nordic action was extended to EU level through the food safety Specific Support Action ERA-NET 'PROFORSAFE' [SAFEFOODERA ERA-NET, 2004. pp. 781. ERA-N ET (2006). European Commission (2008a), p. 22. European Commission (2008b), p. 5. An example is the MARINERA ERA-NET, which found over 50 percent of identified research funding programmes to have eight specific topics in common (MARINERA ERA-NET, 2007a and b). European Commission (2008b), p. 5 These arguments are frequently presented by ERA-NET partners across different research fields. Critical mass in terms of a minimum number of patients, data, biological materials, expertise and facilities is often of crucial importance in medicine (E-RARE ERA-NET, 2007). Food security and organic farmingconstitute other research areas in which there is room for action in order to achieve the necessary critical mass (CORE Organic EKANET, 2006a and b and SAFEFOODERA ERA-NET, 2004, p. 7). This argument was stressed in the field of Antarctic research: 'the total number of the international Antarctic facilities is 65. Twenty-eight of these are European stations. including seven Russian stations [. . .] A common appropriate use of these resources could overcome the cost of carrying out research in Antarctica. which is several times higher than in the Arctic due to the larger scale of infrastructural and logistical support required. These integration potentials are not yet adequately exploited and the European countries act independently on the basis of specific scientific and logistic interests' (EUROPOLAR ERA-NET, 2008, p. 30). This is particularly true for Small and Medium Enterprises (SMEs): 'Participation in trans-national collective research projects enables involvement in the stages of developing new technology and the creation of harmonised norms or standards. while providing a better position in international markets' (CORNET ERA-NET. 2007. p. 12). -

60

European science a n d rechnologj. policy For instance, Baltic Sea research is carried out in all riparian countries of the region by several research institutes, vessels and universities and through hundreds of projects and scientists (BONUS ERA-NET. 2008). An example is the BIODIVERSA ERA-NET. which brings together national funding organizations of different orientation traditionally not cooperating very much. (BIODIVERSA ERA-NET, 2007). TRANSPORT ERA-NET (2005). SURPRISE ERA-NET, 2007, p. 7 and AMPERA-NET (2006a). p. 8. European Commission 2006, p. 2. SURPRISE ERA-NET 2007, p. 6. SURPRISE ERA-NET, 2007, p. 6. European Commission 2004, p. 3. See also Inman (2007), Kristensen (2007). ORGANIC REVISION (2006). Petitet (2007) and Vanhemelrijck (2008). ASTRONET ERA-NET. 2006. p. 10. SNOWMAN ERA-NET, 2005. p. 28. C R U E ERA-NET. 2007, p. 16. SNOWMAN ERA-NET, 2005. p. 14. ASTRONET ERA-NET, 2006, p. 3. HERA ERA-NET, 2007, p. 43. SURPRISE ERA-NET, 2007, pp. 4-8. SURPRISE ERA-NET, 2007, p. 6 ACENET ERA-NET, 2004, p. 4. A good example is ASTRONET (ASTRONET ERA-NET, 2006). SURPRISE ERA-NET, 2007, p. 7. AMPERA ERA-NET (2006a). These barriers were highlighted at a TAFTIE ERA-NET workshop in Malmo (June I . 2006) as indicated in the report of SURPRISE ERA-NET. 2007. p. 8. This barrier was hightlighted at a ERA-NET meeting held in Brussels (06-05-30) as indicated in the report of SURPRISE ERA-NET. 2007. p. 8. SURPRISE ERA-NET. 2007, p. 8. http:Nwww.nova-university.org/ http://www.ncu.nu/ Nordisk Ministerrid (2006). Nordisk Ministerrad (2004). www.nordforsk.org The funding by NordForsk amounted to around €15 million in 2007. The management committee for research within MR-U, the Nordic Council of Ministers for Education and Research. FPR was established in 1982 by the Nordic Council of Ministers and was the advisory body to the Council of Ministers on long-term research policy issues. The task of FPR was to promote Nordic co-operation within research and researcher training and to advise on research policy issues of relevance to the Nordic region. NORFA had the collective responsibility for Nordic Council of Ministers efforts for researcher education and researcher mobility. It advised the Nordic Council and provided financial support to various forms of research education co-operation in the Nordic region and neighbouring areas (cg research networks. research training courses, mobility subsidies and guest professorships). NICe supports Nordic knowledge platforms in areas such as innovation policy; creative industries; biotechnology: micro- and nanotechnology: technology foresight: food safety; innovative building & construction; environmental technology; and venture capital. The total project portfolio of NlCe consists of approximately 120 ongoing projects and networks. f 10.5 million of funding was granted in 2007 (see http://www. nordicinnovation.net). http://www.nordicenergy.org

A European Research Area built hy Memher StutesY

61

The bulk of the N E F budget comes from the Nordic Member States (as opposed to central funding from the Nordic Council of Ministers). Granted project funding amounted to €2.4 million in 2007 (see http:I/www.nordicenergy.net/). Some Nordic research institutes also receive funding from the Nordic Council of Ministers, cg., the Nordic Institute for Theoretical Physics (NORDITA), the Nordic Sami Institut (NSI), the Scandinavian Institute of Maritime Law (NIB) and the Nordic Volcanological Center (NORVOLK). Nordiskt sarnarbete 2007, Regeringens skrivelse 2007. Nordisk Ministerrad (2008). SEK 480 million. http://www.norden.org/forskning/skltopforskning/index.asp Nordic cooperation has been reviewed in the last decade but the focus has been on the overall strategy of Nordic cooperation rather than on single programmes while NordForsk has not yet been evaluated. However, the process of developing NORIA, including the evaluation of Nordic research cooperation in general, is ongoing. Some drawbacks pointed out so far concern the inappropriateness of funding models; the fact that planning processes, timetables, funding instruments etc. are not harmonized across Nordic countries: etc. European Commission (2008b). A cooperation fund involving Czech Republic, Hungary, Poland and Slovakia (http:l/ www.visegradgroup.eu/main.php) Involving Luxemburg and the regions of Lorraine, Rheinland-Pfalz. Saarland and Wallonia.

REFERENCES A C E N E T E R A - N E T (2004), Summary, deliverable n o . 1.5. AMPERA ERA-NET (2006a), Position Paper Proposing Solution.sfor the EJective Removal ofBurriers, deliverable 2.2.1. A M P E R A E R A - N E T (2006b), Report on Best Practices on Management oJ'Transnational Projects, deliverable 2.2.2. AMPERA E R A - N E T (2006c), Report on Ezcisting Complemmturities und Gups among National Programmes, deliverable 2.1.1. A M P E R A ERA-NET ( n o date), A M P E R A BeneJits, accessed a t www.cid.csic.es1

amperalpageslabout-benefits.php. A S T R O N E T E R A - N E T (2006), Report on the Management qf' European Astronomy, w o r k package n o . I . 1. B I O D I V E R S A E R A - N E T (2006), An Assessment of Best Practises in Commissioning

and Mutiaging Biodiversity Research in Europe, and Approaches to Overcoming Barriers to Cooperation. B I O D I V E R S A E R A - N E T (2007), Biodiversu Newsletter, November. B O N U S E R A - N E T (2005), Baltic Sea Research und Funding in 2004, publication n o . 3. B O N U S E R A - N E T (2008), IdentEfication qfurras,for cooperation in existing programmes andgaps, publication n o . 6. C O R E O r g a n i c E R A - N E T (2006a), European Research in Organic Food and Farming, Reports on Organisation and Conduction qf'Research Programrnes in I 1 European Countries, Bonn, G e r m a n y . C O R E O r g a n i c E R A - N E T (2006b), Sharing and Developing Best Prricticefbr thr Evaluation of' Research in Organic Food and Farming (OF&F), deliverable 5.1.

62

European science and technology policj

CORE Organic ERA-NET (2007), Added Valuefrom Working Togrther. CORNET ERA-NET (2006), Quick Reference Guide to Colleo~tiveResearch in Europe. CORNET ERA-NET (2007), How to Organise, Munage and Fund Collective Research - A Practical Guide .for O~vnersand Managers of' Collective Research Programmes, The Hague. CRUE ERA-NET (2007), Good Practice Guide for Rrseurch Progrurnrne Identijcarion, Promotion and Vulidation, CRUE report. work package no. 3. E-RARE ERA-NET (2007). Project Outline, accessed at http://asso.orpha.net/ ERANETIcgi-bin/. ERA-NET (2006), ERA-NET Review 2006, Expert Review Group report. European Commission, December. EUPHRESCO ERA-NET (2007), 'EFSA scientific colloquium on pest risk assessment', presentation, Parma, Italy. European Commission (2004), 'ERA-NET series 11, networking of national research programmes in the European Research Area'. project synopsis. European Commission (2006), 'Summarj report of workshop: "ERA-NET as tool for facilitating Cooperation between Ministries managing RTD programnlrs"', DG Research, Brussels 23 May. European Commission (2007), The European Research Areu: New Perspectives. Green Paper, COM(2007) 161 final, Brussels. 4 April. European Commission (2008a), Impact assessment acconlpanying the cominunicution 'Towardsjoint programming in research: ~ i m k i n gtogether to tackle cotntnon challenges more effectively', SEC(2008) 228 1. European Commission (2008b). Towards Joint Programming in Research: Working Together to Tackle Common Challenges More Eflectiwl~s,COM(2008).468 final. Brussels, 15 July. EUROPOLAR ERA-NET (2008), The Landscape of Europcwn Polar Research: European Polar Capacity - A n Overview of'Research Infrastructures in the Arctic and Antarctic, vol. no. 2. EUROTRANS-BIO ERA-NET (2008), Conference - Workshop P4 Report. GLOSPERA (2004), Global Systems and Policy Design for the European Research Area, final report, July, European Commission FP5 IHPP-STRATA project contract No HPVI-CT-200 1-00007. HERA ERA-NET (2005), Annex I - Description of Work. HERA ERA-NET (2006), HERA Peer Review Report, deliverable 3.2.1. HERA ERA-NET (2007), A Survey of Legul and Administrative Barriers to Joint Research Programmes, synthesis report, deliverable 9.1.1. HY-CO ERA-NET (2006a), Report on Analysis ofDatu Collected, work package 2, deliverable 2.2. HY-CO ERA-NET (2006b), Report on Indicators of Hydrogen and Fuel Cells Research, deliverable 2.5. Inman, A. (2007) 'the importance of coordinating phytosanitary research in Europe', Plant Science, 44 (22) (August), pp. 3-7. INNER ERA-NET (2006). SWOT-Analysis on Common Strategic Issues on Reseurch Programmes on Innovative Energy Systems for the ERA NET INNER. final report, deliverable 4, Paris. Kristensen, E.S. (2007), 'Three years of cooperation in the ERA-NET CORE organic', presentation given 13 September. -

A European Reseurch Area built hy Member Stutes?

63

LERU (League of European Research Universities) (2007), The Future of' the European Research Area, August. MARINERA ERA-NET (2007a), Anthropogenic und Clinmte Chunge Imjxrcts on Murine Biodiversity and Ecosystem Function, report no. 5. MARINERA ERA-NET (2007b), Barriers to Cooperation in MarinERA Purtn~r Stute Marine R T D Programmes, report no. 2. NEW OSH ERA (2007), A Report on Complementarities and Cups in O S H Research Programmes on New and Emerging Risk Factors, deliverable 5.1. Nordisk Ministerrad (2004), NORIA Vithok orn nordiskfimkning ock innowtion, Huvudrapport, TemaNord 2004:502, Copenhagen. Nordisk Ministerrad (2006) Building Nordic Strength Tlzrougl~ More Open R&D Funding Study 3 The Nest Step in N O R I A , TemaNord 2006:576, Copenhagen. Nordisk MinisterrAd (2008), Toppfor.~kning.sinitiuti~~ef - Progrumjurslug till Nordiskti minister8det inom omrddet energi, miljii och klitnar: Frdn Norden till Jorden, 29, accessed at www.norden.org/sv/nordiska-ministerraadet/tvaersektoriel1a-

-

-

aktiviteter/spetsforskningsinitiativet/nordiska-toppforskningsinitiativet. NORFACE (2006), Comparative Anu1ysi.s of' Partner Councils. Extmsion of' the Conzpurutive Ancdysis to the 12 NORFACE Partner C'ouncil.~,deliverable 3.1 . l , Ljubljana. Optimat Ltd, VDIIVDE Innovation, Technik GmbH (2005), E.1-umining the Design oJ'Nutionu1 Rescwrch Programmes. ORGANIC REVISION (2006), Report of the Second Stukrlwlder Workshop (?/'the Organic Revision Prqject, deliverable 6.3. Petitet, E. (2007), 'Preparation of the FP6 success stories. Answers to the EC questions, summary from EUROTRANS-BIO partners', Europcun Rcseurch Headlines. Regeringen (Swedish Government) (2007), Nordiskt sumarhete 2007, Regeringens skrivelse 2007, Skr. 2007108:90. SAFEFOODERA ERA-NET (2004), Annex I Description of Work. SNOWMAN ERA-NET (2005), A S N O W M A N ' S Navigator through Resrurch Funding Programmes Across Europe, deliverable 1. SNS (Nordic Forest Research Cooperation Committee) (2002), Survey oj'Nordic Wood Research A ~ ~ u l y sand i s Recommendutions. SURPRISE ERA-NET (2007), Opportunities and Barriersfor Trans-notionul Joinr Programming- Esisting Knowledge, work package no. 6. Technopolis Group (2001), Bilateral Intemutioncd R&D Cooperation Po1icie.s of'thr EU Member States, Final Report, (European Commission study, contract: ERB HPV2 CT 1999 0009). Technopolis Ltd. (1999) Cross-border Cooperution Within Nutionul RTD Programmes. TRANSPORT ERA-NET (2005), Analysis Barriers to Cooptwrion and Development o f a n Inithl Proposal of Cooperation Procedures, deliverable 1.2. Vanhemelrijck, J. (2008), 'EUROTRANS-BIO, sharing costs, risk and .skills of innovation to stimulate innovation', presentation, Brussels. -

4. The 'European Research Area' idea in the history of Community policy-making Luca Guzzetti 1 THE BEGINNINGS O F COMMUNITY RESEARCH

POLICY On 14 J a n u a r y 1974, t h e very first Council Resolution concerning C o m m u n i t y scientific and technological research (coallsteel a n d nuclear research fields excluded) w a s taken; here is a n excerpt f r o m its text: Whereas the Heads of State or of Government meeting in Paris on 19 and 20 October 1972 expressed their determination to promote the development of a common policy in the field of science and technology and noted that such a policy would require the coordination of national policies within the Community institutions, and the joint implementation of projects of interest to the Community; . . . Article 1. In order to define objectives and ensure the development of a common policy in the field of science and technology involving the coordination of national policies and the joint implementation of projects of interest to the Community the following operations shall be progressively carried out within the Community: 1) the examination and comparison of Member States' national policies in this field, particularly their potential, plans, programmes, projects, budgets, measures and methods in this field; 2) the identification, analysis and comparison of the objectives of the Member States in order to determine the common goals to be adopted and the appropriate ways and means of achieving them; 3) the coordination of national policies on the basis of 1 and 2 above with the aim of: eliminating unnecessary or unwarranted duplication of effort in national programmes, avoiding any divergent tendencies which would be contrary to the interests of the Member States, improving the efficiency or reducing the cost of national and Community projects by sharing of tasks or possibly by the concentration of resources o r research teams, and gradually harmonising procedures for the formulation and implementation of scientific policies within the Community. (Council of the European Communities, 1974: 1) A s can be observed, t h e involvement o f t h e E u r o p e a n C o m m u n i t y in scientific a n d technological research started with t h e project of coordinating national policies, programmes and activities.

The 'European Research Area' idea

65

Research Commissioner Ralf Dahrendorf, following the directives of the Paris Summit, wrote in his Working Programme that 'for Europe, science and education form part of a medium and long-term strategy for the future'; and adopting a subsidiarity principle unte litteram, he added, 'we must be clear in this regard that the mere fact of an action being taken at a European level is not enough. For this reason we should establish standards which will determine what tasks rightly belong at a European level' (Dahrendorf, 1973: 21). Unlike his predecessor, Altiero Spinelli, who had focused his policy initiatives on industrial research, his interest was more in basic research and his main project regarded the creation of a 'European Scientific Area' for stimulating greater cooperation between Member States. Rather than launching new ambitious Community programmes, as Spinelli had wished, Dahrendorf suggested maklng optimal use of the resources already available in the Member States. His proposals were to draw up a register of laboratories and research centres, to increase the mobility of researchers and to facilitate contacts between scientists within the Community. Furthermore, he suggested sharing the use of expensive machines, making particularly expensive installations function to common advantage, to disseminate scientific and technical information on a European scale and to improve coordination between national policies, programmes and projects. For the coordination of national policies, Dahrendorf proposed the creation of a new committee within the Commission and asked for active collaboration by the Member States. National governments should immediately communicate to the Commission all their decisions related to research and development and meetings of the Council of Research Ministers should take place at least every six months to secure a progressive convergence of national policies. Consequently in the 1974 Resolution quoted above, the Council invited Member States to join in a collaboration that would consist above all of their giving the Commission, at appropriate times, all the necessary information. The Resolution also indicated that projects of Community interest could involve non-Member States as well and that some degree of coordination should be established in cases of cooperation with or within international research organizations. The Scientific and Technical Research Committee (CREST), consisting of representatives of the Member States and the Commission, would help both the Council and the Commission to fulfil these tasks. The political relaunch tried in the early 1970s, at the time of the accession to the Community of the UK, Ireland and Denmark, was followed by a sudden crisis: the Yom Kippur War was followed by the oil crisis and by the collapse of almost all political solidarity between the Member States. A common policy in science and technology would start again only in the

66

European science and technology p o l i q

1980s in a different situation and on a different basis, with the high-tech industrial projects proposed by Etienne Davignon. In any case, the Committee proposed by Dahrendorf for the coordination of national policies had been created by the Council with its first Resolution: CREST started operating in 1974 - for an experimental period that lasted from 1974 to 1976 - and has continued its activities till the present day. But from the very beginning, in its preliminary activity of gathering information on national and international research programmes. a necessary step before the coordination of the Member States' research and development (R&D) policies, CREST encountered several problems and difficulties that looked almost insurmountable, as clearly stated in a communication from the Commission to the Council in 1977: 'How is any policy of coordination possible when even in the Member States the planning of research projects and programmes remains partly uncoordinated? How can there be a policy of coordination when the Member States are unwilling to commit themselves in this respect?' (European Commission, 1977: 1 I). All the same, the potential importance of at least the partial coordination of European research activities and policies was very clear to the members of CREST and even more so a decade later in a more prosperous situation. In 1989, after having recognized that Europe had progressed in the field of science and technology - notably with Community programmes such as ESPRIT' and BRITE,' with high energy physics research at CERN (European Organization for Nuclear Research), with fusion research and the construction of the Joint European Torus and with space research at the European Space Agency (ESA) - and having stressed the dbsence of European cooperation in military R&D, CREST wrote in the context of COPOL 88 (an exercise comparing national science and technology (S&T) policies): It is a question of twelve different national policies, independent from each

other, which may converge on some great common initiatives . . . Of course. history and geography have modelled a Europe characterised by very different S&T landscapes, showing in each country diversified ways of intervention. . . . but we should start to think about the question of creating more coherence and synergies in the public policies in this domain. Because in these kinds of questions, the term 'closure' does not only mean an absence of coordination, but also insufficient concertation and even, quite often, complete absence of information about the practices adopted by the partners. (CREST. 1989: 69) The discourse principally regarded new programmes for the stimulation of research, but it could easily be extended to S&T policy at large: 'It is a question of enhancing the homogeneity and the fluidity of the European

67

The 'Europeun Research Area' idtw

scientific area, and of assuring a more systematic valorisation of its results' (ibid).3 Research officially became a Community policy in 1986 with the Single European Act (SEA). In both the SEA and the subsequent Maastricht Treaty, the importance of coordinating national and Community policies was mentioned; but the new centrality given to the Community in I992 now the European Union was also evident in a change concerning the task of coordinating European research activities. Whereas the Act stated that 'Member States shall, in liaison with the Commission, co-ordinate among themselves the policies and programmes carried out at national level', the Maastricht Treaty laid down that 'the Community and the Member States shall co-ordinate their research and technological development activities so as to ensure that national policies and Community policy are mutually consistent' (Maastricht Treaty, Article 130H). As has been shown, over the years, attempts to coordinate national policies had always proved difficult and largely unfruitful, but now, on the basis of the principle of subsidiarity, the existence of a supranational body the competence of which extended into the scientific and technological field - was acknowledged; new attempts could therefore be made to redirect efforts towards coordination, by basing them on reciprocal relations between the Union and the Member States, with a pivotal role played by the Commissjon and CREST in particular, rather than on the goodwill of individual Member States. -

-

2 AFTER MAASTRICHT The Commission guided by Jacques Delors presented in 1993 a White Paper: Growth, Competitiveness, Employmmt. The document started from the assumption that since 1991 the world economy had entered a period of deep recession and that there was no mechanism to guarantee that, in an ever more competitive global economy, Europe could necessarily look forward to vigorous future growth. Still more serious in the Commission's view was the possibility that economic growth, when it came, might not be accompanied by a parallel increase in employment and many factors seemed to indicate that this scenario was quite likely. If growth and employment were the two objectives for which the Commission invited the Member States to draw up effective strategies, the middle factor in the title of the White Paper competitiveness - was the main instrument on which such strategies had to be based. The objective was therefore to create a 'society based on knowledge' or an 'intelligent society'; it was also important to make the best possible use of Europe's cultural and scientific -

68

European science und rechnologj. polic:,

tradition and of the wealth represented by its internal differences by giving education and training a more European slant. Regarding research and technological development, the White Paper put forward three main suggestions: an increase in investment, the more efficient exploitation of new technologies and of the results of research and the coordination of activities at continental level (European Commission, 1993). For the new Commissioner in charge of science, research and development, Antonio Ruberti, the need for effective coordination of national and Community policies looked very obvious: considering the simple fact that the Member States allocated 13 per cent of their public research and development budgets to European cooperation (Community Framework Programme plus funding of such bodies as ESA, ESRF,4 CERN, EUREKA,5 and so on), even partial coordination of the national research activities in receipt of the remaining 87 per cent would have a significant effect on the efficiency and productivity of European research. particularly by limiting the waste of resources arising from duplication. Traditionally, the problem of coordination had been tackled by attempting to persuade national ministers responsible for research and development to keep one another informed of decisions taken independently, but this strategy had proved politically difficult to manage and quite unproductive. The Commission now proposed a more dynamic approach, based on Article 130H of the Maastricht Treaty, quoted above: First, it must be recognised that coordination of national policies cannot be laid down by law. It can only come about through common assent and must become a habit, a state of mind driven by an awareness of its obvious benefits. . . . The approach taken must be multifaceted and flexible, but also practical. Different types of activity will be undertaken at different levels: - on determination of RTD policies, with the objective of providing ministers in the Union with a forum for discussion, with systematic preparatory work to supply the information which they all need; on implementation of research activities, including not only those covered by the Framework Programme for implementing Articles 130K and 130L,6but also the activities under the national programmes in order to make all the efforts more consistent; o n international cooperation. where a stronger presence on the part of the European Union is both desirable and attainable, without impinging on the Member States' prerogatives. (European Commission, 1994a) -

And on the more political dimension of coordination, that of determining research and technological development policies, the Commission added: One thing is clear: Member States decide Community policy together but determine their own national policies. Of course, there is some interaction between the decisions taken at these two levels: although the general guidelines for the Framework Programmes are directed towards action by the Community, they

The 'European Research Area' ideu

69

are based on what is known about the national priorities and have a definite impact on national perceptions and analyses of the situation. This impact in turn depends on traditions and on the level of research attained in the individual Member States. (European Commission, 1994a: 3 4 ) To improve the interaction between the national and Community dimensions, the Commission suggested a series of measures, both technical and political. If coordination was to be effective, it was first necessary to have information on national science policies and standardized statistics on research activities. In October 1994, the Commission presented its first report on science and technology indicators (European Commission, 1994b), to be followed by regular updates.' On a more political note, the Commission suggested that there be many more occasions when national oficials responsible for science and technology could meet to discuss future prospects and that meetings at ministerial level be held on a fairly regular basis. A precedent had already been set by the first informal meeting of ministers for science and technology held at Schwerin, Germany, in July 1994. On that occasion, the ministers met not to take decisions on Community policy but for a wider-ranging and less formal discussion of their national policies and the extent to which they could be pursued at European level (Guzzetti, 1995). According to the Commission, the Council of Ministers could also become the most suitable forum for defining science policies in sectors where organizations distinct from the Community were concerned such as ESA, CERN or EMBLR here too on the basis of a progressive convergence in the policies of the various countries involved. According to Ruberti's assessment of the situation, Community cooperation implied some coordination since, by definition, it imposes a degree of convergence between national and Community research and technological development (RTD) activities. Furthermore, the committees responsible for each specific programme consisted of representatives of the Member States, so that national viewpoints were expressed not only on the major goals, priorities and fields of Community research that is, at the level of the decisions taken by Council and Parliament regarding the Framework Programme and specific programmes - but also on the selection of projects receiving funding. This made it possible to coordinate part of the activities of individual countries with those decided at Community level. While it was clear that there existed some convergence between national and Community RTD activities, now the question was how to go beyond such de facto coordination (Ruberti and Andre, 1995). As part of the Commission project, research activities would also be coordinated through 'complementary programmes' involving only a few Member States and Community 'participation' in programmes initiated by Member States acting together. These two types of cooperation were -

-

70

Europerrn science und technologj polic!

envisaged by the Maastricht Treaty (Article 130K and 130L, mentioned above) but were not put to the test, except in the context of Euratom and in the end no specific form of funding was allocated to them in the 4th Framework Programme (1995-98). On the other hand. for the first time, the 4th Framework Programme distinguished between three types of network, with a structure more solid and lasting than the traditional academic networks: 'consortia', following the pattern adopted for the 'fusion' programme; 'concertation networks', like those organized with Community support for the biomedical programmes; and 'thematic networks', with technological and industrial finalities. A pioneer in the first type of network was the controlled thermonuclear fusion programme of Euratom, which as far back as the early 1960s instituted close collaboration using association contracts to coordinate the activities of existing specialized centres. The model was a planetary one, with the Commission at the hub and the national centres organized around it. Whereas in other fields of nuclear research Euratom was not successful in competing with national programmes, in this case the Commission succeeded in bringing together a European 'invisible college' of fusion and getting the best scientists from all the Member States to collaborate in a truly European project. The Europeanization of the venture subsequently put the scientists involved in this field in a strong position vis-a-vis national governments meeting in Council: the resistances that had prevented the development of so many Euratom research programmes could not develop here. As a result, the fusion programme has always enjoyed massive funding - even in periods of institutional crisis for the Community and it has thus been able to achieve indisputable technical and scientific successes (Guzzetti, 1995). Concertation networks, in addition to stimulating and increasing the efficiency of research in a given discipline, for instance, the medical field, could also help to redirect and improve the organization of some types of research. This was certainly true of the European Laboratories Without Walls (ELWW): given the obvious need for an interdisciplinary approach to the Community's biotechnology programme, it was possible to set up a series of networks based not on distinct disciplines but on the study of a theme or the solution of a specific problem. These proved to be extremely productive. Networks were also a key factor in the stimulation/science programmes and the successive 'human capital and mobility' programmes, helping to mobilize Europe's scientific potential. Although less directly, joint funding contracts for technological innovation were also intended to create European networks: the hope - largely realized - of the originators of ESPRIT and other programmes devoted to new technologies was that participation in Community programmes -

The 'European Research Areu' idea

71

would create fruitful links between companies, research centres and universities and that these links would outlive the programmes themselves, albeit in different forms. Although these probably cannot be called networks, habits of collaboration and stable personal relations developed and an opening up to other countries took place, something that certainly did not exist at European level before the 1980s, neither between research organizations (both public and private), nor between companies operating in the same sector, nor in a large part of the scientific community. Such networks are a consequence of the Framework Programmes that should not be underestimated for the realization of the European Research Area as well. Many organizations have been set up in Europe since the war. The chief purpose of some was to encourage communication and contact within the scientific community. These include the European Science Foundation (ESF), responsible for setting up the first European networks in many areas of basic research in the 1980s, or the European Academy and the All European Academies (ALLEA). Then there were multilateral programmes such as COST" and EUREKA"' which coordinated and funded specific scientific and technological research projects. Finally there were institutions of a sectoral nature, for example the European Organization for Nuclear Research (CERN), the European Space Agency (ESA), the European Molecular Biology Organization (EMBO), the European Molecular Biology Laboratory (EMBL), the European Southern Observatory (ESO), the Institut Laue-Langevin (ILL), the European Synchrotron Radiation Facility (ESRF), the Reading-based European Centre for Medium-Range Weather Forecasts (ECMWF) and other organizations (Krige and Guzzetti, 1997). Most of these bodies collaborated or had collaborated with the Community, but in Ruberti's opinion, much could still be done to set up a truly European network and to establish ongoing, fruitful links between all the organizations working in the field of science and technology. The collaboration agreements concluded by the European Commission with CERN and EMBL in respectively 1994 and 1995 went into this direction. Returning some years later to the question of the de facto coordination of European research activities through EU collaboration (representing at the time 13 per cent of all European budgets dedicated to R&D), Antonio Ruberti said: The European Union research programmes profoundly penetrate in the national structures of research. The networks that they create involve most of the large national research institutions. . . . The impact of the EU programmes on the research activities conducted in Europe is even more important because the Community funds are almost completely utilised for the financing of the

72

European science and ~rcknologj.potiq

research projects, and not for the infrastructures or the functioning costs. like it happens for national funds. All considered, literally interpreted. the figure of

13 per cent tends to underestimate the importance of European scientific and technological cooperation as it is already realised today. (Ruberti. 2003: 151) As has often happened in the history of the European Commission. when one Commissioner concentrates his or her attention on fundamental research, science, networking and coordination, his or her successor will stress more applied research, technology, structures and industrial products and vice versa. This was the case for Ruberti's successor, Edith Cresson, who - as Commissioner in charge of research, innovation and education - focused her activities mostly on industrial research, creating some task forces in the fields of transport, health and education systems. And this was also the case for Cresson's successor, Philippe Busquin, who commemorating Ruberti, who prematurely died in 2000. said: 'As a matter of fact, my political ambition, as European Commissioner in charge of research, is to create a "European research area". Such an ambition finds its deep roots in the thought of Antonio Ruberti' (Busquin, 2003: 223).11 When Philippe Busquin took office as Research Commissioner, the situation in the Commission was very difficult: the Santer Commission had had to resign nine months in advance, after the most violent conflict ever with the European Parliament. Furthermore, at the centre of the scandal had mostly been Edith Cresson, with accusations of nepotism. Therefore, the situation was very delicate and many expected that the new Commissioner would keep a low profile, limiting hidherself to administering the Framework Programme. Instead Busquin, after consultation with many people inside and outside the Commission, took a different attitude, as he himself remembered some time later: Searching in the library of the Directorate General, I realised that the idea of European research area had already been imagined by some among my predecessors . . . But it had remained at the stage of idea. I myself am a politician. And 1 found in the concept of a European research area the pivotal instrument to articulate a really political project. (Busquin and Louis, 1005: 92) With this project in mind, the Commissioner started the work of analysing the situation of European research and the cost of the absence of a European Research Area: Essentially, the non-existence of a European research area is due to the compartmentalisation of public research systems and to the lack of coordination in the manner in which national and European research policies are implemented. . . . Even if most of the measures need to be taken by the public authorities. the measures proposed will have an impact on the whole research system (public

73

The 'European Research Areu' ideu

and private). Centres of excellence will produce knowledge that can be used by companies, which are also among the users of research facilities. And improved systems of indirect support for research and innovation also concern the private sector. (European Commission, 2000: 9) Although the European R&D public budget had reached 17 per cent of the total, many obstacles still existed for the creation of a continental research area, characterized by an open market for researchers and by a European research policy based on the coordination of national ones. For overcoming a situation in which Europe saw 15 + 1 (Member States EU) juxtaposed R&D policies, the Commission proposed systematic benchmarking and making public the results of comparative analyses of national policies, programme and activities in the field of science, technology and innovation; the use of the Framework Programme its main financial instrument for directing national programmes and creating more durable and stable networks of excellence; and a complete liberalization of research activities within the European Union. While Busquin noted that 'an important step in the direction of a European research area would be the opening of national research programmes to the laboratories of other European countries', but that 'to develop a strategy of European opening represents a difficult step for national scientific organisations, because it means a questioning of working methods and most importantly a loss of sovereignty' (Busquin and Louis, 2005: 91). he also stressed that resistance could come from inside the European institutions and from elsewhere as well.'? As a matter of fact, the Framework Programme had become an extremely important source of funding for many countries, disciplines, laboratories, research groups and so on and all these stakeholders were worried by the Commission's project of turning the FP into an instrument for the development of a European Research Area. Multi-level resistance obliged the Commission to indicate clear research priorities for the 6th Framework Programme and to be less radical as far as the exclusive financing of very large projects (integrated projects, networks of excellence, and so on) with more 'European value-added' was concerned.

+

-

-

3 THE LISBON STRATEGY As is well known. at the Lisbon European Council, which took place on 23-24 March 2000, the European Union set itself a new strategic goal, that 'to become 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' (Rodrigues, 2003). The Council

74

Europrun science und rrchr~ologvpoliq.

put R&D a t the centre of its strategy for achieving this ambitious goal by 2010, following the suggestions made in the Commission communication o n establishing a European Research Area. In its conclusion, the Portuguese Presidency invited the Union to: Develop appropriate mechanisms for networking national and joint programmes on a voluntary basis around freely chosen objectives, in order to take greater advantage of the concerted resources devoted to R&D in the Member States, and ensure regular reporting to the Council on the progress achieved; . . . improve the environment for private research investment, R&D partnerships and high technology start-ups, by using tax policies. venture capital and EIB [European Investment Bank] support; encourage the development of an open method of coordination for benchmarking national research and development policies . . .; facilitate the creation by the end of 2001 of a very high speed trans-European network for electronic scientific communication . . .: take steps to remove obstacles to the mobility of researchers in Europe by 2002 and to attract and retain high-quality research talent in Europe: ensure that a Community patent is available by the end of 2001. (Rodrigues, 2003) T h e methods proposed by the European Council for turning Europe into a knowledge-based society were summarized in the idea of a n 'Open Method of Coordination', according to which the Member States would voluntarily converge on the objectives collectively decided. But what did this method mainly consist of? This method has four components: 1 ) the establishment by the Council of common European Guidelines; 2) a reciprocal learning process, which includes benchmarking, peer review, diffusion of best practices and common indicators: 3) given the quoted Guidelines and the learning process. national plans are drawn up by each government; 4) on a regular basis, the Council carries out an evaluation of the results, which can lead to recommendations. The Commission feeds the whole process with its contributions, both in the top-down and in the bottom-up dimensions of the strategy. (Telo, 2002: 250) Although the strategy and the method adopted have been criticized

(Facing the Clzullenge, 2004) and the results of the Lisbon Strategy have not been particularly positive s o far - for instance, the objective that all Member States would invest at least 3 per cent of their G D P in research has not been reached by most Member States - it must be stressed that it was thanks to the Lisbon Strategy that the crucial role of research for the economic a n d social progress of Europe was recognized. This new centrality, formally attributed by the European Council to research and development, attained two major results a t E U level. T h e first is the creation in 2005 of the European Research Council (ERC) with its €7.5 billion budget:

The 'European Reseu~chA r m ' zdcu

75

The European Research Council (ERC) is the first European funding body set up to support investigator-driven frontier research. Its main aim is to stimulate scientific excellence by supporting and encouraging the very best. truly creative scientists, scholars and engineers to be adventurous and take risks in their research. The scientists are encouraged to go beyond established frontiers of knowledge and the boundaries of disciplines. The ERC complements other funding activities in Europe such as those of the national research funding agencies, and is a flagship component of the 'Ideas Programme' of the European Union's Seventh Research Framework Programme (FP7). Being 'investigator-driven', or 'bottom-up', in nature, the ERC approach allows researchers to identify new opportunities and directions in any field of research, rather than being led by priorities set by politicians. This approach ensures that funds are channelled into new and promising areas of research with a greater degree of flexibility. (European Research Council, 2008) The European Research Council is the first strictly European scientific agency, in the sense that decisions of its Council are taken exclusively on the basis of the scientific quality of the proposal without applying the principle of 'justc r r t o u r ' , like, for instance, in ESA, without consideration of political decisions on funding taken by the national governments, like in the many European intergovernmental scientific organizations and without applying formal criteria of multinationality and multisectoriality and of informal political equilibria like in EU programmes. In fundamental research, a European area (part of a larger globalized scientific world) already exists and the creation of the European Research Council is testimony of that. The second major result of the new centrality assumed by R&D in E U policy is the very strong increase in the funding of the 7th Framework Programme. Since 1983, the budget of the Framework Programmes had increased steadily but slowly from €3.75 billion (1983-86) to €4.5 billion (1987-90), to €6.6 billion (1991-94), to €13.215 billion (1995-98), to € 15 billion (1999-2002) and finally, with the 6th Framework Programme, to E l 7.5 billion (2003-06). But the approval in 2006 of the 7th Framework Programme (2007-13) with a budget of over €50 billion for its seven years' duration (so as to adapt to the EU's financial perspectives) marked a deep change in this evolution, with a n increase in annual funds of over 40 per cent. This turns the European Union into one of the most important financial contributors to the development of science and technology in Europe. There is n o doubt that with its new role as financer of both fundamental and applied research, the European Union will play a n even more important role as at least de facto coordinator of national science and technology policies. -

-

European science and technology policy

76

NOTES European Strategic Programme for Research and Development in Information Technology. Basic Research in Industrial Technologies in Europe. CREST had started with COPOL 88. a systematic analysis of national research activities. programmes and policies. but apparently the comparison was not appreciated by some Member States, so that the work was abandoned after the preparation at least as far as I know - of only two publications, relative to Germany and the United Kingdom. European Synchrotron Radiation Facility. http:lleureka.be/home.do (accessed 17 May 2009). Articles 130K and 130L of the Maastricht Treaty concerned the possibility of creating research programmes in which only some countries take part (variable geometry) and the possibility for the Community to participate in programmes launched by Member States. In the 1990s, at least the collection of data about R&D activities turned out to be easier than in the previous two decades. European Molecular Biology Laboratory. European Cooperation in Science and Technology. Pan-European network for market-oriented, industrial R&D. Interestingly, it has been noted that the three European Commissioners who supported the idea of a European Research Area Ralf Dahrendorf, Antonio Ruberti and Philippe Busquin were university professors, close to the scientific community for which the concept of free circulation and exchange sounds obvious while there exists another intellectual family of European Commissioners, like Etienne Davignon, Edith Cresson and Filippo Maria Pandolfi, who were much closer to industry (Andre. 3007). For an interesting discussion on administrative inertia inside the Commission, see Banchoff (2002). The thesis is that the work of preparing and managing the Framework Programme leaves the administration in charge of research with neither the energy nor the time needed for elaborating, proposing and launching neu research policies, like those necessary for the creation of a European Research Area. -

-

-

-

-

REFERENCES A n d r e , M. (2007), L'Pvolution des dispositifs de ~outienu la RDI coopPrative a u j l de la construction d'une Europe de la recherche, Les Rencontres Futuris, Seance d u 2 8 n o v e m b r e 2007, Paris: A N R T . Banchoff, T. (2002), 'Institutions, inertia a n d E u r o p e a n U n i o n research policy'. Journal of Common Market Studies, 40 ( I ) , 1-3 1. Busquin, P. (2003), ' L a ricerca, I'universita, 1'Europa'. in: A . Pascale (ed.), Antonio Ruberti: scienziato, politico, umanista, R o m e : F o n d a z i o n e A n t o n i o Ruberti, p p . 223-3 1. Busquin, P. a n d F. L o u i s (2005), Le dkclin de I'empire scientlfique europeen. Comment enrayer la chute?, Brussels: Editions L u c Pire. C o u n c i l o f t h e E u r o p e a n C o m m u n i t i e s (1974), 'Council Resolution o f 1 4 J a n u a r y 1974 o n t h e C o o r d i n a t i o n o f N a t i o n a l Policies a n d t h e Definition o f Projects o f Interest t o t h e C o m m u n i t y in t h e Field o f Science a n d Technology', Oficiul Journal of the European Communities, C 7 / 2 , 2 9 J a n u a r y 1974. Council o f t h e E u r o p e a n U n i o n (2000). Presidencj. Conclusions of the Lisbon European Council, S N 100100,23-24 M a r c h 2000.

The 'Europmn Research Areu' ideu

77

CREST (1989), L'espuce scirnt~jiqueet technologiqur ruropken duns le conteste internationul. Resources rt condition.~de Iu compktitivitk de 11 Communuuti., COPOL 88, E U R 1 1872. Dahrendorf, R. (1973), Ricercu, scienza e educuzione. Infirmuzione scientifico tecnica. Programma diluvoro del sig. Rulf'Duhrmdororf;memhro dellu Commi.s.sion~, CAB/X//17/73-1. European Commission (1977), Common Policy in the Field of' Science und Technology, Communication to the Council (30 June 1977), supplement 3/77 to the bulletin, Luxembourg: CCE. European Commission (1993), Growth, Competitiveness, Employment: The Clzullmges und Wuys Forwurd into the 21st Century - White Paper, COM (93) 700 final, Luxembourg: Office for Official Publications of the European Communities. European Commission (1 994a), Research and Technological Development: Achieving Coordination through Cooperation, Communication from the Commission, COM (94) 438 final, 19 October 1994, Brussels. European Commission (1994b), The European Report on Science und Technology In(/icutors 1994, EUR 15897, Luxembourg: Office for Official Publications of the European Communities. European Commission (2000), Towurdsu Europeun Reseurch Areu, Communication from the Commission, COM (2000) 6, 18 January 2000, Brussels. European Research Council (2008), http:llerc.europa.eu/index. cfm?fuseaction=page.display&topicID= 12 (accessed 17 May 2009). Facing the Challenge. The Lisbon Strutegy ,for Growth and Development (2004/, Report from the High Level Group chaired by Wim Kok, Brussels, November 2004. Guzzetti, L. (1995), A Brief History of Europeun Union Reseurclz Policy, Luxembourg: Office for Official Publications of the European Communities. Krige, J. and L. Guzzetti (eds) (1997), History of European Scientijic and Technological Cooperation, Luxembourg: Office for Official Publications of the European Communities. Rodrigues, Maria Jozo (2003), European Policies ,for u Kno~vledge Economy, Northampton, MA, USA and Cheltenham, UK: Edward Elgar, pp. 141 and 145. Ruberti, A. (2003), 'L'Europa della scienza e della tecnologia alle soglie del nuovo millennio' (Hendrik Brugmans Memorial Lecture, Bruges, 12 December 1997)', in: A. Pascale (ed.), Antonio Ruberti: scienziato, politico, umanista, Rome: Fondazione Antonio Ruberti, pp. 145-66. Ruberti, A. and M. Andre (1995), Un espace europien de lu science. R@exions .sur la politique europ4enne de recherche, Paris: Presses Universitaires de France. Telo, M. (2002). 'Governance and government in the European Union. The open method of coordination'. in: M.J. Rodrigues (ed.), The New Knowledge Economj, in Europe, Northampton, MA, US and Cheltenham, UK: Edward Elgar, pp. 242-7 1.

5. Analysing Community policies Jim Dratwa This chapter takes stock of European Research Area (ERA) policy actions undertaken at Community level from 2000 to 2008. Although not exhaustive, it delivers a meaningful overview: it is not merely a listing of actions but endeavours to provide indications regarding the results and impacts achieved as well as the limitations or obstacles faced.' This is done with regard to Community-level policies while bearing in mind that ERA is also, crucially, constituted through policies emanating from intergovernmental cooperation and the Member States themselves (in addition to the endeavours of other, regional, supranational, and nonstate actors). Indeed, besides referring the reader to the preceding chapters in Part 1, this Community stocktaking also covers the coordination of actions by Member States and mutual learning processes. Even though the Framework Programme (FP) is crucial for the realization of ERA, the scope of ERA goes beyond the FP and indeed beyond EU initiatives alone. Thus, actions that have to some extent be contributed to ERA can be identified in diverse quarters as shown below. A stocktaking exercise of this nature is a crucial yet thorny undertaking; difficulties abound. Many of the actions supporting the development of ERA are still ongoing or indeed in their early stages. Furthermore, many of the initiatives taken at EU or national level do not have as sole (or main) objective the furthering of ERA. Beyond this introduction, the chapter consists of four main sections. The first section offers a crisp systematic overview of the actions undertaken in relation to each one of the detailed ERA objectives;? the second one carries the stocktaking further by bringing out the main areas of progress as well as those of more mediocre attainment. As regards the ERA objectives themselves - objectives compared to which stock is to be taken and results appreciated the third section reflects on the successive ERA Communications drawing out the evolving and multifaceted nature of those objectives as well as of the tenets underpinning the ERA project. The final section concludes on the appraisal of both the results achieved and the objectives themselves - as well as of the ERA undertaking as a whole considered in that light - and on future developments and perspectives. -

Anulysing Community policies

79

1 ACTIONS, RESULTS, LIMITATIONS In the 2000 Communication on ERA, the ERA'S specific objectives were defined according to seven 'specific themes for action' as follows: a series of material resources and facilities optimized at European level; more consistent use of public instruments and resources; more dynamic private investment; a common system of scientific and technical reference for policy; more abundant and mobile human resources: a dynamic European landscape, open and attractive to researchers and investment; an area of shared values. This section is systematically structured along those headings,' in accordance with the 2000 ERA Communication and (penultimate heading) the 2001 Communication on the international dimension of ERA, each one comprising a short description of key actions undertaken with a summary appreciation of results achieved and limitations encountered. Community actions addressing university-based research and stemming from the work started with the 2003 Communication The Role of the Universities in the Europe of Knowledge (COM(2003) 58) are also included (under the last heading).

1.1

A Series of Material Resources and Facilities Optimized at European Level

As regards networking of centres ofexcellence and creation of virtual centres (that is, 'reducing the fragmentation of the European research system' and 'attaining a critical mass'), FP6 introduced new instruments and aimed at achieving critical mass in terms of research capacity (Networks of Excellence NOES) or resources from various partners (Integrated Projects - IPS). Yet many NOEScorrespond to 'close cooperation' but fall short of the expected research capacity 'integration'. The overall potential impact on de-fragmentation of IPS and NOESis limited by the very small proportion of overall research they account for. As regards the definition of'a European approach to research jucilitit.~,the European Strategy Forum on Research Infrastructures (ESFRI) adopted in 2006 the first strategic 'European Roadmap for Research Infrastructures', which identifies 35 projects for new pan-European research infrastructures an achievement needing to be acted upon. However, the amount of FP7 resources allocated severely limits Community support for the preparatory phase of Roadmap projects. FP7 forms of support do not correspond to infrastructure time-scales, flexibility needs and funding stability requirements. That being said, an important development materialized in 2008 with the Commission proposal on a legal instrument for research -

-

80

European science and technologr policy

infrastructures (European Commission, 200%) designed to facilitate the joint establishment and operation of research facilities of European interest between several Member States and countries associated with the FP. As regards maximizing the potential ofSered 61. electronic netbvorks, 'Communication Network Development' under FP6 gave rise to the panEuropean launch and deployment of a series of e-infrastructures for the research community (GEANT, EGEE and DEISA that is. respectively: the Europe-wide e-Infrastructure, Enabling Grids for E-sciencE, and Distributed European Infrastructure for Supercomputing Applications). These initiatives were successful but limited budget resources hindered further deployment of grid infrastructures for many more scientific communities. -

1.2 More Consistent Use of Public Instruments and Resources

As regards the more coordinated implementation of national and European research programmes (including reciprocal opening, information exchange and evaluation measures), an important development under FP6 was the launch of ERA-NET (European Research Area Network) actions to help nationallregional managers increase mutual coherence and coordination of their respective research programmes - 30 joint calls for proposals were launched in 2006. The key to ERA-NET actions is their 'bottom-up' and 'variable geometry' approach in terms of participating countries. Beyond 71 ERA-NET projects, a first pilot action under Article 169 of the EU Treaty (which covers EU participation in new integrated research programmes undertaken by several Member States) was launched: the European & Developing Countries Clinical Trials Partnership (EDCTP). However, the basic conditions for a successful use of the legal instrument were not met (due to difficulties with IegaVadministrative rules and Member State reluctance to fully integrate their national programmes and make long-term financial commitments), seriously limiting the integration achieved. New Article 169 initiatives under preparation aim to overcome these difficulties. ERA-NET and Article 169 initiatives have enabled Member States to see better the need for optimal coordination. Despite this, the importance of these schemes in terms of volume of research funding in the overall European landscape remains limited (projects launched by the end of 2007 represented only 0.8 per cent of overall ERA public investment in research) and major barriers persist: a lack of nationallregional strategies to differentiate programmes to be opened up to transnational cooperation1 coordination from those where national autonomy should prevail, and very limited progress in reciprocal or unilateral opening up of national -

Analysing Community policies

81

programmes to non-national participants outside the above-mentioned schemes. More encouraging perspectives in this regard are opened up by European Commission (2008b), the Communication on joint programming (as well as by the launch of the new ERANET-Plus scheme). It is also relevant to mention here the development of Technology Platforms, bringing together industrial stakeholders to define and implement Strategic Research Agendas in specific technological fields, which have an increasing coordinating effect on programmes with impacts at EU and national levels. The newly launched Joint Technology Initiatives (JTIs) could mark a quantum leap in that regard across public-private divides. In addition to the above, in some sectors, coordination at a strategic level is ensured through specific fora (for example, National IST-RTD [Information Society Technologies/Research & Technological Development] Directors forum, Standing Committee on Agricultural Research). As regards closer relations between European organizations for science and technology cooperution, see Chapters 1 and 2.

