Subsidiarity and Economic Reform in Europe
George Gelauff · Isabel Grilo · Arjan Lejour Editors
Subsidiarity and Economic Reform in Europe
123
George Gelauff Arjan Lejour CPB Netherlands Bureau for Economic Policy Analysis Van Stolkweg 14 2585 JR The Hague Netherlands
Isabel Grilo European Commission DG Enterprise and Industry Avenue d’Auderghem 45 1040 Brussels Belgium
[email protected]
[email protected] [email protected]
ISBN 978-3-540-77245-3
e-ISBN 978-3-540-77264-4
DOI 10.1007/978-3-540-77264-4 Library of Congress Control Number: 2008922298 c 2008 Springer-Verlag Berlin Heidelberg This work is subject to copyright. All rights are reserved, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilm or in any other way, and storage in data banks. Duplication of this publication or parts thereof is permitted only under the provisions of the German Copyright Law of September 9, 1965, in its current version, and permission for use must always be obtained from Springer. Violations are liable to prosecution under the German Copyright Law. The use of general descriptive names, registered names, trademarks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. Production: le-tex Jelonek, Schmidt & Vöckler GbR, Leipzig Cover design: WMX Design GmbH, Heidelberg Printed on acid-free paper 987654321 springer.com
Preface
Sometimes the European Union is seen as a big success, sometimes it is not. It is successful in integrating the market economies of 27 European countries and 500 million citizens in fifty years time. Now it is the biggest economy in the world, but in many areas other than the Internal Market the role of the EU is much more limited. Besides historical reasons, other motives for the division of competences between the EU and the Member States explain the current role of the EU. In this book the authors discuss, mainly from an economic perspective, this division of responsibilities for economic policy. They concentrate on education and innovation, internal market policy and the common agricultural policy, corporate income taxation, regional policy and transport policy. They ask whether it is efficient to assign these policies to the national or the European level and try to explain the current division of these responsibilities. This leads to interesting conclusions on the role of the European Union and the Member States in economic policy making which are also very relevant for economic reforms triggered by the Lisbon strategy. The chapters in this book are based on contributions presented at the conference Subsidiarity and Economic Reform in Europe in Brussels, November 8-9, 2006.1 The conference was organized by the European Commission, CPB Netherlands Bureau for Economic Policy Analysis and the Dutch Ministry of Economic Affairs.2 The Ministry also provided financial support for the publication of this book, which is much appreciated. We, as editors of the book and organizers of the conference, want to express our gratitude to everybody who supported the conference and publication of the book. In particular, we want to thank Stephan Raes, who was very active in initiating this project within the Dutch Ministry of Economic Affairs, and Jacques Pelkmans for his support in composing the program and for recommending very qualified authors. Moreover, we are grateful to Gert-Jan Koopmans for giving support from the European Commission 1 2
See: www.cpb.nl/goto/subsidiarity/. Note that all contributions in this book are written on a personal title and do not necessarily reflect the views of the European Commission, CPB or the Dutch Ministry of Economic Affairs.
vi
Preface
and for hosting a small workshop, as appetizer for the conference. We enjoyed our work as editors. Our job was made easier by the enthusiastic and prompt responses of the contributors and above all by the accurate lay-out editing work of Jeannette Verbruggen (CPB). The editors, George Gelauff Isabel Grilo Arjan Lejour
Contents
Preface ........................................................................................................ v List of Contributors ................................................................................xiii 1 Subsidiarity for Better Economic Reform? .......................................... 1 George Gelauff, Isabel Grilo and Arjan Lejour 1.1 Introduction ...................................................................................... 1 1.2 An economic perspective on subsidiarity ......................................... 4 1.3 Education and innovation ................................................................. 5 1.4 The Internal Market and agricultural policy ..................................... 9 1.5 Taxation .......................................................................................... 11 1.6 Regional and transport policy......................................................... 13 1.7 Conclusion ...................................................................................... 15 References ............................................................................................ 18 2 Assessing Subsidiarity .......................................................................... 19 Sjef Ederveen, George Gelauff and Jacques Pelkmans 2.1 Introduction .................................................................................... 19 2.2 Fiscal federalism: the basics ........................................................... 21 2.3 Government imperfections ............................................................. 26 2.4 Decision making in legislature ....................................................... 32 2.5 Conclusion: assessing subsidiarity ................................................. 37 References ............................................................................................ 39 3 Who Shall Decide What? Citizens’ Attitudes Towards Political Decision Making in the EU .................................................................. 41 Joachim Ahrens, Martin Meurers and Carsten Renner 3.1 Introduction .................................................................................... 41 3.2 Measuring citizens’ preferences ..................................................... 42 3.3 Attitudes of citizens in the EU15.................................................... 45 3.4 Attitudes of citizens in the new member countries......................... 52 3.5 Conclusion ...................................................................................... 55 References ............................................................................................ 57
viii Contents
4 Subsidiarity and Economic Policy....................................................... 59 Jan Willem Oosterwijk 4.1 Introduction: subsidiarity................................................................ 59 4.2 The European policy agenda: more, less, different ........................ 60 4.3 Political Economy........................................................................... 61 4.4 Public support ................................................................................. 63 4.5 Conclusion ...................................................................................... 64 5 Higher Education Reform and the Renewed Lisbon Strategy: Role of Member States and the European Commission.................... 65 Frederick van der Ploeg and Reinhilde Veugelers 5.1 Introduction .................................................................................... 65 5.2 Challenges of higher education in Europe...................................... 67 5.3 Reforming higher education in Europe........................................... 80 5.4 What is the role of the EU? ............................................................ 86 5.5 Some concluding remarks .............................................................. 93 References ............................................................................................ 95 6 Higher Education, Mobility and the Subsidiarity Principle ............. 97 Marcel Gérard 6.1 Introduction .................................................................................... 97 6.2 Mobile students: from the Production to the Origin Principle........ 98 6.3 Assigning responsibility for attracting mobile researchers........... 105 6.4 Conclusion .................................................................................... 109 References .......................................................................................... 110 7 European Coordination of Higher Education.................................. 113 Sjef Ederveen and Laura Thissen 7.1 Introduction .................................................................................. 113 7.2 Student mobility: developments and determinants....................... 114 7.3 Economies of scale: does size matter?.......................................... 119 7.4 Cross-border externalities............................................................. 121 7.5 Conclusions and implications ....................................................... 124 References .......................................................................................... 126 8 On the Roles and Rationales of European STI-Policies .................. 129 Rahel Falk, Werner Hölzl and Hannes Leo 8.1. Introduction ................................................................................. 129 8.2 Rationales for supranational STI policy ....................................... 130 8.3 STI policy and subsidiarity in practice ......................................... 134 8.4 Conclusion .................................................................................... 140 References .......................................................................................... 141
Contents ix
9 Why European Innovation Policy? ................................................... 143 Albert van der Horst, Arjan Lejour and Bas Straathof 9.1 Introduction .................................................................................. 143 9.2 Innovation policy and the principles of subsidiarity..................... 144 9.3 Public expenditure on R&D.......................................................... 146 9.4 Intellectual property rights............................................................ 153 9.5 Conclusions .................................................................................. 155 References .......................................................................................... 156 10 Integrating Regulated Networks Markets in Europe .................... 157 Jordi Gual 10.1 Introduction ................................................................................ 157 10.2 Strategies for market integration ................................................ 158 10.3 What is different about network services?.................................. 161 10.4 The integration strategy in specific network services: banking, electricity and telecoms .............................................................. 165 10.5 Concluding remarks.................................................................... 172 References .......................................................................................... 174 11 Subsidiarity and the Internal Services Market .............................. 177 Arjan Lejour 11.1 Introduction ................................................................................ 177 11.2 The Services Directive................................................................ 178 11.3 The benefits of integrated services markets................................ 180 11.4 Diversity ..................................................................................... 181 11.5 Political economy issues............................................................. 185 11.6 Conclusions ................................................................................ 186 References .......................................................................................... 188 12 Agriculture Policy: What Roles for the EU and the Member States? ................................................................................................ 191 Harald Grethe 12.1 Introduction ................................................................................ 191 12.2 Overview of EU agricultural policy ........................................... 192 12.3 Costs and benefits of centralization ............................................ 195 12.4 Why is the CAP not in accordance with the subsidiarity principle? Political economy considerations and dynamics ....... 204 12.5 Conclusions on agricultural policy - How to get closer to the ideal? .......................................................................................... 210 References .......................................................................................... 214
x
Contents
13 Tax Policy and Subsidiarity in the European Union ..................... 219 Michael Keen and Ruud de Mooij 13.1 Introduction ................................................................................ 219 13.2 Principles for the allocation of taxing powers ............................ 220 13.3 Tax coordination in the EU ........................................................ 229 13.4 Why tax coordination fails ......................................................... 232 13.5 Administrative measures as a substitute for policy choices ....... 236 13.6 Conclusions ................................................................................ 238 References .......................................................................................... 239 14 Coordinating Corporation Taxes in the European Union: Subsidiarity in Action! ..................................................................... 243 Sijbren Cnossen 14.1 Introduction ................................................................................ 243 14.2 Brief Survey of Corporation Taxes............................................. 244 14.3 Bottom-up approach: tax coordination by Member States ......... 250 14.4 Concluding comment.................................................................. 256 References .......................................................................................... 257 15 Code of Coordination for Corporate Taxation .............................. 259 Albert van der Horst 15.1 Introduction ................................................................................ 259 15.2 Competition in tax rates.............................................................. 261 15.3 Tax base...................................................................................... 265 15.4 Subsidiarity in corporate taxation............................................... 271 References .......................................................................................... 272 16 Corporate Income Taxation and the Subsidiarity Principle......... 273 François Pouget and Eloïse Stéclebout-Orseau 16.1 Introduction ................................................................................ 273 16.2 Corporate taxation coordination in the European Union ............ 276 16.3 Analysing strategic interactions.................................................. 278 16.4 Policy conclusions and areas for further work............................ 286 References .......................................................................................... 289 17 Subsidiarity in Regional Policy........................................................ 291 Iain Begg 17.1 Introduction ................................................................................ 291 17.2 The rationale for EU regional policy .......................................... 292 17.3 Regional policy effectiveness ..................................................... 298 17.4 The Structural Funds in economic governance........................... 301
Contents xi
17.5 A regional policy subsidiarity test? ............................................ 304 17.6 Concluding comments ................................................................ 307 References .......................................................................................... 309 18 Regional Policy as a Means to Curb Immigration......................... 311 Robert Fenge and Volker Meier 18.1 Why does a EU regional policy exist?........................................ 311 18.2 Potential migration and its impact on unemployment ................ 314 18.3 Regional policy against undesired immigration ......................... 316 18.4 Regional policy and subsidiarity ................................................ 320 References .......................................................................................... 321 19 Subsidiarity and Transport Policy in Europe: What EU-Subsidies Do We Need for the TEN? ............................. 325 Bruno de Borger and Stef Proost 19.1 Introduction ................................................................................ 325 19.2 The economics of infrastructure provision by Member States in the presence of spillovers ....................................................... 326 19.3 Pricing and capacity choices by Member States or regions in the absence of EU intervention................................................... 332 19.4 What subsidy rules make sense for TEN investments? .............. 337 19.5 Conclusions ................................................................................ 339 References .......................................................................................... 340
List of Contributors
Ahrens, Joachim Private University of Applied Sciences Göttingen Weender Landstraße 3-7, 37073 Göttingen, Germany e-mail:
[email protected] Begg, Iain European Institute, London School of Economics and Political Science Houghton Street London WC2A 2AE, United Kingdom e-mail:
[email protected] Cnossen, Sijbren University of Maastricht PO Box 616, 6200 MD Maastricht, the Netherlands CPB Netherlands Bureau for Economic Policy Analysis PO Box 80510, 2508 GM The Hague, the Netherlands e-mail:
[email protected] De Borger, Bruno Department of Economics, University of Antwerp Prinsstraat 13, 2000 Antwerp, Belgium e-mail:
[email protected] De Mooij, Ruud Erasmus University, Rotterdam PO Box 1738, 3000 DR Rotterdam, the Netherlands CPB Netherlands Bureau for Economic Policy Analysis PO Box 80510, 2508 GM The Hague, the Netherlands e-mail:
[email protected] Ederveen, Sjef Directorate General for Economic Policy, Ministry of Economic Affairs PO Box 20101, 2500 EC The Hague, the Netherlands e-mail:
[email protected]
xiv List of Contributors
Falk, Rahel WIFO Austrian Institute of Economic Research P.O. Box 91, 1103 Vienna, Austria e-mail:
[email protected] Fenge, Robert Ifo Institute for Economic Research Poschingerstr. 5, 81679 Munich, Germany e-mail:
[email protected] Gelauff, George CPB Netherlands Bureau for Economic Policy Analysis PO Box 80510, 2508 GM The Hague, the Netherlands e-mail:
[email protected] Gérard, Marcel Louvain School of Management, FUCaM, Catholic University of Mons Chaussée de Binche 151, 7000 Mons, Belgium e-mail:
[email protected] Grethe, Harald Department of Agricultural Economics and Social Sciences, Faculty of Agriculture and Horticulture, Humboldt University of Berlin Unter den Linden 6, 10099 Berlin, Germany e-mail:
[email protected] Grilo, Isabel European Commission, Directorate General Enterprise and Industry Avenue d'Auderghem 45, 1040 Brussels, Belgium e-mail:
[email protected] Gual, Jordi IESE Business School Av Pearson 21, 08034 Barcelona, Spain e-mail:
[email protected] Hölzl, Werner WIFO Austrian Institute of Economic Research P.O. Box 91, 1103 Vienna, Austria e-mail:
[email protected]
List of Contributors
xv
Keen, Michael Fiscal Affairs Department International Monetary Fund 700 19th Street, N.W., Washington DC 20431 USA e-mail:
[email protected] Lejour, Arjan CPB Netherlands Bureau for Economic Policy Analysis P.O. Box 80510, 2508 GM The Hague, the Netherlands e-mail:
[email protected]. Leo, Hannes WIFO Austrian Institute of Economic Research P.O. Box 91, 1103 Vienna, Austria e-mail:
[email protected] Meier, Volker Ifo Institute for Economic Research Poschingerstr. 5, 81679 Munich, Germany e-mail:
[email protected] Meurers, Martin Federal Ministry of Economics and Technology, Division Fiscal Policy and Federal Budget, Scharnhorststraße 34-37, 10115 Berlin, Germany e-mail:
[email protected] Oosterwijk, Jan Willem College van Bestuur, Erasmus University Rotterdam PO Box 1738, 3000 DR Rotterdam, the Netherlands e-mail:
[email protected] Pelkmans, Jacques Department of European Economic Studies, College of Europe Dijver 11, 8000 Brugge, Belgium Vlerick School of Management Reep 1, 9000 Gent, Belgium e-mail:
[email protected] Pouget, François University of Paris Dauphine EURIsCO 1 Place du Maréchal de Lattre de Tassigny, 75016 Paris, France e-mail:
[email protected]
xvi List of Contributors
Proost, Stef Department of Economics, Catholic University of Leuven Naamsestraat 69, 3000 Leuven, Belgium e-mail:
[email protected] Renner, Carsten, IDS GmbH - Analysis and Reporting Services, ALLIANZ GROUP Königinstrasse 28, 80802 Munich, Germany e-mail:
[email protected] Stéclebout-Orseau, Eloïse European Central Bank Kaiserstraße 29, 60311 Frankfurt am Main, Germany e-mail:
[email protected] Straathof, Bas CPB Netherlands Bureau for Economic Policy Analysis P.O. Box 80510, 2508 GM The Hague, the Netherlands e-mail:
[email protected] Thissen, Laura ECORYS Netherlands P.O. Box 4175, 3006 AD Rotterdam, the Netherlands e-mail:
[email protected] Van der Horst, Albert CPB Netherlands Bureau for Economic Policy Analysis P.O. Box 80510, 2508 GM The Hague, the Netherlands e-mail:
[email protected] Van der Ploeg, Frederick Department of Economics, University of Oxford Manor Road Building, Manor Road, Oxford OX1 3UQ, UK Veugelers, Reinhilde Katholieke Universiteit Leuven and EC-BEPA Naamsestraat 69, 3000 Leuven, Belgium e-mail:
[email protected]
1 Subsidiarity for Better Economic Reform?
George Gelauff, Isabel Grilo and Arjan Lejour
1.1 Introduction
1
After the successful conclusion of the Internal Market program in 1992, the scope of the European Union has gradually been widened to include areas of public policy that previously remained within the more or less exclusive sovereignty of the Member States. Such areas include monetary and budgetary policy (through the Stability and Growth Pact, SGP and the Economic and Monetary union, EMU), energy and telecommunications, environmental policy, social policy, innovation policy and immigration policy. Although the extent of European involvement widely differs, it seems nevertheless clear that Europe includes increasingly wider elements of the public domain. The causes for this widening of the scope are diverse. Clearly, the introduction of the EMU and the SGP were meant to strengthen the European economy per se by handing over sovereignty in monetary and (partly) budgetary policy to Frankfurt and Brussels. The liberalization trend in Member States implied that (semi) public services (energy, health care) increasingly crossed borders and had to fit within existing European schemes of regulation, competition policy and state aid. With innovation being a target in the Lisbon agenda, European cooperation in innovation policy is being strengthened. Cross border environmental problems in Europe legitimize a common European approach. Other areas are included 1
The views expressed here are those of the authors and should not be attributed to the European Commission or CPB.
2
George Gelauff, Isabel Grilo and Arjan Lejour
in the European agenda to show European citizens that European integration is not simply an ‘economic’ affair, but it also addresses social concerns. Whatever the causes of a more prominent role for Europe are, the consequences are clear: in their public policy making Member States increasingly cooperate. Although this brought new ‘softer’ coordination mechanisms to the fore (such as the open method of coordination), it nevertheless implied that to some extent Member States relinquished part of their national sovereignty in such areas. This process has proceeded considerably. The tendency appears to be that if there is a ‘European problem’ (meaning a problem that regards all Member States) European solutions are to be considered. However, from a strictly economic perspective this Europeanization of public policy is not necessarily justified. The desirability of further European policy actions is guided by the subsidiarity principle. Article 5 of the 1997 Treaty of Amsterdam states: “In areas which do not fall within its exclusive competence, the Community shall take action, in accordance with the principle of subsidiarity, only if and in so far as the objectives of the proposed action cannot be sufficiently achieved by the Member States and can, therefore, by reason of the scale or effects of the proposed action, be better achieved by the Community.” The dynamics of the division of labour between European and national authorities do not always follow the subsidiarity principle. In some cases, the prevailing policy agenda goes beyond what would be desirable from a strictly economic perspective. An examples concerns binding directives on labour standards (maximum working time or minimum rest periods), while empirical support for the risk of social dumping is hard to find.2 In other areas more European involvement may be required than is currently considered. For instance, as part of the Lisbon process the open method of coordination has been applied to the 3% R&D target. However, international R&D spillovers may demand a stronger form of coordination for innovation policy.3 Also in political terms Europeanization has recently gained momentum in the public debate. In some circles, the negative outcomes of the referenda on the European constitution in France and the Netherlands were attributed to perceptions of too much involvement of Europe in national affairs. This would reflect the widely diverging views of European citizens on the desirability of European solutions for different policy areas, the distribution of effects among Member States, as well as the complex and heterogeneous policy momentum in different public domains. However, 2 3
See CPB/SCP (2003) or de Mooij and Tang (2003). See Falk et al. (Chapter 8) and Van der Horst et al. (Chapter 9).
1 Subsidiarity for Better Economic Reform?
3
Ahrens et al. (Chapter 3) confirm the findings of Alesina et al. (2001) that preferences of citizens in the EU15 are ‘on average in line with an economically rational allocation of political domains to national and supranational decision levels’. They mainly observe controversies among citizens of the EU15 on policies relating to immigration, asylum and refugees, defence and the labour market. Yet, Ahrens et al. do find a significant difference between the EU15 and the new Member States, more precisely the EU10. Citizens in the new Member States have a stronger preference for centralization: they prefer EU decision making in 22 out of 25 policy areas, in contrast with 11 out 25 for citizens in the EU15. So as a consequence of enlargement the EU is confronted with increased heterogeneity. The division of competences between the Union and the Member States becomes particularly important in economic reform areas. In these areas substantial policy adjustments take place, which frequently raise considerable challenges for policy makers. The New Lisbon Strategy for Growth and Jobs (EC, 2005) has a large potential to boost Europe’s economic performance.4 Yet, it requires considerable efforts to devise policy measures to realize that potential. That raises the question of subsidiarity: who has the main responsibility for initiation, design and implementation of reform policies, the Union, the Member States in their National Reform Programmes or both? This book reviews subsidiarity in four economic reform areas: education and innovation; the Internal Market and agricultural policy; corporate taxation; and regional and transport policy. These areas cover a substantial part of the key areas of the renewed Lisbon strategy.5 At the same time European policy making concerns such a broad range of policies that it is impossible to give a complete review. The four areas have been selected because of their policy relevance, because they raise interesting new issues concerning subsidiarity and because some economic research is under way to assess subsidiarity in these areas. Moreover, the book takes an economic policy perspective. Authors use economic theory and empirical research in economics to analyse issues of subsidiarity. However, they leave technicalities aside and focus on policy relevant considerations. As such they aim at identifying policy options and policy relevant trade-offs that pertain to subsidiarity. This chapter Gelauff and Lejour (2006) calculate that if Europe would really reach the goals it set in 2000, Europe’s Gross Domestic Product could increase by at least 12%. 5 Agricultural policy is not part of the reform areas covered by the Lisbon strategy although the Common Agricultural policy has been recently reformulated. 4
4
George Gelauff, Isabel Grilo and Arjan Lejour
continues with a brief discussion on the economic analysis of subsidiarity in the next section. Subsequently, in four sections it surveys the main lessons learned for each of the economic reform areas. The last section offers some general conclusions.
1.2 An economic perspective on subsidiarity Ederveen et al. (Chapter 2) present a framework for an economic analysis of subsidiarity. It starts with a functional test for subsidiarity, which centres on the trade-off between preference matching as a motive for decentralization and two motives for centralization: economies of scale and cross-border externalities. Heterogeneity of preferences among Member States is reason for decentralization of decision making in Europe. Yet, when the benefits of acting in common outweigh preference heterogeneity, economic analysis suggests that it is more efficient to shift decision power to a higher level. When policies are costly, for instance due to fixed costs, economies of scale are a reason to combine resources and act together. Common policies in international trade (WTO) or combined funding of large-scale research organizations are a case in point. Cross-border externalities arise when a Member State does not take into account that its policy has consequences for other Member States. These consequences may range from positive spillovers from R&D to negative spillovers from trade barriers in the Internal Market. The differences that sometimes exist between the current practice and the desirable level of coordination according to the functional subsidiarity test raise the issue of the political economy of European coordination. Political economy considerations point at government imperfections or characteristics of the process of decision making in the EU as a second angle to review European coordination. Ederveen et al. include these considerations in their framework to assess subsidiarity. The functional subsidiarity test is based on the premise of benevolent governments and other public actors and on the absence of pressure groups. In reality public actors may not always act in the public interest but may pursue their own objectives, such as expansion of their influence or power. Also they may be (partly) captured by interest groups. Decentralization may then enable citizens to better control public actors or may discipline public actors through policy competition between jurisdictions. In contrast, effective monitoring by Member States may support centralization in the EU. In addition, policy learning may both benefit from experimentation among decentralized authorities and from information exchange and commitment building at a central
1 Subsidiarity for Better Economic Reform?
5
level. Decision making at the centralized level comes with two main risks. Overprovision of locally beneficial policies may result, when decision making takes place under a norm of deference (“I’ll-scratch-your-back-ifyou-scratch-mine”). Common pool problems arise, when Member States have an incentive to draw as much as possible on the common budget for projects that locally provide benefits. Decentralization may prevent these inefficiencies of centralized decision making. Last but certainly not least, European coordination is a dynamic process. In the past various developments, such as increasing pressure on the environment, have called for a coordinated response. In the future, trends may affect the balance between centralization and decentralization in Europe. For instance, internationalization may affect economies of scale in innovation or may increase demand for internationally mobile employees. In addition, dynamic feedbacks exist between policy measures in different fields. Liberalization (for instance of energy markets or health care) brings policy areas within the confines of the Internal Market that previously were delivered by national public providers. Trends and dynamic feedbacks may intensify (diminish) external effects and economies of scale and thus may add to (limit) motives for European coordination.
1.3 Education and innovation In the field of education, subsidiarity primarily concerns higher education. Indeed, students mobility is likely to occur at significant intensity only at this level making potential cross-border externalities or economies of scale only relevant to higher education.6 In addition, higher education relates to innovation, because of the direct linkages between the two areas and their significance for productivity growth. 1.3.1 Higher education Van der Ploeg and Veugelers (Chapter 5) trace out a long series of reform measures in higher education. Governance of universities should provide more room for autonomy, for instance by allowing universities to set differentiated tuition fees. At the same time governance should raise accountability of universities, among other things by providing resources to university on the basis of academic excellence. Higher fees should increase 6
At lower levels of education EU involvement may still be warranted to enhance policy-learning among Member States through open coordination processes.
6
George Gelauff, Isabel Grilo and Arjan Lejour
private investment in higher education, in line with high and rising private rates of return, with income contingent loans providing insurance against the risk of income loss due to unemployment after graduation. As such, these reform measures primarily affect national institutions and are thus principally the responsibility of the Member States. Competition between universities may enhance the quality of education. International competition in higher education points at a potential role for the EU. The main reason is that economies of scale and scope may characterize a European market for higher education (see Figure 1.1). To further integrate the Internal Market for higher education, Van der Ploeg and Veugelers recommend that besides facilitating policy learning, the EU should finance modernization activities by Member States and universities out of the Structural Funds. Moreover, the EU may increase the transparency of the higher education market by urging Member States to implement the Bologna reforms and by promoting cross-recognition of qualifications and a standardized system of quality indicators to evaluate universities. In that way, further transparency may stimulate student mobility, which would create a positive feedback on the process of competition and quality improvement as Figure 1.1 shows.
Competition
scale and scope
among institutes of higher education
Mobility
Quality
of students and researchers
of higher education
transparancy
externalities
Public expenditure on higher education
Fig.1.1. Interdependencies affecting the assessment of subsidiarity in higher education
However, Ederveen and Thissen (Chapter 7) find little empirical support for economies of scale in higher education. The quality of universities in a country is unrelated to the size of the population in that country. Since there is hardly any empirical evidence that larger countries provide better education, extending the higher education market to the EU as a whole is not guaranteed to increase quality.
1 Subsidiarity for Better Economic Reform?
7
Competition between universities only succeeds when students and researchers are sufficiently mobile to move to the best universities (Figure 1.1). Currently, mobility is rather low. Ederveen and Thissen find that geographical and cultural distance plays an important role for students who enrol as regular students in other European countries. Students prefer to study not too far from home. They also react to differences in quality, but the effect of distance is stronger. At the same time international mobility may create free rider problems (Gérard, Chapter 6). Usually in the EU students only pay a (small) part of their education through tuition fees. In that case students, who study abroad and return home afterwards, benefit from investments by the foreign country in higher education, whereas the revenues in terms of increased human capital accrue to the home country. This generates an incentive for the home country to free ride on higher education facilities abroad, which restrains public expenditure in higher education (compare Figure 1.1). The externality disappears when funding follows the students, i.e. when the choice of a student for a specific university abroad results in that university also obtaining the funding to finance the student’s education. Gérard analyses a way to internalize the externality through internationally portable student vouchers. Gérard argues that vouchers could be organized through interjurisdictional cooperation between Member States, limiting the role of the EU to facilitating cooperation, for instance by providing a model treaty. To completely remove the externalities the vouchers would have to cover the full cost of education. However that would give rise to an offsetting externality when after graduation students do not return to their country of origin. In that case the host country benefits from education financed by the home country. Indeed, evidence cited by Ederveen and Thissen shows that student mobility to some extent is a precursor of labour migration. Offsetting externalities may give rise to vouchers that not fully cover the cost of education or to obligations to refund vouchers in case of migration, as mentioned by Gérard. Increasing private returns on higher education that motivate a shift in higher education funding from public to private sources (see Van der Ploeg and Veugelers) also diminish the policy externalities. All in all, intricate interactions between mobility, competition and public expenditure in higher education complicate the assessment of subsidiarity (see Figure 1.1). Currently, student mobility is low. This low mobility may originate from limited transparency or various national institutions that limit competition and investment in higher education. In that sense, pri-
8
George Gelauff, Isabel Grilo and Arjan Lejour
marily national initiatives to reduce these barriers may generate positive feedbacks, which increase economies of scale and externalities and thus ask for further EU involvement. 1.3.2 Innovation Higher education and innovation are linked because graduates from higher education perform R&D and other innovative activities. Gérard (Chapter 6) analyses one of these links in the context of factors mobility. The EU is already active in attracting mobile researchers to conduct research in Europe. However, efforts to stimulate the development of poor regions through innovation may intensify policy competition by Member States in an indirect way, since it will increase regional demand for researchers. That may ask for a stronger EU involvement in promoting innovation in poor regions through regional development policy. Hence, in this case a motive for centralization follows from the complementarity between research policy and regional policy. Falk et al. (Chapter 8) review the extent to which economies of scale or cross-border externalities may manifest themselves in a broad range of policy fields related to innovation and where heterogeneity of preferences is dominant. They provide a taxonomy of policy domains according to these two dimensions and compare the actual allocation of policy competences with the theoretically preferred allocation. Some cases are straightforward, for instance economies of scale clearly concern large-scale R&D projects such as Galileo. Here centralization is warranted according to the subsidiarity principle. Yet, that still doesn’t necessarily imply EU coordination: CERN is a primary example of mutual cooperation in a large R&D project. Also SME innovation policies clearly relate to the national or regional level, which would restrict the role of the EU to supporting policy learning and safeguarding the Internal Market. In some instances, EU initiatives appear to fail a subsidiarity test. For instance, the establishment of a European Institute of Technology is controversial from a subsidiarity perspective. Van der Horst et al. (Chapter 9) search for empirical evidence on economies of scale or cross-border externalities in R&D. For the EU Member States plus the US, Van der Horst et al. find a positive association between the size of the economy and the share of public R&D in government expenditure. This might be an indication for economies of scale in public funding of R&D. The authors find an analogous association between the size of the economy and public funding of private R&D. In addition, they interpret a negative association between public R&D expendi-
1 Subsidiarity for Better Economic Reform?
9
tures and foreign ownership of patents as an indication of spillovers of public R&D abroad. That may point at cross-border externalities, which underscore centralization of public R&D at the EU level. In contrast to these motives for European coordination, Van der Horst et al. draw attention to heterogeneity. This would point at the Member States as the most appropriate level for R&D policy. Heterogeneity pertains to the variety of objectives of public R&D among Member States and to learning from diversity. By consequence, subsidiarity in R&D indeed has to trade off economies of scale and cross-border externalities against diversity among Member States. On one innovation policy field, the Community Patent, the authors clearly argue for centralization from the perspective of subsidiarity. The authors suspect that the difficulties in establishing a Community Patent are the result of protectionism by Member States. In general, Falk et al. and Van der Horst et al. show that it is impossible to assess subsidiarity at the level of innovation policy as a whole. It is necessary to make a careful assessment for each specific policy domain separately. As is the case for many other policies, the devil is in the detail.
1.4 The Internal Market and agricultural policy The contributions on Internal Market Policies concentrate on heavily regulated services such as network industries and banking (Chapter 10) and commercial services (Chapter 11). In addition, agriculture has been included under this heading, because it is linked to the Internal Market. 1.4.1 The Internal Market for services Gual (Chapter 10) analyses the integration strategies in telecommunications, banking and electricity. In all these network sectors the appropriate level of regulation is a fundamental topic. The level of regulation and thus also integration should maintain a level playing field for all competitors and should also protect country-specific strategic interests to varying degrees. This could result in harmonization for some regulations and mutual recognition or host country rules in other domains. In banking, partial harmonization and mutual recognition dominates. In retail banking the use of host country rules could be improved to avoid protectionist devices. In telecommunications, Gual pleads for more use of the mutual recognition principle, which could trigger more competitive deregulation and the spread of best regulatory practices. In electricity, integration has been less
10
George Gelauff, Isabel Grilo and Arjan Lejour
successful. Gual advises more harmonization or regulatory practices with some host country rules in very sensitive sectors. The main message is that there is no single path towards deregulation and integration for the various network industries, and that integration is no goal in itself. Lejour (Chapter 11) analyses the Internal Market for commercial services. At present services markets are hardly integrated although there are no formal trade barriers. The main reason are differences in regulatory practices which hamper cross border trade and foreign direct investment in these sectors. The chapter concludes that further integration is possible but that national practices on regulating services have to be shifted to the EU level to some extent. This shift makes it possible to internalize the negative effects of own regulatory standards on foreign service providers. However the benefits are not equally distributed over the Member States and Member States have different preferences for regulating services. Over time, the benefits of further centralization could increase, among other reasons, due to the globalization trend, the shift towards a services economy, and the tendency to regulate less in Europe. 1.4.2 Agricultural policy The common agricultural policy covers a large part of the EU budget. Grethe (Chapter 12) concludes that the market policies in agriculture should still be a part of the EU budget, because shifting these to a lower level of government would disturb the functioning of the Internal Market for agriculture. The other important budget items are direct payments and rural developments policies. The analysis suggests that there are no clear economies of scale or internalization of externalities if these activities are conducted at the EU level. Because these policies are to a large extent a transformation of past market and intervention policies, it is understandable that these policies were part of the EU budget, but economic reasoning suggests that it is sensible to shift these policies to the Member States in the future. Common pool problems for rural development policies and direct payments in particular are also a reason to concentrate spending at the Member State level. In particular for rural development policies Member States could learn from each other about which policies work and which do not. The advantage of regional and national policies is that regions and countries could experiment with these policies. Involvement of the EU could be useful to create platforms to exchange information, practices and results in these areas.
1 Subsidiarity for Better Economic Reform?
11
1.5 Taxation Keen and De Mooij (Chapter 13) give an overview of the main arguments behind the criteria of functional subsidiarity in the field of taxation. The argument pointing at decentralization – heterogeneity of preferences – is clearly recognized. Indeed, the differences across countries in preferences for government expenditure7 lead naturally to differences in revenue needs and therefore to different choices in taxation. This link is further exploited by Pouget and Stéclebout-Orseau (Chapter 16) who look at the implications of tax competition in the presence of heterogeneous preferences for public goods. On the pro-centralization arguments Keen and De Mooij review the presence of fiscal externalities and its possible forms as well as its likely welfare implications. The existence of scale economies, mainly recognized in the administration and compliance domains, is presented as playing a rather modest role, leaving the bulk of the justification of a possible need for coordination to the presence of externalities. Among the various forms of externalities the one arising from “base snatching” and the consequent tax competition leading to lower rates is seen as deserving particular attention in the context of capital income taxation, due to the high mobility of capital relative to labour. For this reason, the pros and cons of tax competition become a crucial element when discussing the merits of some form of centralization in capital taxation. Keen and De Mooij discuss some of the arguments in favour of tax competition, in particular due to its role as a disciplining device. By confronting the lessons from the theoretical and empirical literature on fiscal federalism with the EU experience in the area of tax policy coordination, Keen and De Mooij give an appraisal of what they see as successes in the EU efforts to coordinate tax policy, namely in the area of the value added tax. However, they suggest that this may have come at the expense of sufficient effort to address the coordination of capital income taxation in earlier stages of the EU. As a lesson to future unions, Keen and De Mooij suggest that capital tax coordination is better dealt with in early stages when membership is more homogeneous and lobby interests are weaker. The three chapters following Keen and De Mooij’s look into detail at corporate taxation, arguments in favour or against tax coordination and the specificities such coordination should take.
7
These differences in preferences for government expenditure items have been discussed for the fields of education and innovation in Chapters 5-9.
12
George Gelauff, Isabel Grilo and Arjan Lejour
Cnossen’s Chapter 14 goes the furthest in dealing with the specificities of corporate tax coordination. Though acknowledging the potential role of tax competition as a disciplining device, Cnossen does not view tax competition as wholly beneficial, in line with Keen and De Mooij, and argues for a gradual, bottom-up and partially reversible coordination over a topdown harmonization approach. He also stresses the importance of a broadly based approach encompassing the taxation of all forms of capital income rather than a narrow coordination of corporate profits. More precisely, Cnossen proposes an agenda for capital income taxation coordination in five sequential steps whose final completion would lead to a European capital income tax. However, he recognizes that this last final step is, in the present circumstances, unfeasible. His agenda’s measures aim at mitigating the distorting effects on corporate financing and investment caused by the differential in corporate and personal income tax rates and at reducing incentives for transfer pricing manipulation and thin capitalization. Van der Horst (Chapter 15) addresses specifically the externalities associated with the two main “parameters” of corporate taxation: tax rates and tax base. The economic effects and the distributional implications of tax reforms are then investigated by assessing the sign and size of spillovers associated with unilateral tax policy and by factoring in the scale economies for firms allowed by tax base consolidation using a general equilibrium model. These exercises lead to an assessment of the benefits/costs of tax rate harmonization and/or tax base consolidation. Van der Horst’s simulation exercises point at limited gains from tax-base coordination. Simulations on the consolidation of the tax base via formula apportionment suggest very uneven economic effects both across and within countries with small overall gains leading to the suspicion that the details in the design of such consolidation may in the end determine whether such gains materialize. In a far-reaching scenario were tax rate harmonization adds to tax base consolidation an overall gain would occur accompanied by large variation across countries with small tax base countries losing. This result suggests that consolidation combined with tax-rate harmonization could pass the subsidiarity test if complemented with a proper redistribution scheme. Tax rate harmonization without tax base consolidation does not seem to be a useful strategy because simulations suggest that, starting from the present situation, further tax rate competition would hardly decrease rates, therefore casting doubts on a “race to the bottom”. Pouget and Stéclebout-Orseau (Chapter 16) devote their attention to the implications that corporate tax competition has for the composition of public spending. The authors question the standard argument that tax competition is harmful because it erodes tax revenues and therefore the govern-
1 Subsidiarity for Better Economic Reform?
13
ment’s ability to produce the optimal level of public good. Instead, they turn to the more subtle concept of distortions in the composition of public expenditure. These distortions are caused by the existence of a second dimension of competition: the provision of a productivity-enhancing public good as a means to attract investment and consequently a broader tax base. Interpreting increased tax competition as the result of easier profit shifting, they find that tighter tax competition leads governments away from productivity-enhancing spending. Conversely, tax-rate coordination would not only lead most countries to increase their statutory corporate tax rates, but also foster a re-allocation of public expenditure in favour of productivityenhancing spending. Although the welfare and policy implications of this corollary are not directly addressed in the chapter, a move away from expenditure in fields such as health and social transfers may have a negative impact on households. However, as the authors recognize, further investigation would be necessary to assess this intuition.
1.6 Regional and transport policy Regional policy takes subsidiarity a step further than the levels of the Union and the Member States to issues concerning decentralization to the regional level. Transport policy has a regional dimension as well. 1.6.1 Regional policy Begg (Chapter 17) discusses the rationale for an EU regional policy. Regional policy is the mainstay of cohesion policy, the aim of which is to improve the competitive position of regions. Although redistribution, in the sense of improving the current incomes of poorer citizens, is not the aim of cohesion policy, in practice the richer Member States transfer resources to the poorer ones in this way, a rather clumsy way to redistribute income, as Begg states. Cohesion policy is, along with the Common Agricultural Policy, the most important budget item of the EU. For some (poorer) Member States the payments can add up to a few percentage points of GDP each year. Structural Funds’ spending is not limited to the poorer Member States, but also goes to poor regions in rich Member States. For the latter case Begg doubts the rationale for EU intervention for two reasons. First, richer countries have the funds to finance regional policy themselves. Second, assignment of regional policy at the Member State level may give a better match with regional preferences and there are lower transaction costs involved.
14
George Gelauff, Isabel Grilo and Arjan Lejour
Begg argues that the case is different for Structural and Cohesion Fund payments to poorer Member States for at least three reasons. First of all, the poorer countries have fewer funds available. The transfer of resources has to be orchestrated by a higher level of government. Second, a higher level of government can more easily handle the competition between regions to improve competitiveness. Third, the higher government level can impose governance conditions such that support is used appropriately. This could also stimulate and help poorer Member States to develop their institutional capacity. Although cohesion policy is in principle temporary - it finishes when poorer countries catch up - Begg foresees a resilience of cohesion policy for constitutional and political economy reasons. Overall Begg suggests that a nuanced approach is needed: there are sound arguments for either or both of the EU and Member State levels having a role. The balance of these arguments is also affected by the disputed effectiveness of cohesion policy, as the empirical evidence is mixed and inconclusive about the ability of regional policy to have an enduring effect on regional prosperity. However, most of these evaluation studies have their limitations as well. Fenge and Meier (Chapter 18) add another argument to the rationale for an EU regional policy, which is not based on altruistic motives of the richer Member States but on their self interest. They argue that EU regional policy can be used as a means to prevent ‘immigration into unemployment’. Transfers to poor EU countries that support infrastructure investment make staying in these countries more attractive and allow them to catch up faster. High wage countries are interested in spending money in this way to avoid immigration that raises unemployment of natives in an inflexible labour market. Concerning the structure of regional policy, donating countries prefer matching grants for investment in infrastructure to unconditional grants or wage subsidies. Investment subsidies lead to faster wage growth in the poor Member States of the EU, which reduces both migration flows and unemployment in the immigration countries. Bilateral agreements between high and low wage countries will not be efficient, because the high wage countries do not take the benefits of less immigration to other high wage countries into account. This externality has to be dealt with at the EU level. Fenge and Meier do not distinguish different regions in the low and high wage country. In their view regional policy does not aim at reducing disparities. The only reason to choose public investment is that this is more effective in their model than redistribution in the form of wage subsidies.
1 Subsidiarity for Better Economic Reform?
15
1.6.2 Transport policy De Borger and Proost (Chapter 19) study the role of the EU in transport policy, in particular in providing transport infrastructure, and the interaction with instruments to price the use of the infrastructure. First, they conclude that the desirability of European funding for infrastructure projects crucially depends on the importance of cross-border transport. Subsidies are not needed from an efficiency viewpoint if there is no transit transport. Second, they show that countries will charge excessively high tolls and strongly under-invest in transport capacity, if they are allowed to determine tolls for the use of their infrastructure. This conclusion only holds if Trans European Networks are interpreted as serial transport corridors (which is often the case for railways and canals) and not as parallel corridors (often the case for roads). To avoid high tolls, insufficient investment and large welfare losses, it is better for the EU not to allow Member States to freely decide on tolls at all in the serial case. The third conclusion is that the EU should also intervene in providing infrastructure for serial transport corridors, because capacity is underprovided. De Borger and Proost suggest linking the provision of financial support to marginal social cost pricing by the Member States. According to their calculations subsidies should depend positively on the share of trans-border transit through the country and on the degree of scale economies in capacity provision, and negatively on the cost of public funds. De Borger and Proost focus on serial corridors. Their conclusions suggest that the EU has a substantial role in developing infrastructure and the pricing of railways and canals. In the case of roads drivers have often more alternatives over parallel roads such that the external effects of Member States policies are much smaller.
1.7 Conclusion Can the application of the subsidiarity principle be helpful for developing economic reform policies in line with the Lisbon agenda? There is no easy answer to this question. An overview of all contributions in this book yields a number of general conclusions for specific policy fields. The pros and cons of acting in common in Europe as well as a number of caveats are also suggested. In particular, the contributions in this book pinpoint the following difficulties: 1. In a number of cases political economy arguments add to the explanation of the current division of responsibilities between the Union and the Member States. For instance, juste retour arguments, lobbies, pol-
16
George Gelauff, Isabel Grilo and Arjan Lejour
icy learning and path dependence to some extent affect agricultural and cohesion policy. 2. Often the devil is in the detail. The assessment of subsidiarity depends on specific characteristics of a policy area and on complex interactions between determinants of economies of scale or cross border externalities (see for instance Section 1.3). Hence, general statements about too much or too little Europe are hard to substantiate. 3. Subsidiarity does not always imply a clear-cut division of responsibilities. In several cases the EU and the Member States may act together. Policy learning through the open method of coordination is a straightforward example. 4. Subsidiarity is a dynamic process: change is driven by various trends. Diversity leading to more heterogeneity is one of these. Diversity relates to the increasing number of Member States and differences in their economic structures. That would require more decentralization to match local preferences. Moreover, the globalization trend could have substantial effects on economies of scale and externalities related to size. 5. Changing circumstances may require abandoning earlier policies for centralization. Yet, this may be difficult, because for some Member States more will be at stake than for others (see for instance Grethe, Chapter 12). That underscores the option value of waiting and the use of sunset clauses (Oosterwijk, Chapter 4). In spite of these difficulties we can draw some conclusions from the application of the subsidiarity principle on education and innovation; the internal market and agricultural policy; corporate taxation; and regional and transport policy. The quality of higher education and thus human capital has to be improved if the EU is to reach its ambitious Lisbon goals. Here is a possible role for the European Union. Increasing mobility of students could lead to more competition and possibly higher quality in education. Moreover, the European Union could act as catalyst for the Member States’ reforms of the higher education sector. The analysis of R&D policies shows that there are clearly defined roles for the Member States and the EU. The EU has a role in funding public R&D and in subsidizing private R&D, because of the externalities involved and the potential benefits of economies of scale. Government support for R&D by small and medium size enterprises can be better conducted at the national or even regional level. These government layers have in general better information about local circumstances and there are no substantial cross-border externalities involved.