1.3 More Dynamic Private Investment As regards the better use of instruments of' indirect support Jbr research (that is, to stimulate private investment in research, particularly in small and medium enterprises [SMEs], with due regard for state aid rules), research and development (R&D) fiscal measures are more common in EU Member States now than in 2000. Best practices in fiscal measures to stimulate R&D were identified and shared among Member States in the framework of CREST (Scientific and Technical Research Committee) working groups. In 2006, the Commission adopted the Communication Towards a More E8ective Use of'Tux Incentives in Favour qf'R&D together with a staff working document offering guidance on the design and implementation of R&D tax incentives. Next steps will focus on the lack of consistent evaluation studies of national R&D tax incentives. As regards state aid issues, the adoption by the Commission in 2006 of a new Community Framework for State Aid for Research and Development and Innovation was an important development. As regards the development of efkctive tools for the protection of intellectual property ( I P ) (including the adoption of a Community patent), there has been no major breakthrough in this area. IP protection through patents and litigation remains too complicated and costly in Europe. As regards the encouragement of risk capital investment und company start-ups, three developments can be noted: first, the adoption by the

82

European science and technology policy

Commission of guidelines on state aid for risk capital in 2001 and, following a review of the text, in 2006; second, the promotion of direct investment into venture capital funds targeting young innovative firms; third. the adoption of the Competitiveness and Innovation Framework Programme (CIP), which includes increased Community support (managed by the European Investment Fund) and awareness-raising actions targeted at potential recipient companies. It should be noted, however, that statistical comparisons with the US still paint a bleak picture of the EU as supply- and demand-side barriers hamper the efficient deployment of risk capital and the Single Market does not operate well as different regulatory and tax environments reinforce fragmentation and inhibit cross-border operations.

1.4 A Common System of Scientific and Technical Reference for Policy Implementation As regards the development of the research needed for political decisionmaking and the establishment of a common system of scientijic and technical references, among the numerous actions in this area, two particularly important ones were the adoption by the Commission in 2002 of guidelines and principles on the collection and use of expertise by Commission services and the development of SINAPSE (a web communication platform developed by the Commission to promote the more efficient use of scientific information and expertise in support of policy-making). The diversity of and inconsistencies in the systems of scientific advice provision in Europe continue, however, to pose a major challenge.

1.5 More Abundant and Mobile Human Resources As regards greater mobility of researchers in Europe and the introduction of a European dimension into scientijic careers (as well as making Europe attractive to researchers from the rest of the world), two key achievements have been the Commission Recommendation on the European Charter for Researchers and on a Code of Conduct for the Recruitment of Researchers, an important instrument for raising awareness and improving researcher career management, and the adoption of the 'scientific visa' package in October 2005 (a directive and two recommendations on the admission and residence of third country nationals to carry out research in the EU). Nonetheless, overall achievements since 2000 remain marginal compared with the importance of what is at stake (see Chapter 13). The central issue remains employers' tendencies to recruit and promote researchers from their local environment without open and transparent

Analysing Community policies

83

procedures. Also, the regulatory frameworks in place fail to recognize researchers as a specific population with specific needs and fail to eliminate regulatory obstacles to career development and mobility, for example, concerning taxation and social security. These obstacles and shortcomings are precisely what the initiative on a careers and mobility partnership (European Commission, 2008a) intends to address. As regards a greater place and role for women in research and giving young people a taste for research and careers in science, cross-country comparisons as well as the identification (and development and dissemination) of best practices have been carried out through several initiatives in both areas. In this regard, the Gender Action Plan (GAP) was an important FP6 initiative to promote gender equality within projects. However, considerable obstacles persist (for example, scientists' perception that measures to increase the participation of women are not compatible with scientific excellence; in the area of science teaching, delays or blockages in transferring innovative methods from the proof-of-concept stage to the classroom). 1.6 A Dynamic European Landscape, Open and Attractive to Researchers and Investment

As regards a greater role of the regions in the European research effort (notably through synergies with structural assistance) and integration q j the scientlJic communities of Western and Eastern Europe, it is estimated that €10.6 billion of cohesion policy funding, notably from the European Regional Development Fund, has been used to support R&D and innovation in the 2000-06 programming period. Such investment plays a significant role in fostering research and innovation activity, particularly in the Community's less developed Member States and regions, especially when the national, regional and private co-financing leveraged by cohesion policy programmes is also taken into account. Cohesion policy programmes offer a platform for regional stakeholders to increase their capacity to undertake excellent research and exploit its results. They are the main EU instrument for fostering research activity in less developed Member States and regions and thus help to address the lack of cohesion and the science and technology (S&T) development gaps identified in the ERA Communication of 2000. The Community Strategic Guidelines, 2007-2013 on economic, social and environmental cohesion assign an even more prominent place to research and development and innovation as drivers of economic growth. The Commission has tried to create a framework for the coordination of cohesion and research policies through the proposals for cohesion

84

European science and technology policy

policy programmes and for the 7th Framework Programme for 2007-13. However, the different levels of governance mean that, in practice, national and regional stakeholders are responsible for coordinating the use of the two instruments and projects. Through its Innovative Actions programmes, cohesion policy has also supported the development of regional strategies in less favoured regions on the theme of knowledge-based technological innovation. Such strategies help regional stakeholders in less favoured regions to implement measures appropriate to their specific context. The regional dimension of the European research effort is also acknowledged in the RTD Framework Programme. The positive results of the Regions of Knowledge initiative launched in 2003 to promote more and bette! investment in research through mutual learning, coordination and collaboration among regional players have led to an extended Regions of Knowledge activity under FP7. In addition, the new FP7 Research Potential action focuses explicitly on strengthening research capacity in 'convergence regions' and 'outermost regions' in terms of physical and human capital.

1.7 An Area of Shared Values As regards tackling sciencelsociety issues on a European scale (notably concerning foresight and the organization of 'Citizens' Conferences' at European level), progress has mainly been made on the identification and dissemination of best practices. As regards foresight, the exchange of knowledge and practice between foresight practitioners and policy-makers was facilitated through several measures: an online guide has been developed to serve as a reference system for foresight and a monitoring system on foresight in Europe with a web portal (www.efmn.info; accessed 18 May 2009) has also been set up. However, foresight has not yet reached a reasonable state of integration and coherence at EU level. Furthermore, the direct impact of foresight on S&T decision-making in the Member States and the Commission cannot easily be identified, except in very specific cases. As regards dialogue and participation, in addition to the identification and dissemination of best practices (for example, the Commission's 2001 Science and Society Action Plan or the 'European Platform of Stakeholders and Experts in Participative Techniques' [CIPAST - Citizen Participation in Science and Technology]), the concrete implementation of participatory techniques has also been effected (for example, two fullsize 'Consensus Conferences' and a new FP7 instrument to support the participation of civil society organizations). Development in this area is

85

Analysing Community policies

still embryonic, nonetheless; in fact the open-coordination process begun in 2001 has not been fully successful as in some Member States there was no counterpart to the Commission's Science and Society pole. Yet the development of participatory processes and of new thinking in relation to policy-making is now gaining increasing currency in several Commission DGs and in the Member States as well as among other actor^.^ As regards development of a shared vision of ethical issues in science and of technology (touching on ethics committees, rules and criteria), actions undertaken have helped to foster dialogue and learning especially for Member States with weak institutional infrastructures for addressing ethical issues. Those actions included the mapping of existing rules and the exchange of information, experience and best practices in the form of both punctual (conferences, studies, workshops) and structural initiatives (creation of the 'National Ethics Councils' [NECs] forum in 2003, of the European Network of Research Ethics Committees [EUREC] in 2005 and of an electronic database of opinions of national councils). The objective of a 'shared v i ~ i o n is ' ~clearly a remote one, however, as ethics is deeply embodied in national cultures and opinions diverge significantly on a number of issues. Finally, it should be underscored that this area, of S&T inland Society, has been a fruitful albeit not widely advertised locus of thinking on the 'governance of ERA', which is a matter of the highest importance, as will be discussed in the final three sections of this ~ h a p t e r . ~ 1.8 Develop an Ambitious and Extensive International S&T Cooperation Programme As regards the objective - set out in the 2001 Communication on The International Dimension oj'the European Research Area (COM(2001) 346) of opening up the European Research Area to the rest of the world, the EU has increasingly encouraged the participation of third countries in its research programmes and concluded S&T agreements with many of them. That being said, the impact of these agreements remains relatively limited, except when focused on cooperation in specific areas (for example, nanotechnologies with the US). S&T agreements' reciprocity clauses give researchers on both sides access to each other's research funding. But given the present absence of a mechanism to fund European participation, full use of reciprocity and real access for Europeans to third country research funds remains elusive. As regards the objectives - set out in the 2001 Communication to focus EU efforts on specific objectives and to step up international 'technology watch' activities, research capabilities in partner regions have

-

-

86

European science and technology policj

been strengthened via the Specific International Scientific Cooperation Activities (INCO) programme, though at an insufficient level to have longer-term and larger-scale institutional effects. No mechanism exists to determine horizontal international cooperation priorities across and between thematic areas of the FP. S&T cooperation actions by the Member States with third countries are uncoordinated, despite some first steps in this direction (for example, some ERA-NETS focus on the international cooperation of EU Member States with some regions and a CREST working group aims to produce an inventory of international S&T cooperation activities conducted by Member States). Technology Platforms, with some exceptions such as the Global Animal Health Technology Platform, have not considered international cooperation in great depth in their Strategic Research Agendas. As regards the 2001 objective to align EU scientific cooperation policies with EU foreign policy and development aid programmes, cross-references are indeed made to the importance of research actions in relevant EU external policy initiatives. However, ensuring coherence is dealt with on an ad hoc basis: no set-up currently exists through which an overview can be maintained and the coherence of potential external policy actions with international research cooperation assessed. It should be noted, however, that a Communication, International ScientiJic Cooperation Activities, is featured as forthcoming in the Commission's Legislative and Work Programme 2008. As regards the objective to enlist EU scientific and technological capabilities to deal with world problems, one should underscore the conclusion of the ITER (International Thermonuclear Experimental Reactor) agreement, which brings together the EU, Japan, China, India. Korea, Russia and the US, and that places Europe at the forefront of nuclear fusion research. In addition, many examples of an EU response to global problems can be found in the F P actions, for example, in relation to health. Yet no framework exists outside the F P to jointly identify the global issues appropriate for an EU response or the organization of such a response. 1.9 As Regards Community Actions Addressing University-based Research

Besides the areas addressed in the 2000 ERA Communication, there have also been Community actions addressing university-based research, which stem from the work started with the 2003 Communication The Role of the Universities in the Europe of Knowledge. With due regard for the fact that the main actors are situated at the national and regional levels or in the universities themselves, the EU is

Analysing Community policies

87

engaged in the coordination of actions with public authorities to support the modernization of European university-based research. This concerns several domains such as granting real autonomy to and requesting real accountability from universities, funding university-based research more on the basis of academic and non-academic research outputs (industrial and/or international partnerships), promoting professional management in universities and the development of needed research management tools (such as transparent research accounting systems), and supporting their innovation capacities. The EU is also collaborating with the institutions themselves to support the modernization of European university-based research in the aforementioned domains, where work with the public authorities is being pursued. Through an increased budget and a range of new actions, FP7 constitutes a major new step in the Community policy to enhance university-based research, including the following three salient points: increased EU funding for research performed by higher education institutions (from 50 per cent to 75 per cent of total eligible project costs and allowing funding on the basis of full economic cost); establishment of the European Research Council, which supports 'frontier' research and stimulates excellence through competition and is due to concentrate funding on top European performers (universities will be able to receive up to 100 per cent of eligible research costs); enhanced support for the establishment of structured links between universities and non-academia through several FP7 instruments (intersectoral mobility in Marie Curie actions, science shops, regions of knowledge, research-driven clusters). In addition, regarding overall support to policy in this domain, one should underline the support for the collection and production of data on research and education in universities.

2 ACHIEVEMENTS The second half of this chapter consists of three sections. To start with, and building on the previous section, the main areas of progress as well as of more mediocre attainment are brought out. Then the objectives and tenets of ERA are appraised. Finally, the conclusions and future perspectives are presented. As concerns the most favourable 'steps on the way', one of the notable developments has been the launch of the ERA-NET scheme.' This has allowed the managers of some national programmes to begin to make an important contribution to the building of ERA by looking at ways and taking action to reduce the inefficiency and fragmentation inherent to

88

European science and technolog? poliq

a system made up of numerous national research funding schemes. The figures in the stocktaking quantifying the response thus far suggest that ERA-NET responds to a real need. Yet the volume of funding involved in the resulting joint activities is still marginal. Moreover, national and regional programme 'owners' are reluctant to restructure their programmes in a way that would enable the development of genuine joint programming. Those are issues that the Commission initiative on joint programming - taken in the framework of the 2007 ERA Green Paper - is paying heed to. Besides, in addition to ERA-NETS and ERANET-Plus, other initiatives with an impact on fragmentation such as Article 169 initiatives (increasing in significance with the advent of a new generation thereof, building on the EDCTP experience), European Technology Platforms and now Joint Technology Initiatives continue to develop. Research infrastructures constitute another area in which good progress has been made. Pan-European infrastructures must play a key role in reinforcing Europe's overall research capacity. A powerful example of how more can be done is the area of health research where the development of a pan-European network of bio-banks representing the diversity of the European population would provide vastly increased analytical power (and perhaps controversy). In several research fields, the situation is simply that no single Member State can afford to develop the required infrastructures. Building on the widespread consensus - that emerged following the 2000 ERA Communication - on the need to forge a more coordinated European approach to key research infrastructures, a first major milestone was reached with the adoption in 2006 of the ESFRI Roadmap. However, the Roadmap will only be a success if the proposed projects are realized and for this to happen there is still a long way to go. New approaches are required, that is, new legal, institutional and financial tools need to be developed, which the new Commission initiative on research infrastructures will perhaps be conducive to. In short, progress thus far in these two areas (coordination of national programmes and infrastructures) demonstrates that the initiatives taken corresponded to a demand, be it latent or explicit. But above all, it underlines the need for new initiatives with more impact, especially at the strategic or governance level. The same can be said of the area of Science and Society, where participative techniques (to allow civil society to take an active part in policy-shaping and decision-shaping discussions concerning S&T) and strategic foresight need to be deployed in ERA on a wider scale. In the area of international cooperation, success stories such as ITER show that Europe has the will and capacity for leadership to address global challenge with partners around the world. But while Europe is increasingly engaged in global science, research and infrastructure initiatives, these

Atzulysing Community policies

89

initiatives are far from systematic and often poorly coordinated with those of the Member States. In some other areas, despite the efforts made, progress achieved has been even more limited. This is the case, for instance, for the labour market for researchers. Human resources in S&T are a key strength of Europe, which produces more S&T PhDs than the US. However, despite the success of important measures aimed at better exploiting human resources (such as the Marie Curie scheme, the European Charter for Researchers and the scientific visa package alongside which one should note the impending 'European researchers' passport partnership' initiative), Europe crucially lacks an open, competitive and attractive labour market for researchers. The exploitation of this strength is sub-optimal and Europe instead suffers from wasted resources: some bright researchers and S&T graduates leave, others do not enter a research career in Europe or exit early, others miss opportunities to move into positions where their capacities could be better used and developed. The instruments used up to now in these areas and the up-take of specially developed ones by the Member States are not commensurate with the importance of this challenge (see Chapter 13). Another problematic area is private investment in research. The latest statistics available d o not show any increase whatsoever in the businessfunded research intensity of the EU compared with 2000. The gap between the EU and its major competitor has not been reduced (see Chapter 11). Compared with what was announced in the 2000 ERA Communication, the range of actions related to the objective of stimulating private investments in R&D was broadened considerably in the wake of the 2002 Barcelona Summit and the 3 per cent Action Plan. Of course, for many of these initiatives, it is too early to see their potential impact on the statistics. This is the case for the new impetus given by the re-launch of the Lisbon Strategy in 2005, with, in particular, the many actions taken in the framework of the National Reform Programmes. While on the one hand, a lot depends on the implementation by the Member States of their National Reform Programmes, on the other hand, ERA-level actions also have a key role to play in achieving the Barcelona targets. Notably, the globalization of research and development and innovation focuses attention on the need to create European poles of excellence capable of attracting internationally mobile private research and development and innovation investments. -

3 OBJECTIVES AND TENETS An appraisal of the ERA roll-out cannot be carried out without due regard for the ERA objectives themselves - objectives compared to which

90

Europeun science and technology policy

stock is to be taken and results appreciated. In fact, since its 2000 (re-)launch, the ERA concept has been subject to gradual changes. The initial focus of ERA was on how to improve the efficiency and effectiveness of research efforts and systems in Europe and on how to get a better return on investment. At this stage, research was given a key role at the Lisbon European Council meeting in achieving the Lisbon Agenda by 2010. Gradually, the scope was broadened to include the need for more public and private investment in research (which found its most prominent expression in the Barcelona objectives of 2002).8 Subsequently, the 2005 ERA Communication broadened it further to encompass the need for more coherence and synergies between research policies and other EU policies in order to achieve the renewed Lisbon S t r a t e g ~ This . ~ was an explicit recognition that ERA is embedded into the 'knowledge triangle' concept - in a context that demands the consideration of the wider impacts of research (innovation, the internal market, financial markets, higher education systems and so on) and not only of the specific research outputs. In the last few years, particularly with the measures implemented under FP6, ERA has been transformed from a theoretical concept into an applied policy approach embodying many different dimensions. However. even though the policy context has evolved, the original ERA preoccupations - how to overcome Europe's S&T weaknesses and fragmentation and achieve a coherent and effective European research policy - are still at the heart of the ERA project. And as for the latest ERA wave with the 2007 ERA Green Paper, it builds on that history to pursue two trends: on the one hand, as will be discussed further immediately below, a refocusing on the core ERA rationale - that is, the fight against fragmentation - and on the other hand, a widespread fanning out across several themes, dimensions or axes due to embrace the sum total every nook and cranny - of the business of Community research policy (yet failing to win this 'allinclusive' game and leaving out altogether crucial piecesL0of the European research and innovation system). The 2000 ERA Communication had no clear definition of ERA other than descriptive, in extenso." Seven years on (or several decades on, see Chapters 1, 2 and 4), the scope of ERA remains unclear. For instance, do the frontiers of ERA mimic the borders of the EU? And does the former enlarge precisely when the latter does? Besides, is the ERA itself conceptually fragmented? That is, is ERA 'one for all and all for one' or is there an ERA of biotech, an ERA of ICT, an ERA of ITER, an ERA of researchers? And what about an ER[I][E]A with regard to research and innovation and education? Or to sum up with the crispest terms: what is in? what is out? We will scrutinize this difficulty further, with a salient case: -

Anulysing Comnzunity policies

91

Some progress has been made since the concept was endorsed at the Lisbon European Council in 2000. The European Research Area has become a key reference for research policy in Europe. However, there is still much further to go to build ERA. particularly to overcome the fragmentation of research activities, programmes and policies across Europe. (COM(2007) 161, p. 2) Such is the terse summary assessment proposed in the 2007 ERA Green Paper, placing 'fragmentation' at the heart of the ERA endeavour. This requires careful attention. 'United in Diversity', such is the motto and the predicament - of the European polity. Well, the crucial relation between diversity and unity, between the parts and the wholes, indeed the careful respect for the former melded with a vocational drive towards the latter, are well encapsulated in the concept of 'fragmentation' and its central position in the ERA project. What is the problem? What is the objective? What is the proposed action? And how solidly are those three elements connected'? Is public intervention justified? Ought it to be taken at Community that is at EU level or rather at regional, national or intergovernmental level? Those introductory questions are a sine qua non in the development of every Community policy. And it would seem that the concept of fragmentation in research bridges the gap between the first and the last of those questions. The first, as fragmentation would be the problem, the thorn in the flesh of Europe's research and innovation. The last, as that problem constitutes in and of itself a call upon intervention at EU level, since it is that said whole that is fragmented (and that is invoked and instituted 'as a pre-existing whole' by the very notion of fragmentation). At this stage, we should proceed with our inquiry by scrutinizing in some detail the components and implications of 'fragmentation'. First, it should be noted that it is a notion with several dimension: fragmentation of research spending/funding/investment; fragmentation of research execution; fragmentation between Member States, between regions; fragmentation between (scientific, technological, industrial) sectors; fragmentation between types of actors (for example, universities, SMEs and so on); fragmentation between individual actors. The above dimensions are often considered only in terms of investment/execution fragmentation. Yet the notion also extends to fragmented vision, evaluation, programming, policy, instruments. Furthermore, the framework conditions and regulatory environment - for research, education, innovation and commercialization are also subject to fragmentation. Ultimately, the Single Market (for knowledge, but also in the classical fourfold perspective) is far from full realization and demand (be it from the consumer base, through public procurement and so on) is naturally fragmented across diverse local markets. -

-

-

-

92

European science and technology policj~

The ceaseless transformations of polities such as Nation States combine diverse dynamics including subnational 'fragmentation' as well as 'consolidations' extending beyond the state such as with regional integration processes. So do the transformations of research and development and innovation. Indeed the essence of research and development and innovation is the creation or elicitation of new questions, answers, facts, products, processes, fields, bodies, avenues, opportunities and actors. Public intervention (public policy, public action) as such is fragmented and so is 'research policy'. It is fragmented very diversely along a combination of two dimensions. Vertically, that is, across the layers of our multi-level governance systems (often reduced to a two-level game or, only marginally less crudely, to a clear set of four levels: regional, national, Community and [wider] international). Horizontally, across policy areas and issueframeworks and indeed across the institutionalized divides between ministries, departments or international organizations. There is no space for more on this here,'? but before moving on, it is worth underscoring one of the key questions for analysis already arising on that basis: what is the optimal level of fragmentation with regard to the above dimensions? Or far more crudely put: is fragmentation really such a bad thing? And here one can, however briefly, flag that there are arguments in favour of fragmentation: diversity of approaches; competitionlemulation; rich and diversified research and innovation ecosystem; regional embeddedness; resilience of the European research and innovation system; widespread absorptive capacity with regard to spillovers. Furthermore, it is also useful, when considering fragmentation and its detrimental impacts, to point to a set of cognate issues. In particular: what are the (alternative) means to address the detrimental impacts of fragmentation? Here a range of options are available: research policy coordination (common visions, common programming and so on); mutual exchange of information on ongoing and planned research activities; mutual opening of national programmes; increasing (the share of) EU-funded collaborative research; developing nationallregional specializations with attention paid to synergies and complementarity (division of labour versus St Matthew effect); networking or reducing the number of funding bodies; networking or reducing the number of research funding beneficiaries; concentration of research activities on a smaller set of prioritieslareas; improving framework conditions and regulatory environment (including as to researchers careers, intellectual property rights, tax incentives and so on). Well, fragmentation - however central and paradigmatic - only serves here as an example. The aim here is obviously not to cover all angles but merely to indicate that these issues require far more conceptual and analytical work than what has so far served to underpin the ERA and to

Analysing Community policies

93

produce justifications that are more akin to slogans or dogmas than to robust rationales. Yet only the latter and the pursuit of the cognate underpinning work can sustain the ERA through less placid climes (for example, convince Member States that it is worth making tougher choices). The above discussion of the unsteady objectives and tenets of ERA brings on a more serious question. Once the cyclical predicament of the ERA project (or indeed 'counter-cyclical' with respect to the FP periodicity and the political imperative of having something to brand[ish]) is recognized, what is the added-value of ERA? Well, plying such a hazy and all-encompassing concept does have one clear advantage: it increases its 'hospitality', its 'carrying capacity' (Dratwa, 2002).'' That is to say, it makes it easier to elicit a consensus and a go-ahead; easier for all actors to subscribe to it and endorse it. This was further compounded in 2000 by the fact that the main instrument for its implementation, the Open Method of Coordination. was strictly voluntary in nature. Member States could thus exercise a good measure of activism in favour of ERA and European research policy without needing to buy in further - nor indeed to increase (especially national) funding. Subsequently, however, having to face the mild dialectic of 'container' and 'content' (in its own right but also with regard to the FP), the ERA project stumbled upon the 'adoption-implementation gap'. In other words, whereas selling Schrodinger's cat in a bag and for free (that is, 'adoption') had garnered cheers from all quarters, things became trickier when the time came - 'implementation' - to meet the cat, feed it and rearrange one's own house to accommodate it. In fact, all metaphors aside, the lack of clarity as to its concept, main objectives and priority instruments have hampered the realization of ERA. Very much in a stocktaking perspective, this error should be redressed rather than replicated. The risk would be that we now merely go through yet another such cycle. Well, the 'cat is out of the bag' now.

4 CONCLUSIONS Regarding the ERA objectives and achievements, five main conclusions can promptly be drawn. First, the objectives set out for ERA do not easily lend themselves to being fully satisfactorily achieved. Second, the rationales underpinning some of the operational objectives - and thus the actions undertaken to fulfil them - are neither fault-proof nor incontrovertible. Third, and not entirely surprisingly given these first two determinations, some of the objectives have evolved and some have at times been lost sight (or track) of since 2000. Fourth, progress towards the objectives

94

European science and technolog! policy

is uneven and, fifth. in areas where progress has been made, the impact on the European research and innovation system remains limited - so that, in other words, progress so far constitutes only the first steps on the way to making ERA a reality. In that regard, efforts up until 2008 to coordinate national programmes and infrastructures confirmed that there was a demand, be it latent or explicit, but showed that corresponding initiatives should have a higher ambition at the strategic (governance) level. In some areas, such as fostering greater researcher mobility and dynamic private investments, progress achieved had been even more modest. The 2008 initiatives that followed the 2007 ERA Green Paper (European Commission. 2008a, 2008b and 2008c) could help remedy these limitations. Although no form of assessment could possibly be posited at this point, they do target relevant avenues and sore spots. The proof of the pudding and indeed the ultimate test as to the success of the ERA endeavour has to rest with the indicators and assessments of progress presented in the third part of this book. Yet Europe faces new challenges (climate change; energy sourcing and production and use; ageing; risks of infectious diseases; changes in the dynamics of globalization and of the production of knowledge and wealth; increased socioeconomic disparities outside and within the EU and so on, awareness of which has also been heightened) that are very diverse but have two commonalities: all of them call upon new S&T work and all require the international pooling of efforts. Community research policy has to rise up to those challenges. In particular through: achieving the move beyond 'a funding programme plus' to a full-fledged policy; achieving the move beyond a fragmented (27 + 1) policy to a genuine European policy; achieving the move beyond an inward-looking European policy to assume its role beyond the EU (including competitiveness and attractiveness at the international level through to joint international undertakings, scientific diplomacy and synergizing internal and external policies). Community research policy is a key objective in its own right as well as an 'enabling policy' vital to the achievement of the Union's other key objectives. The challenges Europe faces -just like the challenge of stepping up the Union's S&T - can only be met by the active engagement of all actors: at international and national and regional level, public and private. And the Commission too has a key role to play in these endeavours, ranging from facilitator/convenor to main protagonist. It is in this connection that lies the making or breaking of ERA. Thus far. ERA has proven itself to be a powerful mobilizing concept, framing several important developments in the European research landscape. However, nine years on, the many challenges faced by EU research actors and the problems of EU science

Anulysing Community policies

95

and technology performance indicate that ERA has yet to achieve its full potential. And yet this review going back to 2000 also highlights research as an increasingly salient protagonist in the jousting and vying pitting together different policy areas and priorities (at EU level as well as at national level). In effect, it highlights research as playing a prominent indeed more and more prominent - role in the various iterations of the Lisbon Strategy, from the initial agenda (see Muldur et al., 2007, Chapter 3 in particular) to the 'renewed' impulse of 2004-05 through to the 2007-08 'last cycle' and 'post-Lisbon' thinking, with the particular place of the 'fifth freedom"" momentum together with joint programming." Looking towards the future in the light of the stocktaking and discussions above, the three most auspicious developments to do with ERA might well be: first, the new impetus given to joint programming (at the heart of the 'mereological' predicament that is the ERA, that is, pertaining to the relations among the parts and between the parts and the whole) provided it is accompanied by the conceptual work necessary to give meaning and applicability to this proposition. Second, the acknowledgment of the importance not only of vertical policy coordination (as indicated notably by the joint programming developments) but also of horizontal coordination, attuned to the evolving role of research (and perhaps indeed of the Commission's DG Research) in the EU policy landscape. And finally, the acknowledgment of the need for a step change in terms of analytical and data-collection work as well as monitoring and evaluation measures in order to chart a course, follow and adapt it, and measure progress on the way - if the ERA is to be more than an episodic watchword. -

-

NOTES 1.

2.

An attempt at a rather exhaustive inventory is made in the stocktaking array in the annex to European Commission (2007a), a valuable reference that the present chapter draws upon. It should be noted that a mere listing of those detailed objectives alone would take more space than this entire chapter (without bringing any added value since they can be found in European Commission, 2000 and European Commission, 2001a). Hence the emphasis on succinctness in that first section. As regards the recent relaunch of ERA initiated by European Commission (2007a), it is cross-cutting in essence and not of the nature of what can be taken stock of at this stage. It is thus not part of the thematic review set out in this section but is considered in the chapter's subsequent sections and concluding discussion. One last yet important caveat: this section succumbs avowedly to its own critique as regards horizontal coordination in that it concentrates on those Community policies that directly involved the Commission's DG Research to the detriment of policies of other departments that might have (indirectly) played a part in the realization of ERA. -

European science and technologj. policy And indeed further refined along the sub-objectives of European Commission (2000). with each one being italicized in the developments under each heading. See Dratwa (2008). On this very notion of shared understanding, see Dratwa (2004). In this regard, one should note in particular the involvement in the framework of European Commission (2001b); the process marked by European Commission (2004) and the process marked by European Commission (2007b). See ERA-NET Review 2006. European Commission (2002). Published as a precursor to the FP7 proposals. the Communication European Commission (2005) foresaw the following objectives of ERA: ( 1 ) delivering on the Lisbon objectives; (2) putting the 'knowledge triangle' of research, education and innovation to work; (3) mobilizing EU financial instruments at the service of knowledge for growth. The overarching action in the research angle of the 'knowledge triangle' was the new FP7. In particular, severely truncating the systemic approach to research and innovation (that is, in short, neglecting the business of the Commission's D G Enterprise, outside the remit of D G Research), focalizing as regards researchers on public researchers alone (at the expense of other research and innovation workers. be they in the private sector or government staff; another example of insufficient horizontal policy coordination), and remaining largely at the level of obstacles and tools with very little in the way of governance (that is, management issues but also the creation of common goals and strategic vision that are shared - and 'bought into' - by all). See European Commission (2000, p. 8): 'How should this idea of a ERA be defined? It should embrace in particular the following aspects: . . .'. Those diverse perspectives, and in particular these last two dimensions, are discussed in detail in Dratwa (2007). Another important benefit should also be underscored: the mobilizing/galvanizing effect of the ERA project. Besides, there is a distinct potential benefit that could have materialized: this provided a valuable opportunity to deepen the reflection (and indeed the research and dialogue) on the policy objectives and instruments: yet this was insufficiently seized upon in the wake of European Commission (2000,2001a and 2002). With reference to the classic four freedoms underpinning the single market, pertaining here to the free movement of knowledge, for example, through mobility of researchers. In fact, the Lisbon project and the ERA project have been and are again faced with another like predicament: the division of labour and responsibilities - or rather the balancing act - between the Commission and the Member States, ensuring not only the latter's acquiescence but indeed their buy-in and ultimately the endeavours' success. Them predicaments are not quite identical. however: the situation is more interesting on the research policy (and investment) side since further transfers to the EU are ponderable - a s explored in Muldur et al. (2007). Besides, both projects, en route to the knowledge economy top spot. have been crippled from the get-go about to pluck the low-lying fruits when stricken with the torment of Tantalus by the bursting of the prematurely inflated ICT bubble and (very differently) of the agbiotech bubble. -

-

-

-

REFERENCES D r a t w a , J. (2002), 'Taking risks with t h e precautionary principle: f o o d ( a n d t h e envir o n m e n t ) f o r t h o u g h t a t t h e E u r o p e a n Commission', Journal of Environmenral Policy and Planning (Special Issue on Risk and Governance), 4 (3), 97-21 3. D r a t w a , J. (2004), 'Social learning a t t h e E u r o p e a n C o m m i s s i o n a n d t h e C o d e x

Analysing Community policies

97

Alimentarius' in: B. Reinalda and B. Verbeek (eds), Decision Muking within International Organizations, London and New York: Routledge. Dratwa, J. (2007), 'Analysing public policies for research and innovation: adding value with Europe' in: K. Piech (ed.), Knowledge and Innovation Processes in Central and East European Economies, Warsaw: The Knowledge & Innovation Institute. Dratwa, J. (2008), 'Public action and collective experimentation: what Europe through proof?' in: C. Eberhard (ed.), Traduire nos responsuhilitc;.~p1unPtuire.s (Translating our Global Responsibilities], Brussels: Bruylant. ERA-NET Review 2006, The Report of the Expert Review Group, Brussels, December 2006. European Commission (2000), Towards a European Research Area, communication from the Commission to the Council, the European Parliament, the Economic and Social Committee and the Committee of the Regions, COM(2000) 6 final, Brussels, 18 January 2000. European Commission (2001a), The International Dimension qf' the European Research Area, communication from the Commission, COM(2001) 346 final, Brussels, 25 June 2001. European Commission (2001b), European Governance: A White Paper, COM(2001) 428 final, Brussels, 25 July 2001. European Commission (2002), More Reseurch for Europe: Towards 3% of GDP, communication from the Commission, COM(2002) 499 final, Brussels, 11 September 2002. European Commission (2004), Governance ofthe European Research Area: Giving Society a Key to the Lab, Luxembourg: Office for Official Publications of the European Communities. European Commission (2005), Building the ERA of Knowledge for Growth, communication from the Commission, COM(2005) 1 18 final, Brussels, 6 April 2005. European Commission (2007a), Green Paper: The European Research Areci: New Perspectives, COM(2007) 161 final, Brussels, 4 April 2007. European Commission (2007b), Science & Governance - - Taking European Knowledge Societj) Seriously, Luxembourg: Office for Official Publications of the European Communities. European Commission (2008a), Better Careers and More Mobility: A European Partnership for Researchers, communication from the Commission to the Council and the European Parliament, COM(2008) 3 17 final, Brussels, 23 May 2008. European Commission (2008b), Towards Joint Programming in Research: Working Together to Tackle Common Challenges More Efectively, communication from the Commission to the European Parliament, the Council, the European Economic and Social Committee and the Committee of the Regions, COM(2008) 468 final, Brussels, 15 July 2008. European Commission (2008c), Proposaljbr a Council Regulation on the Community Legal Framework for a European Research Infrastructure ( E R I ) , COM(2008) 467 final, Brussels, 25 July 2008. Muldur, U., F. Corvers, H. Delanghe, J. Dratwa, D. Heimberger, B. Sloan and S. Vanslembrouck (2007), A New Deal for an Effective European Research Policy The Design and Impacts ofthe 7th Framework Programme, Dordrecht: Springer.

PART 2

Theoretical issues

6.

From the Lisbon Agenda to the Lisbon Treaty: national research systems in the context of European integration and globalization Robert Boyer

The origin of the European Research Area (ERA) has to be traced back to the end of the 1990s. On the one hand, science and technology experts and economists became convinced that a new productive paradigm was emerging: the surge of information and communication technologies (ICT) was the premise of a 'knowledge-based economy'. On the other hand, Europe as a whole was suffering from an increasing gap with respect to the US and Japan as far as the implementation of this new engine of growth was concerned. The EU and the Member States had to redesign their scientific, technological, economic and social policies in order to cope successfully with this new 'American challenge'. The ERA proposal from the European Commission to the Council, the European Parliament, the Economic and Social Committee and the Committee of the Regions (European Commission, 2000) was one of the first steps in the design of the so-called 'Lisbon Strategy' under the Portuguese EU Presidency. Reading again this text, one gets the impression that it was written today. Europe still suffers from the juxtaposition of national research systems with few interactions between them. Academic research continues to be in a good position at world level but is not converted into a dynamic flow of successful innovations, contrary to what is observed in the US, especially in science-based sectors (Soete, 2006). Very few start-ups become large corporations in the high-tech sector and on average Europe's specialization is still in low-medium-tech sectors. Consequently, growth has been slow in the EU since the early 2000s, as it was in the 1990s, compared with the US and still more with China or India. De,facto, the structural weaknesses of the EU remain, in spite of the adoption, implementation and even adaptation in 2005 of the Lisbon Agenda, and new challenges have to be faced. First and foremost, today the 'Asian challenge' has become central.

102

European science and technology policy

Not only has China become the factory of the world for mature and medium-tech products, capturing the benefits of increasing returns to scale, it is also preparing for the next growth regime, which should be based on endogenous innovation (Gu and Lundvall. 2006) progressively shaping an indigenous Chinese innovation system (OECD, 2007). The large pool of highly skilled labour in science and engineering in India is also challenging the past conventional wisdom: innovation is no longer the exclusivity of old industrialized countries and an Indian innovation system is emerging (Gupta and Dutta, 2005). The EU should therefore look carefully at the Japanese strategy: it aims to intensify research and innovation in order to dynamically preserve the relative advantage of Japanese firms over their Chinese competitors, first imitators and now innovators (Kwan, 2002). Second, the mobility of R&D, which is no longer embedded in a web of domestic networks, is an obstacle to the strengthening of European scientific and technological capabilities. European multinationals with research and development (R&D) centres in the US spend more than their US counterparts do in Europe (European Commission. 2007b). Another novelty is precisely that the US research system is itself being confronted with the rise of Asian competitors (Hollingsworth et al., 2008). Thus, the 2000s reinforce the diagnosis made at the end of the 1990s. More than ever the EU needs dynamic research and innovation systems. The purpose of this chapter is to investigate the mutual links between ERA strict0 sensu and the Lisbon Agenda from an institutionalist point of view. It is first of all crucial to investigate why the strategy adopted in the early 2000s has not delivered the expected outcome, that is, a virtuous cycle of scientific advances, innovation, institutional reforms, faster growth and job creation (Section 1). Have the redesign of the Lisbon Strategy in 2005 and the reconfiguration of ERA overcome the previous obstacles to convergence towards a new knowledge-based growth regime? If diversity in national traditions in research and innovation is an asset from an evolutionary perspective, is it also an obstacle for institutionbuilding at European level? Can the Open Method of Coordination IOMC), seemingly invented to help deliver the Lisbon Strategy, be completed or transformed by a larger institutionalization of the processes governing the ERA (Section 2)? But it would be erroneous to attribute the difficult implementation of ERA solely to the complexity of European integration and institutional setting. Actually, in the era of globalization, it becomes increasingly difficult to firmly embed research, innovation and, of course, economic activity in a particular location, be it local, national or continental. That is especially so for basic research (Section 3). A short conclusion wraps up the main findings of the present chapter and suggests

From the Lisbon Agenda to the Lisbon Treatjl

103

a quite pragmatic and nationally differentiated approach to science and technology policies.

1 WHY THE LISBON STRATEGY HAS BEEN

SOMEHOW DISAPPOINTING The severity of the challenges faced by Europe in the early 2000s and the novelty of the tools proposed to promote national reforms in order to reconcile innovation and solidarity have made the implementation of the Lisbon strategy difficult.

1.1

Promoting Innovation-led Growth and National Institutional Reform through a New Method of Coordination: An Ambitious Project

The so-called Lisbon Strategy displayed three major components: An objective: to promote growth and employment by creating a highly competitive European economy. 2. An input: coupling innovation with the preservation of social cohesion and this as a compromise between market liberalization and a social democratic approach under the umbrella of a Schumpeterian vision of innovation. 3. A new method to be added to the conventional Community-level instruments: OMC was designed as a device to overcome the present distribution of competences between the Member States and Brussels and to promote at national level the structural reforms required to meet the Lisbon objectives. 1.

The origin of this institutional innovation was clearly associated with the diverging trends observed between the US and Europe and with the emergence of new pressures on the welfare state (ageing, obsolescence of worker competences and persisting mass unemployment). The collapse of the internet bubble, the emergence of China and India as major players in the world economy and recurring demands by citizens for more security and the related strains put upon the so-called 'European Social Model' suggest that the diagnosis made in the early 2000s is still more valid by the end of the decade (Figure 6.1). Thus, it is no surprise that the more solid critiques of the Lisbon Strategy recognize that the general diagnosis was and still is relevant and that the overall strategy goes in the right direction (Facing the Challenge, 2004; Aghion et al., 2006). Yet a common feeling is that implementation

104

European science and rechnologj.policy

More growth jobs

ore

Enhanced financial capacity

Mainable welfare

Education. mfrastructure

I

More R&D

A

General benchmarking of national configurations

Source:

Author's work

Figure 6.1

The Lisbon Agenda ambition: creating a virtuous cycle via three mutually reinforcing mechanisms

has basically failed. That is why the strategy has been redesigned and according to some analysts downgraded (Pisani-Ferry and Sapir, 2006). The paradox is that the 2005 Spring European Council turned the reformed Lisbon Strategy into a key component of its policy (Rodrigues, 2006). It is then important to investigate some of the obstacles that have been inhibiting the implementation and effectiveness of the Lisbon Strategy. 1.2

Adverse Macroeconomic Trends have Hampered Business Innovation Strategies

In modern theories, research and growth are marked by a two-sided causality. On the one hand, a good economic performance simultaneously generates favourable expectations concerning medium-term growth and the ability to finance R&D out of buoyant firm cash flows and easier public financing via subsidies or tax credits. R&D expenditures therefore tend to be pro-cyclical, especially in typical liberal market economies, but also in the coordinated or mixed economies typical of the EU (European Commission, 2007b, p. 25). On the other hand, such investments in innovation might deliver new products or productivity increases after a maturation period of several years or even one or two decades. Current

From the Lisbon Agenda to the Lisbon Treaty

105

innovation performance is therefore the outcome of past research and development efforts. This standard framework sheds a crude light upon the poor performance of the EU at the aggregate level. Within the context of major uncertainties at world level, the slow European growth observed since the early 2000s is linked to the combination of unsolved structural problems, a common monetary policy directed towards price stability and euro credibility and national policies constrained by excessive deficits. All these factors have limited the expansion of R&D, especially in France, Italy and to some extent Germany. The innovative potential and productive capacities of the EU have therefore probably been affected negatively by the orientation of economic policies in the direction of stability rather than growth. But these are not the only factors, since the very design of the research strategy under the Lisbon Agenda has encountered significant limits. Basically the tools used, such as OMC and European programmes, have major weaknesses. 1.3 The Lisbon Strategy Has Not Been Sufficient to Promote National Reform Policy-makers should learn from the difficulties encountered in the implementation of the Lisbon Strategy. The 2004--05 mid-term review clearly pointed out some of the limits of the actual organization and triggered a reform of the Lisbon Agenda (Rodrigues, 2006). It was basically recognized that strategic objectives were blurred, that the inflation of measures and priorities was detrimental to the implementation of some basic mechanisms and that relevant financial incentives were missing concerning the implementation of the agenda (Table 6.1). The integrated guidelines for growth and jobs contain four major objectives, decomposed into 22 items: clearly they are too numerous. Of course, such a list is the outcome of political bargaining and compromise and is supposed to take into account the complexity of European issues. In contrast with monetary stability and budgetary discipline, which are governed by explicit clauses of European treaties along with 'instruments of compliance', the Lisbon Agenda was not allocated any hard enforcement mechanisms. While this could be a promising method, this lack of conventional instruments of enforcement has been detrimental to the effectiveness of the Lisbon process. Two avenues for improvement can be imagined. On the one hand, peer review could be complemented with some form of 'blame and shame' for the more reluctant Member States (PisaniFerry and Sapir, 2006). On the other hand, in the long run, the design of explicit hard rules at European Community level should not be excluded,

106

European science and technology p o l i c ~

Table 6.1 An assessment of the Lisbon Strutrgj. Criticism

Reply

Possible Reforms

Too many guidelines

A response to the complexity of modern economies. The expression of political compromises On the contrary, a promising method for overcoming institutional and political deadlock Unequal across countries, common to many European issues

Reduce the number of guidelines Replace with mechanisms combining items

Lack of policy instruments to implement the strategy Lack of political will, a technocratic exercise

Low democratic accountability

Little justification for eurozone dimension of benchmarking

Fuzzy criteria in the assessment of National Reform Plans The same reform might have different, sometimes opposite. effects in different countries

More involvement of diverse stakeholders than for other European policies (ECB", competition, etc) Benchmarking as a learning process, a method to overcome institutional deadlock

This is only the first stage of a learning process

It might be an exceptional case

Notes:

a. European Central Bank. b. Stability and Growth Pact. Source: Author's work

Design explicit hard rules at Community level 'Blame and shame' as incentives to reform Better marketing. repackaging of the Lisbon Strategy Make the political objectives more clearly explicit Extend the diversity of stakeholders at the national level Develop another concept of democracy Either the unambiguous re-nationalization of reforms Or taking account of the Lisbon Strategy in the re-design of European instruments (for example SGPbreform) Use the employment1 growth diagnostics Build a genuine methodology Contextual benchmarking Take into account national diversity

horn the Lisbon Agendu to the Lisbon Treuty

107

just to promote a better balance between macroeconomic stability and growth and employment objectives. The Lisbon process is perceived by most outsiders as a typical technocratic exercise that does not call for the expression of strong concerns by policy-makers at either national or Community level. This difficulty can be perceived as a pure communication issue between the policy-makers and the citizens, but the issue is probably deeply rooted in the very process of Europeanization of domestic policies. As such, the Lisbon process is thus part of the general disenchantment with respect to the evolution of European institutions and policies. Within the Lisbon Strategy, ERA has been suffering from the fact that it has been considered a mere component among others of an encompassing structural reform strategy, whereas it should have been considered as one of its founding pillars. The lack of more stringent tools is also a major reason for the absence of a clear inflexion of the EU towards a knowledgebased economy (KBE) and innovation-led growth. 1.4

The Limits of Previous Framework Programmes Concerning the

Organization of Efficient Networking Between the Actors of Research and Innovation Another difficulty is closely related to the nature of the new paradigm in the domain of research. During the 'Golden Age', a linear conception of technical change was prevailing. First came scientific advances, which were then converted into new products and processes and finally propelled economic growth. With information technologies, the process of innovation becomes more interactive between research and development, between universities and firms, between large firms and subcontractors, between producers and users and between design, production and marketing. The quality of these interactions is crucial for the outcome of any R&D activity, as shown by the European experience in aircraft and space research. But it has proven difficult to replicate the success of Airbus and Ariane in successive Community Framework Programmes (FPs). Because of the prevailing emphasis by the Member States on a fair return for each Member State, many programmes seem to have privileged equity in the allocation of European funds at the possible expense of efficiency and effectiveness in the generation of scientific advances. Furthermore, it is not easy to monitor from Brussels the formation and performance of transborder networks in the domain of science and innovation. Even in the case of dense pre-existing links between actors, like in Japan, it has been difficult to build effective cooperation between complementary research centres, firms and competences in order to promote major breakthroughs

108

European science and recl~nolog?polic!.

in computer science (Odagiri et al., 1997) or in biomedical research (Okada et al., 2006). Quite logically, research consortia are more successful when significant spillovers exist among participants, when competition on the product market is moderate, and when basic science is concerned more than applied research. Another obstacle is hindering the full efficiency of the FP. On the one hand, the minimum efficient scale of R&D projects has tended to increase in many sectors, which calls for forms of cooperation organized at European and sometimes world level. On the other hand, the European Commission aims at involving the maximum number of Member States and each of them is preoccupied with getting a fair return for its contribution to the European budget. Consequently, spending is dispersed across too many and too small projects and the chances of success and breakthroughs are reduced accordingly. The European Commission has detected this pitfall in relation to its Networks of Excellence: 'A lesson learnt under the 6th Framework Programme is that such durable partnerships are only possible between a very restricted number of partners pooling a significant volume of resources' (European Commission, 2007a, p. 18). The challenge then is to invent mechanisms to overcome this built-in feature of European integration: the actors defending the equity principle in the allocation of European funds are more numerous than those promoting efficiency and future growth (Le Cacheux, 2005). 1.5 Still Distinct National Trajectories for Research Systems

This defence of equity by each Member State is also an obstacle to an integrated research policy promoting efficiency and future growth. For the sake of efficiency and domestic policy concerns, Member States might prefer to stick to their existing national research systems. Political concerns are not the only factors at the origin of a remarkable inertia of national research systems. First of all, these systems have been elaborated through a long-term historical process and therefore the domain of specialization in research is closely related to the nature of competitiveness in key domestic sectors. It is therefore rational to continue to deepen scientific, technological and institutional advantages through specifically national research and innovation programmes. Comparative studies of national and social systems of innovation have repeatedly demonstrated the strong embeddedness of research systems in the fabric of local institutions such as the organization of education, the degree of concentration of manufacturing firms, the style of public intervention, the enforcement of competition and so on (Freeman. 1987;

From the Lisbon Agenda to the Lisbon Treaty

109

Amable et al., 1997; Amable, 2003). It has even been argued that the deepening of so-called 'globalization' has reinforced the specificities of social systems of innovation among OECD countries (Amable et al., 1997). Therefore, with the exception of continent-wide initiatives concerning transport, energy and defence, which exhibit clear trans-frontier externalities, it might be unrealistic for the EU to promote the same general approach for all Member States, for instance, in terms of reaching a common R&D intensity target. Since the loss of credibility of the so-called 'Washington Consensus', economic policy experts recognize the fallacy of 'the same size fits all' strategy: 'the prevalence of the first-best, optimal policy approach, which implies an excess of orthodoxy, leads to suboptimal growth performance' (Commission on Growth and Development, 2008, p. 30). The idea that local growth trajectories matter is more and more widely recognized (Hausmann et al., 2005). This means that each Member State has to diagnose what the specific obstacles are that limit the performance of its research system and the contribution of innovation to job creation, growth and its successful insertion into the European and world economies. Consequently it is not a surprise that such diverse instruments are used in different Member States for research and innovation policies (European Commission, 2007b, p. 75): even if they tend to evolve under pressure from common factors, they are far from converging. This is a problem for centres of excellence as well since in the EU, these are distributed differently according to discipline. In a sense, the search for ERA should not forget the motto of the EU: 'United in Diversity'. Actually, evolutionary theories suggest that the persisting diversity of dynamic research systems and social systems of innovation might be a source of resilience and even of dynamism at EU level provided some redistributive mechanisms diffuse the relevant benefits across Europe. These limits of the 2000 Lisbon Agenda are more and more recognized (Fucing the Chullmge, 2004; Aghion et al., 2006; Pisani-Ferry and Sapir, 2006) and have led to a significant revision of the European strategy and a proposal for a new configuration of ERA. Is it up to the old and new challenges faced by the EU?

2 IS THE NEW ERA CONFIGURATION A SOLUTION? Quite logically, the new strategy builds upon the pillars of European integration (European Commission, 2007a). A comparison between the US and European integration processes shows a striking analogy: in both

110

European science and technologj.policy

cases the constitution of a Single Market has been the permanent objective and other components of the integration process have frequently been presented as complementary to the principle of competition at continental level (Boyer and Dehove, 2006). It is therefore clever to set the construction of an integrated market for research as a strategic objective for ERA. A second leverage factor for a European research agenda is that the related competences are shared between the Union and the Member States. According to the draft Lisbon Treaty, 'In the era of research, technological innovation and space, the Union has the competence to act, and this includes the definition and implementation of programmes, provided that this competence does not prevent Member States from exerting theirs.' (Lisbon Treaty, Part 1, Title 1, Article 2C, Section 3). This provides legal support for a better optimization of national research programmes and priorities at the initiative of the European Commission (European Commission, 2007a, pp. 22-4). A third pillar is evident: the EU has to strengthen its Framework Programmes in order to respond to the significantly more integrated policies observed in the US, Japan, Singapore (Beffa, 2005) and recently China (Gu and Lundvall, 2006). Let us examine briefly the chances of success of these three tiers of ERA. 2.1

An Integrated Market for Researchers: Would it Provide the Right Balance Between Competition and Cooperation?

All over the world, excellence in academia and research is perceived to be concentrated in the US. At continental level, a rather typical job market for academics and researchers is in place for almost any discipline. This is in strong contrast with the still highly segmented European universities and excellence research centres and this in spite of several ambitious student and researcher exchange programmes promoted by Brussels during nearly two decades. A second key feature of the US research system is its large openness to talented and mobile researchers from all nationalities attracted by US universities. Foreign observers frequently attribute this outstanding competitive advantage to typical market-led competition. A closer look suggests that Europe should not confuse objectives with instruments and causes with consequences and conclude from the US success that only markets can organize the mobility of researchers, both internationally and domestically. First and foremost, the internationalization of research systems concerns mainly small open economies and not so much the medium-sized and leading European economies, which tend to defend their autonomy in this area (Meri, 2007, p. 2).

From the Lisbon Agendu to the Lisbon Treaty

111

Second, there is a huge gap between the simple objective of increasing the intra-European 'mobility' of researchers and the ideal of effective and powerful market 'mechanisms'. It has to be recalled that markets assess ex post the effectiveness and efficiency of heterogeneous productive and research organizations built upon the capture of increasing returns to scale, learning by producing, learning by communicating and so on (Boyer, 1997; Hollingsworth and Boyer, 1997). In the long interim period, 'collectively organized' mobility is the premise for the constitution of such a market, assuming it is possible. Third, the emphasis put upon trans-frontier mobility is justified only if one observes strong 'productivity' differentials when researchers move from one research organization to another. If productivity were exclusively an attribute of each individual, whatever the context, then mobility would have no importance, especially in the epoch of the Internet, electronic reviews and long-distance collaboration. The most convincing part of ERA thus concerns providing support for building new and innovative networks and virtual or physical research organizations (European Commission, 2007a, pp. 15-19). Fourth, under the new paradigm, the probability of success in research and innovation is higher, the more dense and varied the interactions between teaching and research, universities and firms, large and small firms. Therefore, the more important mobility of research is not necessarily geographical but might be thematic and functional (for instance, from research to entrepreneurship, from one domestic research centre to another one). This job mobility seems to correlate with the dynamism of national/local research and innovation systems (Meri, 2007, p. 5). These stylized facts are not a total surprise: are basic scientific advances not based upon 'a subtle mix between cooperation and competition' (David, 2007: 253)? It is therefore probably naive to expect too much from a Single Market for research. Creating research organizations that exhibit strong domestically generated returns from cooperation still seems more important at this stage of ERA. Then an upgrading of the mobility towards these excellence centres can significantly improve the overall performance of the EU. But it has to be remembered that big research institutions are not necessarily producing scientific breakthroughs, at least in biological research (Hollingsworth, 2007). 2.2 Coordination of National Policies andlor Search for Institutional Complementarities at the Local Level? The ERA raises another central issue concerning the level of governance appropriate to foster the institutional reforms required to fulfil its main

112

European science and rechnologj policy

objectives. OMC assumes that coordination among Member States is an important factor in the redesign of economic institutions. The literature points out two externalities at the core of the Lisbon process. If, for instance, slow European growth is somehow related to a restrictive monetary policy based on the belief that labour markets are too rigid. then successful reform reducing structural unemployment in one country may induce change in the European policy-mix. especially if such reform takes place in a large country. There are other forms of 'cross-border externalities'. Similarly, the successful redesign of a national system of innovation is expected to benefit other economies via the conventional positive spillovers associated with technological change. From a theoretical point of view, this would mean that in the long run, the related competences should move to European level in order to internalize at continental level these externalities. According to this view, the Lisbon process would be a method to overcome the present distribution of competences as stated by existing European treaties. Nevertheless the specificity of the Lisbon process is to rely on benchmarking, learning and peer pressure in order to promote economic reform. As mentioned earlier, many social scientists think that this is a major novelty and contribution to significant advances in European integration (Zeitlin, 2005; Zeitlin and Pochet, 2005). Systematic comparisons of domestic policies would therefore be as important as positive and negative externalities to promote a specific process subtly mixing domestic and European concerns within an iterative and long-term process. The experience of recent years suggests that these learning processes, if they exist at all, are quite weak and unable to trigger the emergence of a virtuous cycle, according to which lagging countries emulate the more successful ones, and that would induce a progressive acceleration of European growth and job creation. Quite on the contrary, the abundant literature on capitalist diversity is now confirmed by research about complementarities between various policies: labour market and welfare reform, innovation policy and the policy-mix. The problem is that these complementarities are mainly if not exclusively national. Hence, a possible difficulty of the Lisbon process and especially ERA: the will to cope with cross-border externalities neglects the fact that the crucial issue is frequently the coordination and the sequencing of domestic reforms (Figure 6.2). This remark suggests a possible direction for the reform of the Lisbon process: instead of benchmarking individual measures. why not promote a set of interrelated policies that generate a positive spillover in terms of research, growth and employment, according to a set of complementary mechanisms that would cross the frontiers of various policy domains (legislation, taxation, public spending, finance, labour market, competition

-

F r o m the Lisbon Agenda t o the Lisbon T r e u t y

The main externalities are cross borders

/

Innovation externality

\ Member j

Lower NAIRUa

Growth promoting

The need for coordination is among domestic policies

I

I

Quality of economic

reform

I

industrial policy

reform

Note: u Non-crcceleruting Inficrtion Rute of' Unemploymerrt Source:

Author's work.

Figure 6.2

The needjor cmrdination: across Member States or among domestic policies?

European science and technologj~policy

114

and so on)? Of course, the task is made more difficult but simultaneously far more relevant. Furthermore, there exists a literature on social mechanisms that could be mobilized in order to redesign the current policies. By the way, in some instances, the relevant mechanisms could cross the regional, domestic and European boundaries, thus providing a clear basis for multilevel coordination. If a clear national bias in the organization of research systems still prevails and makes coordination of the Member States dificult and if market mechanisms cannot totally monitor the motivation and mobility of researchers, then a strong impulse from Brussels is required to overcome the quite adverse trends observed for European research as a whole. Many factors might justify such a publicly led initiative. In the 1990s, it was expected that the stiffening of competition in the Single European Market, coupled with financial incentives for private R&D and better education, would automatically trigger more dynamism of research and innovation. This strategy has not succeeded and it is more and more recognized that one form or another of direct public intervention in this domain is required. The more advanced countries have pursued during one or two decades systematic public policies and they now enjoy a form of innovation-led growth: Sweden, Finland, Denmark, Germany and Japan (in spite of poor macroeconomic policy) belong in this category. The US itself derives a large part of its present scientific and technological competitive edge from past large public programmes in the defence, health and information sectors. Clearly the public and private sectors are complementary rather than substitutes as far as R&D is concerned (European Commission, 2007b, p. 75). In the EU, conventional macroeconomic policy orientations - monetary stability, reduction of public deficits, wage austerity, labour market flexibility have been unable to create the surge of productivity and new products that explain the negative growth differential with respect to the US from 2002 to 2007. Many economists now agree that innovation policy is one of the rare tools available to speed up growth in Europe (Sapir et al., 2004; Aghion et al., 2006; Pisani-Ferry and Sapir, 2006). Some macroeconomists even propose to allow a big push in infrastructures and knowledge-based industries and dare to advocate the revision of the implementation of the Stability and Growth Pact (SGP) in order to allow such a recovery of innovation and productivity (Artus, 2008). A drastic shift in EU expenditure from passive support for past and largely obsolete compromises (for instance the Common Agricultural Policy [CAP]) to investment in the future would be welcome, but it is not an easy task to design an effective research and innovation policy -

From the Lisbon Agenda to the Lisbon Treaty

115

at the European level. Let us suggest four criteria for selecting research projects: 1.

They should first of all exhibit 'strong externalities' at the European level but not at world level since this would mean financing typical global public goods. 2. Projects should maximize the expected efficiency of research and innovation without any direct concern for a 'fair return' for each Member State. Therefore some new specific intra-European redistributive mechanisms should be investigated. 3. The size of projects should be sufficient to possibly generate large new world markets and/or significant productivity increases and the related risks should be accepted as a condition for the renewal of sources of European growth. 4. If possible, each project should be embedded in dense European networks in order to prevent a rapid delocalization of the benefits of research and innovation.

A quite ambitious programme indeed! Can it succeed in the present context and given the specificity of research?

3 SOME BASIC LIMITS AND CONTRADICTIONS

AFFECTING THE CONSTRUCTION OF ERA The above has put into perspective the agenda for research with respect to the general process of building institutions and procedures, but clearly the fate of ERA depends upon the adequacy of the processes it proposes for research and innovation given the new scientific trends and geopolitical configuration. Science and technology actually display quite specific features compared with other typical economic activities. 3.1 The Primacy of Medium-term Business Strategies Versus the Long-

term Horizon of European Institution-Building European integration is the outcome of nearly half a century of supranational institution-building that started in the quite permissive international context of the so-called 'Pax Americana': the relative stability of the world system was allowing for a rather long-term horizon for European initiatives and the national strategies of Member States. Policy-makers could therefore shape the expectations and incentives of private actors. They actually oriented investment decisions and gave a chance to an ordered

116

European science and rechnologj~policj.

process of catching up and progressive development of endogenous innovations at the national and European levels. The 2000s do not display any more such a favourable configuration for national public intervention in the area of science and technology. First, the very deepening of the processes of globalization now implies that many large firms have divorced from the domestic sphere out of which they emerged and have become truly transnational corporations. Not only can they invest in almost any place in the world, they also organize their R&D activities at world level by optimizing the use of various scientific and technological capabilities of different countries and regions. In doing so, these multinationals do not react to the expectation of a successful ERA but to the existing international distribution of scientific bases. Furthermore, their time horizon is rarely the same as that of European policy-makers. Consequently, the statement of a future strong ERA does not operate as a self-filling prophecy at all: the benefits should be demonstrated first and only then it would be rational for firms to support the process of constructing a strong ERA. A second obstacle relates to the uncertainties associated with the evolution of the world economy in terms of exchange rates, trade policy, differential growth perspectives of the US, Asia and the EU, natural resources prices and so on. The priorities in research and innovation are thus difficult to set. In theory, when uncertainty is so high, a form of centralization and collective organization of strategic choices is better than total decentralization (Heiner, 1985; Aoki. 1994) but this is not at all the feeling of managers, who tend to ask for less and less public intervention. A third difficulty concerns the intrinsically slow process of European 'institutional benchmarking' and the far more rapid change in 'private organizations', including for R&D, frequently in reaction to brusque changes in world markets. For instance, Sweden, Finland and Denmark appear to have found a quite successful model linking research, innovation, job creation and security, but it has proven quite difficult to imitate it or, more realistically, adapt it to quite different national contexts. Since institutions coordinate a complex web of collective procedures and individual behaviours, the best research systems cannot be expected to be diffused smoothly across Europe. This is a serious challenge for the ERA but it is not the only one.

3.2 Do Not Overestimate the Increasing Returns to Scale Associated with a European Market for Scientists Realizing a single labour market for researchers is the very first proposal for making ERA a reality (European Commission, 2007a, pp. 12-14).

From the Lisbon Agenda to the Lisbon Treaty

1 I7

There is no doubt that the current segmentation of national research systems is detrimental to the efficiency of European research as a whole. But are market mechanisms up to the task of prompting a cumulative improvement in European achievements in science and technology? Not really, and that is why the ERA Green Paper proposes other complementary strategies. But it is important to comprehend why research cannot be regulated by pure market mechanisms alone. First of all, static efficiency - that is, the best allocation at continental level of the existing pool of researchers - should not be confused with dynamic efficiency, which is defined as the ability of those researchers to permanently generate advances in knowledge that may or may not be converted into profitable innovations for firms and society. The same confusion has already been rather detrimental to European integration. Back in the early 1990s, was not the implementation of the Single Market for goods and services supposed to generate a permanent increase in the European economic growth rate as a result of better capturing increasing returns to scale? Implicitly, this assumed the persistence of increasing returns to scale observed during the 'Golden Age'. Unfortunately, they were specific to the Fordist productivity regime associated with the mass production of relatively standardized goods (Boyer, 1990). Modest once-and-for-all gains were observed but no steady increase in productivity. The market does not in and by itself directly generate an alternative productive paradigm. Mutatis mutandis, the same disappointment might occur concerning the Single Market for researchers: a limited increase in the volume of publications and citation impacts can be expected but convergence towards a new model for research organization will not necessarily derive from this larger mobility of scientists and engineers. De facto, this assumes the creation of new positive externalities via, for instance, networking, mixing several disciplines, small size research units and relevant university reform. These seem to be the preconditions for major scientific breakthroughs (Hollingsworth, 2007). A better flow of researchers in terms of disciplines, jobs and countries may then help in favouring the sustainability and performance of this new paradigm and organizational form of research. Markets register the performance of alternative organizations and strategies, but they do not suffice to generate relevant innovative forms. 3.3 Basic Science is a Global Public Good.

..

This leads to a related issue: can pure market mechanisms regulate the output of research? The answer is no since the social value of new knowledge exceeds the private appropriation by researchers. Knowledge, even

118

European science and technology policy

if appropriable, displays some characteristics of a public good: it can be used by many individuals and firms without implying any prejudice for the first user. Society as a whole therefore has an interest in the largest possible diffusion of knowledge. That is especially so for basic research. The tension between the need to provide incentives for investing in research and the interest in the maximal diffusion of its output therefore calls for public intervention usually consisting of public financing of basic research, creation of patents for applied research, tax deduction for R&D expenditures, pooling of patents by firms of the same sector and so on. Most of these public interventions take place within the borders of Nation States. The ERA objective is to Europeanize research by transferring these mechanisms from Member State to EU level. But this strategy encounters two distinct difficulties. On the one hand, science and technology are the subject of national policies that stick to the legacy of past economic specialization. A common European approach might therefore hurt some national research and innovation systems. Hence a possible blocking of the process of Europeanization of science and technology. On the other hand, some basic research, for instance, in particle physics, is so costly that the relevant space for cooperation is the world: Europe is too small and research has to become transnational. 3.4

. . . And Academic Organizations are Typically Transnational

More generally speaking, the research community is highly internationalized and not constrained by the frontiers of Europe, the US, China or India. Under the previous Fordist productive paradigm, this feature had no adverse consequences since the advance of techniques came from the ad hoc invention of equipment, learning by using, and the search for increasing returns to scale by firms. Yet when new goods or equipment are increasingly generated by advances in basic knowledge in physics, biology and other disciplines, it becomes difficult to embed the related competitive advantage within a territory, even one as large as Europe. Another threat is that advances in basic science might be hindered by firms' attempts to patent any new knowledge, which threatens the very logic of open science (David, 2007). The biotech and pharmaceutical sectors already exhibit such a trend of a quite extensive definition of intellectual property rights (Coriat and Orsi, 2002). In other words, ERA has to define its strategy with respect to the transnationalization of research, given both the international nature of the scientific community itself and the attempts by multinationals to internalize within themselves the processes of invention and discovery. Not an easy task indeed!

From the Lisbon Agenda to the Li,shorr Treaty

119

3.5 The Networking of Researchers Follows its Own Logic . . .

A converging body of research points out that, more than typical market mechanisms for researchers and scientists, the density and diversity of interactions within academic networks are crucial determinants of inventiveness and of major advances in science. This is true for successful research organizations, for instance, in biology (Hollingsworth, 2007), but also for local clusters (Brenner, 2007). Of course, for major scientific disciplines, such as physics or biology, the related networks established at the initiative of the researchers themselves have no frontiers. 3.6

. . . And it is Difficult to Monitor at the European Level

Therefore three caveats apply to the ERA objective of building a typical European research community. First, the initiative to establish active networks in response to the intellectual and methodological issues that emerge from a shared paradigm belongs to the researchers themselves. Clearly, public financing might redirect some of those networks but it has to be large and long-lasting and it has to accommodate the fact that most pro-jects may fail. Second, past FPs have exhibited a quite voluntarist approach to the construction of networks, which were supposed to involve as many Member States as possible. The search for an equitable distribution of European research funds seems to have been detrimental sometimes to the scientific relevance of the projects and their outcomes. This means that the ability of Brussels authorities to help build relevant and promising research networks is limited. Last but not least, all these programmes raise the same difficult issue: should the networks include non-Europeans and, if so, under which conditions (for instance co-financing)? It has to be taken into account that most of the networks at the frontiers of research are global.

3.7 Europe is Competing More and More with Emerging Research Systems Back in the early phase of globalization, firms delocalized production via foreign direct investments (FDI) but R&D and marketing, that is, the two extremes of the international value chain, were kept at home, being the major sources of the competitive edge. This inertia of R&D centres is progressively eroded by three inter-related factors. Because of the shift in the scientific and productive paradigm associated with ICT, new regions and countries in the world have a chance to build from scratch research. innovation and education systems well-tuned to

120

European science and technology policy

emerging trends in the world economy. Ireland is a good example of such a rapid reversal of a quasi-secular trend. By contrast, mature industrialized countries experienced a lot of difficulties in reforming their past - and quite successful, but now decaying - research systems. In the past, FDI was first directed towards low-wage countries. Later it took account of the perspective of booming domestic markets in host countries. R&D nowadays tends to localize in countries with an abundant and inexpensive pool of highly skilled professionals, engineers and scientists. This is the basis for the emerging Indian research and innovation system. But the size of the potential local market is a second determinant for R&D localization: one of the pillars of the Chinese innovation strategy consists precisely of attracting multinationals hoping that they will finally diffuse their know-how to domestic firms. The second pillar consists of a strong involvement of public finance in the creation of an endogenous and indigenous innovation system. The combination of these two factors explains the very fast catching up of Chinese technological capabilities now and of its scientific basis in the future. Fast growth generates profits and tax revenues that can be reinvested in the import of frontier technologies and R&D expenditures. A sharp contrast with the slow growth and stagnating R&D/GDP ratio in the EU. Finally, the flow of researchers sustaining the US research system has been reduced after 911 1 and this might well be a turning point in the geographical distribution of researchers (Hollingsworth et al., 2008). Yet who are the beneficiaries of this reversal? Mainly Asian countries and not so much Europe, in spite of its ambitious new ERA objective. Thus, the European strategy suffers not only from an unfinished and inadequate institutional configuration and from the resurgence of the principle of subsidiarity but also from an extremely rapid reconfiguration of research and innovation systems in the two other poles of the Triad.

4 CONCLUSION This chapter has analysed the chances of success for ERA by tentatively integrating the specificities of research systems into the evolution of European governance institutions. It has delivered a quite dialectical assessment: those factors that call for an integrated and more efficient ERA are also the factors at the origin of its difficult implementation and problematic efficiency: 1. Policymakers should learn from the difficulties encountered by the first ERA proposal: as a key component of the Lisbon Strategy its

From the Lisbon Agenda to the Lisbon Treat!,

2.

3.

4.

5.

121

configuration had to be redesigned along with the simplification of OMC in 2005. ERA is now part of a growth and job creation strategy. A continuous learning process is needed. The international context is still more challenging than it was in the 1990s. The competition to attract researchers is now not only with the US but also with very large emerging countries such as China and India. Macroeconomic uncertainties about future growth, exchange rates, trade regimes and financial regulations undermine the credibility of ERA with multinationals preferring to locate their R&D activities in highly performing andlor promising (China) research systems. The relatively poor performance of the EU in converting research into innovation and high value-added at the aggregate level should not hide the fact that some Member States have successfully created (Ireland) or redesigned (Finland, Sweden, Denmark) their research and innovation systems and that other countries have consolidated their previous scientific and technological specialization without necessarily following the high-tech route (Germany). But the benchmarking of research systems and the emulation implied by OMC has not been sufficient to overcome the obstacles encountered by countries such as France, Italy or Spain. Successful regimes usually display complementarities between institutions and organizations and synergies between the various incentives and mechanisms associated with these configurations. One of the major ambiguities in the ERA strategy is to seemingly imply that the cross-border externalities of research are more important than the complementarities between research organizations and economic, financial and social institutions at the national (the core hypothesis of 'national systems of innovation' analyses) or local (emphasized by studies of clusters organizing the complementarity between research, training, innovations and production) levels. The hidden costs of a non-integrated ERA should be evaluated and compared with the returns from alternative stimuli of research and innovation. More research on ERA is urgently required in order to understand the functioning of this multilevel system and the respective effectiveness and efficiency of the five instruments available to sustain it: first, large projects at EU level involving trans-frontier networks; second, monitoring of national reforms of research and innovation systems; third, coordination of national strategies and fourth and fifth, support for the formation of clusters with a high research content at local level, either through direct European intervention (via structural funds and policies) or by nation state policies. For the time being, a theory of such a complex and nested system (Figure 6.3) is not available and

From the Lisbon Agendo

to

the Lisbon Treaty

123

this is a hidden obstacle to the full effectiveness of Community-level interventions. 6. While the overarching objective of ERA is clear to promote innovation, growth and high-skilled, high-wage jobs intermediate indicators of performance are not evident at all. In the long run, should the ERA be fully integrated and promote a form of institutional convergence or is perhaps some diversity to be preserved, along the line of present academic and research specializations of Member States? Does ERA intend to simply remove the duplication of efforts and reach a better efficiency along the present research paradigm or does it aim at the emergence of a totally new principle for generating new knowledge on a permanent and cumulative basis? Does not diversity stand for resilience and performance in the long run, with possible inefficiencies and extra costs in the short-medium term? 7 . An outsider's glimpse at the presentation of the new ERA suggests two issues to be discussed more deeply. First, concerning the mobility of researchers, has the role of labour markets in the generation of high-quality research not been overestimated? Within the difficult mix between competition and cooperation - so crucial for research - should the formation of networks and excellence centres not be privileged? This does not mean replicating the quite voluntarist strategy of past FPs, since for the various disciplines a much more subtle monitoring is required, which could be built on academic research on social networks in research. Second, it is risky to bet too much upon the full efficiency of OMC, even if cleverly reformed. Compared with other Community instruments (competition enforcement, European Central Bank and Stability and Growth Pact, Structural Funds, directives, external tariffs and so on), it is a quite weak tool for stimulating research systems dynamism. Either OMC should evolve progressively towards more formal interventions or it should be complemented with an enlargement of the European budget devoted to large Community research programmes, selected in function of the externalities they manifest at continental level and the competitive edge they could give to the EU in stiff competition with the US and Asian giants. -

-

To conclude, the last decade has taught two crucial messages to economists and science/technology experts. The club of developed countries that seemed to be definitely closed after World War I1 is now largely opened to newcomers, such as China, India and to some extent Brazil. Because of their size and dynamism, they have definitely changed international relations and play a key role in the future and stability of the world economy. The old specialization of the EU is therefore challenged: the increasing

124

European science and technology policy

returns to scale in the production of mature industrial goods are reaped in China, not so much in Europe. A vigorous structural adjustment of the EU is required. A second illusion is progressively dissipating. The North, that is. the mature industrial countries, aims at innovation-led growth and sells the related sophisticated investment and/or intermediate products to the South, which in return supplies mass-produced mature/obsolete goods. This is still the case and explains the recent recoveries of Japan and Germany, but the scientific and technological gap will not last forever. China and India are building their own research and innovation systems, investing massively in research and education to launch their own innovation-led growth regime over the next few decades. A divided Europe should not miss this tipping point in world history.

REFERENCES Aghion, P., E. Cohen and J. Pisani-Ferry (2006), Politique economique et croissance en Europe, Rapport du Conseil dlAnalyse Economique no. 59, Paris: La documentation franqaise. Amable, B. (2003), The Diversify of Modern Capitalisms, Oxford: Oxford University Press. Amable, B., R. Barre and R. Boyer (1997). Les systimes d'innovation a I'ire de la globalization, Paris, France: Economica. Aoki, M. (1994), 'The contingent governance of teams: analysis of institutional complementarity', International Economic Review, 35 (3), 657-76. Artus, P. (2008), 'La profitabilite des entreprises non financieres au coeur des perspectives pour les marches', Flash, 367. Beffa, J.-L. (2005), Pour une nouvelle politique industrielle, Paris: La documentation franqaise. Boyer, R. (1997), 'The variety of unequal performance of really existing markets: farewell to Doctor Pangloss?' in: R. Hollingsworth and R. Boyer (eds), Contemporary Capitalism; The Embeddedness o f Institutions, Cambridge: Cambridge University Press, pp. 55-93. Boyer, R. and M. Dehove (2006). Political Goals. Legal Norms and Public Good.7: The Building Blocks of Europe?, Prisme 8, Paris: Cournot Centre for Economic Studies. Brenner, T. (2007), 'Local knowledge resources and knowledge flows'. Industry and Innovations, 14 (2), 121-8. Commission on Growth and Development (2008). The Growth Report. Strategies for Sustained Growth and Inclusive Development, Washington, DC: The International Bank for Reconstruction and DevelopmentJThe World Bank. Coriat, B. and F. Orsi (2002), 'Establishing a new intellectual property rights regime in the United States: origins, content and problems', Research Policy, 31 (8-9), 1491-507. David, P. (2007). 'Innovation and Europe's academic institutions - second

From the Lisbon Agenda to the Lisbon Treaty

125

thoughts about embracing the Bayh-Dole regime', in F. Malerba and S. Brusoni (eds), Perspectives on Innovation, Cambridge: Cambridge University Press, pp. 251-78. European Commission (2000), Towardsa European Research Area, Communication from the Commission, COM(2000) 6 final, 18/01/2000, Brussels. European Commission (2007a), The European Research Area: New Perspectives, Green Paper, COM(2007) 161 final, 04/04/2007, Brussels. European Commission (2007b). Ke.v Figures 2007: Toward.s u European Research Area. Science. Technology and Innovation, Luxembourg: Office for Official Publications of the European Communities (EUR 22572). Facing the Challenge (2004), The Lisbon Strategy for Growth and Development, report from the High-level Group Chaired by Wim Kok, Brussels, November. Freeman, C. (1987), Technology Policy and Economic Perfbrmance: Lessons from Japan, London: Pinter. Gu, S. and B.-A. Lundvall(2006), 'China's innovation system and the move toward harmonious growth and endogenous innovation', Innovation: Management. Policy and Practice. 8 (1-2), 1-26. Gupta, A. and P.K. Dutta (2005), 'Indian Innovation System: Perspective and Challenges', Ministry of Science and Technology, working paper. Hausmann, R., D. Rodrik and A. Velasco (2005), 'Growth diagnostics', Working Paper, March. Heiner, R. (1985), 'Origin of Predictable Behavior: Further Modeling and Applications', American Economic Review, 75 (2), 391-6. Hollingsworth, R. (2007), 'High cognitive complexity and the making of major scientific discoveries', in A. Sales and M. Fournier (eds), Knowledge, Coinmunication and Creativity, London and Thousand Oaks, CA: Sage Publications, pp. 129-55. Hollingsworth, R. and R. Boyer (eds) (1997), Contemporary Capitalism: The Emheddedness of Institutions, Cambridge, MA: Cambridge University Press. Hollingsworth, R., K . Miiller and E.J. Hollingsworth (2008), 'The end of the science superpowers. Could the end of US world dominance over research mark the passing of national science giants', Nature, 454 (7203). 412-13. Kwan, C.H. (2002), 'The rise of China and Asia's flying-geese pattern of economic development: an empirical analysis based on US import statistics', RlETI discussion paper series, no. 02-EZ-009, July. Le Cacheux, J. (2005), 'Budget europeen: le poison du juste retour', Notre Europe, Etudes et Recherches, 41. Meri. T. (2007), 'How mobile are highly qualified human resources in science and technology?', Statistics in Focus - Science and Technology, 75. Odagiri, H., Y. Nakamura and M. Shibuya (1997), 'Research consortia as a vehicle for basic research: the case of the Fifth Generation Computer Project in Japan', Research Policy, 26 (2), 191-207. OECD (Organisation for Economic Co-operation and Development) (2007), OECD Reviews oflnnovation Policy. China. Synthesis Report, Paris: OECD. Okada. Y., K. Nakamura and A. Tohei (2006), 'Is there a significant contribution of public sector in biomedical research in Japan? a detailed analysis of government and university patenting: 1991-2002', COE-RES discussion paper series no. 175. Pisani-Ferry, J. and A. Sapir (2006), Last Exit to Lisbon. Bruegel policy brief, no. 2, Brussels: Bruegel Foundation.

126

European science and technology policy

Rodrigues, M.J. (2006), 'The Lisbon Agenda and the national diversity', mimeograph 20 March. Sapir, A,, P. Aghion, G. Bertola, M. Hellwig, J. Pisani-Ferry, J. Vinals, D. Rosati and H. Wallace (2004), An Agenda for a Growing Europe: The Sapir Report. Oxford: Oxford University Press. Soete, L. (2006), 'Knowledge, policy and innovation'. in: L. Earl and F. Gault (eds), National Innovation, Indicators and Policy, New Horizons in the Economics of Innovation, Northampton, MA, USA and Cheltenham. UK: Edward Elgar, pp. 198-218. Zeitlin, J. (2005), 'Social Europe and Experimentalist Governance: Towards a New Constitutional Compromise', European Governance paper (EUROGOV), NO. C-05-04. Zeitlin, J. and P. Pochet (eds) (2005), The Open Method of Coordination in Action. The European Employment and Social Inclusion Strategies. Brussels: Peter Lang.

7. The returns to public research funding Kris Aerts and Dirk Czarnitzki Technological progress fosters economic growth and research and development (R&D) activities have become crucial for strengthening modern knowledge-based economies' competitiveness (Romer, 1990). However, R&D entails high levels of risk and uncertainty (Dasgupta and Maskin, 1987). Moreover, knowledge creates positive externalities, that is, it can never be fully protected from free-riders (Nelson, 1959; Arrow, 1962). The public character of information leads to incomplete appropriability of investments aimed at creating knowledge. Knowledge cannot be kept proprietary - not even through today's intellectual property rights systems. Newly created knowledge will always leak out to competitors or others and thus the social benefit will be higher than the private return. Under the assumption that companies intend to maximize profits, it will occur that some R&D projects with high social returns will never be carried out, as private costs are higher than private expected returns. This leads to underinvestment in R&D from the social point of view. Another argument refers to financing difficulties encountered by R&D projects due to asymmetric information among borrowers and lenders. Unlike investment in tangible capital, R&D investment is sunk once it has been expensed and the outcome is highly uncertain. This leads to reluctance on the side of lenders to finance R&D, resulting in financial constraints for R&D investment in the business sector (see Hall, 2005, for a survey of such studies). The arguments referring to the positive external effects of R&D and to financing constraints for R&D investment justify governmental intervention as it is present nowadays in most industrialized countries. A strong education system generating a large pool of high-skilled researchers constitutes the ideal foundation for effective and efficient knowledge generation. An important policy is the implementation of systems for legal protection of knowledge, that is, intellectual property rights such as patents (see Chapter 15), trademarks, industrial designs and copyrights. In this chapter, however, the focus is on other public measures, that is, public funding of R&D activities. Public funding of R&D activities comprises three main

128

European science and technology policy

areas: (1) public funding of basic research institutions, such as universities and other institutions for higher education and public research centres; (2) public funding of R&D in the business sector through direct (missionoriented) subsidies and (3) fiscal measures to stimulate private R&D. When discussing R&D in Europe, the EU Lisbon Agenda defines the current goals for innovation policy of national governments. It postulates that by 2010, 3 per cent of G D P should be allocated to R&D activities. One-third should be financed publicly; the remaining two-thirds should be financed privately. Intermediate evaluations show that this goal still lies far ahead: in 2005 only 1.77 per cent of EU25 G D P was spent on R&D. Conversely, the EU's main competitors performed significantly better: US R&D expenditure amounted to 2.62 per cent of G D P and in Japan this number rose to 3.33 per cent (OECD, 2007). As a consequence, an integrated innovatiodresearch action plan was initiated and calls were launched for a major upgrade of research and innovation conditions in Europe. Mobilizing EU funds and the further development of instruments to support research and innovation are among its key objectives. When considering the large amounts of publicly funded research, a natural question forces itself: how effective is this system of public R&D funding? Despite widespread belief in the merit of these measures in offsetting the market failure inherent to the R&D process, there are indications that they may not bring about the desired added-value. Therefore, it is imperative for policy-makers to gain insights into their effectiveness and to assess the returns to this public research funding.

1 FUNDING OF PUBLIC SCIENCE The economic impact of public science has been discussed among economic scholars, professionals and policy-makers for decades. The most obvious contribution of public science to economic growth is the education of the future high-skilled labour force. Possibly equally important, however, is the fact that public science provides research results for the public domain where insights can be picked up by the business sector so that research results from public institutions can be translated into new production processes and products. This mechanism is usually referred to as spillovers or social returns, and several studies linking publicly funded research and economic growth have shown that such spillover effects may be substantial (see Griliches, 1995 for a survey). Third, more active ways to promote knowledge and technology transfer from academe to industry are, among other channels, collaborations in R&D projects, faculty consulting, spin-off creation by universities, university patenting and licensing

The returns to public researchjiinding

129

of technologies. Those activities are typically summarized as industryscience interactions. Several scholars have demonstrated the positive economic benefits of research results produced in (public) science. Quantifying the total benefits of public science is considered to be basically impossible. Several scholars have tried to estimate some effects or the ultimate effect on productivity growth (see Griliches, 1979 as a seminal paper; Griliches, 1986; and Jaffe, 2008 as a recent reflection). Other studies on more specific or interim benefits have looked at a number of different aspects (see Salter and Martin, 2001 for a survey of different types of studies). For instance, Adams (1990) found that cumulative stocks of academic research stimulate productivity growth in industry. Jaffe (1989) has shown that university research contributes to state-level corporate patenting. Narin et al. (1997) analysed knowledge flows from science to industry by looking at citations of scientific papers in US patents and found that knowledge flows from science to industry tripled in a period of six years from 1987 to 1994. They concluded that US industry relies strongly on research results produced in public science. Lichtenberg (1996,2001 and 2003) linked pharmaceutical innovation to lower hospital costs and increased life expectancy while Toole (2007a and 2007b) concluded that university research makes a significant contribution to drug innovation in the pharmaceutical industry. Mansfield (1991) concluded from a survey of 76 US firms that 11 per cent of product inventions and 9 per cent of process inventions would not have been made in the absence of recent academic research. This picture is supported by the Yale and Carnegy Mellon surveys. Both surveys have shown that universities have a significant impact on new product and process development in firms (Cohen et al., 2002). Further studies have shown that academic scientists contributed significantly to the birth of the US biotechnology industry (Zucker and Darby, 1996; Zucker et a]., 1998) and that academic scientists contribute significantly to firm performance when they venture from academe to industry (Zucker et al., 2002; Toole and Czarnitzki, 2007 and 2008). The studies cited above show a positive relationship between investment in public science and (long-term) economic returns for a variety of dimensions and channels through which academe contributes to welfare. However, all focus on the United States. For the European Economic Area, it has been claimed for almost a decade that a so-called 'European Paradox' exists (see Chapter 12 in this book for details). Scholars have argued that EU countries play a leading role globally in terms of top-level scientific output but lag behind in the ability to convert this strength into wealth-generating innovations (see European Commission, 1995 as one of the first references to the 'Paradox').

130

European science und rechnolog~poliq.

Based on this 'Paradox', many European policies for enhancing technology transfer from science to industry were initiated. Such policies aimed at increasing future industry-science interactions as it was assumed that an attitude focused more on commercialization in public science would result in higher economic benefits. Despite the presumably positive effect on technological progress of the increased commercialization of academic inventions, there are some serious threats. Most important is the peril of the 'culture of open science' at universities through a shift in the content of academic research from basic to applied research that focuses on subsequent commercialization (Verspagen, 2006). Many scholars see the relatively open nature of scientific progress at universities. which unlike corporate R&D is characterized by the sharing of knowledge, data and research results, as a key determinant of the success of university research (Dasgupta and David, 1994). A highly critical review of the 'Paradox' and its subsequent possibly misguided policies can be found in Dosi et al. (2006). The authors demonstrate that no overall 'European Paradox' of leading science but weak downstream firms can be observed. Rather significant weaknesses are present in both European public science and industry. Dosi et al. (2006) argue that the research system of Europe is lagging behind that of the US In several areas and that European industry is relatively weak in some key technology areas that are important for future global competitiveness. For instance, Europe is less present in ICT and biotechnologies, has a lower propensity to innovate and participates relatively weakly in global oligopolies. In combination with policies that require public science to demonstrate the 'usefulness' of its research for the current population of firms and to collaborate with industry, these developments may actually 'hollow out' the quality of basic research in Europe. In the same vein, Czarnitzki et al. (2008) show that the increasing commercialization of university research may not lead to higher economic benefits but rather to a decline of academic research. The authors analysed the patenting behaviour of German professors over a period of more than 30 years and showed that three major policy 'eras' could be identified. In the 1980s, professors owned the property rights to their research results and were free to commercialize. However, they also bore all financial risks associated with that. At least from the mid-1990s onwards, the German government systematically introduced policies meant to increase collaboration between science and industry. Along with declining budgets for public research and the movement towards the 'entrepreneurial university' (see, for example, Etzkowitz, 1998 and 2003), faculty directed its attention towards industry. In 2002, the German government abolished the 'professors' privilege': like under the 1980 Bayh-Dole Act in the US,

The returns to public research funding

131

professors no longer own the property rights to their research results but universities can take out patents, for instance. In addition, the German government even installed state-level patent valorization offices assisting universities without professional technology transfer offices in filing patents. Czarnitzki et al. (2008) show that the quality of faculty patenting declines in line with these changes in policy. In the 1980s, university patents are on average of higher quality than a control group of corporate patents. In the mid-1990s, the quality of faculty patents converges towards that of industry patents. Finally, due to the changing incentive structure for universities after 2002, the quality of faculty patents even drops below the average quality of corporate patents. Czarnitzki et al. (2008) argue that the increased pressure to commercialize distracts researchers from really basic research, which is in line with the critical comments made by Dosi et al. (2006). As a consequence, European policy should carefully review the incentive structures in science and industry before designing future instruments. At the moment, it appears that current policy favours shortterm goals rather than the long-term building of knowledge capabilities in Europe to a large extent.

2

PUBLIC FUNDING OF PRIVATE RESEARCH

While closely related to the financing of public research institutions, government funding of R&D in industry is typically treated separately in the economic literature. Private research is supported by public money in basically every industrialized country. Public authorities can opt between two modes of transferring public funding to the private R&D sector: either direct or indirect. Direct grants in support of R&D and other innovation activities can address various sources of market failure and are, theoretically at least, an appropriate policy response to market imperfections. Induced by R&D grants, the private sector is supposed to invest in R&D activity sufficiently to equalize the private and social returns to innovation, thus maximizing the benefits society receives from new technology. Practically, however, government agencies administering R&D grants are unlikely to possess the necessary information on the incidence and importance of market failures. Ignorance, information asymmetries between innovator and government agency and moral hazard on the part of the inventor or innovating firm make it difficult if not impossible for the government agency to distribute grants so as to reduce or eliminate the gap between the social and private returns to R&D. Furthermore, the administration of R&D grants may pursue objectives other than the stated ones, may be subject to

132

European science and technologj. policy

political pressure, bureaucratic objectives, corruption and incompetence and is costly. The resulting government failure (Winston, 2006) may even be more important than the market failures the grants are supposed to correct for. In contrast to direct grants, tax credits are considered to be a neutral form of encouragement for R&D. All firms incurring eligible R&D expenditures, irrespective of industry, size and innovation activity objective, can claim them. The attractiveness to policy-makers of tax credits lies in the fact that their administration does not involve arbitrary decisions regarding the distribution of R&D support among sectors, regions, industries or firms. They have, however, shortcomings of their own. They stimulate overall R&D activity but do not address the sources of market failures affecting innovation activities. They affect the composition of R&D, favouring activities promising the largest short-term profits. Projects with high potential social rates of return and investment in exploratory projects and development of research infrastructure may be less stimulated by tax credits (David et al., 2000; Hall and Van Reenen, 2000). 2.1

Direct Public R&D Funding: Subsidies

With respect to direct R&D funding, different impacts are assessed in the literature. They can be centred on three levels: (1) input, (2) output and (3) behaviour. Each of these levels will be discussed in the following paragraphs. 2.1.1 Input analyses Starting with Blank and Stigler (1957), a vast body of literature has emerged on the 'additionality' of public R&D subsidies. Whenever a government provides subsidies, they are subject to the danger of moral hazard. Public grants are meant for research projects where the expected private costs are not expected to be covered by the expected private returns but where the social returns may well largely exceed the private ones. However, once a subsidy programme is in place, every private R&D performer has an incentive to apply for support and just substitute public money for private funds. Public funds would thus only crowd out private investment but not lead to increased R&D in the economy. David et al. (2000) have reviewed the literature on such crowding-out effects. While studies at the macroeconomic level most frequently rejected crowding-out effects, microeconomic studies conducted until the early 2000s were more ambiguous. However, David et al. (2000) identified some systematic shortcomings in the econometric methods that had been applied to the question of crowding-out effects at the firm level. Since then, a number of studies

The returns to public resrurch funding

133

using new econometric methodologies have emerged. So far, firm level databases from Austria, Belgium, Denmark, Finland, France, Germany, Ireland, Israel, Norway, Spain, Sweden, Turkey and the US have been subject to an evaluation using state-of-the-art econometric techniques (see Aerts and Czarnitzki, 2008 for a survey). While most studies reject complete crowding-out effects, there is still no consensus. The main reason is the use of different econometric techniques and of different data for a broad set of countries. These data are not harmonized and structured very differently (surveys, administrative data, combinations; cross-sectional versus panel structure), while each country also has its own specific science and technology (S&T) policy.

2.1.2 Output analyses While investigating potential crowding-out effects of public R&D funding on private R&D expenditure and personnel is indisputably highly relevant for innovation policy evaluation, a rejection of such effects does not necessarily imply that increased R&D spending really induces technological progress and subsequently economic value creation. First, R&D wages account for a significant share of total R&D expenditure (in Flanders, for instance, on average around 67 per cent according to Czarnitzki et al., 2006). Therefore, the cost of the input factor R&D personnel, that is, R&D wages, plays an important role in additionality research on R&D employment. R&D wages may impact adversely on the S&T policy measures introduced by government. The input additionality may simply end up in higher wages and inflate positive additionality effects. Reinthaler and Wolff (2004) observed a simultaneous increase in national R&D investment and R&D employment. The increase in R&D staff was smaller, though, which caused them to conclude that scientists' wages also experienced an increase. Goolsbee (1998) concluded that increases in R&D expenditure are mainly allocated to researcher wages and not to research effort. Ebersberger (2004) claimed that the Finnish innovation system was marked by an adequate inflow of researchers and that therefore an increase in R&D investment was fully translated into an increase in R&D employment but he did not put his statement to the test. This was done by Ali-Yrkko (2005), however, who concluded that in addition to a positive effect on the number of R&D employees, R&D subsidies also had a significantly positive effect on researcher wages. Uqdogruk (2004) found indications that in Turkey, R&D subsidies significantly increased researcher wages. Aerts (2008) also found such effects in an analysis of the impact of Flemish R&D subsidies on R&D expenditure, R&D employment and R&D wages. So although there is substantial ambiguity concerning

134

European science und techno log^ polic,

the impact of R&D subsidies on R&D employment. there is consensus on the fact that researcher wages increase when a company receives an R&D subsidy. Therefore, research on input additionality is completed by looking at the impact not only on R&D expenditure but also on R&D personnel and wage structure. Second, the actual reinforcement of private R&D activities may be directed towards more risky and consequently less successful projects. Setter and Tishler (2005) modelled the impact of subsidies on advanced state-of-the-art technologies, which are characterized by substantial initial investment followed by a period of increasing returns and, as the technology matures, decreasing marginal benefits. The authors found that highly uncertain technologies receive a larger share of the budget when budget levels are very high. Czarnitzki and Toole (2007) argued within a real options framework that outcome uncertainty reduces R&D investment. In an empirical study using German data, they showed that firms are indeed sensitive to product market uncertainty. Investment decreases with the degree of uncertainty. If firms receive R&D subsidies, however, they react much less to uncertainty. In fact, the subsidy almost fully offsets the negative effect of uncertainty on investment. That points to the fact that subsidies may result in investment in more risky projects. Third, inefficiencies may arise from research duplication (Irwin and Klenow, 1996; David et al., 2000). The excessive concentration of private expenditure in certain areas, driven by racing behaviour or business stealIng, may push down the social marginal rate of return on the R&D performed by the private sector as some research areas are ignored (David et al., 2000). However, Irwin and Klenow (1996) conclude from their empirical analysis of the US semiconductor industry that R&D subsidies enable companies in R&D consortia to reduce member duplication of R&D expenditure. Furthermore, Dasgupta and Maskin (1987. p. 582) qualify this potential problem as 'parallelism need not imply waste'. Hence, extending additionality research on R&D inputs to include the analysis of the induced innovative and economic output is imperative to get a full understanding of the impact of R&D subsidies. Klette et al. (2000) surveyed the literature on evaluation studies measuring firm growth, firm value, patents and so on. Since then, researchers have also been evaluating measures concerning product and process innovation. More recent research elaborates on the crowding-out question by linking privately financed R&D and publicly induced R&D to innovative activity. For instance, a two-equation model is considered. First, an input analysis is conducted on R&D expenditure and the response to the receipt of a subsidy. In the second equation, a knowledge production

The returns to public research funding

135

function or other output function is estimated relating a measure of innovative output to firm R&D spending and other covariates. The first step allows for disentangling total R&D spending into two components: first, the part of R&D that would have been conducted in the absence of subsidies, that is, the estimated counterfactual situation, and second, the part of R&D expenditure that has been induced by the receipt of subsidies, which comprises the amount of the subsidy itself and the additionally stimulated privately financed R&D. The two components add up to the total observed R&D spending but the decomposition allows for analysing the productivity of purely privately financed R&D and that of the additional publicly induced R&D. See Czarnitzki and Hussinger (2004) and Czarnitzki and Licht (2006) for studies on patent output, for instance.