1 Subsidiarity for Better Economic Reform?
17
With respect to Internal Market Policies, the analyses suggest that the common market for heavily-regulated and commercial services could be improved, because of the external effects of national regulatory practices. Concerning the common agricultural policy, rural development policies and direct payments can be better carried out by national or regional government layers. Although the current divisions of tasks between the EU and the Member States can be understood from the past, EU involvement in rural development and income policies is overall inefficient and could lead to crowding out of EU budgets. EU involvement should be limited to market policies and agri-environmental policies that address cross-border externalities. In earlier stages the EU has been successful in coordinating value added tax policy, but not in the coordination of capital income taxation. This book argues for further coordination of capital income taxation. Although a step by step approach seems to be advisable, the benefit of coordination probably only appears if tax rates and the tax bases for capital income are coordinated. However, tax coordination does not curtail policy competition. Member States are still able to compete with productivity-enhancing public spending such as infrastructure, to attract economic activity. The EU has also a clear role in regional policy, as long it is directed to regions in poorer Member States. These states often lack the financial and institutional capacity to develop the regional economy and the EU can handle competition between regions. It could also be in the self interest of richer Member States to stimulate the economy in poorer ones in order to avoid some undesirable consequences of massive migration. For the stimulus of poor economic regions in richer Member States, these Member States themselves could take the lead. The book also clarifies the role of the EU with respect to transport policy, in particular on Trans European Networks. For railways and canals, the so called serial transport corridors, the EU should take the lead in order to generate sufficient capacity and to avoid excess pricing. For road infrastructure this is different. Because of the many alternative parallel roads the Member States should take the lead. Besides these conclusions, the contributions in this book have indicated at least two main directions for further research: empirics and political economy. Indeed empirically, much is still unknown about the optimal level of decision making in the EU. Often assessments of subsidiarity present a range of theoretical arguments related to the functional subsidiarity test. Yet, determining the position on the trade-off between decentralization and centralization and deriving policy options ultimately is an empirical question. Adding political economy arguments to the trade-off further complicates the assessment. Some authors (for instance, Oosterwijk, Chapter 4;
18
George Gelauff, Isabel Grilo and Arjan Lejour
Gual, Chapter 10; Lejour, Chapter 11; Grethe, Chapter 12; Begg, Chapter 17; Fenge-Meier, Chapter 18) explicitly analyse political economy considerations from a practical or theoretical perspective. Generally, empirical analyses of these motives are way out of sight. Possibly case studies of actual decision making in EU councils and committees may be an interesting first step to follow-up in this direction. All in all, the contributions in this book show that subsidiarity is an intricate and challenging topic, both for economic analysts and for policy makers. The economists can provide policy makers with a range of arguments to support their judgments on issues of subsidiarity. However, they cannot and probably will never be able to provide detailed recipes on how to allo-cate responsibilities between the EU and the Member States. That is all for the best, since in a democracy subsidiarity will ultimately be a political decision, yet economic analysis may contribute to it being as informed as possible.
References Alesina A, Angeloni I, Schuknecht L (2001) What does the European Union do? NBER Working Paper 8647, Cambridge, MA CPB/SCP (2003) Social Europe. In: European Outlook 1, The Hague De Mooij R, Tang P (2003) Four futures of Europe, CPB, The Hague European Commission (2005) Working together for growth and jobs A new start for the Lisbon Strategy, Communication to the spring European Council, COM (2005) 24, Brussels Gelauff G, Lejour A (2006) Five Lisbon highlights: The economic impact of reaching these targets. CPB Document 104, The Hague also published as The new Lisbon Strategy: An estimation of the economic impact of reaching five Lisbon Targets, Industrial Policy and Economic Reforms Papers, no 1. Enterprise and Industry Directorate-General European Commission, Brussels
2 Assessing Subsidiarity
Sjef Ederveen, George Gelauff and Jacques Pelkmans
2.1 Introduction In order to provide inputs for an assessment of subsidiarity in specific policy fields, this chapter reviews the main motives for decentralization and centralization. Several motives underscore decentralization, such as the ability to align policy closer to the preferences of regional constituents or the effect of policy competition to discipline governments. In contrast, centralization may be warranted in a policy field characterized by increasing returns to scale or it may be useful if thereby it could create commitment for economic reform. However centralization may also fail to be useful, for example when local jurisdictions attempt to extract funds from the budget of a higher level of government. In the European context the principle of subsidiarity was only introduced quite recently. Although it is now generally acknowledged as the guiding principle for dividing powers in the European Union, it was hardly ever mentioned in the official European texts until the late 1980s. Fears of centralized power then led to the idea to place the burden of argument with the advocates for further European integration. This idea was especially supported by Britain and Germany. Britain feared European federalism, and the German Länder sought to maintain their exclusive powers enjoyed in the German Federal Republic (Føllesdal 1998). The first time that the principle of subsidiarity is explicitly mentioned is in the European Single Act of 1986 (article 130r ad 4), dealing with environmental policy. Finally, in 1992, the principle of subsidiarity was officially introduced in the 1992 Maastricht Treaty (article 3b), and moved to article 5 in the 1997 Treaty of Amsterdam. The formulation in article 5 comprises three related princi-
20
Sjef Ederveen, George Gelauff and Jacques Pelkmans
ples, the principle of conferral, the principle of subsidiarity and the principle of proportionality: 1. “The Community shall act within the limits of the powers conferred upon it by this Treaty and of the objectives assigned to it therein.” 2. “In areas which do not fall within its exclusive competence, the Community shall take action, in accordance with the principle of subsidiarity, only if and in so far as the objectives of the proposed action cannot be sufficiently achieved by the Member States and can, therefore, by reason of the scale or effects of the proposed action, be better achieved by the Community.” 3. “Any action by the Community shall not go beyond what is necessary to achieve the objectives of this Treaty.” Note that the principle of subsidiarity is neutral about the optimal degree of centralization. In the debate about subsidiarity, people often confuse the concept of subsidiarity with delegating power to the lowest possible level. However, it would be a mistake to think of subsidiarity and lower level decision making as synonyms. The subsidiarity principle involves a careful assessment of the optimal level at which decisions should be taken, which can result in centralization but also in decentralization. The motives for centralization and decentralization follow from the literature on fiscal federalism, including recent analyses that take a political economy perspective.1 Section 2.2 reviews the basic arguments starting from the premise of a benevolent government. Section 2.3 adds motives related to government failure and policy learning. Section 2.4 focuses on the political process between a number of countries that delegate representatives to a central legislature, which makes decisions at the centralized level. It identifies several failures of centralization that originate from negotiations and conflicts of interests in the legislature. Section 2.5 concludes with an overview of the major motives relevant to an assessment of subsidiarity.
1
However, the European setting where centralization has to be decided, step-bystep, from below, differs from the decentralization issues in federations. This contrast between the centralization questions in the integration process and the decentralization ones in federations implies that care is required when adopting fiscal federalism as the analytical framework for subsidiarity in the EU.
2 Assessing Subsidiarity
21
2.2 Fiscal federalism: the basics The economic arguments that guide the debate on subsidiarity are rooted in the theory of fiscal federalism.2 The economic theory of fiscal federalism explores when centralization of public economic functions is welfare improving. This theory finds its foundation in Musgrave (1959), who discusses the optimal level of centralization of several public economic functions. The basic or first generation (according to Oates 2005) theory of fiscal federalism explores the trade-off between preference matching and internalization of interjurisdictional external effects or economies of scale. To arrive at this trade-off, Section 2.2.1 presents the assumptions on government behaviour that underlie the standard model, Section 2.2.2 focuses on preference matching as a motive for decentralization and Section 2.2.3 turns to the motives for centralization. Section 2.2.4 then discusses a functional subsidiarity test to determine the optimal assignment of competences within the European Union. 2.2.1 Government behaviour The point of departure is a region, which consists of a number of jurisdictions. Clear examples include the US, Germany, Switzerland or Australia, all federations. The EU and its Member States are, of course, not a federation but similar questions of assignment of powers to the two levels of government do present themselves in a number of policy areas. The government pursues some public policy. Then the principal question is whether the centralized government (EU) or the local governments (the Member States) should design the policy. The basic theory as elaborated in Oates (1972) is based on three assumptions on the behaviour of governments: Benevolent and perfect government
Each level of government maximizes the welfare of its constituency. This implies that different layers of government always act benevolently. They do not pursue their own interest or fall victim to lobby groups. Moreover, 2
Fiscal policy was the main topic of research in the original literature. Although economic federalism would be a more accurate description nowadays, ‘fiscal federalism’ has become the accepted term and therefore we will use this here. In the EU-context the theory is also referred to as the theory of multi-tier or multilayer government.
22
Sjef Ederveen, George Gelauff and Jacques Pelkmans
the implementation of policies may be more or less costly, but policy design involves no transaction costs nor are there costs of policy learning. Policy uniformity
The central government pursues a uniform policy in all jurisdictions. This assumption is not straightforward.3 A perfectly informed central planner would be able to provide policies in each jurisdiction according to local preferences or local needs. Two motives underlie policy uniformity (Oates 2005). Firstly, information gathering comes at a cost and central governments face higher costs to learn the preferences of local constituencies compared to local governments. Secondly, a central government faces political constraints to treat one jurisdiction more generous than another. Centralized decision making
Each level of government acts as a single decision maker. This implies that the highest level of government consists of a president or small executive council (the single decision maker) elected by all citizens of the union. (Inman and Rubinfeld 1997, 2002) refer to this form of federalism as economic or centralized federalism. 2.2.2 Decentralization: preference matching In the standard model the principal motive for decentralization follows from the diversity of preferences for public policies in jurisdictions. Both preferences for policies and circumstances that affect the demand for policies differ widely in jurisdictions. Concerning preferences, countries may have different views on the role of the government and on the need for centralization or decentralization (for instance, see Dekker and Ederveen 2005). Alesina and Perotti (2004) emphasize the distinction in Europe between the dirigiste attitude, characterized by heavy government intervention in markets, and the more laissez-faire Anglo-American attitude. Cultural characteristics may underlie this diversity of preferences (compare Dekker et al. 2006). In addition, European countries differ in physical conditions, sectoral structure, infrastructure, etc. This may affect the emphasis they put on different policies. For instance, countries with a large agricultural sector may attach a higher weight to agricultural policies than other 3
With the great heterogeneity in the European Union, policy uniformity is often far from optimal (see e.g. CESifo (2003) for an illustrative example of these problems).
2 Assessing Subsidiarity
23
countries. As an other example, a country with a large transport sector benefits relatively more from investment in infrastructure. Hence, diversity in endowments may result in diversity in weights attached to specific goals of economic policy. When preferences differ over jurisdictions, preference matching is the main motive for decentralization. Combining the assumptions on policy making in the previous section with the likelihood of preference heterogeneity, Oates (1972) derived formally the so-called decentralization theorem. The intuition behind the decentralization theorem is that except in the extreme case that preferences across jurisdictions are the same everywhere, decentralization will bring welfare gains as policies can be diversified in accordance with local preferences and conditions. Preference matching as a motive for decentralization has to be weighted against the motives for centralization, which are the subject of the next section. 2.2.3 Centralization: cross-border externalities and scale Cross-border externalities and economies of scale are the two main motives that may call for centralization (Oates 2005). In the EU cross-border externalities arise when a national policy of a Member State has consequences for another Member State that are not taken into account in its decision making process. These externalities can both be positive or negative. Investments in Research and Development (R&D) provide an example of possible positive externalities. When Germany doubles its investments in R&D, other countries, especially the neighbours, could benefit from the spread of knowledge (Ederveen et al. 2005). However, these potential benefits are not taken into account by Germany when it decides about its level of investment. Therefore, from an efficiency point of view countries will generally invest too little in policies that generate positive spillovers. Higher levels of government can internalize these positive externalities. This could justify the centralization of a certain policy. In the same vein it could be argued that policies that create negative spillovers will be overrepresented. A major example is the EU internal market. Barriers to trade or foreign direct investment in one EU Member State prevent companies from other Member States to enter the market or raise costs, because multinational companies have to comply with different regulations or technical requirements in each Member State. At the EU level the barriers hold back competition, which limits incentives for innovation and productivity growth. These negative externalities are internalized in the EU internal market. Internal market policies mainly consist of promoting the free movement of goods, services, capital and labour. As
24
Sjef Ederveen, George Gelauff and Jacques Pelkmans
such these policies restrict the leeway of Member States to intervene in their economies.4 Economies of scale provide another rationale for centralization. When policies are costly, for instance due to fixed costs, decentralized policy making is bound to be suboptimal. In such a case, centralized policies could improve welfare. Examples are a common European policy in international trade negotiations, a uniform European patent system, or European research organizations such as CERN, the particle physics laboratory near Geneva that is supported by many European countries.5 2.2.4 The functional subsidiarity test Combing the motives for centralization and decentralization results in the following trade-off: Decentralization preference matching
<=>
Centralization cross-border externalities economies of scale
Given a certain degree of preference heterogeneity, centralization is only desirable when externalities or economies of scale are sufficiently large.6 Based on this principal trade-off between preference matching versus scale and cross-border externalities, Pelkmans (2005a) derives a functional subsidiarity test. The general test consists of the following three steps:7 x Do cross-border externalities or economies of scale justify centralization? x Is credible voluntary cooperation possible? Hence, in the context of the EU centralization often involves ‘negative integration’, i.e. removing restrictions on the free movement of goods, services, labour and capital. 5 For an overview of 7 comparable intergovernmental research organizations see the footnote on p 307 of Pelkmans (2006). 6 Formally larger externalities imply more centralization, but more preference heterogeneity does not necessarily imply less centralization. This is because more heterogeneity may not only raise the social costs under centralization, but may also raise the costs of inefficient provision of policies when externalities are not taken into account under decentralization (Besley and Coate (2003), p 2616). 7 Pelkmans (2005a) includes another step which determines whether the policy falls in the area of shared competences. To make the test more general, i.e. not bound by the treaty’s text, we exclude this step here. 4
2 Assessing Subsidiarity
25
x At which level can policies be designed and implemented in a cost- minimizing manner? When economies of scale or cross-border externalities exist, the first step, the need-to-act-in-common test is passed. Still there is no need for centralization, if all Member States would voluntarily cooperate on a given policy issue and this cooperation is credible. Credibility is an important criterion to decide upon centralization or decentralization. When credibility is low, cooperation will be unsustainable and a central policy is needed. Credibility will be particularly low when information is highly imperfect or asymmetrically distributed, especially in complex policy areas, because this renders it impossible to monitor compliance. Credibility is also low when the incentives to cheat are strong and the ability or willingness to impose collective sanctions is perceived as minimal. If voluntary cooperation cannot come about or it would not be credible, there is a case for centralization. When there is a case for centralization, it has to be decided how policies should be implemented, monitored and enforced. This is the third and last step of the test. Here the question of proportionality is relevant: no more than what is necessary to achieve the goals of the actions should be done at the central level (compare Section 2.1). Where possible and efficient, Member States should play the primary role in policy implementation. Therefore, in principle coordination and recommendations are preferred over legislation (compare Section 2.4 on EU economic governance). If binding measures are needed anyhow, directives should be considered before EC regulations. In this way maximal discretion is left to the Member States, while internalizing the cross-border externalities. The test described above is a purely functional test of subsidiarity. Such a test is indispensable for a proper assessment of the assignment decision. Taking the EU context explicitly into account, Pelkmans (2005a) provides numerous practical examples where the functional subsidiarity test is directly helpful. However, it is not the full story. The trade-off between preference matching and cross-border externalities or economies of scale is derived under a set of rather strict assumptions regarding the political process. Relaxing the assumptions of Section 2.1 yields a more realistic and at the same time more complex view of the choice between decentralization and centralization. That is the subject matter of the next sections.
26
Sjef Ederveen, George Gelauff and Jacques Pelkmans
2.3 Government imperfections Dropping assumption 1 of benevolent and perfect governments, imperfections in policy formulation come to the fore. Pelkmans (2005a) distinguishes two well-known cases of government failure: Leviathan and lobby. A Leviathan government pursues its own interests in contrast to the public interest. Self-interested governments may exist on a local level and on a central level. Therefore, Sections 2.3.1 and 2.3.2 review the consequences of Leviathan governments in the Member States and the EU, respectively. Successful lobbying may result in governments being captured by organized interest groups. This is the subject of Section 2.3.3. As a third case, policy design may be characterized by imperfections. Also when governments do not deliberately pursue their own interest, they still may struggle with policy formulation. Section 2.3.4 turns to the issue of (de)centralization when policy design is a learning process, which requires appropriate and sufficient incentives to take place effectively. Finally, Section 2.3.5 addresses complementarities between policy fields, which may create dependencies between centralization decisions over policy fields. 2.3.1 Leviathan in the Member States Decentralization may contain Leviathan either through stronger opportunities for citizens to control government (voice) or through the exit option: voting with their feet. In the latter case citizens leave jurisdictions with Leviathan governments for jurisdictions with more reliable governments or policies closer to their preferences. As Tiebout (1956) showed, competition among local regions is welfare improving for society as a whole, as the resulting heterogeneity allows citizens to move to the region that reflects their preferences with respect to taxes and public goods best. However, mobility of labour between European countries is not so large that exit would be an important mechanism (Pelkmans 2005a). In addition, decentralization enhances policy competition (Tabellini and Wyplosz 2004). Policy competition has costs and benefits. One of its benefits may be to contain Leviathan governments. For instance, if governments have distorted incentives to increase spending and taxation, tax competition may raise welfare. Inter-regional competition also forces governments to provide services efficiently. Hence, decentralized provision of policies and the resulting competition may alleviate government failure. These arguments pro decentralization add accountability to the decentralization side of the trade-off. Because they are more responsive to ‘voice’ and constrained by policy competition, decentralized governments
2 Assessing Subsidiarity
27
become more accountable to their constituents. With Leviathan governments ‘decentralization may be preferable even in cases of perfect homogeneity of preferences across local jurisdictions’ (Oates 2005). Adding this motive for decentralization, a trade-off arises between on the one hand preference matching and accountability as arguments for decentralization and on the other hand externalities and increasing returns as arguments for centralization: Decentralization preference matching accountability
<=>
Centralization cross-border externalities economies of scale
Applying the trade-off to concrete policies, the drawbacks of policy competition should be taken into account. An expanding literature argues that policy competition has its limits as a way to contain Leviathan and to improve the efficiency of policy making. Sinn (2003) states that the so-called selection principle fetters policy competition. National governments intervene in the economy to counter market failures. Yet, according to the selection principle, competition between governments through the backdoor may bring back the market failures national policies were meant to resolve. For instance, competitive states may have an incentive to promote national champions, which may be at odds with national competition laws. Moreover, policy competition lacks price signals that guide competition on the market. Hence, if market failures manifest themselves on an international scale, policy competition between countries may be a blunt instrument to enhance the quality of policies and institutions. Precisely because of the double risk of policy competition (a return of market failures and restrictions for the mobility of goods, services, capital or persons between countries) some well-designed degree of moderate centralization can be usefully combined with policy or regulatory competition. That is exactly what the EU is trying to achieve when it fosters mutual recognition in goods markets, and – to a lesser degree – in services markets (see Pelkmans 2005b). The EU level protects the free movement of goods and services in the internal market against restrictions, whilst the principle of mutual recognition may foster regulatory competition between the Member States, but on the condition that the regulatory objectives of the national rules are 'equivalent'. The equivalence condition pre-empts a race to the bottom, hence, prevents market failures through the backdoor.
28
Sjef Ederveen, George Gelauff and Jacques Pelkmans
2.3.2 Leviathan in Brussels In the context of the EU, Leviathan–type behaviour may arise, if the Commission pursues overly centralistic policies and is able to implement these.8 For instance, Pollack (2003) in an empirical study on delegation in Europe concludes: ‘Despite their internal complexity and diversity, EU supranational agents generally behave like unitary actors with preference for further integration’. To display that behaviour, the Commission should have both incentives for centralization and sufficient discretion to achieve centralization. Incentives may come from a relatively pro-centralization stance of people working at the Commission. Through self selection people applying for a job at the Commission may be above average pro centralization. Socialization during their working life with European ideals and interests may add to this attitude. Of course bureaucratic policies may also play a role, such as expanding the power of directorates. Finally, national governments may delegate people with a relatively pro centralization mind-set to counter time inconsistency (Majone 2001). The reason is that national governments know that they may have short run incentives to renege on their commitment to European integration. Delegation to the EU level solves this only if the preferences of the delegates do not simply mirror those of the delegating authority. Otherwise delegated EU officials would give in to the same short-run incentives as national governments. Hence, the presence of pro-centralization officials may not necessarily be a sign of Leviathan in Brussels, but may be a useful counterweight to Member States, in particular to strengthen the role of the Commission as ‘guardian of the Treaty’. Discretion of the Commission to promote centralist policies depends on monitoring and control activities by Member State governments. Free rider problems and diverging interests among Member States may hamper monitoring and control. The free-rider problem is well-known in principal agent models with multiple principals (the Member States) and costly monitoring of an agent (the Commission). Each principal faces an incentive to avoid monitoring and control costs and to free ride on activities by other principals. Diversity among principals enables the agent to exploit different or even conflicting preferences. In some cases it may use some 8
This argument can also be exaggerated. National governments can use the EU as a crowbar or scapegoat to convince their constituency that unpleasant policies are unavoidable. Recent experiences in some Member States have shown that such a strategy may ricochet in undermining the public support for European coordination.
2 Assessing Subsidiarity
29
kind of ‘divide and rule’ strategy. That may enable the Commission to reach its goals through selectively addressing Member States that may oppose its proposals (Scharpf 1999).9 The Commission may first address a single Member State, for instance by acting against protectionism or state support. If it succeeds (in court) that may reduce opposition by other Member States. Moreover, the country that lost its case is likely to become an ally of the Commission. If it has been convinced or forced to act unilaterally and expects to be at a disadvantage to other countries that have not acted yet, it will side with the Commission in proposing a general directive that applies to all Member States. In addition, analogous functional separation within the Commission and the relevant Council formations may weaken control by Member States, because the policy makers involved may share a common view with respect to the policy area under consideration. Both Commission Directorates-General and Council formations are organized along comparable policy areas. For instance, in the Agriculture Council DG for Agriculture and Rural Development and Ministries of Agriculture meet; in the Employment, Social Policy, Health and Consumer Affairs Council the same applies to DG Employment and Ministries of Social Affairs and Employment. Each Council formation operates in relative isolation and representatives discuss issues relevant to their policy area from their perspective. Exceptions are the Competitiveness Council and the European Council of heads of state and government. In particular the latter Council would be able to provide an overall view and exercise countervailing power. However, the European Council discusses only controversial guidelines at length. Many proposals from Council formations do not reach the European Council or are endorsed without much review. Although at this stage of European integration the notion of Leviathan in Brussels remains rather far-fetched and although the 'guardian of the Treaty' role, which the Commission must fulfil and where it has hardly any discretion, is a predominant one, a principal-agent approach brings out a number of risks for possible EU government failure, if monitoring and control by Member States is unsuccessful. For specific policy fields policies may become overly centralistic, depending on the diversity of views among Member States.
9
Successfully pursuing such a strategy requires a strong legal base.
30
Sjef Ederveen, George Gelauff and Jacques Pelkmans
2.3.3 Lobby Lobbying is an other case of government failure. Governments always risk to become captured by interest groups. That risk exists both at the national level and at the EU level. The theoretical literature does not unequivocally support the ‘traditional intuition that local government is more susceptible to capture by lobbies’ (Lockwood 2005). Among others, it depends on whether citizens are better informed at the central level or at the local level and on the strength of the lobby at each level. Moreover, it may be more cost effective to lobby central policy makers, because that involves less players. Furthermore, lobbying may influence the decision to centralize. In some cases capture may prevent beneficial centralization. Pelkmans (2005a) gives the example of national interest groups that through national governments obstruct internal market liberalization. In contrast, with reference to the common agricultural policy he also illustrates that national interests may prompt centralization. Tabellini and Wyplosz (2004) argue that the impact of decentralization or centralization on the efficacy of lobby’s depends on whether the objectives of domestic and foreign interest groups are aligned or not. If interests coincide, centralization means that the foreign lobby obtains an additional channel to influence the domestic government. For instance, through centralization domestic producers lobbying for low consumer rights or limited environmental protection may find support from foreign producer interest groups. Yet foreign interests may also oppose domestic interests, in which case centralization weakens the efficacy of lobby activities. Foreign producers would lobby against domestic producers that attempt to create barriers to entry on domestic markets. All in all, the impact of lobbying on the trade-off between centralization and decentralization is indeterminate. It depends on the specific conditions in a given policy area whether a central or a local government faces a higher risk of being captured. 2.3.4 Policy learning Imperfections in policy design may result from difficulties with policy learning. At times governments face considerable challenges to devise adequate measures in complex policy fields, like for instance intricate transitions in network industries. Even a government that honestly pursues the common good, may not be knowledgeable or creative enough to devise the most suitable policies or may be slow in picking up signals from society
2 Assessing Subsidiarity
31
that reform measures are needed, because institutions have become outdated. Decentralization may stimulate policy learning. Decentralization creates diversity in policy, which yields experiences with policy in one jurisdiction that may benefit another. In that way decentralization facilitates experimentation and creates possibilities for mutual learning. Of course learning only takes places when jurisdictions are aware of experiences in other jurisdictions and have an incentive to heed these experiences. One of the intentions of the open method of coordination (OMC) is to enhance this kind of policy learning. Member States benefit from experiences in other Member States through soft coordination at the EU level (see Section 2.4 for the position of the OMC in EU governance). Under the Lisbon strategy the OMC is applied to a broad range of policy fields, ranging from labour markets and welfare states to education and innovation. The method consists of identifying common objectives for national policy, devising associated performance indicators, drafting policy proposals in National Reform Programmes and mutual assessment of these programmes in joint committees of Member States and EU officials. In principle the OMC seems geared to policy learning, even so whether learning really takes place depends on the effectiveness of the OMC. Here the jury is still out. An important consequence of this form of soft coordination is that incentives to change policies are soft as well (Sapir et al. 2004). In the recent reform of the Lisbon strategy the focus is on ‘delivering’, with the discussion of National Reform Programmes in Member State’s parliaments as one of the measures to intensify incentives for action. Empirical evidence on the impact of the OMC on policy learning is still limited and yields mixed results (Ederveen et al. 2005). The most successful empirical results pertain to evidence about convergence of ideas (Radaelli 2003). In summary, decentralization may enhance policy learning through experimentation, which on a higher level can be strengthened by information exchange and building commitment. However, the inherently soft coordination generates only weak incentives for policy learning, for the simple reason that policy making powers remain at the national level in these areas. 2.3.5 Complementarities Complementarities across policies domains may affect the assessment of subsidiarity. Complementarities imply that a move towards centralization or decentralization in one dimension increases the benefit of moving in the
32
Sjef Ederveen, George Gelauff and Jacques Pelkmans
same direction in other dimensions. For instance, EMU initiated the discussion about the Stability and Growth Pact, because centralizing monetary policy affected the costs and benefits of fiscal centralization (Persson et al. 1996). EMU also provides a rationale for coordinating reforms among euro area countries (Pisani-Ferry 2005). The reason is that in a monetary union, a country that pursues reforms exerts an effect on its EMU partners, because the European Central Bank will adjust the interest rate in response. As another example, sometimes domestic liberalization brings policy areas within the confines of the internal market that previously were delivered by national public providers. Because of complementarities the decision to centralize policy in a specific field may have farther reaching consequences than initially recognized. Once a first step has been taken other policy domains may follow suit. Therefore, a full cost benefit analysis of centralization would have to include both the initial and the complementary policy fields. Another possibility is that complementarities create deadlocks. Pisani-Ferry and Sapir (2006) give the example of complementarities between product market reforms and labour market reforms: ‘a combination of product market regulations that aim at favouring entry and of labour market regulations that aim at preserving existing jobs is a recipe for ineffectiveness. Complementarities have two main consequences for the assessment of subsidiarity: complexity and significance. Theoretically it is very difficult to analyse the decision to centralize in the presence of complementarities. Seabright (1996) models a shift of power to a higher level in case of complementarities and positive externalities. Negative externalities, the main motive for the internal market (see Section 2.2.3), do not fit in the model. Significance means that complementarities increase the importance of an economic assessment of subsidiarity, which identifies the pros and cons in these complex policy dilemmas. That assessment would identify the costs and benefits of centralization in a specific policy domain with and without existing complementarities. In such a way an informed debate can take place about the merits of three options: decentralization, centralization in a specific policy domain, and centralization including a complementary policy domain.
2.4 Decision making in legislature The analysis above assumes that decisions in the union are being taken by a single decision maker, such as a president or executive council, elected by all people in the union. Abstracting from that assumption introduces the
2 Assessing Subsidiarity
33
possibility of conflicts of interest at the centralized level that negatively affect decision making. These drawbacks reduce the benefits of centralization and may shift the trade-off between centralization and decentralization towards decentralization. Abandoning the assumption of a single decision maker, it matters how the decision making process is organized at the centralized level. This section takes the perspective that regional delegates form a legislature, which makes decisions.10 As such it relaxes assumption 3 from Section 2.2.1. At the same time also assumption 2 of policy uniformity is dropped, to create the possibility that representatives engage in centralized policies that specifically benefit their regions of origin. Moreover, policy heterogeneity can be observed in practice. Frequently in federal systems a central government differentiates policies over jurisdictions (Besley and Coate 2003).11 According to Inman and Rubinfeld (2002) the model of decision making in a legislature, which they typify as democratic or majority-rule federalism, closely resembles European governance nowadays.12 In practice, of course, EU governance is more complex than the representation in a formal model. Sapir et al. (2004) distinguish four types of governance arrangements in the European Union:13 1. Full delegation of policies to the Union, such as trade policy (internal market, WTO) or competition policy. 2. Binding commitment among Member States. In this case Member States have agreed on EU surveillance and EU sanctions for policies that remain their ultimate responsibility. An example is state aid oversight by the Commission. 3. Coordination of policies that are decided and implemented by the Member States. Coordination covers: This extends the original theory by focussing on political processes, the behaviour of political agents and distortions arising from asymmetric information. Oates (2005) refers to these extensions as being part of the second generation theory of fiscal federalism. 11 The two assumptions have to be combined, because centralization would always be preferable when discarding policy uniformity while keeping central planning (see Section 2.1). 12 Inman and Rubinfeld (2002) distinguish three alternatives. Economic or centralized federalism concerns the single decision maker from Section 2.2.1. In cooperative or decentralized federalism representatives of the Member States’ governments unanimously decide on central government (union) policies. In democratic or majority-rule federalism representatives of the Member States decide on central policies by (simple) majority rule. 13 These types of governance show that in the EU ‘centralization’ varies from strong (full delegation) to weak (the OMC). 10
34
Sjef Ederveen, George Gelauff and Jacques Pelkmans
- explicit guidelines subject to multilateral surveillance (Broad Economic Policy Guidelines) - collective rules (Single Market regulations) - high-level dialogue (Eurogroup) - mutual information exchange and assessment (Open Method of Coordination) 4. Autonomy of Member States to decide on and implement policies, for instance in the field of direct taxation. Although formal models do not capture all aspects of EU governance, it is useful to review two main consequences of decision making in a legislature derived from such models. Section 2.4.1 touches upon an attitude of deference in the legislature towards each others’ proposals. Section 2.4.2 covers possible conflicts of interest in the legislature. Section 2.4.3 turns to the effects of deference and conflicts of interest on the trade-off between centralization and decentralization. 2.4.1 Deference in negotiations Decision making in a legislature requires negotiations between national representatives about the kind of policies to be addressed at the central level. In single issue bargaining the decision to centralize a policy depends on the bargaining power of countries and on the differences in weights that individual countries attach to that policy and to centralization. In practice negotiations take place in a setting of multiple goals and repeated games. In the European Council national governments continuously negotiate about a broad range of policies. In those cases ‘allowing’ centralization on one specific policy may act as a kind of side payment to obtain support from countries with low weights on centralization for other policies. In particular, if one country attaches a large weight to a specific policy, the other countries may allow centralization on that specific policy. On the one hand this may facilitate finding a solution when a minority strongly opposes a certain policy proposal. Yet, on the other hand this process may ‘get out of hand’. If all countries try to lever their national policies by lifting them to a European level, the process of deference may result in centralization on policies that would not pass a subsidiarity test. The risk of deference in particular applies to democratic federalism, the current system of decision making in the EU. Majority rule decision making is inherently instable. It may cycle from one majority to another without reaching an equilibrium. To cope with instability, Member States may revert to the “I’ll-scratch-your-back-if-you-scratch-mine” legislative norm
2 Assessing Subsidiarity
35
of deference to make decisions. ‘Under this norm locally beneficial but centrally inefficient government policies will be approved’ (Inman and Rubinfeld 2002). An other consequence of deference may be that countries formulate rather broad and vague goals or policies, so as not to interfere with each other’s domestic policies. That may either give much autonomy to central institutions to define the policy actions. Or, more probable, it may reduce the effectiveness of common policies, because it is unclear what is really meant. In particular when countries put each other to the test in a process of peer review, vague phrases may be a way out. Guidelines for individual Member States frequently contain phrases such as ‘promote more adaptable and innovative work organization’ or clauses like ‘where appropriate’. According to Pisani-Ferry and Sapir (2006) the mid-term revitalization of the Lisbon strategy did not succeed in solving these problems. Resistance by the large Member States against naming and shaming had country specific guidelines replaced by ‘no less than’ 24 general guidelines. Each of these guidelines is being applied to all Member States, without distinguishing whether some may be more relevant for a country than others. 2.4.2 Conflicts of interest: common pool problems Discarding policy uniformity (assumption 2 in Section 2.2.1), conflicts of interest are an important reason why centralization may be at odds with preference matching. Representatives in the legislature are primarily answerable to voters in their constituency or region and care less about voters in other regions (Lockwood 2005). Depending on the modelling of the voting process and bargaining in legislation, the literature identifies several consequences of these conflicts of interest. An important drawback of centralization under democratic federalism is common pool budgeting or raiding the commons (Inman and Rubinfeld 2002; Oates 2005). Member States have an incentive to draw as much as possible on the common budget for projects that locally provide benefits. In that way other states co-finance these projects, whereas the benefits mainly accrue to local constituents. EU agricultural policy, cohesion policy and structural funds come to mind here. When considering independent taxation capabilities at the EU level, common pool problems have to be taken into account as well (Rattso 2003). The common pool problem also creates a bias in the selection of projects funded by the legislature (Lockwood 2002, 2005). When there is cost sharing (for instance through the current GDP-proportional contributions to the EU budget) the legislature has an incentive to minimize the costs of
36
Sjef Ederveen, George Gelauff and Jacques Pelkmans
projects, not to maximize their net welfare gains when these are unevenly distributed among Member States. The reason is that all Member States benefit from cost reductions, whereas only one or a few Member States reap the economic surplus of the projects. To some extent this effect may offset the deference problem that creates a bias towards projects beneficial to individual Member States (see Section 2.4.1). Finally, cost sharing of local public goods may induce over-provision of public goods in a centralized system (Besley and Coate 2003). Cost sharing creates an incentive for local voters to strategically delegate by electing representatives with high demand for public spending. If one region elects a delegate to the legislature who places high value on the public good, this delegate will be more aggressive in demanding a higher public good for that region. That benefits citizens from that region, because part of the costs is borne by the other regions. But if all regions act that way the total amount of public goods will be higher than their efficient levels. In contrast to the previous project selection effect, strategic delegation may exacerbate the deference problem. 2.4.3 A shifting trade-off Democratic federalism introduces several inefficiencies associated with centralization. An attitude of deference among the representatives in the legislature may result in approval of inefficient policies and projects, such as transfers that benefit specific regions. Common pool problems may also lead to over-supply of policies at the central level. In contrast, the project selection effect associated with cost sharing may make the central government less sensitive to tastes of the regions. Moreover, in these political economy models of fiscal federalism it is not generally true that the higher the cross-border externalities the higher the welfare gain from centralization (Lockwood 2002). Institutional solutions to these inefficiencies are hard to reach. An option is to reform the legislative process, for instance by increasing the power of the Parliament or giving veto rights to the executive (Inman and Rubinfeld 1997, 2002). In particular the latter will be a rather delicate topic in the context of the EU. Alternatively, adjustment of the institutions of federalism might change the system of direct representation of Member States. In that case groups of local jurisdictions would elect one representative, which might reduce deference. However this comes at the cost of less preference revelation: one representative has to internalize all preferences of several jurisdictions.
2 Assessing Subsidiarity
37
With institutional solutions out of reach, these drawbacks of centralization have to be taken into account when contemplating centralization of a specific policy or not. In some cases inefficient centralized policies may not outweigh the benefits of internalizing cross-border externalities or achieving economies of scale. For instance, considering an independent taxation authority for the Union, Rattso (2003) states that an intermediate situation of broad fiscal responsibilities together with a weak centre is the worst case. It is inferior both to the current situation of a weak centre with limited fiscal authority and the alternative of a politically strong centre with a broad fiscal mandate. Hence, subsidiarity assessment has to take into account these weaknesses of centralization related to the political process of decision making in the Union.
2.5 Conclusion: assessing subsidiarity This chapter has presented many arguments that relate to the degree of centralization of policies in the European Union. Table 2.1 presents an overview. The table starts with the basic trade-off that underlies the functional subsidiarity test from Section 2.2.4. In that trade-off centralization is warranted when increasing returns or cross-border externalities outweigh preference matching. The table shows that several political economy motives affect the assessment of subsidiarity as well. Decentralization provides opportunities for citizens to make governments more accountable to their preferences, which together with policy competition may discipline Leviathan governments in Member States. Effective monitoring by Member States may counter government failures at the EU level and in that way supports centralization. The sensitivity to lobbying depends on whether objectives of domestic and foreign interest groups are aligned or contrasting. If they are contrasting, centralized policies are less affected by lobbying. Policy learning benefits from experimentation among decentralized authorities and from information exchange and commitment building at a central level. Finally, concerning decision making at the centralized level two main risks have been identified. Overprovision of locally beneficial policies may result, when decision making takes place under a norm of deference. Common pool problems arise, when Member States take advantage of the common budget. Hence, decentralization may prevent these inefficiencies of centralized decision making.
38
Sjef Ederveen, George Gelauff and Jacques Pelkmans
Table 2.1 Subsidiarity: motives for decentralization and centralization Functional test
Decentralization
Centralization
Preference matching
Increasing returns Cross-border externalities
Political economy Leviathan Accountability, policy competition Lobby Aligned objectives of lobby groups Policy learning Experimentation Deference Common pool
Effective monitoring by Member States Contrasting objectives of lobby groups Information exchange, commitment for reform
Prevent over-provision of policies Prevent raiding the commons
Dynamics Trends Affect motives over time Affect motives over time Dynamic process Centralization may be irreversible
All these considerations also have dynamic aspects, European integration is a dynamic process. The optimal level of centralization in the Union may shift over time, when trends affect the trade-off between centralization and decentralization. Trends may intensify (diminish) external effects and economies of scale and thus may add to (limit) motives for EU policies. For instance, increasing interdependencies in international trade may raise the weight of a common European position in WTO negotiations. Or trends may affect other aspects of the trade-off, such as lobby intensity or government failures. Furthermore, the decision making process in the European Union is subject to changes. When the decision making process is altered, the assessment of subsidiarity may shift for specific policies, for instance because government failures manifest themselves in another way. By consequence, assessing subsidiarity is no static exercise. In addition, this dynamic process may be asymmetric. Although theoretically it is no problem to change the degree and form of centralization, in practice a decision to centralize certain policies in the European Union may be hard to reverse. That means the decision to centralize has to be taken with care so as not to reduce the option value of waiting. What general picture do all these pro’s and con’s yield? Clearly, the basic trade-off occupies an essential place in the assessment of subsidiarity. Any serious assessment of subsidiarity has to start with identifying the existence of increasing returns and cross-border externalities and weighing them against the heterogeneity in the Member States. This functional sub-
2 Assessing Subsidiarity
39
sidiarity test remains an indispensable tool to make the main arguments pro and contra centralization explicit. However, there may be good reasons for the final subsidiarity test to differ from the functional one. Analysing political decision making is important to further understand the trade-off between centralization and decentralization (see Besley and Coate 2003). At the same time allowing for government imperfections and political economy considerations, makes it difficult to draw general normative conclusions on the appropriate degree of centralization (compare Persson et al. 1996). Because all considerations are case specific, issues of subsidiarity can only be assessed on a case-bycase basis. By consequence a thoughtful answer to the question whether to centralize or not, would demand case-specific careful empirical analysis. The decision to centralize policies in Europe or not is a complex one. Analysts have to weigh all the partly conflicting motives carefully. What the assessment of subsidiarity would entail is to express all arguments in an open debate. This chapter has provided a possible framework to do exactly that.
References Alesina A, Perotti R (2004) The European Union: a politically incorrect view. Journal of Economic Perspectives 18: 27–48 Besley T, Coate S (2003) Centralized versus decentralized provision of local public goods: a political economy approach. Journal of Public Economics 87: 2611–2637 CESifo (2003) Rethinking subsidiarity in the EU: economic principles. EEAG Report: 76–97 Dekker P, Ederveen S (eds) (2005) European Times. In: European Outlook 3, CPB/SCP, The Hague Dekker P, Ederveen S, De Groot HLF, Van der Horst A, Lejour AM, Straathof SM, Vinken H, Wennekers C (2006) Diverse Europe. In: European Outlook 4, CPB/SCP, The Hague Ederveen S, Van der Horst A, Tang P (2005) Is the European economy a patient, and the Union its doctor? On jobs and growth in Europe. CPB Document 80, CPB, The Hague Føllesdal A. (1998) Subsidiarity. The Journal of Political Philosophy 6: 231–259 Hughes-Hallett A, Lewis J (2004) Debts, deficits and the entry of the Accession Countries into the Euro, Money Macro and Finance (MMF) Research Group Conference 2004, no 58 Inman R, Rubinfeld D (1997) Rethinking federalism. Journal of Economic Perspectives 11: 43–64
40
Sjef Ederveen, George Gelauff and Jacques Pelkmans
Inman R, Rubinfeld D (2002) Subsidiarity, governance and EU economic policy. CESifo Forum, 4/2002, Munich Lockwood B (2002) Distributive politics and the costs of centralisation. Review of Economic Studies 69: 313–337 Lockwood B (2005) Fiscal decentralization: a political economy perspective. Warwick Economic Research Papers 721, University of Warwick Majone G (2001) Two logics of delegation: agency and fiduciary relations in EU governance. European Union Politics 2: 103–122 Mooij R de, Tang P (2003) Four Futures of Europe, CPB, Den Haag Musgrave RA (1959) The theory of public finance, New York: McGraw-Hill Oates WE (1972) Fiscal Federalism, New York: Harcourt Brace & Jovanovich Oates WE (2005) Toward a second-generation theory of fiscal federalism. International Tax and Public Finance 12: 349–373, August 2005 Pelkmans J (2005a) Testing for Subsidiarity. BEEP Briefings no.13, College of Europe, Bruges Pelkmans J (2005b) Mutual recognition in goods and service: an economic perspective. In: Kostoris F, Schioppa P (eds) The principle of mutual recognition in the European integration process, Palgrave Macmillan Pelkmans J (2006) European integration, methods and economic analysis, third edition, Pearson Education Persson T, Roland G, Tabellini G (1996) The theory of fiscal federalism: what does it mean for Europe? Paper prepared for the conference Quo Vadis Europe, Kiel Pisani-Ferry J (2005) What’s wrong with Lisbon? Breugel Policy Brief, 2005 Pisani-Ferry J, Sapir A (2006) Last exit to Lisbon. Breugel Policy Brief, 2006/2 Pollack MA (2003) The engines of European Integration: delegation, agency and agenda setting in the EU, Oxford University Press Radaelli CM (2003) The open method of coordination: A new governance architecture for the European Union? Rapport nr 1, Swedish Institute for European Policy Studies, Stockholm Rattso J (2003) Fiscal federalism in the European Union: lessons from decentralized government. Norwegian University of Science and Technology, Working Paper Sapir A, Aghion P, Bertola G, Hellwig M, Pisani-Ferry J, Rosati DK, Vinals J, Wallace H, Buti M, Nava M (2004) An Agenda for a Growing Europe: The Sapir Report, Oxford University Press Scharpf F (1999) Governing in Europe: effective and democratic?, Oxford University Press Seabright P (1996) Accountability and decentralisation in government: an incomplete contracts model. European Economic Review 40: 61–89 Sinn HW (2003) The new systems competition, Basil Blackwell, Oxford Tabellini G, Wyplosz C (2004) Supply-side coordination in the European Union. Report prepared for the Conseil de l’ Activité Economique Tiebout CM (1956) A pure theory of local expenditures. Journal of Political Economy 64: 416–424
3 Who Shall Decide What? Citizens’ Attitudes Towards Political Decision Making in the EU
Joachim Ahrens, Martin Meurers and Carsten Renner
3.1 Introduction
1
The allocation of political powers in the European Union (EU) has been one of the key problems of European integration. The existing pattern of policy responsibilities has emerged rather spontaneously over time. Today, neither a strategic road map nor a consistent and commonly accepted method exists for assigning policy prerogatives clearly and consistently either to the supranational or the national authorities (Alesina et al., 2001a). Economic theory suggests that public goods and services should be supplied at the lowest level of government that is able to do so efficiently (Oates 1999). As a rule of thumb one can argue as follows: Public goods which exhibit international spill-over effects or whose production is subject to scale economies are to be provided at the supranational level. Public goods the demand for which is based on heterogeneous preferences across countries should rather be provided at national or even local levels of policymaking. Regarding the EU, a trade-off resulting from benefits based on the exploitation of scale economies or the internalization of (cross-country) external effects on the one hand and preference costs on the other can be observed (Alesina and Wacziarg 1999; Alesina et al. 2001b, 2001c). These costs arise from the fact that exploiting economies of scale or internalizing external (i.e., European wide) effects necessitates a harmonization of policies. A centralization of policymaking, however, makes it increasingly difficult to account for different political preferences, economic problems, 1
The authors wish to thank two anonymous referees for valuable comments and suggestions. This paper heavily draws on parts of Ahrens et al. (2007).