2.1.3 'Behavioural' analyses A recent stream of additionality research extends the evaluation criteria beyond directly measurable input and output indicators and evaluates how subsidies affect company behaviour, for example, changing collaboration behaviour, sustainability of such networking, changes in companies' R&D management and so on. The concept of 'behavioural additionality' was introduced by Buisseret et al. (1995). Falk (2007), for example, finds that supported companies enhance their innovative capabilities, improve competence building in general and employ new technologies and R&D procedures elsewhere. However, she identifies a major problem inherent to the concept of behavioural additionality: the lack of appropriate measures for the mostly intangible merits of behavioural additionality. Czarnitzki et al. (2007) have analysed R&D spending and patenting as a response to subsidies and collaborative R&D using harmonized survey data (Community Innovation Survey 111) from Germany and Finland. Their study on heterogeneous treatments not only allows for identifying actual input additionality effects but also enables an investigation of potential effects, that is, whether there is additional, promising room for policy. An important difference between Finland and Germany is the share of companies involved in collaborative research, which is much higher in Finland. The authors find that collaboration and subsidies lead to positive additionality in both countries. A remarkable difference, however, is that fostering further collaborative R&D in Finland may not lead to more innovation, while in Germany it would be a promising tool for promoting further innovation. In summary, current European microeconomic studies tend to find positive effects of direct subsidies confirming both positive input and output additionality.

136

European science and technologj~policj~

2.2 Indirect Public R&D Funding: Tax Credits Hall and Van Reenen (2000) have surveyed the literature on the effectiveness of R&D tax credits (see also Van Pottelsberghe et al., 2003). Pioneering research in this domain was initiated in the early 1980s with US data (Eisner et al., 1983). The main reason was that the US was among the first countries to introduce an R&D tax scheme (in 1981). Firm-level evaluations were later carried out in other countries like Australia, Canada, France, Japan, the Netherlands and Sweden but less frequently. Research was also conducted at the macro level (see, for example, Guellec and Van Pottelsberghe, 1999). Different approaches can be found in the literature but in particular the estimation of the marginal cost of R&D has become very popular in recent years in the evaluation of the impact of tax credits. Although different methods and datasets are employed and different schemes apply in different countries (see, for example, Van Pottelsberghe et al., 2003), the conclusions from empirical research leave little ambiguity: R&D tax credits stimulate private R&D spending. A few attempts have also been made to assess the impact of R&D tax credits on private R&D wages. Although in this chapter, the explicit focus is on direct R&D funding, the main results are briefly mentioned. Marey and Borghans (2000) estimated the wage effects of R&D tax incentives in the Netherlands and estimated average elasticities of R&D wages in relation to total sectoral R&D expenditure of 0.52 in the short run and 0.38 in the long run. Lokshin and Mohnen (2008) estimated a short-run elasticity of 0.10 and a long-run elasticity of 0.12 in the Netherlands. Haegeland and Meren (2007) assessed the Norwegian R&D tax credit measure and estimated an elasticity of 0.33. The main critique on the analysis of the impact of R&D tax credits is the relabelling issue. Firms eligible for R&D tax allowances can be expected to label any investment (slightly) related to the area of R&D as R&D expenditure. This may seriously distort the estimates of potential additionality effects.

3 IMPLICATIONS FOR POLICY AND FURTHER RESEARCH 'The positive results identified above for business sector subsidies seem to be encouraging for reaching the goals of the Lisbon Agenda. Reinforcing European innovation policies may indeed have positive effects on R&D carried out in the business sector. However, it has to be understood that the 3 per cent R&D-to-GDP goal of the Lisbon Agenda does not constitute

The returns to public research Junding

137

any theory-based benchmark of R&D activity that would ensure higher welfare. It is just a policy goal to keep up with other industrialized countries like the US and Japan in terms of R&D. If one puts the Lisbon goals aside, it has to be questioned whether direct subsidies for the business sector are the first-best option to invest public funds. There is no study conducting a general equilibrium analysis trying to answer whether public funds should be invested in business R&D or rather into presumably more basic public R&D carried out at universities and similar institutions. For OECD countries. Cincera et al. (2008) conducted a macroeconomic benchmarking exercise on the level of business sector R&D. They analysed the 'trigger' effect of direct subsidies, higher education R&D spending (HERD) and R&D carried out in governmental (research) institutions (GOVERD). They found that both subsidies and HERD induced R&D spending in the business sector. The main goal of the study by Cincera et al. (2008), however, was to identify inefficiencies in national innovation systems. They estimated so-called frontier production functions (both parametrically and non-parametrically using Data Envelopment Analysis) for a time period of 20 years and concluded that there is no unique policy mix avoiding inefficiencies. According to their analysis, the most 'efficient' countries in terms of transforming public R&D, in the form of either subsidies or HERD, are not surprisingly the US, Japan and, depending on the model specification, European countries such as Finland, Germany, the Netherlands and Switzerland. The composition of public R&D spending (as in Figure 7.1 at the country level) reveals that those countries show very different shares of subsidies, HERD and GOVERD. Cincera et al. (2008) concluded that it would be necessary to conduct much more detailed studies, at least at the sectoral level, to identify how well a country's individual policy schemes and industry structure match. The study of Cincera et al. (2008) only conducts an 'industry input' analysis of the R&D process and does not analyse how public R&D spending finally ends up in marketable products and new production processes that would make Europe more competitive. However, a well-functioning link between public science and industrial R&D is the key ingredient to higher welfare in knowledge-based economies and thus the first step to higher total welfare. Returning to the critical statements about the 'European Paradox' cited above, it seems to be highly relevant to balance innovation policy according to a country's individual strengths and weaknesses in science and industry. Conducting a general equilibrium analysis shedding light on the improved allocation of public funds among certain fields of public science and a nation's industry constitutes a big challenge for further research on

European scinzce and tecl~nolog~ poliq

BERD as %GDP HERD as %GDP rest GERD as %GDP 0 %BERD financed bv aovernment %HERD financed bv aovernment

+

Source: OECD, 2007. Figure 7.1

Decomposition o j ' G E R D and shurefinunced hj. the government

t h e returns t o public funding to. Clearly, it would b e most important t o t a k e interactions between science a n d industry i n t o account.

REFERENCES Adams. J.D. (1990), 'Fundamental stocks of knowledge and productivity growth'. Journal of'Politica1 Economy, 98 (4), 673-702. Aerts, K. (2008), 'Who writes the pay slip? D o R&D subsidies merely increase researcher wages?', research report O R 0806, Leuven: K.U. Leuven, Faculty of Business and Economics. Aerts, K. and D. Czarnitzki (2008), 'Econometric evaluation of public R&D policies: literature review and a guide for further research', mimeo, Leuven, Belgium: K.U.Leuven, Faculty of Business and Economics. Ali-Yrkko, J. (2005), 'Impact of public R&D financing on employment', Research Institute of the Finnish Economy discussion paper no. 980, Helsinki. Arrow, K.J. (1962) 'Economic welfare and the allocations of resources of invention', in: R.R. Nelson (ed.), The Rate und Direction of' Inventire Activitj: Economic und Social Factors, Princeton, NJ: Princeton University Press, pp. 609-26. Blank, D.M. and G.J. Stigler (1957). The Demand and Supp1.v of Scientific Personnel, New York: National Bureau of Economic Research. Buisseret, T.J., H.M. Cameron and L. Georghiou (1995) 'What difference does it make? Additionality in public support of R&D in large firms', Internutionul Journal of Technology Managemmt, 10 (4/5/6), 587-600. Cincera, M., D. Czarnitzki and S. Thorwarth (2008). Assessing the EfJiciencj of Public Policies to Support R&D Activities in the EU, report commissioned

The returns to public reseurch funding

139

by the Directorate General Economic and Financial Affairs of the European Commission. Cohen, W.M., A. Goto, A. Nagata, R. Nelson and J.P. Walsh (2002), 'R&D spillovers, patents and the incentives to innovate in Japan and the United States', Research Policy, 31 (8-9), 1349-67. Czarnitzki, D . and K. Hussinger (2004), 'The link between R&D subsidies, R&D spending and technological performance', Center for European Economic Research (ZEW) discussion paper no. 04-056. Mannheim: ZEW. Czarnitzki, D. and G. Licht (2006) 'Additionality of public R&D grants in a transition economy: the case of Eastern Germany', Economics oj'Tran.sition, 14 (1). 101-31. Czarnitzki, D. and A.A. Toole (2007) 'Business R&D and the interplay of R&D subsidies and product market uncertainty', Review of Industrial Organization, 31 (3), 169-8 1. Czarnitzki, D., B. Ebersberger and A. Fier (2007), 'The relationship between R&D collaboration, subsidies and R&D performance: empirical evidence from Finland and Germany', Journal oj'Applied Econometrics, 22 (7), 1347-66. Czarnitzki, D., K. Hussinger and C. Schneider (2008), 'Commercializing academic research: the quality of faculty patenting', Center for European Economic Research (ZEW) discussion paper no. 08-069, Mannheim: ZEW. Czarnitzki, D. (ed.), K . Aerts, B. Cassiman, M. Hoskens, M. Vanhee and R . Veugelers (2006), Research, Development and Innovation in Flandc,rs 2004, Institute for the Promotion of Innovation by Science and Technology in Flanders (IWT) M&A study no. 55, Brussels: IWT. Dasgupta, P. and P.A. David (1994) 'Toward a new economics of science'. Research Policy, 23 (5), 487-521. Dasgupta, P. and E. Maskin (1987) 'The simple economics of research portfolios', The Economic Journal, 97 (387), 581-95. David, P.A., B.H. Hall and A.A. Toole (2000), 'Is public R&D a complement or substitute for private R&D? A review of the econometric evidence', Rrsrurch Policy, 29 (4-5), 497--529. Dosi, G., P. Llerena and M. Sylos Labini (2006), 'The relationships between science, technologies and their industrial exploitation: an illustration through the myths and realities of the so-called "European Paradox"', Resrarclr Polic,~, 35 (lo), 1450-64. Ebersberger, B. (2004), 'Labour demand effect of public R&D funding', Technical Research Centre of Finland (VTT) working paper no. 9, Helsinki. Eisner, R., H.A. Steven and M.A. Sullivan (1983), 'Tax incentives and R&D expenditures', in: Ecole Nationale de la Statistique et de I'Administration Economique and National Bureau of Economic Research (eds), Procecrlings of' the Conference on Quuntitative Studies qf'Reseurch und Development in Industry, no. 2, Paris: CNRS, pp. 375466. Etzkowitz, H. (1998), 'The norms of entrepreneurial science: cognitive effects of the new university-industry linkages', Research Policy, 27 (8), 823 -33. Etzkowitz, H. (2003), 'Research groups as "quasi-firms": the invention of the entrepreneurial university', Research Policy, 32 (I), 109-21. European Commission (l995), Green Paper on Innovation, COM(95) 688, December 1995, Brussels. Falk, R. (2007), 'Measuring the effects of public support schemes on firms' innovation activities. Survey evidence from Austria', Research Policy, 36 (5), 665-79.

140

European science and technolog. policj

Goolsbee, A. (1998), 'Does government R&D policy mainly benefit scientists and engineers?', American Economic Review. 88 (21, 298-302. Griliches, Z. (1979). 'Issues in assessing the contributions of research and development to productivity growth', Bell Journal of Economics, 10 ( I ) , 92-1 16. Griliches, Z. (1986), 'Economic Data Issues', in: M. Intrilligator and Z. Griliches (eds), Handbook of Econometrics, vol. 111, Amsterdam: North Holland. Griliches, Z. (1995), 'R&D and Productivity', in: P. Stoneman (ed.). Hundhook of' Industrial Innovation, London: Blackwell Publishers, pp. 52-89. Guellec, D. and B. Van Pottelsberghe (l999), 'Does government support stimulate private R&D?', OECD Economic Studies No. 29, 1997111, Paris, pp. 95-122. Haegeland, T. and J. Meren (2007), Input Additionalitj~in the Norwegian R&D Tux Credit Scheme, Oslo: Statistics Norway. Hall, B.H. (2005), 'The financing of innovation'. in: S. Shane (ed.), B l a c k d l Handbook of Technology and Innovation Management, Oxford: Blackwell Publishers. Hall, B.H. and J. Van Reenen (2000), 'How effective are fiscal incentives for R&D? a review of the evidence', Research Policy, 29 ( 4 5 ) . 449-69. Irwin, D. A. and P. J. Klenow (1996), 'High-tech R&D subsidies: estimating the effects of Sematech', Journal of International Economics, 40 (3-4). 32344. Jaffe, A. (1989), 'The real effects of academic research'. .4nierican Economic Review, 97 (5), 957-70. Jaffe, A.B. (2008), 'The "science of science policy": reflections on the important questions and the challenges they present'. Journal of Technology Transfer, 33 (2), 131-39. Klette, T. J., J. Meren and Z. Griliches (2000), 'Do subsidies to commercial R&D reduce market failures? Microeconometric evaluation studies'. Research Policj~, 29 ( 4 5 ) , 471-95. Lichtenberg, F. (1996), 'Do (more and better) drugs keep people out of hospitals?', American Economic Review, 86 (2), 384-8. Lichtenberg, F. (2001), 'Are the benefits of newer drugs worth their cost? Evidence from the 1996 MEPS', Health Affairs, 20 (5), 241-52. Lichtenberg, F. (2003), 'Pharmaceutical innovation, mortality reduction, and economic growth' in K.M. Murphy and R.H. Topel (eds), Measuring the Gains from Medical Research: An Economic Approach, Chicago, IL: University of Chicago Press. Lokshin, B. and P. Mohnen (2008), 'Wage effects of R&D tax incentives: evidence from the Netherlands', UNU-MERIT working paper no. 2008-034. Maastricht: UNU-MERIT. Mansfield, E. (1991). 'Academic research and industrial innovation', Research Policy, 20 (I), 1-1 2. Marey, P. and L. Borghans (2000), 'Wage elasticities of the supply of R&D workers in the Netherlands', mimeo, University of Maastricht, Maastricht. Narin, F., K. Hamilton and D. Olivastro (1997). 'The linkages between US technology and public science', Research Policy, 26 (31, 3 17-30. Nelson, R.R. (1959), 'The simple economics of basic scientific research', Journal of Political Economy, 67 (3), 297-306. OECD (Organisation for Economic Co-operation and Development) (2007), Main Science and Technology Indicators, Paris: OECD. Reinthaler, V. and G.B. Wolff (2004). 'The effectiveness of subsidies revisited: accounting for wage and employment effects in business R&D', Center for

The returns to p h l i c researchfunding

141

European Integration Studies (ZEI) working paper no. B21-2004, Rheinische Friedrich-Wilhelms-Universitat, Bonn. Romer. P.M. (1990), 'Endogenous technological change', Journal o f Political Economy, 98 (5), 71-102. Salter, A.J. and B.R. Martin (2001), 'The economic benefits of publicly funded basic research: a critical review', Research Policy, 30 (3), 509-32. Setter, 0. and A. Tishler (2005) Investment Policies in Advanced Dc.fi.nse R&D Programs, Tel Aviv: Tel Aviv University. Toole, A.A. (2007a), 'Does public scientific research complement private research and development investment in the pharmaceutical industry?', Journal of Law ond Economics, 50 (1). 81- 104. Toole, A. A. (2007b), 'The impact of public basic research on industrial innovation: evidence from the pharmaceutical industry', mimeo, Rutgers University, New Brunswick, NJ. Toole, A. A. and D. Czarnitzki (2007), 'Biomedical academic entrepreneurship through the SBIR program', Journal of Economic Behavior and Organization, 63 (4), 716-38. Toole, A.A. and D. Czarnitzki (2008), 'Exploring the relationship between scientist human capital and firm performance: the case of biomedical academic entrepreneurs in the SBIR program', Management Science, 55 (I), 104-14. Uqdogruk, Y. (2004), 'Do researchers benefit from R&D support programs in Turkey?', paper presented at the 10th SMYE Conference, Geneva. Van Pottelsberghe, B., S. Nysten and E. Megally (2003). Evaluation of Current Fiscal Incentiver for B2csine.r.r R&D in Belgium, Brussels: Solvay Business School and Service Public Federal de Programmation Politique Scientifique. Verspagen, B. (2006), 'University research, intellectual property rights and European innovation systems', Journal of Economic Survey, 20 (4), 607-32. Winston, C. (2006), Government Failure versus Murket Failure: Microeconomics Policy Research and Government Perfurmance, Washington, DC: AEl-Brooking Joint Center for Regulatory Studies. Zucker, L.G. and M.R. Darby (1996), 'Star scientists and institutional transformation: patterns of invention and innovation in the formation of the biotechnology industry', Proceedings of the National Academies ofscience, 93 (23), 12709--16. Zucker, L.G., M.R. Darby and J.S. Armstrong (2002) 'Commercializing knowledge: university science, knowledge capture, and firm performance in biotechnology', Munagemmt Science, 48 (I), 138-53. Zucker, L.G., M.R. Darby and M.B. Brewer (1998), 'Intellectual human capital and the birth of US biotechnology enterprises', American Economic Review, 88 (I), 290-306.

8. Scale and scope in research Nicholas S. Vonortas The issue of economies of scale and scope in research is arguably as old as research itself. Recently, however, deep changes in the way we organize research and an improved understanding of the relationship between research and growth have rekindled interest in the subject. A case in point is the concept of the European Research Area (ERA), which appears to be founded on the implicit understanding that economies of scale and scope matter in research funding and execution and that coordination and collaboration (at various levels) are therefore beneficial, whereas fragrnentation and dispersal are inefficient. Hence the emphasis on 'critical mass'. The rest of the chapter is divided into the following sections. Section 1 defines scale and scope economies in research and identifies reasons why empirical verification of such effects may be fraught with difficulties. Section 2 traces the extensive literature on the neo-Schumpeterian hypotheses that laid the foundation for the modern discussion on scale and scope effects in research. Section 3 introduces cooperative research as a reaction to a perceived need for correcting market failure and the inadequacies of individual organizations to go it alone. The section also describes how assumptions of research cooperation are built on the concepts of economies of scale and scope. Finally, Section 4 concludes by relating the discussion on scale and scope, and the related need for government intervention, to the evolution of the European Framework Programmes for Research and the development of the ERA.

1 BASIC CONCEPTS: ECONOMIES OF SCALE AND

SCOPE IN RESEARCH' In this section, definitions are built. Unless otherwise indicated, the following terms are used. 'Output' refers to the outputs of research, such as new concepts, ideas, products, services and production processes and their (imperfect) measures, such as publications, reports, patents, blueprints and prototypeslpilots. 'Production' refers to the transformation of research inputs into research outputs. 'Inputs' refers to the resources

&ale and scope in reseurch

143

utilized in this production process, including both labour (researchers, support personnel) and capital (research equipment and infrastructure). 'Technique' (or 'production technique') refers to the specific technology utilized in this transformation of research inputs into research outputs. In other words, it refers to the specific organizational/managerial structure that is utilized to achieve the most efficient transformation (largest amount of research output from a given level of research inputs). There are several techniques that can be utilized to achieve a certain kind and amount of research output. The choice will depend on both internal factors - capabilities and resources of the organization or group and external factors of organizations that undertake the research relative prices (availability) of inputs, norms in science and technology, external environment (for example, social acceptability of the research) and government policy (for example, regulation). The choice of technique and capacity level represents a long-run decision when all inputs are variable. However, once a certain type of research project of a certain size is put in place, the research enterprise operates under short-term conditions. How easy it is to switch techniques and/or the size of the research project depends on the nature of the research itself. Scale effects may be present at several levels: the research project, the organization, the group of organizations (if cooperative project) and the geographical area (country, region). The concept of returns to scale is used to describe what happens to output when all inputs are increased together (by the same proportion) when a specific technique is in place. An increase in all inputs resulting in a more than proportional increase in research output indicates increasing returns to scale. Increases of the same proportion mean constant returns to scale. If the research output increases by a smaller proportion than all of the research inputs, we have decreasing returns to scale. Economic theory offers three possible reasons for increasing returns to scale: specialization, dimensional effects and indivisibilities. Specialization implies a finer division of labour as the research project grows larger. Dimensional effects refer to the case where a larger unit of capital produces disproportionately more than a smaller unit. Indivisibilities exist when certain inputs are available only in certain minimum sizes: larger research scales may utilize such inputs more efficiently. For instance, professional management could be such an input. On the other hand, the prevailing reason for decreasing returns is the coordination and control complications of large-size operations. Under the assumption that all inputs are in perfectly elastic supply to the organization, the scale effects above translate into cost effects: increasing returns are reflected into economies of scale - decreasing long-term -

-

144

European science and techno log^, policj

average cost - whereas decreasing returns to scale are reflected into diseconomies of scale - increasing long-term average cost. Economies of scope are present when the same one research operation deals with several subjects because of cost advantages. Economies of scope in research may result in situations where several research projects involve at least some of the same management and science and technology (S&T) knowledge, skills and capital equipment, thus allowing for crossfertilization and productive exchange. Economies of scope are distinct from economies of scale. Similarly. diseconomies of scope are unrelated to diseconomies of scale. The extent of cost savings from joint research can be shown with the following example:

where: C(R,) is the cost of undertaking R,, units of research in mobile phones alone; C(R,) is the cost of undertaking R, units of research in mobile phone software alone; C(R, + R,) is the cost of undertaking these two types of research jointly. If ES > 0 economies of scope exist. If ES < 0 then diseconomies of scope exist. The latter are most likely to occur when the proliferation of different research subjects becomes so great that the task of coordinating and controlling the research process (transformation) leads to diminishing returns to management. That is, it would be more efficient if different research subjects were undertaken by different independent organizations. Occasionally one comes across the term 'economies of complementarity' used to indicate cost savings across research projects due to synergies, knowledge transfer, cross-fertilization of ideas between individuals/teams and the like. Economies of complementarity are important, especially for 'virtual' research enterprises such as those described in Section 3 that combine two or more independent organizations pursuing collaboratively specific outcomes. However, economic theory does not clearly differentiate such concepts and we will consider them in this chapter as part of economies of scope. While scale and scope effects have been discussed separately above, in practice they frequently occur simultaneously and are difficult to identify empirically. On the one hand, the input-output relationships at varying

145

Scale and scope in research

sizes of the research enterprise turn out to be much more complicated than implied by simple microeconomic theory. In particular, aspects of research management practice and technique greatly complicate the picture. Consider, for instance, generic, applied and development research in different sectors by different sizes and types of organizations or groups of organizations. On the other hand, established cost accounting systems, designed for different purposes, obscure the existence of these effects in the level of detail in which we are interested. Difficulties include: 0

Research output difSerentiation. The cost disadvantages of a smaller-scale research enterprise may be offset by differentiation of the research outputs in ways that achieve much higher prices for them for instance, high value-added niche markets. Hence, the coexistence in the same technology areas of research enterprises of very different sizes. DzfSerences- between sectors and technology areas. In the presence of significant economies of scale in research, one would expect the size of competing research enterprises to grow and competition to be of classic oligopolistic nature. Such a strategy could, however, prove fatal in the presence of new techniques coming from outside the narrow confines of the industry in question. A case in point is the pharmaceutical industry and the biotechnology revolution. Different inputs and techniques. Only rarely, if ever, would a large research enterprise resemble - in terms of inputs, structure and techniques a much smaller research operation. Therefore, the concept of capital-labour ratios requiring uniform capital and labour units becomes questionable. Equipment and labour skills are usually commensurate with particular scales of the research operation. Broad ranges of eficient research output. In the presence of constant returns to scale in research, the long-term average cost curves are flat rather than U-shaped as required by economies and diseconomies of scale. The shape and position of short-term cost curves will then be the result of managerial decisions and the influence of particular characteristics of the research operationlproject in question. Imhulunces among dzflerent stuges o j research. Imbalances among the different stages of the research enterprise may be the result of deliberate managerial decisions rather than indivisibilities in capital equipment. Excess capacity may be built into the design of a research operation due to anticipated needs. For instance, it is well understood that the cost of successive stages of research from more basic to applied, development, prototyping - increases in a geometric progression. Research managers may build excess capacity in the -

0

0

-

0

0

-

146

0

European science and technology policy

earlier (less costly) stages of research and tolerate the cost of parallel projects in competing approaches in order to eliminate as much uncertainty as possible before moving closer to market. Research output mix. The concept of economies of scope does not take into consideration the demand side, that is, the possibility that a research enterprise is set up to respond to different kinds of research output mixes. This requires some flexibility built into the system, which may make it appear less cost-efficient than what it could be if it were to serve exclusively the current research output mix. Pre-emptive investment in research scale. Occasionally, an organization or group of organizations may invest in a level of research capacity beyond what is currently needed. The reason is dynamic and strategy related: competitors, current and prospective. may be discouraged by the existence of idle excess capacity that can be used quickly when the opportunity arises. They may thus decide to abandon plans for expansion of existing, or establishment of new, research operations in an existing market. Scale and technological change. Efficient scale requirements drastically change over time according to the features of existing technology. This works at two levels. On the one hand, there are very extensive differences across sectors and technological areas. There is, for instance, hardly any resemblance of scale and structure among research operations in the chemical industry and mobile telephony. Even within the same broad sector, there are quite significant differences - see bulk and speciality chemicals. On the other hand, minimum efficient scales of research will change dramatically within the same sector at different stages of the sector's evolution, typically getting larger as the sector matures, products become more standardized and research opportunities decrease. Scope and technological change. Opportunities of economies of scope and complementarity will also differ across technology areas and will change over time on the basis of technology evolution. This creates significant room for managerial intervention and strategic considerations.

2 NEO-SCHUMPETERIAN HYPOTHESES Schumpeter made a distinction between invention and innovation but in his later work considered that both were becoming more 'mechanistic'. He was deeply interested in the dynamics and mechanisms of this change. Long-term economic growth was argued to be based on radical

Scale and scope in research

147

technological advancement whereby successive industrial revolutions resulted in the qualitative transformation of the economy rather than on the simple quantitative growth of extant industries. 'Gales of creative destruction' described the onslaught of waves of radical technological advancement. Generated by entrepreneurial actors, technological innovations would lead to both winners and losers. As the new wave of innovation starts replacing an older one, firms come under competitive threats that eventually destroy established positions. The implication for government policy is that it must deal with a system in constant flux. Perfect competition seemed untenable and, probably, not desirable because although it could allocate resources efficiently at any point in time it would stifle the kind of activity that would allocate resources most efficiently over time.? A firm could restrict its output and capture supernormal profits, at least temporarily, without causing a loss to the economic system. Firms may do so as a strategic response to the market and technological uncertainties created by the anticipated waves of technological change. Schumpeter (19 12, 1942) was principally interested in the differences between the innovative activities of small, flexible, entrepreneurial firms and those of larger, diversified, professionally managed corporations with formal research and development (R&D) departments. While he never really formulated strong, testable hypotheses, followers of his work have extracted such hypotheses from fragments in his writings and put them to empirical test. The empirical literature has tried to test the so-called neoSchumpeterian hypotheses that focus on the effects of monopoly power and large size on innovative efforts. Two main hypotheses were formulated and empirically tested: Hypothesis 1: Monopoly Power and Innovative Activity are Positively Related

Considerations of both ex post market power (anticipated to follow innovation) and ex ante market power (existing before the innovation) underline this hypothesis. On the one hand, firms must anticipate some form of (at least transient) market power, and the consequent realization of extraordinary profits, in order to invest in R&D in the first place. On the other hand, an oligopolistic structure in the market where the investing company already operates may favour innovation because: ( I ) the firm may feel able to extend monopoly power in current product markets to future product markets; (2) supernormal profit enables the firm to respond faster to the innovative efforts of its rivals (which may also operate as a deterrent in the first place); (3) oligopolistic market structure facilitates

148

European science and technology policj

knowledge of rival behaviour; (4) the availability of internal financial resources greatly facilitates both undertaking risky research projects and hiring the best and most innovative people. Ex ante market power also has potential disadvantages for innovation: (1) the monopolist may be more concerned with protecting its current position than with acquiring a new one; (2) the monopolist may be slower in bringing innovative technology online if it replaces existing technologies; (3) the monopolist may regard additional leisure as superior to additional profits (X-inefficiency). The theoretical answer to the question of the optimal level of monopoly power is that the marginal sacrifice in static efficiency resulting from a departure from perfect competition must equal the marginal social benefit from increased innovative activity. Hypothesis 2: Firm Size and Innovative Activity are Positively Related

Large firms may have an advantage in innovation over small firms if: (1) innovative activity is becoming more expensive as Galbraith (1952) argued; (2) there are economies of scale and scope in research because researcher productivity increases with the number of colleagues they have to interact with, because researchers are specialized and must rely on the expertise of colleagues and because of the superior ability of large, diversified firms to exploit the output of their research internally (Nelson, 1959). Firm size may also have disadvantages for innovation: (I) as firms grow large, R&D efficiency is undermined due to coordination problems and loss of managerial control; (2) employee motivation ebbs as the direct link between scientists and entrepreneurs breaks and the ability to benefit directly from innovation diminishes. The neo-Schumpeterian hypotheses attracted significant attention from applied microeconomists, who have tested them empirically for several decades. It must be stressed that such tests have met with significant difficulties regarding the identification of the innovation, the definition of the inputs into the innovation process, the measurement of firm size, the measurement of monopoly power and the undetermined nature of the direction of causality among the variables being studied. The most persistent finding of the empirical literature on the first hypothesis is that the effect of concentration on research intensity depends upon other industry conditions such as technological opportunity, appropriability, technological uncertainty and market and finance conditions (Comanor, 1967; Scherer, 1967; Shrieves, 1978; Scott, 1984, Levin et al., 1985; Wedig, 1990). Schumpeter, however, also argued that the expectation of market power acquired by successful innovation provides an important incentive to

Scale and scope in reseurch

149

undertake inventive activity. Studies have been undertaken to examine whether the ability of firms to appropriate the returns from an innovation encourages R&D investment (Cohen et al., 1987; Levin et al., 1987). Moreover, the idea that market structure is a function of the technology life cycle (or product cycle) has been studied, focusing attention on the conditioning role of the product and technology life cycle (Freeman and Soete, 1997; Tassey, 1997, 2007). Early in the cycle, high technological opportunity stimulates entry, but as the product matures, entry falls off, concentration increases and innovation shifts to more incremental process innovation. All in all, the only simple conclusion to be drawn from this literature is that the links between market structure, innovation and economic welfare are extremely complex (Scherer, 1992). The relationship between R&D intensity/innovativeness and market structure depends on numerous other factors affecting the industry, the firm and the technology. The most important of these appear to be appropriability, technological opportunity and demand structure. There has been significant controversy over the relationship between firm size and innovation. Most tests have actually focused on the impact of the former on the latter. As with the hypothesis on innovation and market structure, however, there seems to be simultaneity: not only may firm size have an effect on the rate of innovation, the reverse must be happening at the same time too. While there is, by now, considerable statistical information on the topic, a good part of it relates to research expenditures or patents rather than innovation, which introduces problems of interpretation (Freeman and Soete, 1997). Several study limitations have been pointed out (Cohen, 1995): Inventive activity has been measured by inputs or outputs. There is no perfect correlation between the two, however, making the interpretation of results difficult (Fisher and Temin, 1973). Selection bias: typically only the most successful firms are represented. Wide variation between studies with respect to their control for various firm characteristics. Hypothesized factors have included cash flow, degree of diversification, complementary capabilities, economies of scale in research and the ability to spread research costs over output. Yet one cannot easily find direct examinations of whether the relationship between size and research is due to any of these factors. Industry effects, such as technological opportunity, have also been widely reputed to impact upon the relationship between firm size

150

European science and technolog), policjs

and innovativeness. Studies that have tried to control for industry effects show that R&D rises more than proportionately with size (for example, Comanor, 1967) or that inventive activity measured in terms of either inputs or outputs increased more than proportionally with size up to a threshold, whereupon the relationship became basically proportional (Scherer, 1965; Philips, 1971; Kamien and Schwartz, 1982). Still, such results have been disputed. 0 Data limitations make it difficult to control for industry effects. Most larger firms are aggregations of business units engaged in a variety of industries. Using two-digit SIC (Standard Industry Classification)level data introduces measurement error to the extent that relevant industry characteristics vary across the four-digit industries that make up the two-digit data. Schumpeter's hypotheses usually make more sense at the level of the business unit rather than at the firm level for large diverse firms (Scherer, 1984; Levin et a]., 1987). a If it exists, the relationship between size and innovation is probably non-monotonic. Threshold effects may be present at the business unit level (Cohen et al., 1987) or at the firm level (Bound et al., 1984; Pavitt et al., 1987). But the threshold effect has not remained unchallenged. Studies like Acs and Audretch (1990) have shown that smaller manufacturing firms accounted for a disproportionately large share of innovations relative to their size and that R&D productivity defined as innovations per unit of R&D tends to decline with firm size. Recent experience with innovation in heavily entrepreneur-based industries such as information and communications technology sectors (particularly service-oriented ones such as Internet and mobile communications). biotechnology and new materials has also tended to corroborate the concept that smaller, entrepreneurial firms and larger, diversified firms are relatively more innovative, depending on the sector. The frequent reversal of results and uncertainty has led to a search for other firm-level determinants of innovative effort. These may include good management, working relationships and intra-firm communication. Analysing firm-level capabilities, however, demands a high degree of understanding of the underlying technologies and other activities within a firm or industry. This has been pursued extensively with examples of studies such as automobile manufacturing, computers and semiconductors, photolithographic alignment equipment, industrial automation, chemicals, energy, biotechnology, software, the Internet and other service sectors (Barras, 1986; Clark and Fujimoto, 1991; Henderson, 1993; Carlsson, 1995; Mowery and Nelson, 1999; OECD, 2000; Steil et al., 2002).

Scale and scope in research

151

Another determinant may be cash flow, seen as a measure of the availability of internal funds for R&D (Grabowski, 1968; Antonelli, 1989). Schumpeter implied that capital markets are imperfect, so the firm would rely on internal funding for its innovative efforts. Many of these studies found that cash flow was positively associated with higher levels of R&D activity. A third alleged determinant is company diversification. The idea here is that a diversified firm is better able to exploit research results. In addition, large, diversified firms are able to exploit the complementarities in their non-research activities with their research activities as well. Scherer (1965) found that an index of diversification was highly significant. Later studies (Grabowski, 1968; Grabowski and Mueller, 1978; Scott and Pascoe, 1987) had ambiguous results. It would be fair to say that the empirical results regarding the relationship of firm size and innovativeness are inconclusive. Perhaps the conjecture of Jewkes et al. (1958) remains as true as ever: 'It may be that there is no optimal size of firm but merely an optimal pattern for any industry, such as distribution of firms by size, character and outlook to guarantee the most effective gathering together and commercially perfecting of the flow of new ideas' (p. 168). It can be concluded that while the empirical investigation of the neoSchumpeterian hypotheses provides a useful introduction to our subject matter, it has not managed to decisively answer the specific question of policy decision-makers and strategists on the existence of economies of scale, scope and complementarity in research and in innovation. If anything, this rather extensive literature points to a heavily qualified answer: the existence or not of such economies heavily depends on the nature of the technical endeavour at hand, the related technological opportunity and appropriability, the stage of the technology cycle and the demand characteristics. In other words, expect differences across sectors/ technology areas and through time.

3 COOPERATIVE RESEARCH An important limitation of the classic discussion on scale and scope effects in research for today's environment is the concentration on individual organizations. One of the most striking features of industrial innovation today is that only a small minority of firms can innovate alone. Most technological advances and significant innovations involve a multitude of organizations. This is especially the case for the most valuable, most knowledge-intensive and most complex technologies. The past three decades have witnessed an

152

European science and technology poliq'

explosion of cooperative innovation agreements involving firms, universities and other research institutes in various combinations. Adapting to an environment of high risks, global competition, increasing complexity of technological advances and rapid generation and diffusion of technical knowledge and know-how, a large number of firms have opted for cooperative relationships. In the presence of technological development that involves a greater array of product and process systems, subsystems and components, no single firm can deploy all of the requisite capabilities and assets at reasonable cost. In this context, a network can serve as a locus for innovation because, for any network member, it provides timely access to external knowledge and resources that are otherwise unavailable, while also testing internal expertise and learning abilities (Powell et al., 1996). Linkages within innovation networks are very complex, involving not only diverse kinds of formal contracts, but also informal exchanges of knowledge, thus increasing opportunities for knowledge transmission. Innovation networks involve not only a wide variety of organizations, but also a wide range of activities including joint ventures, research corporations, joint R&D (for example, research pacts, joint development agreements), technology exchange agreements (for example, technology sharing, cross-licensing, mutual second-sourcing), direct investment, minority/cross-holding, customer-supplier relations, R&D contracts, one-directional technology flow agreements (licensing, second-sourcing), manufacturing agreements, marketing agreements and service agreements. The terms 'strategic alliance' and 'strategic partnership' are often used to describe the building blocks of innovation networks. Innovation networks are increasingly regarded as the dominant organizational mode in the knowledge-based economy. This has created a proliferation of literature with several special issues and individual papers on alliances and networks appearing in the past several years in journals such as Organization Science, Organization Studies, International Studies of Management and Organizations, Strategic Management Journal, Research Policy and World Development. Yet besides case studies, we have few systematic indicators of the nature and dynamics of these innovation networks. Moreover, there remains a great deal of debate about the factors underpinning network formation and e v ~ l u t i o n . ~ The basic rationale for cooperative research in economics has rested on traditional market failure arguments emphasizing insufficient incentives for individual firms to undertake uncertain and imperfectly appropriable research at the socially optimal level. Business arguments have focused on the organization and its strategic considerations such as better access to resources, capabilities and markets and the creation of new

Scale and scope in research

153

strategic options (Hagedoorn et a]., 2000; Hemphill and Vonortas, 2003). Frequently mentioned advantages of research partnerships for private sector partners include:

R&D cost-sharing; reduction of R&D duplication, research synergies; risk-sharing, uncertainty reduction; knowledge spillover internalization; easier access to finance; access to complementary resources and skills; more effective deployment of extant resources and further development of resource base; strategic flexibility, market access and the creation of investment 'options'; promotion of technical standards; and market power, that is, co-opting competition. To the extent that the relevant unit of analysis has shifted from the individual organization to the consortium or network, the conceptualizations of scale and scope in the economic literature must be recast. The relevant research resources, capabilities and strategies are no longer those of the individual organization but those of the group. The research question then becomes whether the incremental benefits obtained by a larger and more inclusive network through the leveraging of larger pools of resources and capabilities overcome the incremental cost of increased coordination needs and ebbing motivation. At the project level, a large consortium or a large budget would, in principle, be associated with improved performance. In terms of both scientific and technological outputs, the efforts and skills of multiple partners in an R&D project would lead to a larger pool of resources and expertise and hence would, ceteris paribus, increase the likelihood for success (Schilling, 2005). Equally important, a large consortium, composed of carefully chosen participants, would increase the heterogeneity of resources pooled together for project use. Increased heterogeneity in skills and experiences among project participants may foster creative problem-solving, promote learning and new knowledge creation, and may thus increase the likelihood of project success.$ Unfortunately, large consortia also have a negative side: the administrative and coordination costs of running the project also increase with size. In addition, large numbers of participants may bring a greater likelihood of social loafing and free-riding, thereby decreasing the extent of learning (Wong, 2004) and hence the likelihood of project s u ~ c e s s . ~

154

Europeun science and technology p o l i q

The trade-off between these effects will have a direct bearing on the consortium's absorptive capacity, that is, the set of capabilities relevant to the acquisition, assimilation, transformation and exploitation of knowledge to finally produce marketable, innovative result^.^ For our purpose, an important consideration is that network structure is emergent in the initial conditions of a specific industry, including the inherent characteristics of the relevant technologies and the norms and institutional factors that help generate rules that guide the competitive1 cooperative behaviour of firms in that industry. There seems to be a fundamental trade-off in network structure between organizational stability and variety. The effort to maintain and strengthen the prevailing relationships tends to freeze the structure of interactions into stable patterns. The more stable the patterns of interaction become, however, the more the network acquires the characteristics of a firm's organization, that is, the more it strives for specialization and the less capable it grows to achieve its fundamental objective of providing variety. Network structure optimality will, at least in part, depend on whether the predominant mode of operation in an industry concentrates on the better exploitation of existing technologies, skills and information or on the exploration of emerging innovations and other changes. Both processes are often needed and pursued simultaneously and compete for limited resources within individual organizations. The optimal allocation of resources among the two will depend on the internal conditions of the firm (resources, capabilities, strategic inclination) and on environmental factors relating to industry characteristics, including both the demand side (conditions for market development) and the supply side (technological opportunity, appropriability, competitive conditions). Analysts have argued for high-density and strong ties for exploitation and for low-density and weak ties for exploration (Rowley et al., 2000). Others have reached the opposite conclusion by using bounded rationality arguments: in dynamic environments with frequently changing conditions and continuous learning by companies, the efficiency of transferring information through bridges in existing networks while avoiding duplication of contacts is argued to become less relevant than openness of contacts, network density and tie redundancy (Hagedoorn and Duysters, 2002). Still others argue somewhere in-between (Gilsing and Nooteboom, 2005): dense networks, redundant ties and variable tie strengths are expected in the case of exploration; less dense, more stable network structures and non-redundant ties are anticipated for exploitation. Moreover, hybrid network forms are anticipated in transitions from exploration to exploitation in the development of a dominant design. In sum, the question of scale and scope in collaborative industrial

Scule and scope in research

155

structures is far from settled. At the level of individual collaborative research projects, it is somewhat easier to tell but the answer will be qualified on the basis of the technology and its projected use. Generalizations are hard to make at this point.

4

SCALE AND SCOPE FOR GOVERNMENT

INTERVENTION IN RESEARCH Public support for science, technology and innovation has been justified by economists on the basis of market failures and system failures. Market failure requires that the social returns to research investment exceed the private returns to the individual organization undertaking the investment. Principal among the culprits are said to be imperfect appropriability and uncertainty. Underinvestment by the private sector is to be expected as a result, raising the need for public investment to achieve a socially optimal level of research effort. More recently, analysts have also vigorously promoted the idea of system failures due to the complexity of scientific and technological advancement and innovation. For instance, one can argue for the existence of 'lock-in' in some technological trajectory, even though an alternative path of technological development might be more efficient. Public intervention may be necessary to make the transition. Or, one can argue for government intervention on the basis of institutional constraints on the utilization and diffusion of knowledge. In addition to supporting the generation of new knowledge, significalit weight must be given to more effective institutional arrangements for its transfer. As a result of the systemic nature of innovation, there are many feedback loops between the various stages of the innovation process. Government intervention may be justified to avoid coordination and institutional failures. The government also has an important role in providing the necessary investments in human capital and in mechanisms to intensify the flows and absorption of knowledge. The issue of investment timing, in particular, has gained attention in the presence of shrinking technology life cycles. Technologies appear, mature and become obsolete in a series of evolutionary phases, which greatly affect R&D decisions. An important economic factor is, therefore, the timing of R&D investments relative to the evolution of a technology. The timing issue has two dimensions: investment decisions directed at attaining market share within a technology's life cycle and those focused on making the transition between life cycles. The transition between two generic technology life cycles presents significant complications. The greater

156

European science and techno log^ polic:,,

the differences between two generations of a technology, the greater the investment risk for individual companies. Transitions to new technology life cycles typically demand different sets of research skills than those of incumbent firms. Incumbents will tend, then, to assign higher technical and market risk valuations to the prospective research programme. with the result that necessary investments are postponed.' This creates a gap. the argument goes, which requires government intervention. Policy thinking regarding research in the European Union has arguably tended to progress from the traditional understanding of market failure to concepts of systemic failure and related scale and scope effects. This may be manifested by the increasing attention not only to collaborative R&D but to the broader concept of the ERA and its vision for increasing coordination among national and regional research policies as well as between research and regional development policies. It also seemingly underlies the philosophy of the Framework Programmes for Research. The justification of the first three or four Framework Programmes primarily reflected market failure arguments based on the need to support pre-competitive research of generic applicability across the Community. The main funding instrument was targeted cooperative research projects of limited time duration. The thinking underlying more recent Framework Programmes, however, has moved visibly towards the arguments of systemic failure. This trend started somewhat shyly with the 5th and moved forward visibly with the 6th and 7th Framework Programmes that have endorsed new funding instruments in research such as integrated projects, Networks of Excellence and Joint Technology Initiatives to advance whole technology platforms. The systemic approach to policy intervention includes notions of scale and scope but it is broader than just these. It calls for a tall research agenda of immediate applicability. Defining and empirically substantiating concepts of critical mass, especially as they relate to collaborative research efforts involving broad networks of participants and stakeholders, becomes of utmost priority.

NOTES We draw on the discussion on returns to scale, scope and learning regarding regular production techniques in Rosegger (1996). 2. This argument is, of course, predicated on the denial of a steady-state condition of an economy: in a stationary state, immediate optimization is synonymous with long-run optimization. 3. For reviews see Powell and Groda1(2005), Malerba and Vonortas (2009) and Chapter 9 of this book. 4. This relates to the notion of technological cognitive distance among project participants (Nooteboom et al.. 2007). I.

Scule and scope in research

157

5. There has been evidence, for example, that the most successful alliances are those

between firms with similar technological foci and/or operating in similar markets. whereas in contrast, distant firms find it difficult to cooperate effectively (Stuart, 1998). This is not a foreclosed subject, however. 6. Zahra and George (2002)building on the work of Cohen and Levinthal(1990). 7. This enhances the dominant firms' tendency to avoid jeopardizing profitable production lines in the prevailing technology life cycle.

REFERENCES A c s Z.J. and D.B. Audretch (l990), Innovation andSmall Firms, Cambridge, MA: The MIT Press. Antonelli, C. (1989). 'A failure-inducement model of research and development expenditure', Journal of'Economic Behavior and Organization, 12 (2), 159-80. Barras, R. (1986), 'Towards a theory of innovation in services', Research Policy, 15 (4), 161-73. Bound, J., C. Cummins, Z. Griliches, B.H. Hall and A. Jaffe (1984), 'Who does R&D and who patents'?', in: Z. Griliches (ed.). R&D, Patents, and Productivity. Chicago, IL: Chicago University Press for the NBER, pp. 21-54. Carlsson, B. (ed.) (1995), Technological Systems and Economic Performance: Thr Case of' Factory Automution. Boston, MA and Dordrecht: Kluwer Academic Publishers. Clark, K.B. and T. Fujimoto (1991), Product Development Perfirmance, Boston, MA: Harvard Business School Press. Cohen, W.M. (1995), 'Empirical studies of innovative activity', in: P. Stoneman (ed.), Handbook o f the Economics of' Innovation and Technological Change, Oxford: Basil Blackwell, pp. 182-264. Cohen, W.M. and D.A. Levinthal(1990), 'Absorptive capacity: a new perspective on learning and innovation', Administrutive Science Quarterly, 35 ( 1 ), 128-52. Cohen, W.M., R.C. Levin and D.C. Mowery (1987), 'Firm size and R&D intensity: a re-examination', Journal of Industrial Economics, 35 (4), 543-65. Comanor, W. S. (1967), 'Market structure, product differentiation, and industrial research', Quarterly Journal ofEconomics, 81 (4), 639-57. Fisher, F.M. and P. Temin (1973), 'Returns to scale in research and development: what does the Schumpeterian hypothesis imply?', Journal of Political Economy, 81 (1), 56-70. Freeman, C. and L. Soete (1997), The Economics of'Industriu1 Innovation, 3rd edn, Cambridge, MA: The MIT Press. Galbraith, J.K. (1952), American Capitalism, Boston, MA: Houghton Mifflin. Gilsing, V.A. and B. Nooteboom (2005), 'Density and strength of ties in innovation networks: an analysis of multimedia and biotechnology', European Management Revicw, 2 (2), 1-19. Grabowski, H.G. (1968), 'The determinants of industrial research and development: a study of the chemical, drug, and petroleum industries', Journal of Political Economy, 76 (2), 292-306. Grabowski, H.G. and D.C. Mueller (1978), 'Industrial research and development. intangible capital stocks, and firm profit rates', Be// Journal of Economics, 9 (2), 32843.

158

European science and technologj- polic:,,

Hagedoorn, J. and G. Duysters (2002), 'Learning in dynamic inter-firm networks: the efficacy of multiple contacts', Organization Studies. 23 (4), 525-48. Hagedoorn, J., A.N. Link and N.S. Vonortas. (2000), 'Research partnerships', Research Policy, 29 (4-5), 567-86. Hemphill, T.A. and N.S. Vonortas (2003). 'Strategic research partnerships: a managerial perspective', Technology Assessment and Strategic Managernenr, 15 (2). . ,, 255-71. Henderson, R.M. (1993), 'Underinvestment and incompetence as responses to radical innovation: evidence from the photolithographic alignment equipment industry', Rand Journal of Economics, 24 (2), 248-70. Jewkes, J., D. Sawers and R. Stillerman (1958), The Sources of Invention, London: Macmillan. Kamien, M.I. and N.L. Schwartz (1982). Market Structure and Innovation, New York and Cambridge: Cambridge University Press. Levin, R.C., W.M. Cohen and D.C. Mowery (1985), 'R&D appropriability, opportunity, and market structure: new evidence on some Schumpeterian hypotheses', American Economic Review Proceedings, 75 (2). 20-24. Levin, R.C., A.K. Klevorick, R.R. Nelson and S.G. Winter (1987), 'Appropriating the returns from industrial research and development', Brookings Papers on Economic Activity: Microeconomics, 18 (3), 783-820. Malerba, F., and N.S. Vonortas (eds) (2009), Innovation Networks in Industries, Northampton, MA, US and Cheltenham, UK : Edward Elgar. Mowery, D.C. and R.R. Nelson (eds.) (1999), Sources of Industrial Leadership: Studies of Seven Industries, Cambridge: Cambridge University Press. Nelson, R.R. (1959), 'The simple economics of basic scientific research', Journal of Political Economy, 67 (3), 297-306. Nooteboom, B., W. Vanhaverbeke, G. Duysters, V. Gilsing and A. Van Den Oord (2007) 'Optimal cognitive distance and absorptive capacity', Research Policy, 36 (7), 101C34. Organisation for Economic Co-operation and Development (OECD) (2000), The Service Economy, Business and Industry Policy forum series, Paris: OECD. Pavitt, K., M. Robson and J. Townsend (1987), 'The size distribution of innovating firms in the UK: l945-1983', Journal of Industrial Economics, 35 (3), 297-3 16. Phillips, A. (1971), Technology and Market Structure: A Study of the Aircraft Industry, Lexington, MA: Heath, Lexington Books. Powell, W.W. and S. Grodal (2005), 'Networks of innovators'. in: Fagerberg J., D.C. Mowery and R.R. Nelson (eds), The Osford Handbook of Innovation, Oxford: Oxford University Press, pp. 56-85. Powell, W.W., K.W. Koput and L. Smith-Doerr (l996), 'Inter-organizational collaboration and the locus of innovation: networks of learning in biotechnology'. Administrative Science Quarterly, 41 (I), 11W 5 . Rosegger, G. (1996), The Economics of Production and Innovution, 3rd ed. London: Butterworth-Heinemann. Rowley, T., D. Behrens and D. Krackhardt (2000), 'Redundant governance structures: an analysis of structural and relational embeddedness in the steel and semiconductor industries', Strategic Management Journol, 21 (3), 369-86. Scherer, F.M. (1965), 'Firm size, market structure, opportunity, and the output of patented inventions', American Economic Review, 55 ( 5 ) . 1097-1 25. Scherer, F.M. (1967), 'Market structure and the employment of scientists and engineers', American Economic Review, 57 (3). 524-3 1.

Scale and scope in research

159

Scherer, F.M. (1984) Innovation and Growth: Schumpeteriun Perspectives, Cambridge, MA: The MIT Press. Scherer, F.M. (1992), 'Schumpeter and plausible capitalism', Journal of'Economic Literature, 30 (3) 1416-33. Schilling, M.A. (2005), Strategic Manugement of' Technological Innowtion, New York: McGraw Hill International Edition. Schumpeter, J.A. (1912), Theorie &r Wirtschuftlichen Entuvic,kl~mg, Leipzig: Duncker & Humboldt. Schumpeter, J.A. (1942), Capitalism, Socialism and Democracy, New York: Harper and Row. Scott, J.T. (1984), 'Firm versus industry variability in R&D intensity', in: Z. Griliches (ed.), R&D, Patents, and Productivity, Chicago, 1L: Chicago University Press for the NBER, pp. 233-45. Scott, J.T. and G. Pascoe (1987), 'Purposive diversification of R&D in manufacturing', Journal of'lndustriul Economics, 36 (2), 193-206. Shrieves, R.E. (1978) 'Market structure and innovation: a new perspective', Journal of Industrial Economics, 26 (4), 32947. Steil, B., D.G. Victor and R.R. Nelson (eds) (2002). Technological Innovation und Economic Performance, Princeton, NJ: Princeton University Press. Stuart, T. (1998), 'Network positions and propensities to collaborate: an investigation of strategic alliance formation in a high-technology industry', Administrative Science Quurterly, 43 (3), 637-68. Tassey, G. (1997), The Economics of R&D Policy, Westport, CT: Quorum Books. Tassey, G . (2007), The Technology Imperative, Northampton. M A , USA and Cheltenham, UK: Edward Elgar. Wedig, G.J. (1990), 'How risky is R and D? A financial approach', The Review of Economics and Statistics, 72 (2), 296-303. Wong, S.S. (2004), 'Distal and local group learning: performance trade-offs and tensions', Organization Science, 15 (6), 645-56. Zahra, S.A. and G. George (2002), 'Absorptive capacity: a review, reconceptualization and extension', Academy oj'Munugement Revie\%,,27 (2), 185-203.

ERA and the role of networks Stefano Breschi and Franco Malerba Over the last couple of decades, the promotion of consortia between firms, universities, research centres and public agencies has gained centre stage in science and technology policy in Europe. Cooperative programmes in the form of shared-cost research and development (R&D) consortia have become the most important source of European Union (EU) funding and institutional support to innovation, international competitiveness and. by way of knowledge exchange and diffusion, intra-European cohesion. Cooperative policies have certainly been pervasive and effective in aggregating public and private institutions from national research communities. However, concerns have been expressed about their effectiveness in raising the level of innovative investments, supporting European competitiveness and providing an efficient mechanism for creating a critical mass of knowledge and competencies whose benefits may extend to laggards. Following recent political debate and the challenges posed by the enlargement of the Union, the implementation of cooperative policies by way of widespread support to a large variety of projects and institutions has undergone significant changes. The European Commission has called for a change in approach that responds to the need for reinvigorating the European research infrastructure and reflects the most recent theoretical and empirical debate about R&D networks. Starting from the Sixth Framework Programme (2002-06), policy actions have been more focused on identifying crucial 'nodes' and networking 'centres of excellence' that would represent the backbone of a truly European Research Area (ERA) and act as catalysts for smaller components or backward areas. The purpose of this chapter is to provide a contribution to the debate about the targets and effectiveness of network policies at EU level. It presents a thorough analysis of the large R&D network that has emerged across Framework Programmes (FPs), building upon recent evaluation studies carried out in this field. After a general overview of the aims and articulation of EU FPs in Section 1, the chapter is organized into three main sections. Section 2 contains a review of the empirical literature that has explored the structural properties and evolution of the research network that stemmed from Research Joint Ventures (RJVs) promoted

ERA and the role qf'networks

161

under the FPs. In Sections 3 and 4, attention is focused on information and communication technology (ICT). In section 3, the EU R&D network shaped by the EU FPs is placed within the context of the global research network. Specifically, it will be discussed to what extent EU FPs have been able to create and attract the major actors in the global network of research collaborations (the Global Knowledge Hubs) and how different funding instruments have been effective in this respect. In addition, we will also investigate to what extent EU R&D policies have been effective in involving the most innovative small and medium enterprises (SMEs) and what might have been the most important barriers to the participation of such companies. Finally, Section 4 focuses on the role of research network infrastructure in fostering the dissemination of innovation-related knowledge at the regional level. In particular, the links and complementarities between research and diffusion networks developed through EU and nationallregional funding will be examined. Section 5 concludes.

1 THE FRAMEWORK PROGRAMME AS A NETWORKING ENVIRONMENT The F P is the main policy instrument of the EU in the area of research and technological development (RTD). The FPs were introduced in the early 1980s to support European competitiveness in key technological areas.' They are agreed upon by the EU Member States and the European Parliament and are implemented over the span of several y e a n 2 They function as an umbrella of specific programmes funding research across a variety of fields such as ICTs, energy, biotechnology, health, advanced materials and manufacturing. Supported RTD projects must have European added-value and are almost exclusively undertaken by consortia of at least three partners representing a minimum of three EU Member States or other affiliated countries. Since their beginning, the FPs have provided a systematic procedure for discussing and agreeing upon priorities, guidelines and budget allocation. The FPs have become the main Community instrument for offering selective support to European companies seeking to undertake collaborative R&D with firms or research institutes in other European countries. In fact, RTD policy has been implemented mostly by supporting shared-cost contractual research, that is, multinational consortia (or RJVs) grouping firms, public agencies, research centres and universities focused in principle on pre-competitive research project^.^ The FPs have undergone significant changes during the past decade and a half, reflecting developments in the socioeconomic context of the

162

European science and technology policj

region and the Community's realization of the programmes' importance. The collaborative approach has been extended to structure new areas of intervention as the budget of the FPs increased and priorities changed. While competitiveness has remained the primary goal of European technology policy, a wider set of objectives has been pursued over time as the Community approach to RTD gained new momentum with the Maastricht Treaty. Since the Third Framework Programme (1990-94), Union RTD policy has been recognized as important - though complementary to other EU policies - for reducing unemployment, accelerating structural change and, at the same time, ensuring greater cohesion. In this perspective, the latest programmes have shifted the emphasis from supply-side factors, central in the design of the first policies, to diffusion-oriented projects and the increase of learning skills and knowledge among Europeans. Indeed, the recent debate about the future of European technology programmes has been characterized by the concern that Europe might not successfully achieve the transition to a knowledge-based economy. The 2000 Communication of the European Commission, Towards a European Research Area, started a broad discussion on the development of research cooperation in Europe and centred around two main concerns: the persistent lower level of innovative investments in Europe compared with the US and Japan and the fragmentation of research efforts, because of which EU investments often add up without much coherence, creating an ineffective dispersed static configuration. The fragmentation, isolation and compartmentalization of national research systems and the disparity of regulatory and administrative frameworks further worsen the effect of the low investment in RTD, limiting the European capacity to produce knowledge and the ability to innovate. A more concerted effort for the development of a real ERA has therefore been called by the Commission an urgent priority on the EU agenda. According to the Commission, for cooperative efforts to produce a longlasting effect, they should primarily aim at changing the organization of research in Europe rather than simply adding up resources and facilities. In this perspective, the focus of European programmes was to change from direct support to a large variety of projects and organizations (that often overlap national incentive schemes without forming a coherent whole) to a more limited number of priorities and measures that exert a 'coordinating, structuring and integrating' effect on European research. Investing in infrastructures, strengthening relations between existing organizat ~ o n sand programmes, improving conditions for political consultation, establishing a common system of scientific and technical reference and promoting greater mobility of researchers constitute the priorities of the newly designed European technology policy. Interventions in these areas

ERA and the role ofnetworks

163

are meant to create favourable conditions for greater public and private R T D investments and for developing a 'critical mass' in major research fields. In other words, action at European level is justified and selected in view of creating a more supportive and coherent European framework for research. In this perspective, policy intervention is directed primarily at stimulating resources and expertise to converge on strategic research areas but self-organization of 'excellence clusters' is expected to follow and to be supported. Within priority areas, clearly defined scientific and technological objectives are to be pursued by mobilizing a critical mass of activities and resources and allowing greater flexibility than in previous programmes for the allocation of resources and management of specific research activities. Aiming at facilitating ERA, the Sixth Framework Programme (2002-06) has therefore had an even stronger focus on research integration than any of its predecessors. The programme has introduced new funding instruments, which combined with more traditional instruments to provide a multitude of opportunities for collaboration. Among the new instruments, Integrated Projects (Ips) and Networks of Excellence (NOES)have played a particularly prominent role, accounting for, respectively, around 61 per cent and 16 per cent of the total cost of projects approved under FP6? 1. IPS are large multi-partner projects involving a wide range of organizations from the research and business communities. They aim at mobilizing a critical mass of resources to attain clearly defined objectives in terms of scientific and technological knowledge and/or results applicable to products, processes and services. The ultimate goal of this instrument is to obtain results with significant and direct impact on European industrial competitiveness or contributing to solving important societal/global problems. 2. NoEs are large multi-partner projects aimed at reinforcing European scientific and technological excellence by pooling and networking a critical mass of resources and expertise. They are primarily intended to combine and cross-fertilize existing strands of research around a common core issue and are also more likely to involve publicly supported research organizations and to have less centralized or hierarchical structures than Ips. Moreover, they aim at creating a progressive and lasting integration of existing and emerging research activities of the network partners. The rationale behind the new policy instruments is that world-class centres of excellence already exist in Europe in a wide range of research

164

European science and technologj policj

fields. However, they are often scattered and only loosely connected while their expertise is not always sufficiently well known across Europe, especially by firms that could usefully join forces with them. The integration of these centres into long-term R&D consortia financially supported by the E U and focused on leading-edge research would contribute to enhancing the European position in strategic fields, attracting new resources and expertise, and, most of all, restructuring the way research is carried out in Europe, favouring the development of an overall more collaborative attitude by public and private actors. At the same time, while the creation of these networks is to be supported with European financing, their activities should not become dependent on this support. European funding is in fact meant to complement resources deployed by the participants and should take the form of a fixed grant for integration (for NOES)and of a percentage grant to the budget (for IPS). Moreover, compared with previous programmes, consortia should enjoy greater freedom in managing their projects and the follow-up by the Commission services should move from detailed monitoring of inputs to a more strategic monitoring of outputs. In other words, European action is meant to be a stimulus for centres of excellence to 'cluster around' common long-term objectives, network on a permanent basis and self-organize the division of task and information flows. 1.1

Policy Evaluation and Networks

Public support for RTD has traditionally been justified in the economics literature on the basis of market failures. The market failure rationale is based on the difference between the benefits to society (social returns) and the benefits to the individuals/organizations undertaking the RTD investment (private returns). The larger this difference, the larger the spillovers from the private party to the rest of society and the lesser the willingness of the private party/sector to invest at the socially optimal level. Although the market failure argument has been the main motivating factor for the implementation of the first FPs, the recent changes in EU RTD policy seem to reflect new theoretical conceptualizations of the innovation process. The latter is conceptualized as a complex nonlinear 'system', which involves interactive learning processes and feedback among a multitude of different organizations. Beyond accounting for inputs and outputs/outcomes, the systemic approach concentrates on the dynamics of RTD and innovation, that is, the processes involved in generating innovation outcomes, providing additional rationales for publicly supporting RTD activities. In particular, the new rationale for public intervention has to be found in 'system' failures stemming from issues such

ERA and the role ofnetworks

165

as path dependence, technological lock-in, investment timing (technology life cycles, trajectories), institutional constraints (general infrastructure), coordination failures (for example, standards), complexity of knowledge bases and the ineffectiveness of mechanisms facilitating knowledge flows (Edquist, 2004). Rather than just providing additional resources, government intervention should therefore focus on fostering collaborative learning, strengthening the linkages among the different types of organizations involved in the innovative process and facilitating a wide diffusion of knowledge. However, policy should take account of the major differences that exist across national, sectoral and local systems (Lundvall, 1992; Malerba, 2002). The consideration of such differences has underlined the concept of 'additionality', which is a useful concept when organizing public support for R T D (OECD, 2006): 'Input additionality': has public expenditure created additional funds to be spent and on what are they spent? 2. 'Output/outcome additionality': has public expenditure generated additional private and social returns? 3. 'Behavioural additionality': has public expenditure created sustainable effects beyond the infusion of resources and outputs such as improving the knowledge, capabilities, organization and strategies of firms? 1.

It is the third aspect of additionality where the network approach can make its greatest contribution. By studying relationships, exchanges, network location and status, network structure and evolution through time, and participant characteristics and roles in the network, this approach provides a new prism to examine important aspects of the longer-lasting, more sustainable contributions of public policy in impacting upon organizational and nationallregional capabilities to innovate. What do we know about these aspects for EU FPs? The following sections will examine these issues in general and for ICTs in particular.

2 STRUCTURE AND TOPOLOGY OF EU FRAMEWORK PROGRAMMES NETWORKS Several studies have been produced over the last years with the explicit aim to map and assess the topological features of the R T D network that stemmed from RJVs promoted under FPs (Jozef Stefan Institute, 1999; Breschi and Cusmano, 2004; Wagner et al., 2005; Roediger-Schluga and Barber, 2006). The main findings of these studies can be briefly summarized as follows:'

166

European science and technologj policj

I.

The R&D collaboration network arising from projects funded by EU FPs exhibits the structural properties of 'small world' graphs, that is, the presence of a giant connected component, a short average distance among participating organizations and a high degree of local clustering; 2. The degree of distribution of participating organizations in terms of both number of partners and number of projects follows a 'power law' distribution, that is, the vast majority of organizations have a very small number of direct links with other organizations (or participating in only one or a very small number of projects) but there is a fat tail containing very few organizations with a very large number of connections to other organizations (or which participate in a very large number of projects), which suggests a 'scale-free' architecture of the graph. 3. Overall, the topology of the network seems to be characterized by a three-layer structure. A 'stable core' of organizations mostly universities, public research organizations and large firms - that take part in a great number of projects under consecutive FPs, collaborate in a recurrent way with the same partners and frequently play the role of project coordinators. Around this persistent core of highly frequent participants, which exhibits a high degree of intra-connectedness, there is first a smaller group of rather frequent but low-profile participants - which enter consortia as partners and take advantage of the programmes for linking with several leading actors - and finally an extremely large number of partners for which participation seems to represent an exceptional event.6 4. The introduction of IPS and NOES under FP6 has led to closer, stronger and denser linkages among organizations, with large firms and research institutes taking on even more central network positions than in earlier FPs. -

What are the implications of such findings for assessing the effectiveness of EU FPs and current trends in EU RTD policies? One has to take account of both positive and negative aspects. On the positive side. the EU FPs have certainly contributed to the creation of a small-world topology in the EU RTD network, which is believed to be effective for both the creation (high cliquishness) and the dissemination (low distance) of knowledge, especially when complex and difficult-to-absorb knowledge is at stake (Watts, 1999; Verspagen, 2006). At the same time, however, the emergence of this structure can be interpreted as the (unintended) consequence of the rules governing the participation in EU RTD projects, which have favoured the formation and consolidation of a relatively restricted core of

ERA and the role oj'netvvorks

167

highly connected organizations or network hubs. In particular, high transaction costs in FPs combined with the typical relative scarcity of resources in smaller firms have been major factors limiting such players from taking a coordinating role in R T D projects. The most feasible way for getting access to FP funding for these companies is often through joining projects led by larger and more reputed organizations. The goal of achieving longrun cohesion and diffusion of knowledge in this field thus seems to have been achieved indirectly by focusing on funding a restricted set of network participants and relying upon their coordination capabilities to attract more and more peripheral organizations. The network that has emerged from EU FP projects is one where a very large number of organizations, frequently of smaller size, float around and are highly dependent upon a small group of core and highly interconnected organizations (hubs). If anything, the introduction of new funding instruments seems to have reinforced the existing structure and patterns of interactions. In this respect, a few potential risks should be noted. First, as long as participation in and funding from EU R T D projects remain conditioned by access to a few anchor companies and institutions, it is unlikely that organizations joining the network late will ever achieve a hub role in the network. Moreover, to the extent that research priorities and network organization are defined by core participants. the risk of lock-in and the resistance to the reorientation of the network towards more productive research areas increase accordingly. The policy implication would be that, besides funding instruments aimed at further increasing linkages among hub organizations (for example, IPS)and among hub and non-hub organizations (for example, NOES),emphasis should be placed on more flexible and manageable instruments that allow smaller organizations to take a leading and coordinating role in RTD projects. If one of the objectives of EU R T D policies is to nurture the development of new European hubs, policies and instruments better tailored to the needs and constraints of non-hub organizations should be promoted.

3 EUROPEAN AND GLOBAL RESEARCH NETWORKS IN ICT Given the network structure outlined above, the 'core' of the network carries the greatest interest for both researchers and policy-makers as the effective amount and quality of knowledge production and transmission within the overall network clearly depend on the resources deployed by the members of this core through their expertise and degree of integration. In this respect, a few questions seem to deserve particular attention. To what

168

Europrun science and t e c l ~ n o l o gpoliq, ~

extent do hub organizations, that is, organizations highly connected in the EU R T D network, play a comparable role within the broader global network of RTD collaborations? To what extent are EU RTD networks and network hubs able to attract the most technologically dynamic small and medium-sized enterprises around Europe? What may be the role of new funding instruments under FP6 and FP7 in nurturing new network hubs and facilitating the effectiveness of existing ones? Providing a satisfactory answer to these questions is not easy as it requires developing and analysing finer-grained data than the ones used to map the structure of the EU RTD network. In the case of ICT, some answers can be derived from a recent study conducted on the Information Society Technology, Research, Technological Development and Demonstration (IST-RTD) Programmes of the 6th Research Framework Programme (Malerba et al., 2006). The study compares in particular the European network formed by organizations participating in FP6 IST projects (partnership network), the global network formed by companies involved in privately funded RTD alliances (global network) and the knowledge network arising from cross-organizational patent citations (knowledge network). Results show that although few European organizations are able to occupy core positions in the global ICT alliances network, EU IST projects are able to attract the vast majority of the most important players in the global research network. More importantly, some of the global players attracted to the IST Programmes serve as gatekeepers, that is, they have a dual role as global hubs and as hubs in the IST-RTD network. These organizations link organizations involved in IST-RTD with the broader global network of R T D collaboration. Their position in both networks puts them at the crossroads of information and knowledge flowing within IST-RTD projects and information and knowledge flowing within the much broader global network of strategic alliances. Moreover, analysis of the patent citations network also shows that IST-RTD hubs and especially gatekeepers, that is, hub organizations in both the IST-RTD and global research networks, are relatively more effective in terms of both producing and diffusing information compared with any other type of organization. As far as funding instruments are concerned, findings show that IPS are highly effective means for the purpose of connecting core organizations (that is, network hubs) in IST-RTD projects and global research hubs with each other and through them to many other IST organizations. The apparent effectiveness of IPS for putting together heterogeneous actors with different and complementary competences can be considered as a strength of this instrument in terms of promoting the ERA objectives. IPS seem to create the scale and ambition necessary to develop technology platforms,

169

ERA and the role of networks

thus propelling some European hubs to positions of global stature. Because they are large and ambitious, IPS also tend to attract global hubs, which provide connectivity to world leaders. Interviews with European experts in the IST field confirmed the fact that IPS are characterized by the participation of larger organizations that are more diversified and have more diffuse research capabilities and a broader market reach. The role of the Prime Contractor was reportedly critical in IPS. Experience with coordinating large projects becomes paramount for success and, it was argued, should be one of the criteria for selecting IP projects. This, in turn, gives these organizations significant bargaining power. On the other hand, NOESseem to be relatively less effective in this specific role; they are more relevant for linking IST-RTD hubs to other more peripheral organizations. They are dominated by Higher Education and Research Organizations and tend to be very large and diffuse. Industry has been hesitant to participate because NOESare perceived to have difficulty with research quality control, do not necessarily involve all excellent partners and are often too big (creating problems for coordination and knowledge diffusion and sharing). Despite the positive assessment of policy changes in the implementation of EU FPs, particularly with respect to IPS, a few concerns have been expressed with regard to their ability to include and integrate in the network the most dynamic SMEs. In this respect, the above-mentioned study shows not only that there is no SME that may be classified as a hub in the IST-RTD network but also that SMEs tend to be on the edges rather than at the centre of the network and are relatively underrepresented in NOES.Even more importantly, a very small fraction of the most dynamic SMEs - either companies holding highly cited patents or fast-growing companies in the relevant technological domains - has been participating in IST-RTD projects.' A relevant policy question arising from these results relates to the possible barriers that limit the participation of many innovative SMEs in IST-RTD projects. Besides the problems related to the often cumbersome and lengthy procedures associated with drafting, submitting and getting approval of project proposals and with the problematic issues arising from the definition of intellectual property rights, especially in relation to large and powerful organizations, the requirement of financial viability has often been claimed to be a major barrier inhibiting the participation of young innovative start-ups lacking a sufficiently long financial track record. Adopting flexible and relatively fast project evaluation procedures - especially in the case of projects coordinated by SMEs and revising the rules defining the financial viability of contractors shifting the emphasis from tangible assets to intellectual and knowledge-related assets - may consequently be seen as policy instruments for increasing the rate of participation of innovative SMEs in IST-RTD projects. With regard to -

-

170

European science and t r c h n o l o g ~policj

incentives to participate in IST-RTD projects, a notable finding is the low rate of SMEs in NOESas compared with IPS. This may be related to the long-term research objectives that seem to characterize NOESand the relative lack of focus on specific products and processes. For most innovative SMEs, growth perspectives are crucially linked to a short time-to-market introduction of new products and services. These characteristics, together with potential problems in the definition of intellectual property rights vis-a-vis larger organizations, may act to deter effective participation of smaller innovative companies in European research projects.

4

EU RESEARCH NETWORKS AND REGIONAL KNOWLEDGE DIFFUSION IN ICT

A key issue much less analysed is how these research networks impact upon regional systems and how they interact with research and diffusion activities carried out at regional level. Are there major complementarities between research networks built through the FPs and diffusion networks built through both dedicated EU-funded programmes and other national and regional programmes that focus more on technology exploitation and development? A preliminary investigation of this issue has been carried out once again for the field of ICT in relation to the projects funded by DG Information Society Technologies (IST) under FP6 (Cassi et al., 2008). The study analysed in particular the linkages between the 'research network' built through FP6 in the thematic area 'Applied IST Research Addressing Major Societal and Economic Challenges' on the one hand and the 'diffusion network' built through two EU programmes, 'eTen' and 'econtent', on the other hand. The former was designed to help the dissemination of telecommunication-networks-based services (e-services) with a trans-European dimension. It focused strongly on public services, particularly in areas where Europe has a competitive advantage. The latter was a market-oriented programme that aimed at supporting the production, use and distribution of European digital content and at promoting linguistic and cultural diversity in global networks. The empirical evidence suggests a few interesting patterns. In the first place, the degree of overlap between the two networks is relatively limited: a very small fraction of all organizations is involved in both types of projects and an even smaller fraction of all links among them depends on participation in both programmes. Yet if one considers the two networks simultaneously, almost all organizations involved in these projects are linked to each other, that is, they belong to a giant connected component and the average distance separating any pair of organizations is very short,

ERA and the vole qf net,r.orks

171

suggesting, in fact. a high degree of connectedness and relatively easy access to information circulating in the network. Second, the network structure outlined above suggests the existence of crucial organizations that act as gatekeepers linking the research with the diffusion network, thus speeding up the process of innovation and technology diffusion and allowing the application of the know-how developed within the research network to specific regional projects. An analysis of the types of organizations playing such a gatekeeping role reveals, quite interestingly, that it comprises both higher education and research organizations, large firms and SMEs. Higher education and research institutes develop and diffuse advanced and frontier knowledge to the network. On the other hand, companies provide links to the market and market information feedback so that research can be more focused on market relevance. Finally, a separate analysis of the research and diffusion networks reveals that both are characterized by a scale-free architecture, that is to say by the presence of a few highly connected nodes or hubs, which ensure the overall connectivity of the graph. A more indepth analysis of these hubs shows a few important differences between the two networks. Higher education and research institutions are more likely to play the role of hubs in the research network while the diffusion network is dominated by private companies and other organizations. Furthermore, the geographical reach of linkages created by hubs also strongly differs between networks. Diffusion hubs tend to act more locally than research hubs. In the research network, private companies are less geographically limited while the opposite is true for the diffusion network. In the diffusion network, the 'other organizations' are those with the most localized links. Their regional links are more than twice the average number. These actors indeed play a key role in diffusion at the regional level. Academic hubs, on the contrary, do not show any differences in the two networks as far as the localization of their links is concerned.

5

CONCLUSION

The present chapter has offered a contribution to the debate about the targets and effectiveness of network policies at E U level by presenting a broad analysis of the large R&D networks that have emerged from the FPs. Specific attention has been devoted to the recent policy changes that have been introduced in the implementation of FP6 and FP7 concerning especially the new funding instruments, that is, IPSand NOES.Such changes have implied a move from direct support for a large variety of projects and organizations to support for a more limited number of large projects and centres of excellence with the objective of mobilizing a critical mass of

172

European science and technology policy

activities and resources, enhancing the organization and quality of research and ultimately European competitiveness. Our assessment of current policy trends has emphasized both the positive aspects and the potential risks involved in them. The FPs have undoubtedly been successful in fostering the formation of new linkages among an expanding set of European public and private organizations thus contributing to the emergence of a cohesive ERA. The topological features of the network built through successive FPs are conducive, at least potentially, to both the creation and the dissemination of knowledge. At the same time, however, this objective seems to have been achieved at a cost. The FPs for RTD have in fact created networks characterized by a wide discrepancy in the criticality of different nodes (organizations) for the network. A few nodes comprising mostly higher education and research organizations and large firms have come to dominate the network, being placed in more critical positions in the network than the large majority of other nodes. If possible, it has been argued, the introduction of new funding instruments is further reinforcing the existing network structure and the position of individual actors in it. This architecture, however, is likely to involve potential benefits but also risks. Large centrally positioned organizations enjoy status and reputation, tend to have timely and easy access to resources and information and thereby attract outside partners, thus connecting the European network to the broader global network of research collaborations. At the same time, one should not neglect the fact that the structural inertia governing the evolution of European RTD networks, further strengthened by new funding instruments, tends to make it more and more difficult to change the existing structure of relationships making it more and more difficult for late entrants, especially small and medium-sized companies, to acquire a prominent role in the network and increasing the risk of lock-ins and resistance to the reorientation of the network towards more productive research areas. The most important policy implication deriving from the analysis carried out in this chapter is that more emphasis should be placed on more flexible and manageable instruments that allow smaller organizations to take a leading and coordinating role in RTD projects. If one of the objectives of EU RTD policies is to nurture the development of new European hubs, policies and instruments better tailored to the needs and constraints of small innovative organizations should be promoted.

NOTES I

When the first FP was launched, in 1984, the 'technology gap', which was perceived to be of the greatest relevance in explaining European declining competitiveness, was the main

ERA and the role of networks

173

concern driving policy action. Along the lines of the Single Market approach, the focus of European RTD policies was primarily directed towards overcoming the fragmented national structure of European industry and markets, permitting economies of scale that could not be achieved at the national level. Accordingly, the preference was for research conducted on a vast scale, projects addressing common interests that could be tackled best through a joint effort, and research contributing to the cohesion of the common market, promoting the setting of uniform laws and standards. The first six FPs were implemented over a five-year period. The Seventh FP will run instead for seven years (2007-13). Moreover, the budget allocated to the Seventh FP is around €51 billion. 41 per cent greater than the budget allocated under the previous FP. The pre-competitive requirement is meant to avoid potential conflict with E U competition policy, which forbids collaboration at the stage of developing products for an immediate market, whereas Articles 85 and 86 of the Rome Treaty allow collaboration for pre-competitive research. Other instruments are Specific Targeted Research Projects (STRePs). Coordinated Actions (CAs) and Specific Support Actions (SSAs). STRePs involve smaller consortia and more narrowly focused research that is innovative within a predetermined workplan. They are self-contained and closest to the typical collaborative research traditionally supported by the FPs. CAs and SSAs provide other forms of support or coordination to ongoing research etforts and areas of policy application in other instruments. It is worth remarking that, due to data availability, currently available studies have mostly focused upon the first five E U FPs. It should also be noted that universities and public research organizations tend to be overrepresented in the core compared with their participation rates In EU FPs. Quite interestingly, the same concerns have been expressed in the so-called Marimon Report, which was given the task of providing an evaluation of the effect~venessof the new funding instruments of FP6. The report produced discusses at length the barriers to participation encountered by innovative SMEs.

REFERENCES Breschi, S. a n d L. C u s m a n o (2004), 'Unveiling t h e texture o f a E u r o p e a n research area: emergence o f oligarchic n e t w o r k s u n d e r EU F r a m e w o r k Programmes', Intrrnational Journal of Technology Management, 27 (8), 747-72. Cassi, L., N . Corrocher. F. M a l e r b a a n d N. V o n o r t a s (2008), 'Research networks a s infrastructure f o r knowledge diffusion i n E u r o p e a n regions', Econon?ic.s of Innovation and New Technology, 17 (7), 651-63. Edquist, C. (2004), 'Systems o f innovation: perspectives a n d challenges', in: J . Fagerberg, D. M o w e r y a n d R . Nelson (eds), Oxfbrd Handbook of Innovation, O x f o r d : O x f o r d University Press. Jozef Stefan Institute (1999), Data Mining and Decision Support for Business Competitiveness: A European Virtual Enterprise, E u r o p e a n C o m m i s s i o n DG I n f o r m a t i o n Society a n d Media. Lundvall, B.-A. (ed.) (1992), Nationul Systems of Innovation: toward,^ [ I Thcory of' Innovution und Interactive Learning, L o n d o n : Pinter Publishers. M a r i m o n , R . (2004), Marimon Report, r e p o r t by a high level expert panel chaired by R a m o n M a r i m o n o f t h e N e w I n s t r u m e n t s o f F r a m e w o r k P r o g r a m m e VI, 21 June. M a l e r b a , F. (2002), 'Sectoral systems of innovation a n d production', Research Policy, 31 (2), 247-64. -

174

European science and technologj policy

Malerba, F., N.Vonortas, S. Breschi and L. Cassi (2006), Evaluation of Progress towards a European Research Area for Information Society Technologies, final report. Organisation for Economic Co-operation and Development (OECD) (2006). Government RTD Funding and Con~panyBekariour: Measuring Beharioural Additionality, Paris: OECD. Roediger-Schluga, T. and M.J. Barber (2006), 'The structure of R&D collaboration networks in the European Framework Programmes. UNU-MERIT working paper series 36, Maastricht. Verspagen, B. (2006) 'University research, intellectual property rights and European innovation systems', Journal of Economic Surveys. 20 (4), 607-32. Wagner, C., J. Cave, T. Tesch, V. Allee, R. Thomson, L. Leydesdorff and M. Botterman (2005), ERAnets: Evaluation of Networks of Collaboration Between Participants in IST Research and Their Evolution to Collaborations in the European Research Area ( E R A ) , final report. Watts, D.J. (1999), 'Networks, dynamics, and the small-world phenomenon'. American Journal of Sociology, 105 (2), 493-527.

10. Transnational collaboration in public research funding and publicly supported research in Europe Henri Delanghe, Brian Sloan and Ugur Muldur One of the approaches taken towards the achievement of the European Research Area (ERA) is the promotion of pan-European transnational coordination, collaboration and integration in terms of both research and development (R&D) funding and execution. Implicit in this approach is the understanding of ERA as a single innovation system. The innovation systems literature emphasizes system-wide science and technology (S&T) connectivity and collaboration within the scope of a single economy. It describes an innovation system as the set of separate but interconnected public and private institutions and organizations and human resources, which either individually or jointly and interactively create knowledge by financing and performing R&D, translate knowledge into innovations and affect the diffusion of those innovations. What matters is the number and quality of systemic links. The innovation system also provides the framework within which government takes action in support of S&T (Capron and Cincera, 1999; OECD, 2002). The understanding of ERA as a single innovation system and the perceived need for increased system-wide S&T connectivity and collaboration raises questions about the current extent of, rationale for and impact of pan-European S&T collaboration. This chapter looks at two kinds of such collaboration. The first relates to transnational collaboration in the public funding of research. Such collaboration is high on the European research policy agenda now as the 2007 ERA Green Paper stated that one of the features of ERA should be 'well-coordinated research programmes and priorities, including a significant volume of jointly-programmed public research investment at European level involving common priorities, coordination, implementation, and joint evaluation' (p. 5) but also noted that

176

European science and tecl~nologjpolic!,

'national and regional research funding (programmes, infrastructures, core funding of research institutions) remains largely uncoordinated' (p. 7). Yet, in spite of its political importance little has been written about the potential and actual benefits of cross-border public research funding cooperation. The second kind of pan-European S&T collaboration this chapter looks at is transnational collaboration in the execution of publicly supported research. Facilitating this kind of collaboration is often one of the main objectives of cross-border public research funding collaboration. Since the launching of the Community Framework Programmes (FPs) in the 1980s it is also one of the mainstays of European S&T collaboration, and it is comparatively well researched and described. Section 1 explains the concept of cross-border collaboration in the public funding of research, examines the current extent of such collaboration and discusses potential benefits. Section 2 provides some indications on the current extent of cross-border collaboration in the execution of publicly funded research and reviews the literature on the benefits and impacts of inter-institutional S&T collaboration, of government support for inter-institutional collaboration and of transnational collaboration via consortia in the execution of publicly supported research. Section 3 concludes among other things with a discussion of the concept of European added-value.

1

TRANSNATIONAL COLLABORATION IN THE PUBLIC FUNDING OF RESEARCH

Transnational public research funding collaboration can take many forms. It can involve pooling funds at EU level via the Community FPs. It can involve Member State governments and research ministries coming together and launching joint programmes through bilateral, trilateral, and so on, agreements, through large-scale intergovernmental schemes or through EU-supported joint programming initiatives such as Article 169 and ERA-NET. Or, still, it can involve collaboration between national research funding agencies (for example, national research councils) either with EU support (COST1,ERA-NET) or without (ESF2)(Table 10A.1). Each of these transnational public research funding schemes differs greatly in terms of source of funding, level of financial integration, level at which programming takes place and extent of joint programming, scope for transnationally collaborative project execution, and so on. The characteristics of some of these schemes are discussed in Chapters 1 and 2. The focus of this section is on transnational public research funding

Trunsnationul collaboration

177

collaboration between Member State governments, which has risen on the EU policy agenda following the 2007 ERA Green Paper, which called for more national and regional research funding coordination, and the 2008 Commission Communication Towards Joint Programming in Research Working Together to Tackle Common Challenges More Efectively, which set out an ambitious new approach for making better use of Europe's limited public R&D funds through enhanced cooperation. The extent of such collaboration currently taking place in Europe today is comparatively small. Barely 15 per cent of European publicly financed civil R&D (or 6 per cent of total R&D investment) is financed in partnership and coordinated at European level, intergovernmental schemes accounting for 10 per cent and the Community FPs for 5 per cent. In spite of well-known barriers to cross-border programme collaboration (Chapter 3), this comparative lack of such transnational public research funding collaboration is surprising given its potential benefits. For instance, the pooling of financial resources allows for overcoming more easily the high barriers-to-entry characterizing some S&T fields, in particular those with high fixed start-up and operating costs depending on 'big science' and large-scale research infrastructures. Such research infrastructures generate sizeable scientific, technological and other impacts (see Box 10.1'). Transnational public research funding collaboration also enables panEuropean public research programme optimization. It can ensure - within the context of scientific and technological uncertainty - that all relevant topics and approaches in a particular S&T field can be researched in parallel (optimizing programme scope) and that sufficient resources are allocated to the research on each topic (by eliminating wasteful crossEuropean programme duplication and increasing programme depth)." Transnational public research funding collaboration can promote scientific excellence. Joint calls with common funding increase the competition for funding and the quality of research proposals through a kind of 'European Research Council' effect. Purely national research programmes usually address domestically based researchers only. Where such programmes operate m the absence of high-quality domestic expertise, and research proposals and the research carried out are of low quality, transnational programmes allow for avoiding the waste of public resources. Transnational public research funding collaboration allows for addressing challenges jointly, developing common solutions and speaking with one voice. It enables several Member States in a particular region (for example, Baltic Sea area), or Europe as a whole, to tackle together commonly faced challenges (for example, climate change, ageing), develop common, standardized solutions (for example, in the area of transport), and speak with

178

European science and technology policj

BOX 10.1 THE IMPACTS OF POOLING RESOURCES TO FUND LARGE-SCALE RESEARCHINFRASTRUCTURES Large-scale research infrastructures have long institutional memories and serve as repositories of knowledge. They generate important structuring effects since they act as hubs that bring together users from different geographical and S&T areas, promote interdisciplinary research, and create pan-European transdisciplinary networks. The use of research infrastructures helps researchers obtain an insight into important new methods and techniques and advance their scientific understanding. Each visit to a facility results, on average, in at least one and probably two papers in peer-reviewed journals and conference communications. A not insignificant number of patents are filed based on work carried out at research infrastructures. And they play an important role in the development of technological clusters. Research infrastructures also have important impacts on industry. In the long run, industry is set to benefit significantly from the scientific, technological, structuring and training effects and impacts already mentioned. Yet the more direct impacts of research infrastructures on industry are also large. Industrial researchers and technicians can receive training in technological and knowledge-related aspects, but also in the complex organizational aspects of technological innovation. At publicly supported facilities, industrially developed materials and products are tested. As a result of the strict requirements of scientific projects carried out at research infrastructures, materials and products are often newly developed or substantially improved, often with commercial potential. Prototypes of important new products are sometimes developed, which later prove commercially viable and can even result in spin-offs. New research methods are developed, and new quality-control, planning and project-management methodologies are developed that can be used in industry. Research infrastructures, in their procurement, often make unusual technical demands on outside designers and constructors of equipment. These 'pushed' technical demands have a surprisingly high beneficial impact on the contracting firms, reflected in increased sales and

Transnational collaborution

179

cost savings, the sum of which can be defined as the economic utility of a contract. Research infrastructures also have large local economic impacts through supplier contracts and employment.

one voice in the international scene. The latter facilitates the interaction of non-European countries with Europe in the field of research. Transnational public research funding collaboration facilitates the coordination of different research programmes (for instance, between the FP, joint programming and intergovernmental schemes) and horizontal policy coordination. The high visibility of large-scale transnational programmes and their results as compared with purely national programmes facilitates the coordination between research policy on the one hand and education, innovation, ICT, energy, environmental, and so on policies on the other hand. Transnational public research funding collaboration can reduce programme management costs. It allows for the reduction of the national management cost of those national programmes that are integrated crossnationally (and thus for providing more support for research). It also enables cross-border policy-learning and improvements in the running of non-integrated national programmes. And it improves the accountability and transparency of public R&D support in the event ofjoint peer review, joint evaluation, and so on. For industry in particular, transnational public research funding collaboration can facilitate access to public research support. To access public research support, multinational enterprises (MNEs) carrying out research in different European countries have to familiarize themselves with a multitude of different national and regional research programmes, each with their own objectives, application, proposal selection and management procedures, and cycles. On the other hand, small and medium enterprises (SMEs) and locally embedded universities and research institutes often have difficulty accessing information on foreign research programmes. Transnational research programmes significantly reduce the number of public programmes operating in Europe, thereby making it easier for MNEs and SMEs to access public research support. They reduce the cost of information and the cost of applying for companies and have a positive effect on the crowding-in factor or leverage effect of public R&D support, resulting in higher private and overall R&D investment. Transnational public research funding collaboration facilitates the development of joint public-private strategic research agendas. Europe's

180

European science and technology polic),

industry is well-networked and able to speak to Member States and to the Community with one voice. On the other hand, however, the multitude of different regional and national research programmes in Europe prevents the public sector from speaking with one voice to industry. Transnational research programmes enable the public sector to speak with a single voice on particular research subjects and this facilitates the interaction and strategic research coordination with the private sector. Last but not least, and as explained in more detail in the next section. transnational public research funding collaboration promotes cross-border project collaboration, which can bring with it a multitude of benefits (and is discussed in the next section). However, with the exception of large-scale research infrastructures and cross-border project collaboration, there have been few studies providing detailed data on and analysis of the impacts of transnational public research funding collaboration.

2 TRANSNATIONAL COLLABORATION (CONSORTIA) IN THE EXECUTION OF PUBLICLY SUPPORTED RESEARCH Whereas the previous section focused on how countries can come together to cooperate in the public funding of research, this section looks at the impacts of transnational cooperation in the execution of such research. Transnational collaboration in the execution of publicly supported research usually involves research teams from different countries coordinating their efforts in an arealproject of common interest. This section will first examine the benefits of executing research via consortia, before going on to look at the effects of government sponsoring of such collaboration and then, more specifically, the impacts of sponsoring transnational consortia.

2.1 The Benefits of Inter-institutional S&T Collaboration Of particular interest within this context is the inter-institutional S&T collaboration literature, which so far has focused mainly on the firm as S&T collaboration initiator. Enterprises collaborate with other enterprises through, for instance, R&D alliances or R&D consortia (Faems, 2006); with universities through, for instance, thesis work, sponsored research, consortia (Finne, 2003); with public research institutes; and so on. Such science, technology and innovation collaboration by European firms is increasing rapidly (Table 10.1).

Trunsnutionul collaborution

181

Table 10.1 Share of' enterprises n~ithinnovation activities engaging in innovation collaboration

Country

1998-2000

2002-04

PT FI SE

23.7 38.6 17.4 19.9 10.2 33.3 24.9 7.9 35.9 23.7 18.5 15.8 51.6 34.0

37.8 42.3 19.2 21.4 18.6 38.8 33.6 11.0 36.7 44.4 18.8 17.7 46.4 47.4

UK

23.0

28.9

BE

DK DE GR ES FR 1E IT LU NL

AT

Source:

Eurostat, New Cronos database.