42
Joachim Ahrens, Martin Meurers and Carsten Renner
and social needs within a heterogeneous union of countries. From a normative view, economic theory suggests that political prerogatives should be anchored on the European (or even the global) level in the areas of foreigntrade policy, competition policy, and enforcing the rules of the Common Market. Concurrent competencies are recommended in the realms of international relations, environmental policy as well as fiscal and monetary policy. Finally, questions and problems regarding education, research, cultural issues, citizen and social protection as well as industrial policy should be addressed by national (or local) authorities.2 The actual allocation of political competencies, however, contradicts this normative prescription. In some areas (e.g., in agricultural and fishing policy) EU activities are too extensive, while in others (such as the environmental policy) EU competencies are too limited. As a result, an inefficient allocation of political prerogatives has emerged. The goal of the following empirical investigation is to analyse whether or not the preferences of EU citizens are consistent with the implications of economic theory, whether they are in line with the actual situation, or whether they would suggest yet another pattern of political power. In doing so, citizens’ attitudes towards EU centralized vis-à-vis national decision-making are analysed in two country groupings the EU15 countries and the ten new members (EU10), which acceded the Union in 2004.3
3.2 Measuring citizens’ preferences The attitudes of citizens with respect to the allocation of political decisionmaking power are reflected in the survey results of the Eurobarometer poll. Being a comprehensive questionnaire on EU-related political issues, household decisions, economic expectations, and overall life satisfaction, it also includes questions referring to 25 policy areas in which the EU administration is currently involved. The questionnaire asks whether citizens
2
3
Of course, this competence allocation is not that straight-forward in reality. While many observers would agree that education as well as innovation policies may be effectively developed and administered at the national level, parts of higher education or innovation and technology policy, e.g., may show crossborder external effects or scale economies and hence should be subject to supranational legislation and administration. Bulgaria and Romania are not included in EU10. Their inclusion, however, would not change the basic outcome of the investigation.
3 Who Shall Decide What? 43
prefer political decisions at the European or national level.4 Since we are interested in current views about EU policies, the introduction of the Euro in 1999 is considered a crucial date, from which onwards citizens have likely become more responsive to European politics. The data used in this analysis cover ten biannual surveys from spring 1999 to spring 2004.5 For the ten new members (EU10), data from 2001 to 2004 were taken into account. We begin our analysis of the differences in citizens attitudes with respect to the individual policy fields (J = 25). For each of these, the Eurobarometer data provide the percentage of respondents in each country who prefer national competencies (%NAT) and EU competencies (%EU), respectively. In order to characterize the preference of the voters of each old member and new member country (i=1,..,15, and i=1,..,10, respectively) over the policy domains (j=1,..,J), we calculate a simple index:
X i, j
100
1 ¦ (% NATi, j ,t % EU i, j ,t ) , t T t
1,..., T
where the number of observations is T = 10 for the EU15 and T = 4 for the EU10. The corresponding standard deviation Si,j is then:
S i, j
Var (% NATi , j ,t ) Var (% EU i , j ,t ) 2Cov(% NATi , j ,t ,% EU i , j ,t )
A first indicator for dominating preferences within the EU15 and the EU10 sample can be obtained by counting the number of countries for which the indices Xi,j are significantly below or above zero according to a simple ttest. This gives a crude first impression about the policy fields with a clear majority of countries for either centralization or decentralization. The difference D of significant pro EU votes and significant pro national votes provides a suitable indicator in the further analysis. Another measure for preferences is the likelihood of a 50% majority L50 for centralized decision-making if each country had one vote. This likeliEach survey consists of approximately 1000 face-to-face interviews per Member States (except Germany: 2000, Luxembourg: 600, United Kingdom 1300 including 300 in Northern Ireland). It is conducted between 2 and 5 times per year, with reports published twice yearly. The exact question text is: “For each of the following areas, do you think that decisions should be made by the (NATIONALITY) government, or made jointly within the European Union”. The results of the polls can be downloaded from http://europa.eu.int/comm/public_opinion/. 5 In 2000 the Eurobarometer poll was only conducted once. 4
44
Joachim Ahrens, Martin Meurers and Carsten Renner
hood is obtained by simulating 10,000 random draws from a Bernoulli distribution over 1 for centralized, and 0 for decentralized decisions, where the probabilities are derived from hypothesized normally distributed preferences of each country based on the previously calculated mean index Xi,j and its standard deviation. Subsequently, it is checked for every draw whether a majority of countries voting in favour of centralization exists. These majorities are counted, and L50 is defined as the percentage of majorities for centralized policies of the total number of experiments (10,000). As an additional piece of information for an evaluation of the robustness of these two indicators, for each policy field j we compute the average across i of the indices Xi,j for the two groups ( X j ) and the standard deviation of the indices within the groups ( S j ). Beyond the preference for the level of decision-making in certain policy domains, the extent of contradicting preferences is reflected in further dimensions: first, in the overall attitude of the individual countries towards the EU, which can be quantified by the number of policy fields for which country i’s citizens would opt for the EU level (Xi,j 0 ); second, in the deviations of an individual country from the preferences of other countries’ citizens (i.e., Xi,j has a different sign than Xk,j for a country k z i ). To obtain a quantitative measure for a country’s preferences in relation to the others, we count the number of countries with the same preference as country i (#Gi,j) and relate it to the number of countries with the majority position (#Mi,j). The degree of accordance with the majority position can then be summarized in the following index: Ai
1 J
¦ j
# Gi, j 1 . #Mj
For a country of the EU15, e.g., Ai may theoretically range from 1 to 1/14. The former indicates that the respective country represents the majority position which is in fact shared by all other countries regarding all policy fields. The latter reflects an outsider position, namely that the respective country never represents the majority position. In that case, this country would be isolated because all other member countries represent a different position with respect to all policy fields.
3 Who Shall Decide What? 45
3.3 Attitudes of citizens in the EU15 First of all, citizens’ preferences with regard to the 25 policy fields are analysed.6 The ranking in Table 3.1 is compiled by sorting the policy fields according to preference measures giving priority to D, L50, and then to the mean of the country-specific indicators. While the indicators D and L50 reflect the preferences in a straight-forward way, the mean and the standard deviation (S.D.) reflect the robustness of these preferences. In order to better understand the motives for the observed distribution of preferences, it is also useful to look at the pair-wise correlations of indicators for distinct policy fields, e.g., between Xi,j and Xi,k for j z k, k J.7 In some cases, the close relationship between certain policy fields might offer a first tentative indication of economic or political considerations that are responsible for the articulated preferences. Another interesting aspect is which coalitions of countries prevail especially for the policy fields without a clear majority. This provides some guidance about geographical factors or past political experiences which might cause common preferences in certain country groups. Remarkably, for 11 of the 25 policy fields a clear majority of countries favours decision-making at the EU level. The desire for centralization is strongest with respect to policies against the Exploitation of Human Beings. Looking at the correlations of the preferences for this policy field with those for other policies, a high correlation exists with Currency, Info about EU, Foreign Policy, Organized Crime and Drugs which are also strongly at the EU level. This agglomeration shows that the preference for centralization might be a reflection of a multitude of interrelated concerns about human rights, economic wealth and security. Beyond the correspondence with the normative considerations about an economically meaningful assignment of responsibilities, this result might also represent a common desire for protection and coordinated decisions to cope with the uncertainties created by a quicker pace of global changes.
Of course, it is not certain that the results of opinion polls reflect national positions, and in fact, the actual official political position may differ from citizens’ attitudes. In the following, all statements relating to country positions reflect citizens’ opinions and not necessarily the current official position of the respective government. 7 The complete correlation matrix is available from the authors upon request. 6
46
Joachim Ahrens, Martin Meurers and Carsten Renner
Table 3.1. Preference indices over policy fields in the EU15 sample Policy Domain
D
L50
Mean
Actual S.D. competence allocation
Exploit. Human Beings 15.0 100% -57.1 11.1 S Information EU 15.0 100% -46.3 12.0 E Foreign policy 15.0 100% -42.2 18.3 L Research 15.0 100% -38.0 18.1 S Drugs 15.0 100% -34.9 11.7 S Regional aid 15.0 100% -28.8 13.5 S Organized crime 13.0 100% -42.8 16.2 S Poverty/social exclusion 13.0 100% -20.5 14.1 L Humanitarian aid 11.0 100% -32.8 23.0 S Currency 11.0 100% -29.4 29.9 E Environment 11.0 100% -17.7 18.5 S Mean EU 13.0 100% -35.5 16.9 Agriculture & Fishing 3.0 71% 0.8 20.6 E Political asylum 1.0 45% -0.7 26.3 S Unemployment 1.0 38% 2.7 19.3 S Acceptin refugees 1.0 35% 4.8 32.3 S Defence -1.0 19% 8.9 34.3 L Immigration -3.0 27% 4.9 31.2 S Mean undecided 0.3 39% 3.6 27.3 Juvenile crime -9.0 0% 17.3 24.5 L Urban crime -11.0 0% 21.7 22.2 L Cultural policy -11.0 0% 23.6 19.7 L Media -13.0 0% 24.8 17.8 L Education -13.0 0% 36.0 16.2 L Health and social welfare -13.0 0% 37.2 25.4 L Justice -15.0 0% 33.4 19.5 S Police -15.0 0% 41.3 15.8 S Mean National -12.5 0% 29.4 20.1 The last column represents the actual allocation of competencies in the EU; E indicating an extensive involvement of the EU, S indicating shared responsibilities of the EU and national governments, and L indicating a limited influence of the EU. Source: Eurobarometer; Busch (2007); own calculations.
The unanimous preference for EU-led Regional Aid seems to be uncorrelated with preferences over other domains. Obviously, a great majority of countries regards the past progress of some of the poorer regions (Portugal, Greece, Spain, Ireland) as a persuasive argument for the continuation of
3 Who Shall Decide What? 47
such a policy.8 In light of the recent political controversies in connection with the military action in Iraq, the strong desire for a common Foreign Policy seems noteworthy. Furthermore, the preferences for EUcoordination of Environment, Humanitarian Aid, and Poverty and Social Exclusion are all interrelated. Hence, the EU perspective can possibly be attributed to the fact that citizens recognize potential coordination failures and free-rider problems of national initiatives. Research seems to be a similar matter. Its strong correlation with currency might reflect economic considerations that EU-coordinated research might allow the exploitation of scale effects. On balance, there are only six policy fields in which the EU15 members’ preferences are controversial. As far as the interrelated fields Political Asylum, Accepting Refugees and Immigration are concerned, citizens in six countries are in favour of a EU decision level Belgium, Greece, Spain, Italy, France, and the Netherlands9 (with respect to political asylum and accepting refugees). The other end of the political spectrum with countries opposed to centralization is represented by Denmark, Finland Ireland, Sweden, the UK, and Austria (with respect to immigration, political asylum and accepting refugees). For policies related to Agriculture and Fishing, the proponents of EU decisions are mainly located in Belgium, Germany, Spain and the Netherlands, whereas the opposition consists mainly of Finland, Sweden, Austria, and the UK In the past, the former group benefited from EU policy measures when it came to common market rules, quality standards and control (e.g., preventing the spread of animal diseases). In contrast to this, Sweden and the UK might fear that such policies are used as measures of protectionism and could potentially harm their national agricultural sector. The preference distributions over a common Defence Policy and EU measures against Unemployment are particularly interesting. Intuitively, foreign policy and defence policy should be strongly connected. This is underpinned by the correlation coefficient of the preference indices of 83%. The preferences over foreign policy, however, are strongly correlated with those of Currency, Justice, Humanitarian Aid, Immigration, Political Asylum, and Info about the EU. Thus, beyond a pure military dimension, 8
9
Alesina et al. (2001a, p. 10) interpret this result as a bias in the answers of the poll “…by the presumption that transferring policy responsibility to the EU may result in net benefits for those countries. In other words, the response in the questionnaire may reflect the perception of personal or country gains rather than a fair judgment about the optimality of allocative criteria”. Germany, Portugal, and Luxembourg are not considered as their preferences for EU decisions are marginal.
48
Joachim Ahrens, Martin Meurers and Carsten Renner
the clear preference for a common foreign policy seems to reflect a rather general desire for a strong position of the EU in the globalization process. Furthermore, it should be considered that defence is already part of the NATO framework, and the open debate about its future role might contribute to the heterogeneous attitudes within the EU countries.10 The question of common measures against the unemployment problem can be regarded as even more controversial.11 Only Italy and Greece are clearly in favour of EU policies. Denmark, Finland, Sweden and the UK are obviously more satisfied with national measures. The correlation between the preference for EU decisions and the tenyear average of the unemployment rate in the member countries suggests that the failure of national policies in the past is likely to account for this distribution of preferences. The outliers Italy and Greece could be partly explained by their relatively weak governance structures. The outlier Finland can be explained by the fact that despite persistent high unemployment, people seem to acknowledge the past success of national authorities in reducing unemployment significantly.12 In particular, the example of Finland, but also the cases of Sweden and Denmark underline the preference of Scandinavian countries for decentralized decision-making that becomes apparent in connection with other policy fields and that is a likely result of the national tradition of decentralized administration.13 From the observation of the country specific differences in opinions we can draw the following intermediate conclusion: Beyond Anderson and Kaltenthaler’s (1996) result that the length of EU membership is important (which might account for the Scandinavian countries’ reservations about EU-based decisions), the preference for EU institutions seems to depend also on the geographical conditions to protect national borders (as indicated by the controversies about immigration, etc.), past costs and benefits of EU policies (agriculture and fishing), and finally the trust in the ability Interestingly, the aggressors of World War II (Germany and partly Italy) and their immediate neighbours Belgium, Luxembourg, Netherlands are relatively committed to an EU defence policy. On the other hand, countries that were involved in national and international military conflicts since the war, like Ireland, the UK, and Greece, favour national decisions in this domain. 11 Despite a majority of three significant votes against EU level decisions, there exists a high probability for a majority of countries in favour EU administration. This can be explained, by the fact that citizens of only 5 countries have a distinct (significant) position regarding this issue. 12 In Finland, unemployment decreased from a peak of 15.4% in 1995 to roughly 8.8% in 2004. 13 Regarding the strong overall preference for respecting subsidiarity see EEAG Report (2003), p 86. 10
3 Who Shall Decide What? 49
of EU institutions to handle problems more effectively than national authorities (e.g., regarding labour market policies).14 The eight domains with clear preferences for national decision-making essentially represent issues of national identity, like Cultural Policy, Media, and Education, or can be clearly regarded most efficient when decentralized decisions are taken, like Police, Urban and Juvenile Crime, and Justice, because local coordination of actions and information gathering are required. Health and Social Welfare can be regarded as matching both criteria. Corresponding to the separation of policy fields into the three groups, the standard deviation of the country preferences is highest on average for those policy fields where a clear majority does not prevail. If the standard deviation is interpreted as an additional measure for heterogeneity, then, independent from the majority situation, one can characterize the scope for controversies that are still present among the EU15 members. Notwithstanding the clear majorities, preferences appear to be most disparate about a common currency and health and social welfare policy, as is indicated by relatively high standard deviations of 30 and 25 index points. Another measure of heterogeneous attitudes relates to country-specific EU optimism or scepticism. The majority of countries in the EU15 sample is in favour of centralizing 15 policy domains (the first 11 domains of Table 3.1 plus Agriculture and Fishing, Political Asylum, Accepting Refugees as well as Unemployment) and of decentralizing the rest. Table 3.2 shows that the centre preference is that of Portugal, which prefers an allocation of powers distinct from the majority position in only one policy field (Agriculture and Fishing). Due to their political capacity, Germany and France can be considered as the EU’s political gravity centre. Their citizens’ preferences differ only in two policy fields from the majority position (Germany with respect to Defence and Political Asylum and France with respect to Defence and Immigration). Similar preference patterns exist in the smaller countries Belgium, Ireland, Denmark, Luxemburg, and the Netherlands. Scepticism about the EU is most apparent in the UK, Sweden, and Finland, where citizens prefer national decisions in the majority of policy issues. A fourth group is constituted by Spain, Greece and Italy.
14
The UK and Sweden are at the lower end when it comes to ”Trust in the EU institutions and bodies” according to recent Eurobarometer surveys.
50
Joachim Ahrens, Martin Meurers and Carsten Renner
Table 3.2. General attitude of the EU15 towards centralization Policy fields with Deviations from Accord Uncertainty Si EU preference majority position index Portugal 14 Germany 15 Belgium 17 France 17 Denmark 12 Ireland 12 Luxembourg 14 Netherlands 16 Austria 13 Spain 19 Italy 18 Greece 9 United Kingdom 8 Sweden 9 Finland 14 Mean 14 Min 8 Max 20 Source: Eurobarometer; own calculations.
1 2 2 2 3 3 3 3 4 4 5 5 6 7 8 4 1 8
0.91 0.92 0.91 0.91 0.91 0.91 0.91 0.90 0.87 0.85 0.82 0.81 0.83 0.81 0.77 0.87 0.77 0.92
12.4 8.1 6.1 7.1 7.5 8.5 10.2 11.2 6.7 9.8 8.0 18.6 8.0 7.4 7.3 9.1 6.1 18.6
Citizens in these countries are quite optimistic about the higher effectiveness of EU-located decision-making; they prefer centralized decisions in many policy fields where the other gravity centre, despite their broad EU commitment, would prefer national administration. Their common characteristic is the preference of EU policies with respect to juvenile crime, which can be seen as an indication that national authorities do not succeed in this domain and consequently the hopes for an improvement are attributed to the EU. This line of reasoning receives support from the specific case of Italian citizens, who seem to be overly EU optimistic in preferring national competences only for issues such as police and urban crime. Empirical support for this argument can be obtained through the World Bank’s Governance Indicators.15 Figure 3.1 shows the scores for three indicators which represent the capacity and competence of national govern-
15
The indicators are compiled as percentile ranks indicating the share of countries in a worldwide perspective that are rated below a particular country. See Kaufmann et al. (2003), http://info.worldbank.org/governance/.
3 Who Shall Decide What? 51
ments. Concurrent with the above reasoning, the governance scores for Italy, Greece and also Spain are at the bottom end of the EU15 sample.16 Furthermore, Table 3.2 reveals that there is a negative correlation between accordance with the majority and the duration of membership in the EU, e.g., if one looks at the indicators for Greece, Austria, and the Scandinavian countries. The latent EU- and Euro-sceptic British certainly represent an exception from this rule. Regarding the uncertainty in citizens’ preferences, the proposed indicator S i , which is an average of the S i , j for all policy fields, seems to be independent of overall EU support or the extent of deviation from the majority position. The only regularity that one might recognize is that smaller countries like Portugal, Greece, Austria, and the Netherlands tend to have a greater variability in their citizens’ preferences. Finally, note that the number of deviations from the majority and the accord index lead to almost identical conclusions. 3 2.5 2 1.5 1 0.5
government effectiveness control of corruption EU15
UK
Sweden
Spain
Portugal
Netherlands
Luxemburg
Italy
Ireland
Greece
Germany
France
Finland
Denmark
Belgium
Austria
0
regulatory quality average governance
Fig. 3.1. Quality of national governance in the EU15 countries Source: Kaufmann et al. (2005). 16
Of course, the link between low appreciation of national institutions and support for EU decision-making cannot explain everything. Other factors such as measures with transboundary effects (e.g., consumer policy, internal market issues) as well as policies exhibiting scale economies (e.g., foreign policy) help explain EU preference. But even if one controls for these factors, the weakness of national institutions remains a significant explanatory variable.
52
Joachim Ahrens, Martin Meurers and Carsten Renner
3.4 Attitudes of citizens in the new member countries Table 3.3 summarizes the preference indicators of the ten new member countries (EU10) with respect to the 25 policy fields. Citizens in EU10 have a much stronger preference for EU-level decisions than in the old member countries. All six policy-fields being controversial among the EU15 are clearly preferred at the central level. Moreover, the three policy domains Health and Social Welfare, Juvenile Crime, and Education, that are overwhelmingly preferred at national levels by the EU15, are also included in the set of policy fields which the EU10 would like to have administered by European institutions. The parallel to the preferences of Italian, Greek and partly Spanish citizens suggests that the scepticism about the competence and capacities of national authorities to provide sufficient security and transparency are the likely causes for this supranational tendency. Again, this is underpinned by the World Bank Governance Indicators. In comparison with the data presented in Figure 3.1, governance indicators for the EU10 are significantly below the EU-average, and only Estonia, Malta and Slovenia lie significantly above the average of indicators for Italy, the country with the weakest institutional performance in the current EU. 1.6 1.4 1.2 1 0.8 0.6 0.4 0.2
government effectiveness control of corruption EU15
Slovenia
Slovakia
Poland
Malta
Lithuania
Latvia
Hungary
Estonia
Czech Republic
Cyprus
0
regulatory quality average governance level of Italy
Fig. 3.2. Quality of national governance in the new member countries (EU10) Source: Kaufmann et al. (2005).
3 Who Shall Decide What? 53 Table 3.3. Preference indices over policy fields in the ten new Member States Policy Area
D
L50
Exploit. Human Beings 10 100% Organized crime 10 100% Drugs 10 100% Information EU 10 100% Research 10 100% Humanitarian aid 10 100% Regional aid 10 100% Foreign policy 10 100% Poverty/social exclusion 10 100% Unemployment 10 100% Environment 10 100% Defence 8 100% Currency 8 100% Juvenile crime 8 99% Agriculture & Fishing 8 98% Political asylum 6 99% Immigration 4 100% Education 4 99% Health and social welfare 4 96% Accepting refugees 4 88% Mean EU 8.2 99% Justice 2 34% Urban crime 0 50% Mean undecided 1 42% Media -4 0% Police -6 2% Cultural Policy -8 0% Mean National -6 1% Source: Eurobarometer; own calculations.
Mean
S.D.
-63.4 -58.6 -58.1 -55.8 -55.7 -48.1 -45.0 -43.7 -40.6 -35.3 -33.6 -30.0 -27.2 -16.8 -14.4 -15.1 -17.2 -11.2 -14.4 -8.1 -34.6 -0.9 -6.4 -3.7 13.8 8.4 19.0 13.7
18.9 19.0 18.1 16.8 17.6 17.1 16.8 18.0 16.5 17.0 12.8 20.7 24.1 15.3 18.8 14.5 16.6 13.8 19.9 15.4 17.4 14.0 18.1 16.0 21.5 14.0 11.5 15.7
Controversies among the EU10 remain about the assignment of Urban Crime and Justice. Unfortunately, the existing alliances with Malta, Estonia, and the Czech Republic in favour of national decisions and Slovenia, Latvia, Poland and Cyprus in favour of EU decision-making do not provide for a starting point to explain this preference distribution. The EU10 share the preferences with the old members for decentralization only with respect to Media related, Police and Cultural Policies. One major difference between the EU15 and the ten new members is that the preferences over the policy fields are not as highly correlated. While the EU15 associate, e.g., defence with twelve other policy areas, the EU10 link defence issues only to Currency, Organized Crime, Drugs and
54
Joachim Ahrens, Martin Meurers and Carsten Renner
Exploiting Human Beings. A striking correlation is the one between Humanitarian Aid, Poverty, Unemployment, Regional Aid, Research, Info about EU, Foreign Policy, Immigration, Organized Crime, Drugs, Exploitation of Human Beings, Health, Accepting Refugees as well as Juvenile Crime. These seem to be the most important policy areas for the ten new members, as they strongly prefer EU responsibility. Comparing the standard deviations of the preference indices of the two groups (last column in Tables 3.1 and 3.3), one cannot find a significant difference regarding the implicit country-specific heterogeneity of preferences within the two groups. There is even a slight tendency for tighter attitudes of the EU10. Interestingly, in both groups the preferences seem to be most disparate with respect to a common Currency. The high withingroup heterogeneity of the EU15 sample with respect to Defence and Immigration is not shared by the sample of new member countries. The similar degree of within-group heterogeneity between the EU15 and the ten new members is also reflected in the country specific preferences of the EU10 group (Table 3.4). The majority position of this group is in favour of centralizing 22 policy domains and decentralizing the 3 domains Police, Media, and Cultural Policy. This position is represented by Latvia. The citizens of Poland, Cyprus, Slovakia, Lithuania, and the Czech Republic are in close vicinity of this position. They prefer the one or the other policy field more or less at the national or the EU level, without any significantly different overall EU confidence. Slovenia constitutes the high end with complete commitment to EU institutions. Table 3.4. General attitude of the EU10 towards centralization Policy fields with Deviations from Accord index Uncertainty EU preference majority position Latvia 22 0 0.96 8.2 Poland 23 1 0.93 8.1 Cyprus 23 1 0.94 10.4 Slovakia 21 1 0.96 13.4 Lithuania 20 2 0.94 8.5 Slovenia 25 3 0.88 16.9 Czech Republic 19 3 0.93 10.1 Estonia 16 6 0.85 5.8 Hungary 17 7 0.81 7.7 Malta 13 9 0.75 10.7 Mean 20 3 0.89 10.0 Min 13 0 0.75 5.8 Max 25 9 0.96 17.0 Source: Eurobarometer; own calculations.
3 Who Shall Decide What? 55
A more sceptic view about the EU seems to persist in Estonia, Hungary, and Malta. These countries prefer a much higher number of policy fields kept with national authorities. Comparing the summary statistics, the similarity of within-group heterogeneity to the EU15 sample is confirmed. The minimum and the maximum number of deviations from the majority position are with 3 to 9 only slightly smaller than in the EU15 sample, and the accordance indices have exactly the same mean, with almost identical upper and lower bounds.
3.5 Conclusion The EU10 are clearly more enthusiastic about centralized EU decisions than the EU15. This is reflected in the new members’ preference for EU decision-making in 22 of the 25 policy fields in comparison to the much smaller set of 11 policy domains favoured by the EU15 countries. A natural explanation for this result is the EU10’s desire to have certain issues administered by EU institutions, where national authorities do not have the capacity yet. This also includes policy fields, like Health and Social Welfare, Education as well as Juvenile Crime, for which decentralized administration is clearly more efficient. Obviously, citizens in the new member countries prefer a slightly inefficient EU administration of policies for security and an improvement of livelihood than relying on their national politics and bureaucracy. As pointed out by Alesina et al. (2001a), preferences of the EU15 group are on average in line with an economically rational allocation of political domains to national and supranational decision levels. In addition, our analysis reveals that the preferences possibly reflect a demand for common EU policies and a joint EU position in response to rapid global changes, which is indicated by the correlation of the preferences over a mixture of policy domains that includes economy, security and international relations. Controversies among the EU15 are most distinct with respect to the policy cluster of Immigration, Asylum and Refugees, Defence and Labour market policies. The reasons for this preference distribution can be mainly traced to geographical characteristics, past political experiences and the perceived failure of national policies. Main political poles can be seen in France, Belgium, Italy, and Spain with their clear preference for centralized decisions regarding these policy areas in opposition to the U.K and the Scandinavian countries, which prefer decentralized decisions. Thus, in spite of possible constraints by the assignment of voting powers, the enlargement by 10 new countries with their clear bias towards centralization tends to
56
Joachim Ahrens, Martin Meurers and Carsten Renner
increase the difficulty for the latter fraction to influence EU politics according to their citizens’ preferences. A final point beyond the simple bipolar question of centralized versus decentralized decisions is the implicit country-specific heterogeneity that is reflected in the indicators S j in Tables 3.1 and 3.3. Notably, the high standard deviation of the preference indices for Currency in both groups could be an indicator for a persistent heterogeneity in the preferences regarding an appropriate monetary policy from each country’s perspective. Therefore, especially in this domain, joint EU decisions are likely to become more difficult to achieve in an enlarged EU.17 The preceding empirical analysis supports the view that greater flexibility is an unalterable precondition to cope adequately with the enlargementcum-deepening challenge and the problem of increasing heterogeneity. The strong preference of citizens in new member countries in favour of centralized policymaking creates the danger that the EU will be overtaxed with tasks, that should be í from a normative point of view í assigned to the national level of government. The current voting schemes, especially in the Council, may be suitable to constrain these demands at the margin. But, as the example of labour market policy shows, in fields in which the EU15’s preferences themselves are relatively diverse, it may occur that even the currently rigid voting rules cannot avoid a tendency toward greater centralization. Finally, EU members with relatively weak domestic governance structures should be encouraged by the introduction of a proper incentive scheme to improve the quality of their domestic economic and political institutions and hence craft a secure politico-institutional foundation of economic policymaking. Before accession the acquis communautaire had constituted such an incentive. After accession this leverage mechanism does not work effectively any more. Thinking about innovative ways to craft suitable incentives to improve the quality of national institutions one may point at distinct areas of policy harmonization, where deepening is obligatory such as monetary integration. In such a case, EU authorities could establish in addition to the Maastricht criteria further institutional requirements, which need to be fulfilled before introducing the common currency. A second way of thinking about incentives is to introduce more political competition, e.g., through fostering the instrument of enhanced cooperation or other mechanisms of subsidiarity. Such an approach towards differentiated integration may induce a competition of clubs within
17
For a discussion, see, e.g., Watson (2004).
3 Who Shall Decide What? 57
the club necessitating individual member countries to modernize their institutions in order to attract citizens and other mobile factors of production. Such an appropriate incentive scheme would enhance citizens’ trust in domestic institutions and confidence in national (and potentially regional) governments. Moreover, such incentives for institutional progress would contribute to an adjustment of citizens’ preferences, so that a greater, economically more efficient devolution of political competencies within the EU could take place and thereby greater flexibility of the European integration process could be realized.18
References Ahrens J, Meurers M, Renner C (2007) Beyond the big-bang enlargement: Citizens’ Preferences and the problem of EU decision making. Journal of European Integration, vol 29, no 4: 447-479 Ahrens J, Hoen HW, Ohr R (2005) Deepening integration in an enlarged EU: A club-theoretical perspective. Journal of European Integration, vol 27 no 4: 417439 Alesina A, Angeloni I, Schuknecht L (2001a) What does the European Union do? NBER Working Paper 8647, Cambridge, MA Alesina A, Angeloni I, Etro F (2001b) The political economy of international unions. NBER Working Paper 8645, Cambridge, MA Alesina A, Angeloni I, Etro F (2001c) Institutional rules for federations. NBER Working Paper 8646, Cambridge, MA Alesina A, Wacziarg R (1999) Is Europe going too far? NBER Working Paper 6883, Cambridge, MA Anderson CJ, Kaltenthaler KC (1996) The dynamics of public opinion toward European integration 1973–1993. European Journal of International Relations 2: 175199 Busch B (2007) Die Europäische Dimension: Deutschland und die Europäische Union im Mehrebenensystem. In: Institut der deutschen Wirtschaft (ed), Föderalismus in Deutschland. Ökonomische Analyse und Reformbedarf. Köln: 195216 European Commission (2002) Public opinion in the European Union. Report no 58, http://europa.eu.int/comm/public_opinion/ European Economic Advisory Group (2003) Report of the European economy 2003. CESifo Janning J, Giering C (1998) Differenzierung als Integrationskonzept der künftigen Europäischen Union. In: Bertelsmann Stiftung und Forschungsgruppe Europa
18
Regarding different concepts for greater flexibility see, e.g., Janning and Giering (1998), Janning and Weidenfeld (1996), and Ahrens et al. (2005).
58
Joachim Ahrens, Martin Meurers and Carsten Renner
(Eds) Systemwandel in Europa – Demokratie, Subsidiarität, Differenzierung. Gütersloh: 4152 Janning J, Weidenfeld W (1996) Das neue Europa. Strategien differenzierter Integration. In: Vorlage zum International Bertelsmann Forum Petersberg, 19– 20, January 1996 (Forschungsgruppe Europa am Centrum für angewandte Politikforschung/Ludwig-Maximilians-Universität München). Internet: http://www.cap.uni-muenchen.de Kaufmann D, Kraay A, Mastruzzi M (2005) Governance indicators for 19962004. Internet: http://info.worldbank.org/governance/kkz2004/ Oates W (1999) An essay on fiscal federalism. Journal of Economic Literature 37: 1120–1149 Watson A (2004) Economic and monetary Union: Of currencies and clubs. Journal of European Integration 26: 2539
4 Subsidiarity and Economic Policy
Jan Willem Oosterwijk
4.1 Introduction: subsidiarity
1
There are a several issues I would like to cover: x the European policy agenda; x the relation between subsidiarity and the political economy of reforms; x and the question whether subsidiarity can contribute to public support for Europe. But first, I would like to take a brief step back and emphasize the dynamic aspect of subsidiarity? Sir Leon Brittan, a former MEP and former European Commissioner probably put it best when he said: “subsidiarity is an ugly word but a useful concept”. Subsidiarity is not an invention of the EU. In fact, as Jacques Pelkmans stressed, the principle has been around for a long time. Today in Europe, subsidiarity is a useful instrument that enables us to maintain the dynamism that the Union needs while growing ever broader en deeper. Because subsidiarity is a dynamic instrument, it is not a static principle. The world changes, Europe changes and the Member States change. Moreover, the nature and scale of the problems that societies face change and hence the policies to accommodate them. It is this dynamism that leads to the shifting shares of responsibility between layers of government. 1
Taken from the closing speech of the conference on Subsidiarity and Economic Reforms in Europe, November 9, 2006, Brussels. The author presented his speech as Secretary-General of the Dutch Ministry of Economic Affairs.
60
Jan Willem Oosterwijk
4.2 The European policy agenda: more, less, different One important observation I would like to make here at the start, is that the EU is not doing a bad job at interpreting subsidiarity. Generally, most will agree on the added value of EU involvement in international trade, competition policy and the internal market. Let me remind you of the importance of a well functioning, extensive Services Directive, which is a priority at Community level. Most will agree too, that issues such as culture, education and social security belong primarily in the domain of the Member States. As Rick van der Ploeg and Reinhilde Veugelers in Chapter 5 set out for higher education: most reforms will have to come from the Member States, although there is a complimentary role for the EU (for example by increasing transparency). Now the question is where, according to the scholars present here today, should the EU have a larger role; and where should we leave more to the Member States? For that purpose we do not only have to discuss the most efficient level of policy making, but also the content of the policies. First, let’s take a look at an issue where I believe most will agree that the EU plays too large a part. Earlier I mentioned the importance of dynamism. The obvious example here is the Common Agricultural Policy or CAP. Post World War II, the fragile system of self-sufficiency with local markets made food supply the overriding priority. Today we have a completely different view: food is a world market. The CAP’s focus has shifted away from stimulating production to direct income support to farmers and the development of rural areas. Such policies can be much better shaped by Member States in line with their particular national characteristics and preferences. As Harold Grethe in (Chapter 12) said: “the assignment of the responsibility of the EU-level in the area is a historical artefact.” But we do need the EU for standards for quality, safety and environment. Are there any areas where I see a larger role for the EU? I see plenty examples on Lisbon related topics. Take innovation policy. We were presented with clear signs of economies of scale in public research and public funding of private research. It seems clear that a European innovation policy can supplement national policies. Secondly, energy policy. It is striking how fast things can change: a few days ago I read an article I wrote on economic policy in Europe maybe some three years ago. To my surprise it didn’t mention the word energy a single time. Energy policy is the cornerstone of a successful Kyotostrategy for sustainability. Moreover: today, in the face of Russian power politics and rising oil prices, it is quite clear there is a new role for the EU to play here, besides the internal integration and harmonization of energy
4 Subsidiarity and Economic Policy
61
markets, one of the vital network industries (c.f. Jordi Gual, Chapter 10). We are taking the first small ‘common’ steps, from Hampton Court to Lahti. It will have to develop, eventually into a real common policy, because otherwise the EU will be marginalized. As Benjamin Franklin put it: ‘we must all hang together or we shall all hang separately”. Now, should our agenda include new areas, areas where the EU currently does not make policy at all? We discussed, on the basis of Michael Keen and Ruud de Mooij’s contribution (Chapter 13) surprise observations of the actual state of tax coordination, a possible EU role in corporate taxes. Corporate tax diversity seems to distort the international allocation of resources, to create fiscal externalities and to entail high administrative and compliance costs. A common tax base is an interesting option. Different views are expressed on this topic in Chapter 14 to 16). We have collected some interesting items for an agenda for redistribution of responsibilities in the economic field. Interestingly, it not only concerns so much the level at which we shape policy, but also the particular content of the policy. That brings me to the second part of my contribution: How do we introduce these changes, these reforms? As the national representative of my country in the Lisbon process, I am all too aware of the permanent struggle of Member States to implement reforms. Introducing reforms at Community level, with 25 (soon 27) distinct nation states adds an additional level of complexity. Can the subsidiarity principle help us here? I am hinting at the subject of Jean Pisani-Ferry’s presentation2 and Sjef Ederveen, George Gelauff and Jacques Pelkmans (Chapter 2): political economy.
4.3 Political economy There are several challenges facing our continent: an expanding Union, globalization, rapid technological development, ageing. And as European Commission President, José Barroso, said: “status quo is not an option”. To face these challenges it is important that we reform: that we look at our pension systems, our health care, innovation systems, energy provision, competition rules, et cetera, et cetera. After five stages of enlargement the Union is more diverse than ever. Due to the increasing number of decision-makers and the increasing heterogeneity of economic structures, financial constraints, societal preferences, and political interests, it becomes more and more difficult to agree on 2
Reforms: is subsidiarity the answer?, see .cpb.nl/nl/goto/subsidiarity/.
62
Jan Willem Oosterwijk
common decisions. I believe that we need – and that there are – new ways to structure European cooperation. Dubbing Barroso’s theme: “one-sizefits-all is not an option”. I would like to take a closer look at three solutions in particular: the open method of coordination; enhanced cooperation; and sun set clauses. Open method of coordination
First, there is the open method of coordination. Its success relies on partnership and commitment. In the framework of the Lisbon Agenda, the EU prepares a Community Lisbon Programme with action at Community level; the Member States have their National Reform Programmes which they tailor to what they believe are their greatest challenges. Member States are stimulated by successful reforms in other Member States. The Lisbon process has created a certain mindset, a spirit of wanting to learn from each other and wanting to inspire each other on how to deal with these challenges. Allow me to give you some examples from my own country. At the top of our policy agenda is boosting innovation; increasing the number of hours worked, particularly by women; and strengthening competition in network industries. Assessing policy options we found inspiration in, for example, the first class Finish innovation framework; the excellent childcare system and labour market regulations in Denmark; and the tradition of open competition in the United Kingdom. Although we could do with a little more political sponsorship, I believe that, all in all, this partnership is a very interesting way forward. I believe it will only increase in the coming years, further enhancing the quality and the pace of reforms in Europe. Let me take the opportunity, by the way, to thank the Commission for its important, high quality contribution to the process. Enhanced cooperation
Greater diversity also imposes limits on the process of further deepening European integration the classical way. So-called enhanced cooperation could cater for Member States that are more ambitious in an area to deepen cooperation between themselves, while leaving the door open to other countries to join them at a later stage. No official use has been made of this provision yet, perhaps because the legal requirements are too strict and cumbersome to fulfil. But it is codification of something we have seen developing successfully in the past: the Euro found its origins in extra Treaty cooperation. As did the Schengen
4 Subsidiarity and Economic Policy
63
agreement. Today we have the North-West European Energy market and the Prüm Convention on cross border cooperation in combating terrorism, crime and illegal migration. And there are other possibilities. For example, maybe this may even be a way to bring the European patent issue forward…! Sunset clauses
The third ‘solution’ concerns abandoning or renationalising a certain policy field. Making policy is difficult, but changing it, or getting rid of it, is even more difficult. There is no possibility for something called enhanced ‘un-cooperation’…! The concept of sunset clauses may be a way of preventing regulation that has outlived its purpose from remaining in force. A sunset clause terminates or repeals all or portions of a law after a specific date, unless further legislative action is taken to extend it. Again nothing new: it was already used by the Romans. The very origins of the EU included such a clause: the European Coal and Steel Community ceased to exist 50 years after its establishment. And the EU Postal Directives have sunset clauses. A year ago the European Commission presented its plan to simplify EU legislation. Part of this plan is a screening and impact assessment of all new Community legislation. To help prevent obsolescence in future, the Commission will introduce in its legislative proposals either a sunset clause or a review clause. Now that I have addressed how subsidiarity helps shape the policy agenda and how new ways of cooperation may be necessary, I would like to return to the issue: can subsidiarity help reconnecting the citizens to the EU?
4.4 Public support Some see subsidiarity as the silver bullet. I feel divided. At the one hand, I see possibilities to use subsidiarity and particular the functional test to better explain to people why it is Brussels setting policy on certain issues. On the other hand, I am all too aware of the practical reality in which the outcome of the functional test is plainly not always reflected. The public notices this: I can imagine that for the proverbial “man in the street” it may feel as if it is always more, sometimes different, never less European rules. Indeed, ‘less’ turns out to be far more difficult. Besides the solutions I just discussed, a more consistent application of the subsidiarity principle may help the European Union to overcome some
64
Jan Willem Oosterwijk
of its problems. It will help building trust with the public as well as with policy makers, that the European Union is more than just another layer of bureaucracy. This will add to the credibility of European institutions. Open, objective discussions on the question whether topics should be tackled at European level or not can contribute to the restoration of trust in Brussels and to the correction of the perception that the EU meddles in just about everything. Before concluding, I would like to put forward some food for thought. Could building support for the EU be an argument for certain policy initiatives? Take for example the globalization adjustment fund. According to Mr Barroso: “this is not a fund to protect sectors that are no longer competitive, but a fund to help people. We want to show that the EU cares". The globalization adjustment fund does not pass the functional subsidiarity test but – as Mr Barroso observes correctly – it can help create public support for the European Union, at a time that it can definitely use some. I favour a practical approach here. Small wrong baits can be useful to catch the right big fish!
4.5 Conclusion Sometimes it seems as if the concept of subsidiarity raises more questions than that it provides answers. But for an open discussion over the distribution of competences it is essential that the subsidiarity question is taken on board. As we saw, subsidiarity is not a static principle but a dynamic instrument against which evolving policies should again and again be tested. It may not be a magic cure but it certainly contributes to keeping the Union healthy 50 years after the treaty of Rome.
5 Higher Education Reform and the Renewed Lisbon Strategy: Role of Member States and the European Commission
Frederick van der Ploeg and Reinhilde Veugelers
5.1 Introduction
1
In a recent survey ‘The Battle for Brainpower’, The Economist of 5 October, 2006 argues that talent has become the world’s most sought-after commodity. The greying of populations and the retirement of the babyboom generation throughout the developed world, the increasing globalization and the shift towards ‘tacit’ jobs makes the shortage of young graduates a serious problem. No economy can afford to be complacent about the war for talent. Alongside the rapid progress of China and India with their huge pools of young talented people and notwithstanding the greying of China, greying Europe faces particular challenges if it does not want to struggle to find enough engineers, scientists, doctors, lawyers and managers. During the last decade Europe’s annual growth in GDP per capita has been about 0.4 percentage points lower than in the US. If this continues for another decade, Europe’s GDP per capita relative to that of the US will fall back by a further quarter. Still, the investment rate and the capital-labour ratio are higher in Europe than in the US. (Aghion 2006) therefore stresses that Europe’s problem is not insufficient saving and investment, but lack of competition, not enough R&D and too little investment in education. As 1
An earlier version was presented at the conference ‘Subsidiarity and Economic Reform in Europe’, Brussels, 8-9 November 2006. We thank the participants as well as the editors for their comments. Veugelers acknowledges financial support from the Belgian Federal Government DWTC (IUAP P5/11/33) and the Research Fund of the KUL (0T/04/07A). The chapter reflects only the views of the authors and does not commit the European Commission.
66
Frederick van der Ploeg and Reinhilde Veugelers
secondary education was crucial to the post-war economy, so higher education has become essential for the development of the knowledge society, which demands increasing levels of supply of highly-educated, highlyskilled people. As Europe approaches the world technology possibility frontier and leaves the era of catching up behind, innovation becomes more essential and the returns on investing in higher education become even higher. European universities have enormous potential. But European universities are not currently in a position to achieve their potential in a number of important ways. As a result, they are behind in the increased international competition for talented academics and students, and miss out on fast changing research agendas and teaching curricula, and on generating the critical mass, excellence and flexibility necessary to succeed. These failures are compounded by a combination of excessive public control, bad governance coupled with insufficient funding. Europe needs universities able to build on their own strengths and to differentiate their activities on the basis of these strengths. Modernization of Europe’s universities, involving their interlinked roles of education, research and innovation, has therefore rightly been acknowledged not only as a core condition for the success of the broader Lisbon Strategy, but as part of the wider move towards an increasingly global and knowledge-based economy. Various policy communications have identified the main items for change.2 At the informal meeting at Hampton Court in October 2005 and the 2006 Spring European Council, R&D and universities were acknowledged as foundations of European competitiveness and stronger action at the European level was called for to drive forward this agenda for universities and research. In the National Reform Programmes based on the Integrated Guidelines for Growth and Jobs, i.e., the renewed Lisbon Strategy, Member States refer generally to these issues. Nevertheless, few make them national priority. Yet these changes are crucial to regenerate Europe’s growth capacity. This chapter will discuss the reforms needed to deal with the challenges ahead. It will investigate which policy actions are needed and at which level, to unleash EU’s potential of its higher education system. Within a subsidiarity perspective, we will focus particularly on what role the EU can have in reforming higher education in Europe, relative to Member States. But first we start with characterizing the major challenges facing EU’s higher education. While the focus of our discussion is mostly on higher education, the interlinkages of education and research cannot be ignored, since universities, 2
E.g. European Commission (2005a) and Council of the European Union ( 2005).
5 Higher Education Reform and the Renewed Lisbon Strategy
67
who are pivotal players in the higher education market, combine both activities. Section 5.2 discusses the facts and challenges of higher education facing Europe today. They concern rising enrolment rates, the rising private returns to education, the importance of access for pupils from less privileged backgrounds, governance problems and the substantial funding problems of institutions of higher education, lack of internationalization and underperformance in research and teaching. Section 5.3 puts forward our proposals for reform of higher education in Europe based on more autonomy for universities, higher tuition fees, more private funding, introduction of income-contingent loans, better governance, more competition among universities and a big leap forward in internationalization. We argue that these reforms will help to address the challenges of European universities. Section 5.4 takes a subsidiarity perspective and asks what the role of the EU can be in reforming the higher education sector. Compared to Member States, the EU provides the scope for creating an enlarged open market for higher education and a EU wide level playing field for universities. Apart from a sustained effort in providing mutual policy learning opportunities, cross recognition of qualifications and furthering the goals of the Bologna reforms, the EU should take a leading role in promoting mobility of students, researchers and teachers and opening up of national funding schemes for applications from other Member States. The EU should also take more initiatives in the context of the renewed Lisbon Strategy to fund research and education infrastructure build up through the Structural Funds and the funds for ‘Competitiveness for Growth and Development’, invest in EU flagships and facilitate global cooperation. The European Investment Bank could be a crucial driver in the higher education market by making income-contingent loans available. Section 5.5 concludes.
5.2 Challenges of higher education in Europe The EU counts almost 4 000 higher education institutions (of which about 2000 are universities strictu sensu who combine education with basic research), over 17 million students and some 1.5 million staff - of whom 435,000 are researchers. This system of higher education and research in Europe today faces a number of key challenges3. 3
Most of the data come from OECD (2006), European Commission (2005b) and European Commission (2005c).