The supporting paradigm is that of 'open innovation'. Developing new technologies internally is time-consuming, expensive and uncertain (Faems et al., 2006). Useful knowledge can be found everywhere and firms should move from a closed innovation paradigm based on company control to one of open innovation under which they make optimal use of ideas generated elsewhere and make others use their internally generated ideas (Chesbrough, 2003). The literature has identified many possible benefits of interinstitutional S&T collaboration. It enables the combination of complementary resources (funding, skills, data - whether complementary, rare or needed for comparison). It encourages the exchange of tacit and codified knowledge and cross-fertilization. It allows the spread of R&D costs (and risks) among different partners. It allows the generation of economies of scale and scope (which decrease technological uncertainty). It allows the avoidance of wasteful duplication (in terms of research efforts and solutions). And it allows the reduction of the appropriability problem (which increases the incentive to pursue R&D) (Sakakibara, 2001; Santoro and Chakrabarti, 2002; Faems, 2006; Faems et al., 2006; Okamuro, 2007). Inter-institutional S&T collaboration has generally been found to be

182

European science and technologj. policj.

beneficial for participants. Empirical studies have found consistently that R&D alliances have a positive impact on partners' innovation performance (Faems, 2006). Motohashi (2004) found that university-industry collaboration has a positive effect on innovation productivity (patents) and productivity and that the younger the company the greater these effects. Chang (2001) found that inter-organizational cooperation is positively associated with the firm's innovation performance. The European Commission (2004) found for both Germany and Finland that a positive impact of collaboration on the propensity to patent in both countries is evident. Cross-border inter-institutional S&T collaboration boosts both the number and the quality of research outputs. The citation impact score of international co-publications is generally higher than that of purely national publications (Gomez et al., 1995; Glanzel et al., 1999; Glanzel, 2001; Adams et al., 2007). Cross-border research collaboration has a positive impact on the quality of patents. Guellec and Van Pottelsberghe (2002) found that research cooperation between inventors bears fruit, both domestically and internationally, and that international cooperation (that is, between inventors with residence in a different country) is even more fruitful than domestic cooperation. And Guellec and Zuniga (2008) found that, overall, cross-border R&D has a technological impact as big as R&D conducted in the home countries (by the same owner country), and often it is much higher. Yet the success of S&T collaboration is conditional. Falvey et al. (2006) argue that the performance of research joint ventures in terms of R&D investment, profit and welfare compared with R&D competition is sensitive to the level of coordination costs and that by ignoring the coordination costs of operating a research joint venture, the anticipated benefits or success of the cooperative project could have been grossly exaggerated. Okamuro (2007) found that whether or not cooperative R&D is successful depends on the structure and content of the cooperation. S&T collaboration is also risky. The failure rate of interfirm collaboration in the form of R&D alliances, for instance, has been estimated to range from 50 to 80 per cent (Faems, 2006). Reasons for the failure of interfirm collaboration in the form of R&D alliances include partner incompatibility, insufficient partner commitment, (fear of) opportunistic behaviour, a lack of trust, differences in national andlor corporate culture, excessive coordination costs, imbalances in bargaining power, unrealistic goal expectations, unexpected environmental changes, inability to manage multiple tensions, and so on (Faems et al., 2004; Faems, 2006; Faems et al., 2006).

Transnational collaboration

183

2.2 The Impact of Government Support for Inter-institutional Collaboration The purported benefits of government sponsoring of inter-institutional S&T collaboration include the reduction of R&D costs and financial risks and increasing the expected benefits (public sector demand signalling effect) and building trust (Lechevalier et al., 2008). Government funding has been found to be instrumental in bringing about R&D cooperation. Busom and Fernandez-Ribas (2008) found that public support significantly increases the chances that a firm will cooperate with a public research organization and that public support also increases the likelihood that a firm will establish private partnerships but to a smaller extent and only when firms have tangible knowledge assets so that public R&D programmes trigger behavioural change in firms' R&D partnerships, alleviating barriers to cooperation. Participation in government-sponsored R&D consortia has generally been found to be beneficial for participants. Sakakibara (1997) found that the net benefits (after accounting for coordination costs) of participation in government-sponsored R&D consortia are positive and that participation in such consortia tends to be associated with higher levels of R&D spending by and also seems to raise the research productivity of participating firms. Branstetter and Sakakibara (2000) found that the outcomes of government-sponsored R&D consortia are positively associated with the level of potential R&D spillovers within the consortium and (weakly) negatively associated with the degree of product market competition among consortium members, and that such consortia are most effective when focusing on basic research. Sakakibara (2001) found a positive impact of the diversity (in terms of core business, technological diversity) of government-sponsored R&D consortia on their participants' R&D expenditure. The results of Lechevalier et al. (2008) are thought-provoking. They found that participation in government programmes has a positive impact on the research productivity of participating firms and that participation in government-sponsored consortia has a greater impact on research productivity than participation in collaborative R&D among firms. 2.3 Transnational Collaboration (Consortia) in the Execution of Publicly Supported Research The positive impact of participation in government-sponsored consortia is also confirmed when it comes to transnational consortia. Ex post evaluations of the EU F P demonstrate that the valorization of high-quality research via peer-reviewed scientific publications is an important goal for

184

European science and technologj. policy

FP participants, especially those from universities and research institutes, now accounting for most F P participations and funding; FP participants generally achieve their publications goals, thus generating large numbers of scientific publications; and publications by scientists participating in the F P have high citation impact (Delanghe and Muldur. 2007). The crossborder project collaboration within the context of the F P enhances the development and use of new tools and techniques; the design and testing of models and simulations; the production of prototypes, demonstrators, pilots and other forms of technological development (Delanghe and Muldur, 2007). It also generates a large number of patents, once more exerting a positive influence on an important S&T indicator (Delanghe and Muldur, 2007). Ex post evaluations of EUREKA confirm the benefits of participating in government-sponsored internationally collaborative research consortia. The 2005-06 Annual Impact Report, for instance, found strong evidence of positive socioeconomic benefits arising from EUREKA projects. It contained encouraging results for obvious success measures (for example, sales of innovative products) but also identified other benefits concerning firm organization, methods, capabilities and competencies, networks, alliances and prestige. It also identified wider social benefits. to the direct customers of the innovating firms, to end-users who benefited from improved environments, safety or better value for money and functionality, and in terms of employment created, safeguarded or enhanced. It found evidence of the role of public funding in enabling, extending or accelerating the work (EUREKA, 2006).

3 CONCLUSIONS European policy-makers appear to conceive of the European Research Area as a single innovation system in need of increased system-wide S&T connectivity and collaboration. Within this context, transnational collaboration between Member States in the public funding of research is high on the political agenda. Yet the extent of such collaboration currently taking place in Europe is comparatively small. This is surprising given the potential benefits of such collaboration. Pooling financial resources allows for overcoming more easily the high barriers-to-entry characterizing some S&T fields. Transnational public research funding collaboration enables pan-European public research programme optimization; promotes scientific excellence; allows for addressing challenges jointly, developing common solutions and speaking with one voice; facilitates the coordination of different research programmes and horizontal policy coordination;

Transnational collaboration

185

reduces programme management costs; facilitates the access of industry to public research support; facilitates the development of joint publicprivate strategic research agendas; and promotes cross-border project collaboration. The above-mentioned benefits are potential and largely theoretical and identified through logical deduction. More ex post evaluation work is needed on concrete transnational public research funding collaboration initiatives in order to better evaluate the scale of these benefits in practice. One of the main objectives of cross-border public research funding collaboration is the promotion of transnational collaboration in the execution of publicly support research, on which comparatively speaking much more literature is available. Based on the 'open innovation' paradigm, inter-institutional S&T collaboration is increasing. Such S&T collaboration, though risky, has many possible benefits and has generally been found to be beneficial for participants, provided that coordination costs are controlled. Government funding has been found to be instrumental in bringing about R&D cooperation and participation in governmentsponsored R&D consortia has generally been found to be beneficial for participants. That is also the case for the subset of internationally collaborative government-sponsored R&D consortia as supported through the Community FPs or schemes like EUREKA. From the above, it would appear that a strong rationale exists for transnational public research funding and research collaboration. An important question then arises: how to organize such transnational public research funding and research collaboration? At what policy level (regional, national, intergovernmental, European) should such collaboration be organized? In transnational public research funding collaboration as well as in transnational research collaboration, there are clear legal bases for Community action. The right for the Community to act in the area of transnational public research funding collaboration is set out in several articles of the Treaty, which make provisions for research coordination and cooperation between Member States and the Community. Article 165 stipulates that 'the Community and the Member States shall coordinate their research and technological development activities so as to ensure that national policies and Community policy are mutually consistent'. It also allows the Commission, in close cooperation with the Member States, to 'take any useful initiative' to promote such coordination. Article 169 allows the Community to make 'provision, in agreement with the Member States concerned, for participation in research and development programmes undertaken by several Member States, including participation in the structures created for the execution of those programmes'.

186

European science and technology poliqs

And Article 171 allows the Community to 'set up joint undertakings or any other structure necessary for the efficient execution of Community research, technological development and demonstration programme^'.^ As far as transnational research collaboration is concerned, Article 163 stipulates that: 1. The Community shall have the objective of strengthening the scientific and technological bases of Community industry and encouraging it to become more competitive at international level, while promoting all the research activities deemed necessary by virtue of other Chapters of this Treaty. 2. For this purpose the Community shall, throughout the Community, encourage undertakings, including small and medium-sized undertakings, research centres and universities in their research and technological development activities of high quality; it shall support their efforts to cooperate with one another, aiming, notably, at enabling undertakings to exploit the internal market potential to the full, in particular through the opening-up of national public contracts, the definition of common standards and the removal of legal and fiscal obstacles to that cooperation'.

And Article 164 states that 'In pursuing these objectives, the Community shall carry out the following activities, complementing the activities carried out in the Member States: (a) implementation of research, technological development and demonstration programmes, by promoting cooperation with and between undertakings, research centres and universities'. It is clear, however, that the existence of strong legal bases for Community intervention in both kinds of transnational S&T collaboration is not in itself a sufficient basis for the Community to take action. A much better evidence-based understanding is required of the best level of government at which to organize such collaboration under which circumstances and for what purposes. This requires not only a better grasp of the effects of transnational S&T collaboration, but also, and equally importantly, an improved understanding of the benefits of purely regional or national funding. For example, it is important to know when such funding can play a role in the promotion of global excellence, efficiency and effectiveness and when it is more adapted to aims such as local research needs or building up basic local research capabilities. These considerations imply a more systematic use of the criterion of added-value across all levels of intervention - regional, national and EU. If not, there is a risk of uncritically accepting purely regional or national funding as the default mode of public research funding, and of only applying the criterion of added-value when it comes to deviations from this default mode.

Transnational collaboration

187

NOTES I.

European Cooperation in Science and Technology.

2. European Science Foundation.

Based on: Schmied, 1975; Schmied, 1977; Schmied, 1982 Ofice of Science and 2003; Technology. 1993: European Commission, 1999; Byckling rt dl., 2000: Autio et d., Autio et ul., 2004; Note by Punel B; Vuola and Hameri, 2006. 4. Several Framework Programme supported ERA-NETS have identified topic coverage gaps and duplication. In the field of Cancer Clinical Practice Guidelines (CPGs) research, many CPGs address the same topic (see the SWOT analysis prepared by Co Can CPG, D 4.5.1, p. 7). Risk of overlaps has also been found in the field of organic food security, especially in animal production system research (Nykanem and Canali, 2006). 5. Treaty establishing the European Community, Official Journal C 325 of 24 December 2002. The Treaty of Lisbon amending the Treaty on European Union and the Treaty establishing the European Community left these articles unchanged. with the exception of Article 165 (see Official Journul C 306 of 17 December 2007). 3.

REFERENCES Adams, J., K. Gurney and S. Marshall (2007), Patterns oj'Interna/ionalCollahorutionfiw the UK und Leuding Partners (Summary Report), a report commissioned by the UK Office of Science and Innovation, Leeds: Evidence Ltd. Autio, E., M. Bianchi-Streit and A.-P. Hameri (2003), Technology TrunsJi~r and Technologicul Learning through CERN's Procurement Activity, Geneva: CERN. Autio, E., A.-P. Hameri and 0. Vuola (2004), 'A framework of industrial knowledge spillovers in big-science centers', Reseurch Policy, 33, 107-26. Branstetter, L.G. and M. Sakakibara (2000), 'When do research consortia work well and why? Evidence from Japanese panel data', National Bureau of Economic Research, working paper no. 7972, Cambridge, MA. Busom, I. and A. Fernandez-Ribas (2008), 'The impact of firm participation in R&D programmes on R&D partnerships', Research Policy, 37 (2), 240-57. Byckling, E., A.-P. Hameri, T. Pettersson and H. Wenninger (2000), 'Spin-offs from C E R N and the case of TuoviWDM', Technovation, 20 (2). 71-80. Capron, H. and M. Cincera (1999), 'The Flemish innovation system: an external viewpoint', Institute for the Promotion of Innovation by Science and Technology in Flanders (IWT) studies no. 28, Brussels, Belgium: IWT. Chang, Y.-C. (2001), 'Benefits of cooperation on innovative performance empirical evidence from the UK's biomedical sector, PREST, discussion paper no. 01-08, Manchester. Chesbrough, H.W. (2003), Open Innovation: The New Imperutiwfbr Creating und Profitingfrom Technology, Boston, M A : Harvard Business School Publishing Corporation. Delanghe, H. and U. Muldur (2007), 'Ex-ante impact assessment of research programmes the experience of the European Union's 7th Framework Programme', Science and Public Policy, 34 (10). 169-83. EUREKA (2006), Annual Impact Report of EUREKA 2005-2006, Brussels: EUREKA. European Commission (1999). Truining and Mobility of'Re.seurcher.s( 1W4-lY98). -

-

188

Europeun science and technology policy

Mid-term Review of the Access to Large-scale Facilities Activity Panel's Midterm Review Report, Brussels: European Commission. European Commission (2004), European Competitiveness Report 2004, Luxembourg: European Commission. European Commission (2007), The Economic Research Area: Ne)r Perspectives, Green Paper, 4 June, Luxembourg. Faems, D. (2006), 'Collaboration for innovation: processes of governance and learning in R&D alliances', PhD thesis at Katholieke Universiteit Leuven, Belgium. Faems, D., M. Janssens, R. Bouwen and B. Van Looy (2004). 'Explorative R&D collaboration: searching for effective and efficient governance mechanisms'. DTEW research report no. 0447, Katholieke Universiteit Leuven, Belgium. Faems, D., M. Janssens, R. Bouwen and B. Van Looy (2006), 'Governing explorative R&D alliances: searching for effective strategies', Manugement Revue, 17 (I), 9-29. Falvey, R., J. Poyago-Theotoky and K. Teerasuwannajak (2006), 'Coordination costs: a drawback for research joint ventures?', working paper (abstract available at ssrn.com/abstract =9O 1804, accessed 20 May 2009.) Finne, T. (2003), 'R&D collaboration: the process, risks and checkpoints', Information Systems Control Journal, 2 (2003), 18-2 1. Glanzel, W. (2001), 'National characteristics in international scientific co-authorship relations', Scientometrics, 51 (I), 69-1 15. Glanzel, W., A. Schubert and H.-J. Czerwon (1999), 'A bibliometric analysis of international scientific cooperation of the European Union (1985-1995)', Scientometrics, 45 (2), 185-202. Gomez, I., M.T. Fernandez and A. Mendez (1995), 'Collaboration patterns of Spanish scientific publications in different research areas and disciplines', in: M.E.D. Koenig and A. Bookstein (eds), Proceedings of the Biennial Conference of the International Societyfor Scientometrics and Informetrics, Medford, NJ, pp. 187-96. Guellec, D. and M.P. Zuniga (2008), 'Globalisation of Technology Captured with Patent Data. A Preliminary Investigation at the Country Level', in: Statistics Sweden Yearbook on Productivity 2007, pp. 109-24. Guellec, D. and B. Van Pottelsberghe de la Potterie (2002), 'The value of patents and patenting strategies: countries and technology areas patterns'. Economics of Innovation and New Technologies, 11 (2). 13348. Lechevalier, S., Y. Ikeda and J. Nishimura (2008), 'The effect of participation in government consortia on the R&D productivity of firms: a case study of robot technology in Japan', The Institute of Economic Research discussion paper series a no. 500 Hitotsubashi University, Tokyo. Motohashi, K. (2004), 'University-industry collaboration and the importance of R&D-focused small and medium enterprises - their implications on Japan's innovation system', Keizai Sangyo Journal, 3. Note by Panel B. (2000), 'Technological innovation, industrial and socio-economic aspects of research infrastructures', Prepared for the conference and incorporating highlights of the debate in Strasbourg on 19 September. Nykanem, A. and Canali, S. (2006), 'An Analysis of O F F Research Topics in CORE Organic Participating Countries Conducted', CORE Organic Project D4.l.a, p. 19. Organisation for Economic Co-operation and Development (OECD) (2002). Dynamising National Innovation Systems, Paris: OECD.

Transnational collahorution

189

Office of Science and Technology (1993), Economic Impacts oj'Hosting International Scientific Facilities, London: HMSO. Okamuro, H. (2007), 'Determinants of successful R&D cooperation in Japanese small businesses: the impact of organisational and contractual characteristics', Research Policy, 36 ( 1 0). 152944. Sakakibara, M. (1997), 'Evaluation of government-sponsored R&D consortia in Japan', in: OECD (ed.), Policj~Evaluation in Innovation and TecI~n~logy Towwds Best Practices, Paris: OECD. Sakakibara, M. (2001), 'The diversity of R&D consortia and firm behavior: evidence from Japanese data', Journal of Industrial Economics, 49 (2), 181-96. Santoro, M.D. and A.K. Chakrabarti (20021, 'Firm size and technology centrality in industry-university interactions', Research Policy, 31 ( 7 ) ,1163-80. Schmied, H. (1975), A Study of'Economic Utility Resultingfkorn CERN Contmrts, Geneva: CERN. Schmied, H. (1977), 'A study of economic utility resulting from CERN contracts', IEEE Transactions on Engineering Management, EM-24, 125-38. Schmied, H. (1982), 'Results of attempts to quantify the secondary economic effects generated by big research centers', IEEE Transactions on Engineering Management, EM-29, 15465. Vuola, 0. and A.-P. Hameri (2006), 'Mutually benefiting joint innovation process between industry and big-science', Technovation, 26 (I), 3-12.

ERA-NET

EU support or membersh~p

Member State bilateral, trilateral etc. cooperation Inter-governmental organizations: European Centre for Medium Range Weather Forecasts (ECMWF) European Molecular Biology Conference (EMBC) European Molecular Biology Laboratory (EMBL) European Organisation for Astronomical Research in the Southern Hemisphere (ESO)" European Organisation for Nuclear Research ( C E R N ) (EC has observer status) European Organisation for the Exploitation of Meteorological Satellites (EUMETSAT) European Space Agency (ESA) (EC-ESA Framework Agreement since May 2004) European Synchrotron Radiation Facility (ESRF) lnstitut Laue-Langevin (ILL) Etc. ERA-NET Article 169 COST (EC supports Secretariat) EUREKAh (EC is full member)

lntergovernrnental Cooperation

NOIC.F: a . Also known as European Southern Observatory. b. Pan-European network for market-oriented R&D

European Science Foundation (EC has observer status)

No EU support or membership

Cooperation Among Research Funding Agencies

Table IOA. 1 Kin& of transnational public researchfunding collaboration

Framework Programme

Cooperation at EU Level

PART 3

Achievements

The EU R&D under-investment: patterns in R&D expenditure and financing Vincent Duchene, Elissavet Lykogianni and Arnold Verbeek 1

THE POLICY CONTEXT: R&D INVESTMENT AT

THE HEART OF THE LISBON STRATEGY Over the past decade, increased policy attention has been paid to the level and growth of research and development (R&D) investment in the various sectors of the European economy. This increased policy concern finds its origin in two main trends. On the one hand, there is mounting evidence that in spite of the recent economic downturn in the US, the EU suffers from a structural growth handicap. Since the mid-1990s and for the first time since the end of World War 11, the Union is no longer catching up with the US in terms of productivity. This may reflect under-performance in the creation, diffusion and utilization of new knowledge over recent years (Denis et al., 2006). On the other hand, with the rapid rise of mainly Asian newly emerging economies, a 'multi-polar world' is developing in which the sources of competitiveness, such as technology and human capital, are more evenly distributed than ever before. In this new context, the EU represents a diminishing share of the world's population, GDP and R&D investment. Newly emerging economies are no longer competing on the basis of low-cost activities alone: China, for instance, is about to overtake the EU-27 in terms of both its share of global high-tech product exports and its R&D intensity (European Commission, 2007a). The divergence between the EU's output performance and the US'S, together with the increasing challenges and new opportunities created by major new players, has led most notably to the initiation of the Lisbon process and its related efforts to encourage governments to launch employment - and productivity enhancing reforms. More in particular, under the influence of the Lisbon Strategy (2000), the Barcelona '3 per cent' objective for more investment in research and development in Europe -

-

-

194

European science and technology policy

(2002) and the renewed Lisbon Strategy (2005), all EU-27 Member States have recently set ambitious targets to significantly increase their share of resources devoted to R&D in the coming years. These commitments to R&D are also reflected in the significantly increased budget for the 7th Framework Programme.' According to the Lisbon and Barcelona objectives, the EU should approach by 2010 '3 per cent' of its G D P being devoted to R&D, of which two-thirds should be financed by the private t business) sector. If each of the 27 Member States would reach the target it has recently fixed for 2010, the EU as a whole would increase its R&D intensity to around 2.6 per cent of G D P in 2010 (in 2006. the EU-27 R&D intensity amounted to 1.84 per cent). In this context, in order to increase the level and 'quality' (impact) of R&D investment and within the broader package of Lisbon-driven reforms, Member States are initiating in-depth reforms to reinforce the mix of policy measures, to foster private R&D investment and, more generally, to better address the perceived weaknesses of their innovation systems. In the context of the re-launch of the Lisbon Strategy, these in-depth reforms were presented in National Reform Programmes that were drawn up by the Member States in the course of 2005 and deal with macroeconomic, microeconomic and employment issues. These reform programmes constituted the beginning of a three-year cycle during which each Member State reports on a yearly basis on the implementation of reforms to the European Commission. The European Commission assesses the progress made and indicates where it deems further action is necessary at Member State or Community level. The second cycle in the implementation of reforms was launched at the end of 2008. The relaunch of the Lisbon Strategy in 2005, reflected in the setting up of national reform programmes impacting on the karious domains of the Lisbon Agenda, is an important milestone because it aims at significantly reinforcing the interaction and coherence between policies at both Member State and European level. In a way, it reflects the need for a systemic approach towards the Lisbon objectives. However, in spite of increased commitment at both national and Community level, recent evidence on trends up to 2006 shows that the E U is not yet on track to meet these targets. Only a small number of Member States (Austria, Denmark, Ireland, Germany and Finland) have over recent years experienced rates of growth in their domestic R&D expenditure, which, if they were maintained, would be sufficient to advance these countries significantly towards their targets. While recent reforms may not yet have induced (visible) effects, it remains important to monitor any further evolution and better understand the mechanisms underlying R&D intensity (R&D expenditure-to-GDP ratio). In this context, the debate on the extent, origin and nature of relative

195

The EU R&D under-investment

R&D under-investment in Europe has gained importance over recent years. Based on the most recent official statistics in this regard, this chapter presents and discusses the main indicators on R&D expenditure and funding in the EU and other major world economies. Where possible, the scope has been broadened as much as possible to include 'non-Triadic' economies such as China or South Korea in the benchmarking. Since the sectoral dimension is key to a better understanding of the origin of the relative EU under-investment in R&D, special attention will be paid to the level and growth of R&D expenditure at industry level.

2 OVERALL INVESTMENT IN R&D 2.1

The Weight of Europe in Worldwide R&D is Shrinking

..

The weight of advanced economies such as the European Union, the US and Japan in global R&D is shrinking. According to the latest OECD data, the EU-27 share has declined from 29 per cent in 1995 to 25 per cent in 2005. Similarly the US and Japan have lost 4 and 3 percentage points respectively of their shares over the same period. Conversely, all emerging economies account for an increasing share of global R&D activity, mirroring the rapid expansion of their science and technology systems. This is particularly true for China and, to a lesser extent, for South Korea and other Asian economies such as Singapore and Taiwan. Consequently, since the end of the 1990s, Asia has taken over Europe's position as the second biggest region worldwide in terms of R&D activity (behind the US). Moreover, the growing share of emerging countries in global R&D is not only due to their booming economic activity, it is also accounted for by the fact that their economies become significantly more researchintensive (OECD, 2008a).

2.2

. ..And the EU's R&D Intensity is Stagnating at a Low Level

Compared with other major world economies such as the US, Japan or South Korea, Europe's R&D intensity remains at a lower level (Figure 1 1.1). After a period of slow but continued growth between the mid- 1990s and 2001, the EU's R&D intensity stagnated in 2001-02, decreased slightly after that and is stagnating again since 2005. According to the latest official data available, in 2006, only 1.84 per cent of GDP was spent on R&D in the EU-27. In Japan, the US and South Korea, the trend over the past decade has been much more positive -- in spite of some minor short-term fluctuations - outpacing Europe's performance in R&D intensity growth.

European science and technology policy

196

J P ~

3.5,

Notes: a. US: Break in series between 1998 and previous years: JP: break in series between 1996 and previous years. b. JP: GERD was adjusted by OECD for the years 1991 to 1995 inclusive. c. KR: GERD does not include R&D in the social sciences and humanities. d. US: GERD does not include most or all capital expenditure. e. CN: Hong Kong is not included. f. Gross expenditure on R&D. Source:

Eurostat, OECD

Figure 11.1

R&D intensity ( G E R D as per cent of G D P ) , 1995-2006"

As a result, the R&D intensity gap with our main competitors has not been reduced at all. On the contrary, if current trends as observed over the past five years continue, Europe's R&D intensity will have declined by 2010 to its mid-1990s level of under 1.80 per cent of GDP. Moreover, new emerging economies such as China are rapidly catching up. It is expected that if current trends persist, China will have caught up with the EU-27 by 2009. The Russian Federation also increased substantially its allocation of resources to R&D between 1995 and 2003 (whereas thereafter it fell back to its pre-2001 level).

2.3 Funding of R&D: Contribution from the Private Sector Remains Low The business sector accounts for the largest part of the overall R&D intensity gap between the EU and its main competitors. The lower contribution of the business sector to the financing of R&D in the EU explains almost 85 per cent of the gap between the EU and the US and an even

The EU R&D under-investment

0.60 0.40 0.20 0.00

Note: Source:

197

-----.__ EU-27

-financed by business enterprise financed by government

-

1995 1996 1997 1998 1999 2000

2001 2002 2003 2004 2005

a. US: GERD does not include most or all capital expenditure Eurostat, OECD

Figure 11.2

Total intramural R&D expenditure (GERD) by sectors o j performance and source offunds

larger part of the gap between the EU and two Asian countries (European Commission, 2007a, p. 24). Despite increased attention from policy-makers, business funding of R&D remains low and has even decreased since 2000. European heads of state decided at the Barcelona Summit of March 2002 to increase not only R&D intensity but also the private-sector contribution to the financing of R&D. In particular, they set the target of increasing the share of R&D expenditure funded by the business enterprise sector to two-thirds of the total by 2010. As shown on Figure 11.2, the private sector contribution to the financing of R&D in the EU has not progressed substantially over the past ten years. R&D financed by the business sector remained at about 1 per cent of GDP in the EU, without any noticeable variation over the decade. In 2004, the private sector financed 64 per cent of total R&D in the US, 67 per cent in China and 75 per cent in both Japan and South Korea, compared with only 55 per cent in the EU. In the US, the trend over the past decade is clearly positive, despite a trend reversal in 2001-02. In China and to a lesser extent Japan as well, the financing of R&D by the private sector has increased at a much faster pace than in the EU. Moreover, since 2000, the private sector contribution to the financing of R&D has actually been decreasing in the EU. As a result, the gap between Europe and the US and Japan has widened significantly over the past decade. R&D funded by government has in general remained very stable and at rather similar levels in the EU and the US, although slightly less stable in the

198

European science and technology policy

US (between 0.6 per cent and 0.8 per cent of GDP). This shows that the business sector is the funding sector that is mainly responsible for the increasing R&D intensity gap between the EU and the US over the past decade. When considering these figures, it is worth mentioning that the level of domestic R&D financed from private sources is slightly underestimated in the EU due to the unavailability of a breakdown between private and public sources in the category 'funded from abroad'. However, since total R&D expenditure funded from abroad represented only 0.16 per cent of GDP in the EU-27 in 2005, this margin of error does not invalidate the observation that the bulk of the R&D intensity gap is caused by the low and stagnating business sector contribution to the funding of R&D.

3 BUSINESS R&D 3.1 Business R&D Intensity is Low in the EU Compared with Other Major World Economies In OECD countries, two-thirds to three-quarters of all R&D activities are carried out by the business enterprise sector.' Therefore, the business sector is not only the principal financing sector of R&D, it is also by far the main performer of R&D. Moreover, within the 'research fabric', the business sector is closest to consumers and therefore best positioned to develop products based on new knowledge (or new combinations of existing knowledge) and exploit them commercially. In the EU, however, mirroring the evolution in overall R&D investment, business R&D expenditure remains low and is stagnating. R&D expenditure in the business sector, at about 1.2 per cent of GDP, remains at a level that is lower than in most other main world regions (Figure 1 1.3). Whereas business expenditure on R&D (as a per cent of GDP) increased in the second half of the 1990s, the trend has been negative since 2001. Conversely, business R&D is increasing at a fast pace in Asia (even though Japan's rate of growth is diminishing) while in the US, the downward trend of 2001-02 has come to an end and growth has turned slightly positive. If these trends are maintained, private R&D investment in China will have reached the same level as in the EU by the end of 2008.

3.2 The EU's Declining Attractiveness for Internationally Mobile R&D Investment In tandem with the overall process of globalization, the 'R&D fabric' is becoming increasingly internationalized. While there has been no drastic

The EU R&D under-investment

199

Notes:

a. b. c.

KR: BERD does not include R&D in the social sciences and humanities US: BERD does not include most or all capital expenditure. CN: Hong Kong is not included.

Source:

Eurostat. OECD

Figure 11.3 Business enterprise expenditure on R&D ( B E R D ) as per cent qf GDP

variation in overall R&D intensities (with the exception of China), there has been a significant shift in the level of internationally controlled business R&D. Based on a survey of 209 multinational enterprises, Reger (2002) concludes that the share of the business R&D budget spent outside the home country by Western European, Japanese and North American multinationals increased significantly between 1995 and 2004, from 26 per cent to 44 per cent (Western European multinationals), from 5 per cent to 15 per cent (Japanese multinationals) and from 23 per cent to 35 per cent (North American multinationals) respectively.' Other, more recent empirical evidence has shown that the pace of internationalization is even accelerating since 2004 (UNCTAD, 2005). Based on a survey conducted between November 2004 and March 2005 among the largest business investors in R&D, an UNCTAD (United Nations Conference on Trade and Development) report concludes that no less than 69 per cent of the responding firms have stated that their share of foreign R&D is set to increase (against only 2 per cent indicating a decline). These survey results are confirmed by the latest available data from the OECD's Activities of Foreign Affiliates (AFA) database (OECD, 2008b). For countries for which data are available, R&D

European science and technology policj3

EU-15 R&D expenditure in the US

Source:

OECD, Activity of Foreign Affiliates database and OECD estimates, January

2008.

Figure 11.4

R&D Expenditure (in million U S $ ) by afJiliates of foreign parent companies, per region within the Triad (flows bet\c-een US-E U )

performed abroad has increased since 1995 relative to R&D performed at home.4 R&D internationalization has traditionally been an intra-Triad phenomenon with the EU and especially the US as major locations. One of the reasons for the EU's low R&D intensity compared with the US is the decision of large European companies to carry out R&D activities in the US rather than in the EU. These companies probably have good reasons for doing so: their principal market may be in the US or they may want to benefit from US technical expertise. Nevertheless, this phenomenon should normally be reciprocal, with US companies deciding to do research in the EU in order to benefit from local expertise or market openings. However, there is evidence that this is not entirely the case. EU companies tend to invest more in R&D in the US than their US counterparts do in the EU. Between 1997 and 2005, US R&D spending in the EU- 15 increased from $9.7 to 17.0 billion purchasing power parity (PPP) while EU-15 R&D spending in the US increased from $9.9 to 19.1 billion (PPP), turning a net outflow of $0.2 billion into one of $2.1 billion (PPP) (Figure 11.4). Although evidence shows that EU companies might benefit from this 'technology sourcing' by means of knowledge spillovers to the parent company resulting in increased marginal productivity at

The EU R&D under-investment E U ~ Japan

70 61

0

20 1

1995

D l Emerging

Asian countriesC C-3 Middle East

2005

ili0te.s:

a.

b. c.

US outward R&D spending refers to R&D expenditure performed by majority-owned (more than 50 per cent ownership) non-bank foreign affiliates of non-bank US parent companies. Data include R&D expenditures conducted by affiliates, whether for themselves or for others under contract; exclude R&D expenditures conducted by others for affiliates under contract. EU-15 up to 2003, EU-25 from 2004 on. China. Hong Kong, India, Indonesia, South Korea, Malaysia, Philippines, Singapore, Taiwan and Thailand.

Sourcc~: OECD, Activities of Foreign Affiliates Database, January 2008

Figure 11.5 Destination of US outward R&D spending in 1995 and 2005" company level in the region of origin (Griffith et al., 2006), such an increasing net outflow reflects the stronger attractiveness of the US research and innovation system compared with that of the EU. Moreover, the internationalization of R&D is no longer limited to intra-Triad flows. More recently, this phenomenon has become more truly global, with many emerging economies becoming important locations for internationally mobile R&D facilities. This more global focus of R&D spending can be seen in the increasing diversification of the US'S own outward R&D investment (Figure 11.5). US firms are targeting all major world regions, especially the emerging Asian economies and, to a lesser extent, the Middle East, the result being that the EU's share in US outward R&D spending has been decreasing significantly since the mid1990s. This trend is expected to continue as new emerging market players continue to build up their science and technology systems and open up their

202

European science and technology policj.

markets to foreign entrants. For instance, a 2007 study by the Economist Intelligence Unit of more than 300 senior executives identified India (26 per cent of respondents), the United States (22 per cent) and China (14 per cent) as the most attractive overseas locations for R&D (The Economist, 2007). The Asia-Pacific region in particular is expected to attract more foreign R&D investment over the next three years: 30 per cent of the respondents planned a substantial increase in their R&D activity in that region. Similarly, in the 2005 UNCTAD survey of the largest R&D spenders worldwide, China ranked third and India sixth among the currently most attractive locations for R&D (UNCTAD, 2005). The findings of the same survey on the most attractive R&D locations for future R&D investment put China at the top of the ranking, followed by the US and India. Large EU Member States such as the United Kingdom, France or Germany ranked fifth, seventh and eighth respectively. Finally, Thursby and Thursby (2006) surveyed in April 2005, US and EU multinationals on their plans to increaseldecrease technical employment (that is, human resources supporting R&D work) in the subsequent three years. They found that emerging countries such as India and China will continue to be the major beneficiaries of R&D expansion in the coming years while the prospects for Western Europe are rather dim (ibid.). 3.3 The Industrial Structure is Key to Explaining the EU-US Funding Gap Key to a better understanding of the EU 'R&D deficit' is the sector-specific analysis of R&D investment. The nature and extent of the 'technological specialization' (that is, internal composition of the industry structure) of a country has recently received increased attention in many policy reports and academic contributions (for a recent overview see O'Sullivan, 2007). According to Mathieu and van Pottelsberghe (2008), 'technological specialization' (proxied by the industry-specific distribution of R&D spending in manufacturing sectors) even explains the variation in R&D intensity across countries much better than any other country-specific characteristic. For this reason, this section analyses the distribution of R&D expenditure across manufacturing sector^.^ A comparison of the distribution of manufacturing R&D across industrial sectors classified according to their level of technology intensity shows that in the US, manufacturing R&D is more concentrated in hightech sectors than in the EU (Table 11.1). In 2003, 55 per cent (0.5611.02) of total EU manufacturing R&D occurred in high-tech sectors compared with 69 per cent in the US (0.8111.18). European industrial R&D is more likely to be concentrated in medium-high-tech and, to a lesser extent, medium-low-tech manufacturing.

203

The EU R&D under-investment

Tuble 11.1 Manuficturing BERD and value-added by type of industry, 2003"."

Total Manufacturing

Of which: High-tech

MediumHigh-tech

MediumLow-tech

Low-tech

Manufacturing B E R D a s '%I of total G D P

EU-27h uS R a t i o US/ EU-27h Value-added a s '%I of total

GDP EU-27h US R a t i o US1 EU-27 Manufacturing B E R D as 'XI of Valueadded

EU-27' US Notes:

a. b.

c.

In the absence of a breakdown for value added between pharmaceutical (high-tech) and other chemical products (medium-high-tech), total chemicals ( i t . pharmaceuticals + other chemical products) has been included in high-tech. EU-27 does not include: BG, EE, LV, LT, LU, CY. MT, AT, PT, RO, S1, SK. 2003 was selected as reference year as it has the widest geographical EU coverage.

Sourw:

Eurostat, OECD (2006) Groningen Growth and Development Centrc.

As shown in Table 1 1.1, high-tech industries show a slightly higher R&D intensity in the US (21.9 per cent) than in the EU (1 8.1 per cent). This, however, may be due to the inclusion of the sector 'total chemicals' in the high-tech category (see note under Table 11.1). 'Total chemicals' is larger in the EU than in the US but in both the EU and the US, it is also less R&D-intensive than high-tech industries. Medium-high-tech and medium-low-tech industries have very similar (or even identical) levels of

204

European science and rechnology polic)

R&D intensity in both the EU and the US. In conclusion, it appears that R&D intensity by type of industry is very similar in the EU and the US.h Therefore, the higher concentration of business R&D in high-tech industries in the US largely emanates from differences in industrial structure between the EU and the US. As shown in Table 11.1, in the US. high-tech industries account for a much larger share of both industrial value-added and GDP than in the EU. In the EU, the industrial texture is more concentrated on medium-high-tech, medium-low-tech and low-tech activities. Although not fully comparable with the Analytical Business Enterprise Research and Development (ANBERD) data used here to analyse the distribution of business R&D across sectors, data from the 2006 ELIndustrial R&D Investment Scoreboard (European Commission. 2007b) on the composition of corporate investment made by the largest R&D spending companies worldwide confirm the differences between the EU and the US. According to the Scoreboard, EU companies considered sector by sector appear to be as R&D-intensive as their US counterparts (ibid., pp. 9-10). The deficit in private R&D spending is mostly due to differences in industry structure and the smaller size of the high-tech sectors. As illustrated in Figure 11.6,67 per cent of US corporate R&D investment is made by companies belonging to high R&D intensity sectors, compared with just 36 per cent for EU companies. Figure 11.6 also illustrates how the ICT sector accounts for a large part of the difference in the sectoral composition of R&D investment by US and EU companies.' 3.4 Which Sectors Account for Most of the EU-US R&D-funding Gap? Since the EU R&D deficit with the US appears to be located primarily in the high-tech manufacturing industry, it is worth examining EU-US differences in the composition of high-tech industry and the relative importance of each sector in the R&D funding gap (Figure 11.7 and Figure 113 ) . The heavier reliance of the EU on medium-high-tech industries justifies a deeper analysis of the composition of this sector. Figure 11.7 shows both the R&D expenditure and the value-added (as percentage of GDP) for each sub-sector of the high-tech and medium-high-tech industries. Figure 1 1.8 shows the R&D intensity of each individual sub-sector. The following observations can be made. The sector 'Chemicals' does not play a significant role in explaining differences between the EU and the US and the higher concentration of R&D in high-tech sectors in the US. This sector is equally large in both economies (somewhat bigger in the EU) and it is as R&D-intensive in the EU as it is in the US (even slightly more R&D-intensive in the EU). The 'Aircraft and spacecraft' industries have equal R&D intensities on both sides of the Atlantic but in the US

The EU R&D under-investment EU 6%

Automobile and parts 24% High R&D-intensive sectors Medium R&D-intensive sectors Low R&D-intensive sectors Very low R&D-intensive sectors

Others "

o

,

"

Sot~rce: The 2006 EU Industrial R&D Investment Scoreboard.

Figure 11.6 Sectoval composition o f R&D investment by EU and U S companies, 2005

European science and technology poliq

206

TOTAL High-techa Radio, televlslon and communicat~onequipment Atrcraft and spacecraft Office, accounting and computing machinery Chemicals and chemical products Medical, precision and optical Instruments TOTAL Medtum-High-techa Motor vehicles Railroad equipment and transport equipment Machinery and equlpment Electrical machinery and apparatus

13BERD as per cent of GDP Value-added as oer cent of GD

TOTAL ~igh-techa

Radlo, television and communicat~onequipment Aircraft and spacecraft Office, accounting and computing machinery Chemicals and chemical products Medical, prec~sionand optlcal instruments

TOTAL Medium-High-techa

Motor vehicles ail road equipment and transport equlpment Machinery and equipment Electrical machinery and apparatus

Notes:

-

a.

b. c.

~

In the absence of a breakdown for value added between pharmaceuticals (high-tech) and other chemical products (medium-high-tech), total chemicals (i.e. pharmaceuticals and other chemical products) has been included in high-tech. EU-27 does not include : BG, EE, LV, LT, LU. CY, MT, AT, PT, RO. SI. SK. 2003 was selected as reference year as it has the widest geographical EU coverage.

Source:

Eurostat, OECD, Groningen Growth and Development Centre

Figure 11.7 High-tech and medium-high-tech industries" - BERD as per cent of GDP and value added as per cent of GDP, EU-2 7b and US, 2003' this sector is almost twice as large as in the EU. It therefore contributes to the higher concentration of R&D in the high-tech sector in the US but only because of its larger size. The 'ICT manufacturing industrie~'~ largely explain the higher concentration of R&D in the high-tech sectors in the US by virtue of both their high R&D intensity and their larger size. 'Office. accounting and computing machinery' is much more R&D-intensive in the US than in the EU but is equally small in both economies. 'Radio. television and communication equipment' is slightly less R&D-intensive in the US but 60 per cent bigger there than in the EU. Finally, the US 'Medical.

The EU R&D under-investment

207

TOTAL High-techa Radio, television and communication equ~pment Aircraft and spacecraft 5'.0

Office, accounting and computing machinery Chemicals and chemical products Medical, precis~onand optical instruments

TOTAL Medium-~igh-techa Motor veh~cles 25 4

6.0

Railroad equipment and transport equipment Machinery and equipment

5.7

Electrical machinery and apparatus

5.2

Nofrs:

a. EU-27 does not include : BG, EE, CY. LU, MT, AT, PT, RO, SI, SK. b. All chemicals and chemical products are included in high-tech c. 2003 was selected as reference year as it has the widest geographical EU coverage. Source:

OECD, Eurostat, Groningen Growth and Development Centre.

Figure 11.8

High-tech and medium-high-tech industried BERD as per cent oJ'GDP and value added, EU-27b and US, 2003' -

precision and optical instruments' sector is twice as R&D-intensive as and almost 50 per cent bigger than the EU one. Two main conclusions can be drawn. First, it is clear that ICT manufacturing industries play a crucial role in explaining the R&D funding gap between the EU and the US not only because they tend to be more R&Dintensive in the US but also because of their larger size. To a much smaller extent, the 'Aircraft and spacecraft' industries contribute to the EU R&D deficit as well. Second, structural differences between the two economies (that is, the larger share of both the ICT manufacturing industries and the 'Aircraft and spacecraft' industries in the industrial texture of the US) seem to be at least as important as 'intrinsic effects' (that is, sector-specific R&D intensities). Similarly, one can examine the sectors that are responsible for the higher concentration of R&D expenditure in the medium-high-tech sectors in the EU. The sector 'Railroad and transport equipment' does not play any significant role in the explanation of the difference: this sector is much more R&D-intensive in the US than in the EU but equally very small in

208

European science and technologj. policj,

both economies. 'Motor vehicles' also plays a rather limited role: it is only slightly bigger and more R&D-intensive in the EU than in the US. The major difference comes from 'Machinery and equipment' and to a lesser extent, 'Electrical machinery and apparatus'. These two sectors have similar R&D intensities in the EU and the US but are twice as big in the EU as in the US. Here again, structural differences and the larger size of sectors seem to account for most of the differences between the EU and the US.

3.5 Dynamics of Industry-level R&D Expenditure If industrial structure and more in particular the size of its high-tech component is key to explaining differences in R&D-intensity between economies, the next question is to what extent this structural composition of R&D expenditure can be modified? In this regard, an analysis of business R&D expenditure (BERD) growth patterns at NACE (French, EU classification system) sector level in the various EU Member States as carried out in the European Commission's Key Figures 2007 is interesting. For instance, Finland, Denmark and Germany are the few Member States for which business R&D intensity has been increasing significantly since 2000. An analysis of growth patterns at the sector level reveals that this increase is largely due to increasing technological specialization in research-intensive industries (European Commission, 2007a. pp. 68-72). In the various Member States analysed, only a very limited number of sectors have played a key role in the growth of BERD. Although their contributions to the growth of BERD vary from country to country, these key sectors are generally the following four: (1) Pharmaceuticals; (2) Motor vehicles; (3) Radio, TV and communication equipment and (4) Computer and related services. Clearly there are technological and market trends that at some point create the conditions in specific sectors leading to very significant increases in BERD. Furthermore, the increase in BERD of a specific sector in a particular country can be due to an increase in the size of that sector or to its specific R&D intensity. The analysis reveals, however, that changes in the industrial structure of the country often play a predominant role in significantly increasing BERD (above changes in R&D-intensity). This confirms once again that the industrial structure and its evolution are key determinants of both the level and the trend of business-funded R&D. From a policy point of view, this means that increasing the (private) R&D intensity and reaching the 'Barcelona targets' should be seen as industrial policy objectives as much as a research policy objectives, which calls for policy tools going well beyond research policy.

The EU R&D under-investment

209

3.6 Fast-growing, Technology-intensive Companies Furthermore, changes in industrial structure and the dynamism of sectors are closely related to (and influenced by) the growth paths of the companies themselves. Small and medium enterprises (SMEs) can indeed grow and become major, critical players in their sector. As previously mentioned, EU companies are sector by sector as R&D-intensive as their US counterparts but tend to be less involved in some very R&D intensive sectorslsub-sectors (especially the ICT sector). Moreover, available data show that in the US more than in the EU, many new R&D-intensive firms active in high-tech sectors (often labelled 'New Technology-based Firms' or NTBFs) were able to grow rapidly and become key economic players. Of the US companies now in the world's top 1000 in terms of market capitalization, 22 per cent were created after 1980, compared with only 5 per cent of their European counterparts (Cohen and Lorenzi, 2000, pp. 122-6).' Of US companies created after 1980 and now in the world's top 1000, 70 per cent are ICT companies. This difference between the EU and the US is not limited to the ICT sector: similar trends can be observed in other emerging high-tech sectors. In the biotech sector, for instance, although the number of companies created is comparable in the EU and the US, the average company turnover and number of employees is much higher in the US than in the EU."' It seems that to a larger extent than the EU economy, the US economy has the flexibility to reorient itself towards promising new sectors, especially through the rapid growth of new R&Dintensive firms. The creation and expansion of new firms in high-technology sectors is essential for allowing Europe to achieve its R&D potential. Ensuring that the right conditions exist to enable NTBF's to flourish in the same way in the EU as they do in the US is therefore of the utmost importance. There is, however, evidence that financial markets are less willing (or able) to fund new sectors and new firms in the EU than in the US (European Commission, 2007a, pp. 36-7). In spite of some recent initiatives (such as the joint EC-EUREKA 'Eurostar' initiative or the EC-EIB Risk-sharing Finance Facility), a key ingredient is still missing here: an EU-wide financing solution for fast-moving European companies."

4 PUBLIC SECTOR R&D AS COMPLEMENT TO PRIVATE R&D Although in both the EU and the US, domestic R&D efforts are largely financed by the business enterprise sector, the role of government in the

210

European science and technology policy

GERD financed by business enterprise as per cent of GDP

Notes: a. IT : 1996; MT, IL : 2002; BE, DK, EL, LU, NL, PT, SE, IS, NO : 2003; BG, DE, EE, ES, FR, CY, RO, FI, UK, HR, TR, US, JP: 2004; AT: 2006. b. CN : Hong Kong is not included.

Source:

Eurostat, OECD.

Figure 11.9

GERD$nanced by business enterprise and by government as per cent of GDP, 2005"

financing of R&D should not be underestimated. High R&D-intensity countries such as Finland, Sweden, Denmark, Austria, Germany, the UK, the US and Japan are characterized by a high level of involvement from the private sector in the funding of R&D activities. However, the level of governmentfunded R&D is also among the highest in these countries, showing that the roles of the private and public sectors are fully complementary (Figure 1 1.9). Moreover, in low R&D-intensity countries, government-funded R&D is higher than business-funded R&D. Government funding of R&D is critical for creating and developing science and technology (S&T) capabilities - a prerequisite for catching up with countries at the technology frontier - and for supporting research projects with high expected social benefits, which the private sector may not find sufficiently attractive. As previously shown in Figure 11.2, R&D funded by government has remained very stable in both the EU and the US but at a lower level in the EU than in the US. Therefore, to allow private R&D activities to develop further and grow on a solid science base, the overall public effort to fund R&D in the EU must be increased as well.