68
Frederick van der Ploeg and Reinhilde Veugelers
5.2.1 Increasing demand for higher education The active population of the EU (25-64 years) has lower levels of higher education attainment than its main competitors in the global economy. The average level of higher education attainment among the active population in the EU is 21%, significantly lower than in the US (38%) and Japan (36%). Figure 5.1 gives the graduation rates for various OECD countries. Furthermore, in comparison with its most important competitors, higher education institutions in the EU attract a lower proportion of secondary school leavers, implying that higher education in Europe is still not an attractive option for a significant part of pupils having completed upper secondary education. About 25% of young people aged 18-24 years were enrolled in higher education in EU 25 in 2002, a much lower share than in the USA (37.7%). In the USA, tertiary students start to study on average at an earlier age than in Europe. Almost 40% of 18-year-olds in the US participate in higher education, compared to about 15% in the EU However, the EU is catching up. Despite low birth rates in the 1980s, the number of higher education students in Europe is increasing as a result of a growth in enrolment rates. The number of higher education students increased in the EU in the period 1997 to 2002 by 16% or on average by 3.1% per year, compared to an annual growth of 2.2% in the USA and only 0.1% in Japan. Figure 5.1 shows the increase in tertiary graduation in most countries (between 2000 and 2004). But at the same time, it highlights the heterogeneity among tertiary graduation rates across countries, particularly within the EU. The Scandinavian countries and the UK reach the highest level of graduation rates while Germany, France, Austria attain the lowest levels. While the EU produces more PhDs overall, it employs only 5.5 researchers per 1 000 employees, which is much less than the US (9.0) and Japan (9.7). The rapid growth of Asian universities, both public and private, is now also challenging Europe – and the US – in terms of the output of doctoral candidates in science and engineering (EC-RTD, Key Figures 2005). This will threaten the position of Europe as the most competitive and innovative economy in the world as stated in the Lisbon objectives. Of course, from the macroeconomic point of view, this new and growing supply of graduates and PhDs from Asia may benefit the ageing US and European economies
5 Higher Education Reform and the Renewed Lisbon Strategy
69
60 % 50 40 30 20 10
Turkey
Austria
Germany
Czech Republic
France
Switzerland
Slovak Republic
Israel
Hungary
Spain
Portugal
United States
Italy 2000
Japan
Sweden
Ireland
United Kingdom
Poland
Netherlands
Norway
Denmark
Australia
Finland
Iceland
New Zealand
0
2004
Fig. 5.1. Tertiary-type A Graduation rates (2000, 2004)4
Demand for higher education depends on interest in study and a variety of cultural reasons, but is also driven to a certain extent by the returns from education. Attaining higher levels of education can be viewed as an economic investment in which there are costs paid by the individual (including reductions in earnings while receiving education) that typically result in higher earnings over the individual’s lifetime. Not only do graduates experience relatively low unemployment rates and good employment prospects, they also enjoy a fast rising skill premium despite a massive increase in the number of graduates. Investment in higher education is thus becoming more and more lucrative. In this context, Figure 5.2 shows that the investment to obtain a university level degree, when undertaken as part of initial education, can produce private annual internal rates of returns as high as 22.6%, with all countries showing a rate of return above 8%. Costs of higher education (45 thousand euros) are much less than lifetime earnings, hence higher education is an excellent investment. From a lifetime perspective students will not be poor and can borrow more. Lifetime earnings in, for example, the Netherlands vary from 1.2 million euros for economics, medical, agriculture and tech4
Year of reference 2003 for: Finland, Denmark and France. Gross graduation rate may include some double counting for the Slovak Republic and Czech Republic. Source: OECD. www.oecd.org/edu/eag2006.
70
Frederick van der Ploeg and Reinhilde Veugelers
nical university male graduates to 0.9 for behavioural and social graduates and 0.8 million euros for arts male graduates (Jacobs 2002). 25 % 20
15
10
5
0 Belgium
Denmark
Finland
Hungary
Korea
males
New Zealand
Norway
Sweden
Switzerland
United Kingdom
United States
females
Fig. 5.2. Private internal rates of return on obtaining a university-level degree from an upper-secondary and post-secondary non-tertiary level of education (2003) Source: OECD. Table 9.6 (www.oecd.org/edu/eag2006).
While on average, the returns from higher education are considerable, the graph also shows heterogeneity in returns among countries. Countries differ significantly in the dispersion of earnings among individuals with similar levels of educational attainment. Also, as Table 5.1 indicates, graduate unemployment rates differ, with evidence that returns from higher education, in the form of expected employment rates, are lower in the EU. Table 5.1. Graduate unemployment rates in 2003 EU 25 USA Japan % Unemployment rate of population aged 25-64 with tertiary education attainment Unemployment rate of population aged 25-64 with less than secondary education Source: EUROSTAT and OECD.
4.2
3.0
3.9
10.3
10.2
6.6
5 Higher Education Reform and the Renewed Lisbon Strategy
71
5.2.2 Improving access to higher education Access to higher education is highly dependent upon successful participation at earlier stages of education. Efforts to improve the quality of higher level education will clearly be in vain if standards at school level are deficient. Therefore upper-secondary completion rates and key competency levels at the end of lower-secondary education are important indicators of the degree to which pupils have been prepared and motivated for higher education after initial schooling. Completion of upper-secondary education is also increasingly important, not just for entry into the labour market, but also for the access it allows to higher education and for paving the way to participation in Lifelong Learning activities. This is why one of the five European benchmarks requires that, by 2010, 85% of 22-year-olds in the EU should have completed upper-secondary education. In 2004 the EU average was 76.4%, which is still lower than levels in the leading non-EU OECD countries.
5.2.3 Governance problems in supplying higher education services European universities often are smaller than their counterparts in comparable OECD countries. To this smaller scale, increasing objectives are projected. Policy-makers have been pushing universities to play a greater role as social actors, and to create ‘social value added’ by extending their role in society. Universities are required not only to play an active role in education, absorbing the increasing mass of incoming students at bachelor level, at the same time being asked to produce high quality basic research and turn this basic research into commercial applications. While in the EU about 2000 universities are also engaged in research and deliver postgraduate diplomas leading to Master and Ph.D. degrees, in the US, out of 4000 higher education institutions, only 500 deliver postgraduate education and only 150 universities are research universities (European Commission, 2005a). US universities thus seem far more specialized than in the EU and most of them only provide bachelor education. Some of these patterns of specialization may be related to stronger competition in the US market for higher education. Beyond an often too small scale and insufficient focus, European universities suffer from bureaucracy and lack of autonomy. The overregulation of university life hinders modernization and efficiency. Ex ante control hinders universities’ capacity to react swiftly to changes in their
72
Frederick van der Ploeg and Reinhilde Veugelers
environment and to engage in international competition. Many parameters are fixed: subsidies per student are fixed, tuition fees cannot be varied, the number of places for each course is often fixed by the ministry of education, and applicants cannot be refused once they have passed their national exams. Nationally defined courses and employment rules for academic staff tend to inhibit curricular reform and interdisciplinarity. Inflexible admission and recognition rules impede mobility. Further details are given in Jacobs and Van der Ploeg (2006). While a number of EU countries have started off reforming their higher education sector (like Denmark, Ireland, the Netherlands and the UK), the governance of universities remains very centralized, state controlled in France, Greece and Italy, or at the level of regions (in Germany, Spain, Belgium). 5.2.4 Funding problems Universities have to cope with a wide range of disparate tasks while being faced with increasing budgetary constraints. 5.2.4.1 Funding for higher education
The EU-25 devotes a much lower share of its wealth to the financing of tertiary education than the US. In 2001, the EU spent 1.3% of its GDP on the financing of tertiary education compared to 3.3% in the US and 1.2% in Japan. Although public funding of tertiary education is also higher in the US than in the EU, the most striking difference between the two regions concerns private expenditure. In relative terms, private expenditure on higher education is nines times higher in the US than in the EU. Table 5.2 indicates that the difference between the EU and the US is less marked when one considers all levels of education. Within the EU, the Scandinavian countries have the highest share of tertiary spending in GDP (most of this spending being public). While Germany, France and the UK spend a bit more than 1%, Italy has an even lower share. Similar gaps show up in yearly spending per student. While the EU is spending on average 8,600 euro per student, the US is spending on average 20,000 euro. For example, the Ivy League universities now charge more than 40.000 US dollars including board.
5 Higher Education Reform and the Renewed Lisbon Strategy
73
Table 5.2. Public and private expenditure on education as % of GDP, 2001a Expenditure
Tertiary education Public Private
All levels of education Public Private
Belgium 1.36 0.21 6.11 Czech Republic 0.80 0.13 4.16 Denmark 2.73 0.04 8.50 Germany 1.12 0.09 4.57 Estonia 1.07 . 5.48 Greece 1.19 0.00 3.90 Spain 1.01 0.30 4.41 France 1.02 0.16 5.76 Ireland 1.24 0.20 4.35 Cyprus 0.81 0.20 4.98 Latvia 1.21 0.79 6.28 Lithuania 0.90 0.54 5.75 Luxembourg 1.34 . 5.92 Hungary . . 3.84 Malta 1.11 0.26 5.15 Netherlands 0.88 0.02 4.47 Malta 1.32 0.28 4.99 Austria 1.35 0.06 5.70 Poland 1.07 . 5.56 Portugal 1.09 0.09 5.91 Slovenia 1.33 0.45 6.13 Slovakia 0.83 0.05 4.03 Finland 2.05 0.06 6.24 Sweden 2.05 0.20 7.31 UK 0.81 0.30 4.69 EU-25b 1.08 0.20 5.10 US 1.48 1.77 5.08 Japan 0.54 0.61 3.57 a Source: DG Research; data: Key Figures 2005, Eurostat. b The values for EU-25 are estimations.
0.44 0.41 0.28 0.98 . 0.23 0.59 0.48 0.35 0.32 1.31 0.70 . 0.00 0.57 0.85 0.45 0.32 . 0.09 0.85 0.12 0.13 0.21 0.81 0.60 2.22 1.17
This spending gap between the EU and its major international competitors can be correlated to the financing mode of higher education. In the EU most of the financing of higher education is public funding, where the State is seen as the provider of education services as public goods with education being mostly ‘free’ with low fees and low private funding through foundations and donations. While in the EU private spending on higher education represents on average 0.2% of its GDP, this is 1.8% in the US.
74
Frederick van der Ploeg and Reinhilde Veugelers
The percentage of funding for tertiary education coming from private sources varies widely across countries, from less than 4% in Denmark, Finland, Greece, Norway to more than 50% in Australia, Japan and the United States, and even above 80% in Korea. In some countries, tertiary institutions are now relying more heavily on private sources of funding than they did in the mid-1990s. In addition, the nature of public funding for education varies considerably across countries and time. Governments rely more and more on (lumpsum) ‘block grants’ with both output and input criteria. Most countries fund on the basis of inputs such as number of enrolled students (Australia, Belgium, France, and New Zealand). Funding in Denmark stresses output, since universities receive funding on number of grade points that students receive (the so-called ‘taxi-meter model’). The Netherlands and Sweden take intermediate positions. About half of funding in the Netherlands depends on the number of diplomas. A similar share of resources depends on number of grade points in Sweden. Germany and the UK differ as funds are allocated on historical grounds independently of the number of students or output criteria, but funding is based on negotiations and enrolment forecasts. However, the UK government puts a growing emphasis on output and performance in teaching and research. With respect to the types of funding (fees, grants, loans, sponsoring, etc.), Figure 5.3 indicates that there exists large heterogeneity. No tuition fees exist in Denmark, Germany and Sweden.5 Other countries have centrally determined tuition fees that may differ between various fields of study (Australia, France, Netherlands, and the UK). Typically, prices charged to students do not depend on costs. In recent years, governments (Australia, Belgium, France, Netherlands and UK) increased tuition rates to maintain resources per student in the face of increasing enrolment. This also happened in the US and New Zealand where institutions are free to set fees. Some countries (Netherlands, UK) have decreased student grants and increased loan facilities. In contrast, Germany, Sweden and Denmark, have increased grants and loan facilities. France only increased grants. New Zealand and Australia both substantially increased loan facilities. Conditions governing student grants have become tighter by linking grants/loans to academic progress in Denmark, Germany, Netherlands, and Sweden.
5
However, Germany has just announced promising new reforms: up to ten new ‘elite’ universities will get an extra 100 million euro for research, greater autonomy for the Länder governments with respect to finance, appointment of professors, management, etc., possibility of introducing tuition fees in 9 of the 16 Länder, and introduction of student loans from private and public banks.
5 Higher Education Reform and the Renewed Lisbon Strategy
75
50 %
45 40 35 30 25 20 15 10 5
scholarships / others grants to households
transfers and payments to other private entities
Japan
Korea
Iceland
United Kingdom
Mexico
Switzerland
Turkey
Poland
Greece
Czech Republic
Spain
Portugal
France
Netherlands
Sweden
Ireland
Norway
Hungary
Germany
New Zealand
Belgium
Australia
Italy
Slovak Republic
Finland
Austria
Denmark
0
students loans
Fig. 5.3. Types of public subsidies for higher education (2002) Source: OECD. Table5.2. See Annex 3 for notes (www.oecd.org/edu/eag2005). 5.2.4.2 Funding for research at universities
Within the Lisbon Strategy, the Barcelona European Council recommended that spending on R&D in the EU should approach 3% of GDP by 2010 and that one third of that should come from the public sector. This also has implications for higher education funding since universities are expected to be beneficiaries of parts of the additional funding for R&D. R&D performed in the higher education sector is on the rise in Europe, Japan and the US. In 2003, higher education expenditure on R&D amounted to 0.44% of GDP in the EU as a whole, well above its 1997 level (0.38% of GDP). Within the EU, the three Nordic countries Sweden, Finland and Denmark showed the highest intensity of higher education R&D in 2003, with values above 0.60% GDP. Austria and the Netherlands were also above the EU average. On the other hand, most of the new Member States (except Lithuania and Estonia) were far below the EU average. In both the US and Japan, higher education expenditure on R&D amounted to 0.43% of GDP in 2003, compared to, respectively, 0.37% and 0.41% of GDP in 1997. Business support for R&D in the higher education sector is substantially higher in the EU (6.6%) than in either the US (4.5%) or Japan (2.6%). In
76
Frederick van der Ploeg and Reinhilde Veugelers
2002, the differences between Europe and its competitors in the levels of government R&D funded by the business sector were even wider. In Europe growth can only be witnessed in the level of higher education R&D financed by the business sector. 5.2.5 Underperformance in delivering quality education and research 5.2.5.1 Quality of university education
A signal of the difficulty of the university systems in the EU to deliver on their education goals, is the high dropout/failure rate as well as the longer duration of tertiary education. At present, too many enrolled students leave the European universities without an academic degree. According to OECD data survival rates in higher education in the 13 EU countries for which data was available amounted to only 66% in 2000, compared to an OECD average of 70% and a rate of 66% in the US, 79% in Korea and 94% in Japan.6 The high survival rates in East Asia relative to those in Europe and the US are also related to specific attitudes towards education.7 Survival rates in Europe vary widely between countries with highest rates in Ireland (85%) and the UK (83%) and relatively low rates in Sweden (48%) and Italy (42%). A tendency to uniformity and egalitarianism in many national systems has ensured that the average quality of European universities, while generally homogeneous, is comparatively good – at least academically. But there may also be deficiencies stemming from insufficient differentiation. Most European universities tend to offer the same monodisciplinary programmes and traditional methods geared towards the same group of academically best-qualified learners, but in many countries there is rapid rise in specialized master programmes.
Survival rates are calculated on the basis of the number of graduates divided by the number of new entrants at the typical age of entrance. 7 Education is among the most important values acknowledged by Asian families. This leads to an attitude favouring high private investment in education in terms of time and financial resources and a strong appreciation of formal degrees. 6
5 Higher Education Reform and the Renewed Lisbon Strategy
77
5.2.5.2 Research quality
In terms of total number of scientific publications, the EU outperforms the US and Japan. However, if one adjusts for population, European claimed leadership in publication disappears, as Table 5.3 suggests. Table 5.3. Publications and citations weighted by population and university researchersa Publications / Population US EU-15 UK Germany France Italy
4.64 3.60 5.84 3.88 3.96 2.58 Citations / Population
US EU-15 UK Germany France Italy
=
=
Top1% Publications/ = Population
a
*
6.80 4.30 6.99 4.77 4.09 5.83
39.75 23.03 42.60 26.82 25.81 16.89
US 0.09 EU-15 0.04 UK 0.08 Germany 0.05 France 0.04 Italy 0.03 Source Dosi et al (2006).
Publications / Researchers
Citations / Researchers
0.68 0.84 0.84 0.81 0.97 0.44
58.33 27.52 51.00 32.98 26.68 38.25 Top1% Publications/ Researchers 0.13 0.04 0.10 0.06 0.05 0.06
Researchers / Population
Researchers / Population 0.68 0.84 0.84 0.81 0.97 0.44
Researchers/ Population 0.68 0.84 0.84 0.81 0.97 0.44
Our calculations based on numbers by King (2004) and OECD (2004a). A number of publications, citations and top 1% publications refers to 1997-2001. Population (measured in thousands) and number of university researchers (measured in full time equivalent) refer to 1999.
Moreover, in science, together with the numbers of publications, at least equally important, are the originality and the impact of scientific output upon the relevant research communities. Two among the most used proxies of such an impact are articles’ citations and the shares in the top 1% of
78
Frederick van der Ploeg and Reinhilde Veugelers
most cited publications. The US is well ahead with respect to both indicators. In particular, controlling for population, EU output is still less than half than the US one. In the second and third column of the same table, the output (i.e., number of publications, citations, and top 1% publications) per population indicator is decomposed into two components: a measure of scientific productivity of university researchers (i.e., output per university researcher) and an index for the intensity of university researchers on population. Table 5.3 clearly shows that US leadership is due to the quality of research published rather than due to the sheer number of researchers. In a different context, Alesina and Spolaore (2003) argue that countries with a large population may benefit from returns to scale and be more efficient in providing public goods and generate higher productivity. Within the context of the market for higher education and research, it is clear from the law of large numbers that in such countries the chances of a genius surfacing in research is larger than for a small country. This is why it is important to engender competition (as well as cooperation) on a European level between the top researchers and degree programmes. However, the evidence so far fails to support that the number of top universities per million inhabitants is an increasing function of the size of the population (Ederveen and Thissen 2008). However, historical empirical comparisons neglect the potential of upcoming countries with a huge population like China and India. Given the intense competition to get into the top universities in China and India, one should not be surprised to see during the next few decades many more top universities in Asia. 5.2.6 Increasing international competition European higher education remains fragmented - between and even within countries - into medium or small clusters with different regulations and, naturally, different languages. But European universities, when attracting students, faculty and funding, are increasingly being faced with an international competitive arena that becomes ever more agile. Students increasingly travel abroad in search of the best study and research opportunities. In 2003, 2.12 million people studying in OECD countries were foreign students, i.e. enrolled outside their country of origin. This represented an 11.5% increase in total foreign students’ intakes reported to the OECD since the previous year. In the UK 13% of students enrolled in higher education is foreign (on basis of country of residence or secondary education). Most notably, Australia, France, Germany, the
5 Higher Education Reform and the Renewed Lisbon Strategy
79
United Kingdom and the United States receive 70% of foreign students in OECD countries. Students from China, India and Southeast Asia comprise the largest numbers of foreign students from partner countries. This global competition among higher education institutes bears out most markedly when international ranking are being published in the press. The Times Higher Education (2006) ranking of the world’s top 200 universities, for instance, considers peer review, recruiter review, international faculty, international students, student/staff ratios and faculty citations scores. Interesting is that Table 5.4 indicates that 41 and 42 of the top fifty universities in, respectively, 2005 and 2006 are from countries with an Anglo-Saxon system of education. Continental Europe (excluding Switzerland) only had three universities in the top fifty in 2005 and this has dropped to only two in 2006. Table 5.4 also gives a summary of country scores in the 2006 rankings. We see that, apart from the elitist Ecole Normale Supérieure and the Ecole Polytechnique in Paris, no universities of continental Europe feature in the 2006 top 50. Heidelberg University was in the 2005 top 50, but dropped out of the 2006 top 50. Australia does again surprisingly well with six universities and the same can be said for New Zealand newcomer Auckland University. China with Beijing University, Tsing Hua University and Fudan University and the China University of Science and Technology and Nanjing University will rapidly catch up in the rankings. The same is true for Indian universities like the Indian Institutes of Technology, the Indian Institutes of Management and Jawaharlal Nehru University. Continental Europe has 48 universities in the 2006 top 200 and especially the Netherlands is catching up. Still, they need to do a lot better to keep up the competition with their US counterparts and the rise in the number of top institutes in China and India. Table 5.4. Number of universities per country in the Times Higher Education (2006) top 50 Top 50 US UK Australia/New Zealand Asia excl. Hong Kong / Singapore Hong Kong / Singapore Canada Switzerland France Germany Total
2005 20 8 6 4 4 3 2 2 1 50
2006 22 8 7 4 2.5 2.5 2 2 0 50
80
Frederick van der Ploeg and Reinhilde Veugelers
5.3 Reforming higher education in Europe The previous analysis has shown that the EU needs to improve access to higher education, increase higher education attainment levels, and increase total investment in higher education, both public, but even more private. European universities also need to catch up with the best universities in the Anglo-Saxon world. It is crucial to improve incentives and generate the funds to be able to compete globally. European universities have much less financial resources per student than in the US, are not competing much with their counterparts at home and abroad and are focused too much on the national rather than the global market. Funding higher education will become increasingly more challenging due to the relentless operation of Baumol’s cost disease8. If the EU has to make an effort to bridge its funding gap on higher education, be it public or private, this can only be realized if at the same time the governance of the higher education system is tackled. This is necessary to increase the efficiency of spending by these organizations, thereby delivering results. To attract more funding, universities first need to convince stakeholders governments, companies, tax payers and above all students – that existing resources are efficiently used and would produce added value for them. Higher funding cannot be justified without profound change. Providing for such change is the main justification and prime purpose for fresh investment. Given the prevalence of overlong study durations, high dropout rates and/or graduate unemployment in Europe, investing more in the current system could be perceived as unproductive or even counter-productive. Yet, combined under-funding and system rigidities are so acute in some countries of the EU that they impede the reform process at universities, who are consequently trapped in a vicious circle. If Member States are to break this vicious circle, they need to combine more and better targeted funding simultaneously with reforms of the supply side, thus creating the necessary conditions to enable universities to improve their performance, to modernize themselves and become more competitive. This implies granting universities much more autonomy while at the same time demanding them to be more accountable for delivering results.
8
Teaching and research basically need to be done by highly qualified people and is more difficult to be replaced by technology. Productivity growth in universities inevitably lags behind that in the manufacturing, so the cost and price of university education inevitably rise over time. This is Baumol’s cost disease applied to higher education (e.g., Jacobs and Van der Ploeg 2006).
5 Higher Education Reform and the Renewed Lisbon Strategy
81
5.3.1 More performance-based funding Most existing funding schemes suffer from ‘ratchet effects’ arising from budgeting and accounting procedures. University governors do not pursue cost-effective policies, because the government cream skims or even penalizes cost savings. Universities are not very aware of how much each programme costs. They do not use rational cost-based criteria and allow for various cross-subsidies. Clearly, European universities have a need for more performance-based funding. Both output and input funding have unintended side effects. Output funding has the unintended disadvantage that it induces grade inflation and reduces incentives to cut costs. Input funding does not induce grade inflation, stimulates efficiency, but leaves monopolistic practices in tact. One thus has to strike a tough trade-off between, on the one hand, avoiding grade inflation and inefficiently run universities, and, on the other hand, curbing monopolistic practices. Countries that rely on substantial output funding therefore often have quality safeguarding committees or make use of external examiners. If there is a lot of uncertainty and efforts correlate little with performance, high-powered incentives become less attractive. Governments should strike the right balance between core, competitive and outcome-based funding (underpinned by robust quality assurance) for higher education and university-based research. Competitive funding should be based on institutional evaluation systems and on diversified performance indicators with clearly defined targets and indicators supported by international benchmarking for both inputs and economic and societal outputs. Funding should be based on less malleable criteria. It is also important to ‘move the post’ every few years, to avoid researchers and institutions to focus too much on the measurable targets only. More important, difficult to measure and unrewarded activities (e.g., pastoral care of students, refereeing articles or helping graduates to get a good job) may be crowded out by easy to measure activities (e.g., peer reviewed publications). Outstanding quality can only emerge from an across-the-board ‘culture of excellence’. Excellence is never a permanent achievement. It always needs to be challenged. It can exist in a few entire universities, but much more widely in individual faculties or teams within institutions or networks. This requires concentration of resources.
82
Frederick van der Ploeg and Reinhilde Veugelers
5.3.2 More private funding Free higher education does not by itself suffice to guarantee equal access and maximum enrolments. This casts the much debated issue of tuition fees in a fresh perspective, isolated from the discussion on access, which is better targeted through other instruments, such as income-contingent loans and scholarships (see Section 5.3.3). The debate on social and private returns from higher education has highlighted its role as an investment, benefiting both the individual (through higher income and status) as well as society as a whole (through higher employment rates, lower social costs and later retirement). If social returns exceed private returns, education causes positive external effects to society and the government should support education. Estimating macro-economic production functions where total output is explained by human as well as physical capital, one obtains macro returns to education at the lower end of the estimated micro returns, suggesting there is no strong case for social returns to be substantially higher than private returns (cf., Blundell et. al. 1999). However, empirical findings suggest that private returns to higher education are substantial (Jacobs and Van der Ploeg 2006). All this evidence suggests more scope for private funding of higher education and in particular for asking students to pay higher tuition fees.9 How should fees be set?
There is evidence that unobserved heterogeneity is at least as important as observable variations in attendance and inputs as class size and number of teaching assistants (e.g., Martins and Walker 2006). Peer effects are important in higher education (e.g., Sacerdote 2001; Willams and Zimmerman 2003). Education is a ‘customer-input technology’, since students are both consumers and co-producers of education. Selecting and attracting the smartest students thus generates a positive feedback loop as it raises the quality and reputation of the institute and thus increases further demand. Having high-quality students improves academic excellence all round and makes it possible to attract much better employees/professors and funding from sponsors and the state. 9
Canada is an interesting testing ground, since provinces levy different fees. Evidence suggests that rising fees by about 2,000 dollar in the 1990’s reduced the probability of participation by persons aged 17, 18 and 19 relative to trend by amount 2 %-points; nevertheless university participation increased dramatically during this period (Johnson and Rahman 2005). Unfortunately, this interesting study did not take account of factors like family income or parental education.
5 Higher Education Reform and the Renewed Lisbon Strategy
83
Without peer group or reputation effects, profit maximizing universities set prices to a mark-up on marginal cost. The mark-up should be particularly high for courses with low price elasticity of demand (e.g., courses followed by local students or courses for which not many substitutes exist). Most students go to their local university, perhaps as they prefer familiar surroundings. The optimal tuition fees are higher for such students, because their price elasticity of demand is lower. If peer group and reputation effects matter, tuition fees are higher for the less able or less motivated students and lower for the smart students (cf. Rothschild and White 1995). Hence, universities should award scholarships or give discounts to the brightest students, especially if they come from less privileged backgrounds. By selling below cost, universities induce permanent excess demand for their courses and can thus select the smartest students and pursue excellence. Unfortunately, the European system with its sometimes not very helpful emphasis on equality, implicitly entails cross subsidies from the smart to the less able students. Europe thus still has a long way to go in this respect. Of course, the main problem with tuition fees in Europe is that they are set centrally and do not vary according to demand and supply or to meet the special needs of universities. Typically, fees are too low and too undifferentiated, thus encouraging ‘fun seeking’ students and an enormous mismatch of students to courses. How to set government subsidies?
A government that maximizes utilitarian social welfare (graduate utility minus tuition subsidies), has no access to non-distortionary taxes and does not have any merit motives for intervening in higher education, would set subsidies equal to zero (Jacobs and Van der Ploeg 2006). The market outcome is efficient. However, the government may support merit studies that are of interest to society as a whole and will not be provided by the market, while generating public benefits (‘educational welfare’). Subsidies should be optimally targeted to fields of study that have the largest social returns. Furthermore, subsidies should be targeted towards the students that appear to generate most social value. Subsidies should be directed towards studies with a large social value, not a large private value. The mere fact that for some disciplines the marginal benefits are mainly non-monetary is not a reason for government subsidies. That will lead to over-investment in those disciplines. Students will take account of immaterial benefits themselves. In general, public subsidy should be high for merit studies, zero for studies with only a market rate of return and negative for studies with rent
84
Frederick van der Ploeg and Reinhilde Veugelers
seeking and signalling externalities. Clearly, subsidies should be low if the cost of public funds is high and high when the elasticity of student demand is low. It follows that, in contrast to current practice, subsidies should be differentiated by type of study and type of degree programme. Clearly, the government has insufficient information about the preferences of individual students and the supply of courses and may wish to use vouchers rather than subsidies to universities. By giving students personal vouchers, which they can use to pay for their courses, the government encourages students to ‘vote with their feet’. This fosters competition between universities. 5.3.3 Addressing accessibility through income-contingent loans Empirical research suggests that the ability of the student and long-run background factors (‘culture’, ‘family’, ‘environment’) are the most important determinants of enrolment in higher education. Increasing enrolment in higher education of children from lower socio-economic backgrounds requires therefore first and foremost intervention in basic and secondary education (de la Fuente 2006). Universities should be accessible to all with sufficient academic capabilities. But this does not imply that higher education should be free from charge, neither does it imply that all should pay the same price, or should pursue the same quality of education. If the purpose of low fees is to guarantee access to universities, and not income equality, an income-contingent loan scheme is sufficient. From a lifetime perspective, the key problem of students is not poverty but insufficient access to credit markets. To tackle student poverty as a barrier to education and to avoid students taking disruptive part-time jobs, students should be allowed to borrow for fees and cost of living. As the Australian experience indicates, incomecontingent loans (ICL) can overcome problems of capital market imperfections with risk-averse students (Jacobs and Van der Ploeg 2006). ICL only require students to pay back principal and interest if their incomes after graduation are high enough. ICL thus offer a combination of loans and social insurance. If income risks of graduates are pooled, fewer subsidies are needed to eliminate risk aversion. Commercial banks and insurers are unable to write contracts based on future incomes, but the government can enforce contracts through the tax authorities and verify earned incomes. By selection and tracking of student performance and denying funds to non-performing students, the government can more easily eliminate the ‘rotten apples’. In principle ICL feature
5 Higher Education Reform and the Renewed Lisbon Strategy
85
no subsidies. Also, interest subsidies should be avoided as this only encourages excessive study and financial arbitrage if necessary with the help of wealthy parents. However, the risks of default may be borne by society. Shifting from grants towards income-contingent loans and at the same time introducing selection at the gate of universities and afterwards are a more efficient and equitable use of scarce public funds. It can avoid subsidies going to the less diligent, less talented students from richer families rather than to the smartest students from poorer families. 5.3.4 More competition among universities In response to scarcer public budgets, a rationalization of the supply side of the higher education market is taking place. The resulting increase in the scale of universities has however generated the danger of creating (local) public monopolies. In the Netherlands the enormous increases in scale and monopolistic practices, especially in institutes of professional higher education, have gone hand in hand with huge increases in overhead and capital expenditures leading to substantial falls in resources for teaching (Jacobs and Van der Ploeg 2006). Such monopolies are more harmful when government shifts to output financing. They then reduce quality (‘grade inflation’), ignore demand of students and employers, and increase overhead costs. Monopolistic price setting may in addition drive up tuition fees and lower both quantity and quality of supply of education, especially if the price elasticity of demand is low. This may be less so if there are increasing returns to scale, but the larger scale have typically reduced rather than increased outlays per student. Barriers to enter the market for higher education should be lowered by abolishing historical funding and barring cross-subsidies that hinder fair competition. Both private and public universities are better able to compete if subsidies are allocated directly to students through vouchers/grants. Students can spend the vouchers on the institution and courses of their preference. A level playing field can open national markets to the international environment, especially if students can get student loans for study abroad and can spend their vouchers abroad. Vouchers will engender competition between institutions. This need not necessarily lead to grade inflation, since the quality of the courses offered is an important way of competing in the market for higher education. To make the higher education market more transparent, it helps if an independent authority publishes yearly performance criteria of universities. These criteria should cover dropout rates, average enrolment durations, average exam marks, student evaluations, quality of scientific publications,
86
Frederick van der Ploeg and Reinhilde Veugelers
evaluations of independent scientific committees, graduate performance in the job market, etc. 5.3.5 Better internal governance of universities Universities will not become innovative and responsive to change unless they are given real autonomy and accountability. In return for being freed from over-regulation and micro-management, universities should accept full institutional accountability to society at large for their results. In many countries this would mean a new approach to policy making, with less ex ante checks and greater ex post accountability of universities for quality, efficiency and the achievement of agreed objectives. For universities, this requires new internal governance systems based on strategic priorities and on professional management of human resources, investment and administrative procedures. Care must be taken not to base governance on the model used in commercial business enterprises. Governance should take account of the fact that universities consist of professionals. Too much external incentives can crowd out intrinsic motivation. A pivotal area of university management is personnel management. Human resources are a core determinant of quality in higher education and research. Universities must therefore work to enhance their human potential, both qualitatively and quantitatively, by attracting, developing and keeping talent in the teaching/research career. Excellence can only emerge from a favourable professional environment based in particular on open, transparent and competitive procedures. Vacancies for professors and researchers should be advertised publicly, and internationally. Compensation should reward quality and achievement in the performance of all tasks. Mobility across national border and between university and industry should be nurtured.
5.4 What is the role of the EU? The agenda mapped out in Section 5.3 is by now, in essence, fairly established in the policy debate on higher education reform in Europe. Action is primarily needed from the public authority and universities, but the students will have to become much more critical and vote with their feet if they are unhappy with the degree programme being offered. On the side of public authorities, Member States (and not the EU) are primarily responsible for the organization of their higher education. The higher education landscape in Europe is and needs to remain in the future diverse with re-
5 Higher Education Reform and the Renewed Lisbon Strategy
87
spect to languages, culture, systems and traditions, thus keeping policy competence in higher education mostly decentralized. But alongside the fundamental local, regional and national roots of universities, the European framework is becoming increasingly important. Our actions suggested for the EU, take full account of the principle of subsidiarity, thus recognizing that the Commission is not a direct actor in the modernization of universities. Competence should only be shifted to the EU level if good arguments suggest that this will indeed improve welfare. There are advantages of scale and of organizing policy at the EU level. The Commission first can play a catalytic role by providing political impetus, policy learning and coordination opportunities and targeted interventions in support of reform and modernization at Member State level. In addition, the European dimension offers the potential benefits of larger scale operation, greater diversity and intellectual richness of resources, plus opportunities for cooperation and competition between institutions in a European wide integrated higher education market. A more integrated higher education market spurs competition between European universities and thus boosts quality of education and innovation in Europe. There is therefore a case for increased mobility of students studying in other Member States. This can be aided with special EU grants and loans and by standardization and harmonization of types of degrees. It is also desirable to have increased mobility of professors and especially young researchers, since this leads to exchange of information and ideas and will benefit the quality of teaching and research. Furthermore, competition for research funds at the EU level will lead to a higher quality of research projects. 5.4.1 Coordinate policy and provide mutual policy learning opportunities The Commission can support a new political impetus for higher education reform via coordinated interaction with Member States through the open method of coordination, identifying and spreading best practice and supporting Member States in their search for more effective university regimes through voluntary cooperation. Beyond providing a forum for coordination and mutual learning among Member States, the EU can also offer added value, by providing supporting surveys and studies analysing the characteristics of best practices. Furthermore, designing and providing a standardized set of relevant indicators to measure higher education performance in terms of funding and outputs will help to assess progress on reforms.
88
Frederick van der Ploeg and Reinhilde Veugelers
5.4.2 Towards an internal market for higher education in Europe An enlarged market for higher education is crucial in order to create more opportunities for citizen choice and mobility, as well as to break national monopolies or tight oligopolies at local supply level. 5.4.2.1 Cross recognition and standardization
An EU internal market for higher education requires sufficient compatibility between the different national regulations. Cross-recognition of qualifications and competencies demands a minimum level of organization at the European level in the form of common references and basic standards. One area of standardization is quality assurance. Quality depends primarily on a ‘culture of quality’ and on an internal quality assurance within universities. But the accountability of universities to society also requires an external system of quality assurance. Europe’s universities need quality seals with international credibility. Obviously, this will facilitate international mobility of students and faculty. In particular, the EU should take the lead in evaluating indicators by which quality of various types of degree programmes can be judged. Such indicators for higher education are essential for a better match between students and courses and for students being able to judge the best place for them to go to. 5.4.2.2 Bologna reforms
Work in the context of the Bologna process is bringing about a convergence in the structure and length of degree programmes towards the Anglo-Saxon degree system. x This convergence will engender competition between a larger number of shorter degree programmes. If students are unhappy with a particular degree programme, they will vote with their feet and go to another programme. The reforms boost international exchange and fuel competition. x It makes the European system compatible with systems of higher education found in UK, US, Canada, Australia, New Zealand, India, Pakistan and much of Asia and Latin America. This enhanced transparency encourages European universities to compete on a global scale. A major effort should be made to implement the core Bologna reforms in all EU countries: comparable qualifications (short cycle, Bachelor, Master, Doctorate); flexible, modernized curricula at all levels which correspond to
5 Higher Education Reform and the Renewed Lisbon Strategy
89
the needs of the labour market; and trustworthy quality assurance systems. This requires targeted incentives from the national authorities responsible in order to ensure proper take-up of the reforms rather than mere superficial compliance with the standards. Curricula in specific disciplines or professions should be renovated, drawing on comparisons and best practice at European level. There is already some evidence that the Bologna process seems to act as a stamp of quality in the sense that programmes that have been restructured according to the Bologna guidelines have attracted more students, especially if they were the leader in their country (Cardosa et. al. 2006). 5.4.2.3 Mobility of students
The Bologna reforms in themselves will not create the conditions for increased intra-university mobility of students. Students, particulary bachelor students seem to have a strong preference for selecting higher education services in close geographic proximity to home (Kelchtermans and Verboven 2006). Students may thus need to be stimulated to go abroad, eg by mobility subsidies. In addition, national grants/loans should be fully portable within the EU. The EU already has about twenty years of experience with the ERASMUS programme for promoting students to spend three to twelve month a year at another university in the EU. Since the inception of this programme, more than 1.2 million students have participated in such exchange visits while remaining enrolled in their university at home (i.e., ‘credit’ mobility). Most of them went to the UK and Ireland (probably to do with the English language and the efficient course structure), but Spain, Germany and Italy are becoming increasingly popular. The number of students going abroad to study full-time at a foreign university (‘diploma’ mobility) is much smaller. Temporary exchange students from the old Member States mainly go for the cultural experience while those from the New Member States value academic quality and learning a foreign language. Also students who take a full-time course abroad are much more motivated by the quality of education (Thissen and Ederveen 2006). It therefore seems sensible to make more funds available especially for those students completing a full Bachelor, Master or Ph.D. degree at universities in other Member States, since these are the students that care most about the quality of higher education and may be the ones that spur competition among European universities. They are also the ones that are more likely to work and stay abroad after migration (Oosterbeek and Webbink 2006).
90
Frederick van der Ploeg and Reinhilde Veugelers
If fees are low and the host country puts up most of the cost of university education (as is the case in most of Continental Europe), there is a possibility of free riding and thus a case for action at the EU level. As mobility increases, it also increases the incentive of universities to raise tuition fees for foreign students. If these students come from the EU, this can only be done by raising fees for home students as well. 5.4.2.4 Mobility of researchers/teachers
The recent Directive on the recognition of professional qualifications10 has made it simpler and quicker to have qualifications for professional practice recognized across national borders. Procedures for academic recognition should also be reviewed to ensure quicker and more predictable outcomes (in particular, by publishing universities’ recognition policies). In addition, full portability of pension rights coupled with the removal of other obstacles to professional, international or inter-sectoral mobility is needed to foster staff and researcher mobility. But perhaps the most important endeavour to increase staff mobility is to increase the attractiveness of the EU higher education landscape.11 5.4.2.5 Opening up of national funding schemes
Beyond mobility of students and staff, free mobility of funds can also contribute to improving the internal market for higher education. If Member States would open up their funding schemes to other EU or nonEuropeans, this would give the opportunity to leverage the efficiency of their funding by drawing on capacities beyond geographic borders. It would enhance competition for funds and thus generate better quality. 5.4.3 Provide funding The Commission should urge national decision makers to close the funding deficit in their higher education. However, the mix of public and private funding and the mix of basic, competitive and output-related funding 10
11
European Parliament and Council of the European Union (2005), the Directive will be implemented from October 2007. In March 2005 the European Commission adopted a European Charter for researchers and a code of conduct for the recruitment of researchers that aims at increasing the attractiveness of research careers and improve mobility and working conditions of researchers across Europe (European Commission 2005d).
5 Higher Education Reform and the Renewed Lisbon Strategy
91
is likely to remain different between countries, reflecting the diversity of cultures, economies and university traditions within Europe. Compared to the funding from Member States, the funding available at the EU level for education and research at universities is negligible. Nevertheless, it can play a catalysing role for enhancing the use of national resources and stimulating the quality of higher education in the EU. The mechanisms within the Financial Perspectives 2007-2013 include not only the Funds for “Competitiveness for Growth & Employment” programmes (the 7th EU Framework Programme for R&D, Lifelong Learning Programmes & Erasmus, Competitiveness and Innovation Programme), but also the Structural & Cohesion Funds. And also the European Investment Bank can provide an important financial impetus for higher education in Europe. 5.4.3.1 Structural funds
Structural and rural development funds offer possibilities to stimulate the modernization of higher education via sectoral measures. These possibilities should be fully developed, since they represent the bulk of funding available at EU level. 12 The Structural Funds can provide funding for the improvement of universities’ facilities and resources, the fostering of partnerships between the academic and business communities and the support of research and innovation relevant to regional or Member State economic development objectives. The Structural Funds’ system of decentralized management enables regional specificities to be taken into account. Member States, regional authorities and universities should take full advantage of these opportunities to improve synergies between education, research and innovation, particularly in the EU's less economically developed Member States and regions. 5.4.3.2 Funds for "Competitiveness for Growth & Employment"
A considerable part of overall student mobility within Europe is supported through Community programmes such as Erasmus, which has funded more than one million students since its inception in 1987/88. The next phase of 12
While the 7th FP, CIP and LLL& Erasmus represent 8.5% of the total EU budget 2007-2011, the Structural Funds represent 35%. The Framework Programme Budget (in total 48081 million Euro for 2007-2011) is split into Cooperation (64%), Ideas (incl ERC (15%), People (Marie Curie Programme (9%), Capacities (Research infrastructures, SMEs, Regions of Knowledge, Research Potential, Science in society, International Cooperation), 8% and Joint Research Center 3%.
92
Frederick van der Ploeg and Reinhilde Veugelers
the scheme (2007-2011) has the ambitious target of three million students to have studied abroad with an Erasmus grant by 2011. This implies that 375,000 students will be participating in the final year of the programme. Funding students through direct scholarships, like Erasmus, should increase in importance. These scholarships should be extended to fund “diploma mobility” and to enable graduates to obtain Master and Ph.D. degrees in other Member States as well. The Marie Curie funds offered for international mobility of post-docs and other academics should be increased. But just as important is a drastically reduction in the bureaucratic nightmare that one has to go through to apply for these scholarships. The administration of this mobility funding should be targeted to individuals rather than to institutions granting such scholarships, so that students can vote with their feet. Similarly, more funding in the FP should be allocated to individual researchers to cover their salaries. This type of funding for researchers, when portable across institutions within the EU, will allow the researchers to vote with their feet and select the best institutions to pursue their research projects, thus instigating more EU wide competition by universities for research talent13. 5.4.3.3 European Investment Bank loans
Finally, higher education is also a priority sector for the European Investment Bank and further expansion of its provision of income-contingent loans is desirable. Organizing income-contingent loans for education at EU level could entail the added advantage of offering more risk-spreading opportunities and coordination on recovery ex post, when graduates are moving across EU and may be tempted not to pay back their loans. 5.4.4 Building capacity at EU scale and the use of flagships In a number of areas where critical mass needs to be built beyond the level of the individual Member State, the EU should support the building of EU wide capacity. A main priority should be European postgraduate/doctoral schools and networks of worldwide calibre in their dual function as the peak of higher education and the first career stage for researchers. The Commission should examine the possibility of providing more support to 13
The VICI scholarships awarded by the NWO are an example of a portable, individualized funding scheme within the Netherlands. Its ambition to select each time the brightest of scholars and allow them to go the best institutions within the Netherlands, is however hampered by the small scale of the Netherlands market.
5 Higher Education Reform and the Renewed Lisbon Strategy
93
such schools and their students/researchers, provided they meet the subsidiarity criterion. Specific support may be envisaged for joint or even ‘European’ doctorates and quality assurance or accreditation at the doctoral level. The Marie-Curie programme for career development and mobility of researchers or the European University Institute in Florence already are cases at this level. In basic research, the European Research Council, will provide an arena for selecting bottom-up research projects and researchers on the basis of EU-wide excellence. When its reputation for selecting quality is established, it may serve as a flagship on which national or regional selection bodies can anchor. The proposed European Institute of Technology could become another example of EU-wide scale building, with its ambition to develop in publicprivate partnership, a flagship in the knowledge triangle of education, research and innovation based on excellence, interdisciplinarity and networking between academic centres and between academia and business in targeted disciplines. 5.4.5 The EU as a facilitator of global cooperation The development of extensive cooperation, mobility and networks between European universities over the past decades has created the right conditions for broader internationalization. Most universities now have experience with multilateral consortia. This should be used to extend international cooperation beyond the EU dimension. By launching the Erasmus Mundus programme the EU has started to promote mobility with third countries. More structured international cooperation through bilateral/multilateral agreements with the EU’s neighbouring countries and worldwide, should be developed at EU level, supported by the necessary financial means. An important prerequisite for international ‘brain circulation’ is to simplify and accelerate legal and administrative procedures for the entry of non-EU students and researchers. Concerning admission and residence of third country researchers, the recently adopted ‘researchers’ visa’ package needs to transposed into national law as soon as possible.
5.5 Some concluding remarks Universities are key players in shaping the future of Europe. They are also key players in the successful transition to a knowledge-based economy and
94
Frederick van der Ploeg and Reinhilde Veugelers
society. However, this crucial sector of the economy and of society needs in-depth restructuring and modernization if Europe is not to lose out in the global competition in education, research and innovation. Implementing this necessary restructuring and modernization requires coordinated action from all parties involved: x Member States need to take the necessary measures with respect to universities, including aspects such as management, granting real autonomy and accountability to universities, funding mix and access to higher education. In particular, they should only subsidize courses whose social benefit exceed the private benefit and make much more resources available for fundamental research. Funding should be based less on inputs or outputs and more on academic excellence. They should also allow universities to set fees independently and to differentiate them by type of student and type of course. The Member States and/or the European Investment Bank should provide student with income-contingent loans and cover default out of general funds. x Universities, for their part, need to make strategic choices and conduct internal reforms to extend their funding base, attract the best students and faculty, enhance their areas of excellence and develop their competitive position. They should clearly state their mission and act accordingly. A wider differentiation of objectives should be allowed to arise, with institutions specializing in research or undergraduate, graduate or post-graduate education. In any case, they must aim to compete with the best institutions elsewhere in the world, if their objective is to pursue excellence. x The Commission can contribute through improving the internal market for higher education, promoting policy dialogue and mutual learning, through financial support to Member States and to universities in their modernization activities, to promote mobility of students and researchers. The Commission can also take the lead in developing and implementing a set of quality indicators for institutions in the EU according to a multitude of criteria. They may also engender transparency of the EU market for higher education and to take action to demolish the power of monopolistic universities if it is used to the detriment of students. x Last but not least, students should adopt a different mindset and choose the degree programme that best suits their needs. Clearly, this need not be the university closest to their family home and may well be a top university abroad. They also need to fund a greater part of their own education and thus be encouraged to demand the highest quality.
5 Higher Education Reform and the Renewed Lisbon Strategy
95
Perhaps, the most important driving force for modernizing higher education in Europe emerges from competition. Increased competition, combined with more mobility and further concentration of resources, will lead universities to offer a more open and challenging environment to the most talented students and researchers, thereby making them more attractive to Europeans and non-Europeans alike. European universities face huge opportunities from reforms.
References Aghion P (2006) A primer on innovation and growth. Bruegel Policy Brief, 6: 1–8 Alesina A, Spolaore E (2003) The size of nations, MIT Press, Cambridge, MA Blundell R, Dearde L, Meghir C, Sianesi B (1999) Human capital investment, the returns from education and training the individual, the firms and the economy. Fiscal Studies 20: 1–23 Cardoso AR, Portela M, Sá C, Alexandre F (2006) Demand for higher education programs: the impact of the Bologna process. Discussion Paper 2532, IZA, Bonn Council of the European Union (2005) Resolution of the council and of the representatives of the governments of the Member States, on mobilising the brainpower of Europe: enabling higher education to make its full contribution to the Lisbon Strategy. Official Journal of the European Union 2005/C 292/01 vol 48, November 24, 2005 De La Fuente A (2006) Education and economic growth: a quick review of the evidence and some policy guidelines. In: Globalisation Challenges for Europe, Report by the Secretariat of the Economic Council – Part I, Prime Minister’s Office Publication 18/2009: 195–212, Finland Dosi G, Llerena P, Sylos-Labini M (2006) The relationships between science, technologies and their industrial exploitation: An illustration through the myths and realities of the so-called ‘European Paradox’. Research Policy 35: 1450–1464 Ederveen S, Thissen L (2008) European coordination of higher education. In: Gelauff G, Grilo I, Lejour A (eds) Subsidiarity and Economic Reform in Europe: 113–128, Springer European Commission (2005a) Mobilising the brainpower of Europe: enabling universities to make their full contribution to the Lisbon Strategy. COM(2005) 152, SEC(2005) 518, April 20, 2005 European Commission (2005b) European higher education in a worldwide perspective, Annex to the Communication from the Commission ‘Mobilising the brainpower of Europe: enabling universities to make their full contribution to the Lisbon Strategy’ {COM(2005)152 final}, SEC(2005) 518 European Commission (2005c) Key figures 2005 on science, technology, innovation: Towards a European knowledge area. DG RTD, Brussels
96
Frederick van der Ploeg and Reinhilde Veugelers
European Commission (2005d) Commission recommendation on the European charter for researchers and on a code of conduct for the recruitment of researchers. Recommendation 2005/251/EC) of March 11, 2005. Official Journal of the European Union L 75/67, March 22, 2005 European Parliament and Council of the European Union (2005) On the recognition of professional qualifications, Directive 2005/36/EC of September 7, 2005. Official Journal of the European Union L 255, 48, September 30, 2005 Jacobs B (2002) An investigation of higher education reform: income contingent loans and graduate taxes in the Netherlands. CPB Discussion Paper 9, The Hague Jacobs B, Van der Ploeg F (2006) Guide to reform of higher education: A European perspective. Economic Policy 47: 535–592 Johnson DR, Rahman FT (2005) The role of economic factors, including the level of tuition, in individual university participation decisions in Canada. The Canadian Journal of Higher Education XXXV: 101–127 Kelchtemans S, Verboven F (2006) Participation and schooling in a public system of higher education. Discussion Paper 5690, CEPR, London Martins PS, Walker I (2006) Student achievement and university classes: effects of attendance, size, peers and teachers. Discussion Paper 2490, IZA, Bonn. Oosterbeek H, Webbink D (2006) Assessing the returns to studying abroad. CPB Discussion Paper 64, OECD (2006) Education at a Glance, Paris Rothschild M, White LJ (1995) The analytics of pricing higher education and other services in which the customers are inputs. Journal of Political Economy 103: 573–586 Sacerdote B (2001) Peer effects with random assignment: results from Dartmouth roommates. Quarterly Journal of Economics 116: 681–704 The Economist (2006) The battle for brainpower, October 7 Times Higher Education Supplement (2006) World University Rankings, October 6 Williams GC, Zimmerman DJ (2003) Peer effects in higher education. Working Paper 9501, NBER, Cambridge, MA
6 Higher Education, Mobility and the Subsidiarity Principle
Marcel Gérard
6.1 Introduction Two issues related to higher education in Europe are addressed in this chapter; one relates to the mobility of students especially the Bologna process, the other to that of researchers, with a special focus on the new starting grants of the European Research Council. In both cases subsidiarity is at stake. However, though in the Bologna framework no central or inter-jurisdictional cooperative institution is in charge of efficiently assigning the task of organizing the financing of mobile students, in the other issue competences have already been assigned to the European Commission, in a way which does not exclude the exercise of similar competences by national or sub-national levels of power. Regarding the Bologna process, we issue and discuss the proposition that, rather than subsidising the schools and universities located in its territory, each government provides the students, whose it is the government of the country of origin, with vouchers. These vouchers can be used either at home or abroad to enrol for credits in schools whose quality has been certified. Providing the students with vouchers may mean giving them vouchers for free, selling them or lending them the vouchers. Regarding the mobility of researchers, we discuss the consistency of the assignment of similar competences to various levels of governments. Our discussion suggests that, if national or sub-national governments have to be permitted to compete for attracting researchers on their territory, it makes sense to also assign the twofold mission of research and regional
98
Marcel Gérard
policy to a central agency, provided that its strong commitment toward the poorer region is well established. Furthermore, the two levels of government need to decide simultaneously on their respective actions since that scenario dominates other settings where the three players interact. The mobility of students is addressed in Section 6.2 and that of researchers in Section 6.3. To both topics we attempt to apply the subsidiarity test suggested by Ederveen et al. (2008). Section 4 concludes the chapter.
6.2 Mobile students: from the Production to the Origin Principle The Bologna process was launched in that Italian city on June 19, 1999 when the representatives of the Ministers of Higher Education of 21 European countries or sub-national jurisdictions signed a common declaration, which intended to achieve the following objectives within the first decade of the new millennium: x adoption of a system of easily readable and comparable degrees, x adoption of a system essentially based on two main cycles, undergraduate and graduate, x establishment of a system of credits - such as in the ECTS system,1 x promotion of mobility by overcoming obstacles to the effective exercise of free movement, x promotion of European cooperation in quality, x promotion of the necessary European dimensions in higher education. Already today students go abroad to get a degree, and that category is expected to increase in the future since the Bologna process should encourage students to get a first degree at home and a more advanced one abroad.2 By extension, some students will interrupt their studies at home, go abroad for either a term or a full year, enrolling there in a local university, in order to get a number of credits that they will further validate on a program of courses in their original institution, where they will enrol again when returning home. We call all those students, Bologna students. CurECTS is an acronym for European Credit Transfer System; those credits measure time spent by students and 60 credits are more or less equivalent to a one-yearfull-time study. 2 On the potential benefits of the Bologna process, see a.o. Jacobs and Van der Ploeg (2006). 1
6 Higher Education, Mobility and the Subsidiarity Principle
99
rently, the country which hosts the students, producing higher education, supports the cost of that production, except if it is permitted to charge them a tuition fee. This is what we call the Production Principle, a principle that we question in this chapter. Two specific features greatly facilitate the raise of the number of Bologna students. First, in most EU countries the tuition fee charged to students, if any, is low and in no case it covers the true cost of studies. Moreover charging a different tuition fee to students resident of the country and to students non-resident of the country violates the non-discrimination principle at the root of the EU process (Del Rey 2001). This is a key difference between the EU and the US where a State University usually charges a different tuition fee to residents and non-residents of the State. UK differs from the standard EU system by charging higher tuition fees to non-EU students; accordingly UK universities are rationally more interested in attracting non EU students. But, in most cases, because of the low tuition fees, young Europeans interested in getting a multicultural skill (Mechtenberg and Strausz 2006) are usually not prevented to do it by a cost barrier – beyond the costs of transport and accommodation. A second feature is that mutual recognition of diplomas awarded in Europe implies that students from one Member State may get a degree abroad and then return to their country of origin and practice their degree in that country: for some students, getting their degree abroad is a way to escape a numerus clausus at home.3 Although it greatly facilitates studies abroad, the Bologna process remains silent about the way those studies will be financed. In particular nothing has been decided as to which country will be in charge of organising their financing: the country where the studies take place or the country of origin of the student? In the former case, we say that the Production 3
An example of that is provided by the following observation: during the academic year 2005–2006, in some classrooms of schools of higher education of the French speaking part of Belgium a large majority of students were coming from France; most were students who failed entrance competition in their own country. They were expected to return home after completion of their degree and some even return home during their period of studies for field training. Those students are not expected to contribute to Belgian GDP in the future so that they are actually a cost for Belgium, while representing a typical free riding, or free lunch, opportunity for France. Similar features may be observed between Germany and Austria. In quantitative terms, the excess of EU incoming students over national outgoing students in higher education amounts to 4.69 per cent of the corresponding overall number of students in Belgium and to 4.42 per cent in Austria, against 0.39 and 1.82 in France and Germany respectively (Gérard and Vandenberghe 2007, based on Oecd and Unesco figures).
100
Marcel Gérard
Principle applies; in the latter case, that the Origin Principle is at work. We do not mention the country where the graduate will work after completing his degree: due to cultural particularities still at work in Europe, that country is most likely to be the same as the country of origin. In this section, we show under which conditions it can be recommended to move to the Origin Principle and to set up a network of bilateral treaties or a multilateral arrangement, in some sense similar to what exists for taxation, social security or health expenditures, which makes the country of origin of the student responsible for organising the financing of his studies abroad – see also Gérard (2006a,b, 2007). Notice that being responsible for organising the financing, does not mean that the country has to use public funds for that purpose, it may require the student to finance part or all of the cost by himself, possibly through a loan. However the state of origin is responsible vis-à-vis the state of production. In this topic subsidiarity is at stake and the test of Ederveen et al. (2008) applies. The answer to their first question on the existence of cross border externalities is positive. In a context where most universities are publicly funded and higher education is deemed to have a beneficial effect on the welfare of the population of the country where the graduates work, students getting their degree abroad and returning back home afterwards generate a welfare gain, a positive externality, for their native country. But they also generate a cost, or a loss, for the country where that additional human capital has been produced, unless they fully pay for the costs of that additional education. In that latter case education is an export product for the foreign country, just as tourism or other services.4 To the second question of those authors, on the feasibility of credible cooperation, the answer is also positive. In the case of Bologna students we will describe and discuss later in this section an institutional arrangement, which may take the form of a network of interjurisdictional agreements, e.g. a network of bilateral treaties, internalising the externalities depicted above. Then, the role of the EU could be to make interjurisdictional cooperation easier by providing a framework arrangement or a model treaty, like the OECD does for international taxation. Alternatively an agency could be set up at the level of the OECD or that of the Bologna area to help signing such agreements. Or the OECD itself could be in charge of that task. 4
The first step of the test also asks the question of the presence of economies of scales as an argument to justify centralization; that issue is beyond the scope of this chapter but Ederveen and Thissen (2008) give a clear negative answer to that question showing that large countries do not have better universities than small country as long as GDP per capita is comparable.
6 Higher Education, Mobility and the Subsidiarity Principle
101
Students who are usual residents of a given EU Member State may enrol in a school of higher education in another Member State and get credits in that country, that they can use to obtain a degree either abroad or at home. Since EU foreign students cannot be charged a higher tuition fee than resident students of the host country, studies abroad is an alternative for studies at home especially for those who want to escape a numerus clausus in their preferred field of study at home. The question then is to find a mechanism such that (1) there is no free lunch for the country of origin of the student, and (2) the sovereignty of the country of origin of the student in terms of numerus clausus in some fields of higher education is or remains enforced. Property (1) is required because some countries may be tempted to systematically send their students abroad in order to save costs and get benefits, being tempted to export students and (re)-import graduates. This is per se an inefficient device since the behaviour of one country causes costs to another jurisdiction. For a symmetric reason the country of production of studies will attempt to limit the number of foreign students and, consequently, the total number of students studying abroad will be too low, compared to the figures generated by a centralized mechanism. A more efficient institutional design could then be to centralize the decisions in matter of studies abroad. However, for the time being, it seems unnecessary to resort to such a solution, apparently not desired by EU citizens, as reflected by Euro-barometers (Cerniglia and Pagani, 2007). Indeed one can argue that centralization can be efficiently approximated by interjurisdictional arrangements and thus the solution discussed in this section seems to be in accordance with Oates (2005) view of the basic theory of fiscal federalism: it is an arguable point on the trade off between preference matching and internalization of external effects. 6.2.1 Proposition: the country of origin provides students with vouchers usable abroad in quality-certified schools Let us set forth the following proposition, which substitutes the current Production Principle with the Origin Principle; the latter has the following four characteristics. First, each government is responsible for financing the studies of its usual residents, the students whose it might be regarded as the government of the country of origin. This means that that government provides those students with vouchers allowing them to enrol for credits in schools of
102
Marcel Gérard
higher education either at home5 or abroad; the government may give the vouchers for free, or it may sell them, or it may propose contingent loans to the students: this is its own responsibility6. Similarly it is its own responsibility to determine the maximum money value of a given voucher. But in any case the government of the country of origin is responsible visà-vis the authorities of the host institution, being committed to pay the amount required from local students by the local institution, up to the value on the voucher. Furthermore, by providing a student with a voucher the government signals that this student is correctly prepared for enrolling in the institution. Thus the voucher is both a sign of payment and a signal of quality. Actually, moving to that system creates a substantial twofold change: on the one hand the responsibility of financing higher education is turned from the country where that production occurs to the country of origin of the students; and on the other hand, the voucher is deemed to be in line with the full cost of education. Actually, nothing will be changed for purely domestic students – for them we already have a system of implicit vouchers in countries where universities and other schools of higher education are financed per capita – although the system will be radically changed for mobile students. Second, in order to validate the higher education policy of its partner countries, e.g. their numerus clausus in some fields of studies – see below –, the government of the host country may commit not to allow foreign students to enrol in a school on its territory without such a voucher. This is a way to recognize and validate the signal sent by the origin government. Remember that the host government is also committed by EU treaties to not discriminate between students from different EU Member States. Third, the government of the country of origin is permitted, and wisely advised, to restrict the use of its vouchers to schools whose quality has been recognized or certified, either by the local government or by a recognized agency. In that respect the instrument designed in this section contributes to the promotion of the quality of the institutions of higher education, which is one of the goals of the Bologna process. Finally, the government of the country of origin may limit the number of vouchers issued in some fields of studies. If it intends to put such a numerus clausus in some fields of studies, it will not deliver vouchers in those fields beyond an amount that it has decided. Therefore it may organ5
6
Vouchers may be used either at home or abroad; however in this chapter we only focus on their use abroad. On the alternative means of financing higher education, see del Rey and Racionero (2005) and Van der Ploeg and Veugelers (2008).
6 Higher Education, Mobility and the Subsidiarity Principle
103
ize a competition or a lottery among prospective students. Compared to the present situation, nationals will no longer have an incentive to go abroad to get a degree if they fail to be admitted in a similar school at home. In other words the territory of the numerus clausus is expanded to the EU territory or even more, to the whole Bologna area. Let us take an example. The French government wants to limit the enrolment in schools for midwifes to M students per year. A term corresponds to 30 ECTS and eight terms or 240 ECTS are required to get that degree. Then the French government issues each year 2M new vouchers, with a value of 30 ECTS. How to select the students who receive those vouchers and how much those students pay for the vouchers is a purely French issue, definitely beyond our scope. A French student can then go to a Belgian school, exhibit the voucher and enrol for one or two terms. The Belgian school is compensated of the money value on the voucher up to the amount requested from similar Belgian students, provided it is labelled as a good quality school. As far as we are informed, some countries already experiment such a system, partially or totally, including the Swiss cantons for students from other cantons, and Norway or the Netherlands for its residents studying at home or abroad. 6.2.2 Discussion That solution rules out the temptation for a government to free ride its neighbour and simultaneously it enables it to expand the area of enforcement of its possible numerus clausus. However that solution is welfare improving only if most students return home after completing their studies abroad. Indeed a student who remains in the host country after completion of his studies, at the expenses of his country of origin, causes costs to that latter – in proportion to the part of the vouchers which has been publicly funded – but gains to the former – in terms of driven general improvement in human capital and in terms of tax revenue. Then another externality is created that increased mobility of skilled workers will inevitably produce. And increased mobility may imply that a series of jurisdictions, all those where the person will actually work during her or his professional career, will gain from his education in a given country. For sure we are not yet in that setting in today’s Europe. But it is a plausible scenario for the future. What to do then? One avenue is to request graduates to refund the part of the voucher that has been publicly funded, at least in proportion of the length of their career spent abroad (possibly
104
Marcel Gérard
through turning a grant onto a loan). That solution has been adopted by some governments for graduates who benefited from fellowships. But professional mobility means that a person going to a given country at the age of 30 may return home at the age of 45. Then, one can imagine that a rational government, uncertain as to the place where future graduates will work, will have little incentives to devote public funds to the higher education of people whose working area is likely to be larger than its own area of sovereignty. Then other avenues make sense. One may consist in assigning the charge of managing the financing of students either to the country of origin in a quite arbitrary way – arguing e.g. that eventually it is most likely that the graduate spends his professional career in that country – or to the European authorities; in that latter case, higher education becomes a centralized competence. Apart from that design, the financing of higher education is likely to shift from the public to the private sector.7 Students mobility is well documented in Ederveen and Thissen (2008) and King and Ruiz-Gelices (2003); its impact on international labour mobility later in life is well investigated in Parey and Waldinger (2007) for the case of German students – see also Oosterbeek and Webbink (2006). Finally a system based on the Origin Principle as depicted above is likely to stimulate the adoption of quality enhancing and cost minimizing behaviours by universities and other schools of higher education. Indeed the former is a necessary condition to remain on the list of schools where the vouchers can be used and the latter allows the schools either to fit into the value range of the vouchers or possibly to generate extra profits enabling them to finance additional research efforts and other activities increasing the reputation and quality.
7
Justman and Thisse (1997) points out the link between mobility and underprovision of publicly provided education. By contrast Stark, Helmenstein and Prskawetz (1997), Beine, Docquier and Rapoport (2001), and Stark and Wang (2002) also consider private investment in education. More recently, both forms of financing are taken into account by Poutvara (2004a,b). For Mechtenberg and Strausz, "the most stable result (…) is that although increasing mobility (...) will lead to higher private investment in education, public provision will decrease. The government will tend to free ride on the education system of other country". According to them, Buettner and Schwager (2004) produces a similar result while Kemnitz (2005) also considers the competition effect for governments providing education to mobile students.
6 Higher Education, Mobility and the Subsidiarity Principle
105
6.2.3 Conclusion on the financing of students To conclude, the solution suggested in this chapter – applying the Origin Principle through vouchers issued by the country of origin and used either at home or abroad but in schools whose quality has been certified – is, for today’s Europe, better than the current application of the Production Principle. Now indeed, governments have an incentive to free ride their neighbour and students who fail at national selection tests have an incentive to study abroad. However, in the future, if mobility increases, it might become necessary to transfer the competence for managing the financing of higher education to a level of government whose area of sovereignty fits that of mobility of students and graduates, or to give up public funding of higher education, totally or partially. From the point of view of university management the reform is likely to stimulate the adoption of quality enhancing and cost minimizing behaviours by universities and other schools of higher education.
6.3 Assigning responsibility for attracting mobile researchers The EU has the ambition to become one of the most research oriented areas in the world – that goal is part of the so-called Lisbon Agenda and it is stressed by the authors of the Sapir Report (Sapir et al. 2003) who mention that "The specific recommendations in this context are to increase government and EU spending in research and post-graduate education, to allocate research grants according to the highest scientific standards, to create an independent European Agency for Science and Research, and to encourage private-sector R&D via tax credits." The recent creation of the European Research Council, in short ERC, obeys that program, and one of its first realizations was to launch in early 2007 a program for young excellent faculty members, providing them with EU funds called starting grants. Any scholar in the world is entitled to benefit from such a grant provided that a University in the EU is ready to welcome her or him.8 Moreover 8
A message to the members of the European Economic Association mentioned: “Eligibility is not limited to any nationality and requires no mobility (nor does it prevent it, obviously). It does require that the principal investigator intends to work in a “host institution” (a university, research institute, etc.) in the EU or the Associated Countries. This means either: (i) being already at such an institution and planning to stay there; or (ii) planning to join (from the EU or else-
106
Marcel Gérard
those grants are portable across EU Member States so that mobility of researchers at stake in this section is both mobility between the rest of the world and the EU as a whole, and mobility within the European Union itself. We have those researchers in mind throughout this section. In addition, we identify regions’ best interest with that of their universities, and we assume that better research induces better higher education and improved human capital.9 Then regions try to attract mobile researchers in their territory, in order to pursue their own goals of economic and social development through the availability of a better scientific and technological environment and of an improved human capital, two incentives for mobile firms to locate in the region. Said otherwise, “talent has become the world’s most sought-after commodity” (Van der Ploeg and Veugelers (2008), quoting The Economist of 5 October 2006). The ERC starting grants mechanism does not primarily intend to stimulate mobility of researchers within Europe: its primary goal is to provide incentives to the worldwide scientific community in order to convince its members to conduct research activities in Europe. Therefore we start with a competence and associated budgetary means assigned to the EU centre. That makes sense from a subsidiarity perspective. Indeed European researchers benefiting from an ERC grant produce additional human capital and knowledge not only where they decide to locate, but presumably, also in neighbouring regions or even for the entire EU, so that they cause interjurisdictional externalities – and to the first question of the subsidiarity test, the answer is “yes”. Should the impact of the presence of a mobile researcher anywhere in the EU, be uniformly distributed among EU Member States, voluntary cooperation among Member States could be a credible device. But presumably the local impact is larger, e.g. through the local spin offs associated to the research centre; then, observing the energy spent by Member States to attract economic activity on their own territory, possibly at the expense of their neighbours – see more generally the abundant literature on interju-
9
where) such an institution. The ERC plans to award two types of grants: “Starting Independent Researcher Grants”, for which the principal investigator should be less than 9 years from the award of his or her PhD, and “Advanced Investigator Researcher Grants”, for which there is no academic age limit. It is important to note that the 2007 budget - given its limited size of around 300 million Euros – will be devoted solely to the Starting Independent Researcher Grants.” Mechtenberg and Strausz (2005) write that “the relation between mobility and human capital has for long been on the agenda of economic research”, mentioning contributions on the brain drain like Grubel and Scott (1966) and Bhagwati and Hamada (1974).
6 Higher Education, Mobility and the Subsidiarity Principle
107
ridictional competition (Wilson 1999) –, we have to say that voluntary cooperation is not a credible device and there is room for centralization. In that setting the EU centre is justified to act and to attract high potential researchers in the territory of the EU wherever they locate or agglomerate in poles of excellence. However, in view of the uneven geographical distribution of the impact of those centres, that competence of the EU conflicts with another competence assigned to that level of government, regional development policy, as well as with own competences of Member States in that field. All that may turn overall stimulation of research in Europe into stronger interjurisdictional competition for attracting researchers in particular locations. In other terms, research and regional policies have become complementary for the EU and the former has become an instrument of the latter for the Member States. In that context one can easily imagine that a central agency - the EU Commission - decides on a budgetary effort to attract researchers in the European Union and then that national or regional authorities engage into horizontal competition to have those researchers located in their own territory. As mentioned above, reality is more complex since the task of the EU Commission is not limited to the promotion of research without concern for its distribution among regions; it is also in charge of improving the welfare, and primarily employment, in the relatively poorer regions of the EU through structural and cohesion funds10 allocated to various programs including the financing of research. It turns out that universities across Europe receive public funds from various levels of government.11 All that explains our focus on the interactions between research and regional policies. In other terms, and in accordance with Ederveen et al. (2008), when the assignment is at the EU level, to what extent should implementation, monitoring and enforcement also be assigned to the EU or to the Member States? In line with that, the ERC grant system lead us to discuss the consistency of a system where similar competences are actually assigned to
Structural Funds and Cohesion Funds are allocated to relatively poor regions/countries and can be used to increase the quality of human capital (Köthenbürger 2002; Riou 2006). At national level a similar pattern is at work and part of the resources allocated by regional authorities may eventually come from the European funds mentioned above. 11 As an example 55 per cent of the resources of the University of Saint-Etienne in France come from the French national government, 36 per cent come from the Region and 9 per cent from more local authorities (Ahues 2005). 10
108
Marcel Gérard
various levels of power. That last question again raises the issue of the link between the two competences of research and regional development. To shed some light on that issue we have developed a non-cooperative game with three public players, two asymmetric regions and a central agency (Gérard et al. 2007). The two regions compete for attracting researchers and eventually mobile firms, interested in a better technological environment and improved human capital permitted by the proximity of research and related higher education. The central agency is in charge of both stimulating overall research and development of the poorer region. Each of the three public agents maximizes a social welfare function à la Laffont-Tirole (1993). That game allows us to compare a series of settings. First, a centralized scenario is examined, where only the centre conducts active policy through a general subsidy to researchers working in Europe and a specific one for those of them locating in the poorer region. That scenario is the most welfare increasing one, as long as the centre is strongly committed to the welfare of the poorer region. Indeed its superiority declines when the commitment of the centre toward the poorer region goes down. However, the regions or Member States are not prevented to interfere with the action of the centre, and therefore scenarios where the Member States are also active players deserve interest. Interestingly, the scenario where the three players move simultaneously welfare dominates that where the centre is inactive – pure decentralization –, again as long as the centre is strongly committed toward the poorer region. This is a strong argument in favour of the implication of the EU Commission in both research and regional policies; the reason is that, if the centre is strongly committed in favour of the poorer region, the joint budgetary effort of the centre and the poorer region together becomes larger, though that of the poorer region is smaller than under pure decentralization. Therefore, if national or sub-national governments are permitted to compete for attracting researchers on their territory, it makes sense to assign also the twofold mission of promoting research and development of poor regions, to a central agency like the European Commission, provided that its strong commitment toward the poorer regions is well established. Furthermore, that scenario dominates other settings where the three players interact, especially those where either the Member States or the centre move first, and thus becomes a dominant player. A scenario where the centre moves first, seems to be in line with usual practice in federations. However, an alternative scenario, where the centre – the EU Commission – is an agent of lower levels of governments – the European Council – makes sense for the European Union. One can show that, compared with simultaneous moves, in the first case, the effort of the centre
6 Higher Education, Mobility and the Subsidiarity Principle
109
will be too low, while in the alternative one the effort of the poor region will be too small.
6.4 Conclusion In the two issues investigated in this chapter, some people are mobile across the European Union. Students are mobile in the first one, researchers in the second one. However in the Bologna framework no central or inter-jurisdictional cooperative institution is in charge of efficiently assigning the task of organizing the financing of mobile students. It turns out that governments of the countries of origin of the students may be tempted to free ride their neighbours, exporting students and (re)-importing graduates, while simultaneously students prevented to study at home because they have failed at entrance examination can try to obtain their degree abroad. Then a proposition might be issued which is consistent with today’s Europe: rather than subsidising the schools of higher education or the universities located in its territory, each government provides the students who are its usual residents with vouchers. These vouchers can be used either at home or abroad to enrol for credits in schools whose quality has been certified. Providing the students with vouchers may mean giving them vouchers for free, selling them or lending them the vouchers. However, if, in the future, mobility of degree holders increases, it might become necessary to transfer the competence for managing the financing of higher education to a level of government whose area of sovereignty fits that of mobile students and graduates, or to give up to a large extend public funding of higher education. In the other issue – that of research and development policies when researchers are mobile and research deemed to be an instrument for attracting firms – competences have already been assigned to the European Commission, but without excluding the exercise of similar competences by national or sub-national authorities. Then the agenda is to discuss the consistency of the assignment of similar competences to various levels of government. The discussion sets forth that, if the national or sub-national governments are permitted to compete for attracting researchers in their territory, it makes sense to assign also the twofold mission of research policy and regional policy to a central agency like the European Commission, provided that its strong commitment toward the poorer region is well established.
110
Marcel Gérard
Furthermore, it is important that the two levels of government decide simultaneously on their respective actions. Indeed that scenario dominates other settings where the three players interact, especially those where either the Member States or the commission move first. In the former, the effort of the centre is too low; in the latter the effort of the poor region is too small.
References Ahues M (2005) Bologna: Learning the French case through an example. A paper presented at the Conference on Higher Education, Multijurisdictionality and Globalization, Mons, December 14–15, 2005 Beine M, Docquier F, Rapport H (2001) Brain drain and economic growth: Theory and evidence. Journal of Development Economics 64: 275–289 Bhagwati J, Hamada K (1974) The brain drain, international integration of markets for professionals and unemployment: A theoretical analysis. Journal of Development Economics 1: 19–42 Buettner T, Schwager R (2004) Regionale Verteilungseffekte der Hochschulfinanzierung und ihre Konsequenzen. In: Franz W, Ramser H, Stadler M Bildung, 33. Wirtschaftswissenschaftliches Seminar Ottobeuren, Tubingen Cerniglia F, Pagani L (2007) The European Union and the Member States: Which level of government should do what? An empirical analysis of Europeans’ preferences. CESifo Working Paper 2067 del Rey E (2001) Teaching versus research: A model of State University Competition. Journal of Urban Economics 49: 356–373 del Rey E, Racionero M (2005) Financing schemes for higher education. A paper presented at the Çonference on Higher Education, Multijurisdictionality and Globalization, Mons, December, 14–15, 2005 Ederveen S, Gelauff G, Pelkmans J (2008) Assessing subsidiarity. In: Gelauff G, Grilo I, Lejour A (eds) Subsidiarity and economic reform in Europe: 19–40, Springer Ederveen S, Thissen L (2008) European coordination of higher education. In: Gelauff G, Grilo I, Lejour A (eds) Subsidiarity and economic reform in Europe: 113–128, Springer Gérard M (2006a) Enseignement supérieur et mobilité, l'interpellation du passager clandestin. In Gérard M, Vandenberghe V (eds) L'enseignement supérieur après Bologne. Reflets et Perspectives de la vie économique 45: 83–89 Gérard M (2006b) Le financement des études supérieures transfrontalières. Revue d'Economie Politique 116: 789–796 Gérard M (2007) Financing Bologna: which country will pay for foreign students? Education Economics 15: 441–454 Gérard M, Gilson N, Ruiz F (2007) Higher education and firms: on the interaction between research and regional policies. Mimeo
6 Higher Education, Mobility and the Subsidiarity Principle
111
Gérard M, Vandenberghe V (2007) Financement de l’enseignement supérieur et mobilité du capital humain en Europe. In: Docquier F, Thys-Clément F (eds) Education et Forces de travail, rapport de la Commission 1 au 17ème Congrès des économistes belges de langue française, November Grubel H, Scott AD (1966) The international flow of human capital. American Economic Review 56: 268–274 Jacobs B, Van der Ploeg F (2006) Guide to reform of higher education: a European perspective. Economic Policy 21: 535–592 Justman M, Thisse J-F (1997) Implications of the mobility of skilled labour for local public funding of higher education. Economic Letters 55: 409–412 Kemnitz A (2005) Educational federalism and the quality effects of tuition fees. A paper presented at the Conference on Higher Education, Multijurisdictionality and Globalization, Mons, December: 14–15 King R, Ruiz-Gelices E (2003) International student migration and the European ‘Year Abroad’: Effects on European identity and subsequent migration behaviour. International Journal of Population Geography 9: 229–252 Köthenbürger M (2002) Tax competition and fiscal equalization. International Tax and Public Finance 9: 391–408 Laffont J-J, Tirole J (1993) A theory of incentives in procurement and regulation, MIT Press Mechtenberg L, Strausz R (2005) The Bologna Process: How student mobility affects multi-cultural skills and educational quality. A paper presented at the Conference on Higher Education, Multijurisdictionality and Globalization, Mons, December: 14–15 Oates W (2005) Toward A second-generation theory of fiscal federalism. International Tax and Public Finance 12: 349–373 Oosterbeek H, Webbink D (2006) Assessing the returns to studying abroad. CPB Discussion Paper 64 Parey M, Waldinger F (2007) Studying abroad and the effect on international labour market mobility – Evidence from the introduction of Erasmus. CESifo Venice Summer Institute, July Poutvaara P (2004a) Educating Europe: should public education be financed with graduate taxes or income-contingent loans? CESifo Economic Studies 50: 663–684 Poutvaara P (2004b) Public education in an integrated Europe: Studying to migrate and teaching to stay? CESifo Working Paper 1369 Riou S (2006) Transfer and tax competition in a system of hierarchical governments. Regional Science and Urban Economics 36: 249–269 Stark O, Helmenstein C, Prskawetz A (1997) A brain gain with a brain drain. Economic Letters 55: 227–234 Stark O, Wang Y (2002) Inducing human capital formation: migration as a substitute for subsidies. Journal of Public Economics 86: 29–46
112
Marcel Gérard
Van der Ploeg F, Veugelers R (2008) Higher education reform and the renewed Lisbon Strategy: role of Member States and the European Commission. In: Gelauff G, Grilo I, Lejour A (eds) Subsidiarity and economic reform in Europe: 65–96, Springer Wilson J (1999) Theories of tax competition. National Tax Journal 52: 269–304
7 European Coordination of Higher Education
Sjef Ederveen and Laura Thissen
7.1 Introduction Although higher education has always been regarded as a national matter, this relation is weakening due to increased student mobility. Moreover the Bologna Declaration introduces the dual Bachelor-Mastersystem, which uniformizes the structure of higher educational programs. This chapter deals with the question of whether recent developments in higher education might justify lifting coordination of higher education to a higher level, i.e. from a national to a European level. It is related to two other contributions in this book. In Chapter 6 Gérard (2008) discusses the issues of mobile students and mobile researchers and the implications for the assignment of competencies in these cases, based on a theoretical model. Van der Ploeg and Veugelers (2008) take a broader view and discuss the challenges and possibilities for reform of higher education in Europe as well as the subsidiarity perspective in Chapter 5. We follow Van der Ploeg and Veugelers in taking a broad view, but we focus more on carefully disentangling all possible arguments for European coordination of higher education and providing empirical evidence of their importance. In addition, we pay due addition to the developments and determinants of student mobility in Europe. European coordination exists in different forms, from top-down governance to voluntary cooperation between the EU Member States (the socalled method of open coordination). As far as higher education is concerned, European coordination can imply coordination of financial matters (funding, tuition fees), but could also include making educational pro-
114
Sjef Ederveen and Laura Thissen
grams within the EU more transparent and comparable in structure or quality, to uniformize admission criteria, and so on. The choice of the appropriate level of coordination in the European Union is based on the subsidiarity principle. The subsidiarity principle states that power may only be shifted to a higher level of coordination when the objectives can, “by reason of the scale or effects of the proposed action, be better achieved by the Community”. A subsidiarity test therefore starts by asking whether coordination is justified by the existence of economies of scale and/or externalities (see the contribution of Ederveen et al. in Chapter 2). Scale economies in the field of higher education may possibly lead to a higher educational quality level or a lower price. Externalities can be subdivided into human capital spillovers caused by mobile students, and human capital spillovers caused by subsequent labour mobility. Sections 7.3 and 7.4 provide an assessment of these arguments. The potential benefits will only materialize if students are internationally mobile. If mobility costs are prohibitive to students international competition for quality will not occur (Vanhaecht and Pauwels 2005). It is therefore essential to understand the motivations of students to study abroad. So before assessing the subsidiarity arguments and discussing its implications, we start this chapter by studying the mobility of students in more detail.
7.2 Student mobility: developments and determinants International student mobility is an increasingly important phenomenon. King and Ruiz-Gelices (2003) report that 1.6 million tertiary-level students all over the world were studying abroad in 1996. This is a rise of almost 20% compared with five years earlier. Half of these international students was studying in Europe, while a third had the European nationality. Stimulated by the EU-financed ERASMUS and Socrates programmes, international student mobility within Europe has increased rapidly over the past decades. In 1987/1988 only a little over three thousand students within the EEA1 went to another (candidate)EU/EEA-country for a limited period of time on an ERASMUS-scholarship. In 2003/2004, their number has risen to almost 136 thousand (European Commission 2005). Notwithstanding the huge increase, the number of mobile students has remained well below the European Commission’s target of 10%. King and Ruiz-Gelices (2003) mention questions of high costs, lack of motivation and difficult organiza1
EEA = European Economic Area, including Iceland, Norway, and Liechtenstein, as well as the EU-countries.
7 European Coordination of Higher Education
115
tion as possible reasons why student mobility has fallen short of expectations. Students may have different motives to study abroad. These reasons will depend on the length of time they spend abroad, which stage of their education they are in, and so on. In this section, we discuss why students go abroad to study. We distinguish two groups: students who go abroad for a limited period of time and who are often already enrolled in an educational program in their country of citizenship (credit mobility), and students who enrol in an educational program abroad (diploma mobility). The latter group will pay the tuition fees and obtain a diploma of the educational institute in the foreign country; the former group mostly will pay tuition and receive a diploma of their home institution. 7.2.1 Temporarily abroad: ERASMUS exchange The European Union introduced the ERASMUS program in 1987. Students enrolled in an educational program in the EEA may study in another member country for 3 to 12 months and receive a grant per month during that period. Currently, almost 2200 higher education institutions participate in ERASMUS. Since the creation of ERASMUS, 1.2 million students have studied abroad under this program (European Commission 2005). Figure 7.1 depicts the motivations of ERASMUS students to study abroad, based on over 4600 surveyed students. All students, except those from the new Member States (NMS), rate ‘Cultural experience’ highest. Students from the NMS aim at improving their language skills and find academic quality to be important. Academic quality does not seem to be an issue for other students in their choice to study abroad, and the availability of specific subjects is discarded by all students. These findings are consistent with other survey results. From their study of groups of University of Sussex students who had spent a year abroad (YA) as part of their Sussex degree, King and Ruiz-Gelices (2003, p 237) conclude: “Summing up it seems, both from our survey results and those of others, that students and graduates see the YA retrospectively (and to a large extent prospectively) mainly in linguistic and cultural terms rather than in terms of its academic value.”
116
Sjef Ederveen and Laura Thissen
5
4
3
2
1 academic quality
NMS
subjects not available Mediterranean
career plans
improve language
Scandinavian
cultural experience
Continental
change of environment Anglo-Saxon
Fig. 7.1. Motivation of ERASMUS students to go abroad Source: Integrated Reporting for International Students (IRIS)2.
7.2.2 Permanently abroad: VISIE scholarship Permanently studying abroad is not nearly as popular as temporarily studying abroad. We look at a small sample (126) of surveyed Dutch students who enrolled in a bachelor program abroad within a year after they graduated from secondary school. They all applied for a VISIE-scholarship, a grant of approximately EUR 300 per month during the entire program, issued by the Dutch government during 1998-2002.
2
Based on surveying 4641 ERASMUS-students. The survey question read: ‘What was the importance of the following factors in your decision to study abroad?’ Scores from 1 (not important) to 5 (very important). Countries of origin are subdivided in NMS = New Member States, excluding Bulgaria and Romania; Mediterranean; Scandinavian countries; Continental (Belgium, the Netherlands, Germany, Austria, and France); Anglo-Saxon (Ireland and the UK).
7 European Coordination of Higher Education other reason 18%
117
subject not available in NL 21%
learn foreign language 7% higher quality of education 12%
international career 14% Dutch is not my mother tongue 11% cultural experience 17%
Fig. 7.2. Motivations of VISIE students to go abroad Source: Nuffic, 2000.3
It is clearly visible in Figure 7.2 that VISIE-students have different reasons to study abroad than ERASMUS-students. The reason mentioned most is that the educational subject was not available in The Netherlands. Over one fifth calls this the deciding factor to study abroad. This is in sharp contrast with the motives for studying temporarily abroad with the ERASMUS program. For both programs the cultural experience and career perspectives are important. However, more than 80% of the VISIE-students who started the study abroad because of the cultural experience involved did not finish it. For the students who continued studying after the first year, quality does seem to matter: more than 20% of these students state the higher quality of education abroad as their main motive. When asked about the decisive factors (more options could be marked) for choosing the specific country, the language spoken in the destination country is most often mentioned (by 70% of the VISIE-students), followed by the quality of education, which is mentioned by almost half of the students. About 30% of the students state culture, distance or specific subject/educational program as an important factor. The latter findings fit with the literature. Rose-Ackerman (1996) finds that students rate curriculum and ideological aspect of a school as important. Tuition or costs of living abroad don’t 3
In the survey, students can only choose one option when answering the question: ‘What was the decisive reason to study abroad?’ Sample consists of 126 surveys.
118
Sjef Ederveen and Laura Thissen
seem to be an important barrier for the VISIE students; 86% of the students say (s)he would have pursued study plans without the scholarship. 7.2.3 Regression analysis: exploring the determinants of student mobility We will conclude our inquiry into the motivations of students to study abroad by discussing the results of an econometric analysis. We will restrict our attention here to a qualitative discussion of the most important results. More details can be found in Thissen and Ederveen (2006). To estimate the determinants of student mobility we used a so-called gravity equation. In the gravity equation, migration flows are expected to depend negatively on distance (both physical and cultural) and positively on the size of the economy measured by population. Furthermore, GDP per capita and unemployment are also included in the specification to account for the economic environment in a specific country. We extended the analysis by including rough proxies for the difference in tuition fees and quality4 between the home and host country. Especially the latter is interesting, as it sheds some light on the issue of whether students do really base their choice on differences in quality. Our results are obtained for a sample of European students who enrolled in another European country in the years 1998-2002 (OECD Education Database). We start by discussing the effects of population and of the traditional macro-economic variables such as GDP per capita and the unemployment rate. The (log of) population is included as a measure for the size of the flow: the larger the population in either country, the larger the flow. This is confirmed in the estimation results. Concerning the economic indicators, we expect that a higher level of GDP per capita or a lower unemployment rate in the country of citizenship reduces the student flow, whereas a higher level of GDP per capita or a lower unemployment rate in the host country is expected to affect student flow positively. All four effects are indeed confirmed by our results. Next, we turn our attention to the effects of quality and tuition. As expected, higher educational quality in the host country relative to the country of citizenship increases the student flow in that direction significantly. A rise of relative quality by one percent increases the student flow by 0.5%.
4
Quality is proxied by the number of universities in the Top 500 normalized by the size of the population. See also Section 7.3.
7 European Coordination of Higher Education
119
A higher tuition in the host country compared to tuition in the country of citizenship is expected to decrease student flow in that direction. This is not what is observed; the regression results suggest that a difference in tuition has a positive effect on the student flow. Tuition is only a small part of the total expenses when studying abroad; housing costs, living expenses, but also scholarships or other funding are involved. This may distort the effect of the measure of tuition that we use (which is zero in many countries in our sample). Also, eligibility of foreign students to scholarships and study contributions may impose difficulties on the interpretation of this effect. Lastly, we consider the effects of distance, both physical and cultural. We have used four distance measures: linguistic, religious, cultural, and geographical distance. We observe that geographical distance has a large negative effect on student mobility. Student flows to a destination 1% further away will be more than 1% lower. This elasticity is much higher than the other effects, like GDP, unemployment and quality. Distance therefore seems to matter a lot for the choice of the destination country. This finding is consistent with results for other countries. For example, students in the US do not prefer being far away from home either (Winston 1999). The other distance measures also have a negative effect on student migration. It seems that students do not prefer a culture different from their own. Summarising, our measures of distance have a large negative effect on the international mobility of students. To sum up, the number of students studying abroad has increased rapidly over the past decades, especially those studying 3-12 months abroad. The cultural experience is often the major reason for studying abroad temporarily. However, for students who enrol as regular students in a foreign university, the motivation is different. For them, the availability of the educational subject is often the deciding motive. In addition, the quality of education is also an important factor to them. Quality of education also shows up significantly in a regression analysis of the determinants of student mobility. However, there is a strong negative effect of distance.
7.3 Economies of scale: does size matter? What matters most for the desirability of European coordination is whether there are significant benefits from the size of a country - economies of scale - and cross-border externalities. The next section deals with crossborder externalities. In this section we consider the relevance of economies of scale in higher education. Two possible advantages of larger popula-
120
Sjef Ederveen and Laura Thissen
tions are lower costs of providing public goods and higher productivity. Both aspects are potentially important for higher education. In the following, we compare the quality of education in countries of different sizes. As a proxy for quality we use the number of universities in the Top 500 of World Universities 2004 composed by the Institute of Higher Education of Shanghai Jiao Tong University normalized by the size of the population. Although this measure has a few limitations, which are discussed in Thissen and Ederveen (2006), it provides insight in the differences between countries. 'quality': number of universities in Top 500/population in millions 1.2 Sweden Switzerland
1
Finland Denmark Norway
0.8
Ireland
Netherlands UK
Belgium Austria
0.6
United States Germany European Union France Italy
0.4 Hungary
0.2
Spain
Greece Portugal/ Czech Republic
Poland
0 0
10
20
30
40
50
60
70 80 90 'scale': population in millions
100
Fig. 7.3. Does educational quality increase with population?