The EU R&D under-investment

21 1

5 CONCLUSION: INTENSIFYING THE PACE OF POLICY REFORMS Together with the EU's recent underperformance in terms of productivity growth compared with the US, the emergence of major new players shows that the 2005 relaunching of the Lisbon Agenda was indeed appropriate. Many countries now accept that solving the EU's growth problem requires a longer-term policy perspective and a sustainable long-term recovery process built upon a Lisbon-inspired structural reform agenda aimed at effectively addressing the fundamental growth challenges posed by the accelerating pace of technological change, globalization and ageing populations. It is particularly essential that the transition of the EU economies towards a knowledge-driven economy -within which education and training, R&D and innovation, ICTs play a critical role - is speeded up. It is therefore necessary to increase the efficiency of R&D, improve the transformation of new ideas into new products, processes, services and solutions and make the overall environment more supportive of firms wanting to increase investment in R&D. While the policy challenge of implementing Lisbon-driven reforms remains a serious one for a large number of EU Member States, it should be clear that the expected gains are considerable. A recent study estimates that the introduction of five key measures of the Lisbon Strategy (that is, the Services Directive, reduction of the administrative burden, improving human capital, '3 per cent' R&D target, increase in the employment rate) can boost the EU's economic and employment growth rates by at least 0.8 per cent per year for more than a decade (Gelauff and Lejour, 2006).

NOTES 1.

FP7 runs from 2007 until 2013; its total budget is 41 per cent higher than the FP6 budget (2004 prices). 2. Partly financed from public sources. 3. Data for 2004 are estimates. 4. Such increased 'R&D performed abroad' may be partially the result of 'incidental' mergers and acquisitions (M&As) having an increased R&D activity abroad as a result (while not being the primary objective of the parent company). The lack of available data making the distinction between 'R&D-driven' investment and pure M&As, however, makes any further assessment of this phenomenon very difficult. 5 . In spite of the fact that the way in which services R&D is reported and classified is not fully comparable across countries, one can estimate that at least three-quarters of total business R&D is concentrated in manufacturing industries in both the EU and the US (European Commission, 2007a, pp. 29-30; Duchkne et al., forthcoming). 6. Even though low-tech industries only account for about 5 per cent of manufacturing

212

European science and technology policy BERD in both the EU and the US and therefore do not play an important role in explaining EU-US differences in business R&D intensity. it is interesting to note that low-tech industries are much more R&D-intensive in the US than in the EU. They also represent a lower share of GDP in the US. R&D investment by these EU (US) companies is not necessarily confined to the territory of the EU (US). 'ICT manufacturing industries' refers to the following three sectors: 'Radio, television and communication equipment'; 'Office, accounting and computing machinery': 'Medical, precision and optical instruments'. This figure does not include the creation of companies through mergers and acquisitions, only ex nihilo creations. Cohen and Lorenzi (2000) note that in 1997. there were 1274 biotechnology companies in the US (compared with 1036 in the EU), which generated a revenue of $15.9 billion and employed 140 000 people (compared with revenue of $2.7 billion and 39 045 people employed in the EU) (p. 126). 'Eurostars' was launched in October 2007. It aims at providing EU and EUREKA (R&D network) member country financial support to research-performing and innovative SMEs. Through this initiative, the EC provides about €100 million of Community funding to support mutually opened national schemes. The Risk-sharing Finance Facility (RSFF) was created on 5 June 2007 by the European Commission and the European Investment Bank (EIB). Sharing the risk between the Commission and the EIB allows the RSFF to produce additional loans for R&D projects (including infrastructure projects) which have a strong European dimension. In addition to bringing in additional, EU-wide financing, both the 'Eurostars' and the RSFF initiatives have a 'demonstration effect': R&D stakeholders (for example, research ministries, companies) and financial institutions will learn to work together, paving the way for further integration of research- and innovation-driven financing.

REFERENCES Cohen, E. a n d J.-H. Lorenzi (2000), Politiques industrielles pour /'Europe, R a p p o r t d u Conseil d'Analyse Economique, N o . 26, Paris: La Documentation franqaise. Denis, C., K . Havik a n d K. McMorrow (2006), 'EU Growth Trends a t the Economy-wide and Industry-Levels', DG E C F I N paper submitted t o the E P C meeting of April 2006, Brussels. DuchCne, V., E. Lykogianni and A. Verbeek (forthcoming), ' R & D in services industries a n d the E U - U S R & D investment gap', Science and Public Poliy. European Commission (2007a), Key Figures 2007 on Science, Technology and Innovation. Towards a European research Area, Luxembourg: Office for Official Publications of the European Communities ( E U R 22572). European Commission (2007b), Monitoring Industrial Reseurch: Analysis of the 2006 EU Industrial R&D Investment Scoreboard, Luxembourg: Office for Official Publications of the European Communities ( E U R LF-NA-22694-EN-C). Gelauff, G . M . M . a n d A . M . Lejour (2006), 'The New Lisbon Strategy: an estimation of the economic impact o f reaching five Lisbon targets', industrial policy a n d economic reforms papers no. 1, T h e Hague: Centraal Planbureau. Griffith, R., R . Harrison a n d J. Van Reenen (2006). 'How special is the special relationship? Using the impact of U S R & D spillovers o n U K firms as a test of technology sourcing', American Economic Review, 96 ( 5 ) , 1859-75.

The EU

R&D under-investnzent

2 13

Mathieu, A. and B. van Pottelsberghe de la Potterie (2008). 'A Note on the Drivers of R&D Intensity', Centre for Economic Policy Research discussion paper no. 6684, London. Organisation for Economic Co-operation and Development (OECD) (2008a). Main Science rind Technology Indicators, 2008i1 edn, Paris: OECD. OECD (2008b), The Internationalisarion of' Business R&D. Evidence, Inzpacts cind Implications, Paris: OECD. O'Sullivan, M. (2007), The EU's R&D Dtlficit and Innovation Policj', report to the 'Knowledge for Growth' expert group of EU Commissioner for Research J. Potocnik, accessed 20 May 2009 at: http://ec.europa.eu/invest-in-researchlpdfl download-enirdd-deficit-report0207.pdf. Reger, G. (2002), 'Internationalisation of research and development in Western European, Japanese and North American multinationals', International Journal of' Entrepreneurship and Innovation Manugerncxt, special issue on Entrepreneurship, Innovation and Globalisation, 2 (2-3), 164-85. Thc Economist (2007), 'Sharing the idea. the emergence of global innovation networks', report by The Economist Intelligence Unit, accessed 20 May http://graphics.eiu.com/fileslad~pdf/eiu~IDA~INNOVATION~ 2009 at: NETWORKS-WP.pdf. Thursby, J . and M. Thursby (2006), Here or There? A Survey of Factors in Multinational R&D Location and IP Protection, Washington, DC: Marion Ewing Kauffman Foundation. U N C T ~ D(2005), World Investment Report. Transnational Corporations and the Internationalization of'R&D, New York and Geneva: United Nations.

12. Does the 'European Paradox' still hold? Did it ever? Giovanni Dosi, Patrick Llerena and Mauro Sylos Labini Since the second half of the 1990s, labour productivity growth has been lower in the Euro area than in the US, as shown in Table 12.1. Small gaps end up producing large differences and policy-makers are rightly concerned about the EU's poor economic performance. What went wrong with the European economy and how can the poor performance be explained? Classical economic growth theory stresses the importance of capital accumulation and savings rates. Unfortunately, however, this approach is not very helpful here since the capital-labour ratio and investment rates are still higher in Europe than in the US. A different often-mentioned explanation refers to Europe's failure to reform its product, service and labour markets. Though such reform can be important, differences in market regulation between the two sides of the Atlantic are not new and already existed when Europe was growing much faster than the US. Another very popular interpretation of the widening EU-US gap is the one that inspired the so-called Lisbon Agenda. One version of this story can be summarized as follows: by the late 1980s, after a long phase of catching up, Europe could no longer rely on capital accumulation and technological imitation as its principal sources of economic growth. At the same time, the information technology revolution was finally producing positive effects on the US economy. Nevertheless, European countries' institutions and policies did not allow them to fully benefit from these new technological paradigms. The same institutions and policies that did a good job in fostering economic development through catching up and imitation - the story goes - were unfit to foster economic growth, stemming from the complex relationship between new scientific discoveries, novel technical innovation and their industrial exploitation. In our view, such an explanation is incomplete and in some ways misleading. It is indisputable that Europe does not invest enough in research and development (R&D) and that its knowledge economy however defined - is weak. Nevertheless, the dismal R&D figure, rather than an -

215

Does the 'European Paradox' still hold:)

Table 12.1 Labour productivity growth

E u r o area United States

199I --95

1996-2000

2001-05

2006-08

1.9 1.3

1.2 2.0

0.5 2.0

0.9 1.2

The table displays multiple years averages. Euro area comprises Austria, Belgium, Finland, France, Germany. Greece, Ireland, Italy, Luxembourg, Netherlands, Portugal, and Spain.

Notes:

Source:

Our elaborations on OECD data.

explanation itself, is likely to be the consequence of some deeper determinants. More specifically, this chapter ventures that what does not work in Europe along the science-technology industry also relates to the quality of knowledge production. A conventional view is that Europe plays a leading role worldwide in terms of top-level scientific output but lags behind in its ability to convert this strength into wealth-generating innovations (European Commission, 1995). We have already critically discussed this conjecture - known as the 'European Paradox' - elsewhere and concluded that the existing pieces of evidence do not support it. On the contrary, what a number of indicators show is that European weaknesses reside in both the European system of scientific research and a relatively weak industry especially with respect to the activities based on the newest technological paradigms (Dosi et al., 2006). This chapter refines and adds more evidence to the analysis cited above.

1 THE MYTH OF EUROPEAN LEADERSHIP IN SCIENCE A central piece of evidence that should support the notion that Europe's major weakness stems from the difficulties it encounters in transforming the findings of its excellent research system into innovations and competitive advantages concerns the width, depth and originality of European science. The stage was set by the 1995 EU Green Paper on Innovution, which measured the scientific impact of Europe using as an indicator the number of publications per euro spent in non-business enterprise research and development (R&D) (European Commission, 1995). This ratio was slightly higher in Europe but this was mainly due to the fact that the total

216

European science and technology policy

Table 12.2 Publications and citations weighted by population and university researchers -

Publications US EU-15

Citations

us EU-15 US

EU-15 Votes:

-

Population 4.64 3.60 Population 39.75 23.03 Top 1 % publications Population 0.09 0.04

Publications Researchers 6.80 4.30

-

Citations

-

Researchers 58.33 27.52 Top 1% publications Researchers 0.13 0.04

Researchers Population 0.68 0.84 Researchers Population 0.68 0.84 Researchers Population 0.68 0.84

Number of publications, citations and top 1 per cent publications refers to 1997-

2001. Population (measured in thousands) and number of university researchers (measured In full time equivalent) refer to 1999. source:' Our calculations based on numbers reported by King (2004) and OECD (2008).

number of publications was (and indeed is) higher in Europe than in the US.' Relying on this evidence, the document concluded that '[clompared with the scientific performance of its principal competitors, that of the EU IS excellent' (European Commission, 1995, p. 5). More recently, although the claims concerning a supposed paradox have become feebler, the European Commission still suggested that high-quality European publications face more obstacles in their translation into technological applications than comparable scientific output in the US (European Commission, 2007, p. 49). The basic indicator used by the Green Paper is largely inaccurate and does not take into account that only a small number of total publications have a significant effect on the advancement of knowledge. This is suggested by the fact that only a fraction of articles are cited while the overwhelming majority receives zero citations. Two of the most used proxies of the impact of any scientific output upon the relevant community are articles' citations and the shares in the top 1 per cent most cited publications. As shown in Table 12.2, once one controls for population, the US is well ahead with respect to both indicators. In particular, the outstanding EU scientific output is still less than half the US one. In the second and third column of the same table. numbers of

Does the 'European Paradox' still hold.?

217

publications, citations and top 1 per cent publications are decomposed into two parts: a measure of university researchers' productivity (that is, output per university researcher) and a ratio of university researchers to p ~ p u l a t i o nThe . ~ table clearly shows that US leadership is due to the quality of research published rather than to the sheer number of university researchers. To further investigate whether the unsatisfactory state of European science is confirmed by other bibliometric indicators, we exploit the data on highly cited researchers (HCRs) collected by Thomson Scientific. More specifically, Thomson identifies those researchers whose collected publications have received the highest number of citations across the past two decades in 21 scientific disciplines.' They comprise less than one-half of 1 per cent of all publishing researchers and can therefore be considered as those researchers who have made fundamental contributions to the advancement of science and technology in recent decades. Of the 6395 affiliated researchers identified, 4232 (about 66 per cent) are currently employed by a US institution and only 1434 (about 22 per cent) by an EU-17 0ne.l As shown in Table 12.3, one observes large disparities across European countries. For example, if the total number of HCRs is normalized for population, Switzerland (which, incidentally, is not an EU member) ranks remarkably well and its scientific impact is similar to the US.' On the other hand, the overall European performance is quite poor. Germany and France in particular have a normalized number of HCRs that is less than one-fifth of the US. Hence, if one replaces a very rough indicator such as the total number of publications with more demanding ones, the European leadership in science becomes a myth and the structural gap with the US increases with those measures that focus on topimpact researchers. Of course, there is substantial interdisciplinary variation in the revealed quality of European research. According to King (2004), when the total share of citations is considered, the superiority of the US is remarkable in the life and medical sciences, while Europe performs slightly better in physics and engineering. However, if one focuses on Institute for Scientific Information (ISI) data on top researchers, a less diversified picture emerge^.^ First, as shown in Table 12.4, the only European country that displays significant strengths in physical sciences and engineering is Switzerland. Second, the only field in which the EU-17 has a higher number of HCRs is pharmacology. However, even in this discipline, the US performs better when one adjusts for population. Third, the gap is indeed larger in the life and medical sciences, where the number of HCRs is around six times higher in the US. The general message stemming from the above evidence is therefore far

European science and technology policy

218

Table 12.3 Highly cited researchers -

Total Number

World Share

Number Per Capita

EU 17 Switzerland

United States Israel United Kingdom Sweden Netherlands

Canada Australia Denmark New Zealand Belgium Finland Norway Germany France Ireland Japan Austria Singapore [taly . ..

Spain Korea Russia Total number refers to most highly cited researchers within 21 scientific fields for the period 1981-99. See www.ISIHighlyCited.com for details (accessed 24 May 2009). The top 20 countries (plus Spain, Korea, and Russia) in terms of number of HCRs per capita are displayed. Researchers highly cited in more than one field have been counted several times. EU-17 is EU-15 plus Norway and Switzerland. Population (in millions) is calculated In 1999. .Votes:

Source:

Our calculations on IS1 Thompson Highly cited scientist data

from suggesting any European leadership in science. On the contrary, one observes a structural gap in top-level research output and in the number of top researchers per capita. These basic facts suggest that one of the likely underlying determinants of the dismal performance of the so-called 'science, technology, innovation systems' in Europe is the weak European scientific output of 'excellence'.

Does the 'European Paradox' still holu':' Tuhle 12.4 HCRs in EU-I 7 and U S hy$elds Total Number Switzerland United States Finland United Kingdom Germany EU17 France Switzerland United States Denmark Germany EU17 France Sweden United Kingdom Switzerland United States EU17 Germany France Sweden United States Belgium Switzerland Ireland Germany EU17 France

United States United Kingdom Finland Netherlands Denmark EU17 France Germany

World Share

Number Per Capita

Phy.sic:s 0.065 0.545 0.007 0.065 0.090 0.317 0.021 Enghwering 0.029 0.683 0.008 0.038 0.169 0.017 Phurrnucology 0.03 1 0.177 0.014 0.379 0.450 0.075 0.044 Biology & Biocherni.str-J 0.026 0.636 0.015 0.007 0.004 0.044 0.195 0.026 Clinicul Medicine 0.752 0.105 0.008 0.023 0.004 0.194 0.012 0.004

Notes: Total number refers to most highly cited researchers within 21 scientific fields for the period 1981-99. See www.ISIHighlyCited.com for details (accessed 24 May 2009). Top five countries plus EU-17, France and Germany are displayed. Researchers highly cited in more than one field have been counted several times. EU-17 is EU-15 plus Norway and Switzerland. Population (in millions) is calculated in 1999.

Source:

Our calculations on IS1 Thompson Highly cited scientist data.

220

2

European science and technology policj

EU UNIVERSITIES IN COMPARATIVE PERSPECTIVE

European universities are obvious candidates to be the culprits in explaining EU weaknesses in scientific productivity. To be sure, comparing the European system of higher education with that of its major competitors is a difficult task and this for a number of reasons. First, notwithstanding recent attempts to converge towards a common EU model, European countries still have quite different academic institutions. Second, reliable cross-country indicators are surprisingly scarce. Nevertheless, important insights can be obtained from the huge case study literature together with several quantitative indicators drawn from the IS1 Highly Cited Researcher database and the Shanghai Academic Ranking of World Universities. The latter ranks top universities using several proxies of academic and research performance including alumni winning Nobel Prizes and Fields Medals, highly cited researchers, articles published in Nature and Science or indexed in major citation indices and the academic per capita performance of a given institution. Scores for each indicator are weighted to obtain a final general score for each listed university.' Of course, the Shanghai ranking is far from perfect. For example, it does not control adequately for sheer size and its weights are somewhat arbitrary. However, its big advantage with respect to other rankings is its transparency and reliance on publicly available information. To begin with, let us evaluate the overall performance of European universities using the data provided by the Shanghai exercise. To transform the university ranking into a measure that allows for country comparisons, we follow the methodology of Aghion et al. (2007). More specifically, we build three distinct country rankings - Top 50, Top 100 and Top 500 - assigning to each university listed in the original ranking, a decreasing score corresponding to the rank. So, for instance, in the Top 100 ranking, the first university gets 100 points, the second 99 and so on. Then, for each ranking, we sum the points obtained by all universities within each country group and divide the obtained sum by country population. Finally, we divide each country's score by the US one, in order to interpret the final result as a fraction of the US per capita performance in each ranking. The above methodology is applied to each edition of the Shanghai exercise. As shown in Table 12.5, the US outperforms European countries in the Top 50 ranking of universities, with the notable exceptions of Switzerland and, to a lesser extent, the UK. However, the overall gap closes as one moves from the Top 50 to the Top 500. Finally, although the Shanghai data are available only since 2003, as displayed in Table 12.6, across time,

22 1

Does the 'European Paradox' still hold?

Tuble 12.5 Country perfi~rmanceindex according to the 2007 S/zang/zai ranking Country

Top 50

Top 100

Top 500

Switzerland

0.946

United Kingdom

0.697

Netherlands France Germany Italy

0.162 0.051 0 0

1.62 0.813 0.487 0.180 0.156 0

2.14 1.18 1.33 0.479 0.683 0.378

EU-17

0.141

0.282

0.720

Note,

US

=

1 1s the reference. Population

1s

calculated In 1999

Sourw: Our computations based on the Shanghai Jiao Tong University Academic Ranking of World Universities (2007).

Table 12.6 EU-17perjormance ucross time Year

Top 50

Top 100

Top 500

2003 2004 2005 2006 2007

0.162 0.156 0.137 0.135 0.141

0.180 0.299 0.279 0.281 0.282

0.686 0.718 0.695 0.7 16 0.720

Notes:

US

=

1 is the reference. Population is calculated in 1999

Source: Our computations based on the Shanghai Jiao Tong University Academic Ranking of World Universities (2003-07).

European universities seem to have lost ground in the top tiers ranking while gaining 'strength' in the Top 100 and Top 500 ones. Similar to what we found on the basis of different bibliometric indicators, the gap is larger (and increasing) when one focuses on the upper tail of the distribution of science performers whereas it tends to narrow when also average performers are considered. Having assessed the relatively poor scientific performance of European universities, the task is to explain why. A first important aspect regards the amount of money spent on higher education. According to OECD data, the US outperforms the EU in expenditure on tertiary education. As displayed in Table 12.7, in 2004, total (public and private) spending on higher education in the EU-19 accounted for barely 1.3 per cent of GDP, against

222

European science and techno log^. policjx

Table 12.7 Expenditure on tertiary education as a percentage of GDP

United States EU-19 average United Kingdom Germany France Italy Spain

Private

Public

Total

1.O 1.1 0.8 1.O 1.2 0.7 0.9

1.9 0.2 0.3 0.1 0.3 0.3 0.3

2.9 1.3 1.1 1.1 1.3 0.9 1.2

Source: OECD (2007)

2.9 per cent in the US. The gap is similar if one observes expenditure per student, with an annual amount of $7192 in the EU-19 vs. $19 842 in the US (OECD, 2007). Although these data do not specify whether those expenditures are allocated to research-oriented or to teaching-oriented institutions, Aghion et al. (2007) show that there is a strong positive correlation between expenditure per student and country performance in the Shanghai ranking. However, money is only part of the story. Another important aspect concerns the institutional differences that one observes across the two sides of the Atlantic. First, although research universities emerged for the first time in mid-19th-century Prussia - with the so-called Humboldt model - universities seem to occupy a less significant position among researchproducing institutions in today's Europe. As highlighted by Table 12.8, in many European countries, with the notable exception of the UK, a relevant portion of top-quality research is performed by non-university institutions. For example, in France, CNRS (National Institute for Scientific Research), INRA (National Institute for Agricultural Research) and the Institut Pasteur are top research performers and, similarly, the majority of the best German researchers is affiliated with the Max Planck Institutes. On the other hand, in the US, after World War 11, also influenced by the guidelines of the Vannevar Bush Report (1945), research universities are seen as the central institutional locus for top research. As shown in Table 12.8, four of the top five institutions today are research universities with Harvard and Stanford as clear leaders. Even if these EU vs. US differences do not explain per se the weaknesses of European universities, they are likely to be an important part of the explanation given the strong complementarities between basic research and (especially graduate) teaching activities. Second, university systems in most European countries are characterized

223

Does the 'European Paradox' still hold.? Tuble 12.8 Top institutions in terms of number oJ'HCRs by country Number of HCRs

Name United Stcite.\

Harvard University Stanford University National Institute of Health University of California, Berkeley Massachusetts Institute of Technology United Kingdom

University of Cambridge University of Oxford Imperial College London University College London King's College London Germany

Max Planck Institute Technische Universitat Miinchen Universitat Wiirzburg Universitat Hamburg Deutsches Krebsforschungszentrum France National Institute for Scientific Research (CNRS) Universite Pierre et Marie Curie National Institute for Agricultural Research (INRA) Institut Pasteur Universite Louis Pasteur Italy

National Institute of Astrophysics (INAF) University of Milan University of Pisa University of Florence Menarini Ricerche S.p.A. Spain National Institute of Scientific Research (CSIC) Universidad Autonoma de Madrid Almirall Prodesfarma Research Center Universidad Politecnica de Valencia Universitat Pompeu Fabra Notes: HCRs are top cited researchers in different fields of science for the period 198 1-99, See http:llisihighlycited.coml for details (accessed 24 May 2009). Researchers highly cited in more than one field have been counted several times. Source:

Our calculations on IS1 Thompson Highly cited scientist data.

224

European science and technology policy

by centralized control and the political inclination among most continental Europe countries has been to resist differentiation between different types of universities. This has prevented a sharp distinction between research-cum-graduate teaching universities, undergraduate liberal art colleges and technical colleges similar to the one that emerged in the US. In fact, in Europe - especially Continental Europe - most universities offer a confused mix of the three. Anecdotal evidence suggests that this is neither good for research nor for mass-level training. Moreover, Aghion et al. (2007) find that university autonomy in terms of budgets, hiring and remuneration increases the efficiency of both public and private educational investment. In this respect, the intra-European harmonization of higher education systems could also provide an opportunity for enhanced competition among European universities provided that it occurs on the ground of 'virtuous' criteria for selection and resource allocation. For example, as suggested by Mas-Cole11 (2003), it would not be acceptable to base the competition for students on the ease of obtaining a degree.*

3 POORER TECHNOLOGICAL PERFORMANCE: R&D INVESTMENTS AND THE LISBON AGENDA Let us now turn to the assessment of European strengths and weaknesses in terms of innovative output and underlying investments in research and development. The so-called Lisbon Agenda, which according to the pompous statement of the Lisbon European Council aims at making the EU 'the most dynamic and competitive knowledge-based economy in the world', fixed two important targets with respect to these measures: first, the ratio of EU R&D expenditure to GDP was supposed to reach the ambitious target of 3 per cent of GDP by 2010. Second, the share funded by business ought to rise to around two-thirds (European Commission, 2002). Europe is likely to miss those targets by far. In fact, to begin with, Figure 12.1 shows that at the aggregate level, the EU under-invests in R&D with respect to both the US and Japan; second, the gap is not shrinking and, third, it is, of course, very unlikely that the 3 per cent target will be reached in the next two years. Likewise, as shown in Figure 12.2, the share of R&D financed by the business sector is only increasing very slowly. In any case, more informative international comparisons of R&D intensities should consider the industrial specialization of each country given the sectoral specificities of R&D incentives (Dosi, 1988; Mathieu

225

Does the 'European Paradox' still hold?

1990

2000

Year Note: Source:

The lines depict gross expenditure on R&D as a percentage of GDP OECD Main Science and Technology Indicators.

Figure 12.1 R&D intensity on GDP ...................... EU-15

Japan

Japan

,'

Year Note: Source:

Ratios of industry-financed R&D on total R&D. OECD Main Science and Technology Indicators.

Figure 12.2 Share o f industry-financed R&D

'..-.--.,' US

226

European science and technology polic)~

Table 12.9 Decomposing 2001 governmenr funded R&D

Country

BERD Total expenditures

Vote: Source:

on GDP (YO)

Non-BERD Total expenditures

on GDP (%)

Gross expenditures are expressed in million 2000 dollars - constant prices and PPP. Our calculations based on OECD (2008).

and van Pottelsberghe de la Potterie, 2008). In fact, first, specialization explains a good deal of the difference in R&D intensities across countries. Second, when one controls for specialization patterns, only a few countries (that is, Sweden and the US) unequivocally show remarkably higher R&D intensities. Note that as a consequence of the sectoral specificity of the propensity to invest in R&D, the 'Lisbon Targets' are quite arbitrary and in any case cannot be reached without quite sweeping industrial policies aimed at radically changing the European industrial structure. Granted the aforementioned sectoral specificities, how do governments influence business-performed R&D? And how do they allocate their research funds in the first place? To begin with, it can be useful to compare how countries spend their publicly funded R&D. The latter may be decomposed into several categories. As shown in Table 12.9, compared with EU governments, the US government spends more on both research and development carried out by firms (business expenditure R&D [BERD]) and other forms of R&D. The bulk of the difference, however, is in the former. Three broad categories of public support for industrial technology can be identified: first, those programmes designed to encourage industrial firms to carry out R&D by reducing its costs through grants, loans and fiscal measures; second, government payments to industrial firms financing R&D as part of procurement programmes, for example, as part of defence or space objectives and, third, public support to 'research infrastructures' specifically aimed at industrial development not involving, however, any financial transfer to private firms (for example, applied research undertaken in public institutes and universities). To be sure, publicly funded BERD underestimates the full amount of public support for industrial technology: it does not include fiscal incentives, some categories

Does the 'European Puradox' still hold?

227

of loans and also R&D carried out in other sectors, which governments finance with the aim of supporting industry. Unfortunately, international statistics on the above are hardly available, even for industrialized countries. However, Young (2001), exploiting the data from a pilot study run by OECD, finds not surprisingly that the pattern of support varies considerably across countries. In particular, US federal support for industrial technology is almost entirely paid to firms (public institutes and universities do not seem to receive public funds for industrial technology!), the largest share being in the form of missionoriented contracts and procurement. As far as EU countries are concerned, in France and the UK, mission-oriented contracts are also relatively important while in Germany and the Netherlands, funds are distributed evenly across the three categories. What are the plausible links between public policies and privately financed R&D - indeed one of the 'Lisbon objectives' of the EU? An implication of the 'paradox diagnostics' is that an effective channel to foster private investment in R&D is increasing university-industry links. If this were correct, however, one would observe that those countries where university-industry links are stronger are also the ones where the industryfinanced R&D intensity is higher. Taking the share of higher education expenditures on R&D (HERD) financed by the business sector as a proxy of the strength of university-industry links, Figure 12.3 shows that there is no cross-country correlation in line with the 'paradox conjecture'. In the opposite direction, several authors have argued that private investment in R&D and its localization are likely to be stimulated by the quality and size of academic research (Mansfield, 1995; Cohen et al., 2002; Abramovsky et al., 2007). Possibly, the single most important factor behind the importance of academic research is the supply of a qualified and skilled labour force. However, survey evidence suggests that scientific output dissemination public conferences and meetings, informal information exchange and consulting - may also be important. As shown in Figures 12.4 and 12.5, cross-country comparisons reveal that industry-financed R&D is indeed positively associated with both the per capita number of highly cited researchers and expenditure on higher education R&D. Of course, such correlations do not control for many country-specific characteristics likely to be associated with both measures (including the sectoral specialization of industry, discussed above) and, therefore, we would not dare interpret such correlations in causal terms. However, to check for their robustness, we first verify whether they also hold within sectors. Second, we exploit the longitudinal dimension of R&D expenditure data to control for those unobserved country characteristics that are constant over time. -

European science and technologj~policy 0 Israel

0 un~tedStates

0 Germany

'wknada o ~ ~ ? a ! ! $Norway ' nl P ed Kingdom

(f@wal

Netherlands

0 Turkey

Poland Greece 0810vak &publfc

Share of HERD financed bv ~nduslrv

.Vote: Source:

R&D data refer to 2007 or most recently available year. OECD Main Science and Technology Indicators.

Figure 12.3 Industry-financed R&D and share of HERDfinanced by industry

Unlted Stales

0

to

5

15

HCRs per caplta

Notes: R&D data refer to 2007 or most recently available year. Data on HCRs are drawn from www.ISIHighlyCited.com.

Source: OECD Main Science and Technology Indicators and IS1 Thompson Highly cited Scientist data.

Figure 12.4 Industry-financed R&D and Highly Cited Researchers per capita

Does the 'European Paradox' still hold:' Israel

.

Japan

.Sweden

Korea

Finland

.United States Germany

OECD Total

Denmark

0 Luxembourg

Austria

China

0

0.2

lreland

Netherlands Unlted K~ngdam Noway

0.4

0.6

.Canada

0.8

HERD as per cent of GDP

Note: Souuce:

R&D data refer to 2007 or most recently available year. OECD Main Science and Technology Indicators.

Figure 12.5 Industry-jinanced R&D and H E R D Figures 12.6, 12.7, 12.8, 12.9 and 12.10 reveal that in a number of industrial sectors, R&D intensity is positively correlated with the quality of academic research in selected related fields.' Consistent with the findings of Abramovsky et al. (2007) and Cohen et al. (2002), the strongest correlation is in pharmaceuticals. However, in other industries as well, those countries with the highest per capita number of highly cited scientists in relevant fields display the highest R&D intensities. As far as the correlation between business R&D and HERD is concerned, we calculate for each country, the differences between the first and the last year for which the two variables are available.'" This allows for removing all correlation due to unobserved country-specific time-fixed effects. As shown in Figure 12.11, the positive correlation still holds: those countries that increased their effort more concerning HERD to GDP also performed better in terms of the variation across time of business-funded R&D intensity.

European science and technology policy

230

HCRs per caplta In blalogy,chem~stryand medical flelds

.Votes: R&D data refer to 2007 or most recently available year. Data on HCRs are drawn tiom www.HighlyCited.com.

Source: OECD Main Science and Technology Indicators and IS1 Thompson Highly cited Scientist data.

Figure 12.6

Industry-financed R&D in pharmaceuticals and HCRs per capita in selectedJields

1

0 Japan

0 Germany

0

France

* Unlted States weden FlnlanbS 0 United Kingdom 0 Netherlands

0 Australia

Canada

0

0.5

1

1.5

2

HCRs per caplta in chem~stryand material sciences

Notes: R&D data refer to 2007 or most recently available year. Data on HCRs are drawn from www.HighlyCited.com.

Source: OECD Main Science and Technology Indicators and IS1 Thompson Highly cited Scientist data.

Figure 12.7 Industry-financed R&D in chemicals and HCRs per capita in selectedfields

Does the 'EuropeanParadox' still hold.7

20j 0

.'

.Canada

Norway

.Belgium Ftnland

Germany

e Italy

s

Unlted Klngdom *. Australla fenmark

.Netherlands 1

0

2

3

HCRs per caplta ~nenglneerlng computer and materlal scwnces

Notes:

R&D data refer to 2007 or most recently available year. Data on HCRs are drawn

from www.HighlyCited.com. Source: OECD Main Science and Technology Indicators and IS1 Thompson Highly cited Scientist data.

Figure 12.8 Industry-$nanced R&D in radio, TV and electrical equipment and H C R s per capita in selected fie1d.s

e France. Finland

0

1

Spaln

0

Australla 0 Canada

*Italy

HCRs per caplta In engmeerlng and rnaterlals science

Notes: R&D data refer to 2007 or most recently available year. Data on HCRs are drawn from www.HighlyCited.com. Source. OECD Main Science and Technology Indicators and IS1 Thompson Highly cited Scientist data.

Figure 12.9 Industry-financed R&D in machinery and equipment and HCRS per capita in selectedfields

European science and technology policy

132

i

Sweden

Norway Germany

P

1

Japan Un~tedStates France

i/ '""

Australia

Netherlands

I eSDain

J

'

Finland '*lg'um

Canada

HCRs per capita in engineering and materials science

Nores: R&D data refer to 2007 or most recently available year. Data on HCRs are drawn from www.HighlyCited.com. Source: OECD Main Science and Technology Indicators and IS1 Thompson Highly cited Scientist data.

Figure 12.10 Industry-$named R&D in motor vehicles and HCRs per capita in selectedfields

4 WRONG DIAGNOSES AND MISGUIDED POLICIES: SOME MODEST ALTERNATIVE PROPOSALS The European picture indeed shows worrying signs of weakness with respect to the generation of both scientific knowledge and technological innovation. However, no overall 'European paradox' with a lead in science but weak 'downstream' links is observed. On the contrary, significant weaknesses reside precisely 'upstream', that is, the European system of scientific research lags behind the US in several areas. In turn, such a picture calls for strong science and higher education policies. However, this is almost the opposite of what has happened. The belief in a purported paradox together with the emphasis on 'usefulness' of research has led to a package of policies whereby EU support to basic research and research universities is still very scarce. As Keith Pavitt put it a few years ago:

Does the 'European Pararlo.~'still hold.?

"

233

* Japan

6

1

Slovak Republic

-1

-0 1

0

0.1

0.2

0.3

0.4

Varlatlon HERD on GDP. percent

no re.^: R&D data refer to the 1981-2007 period. For each country the variation is calculated according to data availability.

Source:

OECD Main Science and Technology Indicators

Figure 12.11

Variation in industry-financed R&D and vnviution in HERD

[rlesearch proposals are expected to identify possible practical as well as scientific benefits; higher priority is being given to user involvement (including partial funding). universities are being invited to extract more revenue from licensing their intellectual property, and substantial public funds have been spent on 'foresight' exercises designed to create exchange and consensus around future opportunities of applications. (Pavitt, 2001, p. 768) The 'Framework Programmes' have all been conceived with such a philosophy in mind, which in the most recent one is pushed to the extreme with the 'Networks of Excellence': they not only do not support research, they explicitly prohibit the use of EU money for that purpose (for example, the spirit of the Network of Excellence, sponsored by the European Union Framework Programme). Similarly, with regard to industrial R&D, the focus on 'pre-competitive' research has meant the emergence of a sort of limbo wherein firms often in combination with academics - try to tap Community money in areas that are marginal enough not to justify the investment of their own funds. Moreover, the networking frenzy has gone hand in hand with growth in the number and power of research bureaucrats (at both European and national levels) and academic intermediaries whose main competence is -

134

European science and technology p o l i q

precisely in 'networking', 'steering', writing lengthy reports and demandIng researchers do the same. If our diagnosis is correct, this state of affairs IS bad for research, wasteful for society and also bad for business. Given this state of affairs, what can be done? Let us conclude with some policy implications of the foregoing analysis, which we have partly outlined in Dosi et al. (2006) as well. First, increase support to high-quality basic science, through agile mstitutions much like the US National Science Foundation (NSF) relying on world-class peer review (and also physically located far away from Brussels - as May, 2004 suggested). In that regard, the recent constitution of a 'European Science Council' is a very welcome development. Second, fully acknowledge the difference within the higher education system between research-cum-graduate teaching universities and other forms of tertiary education discussed above. European higher education mstitutions offer a confused blend that is neither good for research nor for mass-level training. The well-placed emphasis on the role of the first type of institutions often comes under the heading of 'Humboldt model' as pioneered by Germany more than a century ago. However, nowadays the practice is mostly American while in Europe (especially Continental Europe), one observes that top-quality research is carried out outside universities. Third, build ambitious, technologically daring missions justifiable for their intrinsic social and political value. Our data show that Scandinavian countries and Switzerland are able to mobilize considerable resources for high-quality basic and applied research even without the massive defence and health expenditures of the US. Big European countries and the European Union itself have indeed much to learn from them. At the same time, one should not overlook the importance of large-scale far-reaching European programmes with ambitious and technologically challenging objectives in the fields of for example, energy conservation, health care and environmental protection. Finally, we believe that an increased rate of technological innovation in Europe will not be achieved without explicit industrial policies aimed at strengthening European participation in production based on new technological paradigms in ICT and life sciences. But this goes well beyond the diagnostics and possible remedies of the European weaknesses so to speak, 'upstream', in pure and applied research. We certainly do not believe that even good science and technology policies can alone redress the European economic performance. However, we trust they are part of a broader policy package involving also macroeconomic policies and institutional changes, which, of course, are well beyond the scope of the present chapter.

Does the 'European Paradox' still hold?

235

NOTES As reported in European Commission (2007), the EU is still the world's largest producer of scientific output as measured by its share of the world total of peer-reviewed scientific articles in 2004. It represents 38 per cent scientific output versus the 33 per cent of the US. This figure is based on the assumption of equal research production of non-academic researchers. Their names, afiliations and countries of residence are freely available online at www. ISIHighlyCited.com (accessed 21 May 2009). The same data set has also been analysed by Bauwens et al. (2007). The small differences with our elaborations are due to the fact that we counted top researchers ranked in multiple fields several times. CERN (European Organization for Nuclear Research) highly cited researchers are 8 (about 7 per cent of the total) and therefore do not fundamentally drive the outstanding Swiss performance. As mentioned above, data on HCRs are also likely to reflect a smaller number of research outputs compared with the top 1 per cent cited articles. More details on scoring procedures are available at http://ed.sjtu.edu.cn/ranking.htm (accessed 21 May 2009). See also the evidence reported by Bagues et al. (2008) concerning the Italian case. To define the relevance of different scientific disciplines for any industrial sector we make use of the 1994 Carnegie Mellon Survey, which displays for each sector the importance of several science fields: biology; chemistry; physics; computer science: materials science; medical and health science; chemical engineering; engineering and mathematics. A discipline is chosen to be relevant for a given industrial sector if it is rated moderately or very important. Results are confirmed by panel data analysis. Tables are available from the authors upon request.

REFERENCES Abramovsky, L., R. Harrison a n d H. Simpson (2007). 'University Research a n d the Location of Business R&D', Economic Journal, 117 (519), 1 14-41. Aghion, P., M . Dewatripont, C . Hoxby, A . Mas-Cole11 a n d A. Sapir (2007), Wh,v Rtlfornz Europe's Universities?, Bruegel Policy Brief N o . 4, Brussels: Bruegel Foundation. Bagues, M., M . Sylos Labini a n d N. Zinovyeva (2008), 'Differential Grading Standards a n d University Funding: Evidence f r o m Italy', CESifi, Econonzic Studies, 54 (2), 149-76. Bauwens, L., G . Mion a n d J. Thisse (2007), 'The Resistible Decline of European Science', C O R E Discussion Paper N o . 2007192, Louvain-la-Neuve: T h e Center for Operations Research a n d Econometrics, Universite Catholique d e Louvain. Bush, V. (1945), Science; The Endless Fronfier, Washington, D C : Government Printing Office. Cohen, W.M., R . R . Nelson a n d J.P. Walsh (2002) 'Links a n d Impacts: T h e Influence of Public Research o n Industrial R&D', Management Science, 48 (1 ), 1-23. Dosi, G . (1988) 'Sources, Procedures, a n d Microeconomic Effects of Innovation', Journal oj'Econotnic Literature, 26 (3), I 120-71. Dosi, G., P. Llerena a n d M . Sylos Labini (2006), 'The Relationships Between

European science and technology policy

236

Science, Technologies and their Industrial Exploitation: An Illustration Through the Myths and Realities of the So-called "European Paradox"'. Research Policy, 35 (lo), 1450-64. European Commission (1995), Green Paper on Innovation, COM (1995) 688, Communication from the Commission, 20 December 1995, Brussels. European Commission (2002), The Lisbon Strategy Making Change Happen, COM (2002)14, Communication from the Commission, 15 January 2002, Brussels. European Commission (2007), Key Figures 2007. Towards a European Reseurch Area Science, Technology and Innovation, Luxembourg: Office for Publications of the European Communities. King, D.A. (2004), 'The Scientific Impact of Nations', Nature, 430 (6997). 311-16. Mansfield, E. (1995), 'Academic Research Underlying Industrial Innovations: Sources, Characteristics, and Financing', The Reviebt,of Economics and Statistics, 77 (I), 55-65. Mas-Colell, A. (2003), 'The European Space of Higher Education: Incentive and Governance Issues', Rivista di Politica Economica, 93 (1 1-12), 9-27. Mathieu, A. and B. van Pottelsberghe de la Potterie (2008), 'A Note on Drivers of R&D Intensity', CEPR Discussion Paper No. 6684, London: Centre for Economic Policy Research. May, R.M. (2004), 'Raising Europe's Game', Nature, 430 (7002), 831-2. OECD (2007), Education at a Glance, Paris: OECD. OECD (2008), Main Science and Technology Indicators, Paris: OECD. Pavitt, K. (2001), 'Public Policies to Support Basic Research: What Can the Rest of the World Learn from US Theory and Practice? (And What They Should not Learn)', Industrial and Corporate Change, 10 (3), 761-79. Young, A. (2001), 'Improving Measures of Government Support to Industrial Technology', Science Technology Industry Review, 27 (Special Issue), 147-83. -

-

13. The European Research Area and human resources in science and technology Wendy Hansen For the times they are a-changin'. (Bob Dylan)

1 GLOBALIZATION AND HUMAN RESOURCES IN S&T Human capital has a powerful and critical role in the knowledge-based economy. Highly skilled scientists, engineers and researchers drive job creation through innovation and support economic change through the flexibility of skill sets and inter-occupation, sector and country mobility. The exploration of the links between human capital and performance suggests not only that PhDs in science and engineering ( M E ) contribute to technological performance but also that variations in PhD 'strength' (of numbers) seem to be more important than variations in research and development (R&D) expenditures in high- and medium-high tech manufacturing industries.' In today's global marketplace, enabled by technology advances and especially in the case of information technologies, multinationals have the world to choose from for R&D activities. 'Firms now view parts of the developing world as key sources not only of cheap labour, but also of growth, skills and even new technologies.'? Labour costs are an important factor for business planners but the availability of a skilled scientific, technical and engineering workforce is a critical element in a location decision strategy that also looks at finance, the regulatory framework and market access. Western knowledge-based economies face growing competition for job creation as countries like China and India evolve from the role of providing low-paid, low-skilled workers for manufacturing and services off-shoring to that of providing a physical environment for innovation and R&D, an environment replete with highly skilled scientists, engineers, researchers and technically trained workers. Non-OECD countries like China and India are challenging developed countries' hold on high skill high value jobs. 'China in particular is moving up the value chain and thus seems to compete directly with OECD countries'.' China and India are developing a science and technology (S&T) workforce with the knowledge and skills to generate new knowledge and

238

Europeun .science und technology policj,

Table 13.1 Top ten prospective locations for future R&D, 200549 Per cent of responses 1. 2. 3. 4. 5.

6. 7. 8. 9.

10.

China US India Japan UK Russian Federation France Germany The Netherlands Canada Singapore Taiwan Province of China Belgium Italy Malaysia Republic of Korea Thailand

60.9 40.6 29.0 14.5 13.0 10.1 8.7 5.8 4.3 4.3 4.3 4.3 2.9 2.9 2.9 2.9 2.9

Nore: Several countries rank together at position 9 and position 10 Source: Occasional note, UNCTAD survey on the internationalization of R&D. current patterns and prospects on the internationalization of R&D, UNCTAD/WEB/ITEI IIA/2005/12, 12 December 2005.

drive innovation. India is considered one of the top desirable locations for R&D for two reasons: the S&T work force in India is made up of, in part. graduates from the West who bring back not only S&T skills but also entrepreneur skills; and, India has access to supplies of S&T graduates from top universities across Asia. As the ranking in Table 13.1 shows, countries in developing Asia are most attractive for foreign R&D along with the US. The UK, France and Germany ranked among the top ten preferences albeit with considerably less popularity. Changes and opportunities brought about by new technology and the globalization of manufacturing, business services, education and research underscore the need to ensure that a sufficient supply of high calibre human resources in S&T (HRST) remains a priority for policy and decision-makers in Europe. A highly skilled and innovative S&E workforce will help ensure that Europe's economy is competitive as countries vie for financial and human capital for R&D and knowledge-based product development.

The Europeun Reseurch Arru und human resources

239

The challenge for European policy and decision-makers is to develop a strategy to ensure an adequate supply of HRST for the European Research Area (ERA) as foreign influences exert pressures on domestic supplies hitherto unseen brought about by the internationalization of R&D and HRST mobility.

2

THE UNIVERSITY PIPELINE IN THE GLOBAL

LANDSCAPE Countries like Australia, Canada and the US have always relied upon foreign students to boost participation in S&E programmes and S&E degree output. Overall, universities in the EU have success in attracting foreign students albeit mainly from within EU Member States. Estimates are that in 2006, there were 458 000 EU-27 students studying abroad across the EU-27, European Economic Area (EEA) or candidate countries with EU-15 countries accounting for two-thirds of them. Germans were the most 'mobile', accounting for 12.6 per cent of visiting students. French students were second (10.3 per cent studying elsewhere in the EU) followed by Greece (8.1 per cent), Italy (7.6 per cent) and Poland (7.4 per cent) (Table 13.2). A number of countries (marked by * in Table 13.2) show a declining interest in international studies, Greece being the most notable. The U K enjoys particular success in attracting foreign students. From 1995 to 2006, UK universities saw foreign student S&E undergraduate enrolment grow from some 34 500 students (8.8 per cent share of total undergraduate S&E enrolment) to almost 56 000 (10.4 per cent share of total undergraduate S&E enrolment). The number of foreign graduatelevel students in S&E rose from almost 29 000 to more than 67 600, which brought their share from 28.9 per cent of S&E graduate enrolment to 43.3 per cent.? Lack of university spaces and growth in high skill industries are two factors driving Chinese students overseas. It is estimated that more than 350 000 Mainland Chinese students are studying for degrees at foreign universities and this figure is expected to surpass 600 000 within two decade^.^ Chinese students are driving growth in S&E enrolment in countries in Europe and the US. In Japan in 2005, Chinese students accounted for 72.9 per cent of the foreign students in S&E fields at the undergraduate level and 54.6 per cent at the graduate level. In 1995, there were fewer than 1500 Chinese students in S&E programmes in the UK, a decade later there were almost 20 000, with more than half of them at the graduate level of study. This means that within the foreign student contingent in S&E at the graduate level, Chinese students' share rose from 4.6 per cent

240

European science and technology policj

Table 13.2 Estimates of students studying abroad in Europe (country as share in EU-27, EEA and candidate countries). 2000 and 2006 (per cent)

AT* BE* BG

CY

cz DE DK* EE EL* ES* FR* HU* IE IT* LA LT LU* MT NL* PL PT RO SE* SI* SK UK* Source: Based on Eurostat data table: 'Mobility of students in Europe', accessed 25 November, 2008, at http://epp. eurostat.ec.europa.eu

to 14.9 per cent. At the graduate level, students from China (and India) favour graduate studies in engineering compared with foreign students overall. More and better challenging career opportunities in Europe can help ensure ERA can compete against countries like the US to augment the domestic HRST supplies by bringing foreign students to Europe to study and work. An evaluation of US admission policy looked at ten countries, including four EU countries, and developed a ranking based on key policy

The European Research Area and human resources

24 1

elements including employment opportunities and retention after graduation. France, Germany and Austria ranked second, third and fourth in terms of restrictive policies (Switzerland ranked first) and above the ten-country average. Sweden ranked below the average and below the US posltion and the UK was the least restrictive." An administrative and regulatory environment free of impediments for foreign student enrolment and integration once they graduate is an important consideration to minimize barriers to international mobility of HRST. A second option to expand ERA'S HRST pool is to have EU citizens pursue studies at top universities around the world. When they return home, they bring to ERA not only their world-class degrees in S&E but also their networks and the international experience they gained during their overseas studies. Of course, the key here is that ERA has to be in a position to draw Europeans back to ERA to pursue careers in S&E once they have been exposed to international opportunities in countries like Australia and the US. In 2005, foreign students accounted for 14.5 per cent of the 329 600 or so S&E undergraduate students in Australia. In programmes in mathematics and computer science, the representation of foreign students was as high as 39.7 per cent. At the graduate level, the foreign student share stood at 31.9 per cent of the 101 644 students enrolled in S&E and 41.6 per cent of them were enrolled in mathematics and computer sciences.' Europeans may go to Australia to study but they overwhelmingly choose programmes in management and commerce over S&E. In 2007, more than one-quarter of the 3259 German students and 1140 UK students, one third of the 1052 Swedish students and half of the 1223 students from France were in management and commerce programmes. In 2006, some 12 200 EU students were enrolled in Canadian universities with three European countries (France, Germany, UK) ranked among the top ten source countries accounting for 61.6 per cent of those from the EU. At least two in five Europeans were in S&E studies, although just like in the case of Australia, Europeans studying in Canada are less inclined to pursue studies in S&E compared with other foreign students. In 2007, there were almost 266 000 foreign students enrolled in graduate programmes in the US, more than half of them (53.3 per cent) in S&E fields of study. Students from India and China show greater interest in pursuing degrees in S&T compared with other foreign contingents. In 2007, among the graduate level foreign students, 70.2 per cent from India and 66.7 per cent from China were in S&E programmes. Among Romanian students in graduate programmes, 58.5 per cent were in S&E and lower shares were reported for other European countries such as Germany (45.5 per cent) and France (43.8 per cent) and the UK (37.9 per cent).'