Figure 7.3 depicts the quality of education in a particular country relative to its population. The use of population as a proxy for size is common in comparable empirical assessments of scale effects (Rose 2006). If economies of scale are present, educational quality is expected to increase with population size. In Figure 7.3, it clearly doesn’t; countries in the top left of the figure, i.e. Switzerland and the Scandinavian countries, are seemingly able to provide high educational quality without having a large market size. What we do observe, is distinct ‘regional’ categories per quality range: in the top left of the figure, Switzerland and Scandinavia; with about the same population levels but somewhat lower quality we find Belgium, The Netherlands, Austria, and Ireland; and in the lower left corner are three East-European countries, and Spain, Portugal, and Greece, not too far from two other Southern-European countries: Italy and France. These regional
7 European Coordination of Higher Education
121
categories seem to bear a resemblance to GDP per capita-level: countries with a high level of GDP per capita seem to have a high educational level, whereas poorer countries have lower quality. In short, Figure 7.3 does not provide evidence for the existence of economies of scale at the country level. Rose (2006) reaches a similar conclusion. He explores possible scale effects for a wide range of indicators and concludes that small countries are not systematically different from large countries. With respect to education he considers the literacy rate, primary school completion and secondary school enrolment and concludes that they all fall with country size. There is therefore hardly any empirical evidence that larger countries provide better education. Economies of scale could also reveal themselves in lower costs in providing a certain quality level or in larger schools having a higher educational quality level. However, empirical evidence does not support these arguments. Expenditure per student on tertiary education for instance does not fall with country size (OECD, 2007). Also the quality of the university bears no relationship with the number of students enrolled (Thissen and Ederveen, 2006).
7.4 Cross-border externalities This section gives an overview of the second category of motives why European coordination of higher education may be theoretically justified: cross-border externalities. The externalities of higher education can be subdivided in human capital spillovers of (mobile) students or (mobile) employees. This section briefly describes the theoretical mechanisms, and provides empirical underpinning. 7.4.1 Human capital spillovers of student mobility The existence of externalities provides a rationale for investing public funds in the field of higher education. If the social returns of higher education exceed the private returns, then individuals will invest too little in higher education from a social welfare perspective and the government should step in. However, at current levels of government support, there seem to be no excess social returns to education (Canton et al. 2005). Jacobs and Van der Ploeg (2006, p 571) conclude that “the empirical evidence does not suggest persuasive externalities of human capital as the macro returns to education are (at most) equal to the micro returns”.
122
Sjef Ederveen and Laura Thissen
This refers to the returns to education on a national level. What matters for an assessment of the desirability of European coordination, is whether there are international spillovers associated with higher education. When studying abroad, knowledge from visiting students may be transferred to students in the host country5. Likewise, human capital accumulated in the host country can be transferred to the country of citizenship. As an example, Baláž and Williams (2004) evaluate the experiences of Slovakian students who had studied in the UK for at least three months. Afterwards, these students indicated that with respect to learning competences they mostly acquired new approaches to work and new ideas during their stay in the UK. Even though these spillovers of education are intuitively straightforward, empirical evidence is rather scarce. A modest empirical literature focuses on so-called peer effects in higher education, i.e. the effects that students’ characteristics and behaviour have on other students’ behaviour. Comparing the influence of room mates’ SAT-scores on a student’s SATscore, Winston and Zimmermann (2003) find some evidence that strong students tend to increase peers’ academic performance and weak students tend to reduce it. The implications of peer effects for the desirability of European coordination of higher education are not unambiguous. Attracting bright foreign students may enhance the performance of the domestic classmates, but at the expense of the former foreign classmates. Some recent evidence points at the possibility that students benefit from having less variation in quality levels (Ding and Lehrer 2007). That may provide a rationale for creating conditions for European top universities. 7.4.2 Free-riders People in countries with high tuition fees (e.g. the Netherlands) could move to a country for educational purposes without tuition fees (e.g. Germany). In this case, the issue of free-riders arises. In this example, the Netherlands would benefit from state-subsidized education in Germany. As a result, Germany would underinvest in education because part of its investment is not beneficial to its own citizens but to the Dutch. European 5
Studying abroad may also facilitate cultural encounters and have important effects on an individual’s attitude towards Europe. There is some evidence that these students are more likely to consider themselves at least partly European (King and Ruiz-Gelices 2003), but it is not clear whether this is the result of studying abroad. These effects are implicitly considered when discussing the effect of studying abroad on the likelihood of subsequent labour migration below.
7 European Coordination of Higher Education
123
coordination of higher education could prevent this. A solution to freeriders might be to directly finance students instead of financing educational institutes (see Gérard 2008). In practice, free-riders are only an issue when foreign students are eligible to enrolment in an educational program in another EU-country (and not, for example, restricted by language deficiency), and if they are eligible to the same compensation of tuition fees or scholarships as nationals. The recent verdict by the European Court of Justice in the Bidar case has eased the conditions on eligibility by ruling that EU students, residing legally in another EU country and being able to prove that they are ‘integrated sufficiently’ in that host country, cannot be refused access to social support: they have to be treated equally to the nationals of that country. If foreign students would be eligible to student loans in the country in which they study, problems could occur when they leave the country without repayment of the loan. In order to prevent this, countries should make arrangements, e.g. the debt could be transferred to the country of citizenship of the student when he leaves the host country without repaying his debt. Agreements to collect debts abroad already exist (CPB 2003). 7.4.3 Human capital spillovers through labour mobility Returns to education can also turn out in favour of the host country in case of skilled labour mobility. Although labour mobility within the European Union is known to be rather low, the group of people being most mobile are the highly educated (Antolin and Bover 1997). Not only are students who have studied abroad more likely to pursue a professional career in that country (Tremblay 2002), they may also be more likely to start their career in another foreign country. If the student is educated in his/her country of citizenship, and finds employment abroad, the host country will benefit. Consequently, the country of citizenship underinvests in education, since the returns to education leak away to other countries. Justman and Thisse (1997) show that a government that maximizes the utility of immobile residents indeed will reduce investment in public education when the educated become mobile. This provides a motive for European coordination. Data indeed suggest that student mobility may be a precursor for labour migration. A study carried out in the United States of a sample of 4200 temporary immigrants holding an H1B visa shows that some 23 per cent of them previously held a student visa (US Immigration and Naturalization Service 2000). King and Ruiz-Gelices (2003) find that students who studied a year abroad were roughly twice as likely to have migrated abroad since graduation compared to students who did not study abroad. A prob-
124
Sjef Ederveen and Laura Thissen
lem in these kind of comparisons is that they can not correct for possible intrinsic differences between both groups, like differences in their international orientation. One rare example of a study that tries to control for these differences is Oosterbeek and Webbink (2006). They estimate that students who studied abroad are 15 to 18 percentage points more likely to live abroad. If selection issues are taken into account, the regression results even suggest that 7 to 9 months of studying abroad increases the probability of living abroad by more than 30 percentage points.
7.5 Conclusions and implications The theoretical motives for European coordination of higher education can be grouped in economies of scale and cross-border externalities. Regarding the first group, we analyse the relationship between educational quality in a country and the population of that country. We find little evidence for economies of scale. What we do observe is distinct ‘regional’ categories per quality range: the Scandinavian countries and Switzerland have high educational quality and relatively small populations; the Southern European countries and new Member States have lower educational quality. Regarding the second theoretical rationale (cross-border externalities), we examine spillovers of mobile students and of mobile employees, and free-riders. Mobility is of vital importance in order to benefit from human capital spillovers. We find empirical evidence for externalities is rather scarce. There is little evidence for human capital spillovers, and almost no information on free-riders. However, there is support for the idea that higher education has indirect spillovers. A number of studies suggest that student mobility is a precursor for labour migration. As labour mobility within the EU is low, we should not expect too much from these indirect effects. These effects may however justify making degrees more comparable within the European Union in order to remove some of the barriers for the mobility of high skilled labour. A necessary condition for human capital spillovers and economies of scale is mobility of students. If students do not base their choice on educational quality, a single European market for higher education will not lead to more competition and more quality. We find that quality does matter for students who enrol as regular students. Distance seems to matter much for the choice of the students’ destination. This imposes limits to what possibly can be achieved with European coordination.
7 European Coordination of Higher Education
125
Concluding, we find little empirical support for the need for European coordination of higher education. Distance still matters more to students than quality, economies of scale are absent and the empirical underpinning of the importance of external effects is scarce. Still, there are potential benefits through indirect spillovers of human capital through labour mobility. Therefore this does not imply that uniformizing the structure of higher education in the EU, and making educational programs more transparent, is to no avail. Quality does matter for students, and student mobility is an ongoing process. This may be beneficial to labour mobility. What does this imply for the potential of the Bologna agreement? Jacobs and Van der Ploeg (2006) list four potential benefits of the introduction of the dual Bachelor-Mastersystem in all European Member States. They claim that it (i) encourages students to complete their studies more quickly, (ii) reduces the risk of choosing the wrong course of study, (iii) stimulates product variety and (iv) can strengthen competitive pressures and enhance transparency. Of these benefits, only the last has a clear international dimension and could be a possible justification for European coordination. Increased transparency of higher education would certainly be beneficial. However, we should not expect too much from these changes. First, the positive effects of the introduction of the dual Bachelor-Master system on student mobility are not guaranteed. The shorter study length may make it more difficult to spend half a year studying abroad compared to the present system. Second, real competition between European universities is still far away. As Jacobs and Van der Ploeg (2006, p.556) note, a prerequisite for the potential advantages to materialize is “a revolutionary change in mindset” as currently most students “go to their local university or college near to the home of their parents even if this is evidently a bad match with their talents or their demand for education.” The empirical evidence that we have presented in this chapter shows that this revolution seems nowhere near. Distance often proves a prohibitive obstacle to student mobility. In addition, simply enlarging the scale will probably not lead to more competition between universities and higher quality levels, as the incentives for competition are reduced by existing national institutions such as financing arrangements for universities, lack of selectivity and student compensation. To achieve higher quality through competition, national governments should first reconsider the structure of the system of higher education in their own countries.
126
Sjef Ederveen and Laura Thissen
References Antolin P, Bover O (1997) Regional migration in Spain: the effect of personal characteristics and of unemployment, wage and house price differentials Using pooled cross-sections. Oxford Bulletin of Economics and Statistics 59: 215–235 Baláž V, Williams AM (2004) Been there, done that’: International student migration and human capital transfers from the UK to Slovakia. Population, Space and Place 10: 217–237 Bassanini A, Ernst E (2002) Labour market institutions, product market regulation, and innovation: Cross-country evidence. OECD Economics Department Working Papers 316 Canton E, Minne B, Nieuwenhuis A, Smid B, Van der Steeg M (2005) Human capital, R&D and competition in macroeconomic analysis. CPB Document 91 Coe DT, Helpman E (1995) International R&D spillovers. European Economic Review 39: 859–887, May 1995 CPB (2003) Een sociaal leenstelsel: studiefinanciering volgens het boekje, Chapter 6. Macro Economische Verkenning 2004, CPB, The Hague Ding W, Lehrer S (2007) Do peers affect student achievement in China’s secondary schools? Review of Economics and Statistics 89: 300–312, May 2007. Ederveen S, Gelauff G, Pelkmans J (2008) Assessing subsidiarity. In: Gelauff G, Grilo I, Lejour A (eds) Subsidiarity and economic reform in Europe:19–40, Springer European Commission (2005) Actual Number of ERASMUS Students by Country of Home institution 1987/88-2003/04 Jacobs B, Van der Ploeg F (2006) Guide to reform of higher education: a European perspective. Economic Policy 21: 532–592 Justman M, Thisse J-F (1997) Implications of the mobility of skilled labour for local public funding of higher education. Economics Letters 55: 409–412 Gérard M (2008) Higher education, mobility and the subsidiarity principle. In: Gelauff G, Grilo I, Lejour A (eds) Subsidiarity and economic reform in Europe: 97–112, Springer King R, Ruiz-Gelices E (2003) International student migration and the European ‘Year Abroad’: Effects on European identity and subsequent migration behaviour. International Journal of Population Geography 9: 229–252 Nuffic (2000) Over afzien en genieten, inschatten en beslissen, Evaluatie van het VISIEbeurs programma OECD (2007) Education at a glance, Paris: OECD. Oosterbeek H, Webbink D (2006), Assessing the returns to studying abroad. CPB Discussion Paper 64 Rose A (2006) Size really doesn’t matter. In: Search of a national scale effect. Journal of the Japanese and International Economies 20: 482–507 Rose-Ackerman S (1996) Altruism, nonprofits, and economic theory. Journal of Economic Literature 34: 701–728
7 European Coordination of Higher Education
127
Thissen L, Ederveen S (2006) Higher education in Europe: time for coordination on a European level? CPB Discussion Paper 68 Tremblay K (2002) Student mobility between and towards OECD countries in 2001: a comparative analysis. International Mobility of the Highly Skilled: 39–67, OECD, Paris Van der Ploeg F , Veugelers R (2008) Higher education reform and the renewed Lisbon strategy: role of Member States and the European Commission. In: Gelauff G, Grilo I, Lejour A (eds) Subsidiarity and economic reform in Europe: 65–96, Springer Vanhaecht E, Pauwels W (2005) University competition: symmetric or asymmetric quality choices. Working paper / University of Antwerp; 2005: 21 Winston GC (1999) Subsidies, Hierarchy and Peers: The Akward Economics of Higher Education. Journal of Economic Perspectives 13: 13–36 Winston GC, Zimmermann DJ (2003) Peer effects in higher education. In: Hoxby C (ed) College Decisions: How Students Actually Make Them and How They Could, NBER, University of Chicago Press
8 On the Roles and Rationales of European STIPolicies
Rahel Falk, Werner Hölzl and Hannes Leo
8.1. Introduction EU enlargement has increased the economic diversity of the European Union substantially, also with respect to capacities in the fields of science, technology and innovation (STI). The increasing disparities have (re)kindled (old) debates on the role and scope of EU policy intervention. The conflicts of interest revolve mainly around the size of the budget, but also on the composition of funding and – within the single chapters of funding- around the content of programmes. In a nutshell, the net paying Member States aim to limit the redistributive role of EU funds and strive for more excellence in European STI performance. The net recipient countries in turn insist on Article 2 of the Treaty on the European Union which states that cohesion is one of the core objectives of the European Union. The innovation policy agenda of the EU has grown substantially in the last decades. Since the 1986 Single European Act STI policy is a policy priority at the European level.1 The Single European Act has established competencies for a common STI policy at the EU level, and gave the Commission a procedure for implementing multi-annual Framework Programmes (FPs). Since then STI policy in Europe has become a multi-level policy area (cf. Borras 2003). The ratification of the Maastricht Treaty in 1992 provided an even stronger base for STI policy at the European level. It gave the Commission competences to take initiatives to ensure coordination between Member States’ and EU activities in STI. These competen1
Although early pan-European research cooperation in science and technology were in place, e.g. COST (Co-operation in the field of Science and Technical Research) since the mid 1970s or ESPRIT (European Strategic Programme for Research and Development in Information Technology Programme) since the earlier 1980s.
130
Rahel Falk, Werner Hölzl and Hannes Leo
cies are seen also in the ambitious goal to establish a European Research Area (ERA). It is believed that less dispersion and overlap in national research programmes help to better exploit Europe’s scientific and technological potential. If successful, ERA gives the European Commission more autonomy to initiate projects and programmes that affect national STI players. This creates overlap with national competencies and policy making. Even if the largest part of STI policy is still pursued and funded at the national level and Member States pay close attention to retaining their individual decision-making powers (Banchoff, 2002), budgets and competencies have become increasingly dispersed and complex. However, any argument in favour of a coordinating role of the EU citing the complexity of European STI policies would put the cart in front of the horse. Such arguments give the EU a role but do not provide a genuine rationale for STI policies at the EU level. Against this background it seems important to discuss and (re)define the rationales and roles of European STI-policy. In which fields should the EU set the policy agenda, pursue active policy making or maintain coordinating tasks only and what aspects of STI-policy should remain within the exclusive authority of national policy makers?
8.2 Rationales for supranational STI policy While there is much literature on the rationales of government intervention to foster innovation, technological development and science, not that much is known from an economic perspective on the rationales for a supranational STI policy. 8.2.1 Rationales for government intervention The conventional rationales for government intervention in STI are derived from either market failure arguments or from system failure arguments. Market failure arguments are based on traditional neoclassical welfare economics and are related primarily to the fact that market failures due to imperfect appropriability, uncertainty and risk, asymmetric information and economies of scale lead to a suboptimal level of innovative activities. Government action in the form of innovation policy should then establish an optimal level of innovative investment. The argument of system failures takes into account that knowledge generation and technological change is essentially a learning process that takes place in networks of knowledge. Missing coordination between actors reduces the diffusion of knowledge
8 On the Roles and Rationales of European STI-Policies
131
and hence leads to a suboptimal innovation performance. System failures are related to missing bridges between organizations, related to dysfunctional institutions and provide a different set of rationales for public intervention in the area of innovation policy. 8.2.2 STI policy in multilayer policies The rationales for national government intervention apply of course also to supranational STI policy. However, there is also an additional set of requirements that supranational STI policy needs to fulfil. The main criterion to assign policy competences to supranational institutions is that the outcome of this assignment must have positive effects. The main rationale for supranational policies is that they enable to take on tasks collectively that would not have been tackled independently. The question of what and how much of public goods provision should be centralized in the European Union is subject to the subsidiarity principle. The precept is that public policy and its implementation should be assigned to the lowest level of government with the capacity to achieve the given objective. There is a close relationship between the subsidiarity principle and the economic theory of federalism, the main task of which is to define the assignment of allocative responsibilities to decentralized government levels.2 The criteria for the centralization or decentralization are related to (i) the heterogeneity of preferences, (ii) the existence of economies of scale, (iii) the internalization of external effects and (iv) policy learning: 1. A high diversity of regional preferences leads to a strong case in favour of decentralization (Oates 1972) as smaller units can better account for the preference of its constituencies. With STI policy this is primarily related to national and regional innovation policy. Here, the size of the jurisdiction is a decisive factor in evaluating heterogeneity and for realising economics of scale: The larger the jurisdiction, the more responsibilities should be assigned to the lower level. 2. The realization of economies of scale is a main decisive factor for the assignment to a central level of decision-making. The argument of “creating critical masses” is key to the EU’s self-conception of appropriate cases for supranational policy intervention. In case of physical research infrastructure, if only the cross-country pooling of 2
Ederveen et al. (2008) spells out the basics of fiscal federalism. See also Oates (1972, 1999), Breuss and Eller (2004) or Alesina et al. (2005).
132
Rahel Falk, Werner Hölzl and Hannes Leo
financial resources makes the research feasible, the subsidiarity principle is satisfied. 3. It is well known that STI has substantial interregional spillover effects introducing a coordination argument in favour of centralized decision-making. A first example for the internationalization of external effects is related to global public goods (GPGs). GPGs are public goods whose benefits reach across borders, generations and population groups (Kaul et al. 1999). Scientific research is an important GPG. Other examples are climate change, epidemic and commutable diseases. With global public bads the rationale for transnational policy intervention is straightforward: if the prevention of global public bads requires coordinated action among states to escape a prisoner's dilemma scenario. Scotchmer (2005, p 347) argues that the dearth of international efforts to coordinate public spending on STI on global public goods led to an inefficient shift from public sponsorship of research towards the private sector via intellectual property rights policies. Another example for externalities are cross border STI activities. On the one hand there may be a risk of wasteful locational policy competition when governments are willing to subsidize inflow of STI activities. Here, supranational policies can help to avoid wasteful rent seeking. On the other hand a problem may arise with regard to cross-border STI collaborations. If most countries restrict the use of subsidies and tax-credits to domestic STI, the leakage of public funds can be considered to be minimal (cf. Van der Horst et al. 2008). However, restricting the choice of collaboration partners to domestic partners may lead to suboptimal outcomes. 4. The fourth argument is the policy learning issue. It is closely related to the externality issue, and is close to the concept of interjurisdictional competition. Policy learning requires decentralization (or the ability by the central player to set locally different policies) and at the same time an arena where the experiences with different policies can be shared. Most policy learning initiatives in the different EU programmes are related to policy learning at the regional level or at the level of very specific innovation policies. In the field of STI policy the European Commission has a role in promoting a coordinated European STI policy agenda. The Commission tries to lead the European policy discussion by presenting communications, and supporting mutual (policy) learning by establishing and promoting networks (e.g. open method of coordination).
8 On the Roles and Rationales of European STI-Policies
133
The choice on what and how much STI policy should be centralized is essentially subject to a basic trade-off arising from economies of scale or externalities and the costs of harmonising policies in light of the heterogeneity of preferences. The arguments in favour of supranational European STI policies emphasize the internationalization of STI externalities through transnational spillovers, and the provision of transnational public goods. In fact, “policies where economies of scale and/or externalities are predominant should be allocated at the union level, or even at the world level” (Alesina et al. 2005: 276). This makes clear that the subsidiarity principle is vindicated for innovation policies with an exclusive national and regional orientation. However, there are more policy alternatives than complete centralization or decentralization. An intermediate solution is mutual cooperation between nation states. 8.2.3 The assignment of policy domains in theory Let us now summarize in a stylized taxonomy the assignment of STI policy domains to the EU level, the national or regional level and to mutual cooperation based on the dichotomy between heterogeneity of preferences and spillovers/scale effects. Table 8.1 shows that policy fields where scale effects and spillovers are low should be allocated to a national or a subnational level. This is the case for innovation and development activities that have an exclusive regional or national focus. In contrast, policy fields that have a low heterogeneity of preferences and high spillovers and/or scale effects should be allocated at a supranational level. A primary example for this is research that has high fixed costs, high uncertainty and provides results that have the potential to be beneficial to all Member States (global public good character). The need to assign this policy area to the EU is weakened if cooperation among the Member States is possible and credible. CERN is a example of mutual cooperation. Credible cooperation is weakened – even if the preferences are quite homogeneous - once the benefits of common actions are ‘non-excludable in consumption’, i.e. once some form of free-riding becomes a possibility. An example is research in vaccine for commutable diseases. In this case it is appropriate to assign the policy competences to the supranational level. Coordination projects that promote policy learning on the national/regional scale could also be classified in this area, as they provide platforms for policy learning. However, as policy learning platforms do not assign competences in STI policy, these initiatives are better located in the upper right quadrant, i.e. they could be implemented at different levels. In the upper right quadrant high spillover and potential scale effects com-
134
Rahel Falk, Werner Hölzl and Hannes Leo
pete with a high heterogeneity of preferences. These policy fields are politically highly contested. Policy areas that fall in this quadrant are inherently political as the preference heterogeneity is usually revealed through the political process (Pelkmans 2006). From a normative perspective the overall welfare effects of centralized and decentralized provision of the policy need to be assessed in detail using the compensation principle. In this instance the compensation principle would argue in favour of centralization if and only if the welfare gains are large enough to compensate the losers. For the issues that pass the compensation principle supranational governance in one or the other form is needed in order to reduce negative externalities or to take advantage from positive externalities and scale effects. Table 8.1. A simple taxonomy for the assignment of policy domains Heterogeneity of Preferences Low 1
High
High EU / mutual cooperation
EU / mutual cooperation / national / regional (according to the possibilities of mutual cooperation and the compensation principle)
e.g. promotion of scientific research on global public goods, research that has large scale economies (e.g. energy fusion), IPRs (except for language issues) 1 Low national/regional
e.g. policy coordination and learning at the European level, regulations (e.g. state aid) national/regional
e.g. regional innovation policies with e.g. sectoral/regional innovation poliexclusive regional focus cies, where specialization patterns are important 1 spillovers and scale effects.
8.3 STI policy and subsidiarity in practice In order to apply the ideas presented in Section 8.2 to EU STI policy we use the Communication "Putting knowledge into practice: A broad-based innovation strategy for the EU" EC (2006a). This choice is motivated by the fact that European STI policy cannot be reduced to the Framework Programmes for Research and technological development. Regulations to foster the Single Market play also an important role. This pragmatic limita-
8 On the Roles and Rationales of European STI-Policies
135
tion to issues mentioned in the latest Communication needs not be a limitation, as the Communication intents to set the ground for a broad based innovation strategy. In fact, the topics raised in the Communication are very broad (see Table 8.2). The Commission sees a European innovation policy as a broad horizontal collection of initiatives that cut across different policy fields which are under the auspices of different directorates (ministries) at the European (national) level. While all themes touched upon are important for STI policy in Europe, importance as such is not a rationale for shifting competences to the EU level as argued before. The main rationales for the delegation of competencies to the European level should be efficiency arguments related to externalities, scale economies and heterogeneous preferences. Table 8.2 summarizes our assessment of the rationales for assigning STI policy competences at the different layers of policy making. Let us begin with the first topic, higher education. It is well known that European higher education systems are quite different. The relevant economic research does not show that there is a large role for EU involvement in the designing or implementation of policies due to cross-border externalities or scale effects (e.g. Breuss and Eller 2004, Ederveen and Thissen 2008). However, as the politics of the reform of the higher education system are quite complex, the Commission has a role to play as facilitator of policy learning. Important is in this respect the regulation of the recognition of degrees and qualifications, and the EU-wide adoption of a system of comparable degrees. Such regulation is necessary to guarantee the free movement of students and graduates. If a voluntary agreement between Member States (such as the ongoing Bologna Process) fails this establishes a strong rationale for assigning regulatory competences to EU. Beside this the role of the EU in education and higher education policy should mainly be restricted to the issuing of non-binding recommendations and to the role of facilitator of policy learning.3
3
However, as long as higher education is largely funded by the state there may arise an incentive problem when European Single Market regulations are stretched in a way to prohibit discriminatory actions against students from other countries. The incentive to invest in higher education may be weakened in the target country when a large fraction of the students are foreigners and the incentives may also be weaker in the originating country as students can study abroad (see Gerard 2008).
136
Rahel Falk, Werner Hölzl and Hannes Leo
Table 8.2. Policy assignments in EU STI policy Responsibility for
Topic
designing implementing COM MS COM MS COM MS cross- scale policy Heteroborder exter- geneity externalities of preference nalities X X X X X X X X X
1 Higher Education 2 Public funding of Research Research with GPG X character and/or cross-border externalities Research with scale X effects European Institute X1,2 of technology 3 Mobility of reX searcher 4 Knowledge transfer 5 Cohesion policy X 6 Financing of innovation Financing of innoX vation: state aid guidelines 7 IPR policy X 8 Regulation (single X market) 9 Lead markets if GPG character X and/or cross-border externalities 10 Public Procurement X 1 2
Rationales for involvement of
X
X
X
X
X
X
X1,2 X
X
X X
X
X
X1,2 X
X
X (low)
X
X (low)
X1
X1,2 X
X
X X X
X
X X X
X
X
X
X X
X (low) X
X X
X
X
X X
X
X X
X X
X X
X
If the EIT takes the form of an institution for research fields with substantial scale effects and there is no voluntary cooperation among Member States. If the EIT takes the form of a funding agency.
The second topic is the funding of research. In contrast to education, research can have direct cross-border externalities and substantial scale economies. However, there are not many indications that public funding of
8 On the Roles and Rationales of European STI-Policies
137
research needs to be centralized unless the research topics have a global public good character or the research projects have substantial cross-border externalities. In addition, there is considerable evidence for international cross-border spillovers of public and private research (e.g. Keller 2004). This suggests clearly a role for the EU in financing R&D in the Single Market. The case for EU support for basic research is particularly strong, as science is itself a global public good. Here, the establishment of the European Research Council with the 7th framework programme is a first step in the right direction. The main instrument to finance R&D at the EU level are the framework programmes. The research priorities selected in the current 7th framework programme largely fulfil the criteria for direct involvement of the European Commission, because (i) the nine thematic priorities of the cooperation programme as well as the programmes ideas and people relating to the European Research Council and the mobility of researchers essentially reflect issues which are related to EU-wide public goods or are directed towards issues with large potential external effects for the EU (for more detail see e.g. Hölzl 2006 and Van der Horst et al. 2008). From a subsidiarity position the establishment of the proposed European Institute of Technology (EIT) is controversial. The proposal by the European Commission presents four possible options to establish the EIT (EC 2006b): On the one hand of the spectrum is the option to establish the EIT as a centralized institute with one location, at the other end of the spectrum is the option to establish the EIT as a funding agency. An assessment of the options requires information on the selection of research fields. For example, if the EIT is established following the preferred model of the European Commission, then from a subsidiarity perspective the EIT is justified only for research fields that have substantial EU-wide economies of scale. This assessment is based on the fact that there are not many rationales to centralize EU-wide capacity building, except when substantial scale economies and EU-wide externalities exist. The decision about where to locate the EIT (or its centres) leads to a high heterogeneity of preferences, as at least some of the benefits of the EIT have local character. Establishing the EIT as science and research funding agency receives a more favourable assessment, as there is widespread scepticism regarding the usefulness of centralizing R&D capacity building. Overall our discussion establishes a clear role for the European Union to be active in funding research and to implement mission oriented research programmes that are relevant for all Member States due to scale effects or cross-border externalities. Here another role of the European Commission comes into play. The European Commission can act as coordinator to increase R&D spending in the Member States to socially efficient levels, as
138
Rahel Falk, Werner Hölzl and Hannes Leo
spillover effects benefit all EU countries. However, the heterogeneity in preferences results in reluctant decision taking. This makes it very difficult to facilitate mutual cooperation among Member States and to monitor progress without having the possibility to create binding agreements or using disciplinary devices to inhibit free-riding behaviour. A failure to reach an agreement by voluntary cooperation strengthens the case for a centralization of STI-funding, but again only for research fields that have substantial cross border externalities and/or scale effects. For localized networkoriented (regional) innovation policy the case for centralization is very weak, and limited to initiatives that foster policy learning and voluntary cooperation among countries and regions. The mobility of researchers (topic 3) is already an important part of EU R&D policy and performed via the programme "People" in FP7. The aim is to foster the diffusion of knowledge within the EU. This is an integral ingredient for any strategy to create a European Research area. Externalities give a clear rationale for coordinated or even centralized support of researcher mobility. Knowledge transfer (topic 4) between research institutions and industry and society is an important element for any innovation policy strategy. However, the often localized character of these spillovers and the largely missing scale economies do not establish a rationale for assigning competences to the EU. The heterogeneity of preferences shows up in a particularly clear way in the case of Cohesion policy (topic 5). The European Commission aims to make structural interventions more targeted toward the strategic priorities of the Lisbon agenda. Expenditures for research, technology and innovation averaged 5.5 % of the total Structural Funds support from 2000 to 2006. Thus the R&D and innovation related expenditures in the structural funds are a relevant part of EU innovation policy. However, Cohesion policy is a special policy field, as its rationales are related to efficiency effects resulting from redistribution.4 The primary aim is to create a more equal playing field by fostering the convergence of less favoured Member States and regions within the EU. The arguments in favour of Cohesion policy are not entirely persuasive as redistribution can stand in the way of needed regional adjustment and have a structure-preserving instead of a structurechanging effect. The main rationale behind the expenditures for research and innovation in the Structural Funds is regional/national R&D capacity building. Thus, the decision-making about how to allocate the resources needs to be allocated at the regional/national level. The main challenge of 4
Casella (2005) suggests that such arguments are stronger when there are substantial barriers to labour mobility.
8 On the Roles and Rationales of European STI-Policies
139
research-oriented Cohesion policy is to assist to adapt local institutions and capacities in order to enhance workable innovation systems and focus on activities that reflect and reinforce the comparative advantages of regions and countries (Midelfart-Knarvik and Overman 2002). The evaluation of the Structural Funds suggests that science parks and technopolies that are not connected to existing regional networks and regional competitive strengths are not functional and become 'cathedrals in the desert' (von Tunzelmann and Nassehi, 2004). This clearly shows that attempts to centralize network-oriented innovation policies that have a clear regional/national orientation beyond the regional and national level is very unlikely to enhance efficiency. The funding of innovation (topic 6) is closely related to the funding of research and science (topic 2), knowledge transfer (4) and cohesion policy (topic 5). Therefore, the arguments from before are valid: substantial EUwide scale economies or EU-wide externalities need to be present in order to justify EU action. As example let us consider innovation policy for SMEs. While there are good arguments and rationales for an innovation policy for SMEs, it is also clear that the funding of innovation projects of SMEs or provision of public risk finance should be primarily a national issue. There is not much evidence that SMEs generate substantial externalities (Holtz-Eakin 2000) nor that scale effects are important (Van der Horst et al. 2008). However, there is a role for EU action with regard to indirect policies related to regulation affecting SME performance. In fact the Single market between EU Member States is one of the first principles upon which the EU has been built. With regard to regulation the goals of subsidiarity may conflict with the aim of creating new legislation at the EU level: The rationales to transfer regulatory authority to Brussels may not always be very strong. In our view here priority should be given to the completion of the Single Market. It should thus be a priority to delegate competencies to the European level in policy fields where a common legal framework reduces trade barriers, regulatory burdens and creates a single regulatory framework instead of 27 different legal frameworks. This applies to all issues connected with the establishment of the Single Market and in particular also to the policy domains of intellectual property rights (topic 7), competition policy, rules for state aid (topic 8) or public procurement (topic 10). Consider first IPR policy. European coordination is still missing for patents, while for the most other formal methods to protect intellectual property European regulations exist. In the case of patents scale economies clearly suggest that the establishment of a Community patent would be in line with the subsidiarity principle (cf. Van der Horst et al. 2008).
140
Rahel Falk, Werner Hölzl and Hannes Leo
The regulation of state aid is an important element of competition policy at the EU level. Public support of private research is an important element of innovation policy. Here, the new guidelines for State aid relax the constraints on the provision of public risk capital to SMEs. The drafting of the guidelines should be allocated to the EU level, so that the regulation holds for all Member States. But given the arguments regarding the innovation policy for SMEs the creation of risk capital instruments is a national task. Lead markets (topic 9) are a new concept in EU innovation policy and cover quite some space in the Communication “Putting knowledge into action”. The goal is to facilitate the creation of markets for innovative products and services in promising areas. Lead market strategies can be considered to be an elaborate version of “forward-looking” industrial policy. The integration of research funding, regulatory action and persuasion (policy learning) shows that this concept provides a fresh way of looking at mission-oriented and cluster policies that goes beyond funding research. However, as the old industrial policy also the concept of lead markets is a controversial topic. In the case that lead market strategies are oriented towards areas which have a global public good character and substantial crossborder externalities the previous discussions have shown that there are rationales to allocate policy competences to the EU level, if no voluntary agreement between Member States is likely. These rationales are not related to the concept of a lead markets as such. Thus, for lead markets that do not have a GPG character or substantial cross-border externalities no case for the allocation of policy making competences at the EU level can be established.
8.4 Conclusion For a number of different policy areas in STI policy it is possible to find rationales to assign competences to the EU level. However, these rationales arise in a strong form only for a small subset of policy issues. In the case of STI policy it is difficult to allocate competences exclusively to one level. This presents the danger that the responsibilities become blurred if too many levels of government are involved in a policy area. Thus the boundaries have to be carefully crafted in order to facilitate a coordinated interplay between the European Union and the Member States and regions. Our analysis suggests that the EU plays two quite distinct roles in the European STI policy arena. First, the Union should act as policy maker, regulator and/or program owner in policy domains where there are clear rationales for doing so. This requires an explicit allocation of policy com-
8 On the Roles and Rationales of European STI-Policies
141
petences. But this alone is likely not enough when it comes to managing and coordinating complex horizontal policy fields such as STI policy. Here the second role of the European Union as an institution with the "right" to fuel discussions (e.g. through the open method of coordination), to propose coordinated solutions and to monitor the Member States progress on the agreed goals comes to the fore. A primary example for this role is to encourage Member State to increase STI spending by committing themselves to the Lisbon targets. This role is political. It relates to the job to stimulate activities in areas where the Commission as "guardian of the treaties" has no clear mandate to act, due to missing rationales from a subsidiarity perspective.
References Alesina A, Angeloni I, Schuknecht L (2005) What does the European Union do? Public Choice 123: 275–319 Banchoff T (2002) Institutions, inertia and European Union research policy. Journal of Common Market Studies 40: 1–21 Borras S (2003) The Innovation Policy of the European Union. Cheltenham: Edward Elgar Breuss F, Eller M (2004) Efficiency and Federalism in the European Union: The optimal assignment of policy tasks to different levels of government. Constitutional Political Economy 15: 27–76 Casella A (2005) Redistribution policy: a European Model. Journal of Public Economics 89: 1305–1331 Ederveen S, Gelauff G, Pelkmans J (2008) Assessing Subsidiarity. In: In Gelauff G, Grilo I, Lejour A (eds) Subsidiarity and Economic Reform in Europe: 19– 40, Springer Ederveen S, Thissen L (2008) European Coordination of Higher Education. In Gelauff G, Grilo I, Lejour A (eds) Subsidiarity and Economic Reform in Europe: 113–128, Springer European Commission (2006a) Putting Knowledge into Action: A broad-based innovation strategy for the EU, Communication from the Commission, COM(2006) 502 final European Commission (2006b) Impact Assessment, Commission Staff Working Document accompanying the Proposal for a Regulation of the European Parliament and the Council on the European Institute of Technology, COM(2006) 604 final Gérard M (2008) Higher Education, Mobility and the Subsidiarity Principle. In: Gelauff G, Grilo I, Lejour A (eds) Subsidiarity and Economic Reform in Europe: 97–112, Springer Hölzl W (2006) Cohesion and Excellence: Two ways to a better Europe. TIP Study, Vienna: WIFO - Austrian Institute of Economic Research
142
Rahel Falk, Werner Hölzl and Hannes Leo
Holtz-Eakin D (2000) Public Policy toward entrepreneurship. Small Business Economics 15: 283–291 Keller W (2004) International Technology Diffusion. Journal of Economic Literature 42: 752–782 Kaul I, Grundberg I, Stern M (1999) Global public goods – international cooperation in the 21st century, New York: Oxford University Press Middlefart-Knarvik KH, Overman HG (2002) Delocation and European integration: is structural spending justified? Economic Policy 17: 321–360 Oates WE (1972) Fiscal Federalism, New York: Harcourt, Brace, Jovanovic Oates WE (1999) An essay on fiscal federalism. Journal of Economic Literature, 37: 1120–1149 Pelkmans J (2006) Testing for Subsidiarity. BEEP Briefing no 13, February 2006 Scotchmer S (2005) Innovation and Incentives, Cambridge MA: MIT Press Von Tunzelman N, Nassehi S (2004) Technology policy, European Union enlargement, and economic, social and political sustainability. Science and Public Policy 31: 475–483 Van der Horst A, Lejour A, Straathof B (2008) Why European Innovation Policy? In: Gelauff G, Grilo I, Lejour A (eds) Subsidiarity and Economic Reform in Europe: 143–156, Springer
9 Why European Innovation Policy?
Albert van der Horst, Arjan Lejour and Bas Straathof
9.1 Introduction
1
In March 2000 the Lisbon European Council defined the goal for the EU to become the ‘most competitive and dynamic knowledge-based economy in the world by 2010’. This has raised the question of how Member States can intensify investments in R&D, or more generally, of how they can become more innovative. One part of this question, which has been extensively investigated by Cornet et al. (2006), is which policies are likely to be successful in stimulating private R&D or in raising productivity. They showed for the Dutch economy that, among other policies, expansion of the provisions for starting innovating companies and expansion of public support for funds that supply small amounts of venture capital loans are likely to be successful. The second part of the question on how innovation can be stimulated is whether or not innovation policy should be coordinated at the European level, or should be left to the Member States. This is the central question of this chapter. The main reasons for coordinating innovation policy are to benefit from economies of scale and to internalize the effects of crossborder externalities, such as knowledge spillovers (Ederveen et al. 2008). Cross-border externalities reduce the incentives for national policy and thus lead to so-called policy externalities. European coordination can lead to a more efficient innovation policy as it can internalize these policy externalities. The key disadvantage of coordination is, however, that a Euro1
This chapter is based on Van der Horst et al. (2006). We would like to thank George Gelauff and Isabel Grilo for their comments.
144
Albert van der Horst, Arjan Lejour and Bas Straathof
pean policy is less able to deal with location specific factors. This chapter is an attempt to weigh the pros and cons of European coordination in innovation policy. In other words: is European innovation policy consistent with the subsidiarity principle, which states that a policy should only be adopted by the European Union if Member States themselves cannot achieve equivalent or superior results? The chapter is structured as follows. Section 9.2 outlines how the principles of subsidiarity as formulated by Ederveen et al. (2008) apply to innovation policy. These principles are applied to two policy instruments. Section 9.3 focuses on public expenditure on R&D and Section 9.4 on intellectual property rights. In Section 9.5 we draw conclusions on the benefits of European coordination of both policy instruments.
9.2 Innovation policy and the principles of subsidiarity Should innovation policy be coordinated in the European Union or should it be left to the Member States? Ederveen et al. (2008) list several reasons for centralization, but also for decentralization, of policy at the European level. The main reasons for centralization are economies of scale and policy externalities. Economies of scale in innovation policy can arise if designing and executing policy involve substantial fixed cost. Policy externalities arise if a country’s innovation policy has unintended effects on other countries. Policy coordination can internalize these policy externalities. In fact, this internalization is a kind of scale economy as innovation policy at a larger scale is less vulnerable to leakages across the border. Among the reasons for country-specific innovation policy are differences between Member States in preferences regarding innovation and innovation policy and heterogeneity in innovation inputs. In general, centralization of policy might be more efficient than national policy because the fixed cost of public administration have to be incurred only once, instead of for every Member State. For example, if a policy needs to be implemented only at the European level, this will save the cost of implementation for each individual Member State. The same argument applies to the monitoring and enforcement of regulation. In particular, the selection and evaluation of research proposals require expertise that can be maintained more efficiently at a larger scale. The European Patent Office (EPO) is an example of how centralization can reduce the cost of maintaining expertise in a wide range of fields. Mulligan and Shleifer (2005) present evidence on economies of scale in public administration. For education and health, evidence of scale economies is reported by Dao (1995).
9 Why European Innovation Policy?
145
Economies of scale are not only present in public administration, but also in R&D itself. If more researchers work in the same field, this stimulates competition and specialization. In addition, a larger knowledge base is created, letting researchers benefit from each other’s discoveries. Cohn et al. (1989) present some evidence on economies of scale in public research and graduate teaching. Centralization of policy may benefit from these economies of scale. Decentralized education and research policies could at least act as a barrier for the international integration of national research communities. The presence of policy externalities provides a second rationale for European coordination of innovation policy. Policy externalities arise when national policy of a Member State has unintended consequences for another Member State (Ederveen et al. 2008). A national R&D subsidy, for example, can benefit research beyond the borders of the domestic economy. If a government ignores the favourable effects of its subsidy on other countries, the amount of subsidy is too small from a European perspective. A national policy might also have a negative effect on other countries. For example, a country might fail to protect the intellectual property of foreign firms, thereby facilitating imitation by domestic firms. In the context of innovation policy, policy externalities can arise if knowledge diffusion does not stop at the border and if foreign buyers benefit from domestic R&D. Without European coordination, Member States likely ignore the positive effects on foreign buyers in determining the size of the policy. In addition, governments will be less inclined to prevent domestic firms from infringing on the intellectual property of foreign firms. European cooperation on public R&D, public funding of R&D and the protection of intellectual property rights can prevent these undesirable effects. National policymakers face a free-rider problem when reducing market failures, in particular for innovations applied in high-trade sectors. Suppose a subsidy on R&D-expenditure is given to a firm in order to stimulate innovation. Part of the benefits of the subsidy will accrue to customers of the firm in other countries if the firm is not able to capture the entire added surplus. In an integrated European market, there could be an incentive for Member States to reduce or abolish subsidies to R&D while still profiting from the innovations subsidized by other countries. Centralization and cooperation can be used to internalize this policy externality. So far, we focussed on arguments in favour of centralising innovation policy at the European level. However, centralization also has its cost. Keeping innovation policy at the national level has three potential benefits: adaptation to local circumstances, learning from a diversity of experiences
146
Albert van der Horst, Arjan Lejour and Bas Straathof
and better incentives for policymakers through policy competition (Ederveen et al. 2008). In the context of innovation policy, local circumstances are important not just because preferences tend to differ from one country to another, but also because of differences in innovation systems (see Carlsson 2006; Foray 1995). As Falk et al. (2008) point out, the recent enlargements of the EU have led to a substantial increase in diversity. A second benefit of decentralization also has to do with maintaining diversity. Governments adopting different innovation policies can learn from each other about effective policies. While centralization would lead to a complete loss of diversity within the EU, a degree of diversity is still possible in case of multilateral cooperation. Greater accountability of policymakers is a third potential benefit of decentralization. Tiebout (1956) presents a theoretical model in which policy competition arises because voters can migrate from one region to another (see also Pelkmans 2006). Instead of voters, also multinational companies may trigger policy competition. Member States with a more effective innovation policy might attract more foreign investment, such that underperforming Member States will be under pressure of adjusting their innovation policy. The availability of benchmarks from other Member States provides an additional incentive for policymakers to improve.
9.3 Public expenditure on R&D Public expenditure on R&D can take two forms. Governments can choose to fund public institutes like universities to perform R&D or they can subcontract or subsidize private firms. These two types of public expenditure are not entirely similar when it comes to subsidiarity and are therefore treated separately. The section concludes with a discussion of the Commission’s Seventh Framework Programme – the single most important EU fund for research – which encompasses both public R&D and publicly funded private R&D. 9.3.1 Public R&D Public R&D is the most direct form of innovation policy that one can imagine. Both the funding and execution of R&D are performed by the government. Public R&D encompasses research and development by publicly owned enterprises in sectors like healthcare and defence. In addition, we include research at public universities (or by higher education in gen-
9 Why European Innovation Policy?
147
eral) in public R&D as both the expenditure and execution is governed by the government. Does the scale of public R&D affects its efficiency? The presence of fixed cost provides a source of scale economies. Some research projects are too large and risky to be undertaken by a single country. Multilateral cooperation then enables R&D projects that would otherwise not have been undertaken. A concrete example of a large European public R&D project is the Galileo satellite navigation system, which is a joint venture of the EC and the ESA with a budget of several billions of Euros. In addition, European coordination of public R&D can foster competition between researchers, induce specialization among them and reduce the risk of “reinventing the wheel”. Figure 9.1 shows a scatter plot of the relative size of public R&D on the vertical axis against the size of the economy measured in terms of GDP. The relative size of public R&D is measured as the share of public R&D expenditure in total government expenditure.2 The absolute size of GDP is measured in billions of euros purchasing power parity (ppp). The horizontal scale is logarithmic because the differences between countries increase with the level of expenditure. A regression-line has been added based on OLS-estimation. There appears to be a positive association between the size of the economy and the amount of public R&D. A country, say France, which is ten times larger (in terms of GDP) than another country, say Ireland, has on average a 1%-point larger share of public R&D in government spending. The exceptions to the rule that larger countries spend relatively more on R&D are also clearly visible from Figure 9.1. Despite their limited scale Finland, Sweden and Iceland spend relatively much on public R&D. Figure 9.1 suggests that large countries/economies spend relatively more money on public R&D which could be an indication for the existence of economies of scale. A tentative implication is that the European Union has some potential to exploit these possible scale economies in public R&D. Its share of public R&D in government expenditures is below the regression line, let alone the gap with the United States. The European
2
Alternatively, the share of public R&D in GDP can be used. This would reduce the vertical gap between the EU and the USA. The share of public R&D in government expenditure has been chosen in order to take into account heterogeneity in preferences for the size of government across countries. If the inhabitants of a country tend to have an appetite for a small government it is likely that a larger part of R&D expenditure will come from private sources. In particular, a lot of fundamental research in the USA is privately funded.
148
Albert van der Horst, Arjan Lejour and Bas Straathof
Union might be able to internalize a larger share of the return to R&D than each individual Member State.
Share of public R&D in gov. exp. (%)
6
ISL
5
FIN USA
4
DEU SWE
HVR
FRA
NOR
3
SVN LTU
ESP IRL
CYP
LVA LUX
ITA GBR
EU15 EU25
GRC BEL
HUN PRT CZE
BGR
1
NLD
DNK
EST
2
JPN
POL
SVK
MLT
ROU
0 1
10
100
1000
10000
GDP (bln eur ppp), log scale
Fig.9.1. Economies of scale in public R&D (2003) Source: Eurostat.3
The presence of cross-border externalities could lead to a suboptimal level of public R&D because Member States do not take the effects of their public R&D on other Member States into account. This is the second motivation for European coordination of public R&D expenditure. The main reason why public R&D benefits not only the home country, but other Member States as well, is cross border knowledge diffusion. In open economies it can be expected that the benefits of public R&D are more likely to ‘leak’ to other countries than in closed economies. Therefore governments of countries with an open economy might be less inclined to spend on R&D compared to the optimal level of R&D from the perspective of all countries together.
3
EU 15 and EU 25 have been omitted from regression.
9 Why European Innovation Policy?
149
5.0 FIN
4.5 Share of public R&D in gov. exp. (%)
USA JPN
4.0
DEU SWE
3.5
NOR EU25
EU15
FRA NLD
3.0
DNK
ITA
ESP
LTU
GBR IRL
2.5
BEL POL CZE
2.0 1.5
SVK CYP LUX
1.0
ROU
0.5 0.0 0
20
40
60
80
100
120
Domestic patents granted by USPTO with foreign ownership (% of total domestic patents)
Fig. 9.2. Public R&D is negatively related to the share of foreign-owned patents (2003) Source: Eurostat.4
Figure 9.2 shows a negative association between public R&D expenditure and openness5 as indicated by the proportion of patents owned by foreigners.6 Remarkable is the extremely high proportion of foreign ownership for the former communist countries. After the Iron Curtain was drawn, western multinationals en masse patented inventions which was not possible for them before that time. The negative association between public R&D expenditure and foreign ownership of patents suggests that governments are indeed sensitive to ‘leakage’ of benefits from public R&D abroad. The leakage is larger at the Member State level (also for larger countries) than for the EU as a whole, because the leakage between EU Member States remains within the EU. Hence, European centralization of public R&D expenditures will reduce sensitivity to leakage.
4 5
6
EU 15 and EU 25 have been omitted from regression. In this context openness refers the degree of internationalization of research, in particular patenting. In order to ensure international comparability, the proportion of foreign-owned patents is based on patents granted by the United States Patent and Trademark Office (USPTO).
150
Albert van der Horst, Arjan Lejour and Bas Straathof
Figure 9.2 suggests that leakage could indeed reduce expenditure on public R&D. The share of expenditure on public R&D in the EU is small compared to the US and Japan – possibly as a result of a free-rider problem in the EU. Assuming that the association between relative expenditure on public R&D and the share of foreign-owned patents reflects a causal relation, centralising public R&D policies at the European level could raise the European spending for public R&D to a share comparable to that of the United States or Japan because all three regions have a similar, modest share of patents from outside the region. Even though the possible scale economies and externalities provide arguments in favour of European coordination of public research, national policies might still be better at cooping with heterogeneity. All three arguments for heterogeneous national policies discussed in Section 9.2 apply to public R&D. First, at least part of public R&D tries to solve local problems, like research for dike construction in the Dutch province of Zeeland. This does not exclude, however, the possibility that this knowledge spills over to other countries, like to New Orleans in the United States or Venice in Italy. Still, it limits the scope for European coordination, as most other Member States are much less interested in dikes. Second, heterogeneity allows different governments to seek different answers to similar problems. This facilitates learning from each other. Third, diversity in public R&D may in the long run lead to the most efficient type of R&D. In particular, if multinationals are among the recipients of the public innovations, governments may be induced to seek the highest efficiency in public R&D. Public R&D differs from country to country; some spend a lot, others spend a bit. However, for the question of subsidiarity also the composition of public R&D matters. An indication of the variety in composition is provided by Figure 9.3, showing the socio-economic objectives of government expenditures on R&D in a selection of OECD countries for which these data are available. At first sight, the figure points at strong heterogeneity in the objectives of public R&D. Part of this heterogeneity can be explained by the industry structure of the countries. Public R&D is aimed at activities or sectors in which a country is ‘large’. In industrial countries like Germany and Japan, a large share of public R&D expenditure is focused on industrial production, whereas Denmark and Portugal spend higher fractions on research oriented towards agriculture.
9 Why European Innovation Policy?
151
Public R&D expenditure (GBAORD, % GDP)
1.60 1.40 1.20 1.00 0.80 0.60 0.40 0.20
exploration agriculture non-oriented
environment social structures defence
health space
N
IS L
FI
R G BR EU 15 D N K D EU SW E N LD U SA M LT FR A
N O
E JP N
C H
A
T
infrastructure industry other civil
IT
AU
T ES P
BE L
PR
SV N
C SV K
IR L
G R
LT U
R O
U LV A
0.00
energy university
Fig. 9.3. Heterogeneity in socio-economic objectives of public R&D expenditure (2000) Source: Eurostat.
Specialization of public R&D in domestically large sectors maximizes the domestic return and minimizes the leakage of knowledge to other countries. It reduces therefore the need for European coordination. In addition, the mere presence of heterogeneity reduces the scope of European involvement in public R&D. 9.3.2 Public funding of private R&D Besides funding public R&D, governments can also pay private firms to do (more) R&D. The size of publicly funded private R&D is quite limited in most European countries, ranging between less than 0.01% of government expenditures in Bulgaria and Cyprus to 0.6% in Great Britain, France and Sweden. Quite remarkable is the large share of public funding in the United States of more than 1% of the government budget, mainly attributable to innovation by private firms for the US Department of Defence. These numbers only reflect subsidies, not tax credits. Van der Horst et al. (2006) find some evidence of scale economies and external effects in the public funding of private R&D. The data show that larger economies provide relatively more support for private R&D. These economies of scale could be due to fixed costs in providing public support for private R&D, such as monitoring or establishing an office or govern-
152
Albert van der Horst, Arjan Lejour and Bas Straathof
ment agency for that purpose. Cross border externalities are significant in private R&D, but the likely implication that governments of open economies spend less on public funding is not supported by the available data. Van der Horst et al. (2006) do not have specific data on the objectives of Member States regarding the public finance of private R&D, but the data on the objectives of publicly funded R&D as well as the literature suggest that the heterogeneity is large. Weighting these arguments they arrive at a similar conclusion as for public R&D. As there seems to be some evidence for scale economies and external effects, EU involvement can be recommended in areas where national policy objectives match. In other cases, national governments can better take account of the particular characteristics of the country regarding the sector structure, the innovation system, and preferences. 9.3.3 Seventh Framework Programme (FP7) The Seventh Framework Programme (FP7) is an initiative of the European Commission run by DG Research under which various subsidies are granted for both public and private research. The budget of FP7 is 50.5 billion euros for the period 2007-2013. This amounts to an average yearly budget of 7.2 billion euros, which is substantial when compared to the 65 billion euros spent on public research by the Member States of the EU in 2003. FP7 consists of four programmes: Cooperation (32.4 billion), Ideas (7.5 billion), People (4.8 billion) and Capacities (4.1 billion). The main instruments of the four programmes are funding of public and private research. The theoretical arguments for policy centralization of public research are economies of scale in research and in public administration. Both arguments are reflected in the objectives of the majority of the FP7-programmes. Cooperation and People should foster the diffusion of knowledge within Europe, while Ideas can benefit from economies of scale in public administration by avoiding duplication and by a more efficient allocation of funds. Our explorative empirical analysis has shown that the proportion of public R&D in total government expenditure depends positively on GDP. The economies of scale are probably mainly caused by the internalization of policy externalities. Countries in which a high proportion of domestic patents is owned by foreign companies spend less on public R&D than countries in which this kind of ‘leakage’ is limited. Together, the theoretical and empirical analysis provide sufficient ground for European coordination of public research.
9 Why European Innovation Policy?
153
The second policy instrument available in FP7 is public funding of private R&D, where economies of scale are likely to occur, but where indications of policy externalities are lacking. The main purpose of this instrument in FP7 is to promote economic integration in innovative sectors, where economies of scale in research can indeed be exploited. Economies of scale and policy externalities are present in the public funding of R&D and tentatively support European innovation policy. However, the diversity of European economies is insufficiently incorporated in this tentative conclusion. It may motivate national responsibility of some parts of the FP7-programme like Capacities, which is aimed at research infrastructures, small- and medium-sized enterprises and backward regions.
9.4 Intellectual property rights When it comes to technological innovation, intellectual property rights are usually awarded through patent systems.7 A patent system can contribute to innovation in two ways. First, it increases the rewards to research by granting a temporary exclusive right to use the invention. Second, it stimulates the diffusion of knowledge as patents contain a complete and publicly accessible description of an invention. Not only does a patent system reduce duplication of research efforts in this way, it also provides inspiration for new inventions. Both functions can be performed better if the scale of the patent system increases. Scale matters for R&D-incentives as a patent enjoying protection in all Member States of the EU is worth more than a patent valid in only one Member State. The diffusion of technology is also stimulated by the size of the jurisdiction of a patent system. A patent system spanning several countries systematically cross-references patents granted in different countries, whereas this does not apply to patent systems restricted to single countries.8 From an administrative point of view a larger scale also might have drawbacks as each patent application has to be compared with a larger number of granted patents. In practise, of course, an increase in scale most likely leads to a more narrow and specialized search among earlier patents, thus keeping administrative cost within bounds.
7 8
For software, copyright protection is also relevant. Officially, European patent clerks have to search all patent databases in the world in order to establish newness. In practice, this is a time-consuming task – in particular because patents have to be translated.
154
Albert van der Horst, Arjan Lejour and Bas Straathof
Despite the obvious advantages of scale, the European Union still does not have a single patenting system, a system usually referred to as the “Community Patent”. There are probably four major reasons why such a Community Patent has not been established to date. First, a Community Patent will expand the average geographical coverage of a patent, thereby promoting the diffusion of knowledge via patent citations while simultaneously discouraging outright imitation of foreign inventions. Both consequences can trigger protectionist behaviour by national governments. A country with a strong knowledge base might fear an increase in knowledge spillovers abroad, while a country in which foreign inventions are being imitated on a large scale is not eager to see improvements in the IPRs of foreign companies. Both elements give rise to a policy externality. If a country makes it easier for foreign firms to patent their inventions, for example by reducing translation requirements, this has a positive effect on the competitiveness of foreign firms vis-à-vis domestic firms. As a result individual governments will not open up their patent systems. A Community Patent internalizes most of this policy externality as all Member States simultaneously reduce their barriers to foreign firms. A problem arises when the net benefit of cooperation is negative for some countries. In that case, agreement over the establishment of a Community Patent can be frustrated by a minority of Member States. Second, there is the problem of language. The cost of translating a patent are substantial – especially when (a part of) each patent should be translated in all languages of the European Union. Only allowing patent applications in English, German or French reduces translation cost, but at the same time would introduce barriers for countries in which one of these languages is not widely spoken or understood. Third, the Community Patent seems to be affected by a conflict of interest between SMEs and large companies. SMEs are in favour of a Community Patent if it would reduce the cost of filing and litigation. In particular, SMEs tend to be in favour of an English-only system. Larger companies tend to oppose the formation of a European Patent Court because they are concerned that such a new court will have insufficient experience in handling litigation procedures (European Commission 2006). These positions confirm the view that large companies benefit from maintaining the status quo, which, coincidentally, imposes barriers to multi-country patenting – especially for SMEs.
9 Why European Innovation Policy?
155
Fourth, a Community Patent is not in the interest of some of the patent offices and, especially, patent courts of individual Member States, because of a likely loss in employment. As cooperation of national offices is important for the establishment of a Community Patent, this hurdle is a difficult one to overcome.
9.5 Conclusions A European innovation policy is to be preferred over national policy if scale economies and externalities are large and if countries are not too different from each other. Some degree of European involvement is suitable for the policy instruments examined in this chapter – public R&D, public funding of private R&D and protection of intellectual property rights. For public R&D, European coordination could be beneficial because policy externalities seem to hamper spending on public R&D assuming that the empirical correlations presented before represent a causal relation. If the Member States of the European Union agree to increase expenditures on public R&D, all Member States could benefit. The European Commission should also be involved in public R&D projects where the scale really matters – such as the ITER-programme on nuclear fusion – and to basic science where international specialization is important. Extensive framework programmes could be helpful here. This does not imply that all public R&D should be performed at the European level. Given the differences in national socio-economic objectives of public R&D it seems reasonable that the Member States themselves determine their objectives to some extent. For public funding of private R&D Van der Horst et al. (2006) find some evidence of scale economies and external effects in the public funding of private R&D. The data show that larger economies provide relatively more support for private R&D. The economies of scale are a motivation for greater involvement at the European level in the funding of private R&D. However, one has to be aware that the diversity in type and amount of public funds for private R&D across Member States is large. This justifies a substantial role for national governments in handling this diversity. Greater European involvement could be fine, but priority should remain with national policy. From the viewpoint of subsidiarity, the Community Patent would be desirable, because of scale economies stemming from a better protection of IPRs and a wider diffusion of knowledge. A Community Patent internalizes the incentive of national governments to keep their patent systems relatively unattractive to foreign firms. Heterogeneity in languages pro-
156
Albert van der Horst, Arjan Lejour and Bas Straathof
vides a powerful political argument against the Community Patent – although this argument apparently is not supported by SMEs in most Member States. A more likely explanation for the absence of a Community Patent is protectionism by Member States.
References Carlsson B (2006) Internationalization of innovation systems: a survey of the literature. Research Policy 35: 56–67 Cohn E, Rhine SLW, Santos MC (1989) Institutions of higher education as multiproduct firms: economies of scale and scope. Review of Economics and Statistics 71: 284–290 Cornet M, Huizinga F, Minne B, Webbink B (2006) Kansrijk kennisbeleid. CPB Document 125 Dao MQ (1995) Determinants of government expenditures: new evidence from disaggregative data. Oxford Bulletin of Economics and Statistics 57: 67–76 Ederveen S, Gelauff G, Pelkmans J (2008) Assessing subsidiarity. In: Gelauff G, Grilo I, Lejour A (eds) Subsidiarity and economic reform in Europe: 19–40, Springer European Commission (2006) Future Patent Policy - Public Hearing 12 July 2006, http://ec.europa.eu/internal_market/indprop/patent/hearing_en.htm Falk R, Hölzl W, Leo H (2008) On the rationales and roles of European STIpolicy. In: Gelauff G, Grilo I, Lejour A (eds) Subsidiarity and economic reform in Europe: 129–142, Springer Foray D (1995) The economics of intellectual property right and innovation systems: the persistence of national practises versus the new global model of innovation. In: Hagedoorn J (ed) Technical change and the world economy: convergence and divergence in technology strategies: 109–133, Edward Elgar, Aldershot, England Mulligan CB, Shleifer A (2005) The extent of the market and the supply of regulation. Quarterly Journal of Economics 120: 1445–1473 Pelkmans J (2006) Testing for subsidiarity. BEEP Briefings 13, College of Europe, Bruges Tiebout CM (1956) A pure theory of local expenditures. Journal of Political Economy 64: 416–424 Van der Horst A, Lejour A, Straathof B (2006) Innovation policy: Europe or the Member States? CPB Document 132
10 Integrating Regulated Networks Markets in Europe
Jordi Gual
10.1 Introduction
1
The process of creating a single European Union (EU) market for goods and services has gone through several stages, from the early removal of tariff barriers to trade in merchandises, to an increasingly complex period of integration in services provision. The integration of services provided over networks is probably one of the last and most sophisticated phases in the long journey towards the integration of the European market. This chapter argues that in regulated network markets, integration should not be an end in itself. The conventional gains from trade or freedom of establishment, which are at the root of the traditional case for economic integration, may be outweighed by significant welfare losses in regulated markets, if integration involves the choice of a misguided deregulation model. Moreover, the design of the integration process will affect the distribution of the gains from integration, and this may be unacceptable to some of the countries and/or social groups involved. The chapter suggests that depending on the industry under consideration, the design of the deregulation cum integration process should ensure the maintenance of a level playing field and the protection of countryspecific strategic interests, to varying degrees. This may lead to the use of a variety of integration tools, combining sometimes strong harmonization
1
Comments by conference participants and an anonymous referee are welcomed.
158
Jordi Gual
of some regulations with a large degree of freedom at the Member State level in other domains. The chapter is organized as follows. Section 10.2 reviews the ways in which product and service markets may be integrated and considers the experience of EU integration regarding product markets and unregulated services markets. Section 10.3 highlights the distinctive features of network markets and discusses how those peculiarities affect both the evaluation of the gains from integration and the choice of the most appropriate integration strategy. This general discussion is illustrated in Section 10.4 with a brief analysis of the experience so far in the integration of three key network markets: banking, telecoms and electricity. Based upon this industry-specific analysis, Section 10.5 concludes with recommendations for the strategy of the EU in the integration of this type of service industries.
10.2 Strategies for market integration The achievement of a single market, beyond the simple formal elimination of trade barriers and the impediments to the establishment of foreign providers, requires that the authorities decide upon the general rules that will be applied to the products or services exchanged in the marketplace and to the firms that provide them. This is the case even in lightly-regulated product and service markets for two reasons. First, because even for the less regulated goods, there may be technical or safety standards that have to be satisfied for the general acceptance of a good in trade. And second, because the existence of these minimal regulations may in fact be used by countries as protectionist devices. That is, as artificial obstacles to trade that maintain market segmentation. These basic regulations or standards may be of two types: product and process standards. Product standards refer to conditions imposed on final goods, say for reasons of safety or technical compatibility. In services, these standards may refer to the quality of the service being provided, for example. Process standards, on the other hand, deal with the conditions that have to be satisfied within the firm, for example in terms of taxes, environmental controls or the use of some types of inputs (labour). 10.2.1 Strategies The creation of a single market may be achieved by a full harmonization of all these rules. This is, of course, the most stringent type of integration strategy. It guarantees the free flow of goods and services and the subse-
10 Integrating Regulated Networks Markets in Europe
159
quent gains from the increased possibilities of exchange, but it imposes substantial costs in the countries engaged in integration to the extent that some of them will have to sacrifice local regulations and adapt to the regulations of other countries. If existing rules (for example in terms of safety standards) reflect local preferences, this is a real welfare cost. Obviously, the rules may also be pure artificial restrictions to competition imposed by local providers or other pressure groups A less demanding type of integration strategy is the mutual recognition of regulatory regimes. This integration method implies that countries engaged in the integration process accept each other’s regulatory framework. That is to say, the country receiving a foreign product or firm, the host country, recognizes the validity of the regulations imposed on the firm by its home country. This is a very powerful integration method,2 short of the harmonization of rules. It requires a large degree of mutual confidence between the countries involved and a strong political willingness to integrate because mutual recognition may trigger a process of competitive deregulation. If foreign firms are subject to a regulatory regime that diminishes costs, mutual recognition is likely to increase the pressure for deregulation in the host country. Finally, another strategy of market integration is the use of the host country rule. This method implies that the host country need not alter its product and process standards, but, in the spirit of the well-known “national treatment principle” of trade, it is obliged not to discriminate between foreign and domestic providers. Obviously, this strategy leads to a lesser degree of integration to the extent that the rules of the host country may remain intact and foreign providers have to comply with them. 10.2.2 Products and services markets Within the European Union the process of integration of product and lightly-regulated service markets has been based upon the use of a variety of integration strategies. In product markets, integration requires free trade and freedom of establishment without restrictions. Public policy goals, such as the preservation of consumer safety and environmental protection, have been achieved for 2
Using a simulation model, Kox and Lejour (2005) show that the use of mutual recognition could dramatically increase trade in services within the EU. However, as highlighted by Ilzkovitz et al. (2007 pp 61-62) mutual recognition is hampered by legal uncertainty in practice and further legislation may be needed so that regulatory risk is reduced. This seems to be one of the goals of the recent initiative of the Commission (see European Commission 2007).
160
Jordi Gual
high-risk products by harmonization of product standards. For the rest of products mutual recognition is the dominant rule. The existing degree of political and institutional integration within Europe allows the mutual recognition of regulations for low-risk products, as well as the mutual acceptance of (moderately different) process standards (for example in terms of tax and social costs). Things are slightly different for services, even if we talk about lightlyregulated services, as shown by the complex political process lived in the EU for the approval of the directive which sets the basis for integration (the so-called Services Directive3). First, the legislation has excluded a large number of activities (such as audiovisual and private security) even though it is hard to claim that these sectors are as heavily regulated as network services. Second, even if mutual recognition has been accepted as the method of integration for service standards, host country rule has been maintained for the key process standards. That is to say, countries are obliged to accept foreign providers of services, but they may require from them the compliance with key domestic process regulations, and in particular with labour conditions and other items that affect social costs. The high labour content of services makes this provision highly restrictive in practice. Moreover, countries are allowed to maintain other local regulations on service providers (i.e. economic needs tests, fixed tariffs, restrictions on the number of outlets) which further restrict integration. The goal of the directive, however, is that these local regulations should be subject to a process of mutual evaluation, whereby best practice regulation is promoted and unnecessary restrictions are lifted. All in all, however, it is clear that host country rules dominate and that, due to domestic preferences and/or the pressure of local groups, the integration of lightly-regulated services with this institutional basis is not likely to advance very far. For both, products and lightly-regulated services, the process of integration is expected to yield substantial economic gains. These have been well researched and documented in the literature.4 Apart from the comparative advantage gains, they include those arising from increased competition (lower prices, larger number of varieties and increased consumer welfare) and from the achievement of a larger market (increased exploitation of scale economies, both statically and dynamically). Most importantly, however, the integration process in products (and to a lesser degree lightlyregulated services) is likely to have limited adverse effects which may thwart its political viability. The process of integration is mostly market 3 4
European Commission (2006). For a recent review of the gains from economic integration see Ilzkovitz et al. (2007) and the references therein.
10 Integrating Regulated Networks Markets in Europe
161
driven and its distributional effects are likely to be moderate with a relatively disperse impact on different social groups. Moreover, if the product and process standards are guaranteed in the key high-risk markets, the potential adverse effects on quality, variety and social conditions at large, are also -in principle- moderate. As we have seen for the case of services, the integration strategy is substantially more difficult when the areas involved (East and Western Europe) and the markets to be integrated (services which are intensive in unskilled labour) are perceived to lead to significant impacts in terms of the geographical distribution of the gains and losses of integration.
10.3 What is different about network services? Economic services provided over networks share some economic and socio-political features which make them distinctive in terms of the analysis of the conditions for and the consequences of market integration. Achieving a single market for network services (such as electricity, gas, water, telecoms, transport and retail banking) involves much more than the simple introduction of rules that liberalize entry or restrict discriminatory behaviour by domestic authorities. 10.3.1 Market characteristics The list of economic features of network service industries that lead to complex market structures, almost always oligopolistic, and pervasive government intervention is long and starts with the network concept itself. Network industries often involve direct network externalities whereby the value of the service to a client depends on the overall number of users of the network (telecoms, automatic teller machines, credit cards) or the number of connections (transport networks). These externalities affect the pricing of companies and very often lead to concentrated market structures. Network industries involve sometimes the use of physical distribution networks, and this implies a cost structure that again favours size and makes marginal cost pricing difficult if at all possible, due to large scale economies and huge sunk costs (well known examples are the copper networks in telecoms, the distribution grid in electricity and gas, and the networks used for credit card transactions). Needless to say, these networks pose other difficult problems of coordination and the management of congestion in some cases (electricity, transport). Finally, two additional features figure prominently: the peculiar structure of the value chain and, for
162
Jordi Gual
certain industries, the rapid pace of technological change. The value chain is important to the extent that, in many industries, infrastructure and service provision may have totally different “natural” market structures; moreover, the value chain will often include important bottlenecks at the infrastructure level, which affect the nature of competition at the service level (the local loop in telecoms, access to the network level in credit cards, access to pipelines in gas, etcetera). Technology is important because network industries are capital intensive and rapid technological change may create problems in terms of the incentives to pay for the new infrastructure that will substitute the old. Technological change is important in industries such as telecoms and banking, but less so in electricity, gas and water. On top of the economic features, network services have peculiarities from a social and political perspective. Most of these services are considered services of general interest and as a consequence their supply is subject to restrictions in terms of coverage and affordability (regulations on universal pricing and its funding) and to an on-going debate about the definition and scope of the basic service, particularly in fast changing industries such as telecoms. Similarly, these services are also considered as strategic by governments, in the (limited) sense of the word strategic that focuses on the need to guarantee uninterrupted supply, due to the general systemic consequences of a serious disruption in the provision of the service (obviously, banking is again similar due to the potential systemic risks arising from banking defaults). These social and political features of network industries, together with the natural monopoly features emphasized above, have always led to a complex regulatory framework that has been put in place by the state. It leads also very often to the direct intervention of the government, either by direct provision of the services (state-owned companies) or through the financial support of regulated private firms that satisfy political goals (state aid or favourable regulatory conditions). It must be obvious at this stage that the integration of this kind of very complex services is not at all simple. How should it be done? The answer to this question may benefit from a preliminary discussion of something that we take for granted in goods and lightly-regulated services, but is not that obvious in regulated network markets. Why should we integrate those markets at all? Are we sure that integration in these markets leads to an overall welfare gain, with limited adverse effects in terms of distribution? A positive answer to this more fundamental question is not obvious in regulated industries which are in the process of being deregulated. For example, some of the gains come from increased competition, but do we know how large are they? And, more importantly, do we know which is
10 Integrating Regulated Networks Markets in Europe
163
the proper target for the degree of competition? Both from a static and a dynamic perspective, these are industries where the ideal of perfect competition need not be desirable, and may not even be attainable. The same problems arise in terms of the potential gains from static and dynamic efficiency (innovation and learning by doing). There is much uncertainty both on what type of deregulation triggers a more effective increase in competition and on the market structure which is more conducive to the achievement of dynamic efficiency gains.5 Two additional characteristics of network services complicate the assessment of the net gains from integration. First, these are industries where consumer surplus is determined not only by final prices but also by quality measures; and industries where the total absence of regulation is in many cases still very far away. Clearly, under these conditions the specific deregulation road and the final regulatory framework are key determinants of the welfare gains. Second, it is not at all clear that the distributional effects of achieving the single market can be neglected in network services as we often do in goods markets. This is so because in network services integration may involve the adoption of completely different regulatory frameworks and the appearance of totally different and concentrated market structures at the EU level. These changes are likely to have strong distributional consequences, particularly across borders (the control of the remaining companies may remain in a few centres across the EU), and they may jeopardize the political viability of the integration process. 10.3.2 Integration I would argue, therefore, that the gains from integration in regulated network services are less clear-cut that in other more conventional markets. More importantly, the previous remarks highlight the need to be extremely careful in the integration process, and the importance of adopting an integration method or strategy that appropriately balances all the conflicting issues that arise. These issues may be easily summarized. It is clear that if policy-makers wish to integrate these services, they will need to ensure access and interconnection to guarantee cross border provision and free entry, but this immediately begs the question of the regulatory conditions faced by the different competitors in their home country. This is, in the end, a question of product and process standards in a rather complex setting. Even if competitors face similar (non-discriminatory) conditions in 5
On these issues see, for example, Sutton (2007) and Armstrong and Sappington (2007).
164
Jordi Gual
the host market, the key question often revolves around the (regulated process) conditions faced in the home market, which is far more important and ends up determining the overall cost competitiveness of the companies. The question of the level playing field is complex for several reasons. First because, as hinted above, we face considerable uncertainty as to the optimal path of deregulation, and countries may (reasonably) disagree on that choice. Secondly, due to the fact that countries may be tempted to use (favourable) regulatory conditions as a strategic tool to place their domestic competitor in an internationally advantageous position. And it will often be impossible to classify this support as state aid. Finally, because deregulation is intrinsically multidimensional and the extent of deregulation varies across countries – both in intensity and form – due to a variety of reasons, such as legal history, the preferences of electorates and the lobbying of affected parties. Apart from guaranteeing a level playing field, two other issues appear as critical in the process of integrating network markets. The first, is the choice of the proper regulatory framework for the integrated market (and the associated deregulation path). And, in fact, given the long-lasting impact of regulatory decisions on business, the degree of flexibility of this deregulation process. The second, given the strong social and political underpinnings of these industries, is the choice of an integration process that takes due account of the diversity of local preferences and strikes an adequate balance between the progress of the integration and its gains, and the preservation of vital domestic interests. In the absence of appropriate compensation mechanisms, the political viability of integration hinges upon the satisfaction of these restrictions. Indeed, I will argue that the integration strategy for network industries should be based upon a combination of methods (ranging from harmonization to host country rule) which may be different across network services, but which will depend on the relative importance of these key goals in the integration process: the preservation of the level playing field, the flexibility of the deregulation path, and the guarantee of respect for local preferences. The discussion of the alternative integration methods in Section 10.2 and the preceding description of the conflicting goals of the integration process make it clear that each goal is most suitably achieved by some integration methods, while others may be totally inappropriate. Table 10.1 summarizes these relations between goals and integration strategies. The table is self explanatory and just a few remarks are in order. Harmonization does a good job at levelling the playing field, but note that this can also be achieved through mutual recognition, although this could lead to
10 Integrating Regulated Networks Markets in Europe
165
the disappearance of regulation altogether. Regulatory flexibility is achieved clearly by host country rules, but this comes at the cost of sacrificing integration. Finally, ensuring the respect of local preferences (including the supervision of foreign providers for strategic reasons – such as financial stability and energy security) requires host country rules, unless the degree of political integration is already very high. In general, the choice of integration method will depend upon the importance of each key goal for each industry. Table 10.1. Integration strategies and the (conflicting) goals of integration Integration Strategies Key goals in the integration process
Host country rule
Mutual recognition
Harmonization
Preventing regulatory rigidity Levelling the playing field Ensuring respect for local preferences
9
9
8
8
9
9
9
8
8
10.4 The integration strategy in specific network services: banking, electricity and telecoms The examination of the integration strategies adopted by the EU in the landmark cases of banking, electricity and telecoms and the divergent degrees of integration achieved in these markets provide an interesting background for the examination of why the most appropriate route to integration need not be the same across industries and need not coincide with the route taken so far. As argued above, there are many dimensions to regulatory intervention. For the purposes of our analysis of the comparative integration strategy followed by the EU in different EU network service industries, I have classified them in two broad categories and eight types of regulations, as shown in Table 10.2. Table 10.2 follows the tradition of industrial organization. It distinguishes between regulations that affect the behaviour of companies, and those that affect the structural conditions of the market. What is novel is the introduction, under conduct regulation, of the interconnection conditions that affect cross border services provision.
166
Jordi Gual
Table 10.2. The key dimensions of regulatory intervention Regulation Conduct regulation Technical and other product standards or regulations Market power Cross-border interconnection Structure regulation Entry Vertical restrictions Line of business restrictions Access to infrastructure Mergers and state aid
Examples Standards of safety, information, etc. that have to be satisfied by the product or service. General controls on prices, bundling of services, spending on advertising, etc. Limits on dominant companies Feasibility and conditions of interconnection Limits on entry (minimum size of new companies or other restrictions) Limits or restrictions on vertical integration Limits or restrictions on ownership of horizontally related business units (i.e. Banking and insurance, fixed and mobile, etc.) Conditions of access and interconnection, including organization of wholesale markets Restrictions on mergers and state aid
Regulatory divergences matter even if, as it is the norm in the EU, there is no discrimination between domestic and foreign providers. They matter both for the cross-border provision of services and for the provision through the establishment of foreign affiliates. This is so because favourable home country regulations may provide companies with a competitive advantage when facing other companies in third markets. As we will see next, it is the willingness to tackle regulatory divergences in these areas that is at the basis of the achievement of different degrees of integration across industries. Tables 10.3 to 10.5 provide a broad assessment of the way in which the EU has advanced in the process of integration of the markets for banking, telecoms and electricity. For each market and each type of regulation, the key integration method is highlighted in the tables, although in many cases it is necessary to comment and qualify the general characterization. The tables are self-explanatory, and from them it is possible to extract a general perspective on the integration strategy followed in each market.
10 Integrating Regulated Networks Markets in Europe
167
10.4.1 Banking In banking, the general principle of integration has been that of mutual recognition.6 However, the danger of a race to the bottom on crucial issues, such as the rules on prudential matters, has led to the establishment of harmonized rules in this dimension. Similarly, even if the general principle is that of mutual recognition, the rule of host countries has remained prevalent on sensitive issues that relate to local preferences (rules on consumer information) or the stability of the domestic financial market (the supervision of foreign providers for monetary policy reasons). The banking industry in fact provides an example of the simultaneous use of two integration methods: mutual recognition and host country rules, for the same type of business, the local provision of financial services. This provision can be undertaken by means of a subsidiary or by a branch, and only in the second instance the more advanced integration method of mutual recognition (the so-called single passport which implies that home regulations apply with the exceptions on monetary policy and consumer protection highlighted above) is used. Overall, the extent of integration in retail banking across the EU has been substantial. It is clearly the case that more could be done. For example, the mutual recognition of competition policies has been shown to be insufficient, since some countries have used local regulations as protectionist devices. Similarly, the host country regulations that apply to teller networks and settlement systems could be also mechanisms that prevent integration. But the same could be argued with regards to the rules on consumer information and protection. The harmonization of these rules and systems would certainly allow a fuller exploitation of scale economies across the EU, both in terms of the design of financial products and in terms of the efficiency of cross border financial transactions. Nevertheless, these gains have to be balanced with the potential losses, in terms of the welfare of specific countries with idiosyncratic payment systems or preferences regarding the protection of consumers. This trade-off is exemplified by the proposal of the Directive on Consumer Credit Loans (IP/07/687, 21 May 2007). The new Directive would go beyond the current minimum requirements, which de facto have led to a host country rule regime (a patchwork of different rules across the EU), to a partially harmonized framework whereby all EU countries would require from credit providers comparable terms with regards to key issues such as advertising, the calculation of the annual percentage rate of charge, the
6
See Gual (2004) and Barros et al. (2005) for a detailed discussion.
168
Jordi Gual
right to withdraw from a contract within a limited time period, and the terms of compensation to creditors for early repayment. Table 10.3. Regulatory dimensions and integration method in banking Regulation
Integration method
Comments
Conduct Regulation Technical and other prod- Host country rule uct standards or regulations Market power Mutual recognition
Rules on information and consumer protection Of competition policies
Cross-border interconnec- Host country rule tion
Only recently harmonization for international transfers
Structure Regulation Entry Vertical restrictions
Partial harmonization and mutual recognition Mutual recognition
Line of business restrictions Access to infrastructure
Mutual recognition Host country rule
Mergers and state aid
Mutual recognition
Minimum requirements for prudential objectives Countries free to choose the degree of integration Countries free to choose any restrictions Access to ATM network and settlement systems Through the mutual recognition of competition policies, even if these are not fully equivalent (exc. Italy and Portugal)
In terms of Table 10.3, this implies moving from host country rule to partial harmonization. As argued, this is fine to the extent that different domestic rules are being used as protectionist devices and do not reflect genuine local preferences. As a general rule, however, domestic regulations that protect consumers should be respected if it can be shown that they do not imply the “de facto” discrimination of foreign providers, and that no equivalent protection is guaranteed by the home country regulations imposed to those providers. Finally, it is worth stressing that the extent of integration itself need not be the same in all markets even when all barriers to integration have been removed. Some markets, such as retail banking, are likely to remain geographically segmented because the basic economics of the industry (asymmetric information) lead to a fragmented market structure on a regional basis. In this type of markets, policies that artificially promote market integration beyond its natural level are probably ill-advised (for example, in markets such as mortgage credit or pension
10 Integrating Regulated Networks Markets in Europe
169
funds, where contracts and product information and regulations are closely related to local commercial and tax law, and the overall structure of the social security system).7 10.4.2 Telecoms The telecoms industry (see Table 10.4) provides an example of an industry where the EU has used an alternative approach to integration, but with results that have been fairly limited so far. The basic integration strategy used in this industry has been the establishment of limits to the rules introduced by the domestic regulators. Following our taxonomy, we refer to this as “host country rules within limits”. The EU ‘deregulation cum integration’ strategy in telecoms was set up in the framework directives back in 2003 and is currently in the process of being revised. The basic tenet of the strategy is the gradual liberalization of the industry, by way of establishing competition benchmarks which allow, in a consistent manner, the removal of regulatory restrictions and the use of general competition policy to assess market problems. Obviously, a full fledged discussion of this regulatory strategy is beyond the scope of this chapter, but a few general remarks are needed.8 First, the strategy has led to a substantial amount of guidance on the key issue of the deregulation strategy for fixed telephony and wireline broadband, while leaving lots of leeway in other dimensions which are in practice of no less importance (i.e. the regulation of alternative networks). Indeed, the guidance provided in broadband has possibly reduced potentially useful regulatory experimentation, and the lack of stronger market integration rules in mobile (the integration of spectrum) has artificially thwarted the crossborder growth of this market. On the positive side, the regulatory framework and the increased coordination of domestic regulatory bodies has gradually ensured that (within the domains contemplated by the framework) the application of the rules across borders has been quite consistent, thus limiting to a reasonable degree potential problems of uneven competitive conditions faced by the different competitors.
Alternatively, there is potential for integration in markets such as point-of-sale consumer credit or in saving products. Nevertheless, in this last instance it is not obvious whether the integration will take place directly (through cross-border sales) or indirectly (through the local provision of saving products that invest in EU-wide assets). 8 See Bergman et al. (1998) and Gual (2002). 7
170
Jordi Gual
Table 10.4. Regulatory dimensions and integration method in telecoms Regulation Conduct Regulation Technical and other product standards or regulations Market power
Integration method
Comments
Host country rules (within some limits)
Limits on controls that may be imposed by national authorities
Limits on how abuse of market power is defined and what remedies are applied Cross-border intercon- Host country rules for In mobile, the GSM standard fanection mobile. Harmonization cilitates interconnection, but for fixed telephony. spectrum is allocated on a national basis Structure Regulation Entry
Host country rules (within some limits)
Host country rules (within some limits for certain segments)
Vertical restrictions
Host country rules
Line of business restrictions
Host country rules
Access to infrastructure
Host country rules (within some limits)
Mergers and state aid
Mutual recognition
Service provision to be subject to authorization instead of licensing. No limits on domestic rules on entry conditions for mobile operators No common rules on, for example, virtual mobile network operators, or fixed network unbundling No common rules on crossownership of, say, cable and ADSL The EU establishes limits on the Member State rules on access pricing Of national competition policies
10.4.3 Electricity Finally, the electricity market (Table 10.5) provides an example of yet another integration strategy, based upon a loosely shared deregulation framework, but without significant limits on the regulations that may be established by local authorities. This last feature, together with the limited efforts undertaken in the achievement of physical interconnection facilities, explain the very limited degree of integration achieved in this market. In particular, the large room for manoeuvre left to national authorities has led to substantial differences in terms of the regulatory conditions faced by
10 Integrating Regulated Networks Markets in Europe
171
domestic competitors, and thus placed some of them in a clear competitive advantage in the European market for corporate control. Table 10.5. Regulatory dimensions and integration method in electricity Regulation Conduct Regulation Technical and other product standards or regulations Market power Cross-borderinterconnection
Integration method
Comments
Host country rules Restrictions (time limits) on the (within some limits) rythm of liberalization Host country rules (Limited) Harmoni- And limited physical connections zation
Structure Regulation Entry Vertical restrictions
Host country rule Host country rule Some limits on vertical restrictions (within some limits) (at least accounting separation but no common rules on ownership) Line of business re- Host country rule Subject to common competition polstrictions icy, no restrictions on ownership of gas and electricity assets Access to infrastruc- Host country rule Loose limits on third-party access ture (within some limits) and the reorganization of wholesale markets Mergers and state aid Mutual recognition State aid to public companies, and (with many excep- golden shares tions)
As with telecoms, the ‘deregulation cum integration’ strategy in electricity has involved a fairly definite choice of a regulatory path: in this case based upon the idea of the gradual opening of wholesale markets, depending on client size, and a second stage where competition reaches the residential consumer.9 The deregulation strategy is to be implemented across all the EU Member States, with mutual opening to foreign providers to the extent that each market has advanced in the route towards liberalization. The choice of a fairly rigid deregulation path applies to countries with very different sources of primary energy and previous regulatory setups. It is, therefore, not at all clear that such a strategy will favour the adoption of liberalization measures, particularly in an industry where political and strategic concerns are crucial. 9
See Newbery (2004). On the international experience in this area see Wolak (2005).
172
Jordi Gual
10.5 Concluding remarks The analysis of the integration strategy followed by the EU in key network services such as banking, telecoms and electricity makes it clear that the differences in the approach to integration and the progress made thus far are not only the result of institutional and political restrictions. They also stem from a lack of clear priorities in terms of the key integration goals to be achieved in each market. This chapter has argued that the goals may be quite different across industries. In an industry as dynamic as telecoms, probably the overriding concern should be the prevention of excess regulatory rigidity, since the costs of implementing an inadequate framework may be particularly high when technology changes fast. Worries about differential competitive conditions should be a second-order concern. Similarly, with the increased variety of telecommunications services and their increased competitive supply, the risk that the more competitive and homogeneous market impacts negatively on local preferences appears to be secondary. In banking and electricity, however, the concern about the level playing field and local preferences is critical. This implies that the choice of a deregulation cum integration strategy should be based on well-known processes which guarantee that no competitor enjoys unfair competitive advantage and that the stability of national banking and electricity systems is guaranteed, both in terms of systemic issues and of the information and guarantees offered to consumers. Technology in these industries changes relatively slowly compared to the fundamentals of the service provided to users, and the integration process should never put in jeopardy the broader social goals that banking and energy systems satisfy. The recognition of these differences in goals and the evaluation of the most proper integration method for each industry should be used in any revision of the integration strategy for network markets. In this regard, the new policy approach unveiled by the European Commission in February 2007 provides new instruments that would be useful in the pursuit of a more nuanced and differentiated integration strategy. In particular, the 2007 Commission Communication (European Commission 2007) refers to the right balance between harmonization and mutual recognition. As argued in this chapter, this balance is key in network services. For example, in electricity there is probably a need for further harmonization, but mutual recognition appears to be a better instrument in telecommunications. The Communication includes also useful tools such as the exchange of best practices, self and co-regulation, the pro-active enforcement of competi-
10 Integrating Regulated Networks Markets in Europe
173
tion policies and EU law, and cooperation mechanisms for the compatibility of national legislations. In banking the strategy used so far has been based on the joint implementation of partial harmonization and mutual recognition. This approach has built upon a long tradition in international banking regulation and requires a fair amount of mutual confidence between regulatory bodies and policy makers across the single market. The process has proved successful and the scope for improvement seems to be limited, at least in retail banking, which is a business naturally confined to regional markets. Improvements could be obtained by a more systematic treatment of competition policy rules across Member States, and also by a permanent assessment of host country rules for consumer protection, ensuring that they are not used as protectionist devices. In telecoms the EU approach has been very much determined by the inertia of old regulatory regimes and the disparities across countries in terms of regulatory agencies and governments with different competencies and degrees of involvement in the industry. A broad overall assessment, would lead us to conclude that there is probably increased room for the use of the mutual recognition principle, which would trigger more competitive deregulation and the spread of best regulatory practices. In some areas (mobile telephony) additional policy steps should be taken to promote the integration of spectrum. Finally, electricity is probably the market, among those examined in this chapter, where the integration strategy pursued by the EU has been less effective. Apart from the obvious importance of political economy and national security considerations, it is clear that the EU approach has not been pro-active. Rather, we can see it as the EU level reaction to a piecemeal process of deregulation that has taken place all over Europe, as a follow-up of the early deregulation experiences in the Anglo-Saxon countries. The EU integration approach has been an attempt to cope with these deregulation processes and make them compatible with a gradual elimination of barriers to the cross-border provision of services. Looking ahead, the integration approach in electricity should combine the increased harmonization of key regulatory cost drivers, which today are loosely controlled and where basically the rules are set by the home country (ie. very loose limits on vertical ownership) with the maintenance of host country rules with few limits in very sensitive market segments (retail customers). At the same time, since in the electricity markets most of the benefits from integration arise at the wholesale level, the previous strategy could be combined with a more aggressive integration policy at that level, with increased interconnection facilities and competition enforcement for the generation market and the distribution to large industrial customers.