242

European science und rechnologj~policy

Table 13.3 Field ofstudy of EU-27 US doctorate recipients, 2006 Per cent Phys scs Humanities soc scs Life scs Eng Ed Other Based on NSFINIHIUSEDINEHIUSDAINASA, 2006 Survey of Earned Doctorates, September 2008

Source:

Europe remains a world leader in generating university degrees. In 2004, EU-15 universities granted more than 55 000 doctorates in S&E compared with 34300 or so granted in Asia and the almost 26 300 awarded in the US. In fact, France, Germany and the UK taken together (26929 awarded in 2004) surpass the US figure. An additional 26000 or so doctoral degrees in S&E were awarded across countries in the European Free Trade Area (EFTA) and Central and Eastern Europe. In 1999, China awarded 6778 doctorates in S&E; just four years later the figure was close to doubling at 12238. This compares with growth in India's S&E doctorates of 14.8 per cent, a 7.0 per cent increase in Japan and a 15.6 per cent rise in the EU-15. The US suffered a slight decline over this time period but the numbers are since recovering. A total of 1243 of the 45600 or so US doctorate recipients in 2006 reported a bachelor degree from a university in the EU-27 (about three quarters from the EU-15 and one quarter from the EU-12).9 Most of the US doctorates earned by persons from the EU-27 were in physical sciences that include mathematics and computer science, skills and learning for high growth occupations (Table 13.3). Among the US doctorates earned by EU-27 students, more than two in five were earned by women, with women from the EU-12 showing a higher share compared with their counterparts in EU-15 countries. Three-quarters of EU-27 US doctorate recipients were in the US studying under temporary visas while one-quarter of them had already obtained permanent visas. The US has the capability to keep a large share of foreign-born nationals that earn their doctorate in US universities. At the same time, few US-born doctorate graduates are tempted to pursue their careers outside of the US. Europeans earned almost 6600 US doctorates in S&E between 2002 and 2005 and three-quarters of them had plans to stay in the US

The European Research Areu and humun uesourcrLs Table 13.4

243

Share of Europeun US doctorate recipients w,ith$rm plans to stuy in the US. 2002 to 2005

Per cent UK

Germany Italy Greece France Spain Source.

Based on NSF, Science and Engineering Indicators (2008). Appendix Tablc 2-33

after graduation and more than one-half reported definite plans to remain in the US after graduation." It was students from the UK and Germany that reported the largest shares with firm plans to remain in the US following graduation, and there were also significant shares from Greece (46.6 per cent), France (46.0 per cent) and Spain (42.6 per cent) (Table 13.4). A somewhat ominous signal for ERA is not only that large shares of Europeans who earn their doctorates in the US plan to stay in the US but also that the share reporting intentions to stay in the US seems to be on the rise. Location plans following graduation vary by field of expertise, career opportunities and research conditions at home and abroad. Among the 480 UK doctorate recipients in the US between 2002 and 2005, physical scientists reported the highest share with definite plans to remain in the US, whereas among the 389 French US doctorate recipients, it was people from mathematics and computer sciences. Among the 750 Germans and the 494 Italians, it was persons from social and behavioural sciences that showed the highest share with definite plans to stay in the US whereas among the 258 US doctorate recipients from Spain, it was agriculture and biological sciences that reported the highest share with firm plans to remain in the US (Table 13.5).

3 TEACHERS FOR THE NEXT HRST GENERATION University teachers instruct the next generation of scientists and engineers; they conduct leading-edge research. In the U K in academic year 2006-07, 63.8 per cent of full-time employed academics had teaching and research

244

European science and rechnolog~policj

Table 13.5 Share of European US doctorate recipients ~ i i t h j r r nplans to stay in the US by$eld of degree, 200245 Phys Scs

AgIBio

MathIComp

Eng

Social Sc Per cent

France Germany Greece Italy Spain UK Source:

46.2 51.9 42.1 42.1 35.4 63.3

53.2 53.7 48.6 44.4 49.1 56.2

57.1 50.9 51.6 53.7 42.1 57.1

36.8 42.9 44.9 51.0 39.3 48.1

46.0 55.6 45.9 54.9 45.1 60.2

Based on NSF, Science and Engineering Indicators 2008, Appendix Table 2-33

responsibilities and 27.9 per cent reported being engaged in research only with no teaching duties." In the US, 47.2 per cent of the full-time professional staff in degree-granting institutions was engaged in teaching and research in 2005, and when part-time professional staff is included, the share carrying out research and teaching rises to 52.4 per cent. In 2004-05,34.0 per cent of the core academic staff in the UK was aged 50 and over. In Canada, the share of full-time faculty aged 50 and over was even higher at 48.4 per cent with men showing a considerably higher share aged 50 and over than women (51.9 per cent compared with 41.1 per cent, respectively). Estimates are that academic recruitment needs in England could rise by as much as 25 per cent over the time period 2004 to 201 1.12 The ageing of the faculty presents different challenges for different departments. In the UK, 44 per cent of the academics in physics and engineering were aged 50 and over and in Canada, 42 per cent in mathematicslphysical sciences were aged 50 and over. The greying of faculty ranks is adding pressure on the supply of doctorate graduates and has faculties in competition with the private sector for these scarce resources. This includes competing for PhD graduates as well as recruiting seasoned instructors Demand for university faculty in key S&E fields such as computers and physical sciences continues to grow and just as business enterprises turn to foreign sources to fill S&E labour ranks, universities too turn to foreign supply for their faculty. In 2005-06, 13 per cent (41 16) of the 31 575 fulltime faculties at Canadian universities reported earning hislher doctorate in the EU-27. A higher share of EU-27 men had earned their doctorate in the EU prior to coming to Canada to teach than women. Reliance upon Europe for university faculty varies by department. In engineering1

The European Reseurch Areu and hurnun re.sources

Tuble 13.6

245

Full-time faculty in Canadian universities with u doctorate turned in EU-27, srlectedjields, 2005-46 Per cent

All fields

Eng/app scs Eng Health

Mathlphys Comp scs Chemistry Geology Physics Source:

13.6

Based on faculty data of Statistics Canada. special tabulations, 2008

applied science faculties, 15.1 per cent of faculty reported an EU doctorate and 16.2 per cent among faculty in mathematics/physical sciences (Table 13.6). Sources of EU-27 doctorate-qualified faculty fa11 to two countries: together the UK and France account for three-quarters of EU-27-born faculty. When asked about country of previous employment, about half of the faculty from the EU-27 reported a country in the EU as country of previous employment, suggesting that Canadian universities are recruiting seasoned faculty from the EU to replenish their academic ranks as well as recently PhD-qualified graduates from European universities. The UK has a successful history of attracting academics from overseas. Recent figures show that non-UK nationals account for 14 per cent of core academic staff compared with only 8 per cent or so in the mid-1990s. Most of the UK's foreign academics come from other countries in Europe, although some one in five come from the US, Canada, Australia or New Zealand. Two factors are credited with the UK's ability to attract academics from overseas including the UK's strong and successful research base and the use of the English language, a user-friendly language of science." In the US, reliance upon foreign-born faculty is also on the rise. In 2003, 10.7 per cent of the full-time doctoral instructional faculty were non-US citizens and another 13.3 per cent were foreign-born with US citizenship. A decade earlier, only 8.3 per cent were non-citizens and 9.2 per cent were foreign born with US citizenship. The figures for fields in S&E fields are even higher. In 2003, 12.7 per cent of the doctoral faculty were noncitizens and 13.3 foreign born with US citizenship. In research institutions,

246

European science and technology policy

16.4 per cent of the S&E doctoral faculties were non-citizens. Dependence on foreign-born doctoral faculty varies by field of specialization. Among the doctoral instructional faculty at research institutions, more than onequarter in physical sciences/mathematics/computer sciencelengineering was non-citizen foreign born and another one in five were foreign born but with US citizenship obtained. The US Bureau of Labor predicts that college/university teaching ranks will rise from 1.2 million in 2006 to some 1.5 million in 2016. The number of full-time faculty in the US (instruction and research) alone increased by more than one-fifth between 1995 and 2005, bringing the total full-time instruction and research staff to 675 624.

4 THE GLOBAL GREYING OF HRST AND

IMPLICATIONS FOR HRST OCCUPATIONS Across all occupations, across all sectors and around the world, baby boomers are soon (if not already) readying for retirement. In the US, 44 per cent of the HRST with doctorate-level skills are aged 50 and over. On average, 39.6 per cent of HRST in the EU are aged 45-64 as of 2006. This varies by country, with Bulgaria showing the highest share in this age cohort and Ireland and Spain reporting some of the smaller shares. In ten countries in the EU, at least two in five scientists and engineers are aged 45-64 (Figure 13.1). Employment in HRST occupations has been outpacing employment growth overall across OECD countries.14 Recent projections for occupations in S&E in the US show significant growth is expected in the next decade or so. In occupations of network systems and data communication analysts alone, an increase of 53.4 per cent is suggested, meaning an additional 140 000 job openings due to growth between 2006 and 2016. Then there are high-growth occupation estimates that show a 44.6 per cent increase for computer software engineers (226 000 jobs), 25.4 per cent for environmental engineers (14 000 jobs) and 21.1 per cent for biomedical engineers (3000 jobs) (Table 13.7). Countries in Europe have to compete for scarce HRST resources with countries within the EU like the UK that successfully recruits from other E U countries and non-EU countries like the US, Canada, Australia. When HRST growth needs cannot be met by domestic supply, the ability to attract foreign-born HRST is key. Countries like the US, Canada and Australia have proven track records in attracting highly skilled European HRST whereas Europe has struggled to attract and retain highly skilled researchers from outside of EU borders.

The European Research Area and human resources

247

BG, CZ, DK, DE, EE, IT, LV, LT, HU, RO

40% or more scientis engineers aged 45-64

BE, IE, EL, ES, FR, CY, LU, NL, AT, PL, PT, SI, SK, FI, SE, UK

engineers aged 45-64

average 38.1%

Source:

Based on table, Statistics in Focus, Science and Technology, 2612008

Figure 13.1 Age groupings of' scientists and engineers in the EU, 2006 Table 13.7 Projected grocllth of' selected HRST occupations, US. 2006 to 2016 Per cent Network systems and data communications analysts Computer software engineers Computer systems analysts Computer systems software engineeers Environmental engineers Biomedical engineers Industrial engineers Civil engineers Computer and information systems managers Life scientists

53.4 44.6 29.0 28.2 25.4 21.1 20.3 18.0 16.4 15.3

Source: Based on figures of Bureau of Labor Statistics, 2006-16 matrix data.

I n 2006, t h e UK, F r a n c e a n d G e r m a n y provided 15.1 per cent of highly skilled temporary workers t o C a n a d a , o r s o m e 26 000 o f t h e total skilled temporary workers. In 2007, a m o n g the almost 462 000 skilled workers t h a t entered t h e US u n d e r a n H-1B permit (used f o r recruiting speciality skill workers), a r o u n d 8300 (some 6 per cent) were f r o m t h e EU.

248

European science and technology policy

Among those from the EU, taken together France, Germany and the UK accounted for three in five of them. Multinational enterprises have the ability to move skilled HRST about the globe. Intra-company transfers to the US reveal the volume of highly skilled Europeans, managers and other specialists moving in response to multinational needs. Among the more than 363 500 going to the US on intra-company transfers, almost 21 000 were from the EU (about 14 per cent). France, Germany and the UK accounted for two-thirds of those from the EUI5 (Table 13.8). Internationalization of R&D and vying for financial and human capital at the global level have companies solving domestic HRST challenges by ofr-shoring to Asia. It is not just about cost savings but also about access to a highly skilled labour force: 'the following motives for off-shoring: experience accumulated through off-shoring of manufacturing and service activity, the need for a shorter R&D cycle, the increased efficiency and effectiveness of R&D activity, and the growing R&D talent in certain areas'.16 Globalization brings an urgent and sharper focus on human capital in S&E. HRST is a key resource and an important and highly desirable commodity trading at the global level. The extent to which R&D is carried out internationally varies by industry. Among the world's largest R&D investors surveyed (UNCTAD 2005a and b), chemical, pharmaceutical/biotechnology and IT hardware industries had the higher shares of R&D employees abroad (Table 13.9).17

5

POLICY REACTIONS TO ADDRESS GROWING HRST DEMAND AT THE GLOBAL LEVEL

The EU, and with it other developed and developing countries, is shifting policy gears in reaction to pressure on the supply of highly skilled HRST and the changing landscape brought about by globalization. Increasing HRST is a key target of the Lisbon Agenda. For IT industries alone, there are predictions of shortages of 300 000 engineers for Europe. Microsoft has turned to off-shoring, including carrying out software development on the west coast of Canada claiming US visa limits prevent it from hiring the engineers it needs.I8 With a Canadian policy that facilitates entry of software developers and with absence of quota limits, the local economy in British Columbia is able to reap the benefits. The 'competitive' elements of economic immigration policies for foreign workers are about ease of entry and duration - the conditions for recruitment and the conditions for the stay of the foreign worker. Although not without its problems, the 'Researchers in Europe 2005' initiative

249

Thc European Research Area and human resources

Tahltr 13.8 Nun-immigrant admissions to the US, H-1 Bpermit and intracompany transfers by selected EU country, 2007 ('56)

France Germany Italy Spain UK

H- 1B Visa per cent

Intra-company per cent

14.6 15.1

19.3

10.9

7.2

4.9 4.4 40.9

7.1

29.5

Intra-company transfers include family members of the principal applicant.

Note: Sourw:

Based on figures of Yeurhook of'lrnrnigrc~tionStutisrics. 2007, Homeland Security.

Table 13.9 Share of R&D employees abroad, selected industries Per cent Chemicals Pharmalbio IT hardware Automotive - -

Sourtr

45.4 39.5 37.5

23.2

-

U N C T A D (2005a)

addressed the need to raise Europe's profile as an attractive place to pursue a career in research. It provides for fast-tracking of admission procedures for researchers and options to extend the stay of researchers without them having to return to their country of origin to reapply. As mentioned earlier, countries like Australia, Canada and the US have a long history of using immigration policy to leverage supplies of foreign highly skilled HRST. Countries in the EU like the UK and Germany have also reacted by putting policies in place to facilitate the import of highly skilled workers. Japan has eased immigration restrictions in order to augment its HRST work force, and countries like China and India are introducing policies to encourage their S&E trained nationals to pursue S&E careers at home. Table 13.10 provides information on some of the more recent policy responses to provide access to foreign-born seasoned workers as well as some examples of easing foreign student transition into the domestic labour market.

250

European science and technology polic~s

Table 13.10 Examples of recent policy schemes to acquire foreign-born HRST, selected countries Key Conditions for Entry

EU

Denmark

France

Germany

Ireland

Italy

Netherlands

Blue Card Scheme (proposed). Criteria include: Education: university qualification Contract (e.g., waiting in the EU) Experience - five years minimum Salary (minimum threshold) Duration: two or three year work and residence permit; extension possible. Green Card. Criteria: points system based on: Education Age Language Work experience Adaptability Bonus points for experience in occupations identified as suffering shortages Employer applies Work permits in targeted occupations no longer require labour market testing (fast tracking targeted occupations) Simplified transition for foreign students to work permit (e.g., IT occupations) Duration: one year; renewable Green Card - launched for IT demand Simplified transition for foreign students to work permit Duration: one year; renewable Green Card based on criteria including: Occupation identified as suffering shortage of skilled workers Job offer from company active in Ireland Duration: two year; renewable Reverse brain drain programme to bring Italian scholars and researchers back to Italy and to attract foreign HRST University contracts of two to four years to returning and/or foreign researchers. Bilateral agreements on S&T include researcher short-term mobility schemes Targeting of Asian HRST: Marco Polo programme provides facilities for Chinese students Employer applies Duration: one year; renewable

The European Research Area and h u n ~ a nresources

25 1

Table 13.10 (continued) Key Conditions for Entry - - ~-

UK

Non-EU Australia

Canada

Japan

Korea

uS

--

New points based system to target foreign workers that will contribute to economic growth No prior job needed; open access to U K labour market Skilled workers nominated by employer (unless three months); availability of domestic workers ground for refusal Criteria: Points system; age exclusion (must be 1.5) the international impact standard of the field. It is also possible to calculate CPPIFCSm for a specific country or for the EU or Europe as a whole. Table 14.2 shows the CPPIFCSm value for the EU-27 Member States. the three candidate countries, and the six major non-EU countries Brazil. China, India, Japan, Russia and the US. This indicator can be seen as a measure of the impact, that is, the 'scientific strength' of a country, regardless of size and averaged over all fields. Numbers are given for the ten-year period 1998-2007 and thus Table 14.2 also shows the trends in countries' international scientific strength. Most 'older' EU-27 Member States show an increase in scientific strength in the ten-year period. In recent years, Austria, Belgium, Finland, France, Germany, Ireland, Sweden, the UK and particularly Denmark and the Netherlands are above international level, while Luxembourg, Spain and Italy are about international level, Portugal is approaching international level and Greece is below international level but has increased considerably its scientific strength over the last ten years. Countries with a relatively strong increase in the ten-year period are Austria, Belgium, Denmark, Luxembourg, Portugal and Spain. The larger other 'older' EV-27 Member States have more or less stabilized their position at a high

-

Table 14.2 Scientific strength measured by the biblionietric crown indicator CPPIFCSm - o f t h e EU-27 Meniber States, three candidate countries and six major non-EU countries; trend analysis ivith overlapping four-yeur blocks in the period 1998-2007

Source:

Author's work.

Table 14.2 (continued)

Performance of Europem science

265

level. For most of the recent EU-27 Member States, a relatively strong increase is observed, particularly for Bulgaria, Czech Republic, Estonia, Hungary and Latvia. A relatively strong increase in scientific strength is also found for the Candidate Countries Croatia and Macedonia. While a strong increase in scientific output was noted for Turkey, the impact of this output has increased less strongly. The scientific strengths of the non-EU Member States discussed in this chapter are very different. It will not be a surprise that the US has by far the greatest scientific strength, but this strength is decreasing. Japan shows no significant increase in strength and remains below international level, which is disappointing given its considerably long scientific history. Russia is increasing its strength, but given the old and rich scientific tradition of this country, it is almost shocking to see that its strength is and remains significantly below that of developing countries like Brazil, China and India. Brazil improved its scientific strength, but not in a pronounced way. China almost doubled its strength, which constitutes a remarkable performance given the large increase in scientific size. The strength of India increases considerably, but less than that of China. So in terms of both scientific size and scientific strength, China is considerably stronger than India. We expect that within a few years, China will overtake Japan as the dominating Asian power house in science.

3 INTERNATIONAL COLLABORATION WITHIN THE EUROPEAN UNION Science is a global human endeavour not limited to national and cultural boundaries. The international nature of scientific research is driven by different factors such as the short- and long-term mobility of researchers and international collaboration in both 'big science' and smaller-scale research. Often, but certainly not always, international scientific collaboration becomes tangible through joint publications of research groups in different countries. A first indication of the extent to which researchers in a country cooperate with colleagues outside their own country is thus the share of internationally collaborative publications in the total scientific output of a country. The highest impact is generally observed for internationally collaborative publications (van Raan and van Leeuwen, 2002). Table 14.3 shows the percentage of international publications for all EU Member States. the three Candidate Countries and the six major non-EU Member States Brazil, China, India, Japan, Russia and the US. This indicator can be seen as a measure of the international orientation of a country. It should be

TabIe 14.3 International orientation - measured by the percentage of international publications - of the EU-27 Member States, three candidate countries und six major non-EU countries for the period 1998-2007

Source:

Author's work

168

European science und tec./zno/ogypolicj.

stressed, however, that the complementary percentage, that is, the share of non-internationally collaborative publications, is an indicator of the size of a country's own scientific basis and of its 'scientific independence'. A first observation is that for the 'older' EU-27 Member States, the share of international publications was in the 30-50 per cent range ten )ears ago and increased to 40-60 per cent nowadays. For the Netherlands. the UK, Greece, Spain and particularly Ireland, Italy and Luxembourg, the share is below 50 per cent while the lowest percentage in the EU-27 is found for Portugal. For most of the recent EU-27 Member States. a similar increase is observed though to a level that is generally still lower than that in the 'older' Member States, particularly in Poland and Cyprus. For the Candidate Countries, a considerable increase in international collaboration is observed in Turkey and Croatia, but a strikingly low level and no positively evolving trend in Macedonia. Among the six non-EU countries, both Brazil and China show a somewhat decreasing percentage of international collaboration. This does not necessarily mean that these countries cooperate less internationally but simply that the number of scientific papers produced in the country itself, without international cooperation, is growing more rapidly than the number of papers resulting from international collaboration. This s~tuationis typical for countries with a quickly growing science system. For the other countries -particularly Japan and the US, to a lesser extent Russia - a clear increase is observed in the share of papers resulting from international collaboration in total scientific output. A further step in the study of international collaboration is to break the total number of international publications of a country down into numbers for different collaboration partner countries. This can be done by constructing a matrix in which countries make up the rows and columns of the matrix and in which the cells record the number ofjoint publications between any pair of countries. To investigate collaboration between European countries. one can restrict the matrix to these countries. Several data analysis techniques including proper normalization exist for transforming matrices into landscapes of relations. An example is mapping scientific fields by co-word or co-citation analysis (van Raan, 2004). In these mapping techniques, clustering and multidimensional scaling procedures are used. Another approach is network analysis. The increase of the past few years in network analysis has been stimulated strongly by the development of software programs that include the mathematical models of network theory. Examples of these software programs are Pajek (Batagelj and Mrvar, 2006) and Netdraw (Horgatti, 2002). In network analysis and visualization research, many studies are based on co-publication activities (see, for instance, Newman, 2001a and 2001b; Borner et al., 2005; Calero et al., 2006).

Performance o f European science

269

'EE

PT

a)

Figure 14.1

Collaboration network for all research of the EU-27 Member States and the three Candidate Countries for two successive periods ( a ) 1998-2002 and ( h ) 2003-07'

In Figure 14.1. Netdraw network maps are presented of the EU collaboration matrix for all research of the EU-27 Member States and the three Candidate Countries for two successive five-year periods, 19982002 and 2003-07. While Figure 14.1 looks at the scientific production of countries regardless of the field of research, Figures 14.2 and 14.3 present

270

European science and technology polic~ .MT

Figure 14.2

Collaboration network for 'Clinical Medicine' research of the EU-27 Member States and the three Candidate Countries for two successive periods ( a ) 1998-2002 and ( b ) 2 0 0 3 4 7

for the same two five-year periods, the collaboration networks of the EU collaboration matrix for the main fields 'Clinical Medicine' and 'Physics' respectively. The position in the maps of the countries is based on a spring-embedded algorithm included in the Netdraw software. Its effect is to distribute the vertices in a two-dimensional plane with some separation while attempting

Performance of European science

271

Figure 14.3 Collaboration network for 'Physics' research of the EU-27 Member States and the three Candidate Countries for two successive periods ( a ) 1998-2002 and ( b ) 2 0 0 3 4 7

to keep connected countries reasonably close together. The edges can he seen as springs pulling vertices (countries) together, though never too close (de Nooy et al., 2005). The algorithm 'pulls' vertices to better positions until they reach a state of equilibrium. In Figures 14.1 to 14.3, this layout means that countries that are linked or have links in common will be closer in the map. In order to make the network maps easier to visualize, only

272

European science and technology policj Country

Clinical medicme

( CPPIFCSm)

Austria Belgium Bulgaria Cyprus Czech Denmark Estonia Finland France Germany Greece Hungary Ireland Italy Latvia Lithuania Luxembourg Malta Netherlands Poland Portugal Romania Slovakia Slovenia Spain Sweden United Kingdom

1

C r o a t ~ a(0 48) Macedonla (I 03) /j/////////////////A Turkey ( 0 47) 0

1

I

I

5

10

I5

. I

I

20

25

I 30

Share of the national output ('I/;,)

) Impact: hote:

Source:

Low

=

Low

Average

High

CPPIFCSm < 0.8: High CPPIFCSm > 1.2

Author's work.

Figure 1 4 . 4 ~ Fieldprojile of 'Clinical Medicine' for the EU-27 Member States and the three Candidate Countries, recent five-year period 200347, based on all WoS covered publications. Countries are ranked alphabetically co-publication links, of which the strength exceeds a certain threshold (haries from one figure to another), have been displayed. In all figures about 50 per cent of the co-publication matrix is represented. In the figures, the central position can be observed of most of the larger and 'older' EU-27 Member States. Most of the more recent Member States

Prformance of'European science Country (('PPIFCSm)

~ u l g a r i a(0.77) Cyprus (0.56) Czech ( I .O3) Denmark (I .49) Estonia (0.87) Finland (1 35) France (I 2 3 ) Germany ( I .24) Greece (0.81) Hungary (I .02) Ireland (I .21) Italy (1.09) atv via (I .73) Lithuania (1.10) Luxembourg (1.00) Malta(l.16) Netherlands (I .36) Poland (1.00) Portugal (1.10) Romanla (1.29) Slovak~a(0.87) Slovenia (0.64) Spain (1.22) Sweden ( I .29) United Kingdom (1.21) Croaha (0.48) Macedonia ( 1.12) Turkey (0.48)

I

I I

0

5

Impact:

Note. Source:

Low

=

:

Y////////////////' i Zj I

I

Low

Average

.

I

1

25

10

I

10 15 20 Share of the nat~onaloutput ('XI) High

I

CPPIFCSnl < 0.8: High CPPIFCSm > 1.2.

Author's work.

Figure 14.4b Field profile of 'Clinical Medicine'jor the EU-2 7 Member States and the three Candidute Countries, recent five-year period 200347, based on English WoS covered publications only. Countries are ranked alphabetically

as well as the Candidate Countries are positioned on the periphery of the network. Figure 14.1 shows that several countries - for instance, Austria, Belgium, Greece, the Czech Republic and Poland - have moved to positions more inside the network. The configurations are generally stable over time, which means that it takes a long time to change significantly

European science und technology policy

274

Country

Physics

( CPPIFCSm)

I

I urkey (U 6 6 )

0

I

I

I

5

10

15

.

I

I

20

25

Share of the national output ('%I) Impact: Aote:

Source:

Low

=

Low

Average

High

CPPlFCSm < 0.8: High CPPIFCSm > 1.2

Author's work

Figure 1 4 . 4 ~ Field profile of 'Physics'for the EU-27 Member States and the three Candidate Countries, recent five-year period 200347, based on all WoS covered publications. Countries are ranked alphabetically international relations in science. The comparison of Figures 14.2 and 14.3 reveals the differences in international relations across different fields. in this case 'Clinical Medicine' and 'Physics'. For instance, Poland occupies a more central position in the network in 'Physics' than in the one in 'Clinical Medicine', particularly in the recent period 2003-07.

4 RESEARCH PROFILES A further important step is the breakdown of the output of EU Member States and Candidate Countries into research fields. This provides a clear indication of the research scope or profile of countries. There are two types of profiles. First, a field profile, in which output and field-normalized measured impact (CPPIFCSm) are ranked by country. Results are presented at the aggregation level of 'major fields', a major field being a set of related fields. These major fields are the same ones as used in the Netherlands Observatory of Science and Technology (Tijssen et al., 2008). Second, a country profile, in which output and field-normalized measured impact (CPPIFCSm) are ranked by field. In field profiles, output and impact in one specific field of research (for example, clinical medicine) are thus displayed for all countries, while in country profiles, output and impact in all major fields of research are displayed for one specific country. In Figures 14.4a to 14.4c, the profiles are shown for two major fields 'Clinical Medicine' (in two modalities as explained hereafter) and 'Physics' for the recent five-year period 2003-07 for the EU-27 Member States and the three Candidate Countries. Countries are ranked alphabetically. In 14.5a to 14.5g, the country profiles are shown for three EU-27 Member States, one of the three Candidate Countries, China, India and the US as examples of non-EU countries. First, Figures 14.4a to 1 4 . 4 ~are discussed. The length of the bars represents the share a major field accounts for within the national scientific output. For instance, 'Clinical Medicine' accounts for 22 per cent of Austria's total scientific output as far as publications covered by the WoS system are concerned. The impact is indicated by CPPIFCSm classes: a white bar stands for performance significantly below the international average for the field, that is, CPPIFCSm < 0.8; a shaded bar stands for performance around the international average, that is, 0.8 < CPPIFCSm < 1.2 and a black bar stands for performance significantly above the international average for the field concerned, that is, CPPIFCSm > 1.2. The actual CPPIFCSm values are given between parentheses immediately after the name of the country. It can thus be observed that the (average) impact of 'Clinical Medicine' research is 1.08 in Austria, 1.02 in Germany and 0.99 in France. These indicator values are based, as usual, on all publications of these countries covered by the entire WoS. It has been known for quite a long time now (van Leeuwen et al., 2001) that the WoS data system covers a number of journals in non-English languages, particularly in German and in French. Publications in these non-English language journals are counted as part of the output of countries but generally have very low impact as very few scientists outside Germany, Austria and Switzerland are able to read -

-

276

European science and technology policy

German. The same applies to French. These non-English publications will therefore 'dilute' considerably the average impact of countries such as Germany, Austria and France. This is particularly the case for more application-oriented fields such as 'Clinical Medicine' and much less the case for 'Physics'. Through the comparison of the 'Clinical Medicine' profiles for all WoS covered publications (Figure 14.4a) with those for only-English WoS covered publications (Figure 14.4b), a quite dramatic language effect can be identified (van Raan et al., 2009). The differences are striking: the impact of Austria changes from 1.08 to 1.17, that of Germany from 1.02 to 1.24 and that of France from 0.99 to 1.33! But the language effect also influences to a lesser extent the impact of many more other countries in which scientists write a certain number of publications in French or German, particularly Belgium, the Baltic Countries and the Czech Republic. In addition, also for Spain a considerable language effect is noted; here the impact changes from 1.01 to 1.22. As can be expected, for the English-speaking countries the UK and Ireland, the effect is zero. It is also clear that for countries that achieve a higher impact when excluding the non-English publications, the relative output in 'Clinical Medicine' decreases for the same reason (excluding non-English publications). As always, interpreting the figures requires caution for statistical reasons. It is clear that in a smaller country, a few institutes or even one very good research group in a specific field will determine the scientific size and strength of the whole country for that field. Conversely, in a larger country, the performance of excellent groups in a field may completely 'disappear' in the average for that country because there are many reasonably good but not that excellent other groups. Some striking features are observed, for example, the large difference in output (share of national output) in 'Clinical Medicine' between Turkey (large output) and Bulgaria (small output). In both cases, however, the impact of 'Clinical Medicine' research is not high and for Turkey particularly low. This does not mean, however, that there are no high-impact research groups in 'Clinical Medicine' in Turkey. They may just have 'disappeared' into the average value for the whole country. This clearly shows that further more, detailed analysis is necessary to identify the strengths and weaknesses in each country at, for instance, university level. For Latvia and Romania as well, a small output is observed for 'Clinical Medicine', but it has high impact. These striking differences are important for understanding the science systems of the countries and more detailed bibliometric research will enable us to find an explanation for the observed differences. Belgium, Denmark, Finland, France, Germany, Ireland, Sweden, the Netherlands, Spain and the UK also have strong positions in 'Clinical Medicine': a high output with a high impact (strength) (Figure 14.4b).

Performance of European science

277

Some features in the 'Physics' profile (Figure 1 4 . 4 ~contrast ) with those of the 'Clinical Medicine' profile, for instance, the output of Bulgaria and Romania. These observations once more illustrate the differences between national science systems, particularly the different degree of specialization in specific fields of research. 'Physics' accounts for the largest share of national research output in Romania. Germany occupies a remarkably strong position in 'Physics' (in terms of relative output and impact) while in the Netherlands, 'Physics' accounts for a comparatively small share of national research output but achieves high impact. In Figures 14.5a to 14.5g, country profiles are presented. Presenting profiles for all EU-27 Member States, the three Candidate Countries and the six non-EU countries would make this chapter too lengthy. Therefore a choice was made that allows for discussing the kinds of observations that can be made through these profiles. The countries selected include Denmark, a smaller, 'older' NorthWestern European EU-27 Member State with a historically strong scientific tradition; Greece, a smaller, relatively 'recent' Mediterranean EU-27 Member State; Poland, a recent larger Eastern European EU-27 Member State; Turkey, one of the Candidate Countries; and China, India and the US as examples of non-EU countries. Moreover, these countries do not 'suffer' from the language effect. Indicator values are based on all WoS-covered publications. The profiles cover the recent five-year period 2003-07. Fields are ranked by their relative output. The profiles cover all (major) fields accounting for more than 1 per cent of the total scientific output in WoS-covered publications. Similarly to Figures 14.4a to 14.4c, the length of the bars represents the share accounted for by (major) field within the national scientific output while the shading of the bars provides an indication of the impact. Immediately the main features of these countries' national science systems can be observed: the fields making up these countries' 'scientific core business' and their performance in these fields. For instance, for Denmark, 'Clinical Medicine' accounts for the largest share (19 per cent) with impact CPPIFCSm = 1.48, followed by 'Basic Life Sciences' (1 1 per cent) with CPPIFCSm = 1.22 and 'Biomedical Sciences' (almost 10 per cent) with CPPIFCSm = 1.03. The best performance for Denmark is in 'Materials Science' (at least among the fields covered by this profile) with CPPIFCSm = 1.65. The worst performance is found in 'Biomedical Sciences', although this performance is still around the international average for this field. For Greece, the three largest fields are 'Clinical Medicine', 'Physics' and 'Chemistry' with each field's impact around the international average. In particular 'Civil Engineering' and 'Chemical Engineering' achieve impact significantly above the international average with CPPIFCSm = 1.43 and

278

European science and rechnologjt policj Field

Denmark

( CPPIFCSm)

Clln~calmedlcme ( l 48) Bas~chfe saences ( l 22) B~omed~cal sclences (1 03) Physlcs ( I 46) Agr~cultureand food sclence (1 24) Chemtstry (I 38) B~olog~cal sclences ( l 1 7) Environmental sclences ( I 26) Earth sciences ( l 22) Pharmacology ( I 14) Health sclences (I 12) Computer sclence ( 1 45) Materials sclence ( 1.65) Mechan~calengmeerlng ( 1 39) Mathematics (1 52) Astronomy ( I 26) Economics, busmess and management (1 07) CIVIJenglneermg (1 33) Electr~calengineering ( l 37) Fuels and (nuclear) energy ( 1.21 )

0

5 10 15 20 25 Share of the output ('YO)

I Impact: 0 Low fife:

O Average W High

Fields are ranked by share of national scientific output.

Source: Author's work.

Figure 1 4 . 5 ~ Country projile for Denmark, recentjve-yew period 200347

1.22, respectively. Bad performance is found in 'Biomedical Sciences', 'Computer Science', 'Mathematics' and 'Pharmacology' with CPPIFCSnl between 0.60 and 0.70. For Poland, the three largest fields are 'Physics', 'Chemistry' and 'Clinical Medicine'. The performance in the latter field is around the international level while for the two other fields, performance is below international level. Once again the problem of statistics discussed earlier should be emphasized. It is very possible that there are excellent research groups in 'Physics' and 'Chemistry' in Poland but that the performance of these groups disappears in the average of the entire country. So the profile provides us with information on the general state of art of a research field

Performance of European science Field

Greece

( CPPIFCSm)

Clinical medicine (0.81) Physics (0.98) Chemistry (0.99) Biomedical sciences (0.68) Basic life sciences (0.80) Computer science (0.67) Agriculture and food science (0.99) Electrical engineering (0.82) Environmental sciences (0.94) Materials science (0.83) Biological sciences (0.71) Earth sciences (0.81) Mathematics (0.65) Mechanical engineering (0.95) Pharmacology (0.69) Civil engineering ( 1.43) Fuels and (nuclear) energy (0.80) Chemical engineering (1.22) Astronomy (0.70) Health sciences (0.90) Other engineering sciences (0.98) 0 Share of the output ('XI) Impact: Note:

Low El Average H High

Fields are ranked by share of national scientific output

Source: Author's work

Figure 14.5b

Country profile for Greece, recent jive-year period 200347

as a whole in a country and does not provide any information on the possible existence of 'crown jewels' in that field. A more detailed bibliometric analysis is necessary to identify these crown jewels. Good performance for Polish science is found in 'Astronomy' with a CPPIFCSm around the international level (1.01) while in 'Chemical Engineering', performance is particularly low (CPPIFCSm = 0.33). The results for Turkey show a country with a developing science system. The three largest fields are 'Clinical Medicine', 'Chemistry' and 'Biomedical

European science a n d technology policy Field

Poland

( CPPIFCSm)

Phys~cs(0.77) Chem~stry(0.57)

h n ~

Clinical medicine (I .00) Bas~chfe sclences (0.52) Mater~alssclence (0 50) Biomed~calsclences (0.56) Agriculture and food science (0.48) Biological sciences (0.48) Computer science (0.66) Mathematics (0.65) Chemical engineering (0.33) Environmental sciences (0.56) Pharmacology (0.66) Mechanical engineering (0.56) Earth sciences (0.61) Astronomy ( I .Ol) Electrical engineering (0.53) Fuels and (nuclear) energy (0.60) Instruments and instrumentation (0.73)

0 Impact: Note:

.

5 10 15 20 Share of the output (%) Low El Average

25

High

Fields are ranked by share of national scientific output.

Source: Author's work.

Figure 1 4 . 5 ~ Country pro$le for Poland, recentjve-year period 200347

Sciences', all with performance below international level. Good performance at international level is found for 'Chemical Engineering', 'Fuels and Energy', 'Mechanical Engineering', 'Computer Science', 'Civil Engineering' and 'Electrical Engineering'. It appears that the engineering fields constitute the strengths of the Turkish science system. Improvement is necessary for the bio- and medicine-related fields. For China, a focus on the natural sciences and engineering is clearly observed, impact being around international level. The impact of 'Mathematics' is above international level. Somewhat behind in strength are

Performance of European science Field

Turkey

( CPPIFCSnz)

- 1

Clinical medicine (0.47) Chemistry (0.63) Biomedical sciences (0.45) Physics (0.66)

Agriculture and food science (0.65) Basic sciences (0.55) Materials science (0.72) Environmental sciences (0.68) Pharmacology (0.57) Chemical engineering (0.90) Biological sciences (0.46) Fuels and (nuclear) energy (0.81) Mechanical engineering (0.97) Computer science (0.81) Mathematics (0.68) Civil engineering (0.88) Earth sciences (0.77) Electrical engineering (0.82) Dentistry (0.70) Health sciences (0.59) 0

I Impact: Note:

5

.

10 15 20 25 Share of the output ('%I) Low 0 Average

30

1

~igh

Fields are ranked by share of national scientific output

Source: Author's work.

Figure 14.5d

Country profile,for Turkey, recent jive-year period 200347

the 'Basic Life Sciences' while in 'Clinical Medicine', performance is better. We also observe a relatively good impact position for 'Environmental Sciences' and 'Agriculture & Food Science'. The profile of India shows that 'Physics', 'Astronomy', 'Chemical Engineering' and particularly 'Civil Engineering' are around international level. Far below international level is India's performance in the life sciences, particularly 'Biomedical Sciences', and also 'Clinical Medicine' as well as 'Agriculture & Food Science'. Given the large social and in particular health problems in this country, one can

European science and technology policj. Field

China

( CPPIFCSm)

Physics (0.83) Chemistry (0.85)

Computer science (0.86) Mathematics (1.02) Biomedical sciences (0.8 1) Electrical engineering (0.86) Mechanical engineering (0.95) Biological sciences (0.65) Earth sciences (0.86) Environmental sciences (0.86) Pharmacology (0.71) Agriculture and food science (0.94) Chemical engineering (0.80) Fuels and (nuclear) energy (0.85) Civil engineering (0.99) Other engineering sciences (1.17)

L. Share of the output (%I)

1 Im~act: Note:

Low E2 Averaee

~ i e1h

Fields are ranked by share of national scientific output.

Source: Author's work

Figure 14.5e Country profile for China, recent five-year period 200347 ask whether the Indian government should not change its science policy. In the US, biomedical research is dominant in size but the natural sciences dominate in strength, particularly 'Materials Science'. Less strong but still about the international level are fields such as 'Pharmacology'. As discussed above, statistical constraints limit the explanatory value of our profiles. Nevertheless these profiles provide first indications of where countries can improve their science system. Notwithstanding their shortcomings, bibliometric profiles provide important instruments for science policy and research management, also at the national level.

Performance of'European science Field

India

( CPPIFCSm)

Chemistry (0.65) Physics (0.83) Clinical medicine Materials science Basic life sciences Agriculture and food science Biomedical sciences Biological sciences Pharmacology Environmental sciences Earth sciences Mechanical engineering Chemical engineering Fuels and (nuclear) energy Multidisciplinary sciences Computer science Electr~calengineering Mathematics Astronomy

.

1

,

I

(0.52) (0.79)

(0.53) (0.47) (0.43) (0.5 1 ) (0.65) (0.68) (0.54) (0.79) (0.95) (0.37) (0.12) (0.75) (0.52) (0.56) (0.81)

Civil engineering (1.10)

0

2

I Impact: Note:

4 6 8 10 12 14 16 18 20 Share of the output ('YO)

Low

Average

High

1

Fields are ranked by share of national scientific output

Sourer: Author's work.

Figure 14.5f' Country projile for Indiu, recentjive-year period 2003-07

5 WHAT ABOUT THE SOCIAL SCIENCES A N D THE HUMANITIES? Based on the above, one could conclude that the social sciences and humanities are 'invisible' in applied bibliometric approach. Indeed, the number of publications in international journals is much lower in those fields than in the natural sciences and medical fields. Therefore, social sciences and humanities fields show up in the tail of country profiles and not in the first 30 or so positions. More than in the natural sciences and in the medical fields,

European science and technology p o l i q Field

US

( CPPIFCStn)

Clinical medicine Basic life sciences Biomedical sciences Physics

(1.32) (1.30) (1.26) (1.58)

Biological sciences (1.23)

Earth sciences Computer science Pharmacology Materials science Electrical engineering Mathematics Economics, business and management

(1.31) (1S 5 ) (1.13) ( 1.76) (1.40) (1.32) (1.43)

0 2 4 6 8 101214161820 Share of the output (%)

I Impact: .Vote:

Low el Average W High

Fields are ranked by share of national scientific output.

Source: Author's work.

Figure 14.5g

Country profile for the US, recentjive-year period 200347

national orientation (Kyvik and Larsen, 1994) may play an important role in the social sciences and humanities - the same is often the case for engineering and other applied research - but the social sciences and humanities generally also have a strong international orientation. More important is the difference in research and communication cultures between the natural sciences and the medical fields on the one hand and the social sciences and particularly the humanities on the other. An exception is psychology, where communication practices are comparable to those in the exact sciences. Such differences in communication practices reflect differences in the nature of the social sciences and humanities. For instance, in the humanities, consensus is less often found on what constitutes a successful scientific approach.

Perjormunce of' European science

285

In many parts of the social sciences and humanities, scientific publication practices are less standardized than in the exact sciences. The structure of the written scholarly communication system often lacks a clear core-periphery structure, English is not always a dominant language and the role ofjournals is less important than in the exact sciences. What are the consequences of these considerations for the application of bibliometric methods to the social sciences and humanities? Without going into too much detail, some important points of consideration for further approaches are presented here. In the social sciences and humanities, the number of citations is generally an order of magnitude lower than in the natural and medical sciences, which exacerbates the statistical problems. For most social sciences and humanities fields, it is necessary to use a considerably longer citation window (for example, five to six years or longer, even 'lifetime'; see Linmans, 2008) than in the natural sciences and medical fields (mostly 4 years). Within the social sciences and humanities, publication and citation characteristics may vary widely across different fields. As discussed above, these differences often relate to research cultures, for example, teamwork versus individual research. For instance, the internationally oriented publication culture of experimental psychologists sharply contrasts with the often 'locally' oriented sociologists. Meertens et al. (1992) found that books and book chapters account for about onethird of all Dutch social psychology publications. These 'non-WoS publications', however, can well be cited in articles in WoS-covered journals. Through appropriate analytical procedures, their impact can be estimated. The persistent characterization of the social sciences and humanities as 'bibliometrically inaccessible' should therefore not be too easily accepted (van Leeuwen, 2006). Furthermore, comparison with a European benchmark is an effective means of coping with a possible US bias in the WoS, particularly for the social sciences and humanities. As a general principle, it can be stated that monographs, doctoral theses and multi-authored books are important sources of written communication in many disciplines of the social sciences and humanities and should therefore not be omitted in a comprehensive assessment of publication output (Moed, 2005). Whereas a standard bibliometric analysis takes account only of citations to publications in WoS-covered journals, an expanded analysis also determines as discussed above the citation impact of non-WoS publications, in particular books and contributions to edited volumes such as conference proceedings. For the humanities, the importance should be stressed of the development of alternative, non-citationbased measures of impact or availability, for example, the extent to which books published by an academic staff member are included in collections of research libraries all over the world (Torres Salinas and Moed, 2008).

186

European science and technology policy

Although many databases are available that cover scholarly output in the social and behavioural sciences and humanities, these databases have several properties that make them less suitable for calculating bibliometric indicators: It may be unclear which bibliographic sources are processed. The criteria for the selection of sources may be unclear. The database may have national or geographical biases. Large fractions of processed documents do not mention the institutional affiliations of authors. Even if documents contain author addresses, database producers may not include these in the database. Important data elements - even journal titles and country names may not be standardized. To the best of our knowledge, none of the major databases includes cited references. Many databases are available only through host computers that offer only limited counting and statistical facilities. Their use may be expensive.

-

The following steps can be made to assess research performance in the social sciences and humanities. Major parts of the social sciences and humanities - such as experimental psychology, economics and linguistics - are more and more approaching the publication behaviour of the natural sciences (Nederhof, 2006). To a considerable extent, therefore, the problem of the applicability of bibliometric analysis in the social science and humanities boils down to WoS journal coverage. The extent to which cited references of WoS-covered publications are themselves WoS-covered publications provides an indicator of WoS coverage for the fields concerned. Table 14.4 shows coverage percentages of large disciplines (Moed, 2005). Therefore, for all social science and humanities fields, the WoScoverage has to be examined. A trend analysis is particularly important to observe changes in 'publication culture' (in terms of WoS coverage). Furthermore, other databases than the WoS can be explored, particularly Scopus and Google Scholar (both providing citation data) as well as field-specific databases such as ECONLIT, Psychological Abstracts, Sociological Abstracts. A new and important development is the creation of national research databases in which publications in all fields of sciences, including the social sciences and humanities, are covered on the basis of field-specific quality criteria, regardless of whether a publication is covered by WoS or Scopus and of document type. A good example of this develop-

Perfornzunce oj European science Tuble 14.4 Znternul WoS coveruge percentages

Internal Coverage Percentage 80- 10O1%1

60-80%

40--60%)