174
Jordi Gual
In sum, it should be emphasized that each network service market is different, and even if conceptually there are relevant general rules that proper regulation should take into account (see Armstrong and Sappington 2007) the choice of the deregulation cum integration strategy should be industryspecific. This chapter has attempted to provide a general framework that specifies the integration parameters that have to be chosen and highlights the key trade-offs. The main message, however, is that not only there is no single path towards deregulation and integration, but also that integration should never be an end in itself, since –given the nature of these industrieseconomic and social welfare is going to depend also, and crucially, on getting the deregulation process right.
References Armstrong M, Sappington D (2007) Recent developments in the theory of regulation. In: Armstrong M, Porter R (eds), Handbook of industrial organization vol 3, North Holland: 1557–1700 Barros PP, Berglöf E, Fulghieri P, Gual J, Mayer C, Vives X (2005) Integration of European banking: the way forward. In: Monitoring European deregulation 3, CEPR Report, London Bergman L, Doyle C, Gual J, Hultkrantz L, Neven DJ, Röller L-H, Waverman L (1998) Europe’s network industries: conflicting priorities: Telecommunications. In: Monitoring European deregulation 1, CEPR/SNS Report, London European Commission (2006) Directive 2006/123/EC of the European Parliament and of the Council of 12 December 2006 on services in the internal market European Commission (2007) A single market for citizens. In: Communication from the Commission to the Council, the European Parliament, the European Economic and Social Committee and the Committee of the Regions COM(2007) 60 final, Brussels 21.2.2007 Gual J (2002) Regulation and the development of electronic communications in Europe. INFO 4 no 3, August Gual J (2004) The integration of EU banking markets. In: Gual J (ed) Building a dynamic Europe: the key policy debates, Cambridge University Press, Cambridge, UK Kox H, Lejour A (2005) Regulatory heterogeneity as obstacle for international services trade. CPB Discussion Paper 49 Ilzkovitz F, Dierx A, Kovacs V, Sousa N (2007) Steps towards a deeper economic integration: the Internal Market in the 21st century A contribution to the Single Market Review. DG EcFin Economic Papers, 271, Brussels Newbery D (2004) Integrating and liberalizing the market for network services; gas and electricity. In: Gual J (ed) Building a dynamic Europe: the key policy debates. Cambridge University Press, Cambridge, UK
10 Integrating Regulated Networks Markets in Europe
175
Sutton J (2007) Market structure: Theory and evidence. In: Armstrong M, Porter R (eds) Handbook of industrial organization vol 3, North Holland: 2301–2768 Wolak F (2005) Lessons from international experience with electricity market monitoring. World Bank Policy Research Working Paper 3692
11 Subsidiarity and the Internal Services Market
Arjan Lejour
11.1 Introduction The assignment of Internal Market policy to the European Union is probably one of the most undisputed ones. Internal market policy cannot be shifted to the Member States without disrupting the free movement of goods, services, capital and labour. However, full market integration also affects national policies, because many domestic policies hamper market integration, at least indirectly.1 Then the question becomes to which extent the Member States are willing to switch authority to the EU level or are prepared to coordinate national policies in order to curtail the trade hampering effects of these policies. From the subsidiarity principle the question is whether the benefits of coordinating these policies outweigh the loss of less targeted national policies to national circumstances and preferences. I will discuss this case for commercial services. Commercial services represent about 40 percent of the European economies, but are rather closed to foreign competition. Trade in commercial services is at most one-fifth of total world trade in goods and services. Compared to manufacturing and agriculture, services sectors are less open to international trade. This difference with the goods-producing sectors has two basic causes. The first is that services production and delivery often needs the proximity of producer and consumer, so that the services supplier has to move abroad.2 This makes cross-border trade relatively diffi1
2
One example is national corporate income taxes, which will be discussed in Chapters 13 to 16. In the case of tourism and travel the consumer moves abroad.
178
Arjan Lejour
cult compared to goods, where normally only the product is shipped abroad. The second is that regulation-caused non-tariff barriers limit the international integration of services markets. Regulations for services suppliers, for foreign investors and for the services products themselves often are primarily established for domestic purposes without taking account of the interests of foreign service providers. The fact that a national service market is regulated need not be in itself a barrier to international services trade. The problem arises because each country has different regulations in place and does not recognize qualifications in a foreign firm’s home country. Then the qualification costs to enter a foreign market become cumulative fixed costs. Because the costs are country-specific, they are in fact sunk market-entry cost for a country market. This hampers exports and investment.
11.2 The Services Directive The Services Directive (SD) introduced by the European Commission (2004, 2006) intends to make headway with a common market for services, by reducing the negative impact of heterogeneity in regulation, by ruling out national measures that explicitly or implicitly discriminate against foreign service supplies, and by calling for measures facilitating trade and investment in services. All services
Traded
Non-traded, not SD-covered - government services - health, education - localised services (haircut)
Non regulated, covered by SD
Regulated, covered by SD
Regulated, not covered by SD
Network services - tourism - wholesale trade - consultancy & analysis - voluntary tests, certifying - data processing - software - logistics - market research - simple cleaning
- professional services - advertising, cleaning, - engineering, architects - entertainment, audiovisual - construction - recruitment - hotels, cafes, restaurants - retail trade - obligatory tests,certifying - repair, maintenance
post, telecom, gas, water & electricity distribution, rail, airports, broadcasting
Other EU regulatory frames
banking, insurance, road & sea transport, investment
Fig. 11.1. Sectoral coverage of Services Directive
11 Subsidiarity and the Internal Services Market
179
Figure 11.1 tentatively shows the sectoral coverage of the Services Directive. The regulated services sectors are the ones which are most affected by the directive. Gual (2008) concentrates in Chapter 10 on the (heavily) regulated network industries which are not covered by SD. In this chapter, the regulated services covered by the SD are the topic of analysis. Gual discusses the various strategies to deal with the differences in regulation in order to integrate services markets: full harmonization, mutual recognition and the host country rule. The host country rule guarantees the preferences in regulation of the host country, but could disturb trade because services providers have to comply with regulation in the host country, which differs in most cases from regulation in the home country. This problem could be solved by harmonization or the mutual recognition principle, but these strategies do not address the various preferences of the Member States completely. The original proposal for the SD (European Commission 2004) introduced the country of origin principle. This principle implied that the host country had to accept the regulation of the home country. This could be interpreted as an extreme case of the mutual recognition principle (Nicolaïdes and Schmidt 2007). The amended proposal (EC 2006) more or less skipped the country of origin principle and places much more weight on the host country rule and the principle of mutual recognition. The country of origin principle requires trust in each others regulation, cooperation between national authorities and runs the risk of competition in regulation to support national firms. As a consequence, people were afraid of lower regulatory standards, which would hurt consumers in particular, although SD applies mostly to business services. In contrast to the country of origin principle, mutual recognition contains explicit derogations (see Pelkmans 2006), which are conciliatory to the various preferences of Member States. If the host country imports a service for which it regulates the quality in order to protect the consumers and such regulation does not exist in the exporting country, the host country is legitimated to ask for compliance to this regulation. However, the principle of mutual recognition is not widely understood and gives some leeway to country specific interpretation as appears from the application to trade in goods. There are many complaints listed on the application of the principle by the Member States. Therefore the EC (2007) introduced some new proposals to enhance the implementation of this principle, to improve information dissemination and to reduce the uncertainty for producers about its application. Ilzkovitz et al. (2007) suggest that trade in goods for which this principle applies could at least increase by 20% due to these measures, resulting in a GDP increase of about 0.2% in Europe.
180
Arjan Lejour
The amended SD thus takes more account of the local circumstances in the Member States at the expense of less market integration as will be illustrated in Section 11.3. The question to ask is what are the limits of internal market policy for services. Seen from the subsidiarity principle spelled out by Ederveen et al. (2008), internal market policy is welfare improving as long as the benefits of economies of scale and of taking care of the externalities outweigh the benefits of decentralized policies that better target the preferences and local circumstances in the Member States. Does the amended SD strike the right balance or does the subsidiarity test suggest welfare improvements from further market integration?
11.3 The benefits of integrated services markets The most important reason, from an economic point of view, to centralize internal market policies is the coordination of external effects induced by national policies. Internal market policies slash down trade-hampering barriers, which have a negative effect on firms in other countries. A larger market also increases the possibilities for firms to exploit economies of scale. With its initial SD proposal (EC 2004) the European Commission wanted to accomplish a European Single Market for a large part of the services sector. The proposals would eliminate important obstacles to the freedom of establishment and the free movement of services, while strengthening mutual trust between the EU countries on their regulatory regimes. Using the 2004 proposals as a point of departure, several studies have quantified the potential effects of the proposed measures. Here we summarize their main results. Kox and Lejour (2006a) approach the issue by quantifying the sunk market-entry costs of country-specific regulations, accounting for differences in product-market regulations between each EU country pair. Analysed in gravity equations for bilateral services trade in the EU, regulatory heterogeneity in policy areas like competition and trade regulation appears to have a robustly negative trade impact. At a detailed level they subsequently estimate the extent to which the Commission's SD would affect bilateral heterogeneity in regulation. Overall trade in commercial services (excluding transport and travel) could increase by 30 to 62 percent; and intra-EU FDI in services could increase by 18 to 36 percent. De Bruijn et al. (2008) have fed these estimated trade impacts (not the FDI effects) into a large CGE model (WorldScan) in order to estimate the macro-economic effects. They conclude that average European consump-
11 Subsidiarity and the Internal Services Market
181
tion could increase by 0.5 to 1.2 percent and that the country-of-origin principle (or shortly: CoOP) accounts for about one-third of these effects. This illustrates the point made in Section 11.2 that eliminating the proposed home country rule from the SD reduces the positive trade and economic effects. Breuss and Badinger (2005) use these quantitative estimates of the SD’s services-trade effects to estimate how much additional trade will erode profit mark-ups in EU services markets. They find that more entry does not directly affect productivity, but for the sample of service industries covered by SD, they find significant positive effects of trade on competition. The latter has significant positive impacts on productivity, employment and investment in EU services. Copenhagen Economics (2005a, 2005b) analyses the welfare effects of the EU proposals with a CGE model which also considers the FDI effects. They assess that overall consumption in the European Union would increase by 0.6 percent. According to their analysis the CoOP contributes only about 10 percent to the total welfare effects (including the FDIinduced effects). Vogt (2005) considers the estimates of Copenhagen Economics to be conservative, because all dynamic effects of extra competition on productivity and innovation are left unconsidered. This criticism also holds for the studies of Kox and Lejour (2006a) and De Bruijn et al. (2008). Both studies concentrate on the static (one-off) effects of opening up European services markets: economies of scale, efficiency gains through inter-sector supply linkages. Kox and Lejour (2006b) conclude that the upshot from the available scarce evidence is that the prime dynamic gains from services liberalization will come from more new market entry by firms based in other EU countries. Improved market access will stimulate competitive selection and productivity growth. Competitive selection will lift average productivity, bolster the role of SME firms in exports, intensify knowledge spillovers, and strengthen innovation by incumbent firms. Moreover, increased FDI in liberalized services markets will also increase average productivity. This can be expected to be beneficial for the variety of available services, for service quality, and for the price of services.
11.4 Diversity There are clear economic benefits of integrating (lightly) regulated services markets. However integration also comes at a cost. In the early process of market integration the Member States eliminated direct barriers to
182
Arjan Lejour
trade such as tariffs, quotas in goods trade, prohibitions or discrimination based on nationality. These barriers had the explicit purpose of hampering trade. By giving up these national trade policies all Member States benefited from a more integrated market. With respect to regulation in services this is far more difficult. Often regulations are implemented to serve other purposes than trade protection. In general the regulation in place has several dimensions. If countries agree to give up some autonomy in regulation, not just protection but also the other policy dimensions are affected. In fact the freedom of countries to deal with these other policy dimensions is curtailed. If Member States would have the same preferences for regulation or have the same economic structure, harmonization could be a solution, but this does not match with reality. This section argues that Member States have different preferences for regulation and different economic structures which imply different effects of services trade liberalization within Europe. 11.4.1 Diversity in preferences for regulation The OECD has developed indicators for product market and labour market regulation. Inspection of these indicators shows that on average AngloSaxon countries are less regulated than Scandinavian countries. The latter group of countries regulate their product and labour markets less than continental countries in Europe and Mediterranean countries seem to have still a bigger appetite for regulation. Also within these four groups of countries the level of regulation differs. The differences do not only appear for product market regulation in general, but also for regulation in specific services sectors. Below I review the level of regulation in two regulated service sectors: retail distribution and professional services. The level of regulation is interpreted as a kind of revealed preference for regulation Retail distribution
The retail sector is often subject to numerous regulations that weaken competition. Conway and Nicoletti (2006) observe large differences in the indicator of retail regulation across OECD countries, suggesting very different policy approaches, see Figure 11.2. In Sweden there are only some operational restrictions, while most other countries also have entry barriers and some price controls. In particular in Greece and Belgium the level of regulation is high in the retail sector. Barriers to entry definitively hamper market access of foreign providers, but operational restrictions are often
11 Subsidiarity and the Internal Services Market
183
also discriminatory, because these regulations often require a different behaviour and conduct than in the home country. 5
4
3
2
1
0 Sweden
Ireland
Netherlands
United Kingdom
Portugal
Italy
Japan
barriers to entry
Denmark
Finland
United States
operational restrictions
Germany
France
Austria
Spain
Greece
Belgium
price controls
Fig. 11.2. Regulation in retail distribution in 2003 Source: Conway and Nicoletti (2006). Professional services
Conway and Nicoletti (2006) argue that regulation in the professional services often limits competition by restricting entry, allowing for price fixing, granting exclusive rights to perform particular services, and restricting advertising and business structures. These regulations are claimed to be in the interest of consumers because they improve quality of services and overcome information asymmetries. There is little empirical evidence that indicates at a positive impact on consumer welfare. The indicators of regulation in professional services (Figure 11.3) suggest that in Scandinavia, the Anglo-Saxon countries, and the Netherlands barriers to entry into professional services are less strict compared to other EU countries and Japan. Differences among restrictive countries mostly reflect differences in conduct regulation, that is, restrictions on price setting, advertising, form of business, and inter-professional cooperation. In some countries conduct regulation is nearly absent.
184
Arjan Lejour
8 7 6 5 4 3 2 1 0 Denmark
Sweden
Finland
United Kingdom
Ireland Netherlands United States
Austria
entry regulations
France
Belgium
Japan
Portugal
Spain
Greece
Germany Luxembourg
Italy
conduct regulation
Fig. 11.3. Regulation in professional services in 2003 Source: Conway and Nicoletti (2006).
11.4.2 Diversity in economic structure The importance of the (lightly) regulated services sectors varies widely across the Member States. Using the value-added share of other commercial services (including all traded services but excluding network services, see Figure 11.1) as an indicator for its economic importance, it ranges from 26 percent in Czech Republic and Slovenia to about 55 percent in Austria and Germany (in 2001). As a general pattern, the size of other commercial services is relatively smaller in the new Member States and Scandinavian countries and larger in the old continental Member States, see De Bruijn et al. (2008) for more details. Most continental and Anglo-Saxon countries are specialized in exporting other commercial services, while most other countries are not. The simulations result of De Bruijn et al. (2008) show that the GDP effects of implementing SD may vary between 3 and 5 percent for most of the new Member States, whereas they equal about 0.5 percent for some older Member States. The variation in country effects depends on the reduction in regulation (which differs for each trade relation), the terms-of-trade effect and shifts in specialization. Some of the original EU Member States increase their relative specialization in commercial services due to the
11 Subsidiarity and the Internal Services Market
185
more open borders. The new Member States, however, reallocate resources to manufacturing. The variation in effects also suggests diverging interests of Member States in integrating services markets. As preferences for regulating the economy and economic structures differ between Member States, harmonized regulation of services markets would not meet particular national circumstances. However, the differences in national regulation hurt services trade and the benefits of integrating European services markets are overwhelming. The home country rule expressed by the country of origin principle overcomes these trade barriers but does not meet national circumstances and preferences. Moreover it requires much trust between Member States. In that respect, the principle of mutual recognition seems to be a better instrument to balance the benefits of an integrated services markets and national diversity. However, the balance between the benefits of a centralized internal market policy for services and the benefits of national-targeted policies could also be affected by other considerations which are dealt with in Section 11.4. Furthermore, the balance could shift over time as is argued in Section 11.5.
11.5 Political economy issues Ederveen et al. (2008) present various explanations for the discrepancy between the assignment of policies in the EU and the outcome of the functional subsidiarity test. Most of the explanations are political economy arguments. Some of them could explain why a policy is not assigned to the most efficient level, others give a good reason why it should not be assigned to the most economically? We will discuss some of these arguments with respect to the internal market for services.3 From a political economy and public choice point of view there are several reasons why Member States hold on to the authority to regulate services markets. The first is self interest of the government (Leviathan behaviour). If Member States regulate services, citizens (consumers and firms) can choose to settle in a Member State with their preferred regulation. However, labour mobility is low in Europe and the movement of firms as well, so in practice this argument is hardly convincing for decentralization. A second argument for decentralization is that it enhances policy competition with home country regulation. This could prevent an overkill of regulation that slows down growth. However, this mechanism could also lead 3
To be clear I do not try to explain the political discussion on the Services Directive, see Pelkmans (2007) for an overview.
186
Arjan Lejour
to an under-provision of regulation in services. Then the competition between national governments in attracting services could lead to a bigger government failure than the original market failures which were reason for government intervention in services markets. Noting that still more than 90 per cent of all service firms only serve the home market, it is questionable whether this kind of policy competition is effective. There are also arguments to centralize policies in services markets. The first is that the treaties proclaim the free movement of services. Without common regulation or less regulation there will be no effective internal market policy for services. These benefits are to some extent already covered by the external effects discussed in Section 11.3. The issue of common policies or national policies may be also solved to some extent by choosing the appropriate type of governance. Internal market policy requires full delegation to the central level, but coordination is a softer form which could help to some extent. Coordination could concern implementation, collective rules and mutual information or exchange initiatives like open shops for foreign firms. These at least reduce the transactions costs for services firms to establish themselves abroad. It gives Member States some leeway to take care of national circumstances while trade hampering effects are reduced. However, leeway also creates the opportunity for Member States to protect their own firms. In general, as is also the experience with applying the principle of mutual recognition to goods trade, there is some uncertainty and even leeway to interpret this principle, its derogations and its implementation. So, there is room for manoeuvre for the Member States and to serve their own interest in that respect. Another argument to plea for centralized policies is lobbying. With respect to lobbying, regulation at a central level is probably in the interest of large firms. This is not the case for SMEs that dominate services industries. They are at a more distanced level from the EU. This suggests that lobbying at national level is probably more effective than at the central level in services. The discussion on SD, however, showed lobbying at all levels (Pelkmans 2007).
11.6 Conclusions This chapter sketches a trade-off between a common internal market policy and national regulatory practices in services. Dekker et al. (2007) conclude that the benefits of further market integration in services can be substantial, but that this requires a shift in policymaking from the national level to the European level with respect to regulatory practices which af-
11 Subsidiarity and the Internal Services Market
187
fects trade in services. Such a shift would not benefit each Member State equally. First of all, the distribution of the benefits is unequal because some countries benefit more from trade and changes in allocative efficiency than others. Second, countries differ in their preferences to regulate services. Over time we could expect that the balance between the benefits of further integrating services markets and less possibilities for tailored national regulation could shift towards integration. There are several arguments for this claim. First of all, the process of splitting production chains continues. Related to this is the persistent declining trend of transport and communication costs. As a consequence, outsourcing of tasks becomes more and more important. Some economists already argued that trade will be characterized by trade in tasks instead of goods. As a consequence, services become more tradable. Second, the trend towards a services economy continues. Services become more and more important for productivity growth and trade in the economy, which could underpin the need for large services markets. These two trends could increase the benefits of integrating services markets in Europe and as such will also illustrate the trade hampering effects of national regulation in services more and more clearly. Another relevant trend is deregulation of product and services markets in general in the Member States in recent decades, albeit at different speeds and different points in time. If this trend continues it could become easier to agree on harmonization or less regulation in services markets, interpreting the actual level of regulation as an expression of preference. Further on, future decisions by the European Court of Justice could stimulate the integration of services markets as in the past cases like Dassonville and Cassis de Dyon have been important breakthroughs for goods market integration in Europe. Most trends suggest a shift to closer integration of national services markets because the benefits increase or the relevance of regulation decreases. Whether this shift is also acceptable given the fact that differences in national circumstances will be less weighted in the policy decisions depends to a large extent on the details of future common (market) policies.
References Breuss F, Badinger H. (2005) The European single market for services in the context of the Lisbon agenda: macroeconomic effects. Wirtschaftsuniversität Wien, Vienna Bruijn R de, Kox H, Lejour A (2008) Economic benefits of an integrated European market for services. Journal of Policy Modeling, forthcoming
188
Arjan Lejour
Conway P, Nicoletti G (2006) Product market regulation in the non-manufacturing Sectors of the OECD countries: measurement and highlights, OECD Economics Department Working Papers 530, Paris Copenhagen Economics (2005a) Economic assessment of the barriers to the internal market in services, commissioned by the European Commission, www.copenhageneconomics.com Copenhagen Economics (2005b) The economic importance of the country of origin principle in the proposed services directive, commissioned by UK presidency of the EU, Copenhagen Dekker P, Van der Horst A, Lejour A, Straathof B, Tammes P, Wennekers C (2007) Market place Europe Fifty years of public opinion and market integration in the European Union. In: European Outlook 5, CPB/SCP Ederveen S, Gelauff G, Pelkmans J (2008) Assessing subsidiarity. In: Gelauff G, Grilo I, Lejour A (eds) Subsidiarity and economic reform in Europe: 19–40, Springer European Commission (2004) Proposal for a directive of the European Parliament and of the Council on services in the internal market, SEC (2004) 21, Brussels European Commission (2006) Amended proposal for a directive of the European Parliament and of the Council on services in the internal market, Document Com(2006) 160 Final, Brussels European Commission (2007) Proposal for a regulation of the European parliament and of the Council laying down procedures relating to the application of certain national technical rules to products lawfully marketed in another Member State and repealing Decision. 3052/95/EC, Communication COM(2007) 36 final, Brussels Gual J (2008) Integrating regulated network markets in Europe. In: Gelauff G, Grilo I, Lejour A (eds) Subsidiarity and economic reform in Europe: 157–175, Springer Ilzkovitz F, Dierx A, Kovacs V, Sousa N (2007) Steps towards a deeper economic integration: the Internal Market in the 21st century. A contribution to the Single Market Review. European Economy 271 Kox H, Lejour A (2006a) The effect of the services directive on intra-EU trade and FDI. Revue Economique, vol 57 no 4: 747–769 Kox, H, Lejour A (2006b) Dynamic effects of European services liberalisation: more to be gained. In Global challenges for Europe, Report by the Secretariat of the Economic council, PART 1, Prime Minister's Office Publications 18/2006: 313–142 Lejour A (2007) The Internal Market for services: patchwork. In: Hamilton D, Quinlan JP (eds) The sleeping Giant, Center for Transatlantic relations, John Hopkins University, Washington Nicolaïdes K, Schmidt S (2007) Mutual recognition on ‘trial’: the long road to services liberalization. Journal of European Public policy vol 14 no 5: 717– 734 Pelkmans J (2006) European integration: methods and economic analysis (third edition), Pearson Education
11 Subsidiarity and the Internal Services Market
189
Pelkmans J (2007) The internal services market between economics and political economy. In EIPASCOPE, autumn, Maastricht, European Institute of Public Administration Vogt, L., 2005, The EU’s Single Market: at your service? Economic Department Working paper ECO/WKP(2005)36, OECD, Paris
12 Agriculture Policy: What Roles for the EU and the Member States?
Harald Grethe
12.1 Introduction
1
Agricultural Policy has been a core element of the European Economic Community since its early beginnings (Tracy 1989, 1997). The assessment of the role for the European Union (EU) and the Member States in agricultural policy in this chapter is based on the principle of subsidiarity as defined in Ederveen et al. (2008). The Treaty of the European Community (1997) states that "the Community shall take action, in accordance with the principle of subsidiarity" only for "areas which do not fall within its [the Community's] exclusive competence" (Art. 5, par. 2). This chapter, however, goes beyond this definition and reviews the role of the EU and the Member States in agricultural policy design and funding independently from the criterion of whether they fall under exclusive responsibility of the EU, such as large parts of agricultural policy in the EU. The basis for the assessment of the role of the EU and the Member States is the functional test for subsidiarity developed in Ederveen et al. (2008). 1
The author is grateful to Alan Matthews, Christian Schleyer, and Harald von Witzke and to two anonymous reviewers for valuable comments on draft versions of the chapter. Furthermore, the chapter has benefited from helpful comments made by Dieter Kirschke, who was a discussant, as well as from participants of the Brown Bag Seminar of the Institute of Agricultural Economics and Social Sciences at Humboldt University of Berlin. All remaining deficiencies of the chapter, of course, are the responsibility of the author.
192
Harald Grethe
The chapter is organized as follows: An overview of the Common Agricultural Policy (CAP) of the EU is given in Section 12.2, and costs and benefits of centralization are reviewed in Section 12.3. Section 12.4 takes a political economy perspective and discusses why today's policies look different from the ideal from a subsidiarity point of view, which would prevail if the EU were governed by a benevolent central planner with no other interest than to optimize social welfare in the EU. Finally, in the concluding Section 12.5, options are explored for how to move closer to the ideal.
12.2 Overview of EU agricultural policy The strong role of the CAP in the EU is reflected in its high share in the total EU budget as depicted in Figure 12.1. 120
100
2005 billion euro
80
60
40
20
0 1991
1992
1993
1994
1995
1996
1997
EU budget
1998
1999
2000
2001
2002
2003
2004
2005
CAP budget
Fig. 12.1. Total and agricultural EU budget expenditure, 1991 to 2005 (in €)
The EU budget has increased over the period 1991-2005 by about 50% in real terms. Over the same period, the CAP budget increased by only 23% but still had a share of almost 50% in the EU budget in recent years. Relative to the Gross National Income (GNI) of the EU, the total budget has declined slightly from 1.1% to 1.0%, and the CAP budget has declined from 0.7% to 0.5% (European Communities 2006). This reflects the eco-
12 Agriculture Policy
193
nomic growth of the European Union, including successive enlargements from 12 Member States in 1991 to 25 Member States in 2005. Although total agricultural expenditure of the EU is rather stable, its composition has changed tremendously in recent decades. In the first decades of the CAP, its main aims were an increase in production in order to provide food security for a growing population and income support to an agricultural sector that was shrinking in relative terms in the process of economic development. This aim was mainly pursued by a highly protectionist system of price support through high tariffs and domestic measures such as intervention price systems. During the 1980s the Community turned into a net exporter for most agricultural products and increasingly relied on export subsidies, which brought growing concerns from trading partners. In the context of increasing external pressure on the CAP during the Uruguay Round of multilateral trade negotiations in the GATT, the EU substantially reformed the CAP by reducing support prices for cereals and beef and introducing a system of compensatory direct payments from 1993 onward (McSharry Reform 1992). This principle was continued by further reduction of support prices and the extension of direct payments as part of the Agenda 2000 reforms agreed upon in 1999. Direct payments were increasingly decoupled from actual production, and this principle has culminated in the 2003 Reform of the CAP under which about 90% of current area payments in the EU-15 are part of the Single Farm Payment (SFP) and, thus, are not linked to the production of any specific product. For eight Member States out of the EU-10, direct payments are fully decoupled from production under the Simplified Area Payment Scheme (SAPS). Along with the increasing reliance on direct payments since the McSharry reform came an increase of the budget for rural development measures, which include heterogeneous policies (such as agrienvironmental programs, investment subsidies, and early retirement programs) and which are co-financed by the EU Member States. The main changes in the CAP are reflected in the composition of the CAP budget as shown in Figure 12.2. The decreasing relevance of market intervention is reflected in the declining shares of expenditures for export subsidies and the category "other policies," which mainly includes market interventions. Together these policies accounted for almost 80% of the CAP budget in 1991 and declined to less than 20% in 2005. Over the same period, the share of expenditures for direct payments increased from 15% to 63% while increasing from 6% to almost 20% for rural development measures. However, the increase in the share of rural development funding in the CAP recently came to a halt when the Council decided to set expenditures substantially below
194
Harald Grethe
the Commission's proposal for the financial perspective 2007-2013. As a result, future rural development funding will be substantially below the 2006 level (Agra Informa 2006a). 100%
80%
60%
40%
20%
0% 1991
1992
1993
1994
1995
1996
export subsidies
1997
1998
direct payments
1999
2000
rural development
2001
2002
2003
2004
2005
other
Fig. 12.2. Composition of the EU CAP budget, 1991 to 2005 Sources: European Commission (2006a; various issues); European Communities (2006); OECD (2006); own calculations.2
It is not only in terms of budgetary expenditures that agricultural policy in the EU displays a high degree of EU involvement: the CAP is also noteworthy with respect to the regulatory density. Alesina et al. (2002) calculate the shares of various policy domains in total EU legislation (Regulations, Directives, and Decisions) for five-year periods from 1971-2000. EU agriculture and fishery policies account for 37% to 57% of EU legislation during these periods. Much of this legislation refers to the definition of common product standards, such as food safety standards, and common standards of agricultural production, such as agro-environmental or animal welfare standards.
2
Rural development includes expenditures from the Guidance as well as the Guarantee section of the EAGGF. Direct payments before 2000 are extracted from OECD (2006) due to a change in the data classification in published EU budget figures.
12 Agriculture Policy
195
The CAP is subject to an ongoing reform process. As part of the initiative to simplify the CAP, some technical simplification is envisaged (European Commission 2006b). But the "Health Check" in 2008 foresees a thorough general review of the CAP, including a view to the CAP design after the period of the financial perspective 2007-2013. In addition, a fundamental review of the composition of the EU budget is envisaged in the "Budget Review" in 2008-2009. Main drivers for the development of the CAP are budgetary pressures, international trade negotiations, and changing expectations of consumers and taxpayers.
12.3 Costs and benefits of centralization 12.3.1 Market policies Market policies are almost exclusively designed and financed at the EU level with their implementation, for the most part, delegated to the Member States. This is the case for the most important policy component, external protection by tariffs, as well as for export subsidies and domestic policies, such as intervention prices (cereals, dairy products), storage subsidies (sugar, dairy products), product subsidies (milk powder for feed use, dairy products), and supply control measures (set aside and production quotas for milk and sugar). The assignment of the responsibility for market policies to the EU level is the logical implication of the common market. And this common market is not questioned - it allows for gains from trade due to resource allocation according to comparative advantage and positive effects of scale. In a common market, assignment of the responsibility for market policies to Member States would be, in part, technically infeasible (e.g., differing intervention prices among Member States), and would partly distort competition among Member States (e.g., product subsidies). In addition, delegation of implementation to the EU level would result in considerable transaction costs due to the requirement of a parallel administration, which holds in general for the implementation of agricultural policy, residing for the most part with the Member States. Thus, the current system of responsibilities for policy design and funding at the EU level and implementation at the Member State level is appropriate. What may be questioned, however, is the high complexity of a system that evolved in a period in which market price support was a much more important policy component. This complexity involves transaction costs in design and administration as well as private sector actors, who
196
Harald Grethe
must cope with the system. In light of the declining economic relevance of market price support, simplifications as envisaged under the current EU action plan (European Commission 2006b) and beyond are desirable. Candidates for simplification are, among others, the import regime for NonAnnex I products, the ever increasing complexity of the EU system of agricultural trade preferences (Grethe 2005a), as well as the complex but partially redundant EU import system for fruits and vegetables (Götz and Grethe 2007). Finally, many product and production standards are set at the EU level in order smooth the functioning of the single market. For product standards, such as food safety standards, allocation of responsibility at the EU level is unambiguous: the single market makes different national product standards impractical, as the free movement of goods would not allow the market to ensure a higher standard in any Member State than the standard prevailing in the Member State with the lowest level. In some cases, Member States handle product standards differently. For example, EU importers of organic products from non-EU countries report that compliance with the EU standard for organic products is less strictly monitored in certain Member States than in others. This makes no sense as the imported products move freely in the EU, and the role of the EU in ensuring uniform implementation of the import standards should be strengthened. For process standards, the assessment is more complex. Some standards refer specifically to the quality of local environmental goods, such as standards with respect to soil quality. In such cases, there is no reason to apply standards at the EU level. But some standards refer to the protection of environmental goods with a transboundary character, such as the protection of surface waters, or are based on ethical values that are decided upon at the EU level, such as animal welfare standards. In such cases, the EU level can be the right level for setting the standard in order to ensure that objectives are reached in all Member States and to avoid the distortion of competition due to unequal standards among Member States. 12.3.2 Direct payments The historical origins of direct payments are the market policies through which the EU granted support to producers in the past. Direct payments have been granted as compensation for the reduction of institutional prices. The fact that this compensation was designed and financed at the EU level can be explained by the fact that its predecessor, the price support system, as well as its dismantling were designed at the EU level. There were also good reasons to design and finance direct payments at the EU level from
12 Agriculture Policy
197
an economic point of view: early direct payments were much more coupled to production than they are today, and, thus, had clear externalities they were production distorting. The situation has changed. For the EU-27, on average, direct payments are already decoupled from production by about 90%. In addition, their full decoupling is currently discussed (Fischer Boel 2006) and would involve advantages, such as allowing elimination of the blue box in the WTO. This also would allow abolition of the obligatory set aside, for which the main rationale is to fulfil the blue box criterion that direct payments are only eligible for this exemption if they are embedded in production-limiting programmes. Furthermore, full decoupling would do away with the production distorting effect of the differing degrees of decoupling among EU Member States. Thus, the economic nature of direct payments has changed fundamentally: they have changed from a product(ion) subsidy to a sectoral and personal income policy.3 But sectoral and personal income transfer policies, such as income tax systems, social security systems, or sectoral subsidies (e.g., for coal mining), are generally designed and financed at the Member State level, not at the EU level. There are good reasons for this. Preferences with regard to sectoral and personal income distribution differ widely among Member States. As there are neither externalities nor economies of scale involved in granting fully decoupled direct payments to farmers, the assignment of responsibility for direct payments and their funding at the EU level is an historical artefact that is in clear conflict with the subsidiarity principle.
3
Sometimes, direct payments are seen as a lump sum compensation for the multifunctional non-market outputs of agriculture in the EU that would be implied by the cross-compliance mechanism. This, however, is not convincing as the level of direct payments is in no way linked to the level of output of public goods but rather based on historical production and price support. In the case of the historic system being applied for the SFP, and in the case of trade of entitlements to the SFP being allowed, payments often differ wildly among farmers providing a similar level of public goods. Furthermore, the level of direct payments is too high to be justified by the production of public goods and tends to capitalise in land prices to a large extent (Duvivier et al. 2005; Kirwan 2004; Brümmer and Loy 2001; Chatzis 1997).
198
Harald Grethe
12.3.3 Second pillar policies 12.3.3.1 General considerations
The second pillar of the CAP consists of rural development policies as laid down in EU Regulation 1257/1999. These comprise a heterogeneous menu of policies, such as agri-environmental policies, support for less favoured areas, investment subsidies, training programs, and others. Regions can choose freely from this menu, yet, agri-environmental measures are compulsory. To receive funds, regions must prepare rural development plans that are to be appraised and approved by the European Commission. Rural development policies are financed from the EAGGF Guarantee and Guidance sections and are subject to minimum national co-financing rates of 25% in Objective 1 regions, 20% in the new Member States, and 50% in all other regions. On average for the EU-15, the national rate of cofinancing was 55% in 2005 (European Commission 2006a). Thus, policy design and funding is a shared responsibility of the EU and the Member States/regions, whereas the responsibility for implementation is with the regions. From 2007 on, rural development policies will be governed by the EU Regulation 1698/2005/EC. Funding will be summarized in one instrument, the new European Agricultural Fund for Rural Development (EAFRD). In addition, the conceptual and administrative linkages between the European Commission, the Member States, and the regions will become closer: the Commission has established strategic guidelines that should be the basis for national strategy plans for rural development. Once these national strategy plans are approved by the Commission, regions, as under the former system, will submit their rural development plans that must be in line with the national strategy plans. Most of the rural development measures are of a rather local nature regarding their effects and the problems they address. Furthermore, preferences for the effects of such measures differ widely among Member States, especially in an enlarged and increasingly heterogeneous EU. It is therefore difficult to see why the responsibility for design and funding of such rural development measures should be at the EU level. For example, this holds for the support of less favoured areas, early retirement, young farmer programmes, training and the improvement of rural infrastructure, as well as investment subsidies. Further, public funding of many of these measures - such as investment subsidies - is questionable. The Advisory Board at the German Ministry of Consumer Protection, Food, and Agriculture concludes that rural development policy should focus exclusively on the pro-
12 Agriculture Policy
199
vision of public goods, which would be the only valid long-term motive to justify budget transfers to farmers (Wissenschaftlicher Beirat 2005). For environmental policies that are part of the second pillar of the CAP, the situation is different. These policies mainly aim at the provision of public goods and have, at least partially, transboundary spillovers. Because of the heterogeneity of other rural development policies, as well as their generally questionable motivation, the next section focuses on environmental policies. 12.3.3.2 Environmental policies
Environmental policies4 are the most important single policy group in the second pillar: agri-environmental measures accounted for 45% of rural development spending in the EU-15 from the Guarantee section of the EAGGF in the period 2000-2005 (European Commission 2006a). Most of the environmental policies under the second pillar of the CAP address predominantly local or regional environmental problems and have, by and large, local or regional effects. This holds for policies such as those addressing the quality of ground water, the shape of landscape, and soil erosion (Rudloff 2002: 243; Döring 1997: 244-257). In addition, local preferences for these public goods as well as the costs of providing them may vary substantially among regions. This would suggest attributing responsibility for their design, funding, and implementation to the Member States rather than to the EU. In most cases, regional or local administrations may be even more appropriate. More decentralized and flexible concepts of spatial administration are suggested in order to address the specific spatial extent of the respective public goods, such as Frey's concept of 4
Environmental policies discussed here are those that are part of the EU rural development policy. From a farmer's perspective these are voluntary and incentivebased policies. In addition, the agricultural sector is subject to a wide range of environmental command and control policies, such as the Flora-Fauna-Habitats Directive 92/43/EEC, the Wild Birds Directive 79/409/EEC, the Water Framework Directive 00/60/EC, and the Nitrate Directive 91/676/EEC. These policies are discussed here only in case of overlap with rural development measures, such as the compensatory payments for areas that are subject to restrictive environmental regulation as foreseen in Article 16 of Regulation 1257/1999/EC. Analyses of EU environmental policy can be found in Döring (1997) and Jordan and Jeppesen (2000). Holzinger et al. (2006) hint at the turn toward EU "Soft Modes Governance" such as framework regulations. Moss (2004) analyses the implementation of the Water Framework Directive in Germany and describes its importance as a catalyst to improve the cooperation among national, regional, and local authorities in order to set up river basin management systems.
200
Harald Grethe
"Functional, Overlapping, and Competing Jurisdictions" (Frey and Eichenberger 1999). Specific proposals for agri-environmental policies in Germany are made by Urfei (1999), Zeddies et al. (2000), and Hampicke et al. (2000). Petrick (2006) reviews arguments as to why central governments should not directly finance and provide public goods. Arguments include the potential ineffectiveness of central governments in meeting residents' preferences: sorting and voting of residents may result in the provision of public goods that is closer to their true preferences (Tiebout 1956; Dowding et al. 1994). Furthermore, central governments may be less effective than community governance due to information problems, rent-seeking, and lack of accountability (Baland and Platteau 1996; Westholm et al. 1999; OECD 1998). Finally, private initiative may be crowded out by public measures (Andreoni 1993, 2006). Other recent work also highlights the potential for local and private coordination mechanisms, such as environmental cooperatives, to address the provision of public goods in rural areas (Hagedorn et al. 2002; Moseley 2003). In contrast, Flynn (2000) points to the often poor capability of regional or local governance systems to address environmental problems for reasons such as inadequate administrative, human, and financial resources and the danger of regulatory capture. To sum up, empirical evidence suggests that the current system of agrienvironmental policies is too centralized. Therefore, the responsibility for design as well as funding of a large part of today's agri-environmental policies should be attributed to the Member States or lower administrative levels (Ewers and Henrichsmeyer 2000; Wissenschaftlicher Beirat 1998; Petrick 2006; Niedersächsische Regierungskommission 2001). For policy design, a large share of responsibility is already with the regions and Member States that prepare rural development programmes. For funding, however, the EU is heavily involved through substantial co-financing, which hurts the principle of fiscal equivalence. In contrast to the measures above, which predominantly address local and regional environmental goods, some agri-environmental measures address environmental goods that have a transboundary character. These are measures that address global warming, the pollution of surface waters flowing into transboundary rivers, and some measures addressing biodiversity. In addition, some agri-environmental policies may involve economies of scale, such as the establishment of transboundary habitats (e.g., through the Flora-Fauna-Habitats Directive and complementary compensation payments as part of the rural development package) or the provision of resting areas for migratory birds. In such cases, EU coordination of policies, or policy design and funding, can be justified. When voluntary cooperation is impractical or inefficient
12 Agriculture Policy
201
because of economies of scale, responsibility should be assigned to the EU level. The decisive criterion for the EU to distribute funds should be where to get the most environmental goods for the same money. This could be realized, for example, in a tender system under which local or regional administrations or other institutions could apply. The rates of co-financing should not be fixed a priori: a region may have an urgent transboundary agri-environmental problem and be able to address this at relatively low costs but have very limited means for co-financing. Unfortunately, it is not always possible to assign a clearly defined spatial effect to a specific agri-environmental measure. Many measures affect local as well as regional and cross-border or even global environmental goods. Take the example of planting and maintaining a hedgerow. This will have a predominantly local effect on the landscape. If the region’s local recreation or tourism is connected with the hedgerow landscape, this may have regional or even national or international spillovers. Furthermore, the hedgerow may prevent soil erosion that could result in reduced nutrient loads reaching surface waters, which may be transboundary, before finally flowing into the sea. Also, the hedgerow may become a habitat for endangered species and therefore contribute to global biodiversity. The current system of co-financing agri-environmental policies seems to address this complexity as both Member States and the EU contribute to finance measures that aim at local as well as transboundary public goods. This link, however, is weak, as there is no link between the rate of cofinancing and the degree to which measures have transboundary spillovers. Instead, the current system of co-finance has various significant disadvantages. 12.3.3.3 Drawbacks of the co-finance system for the second pillar of the CAP
First, the current system of co-financing the second pillar of the CAP results in policy design that is often determined more by the intention to generate budgetary flow-back from the EU than by preferences of the local or regional residents for environmental goods - a problem described as the "shared lunch" or "common pool budgeting" problem (Weingast et al., 1981: 651; Inman and Rubinveld 2002: 9). Second, funds are not concentrated on spots with the most urgent agrienvironmental problems. Instead, funds are distributed among Member States according to the size of their agricultural sectors, their historical negotiating power, and their capacity and willingness to co-finance EU funds. Within Member States, funds are distributed among various meas-
202
Harald Grethe
ures according to regional preferences and the influence of lobbying groups. No clear criteria exist for the distribution of rural development funds in general; yet, the EU has established agricultural area, employment in agriculture, and GDPPPS per capita as criteria for the allocation of modulation funds, which is "new second pillar money".5 The same criteria were used by the European Commission for the allocation of SAPARD (Special Accession Programme for Agriculture and Rural Development) funds to the Central European accession candidates and for the allocation of rural development funds for the EU-10 as well as Bulgaria and Romania (European Commission, 2002: 5). Summarizing these criteria in order to reach a weighting scheme for the distribution of second pillar funds among Member States could happen according to:
Weight c
0.65 area ratio c / EU 15 0.35 employment ratio c / EU 15 (1 GDPPPS ratio c / EU 15 ) / 2
(1)
with the resulting share in rural development funds for each country (c) being
Sharec
Weightc
¦Weight
(2) c
c
Table 12.1 compares the current distribution of expenditures for rural development from the EAGGF (European Agricultural Guidance and Guarantee Fund) sections Guidance and Guarantee among the EU-15 countries for the period 2000-2005 with a distribution that would result from applying formulas (1) and (2).
5
EC Directive 1782/2003, Official Journal of the European Communities (OJ) L 270, 21.10.2003: Modulation is the reduction of funds for direct payments and the transfer of these funds to the rural development budget. Obligatory modulation was introduced at a rate of 3% in 2005 and is at 5% since 2007. In addition, EU Member States may increase the modulation rate voluntarily.
12 Agriculture Policy
203
Table 12.1. Current distribution of rural development funds in the EU-15 (2000 to 2005) and distribution according to measures promoted by the European Commission Member State
Actual Share as deDeviation Actual share in share fined by (2) (perc. points) agri-env. funds
Share in agr. area
% Belgium 0.7 1.0 -0.3 0.6 Denmark 0.7 1.7 -1.0 0.9 Germany 18.2 13.0 5.2 18.5 Greece 4.8 6.2 -1.4 0.5 Spain 17.1 18.3 -1.2 6.1 France 12.4 19.8 -7.5 10.0 Ireland 5.6 2.5 3.1 7.1 Italy 14.4 13.1 1.3 18.2 Luxembourg 0.2 0.1 0.1 0.3 Netherlands 1.1 2.0 -0.9 0.5 Austria 7.3 2.6 4.7 15.2 Portugal 6.2 6.3 -0.2 3.8 Finland 5.5 1.7 3.7 7.9 Sweden 2.8 2.1 0.7 6.9 UK 3.2 9.6 -6.4 3.4 EU15 100 100 100 Sources: European Commission (2006a, various issues); own calculations.
1.1 2.0 13.1 3.0 19.4 22.6 3.4 11.6 0.1 1.5 2.6 2.9 1.7 2.4 12.6 100
Table 12.1 shows that formulas (1) and (2) serve quite well to explain the distribution of rural development funds; the coefficient of correlation between actual and predicted expenditures is 84%. Notable differences exist, however; for example, Austria received almost three times as much funding as predicted, whereas the UK received only about one-third of predicted funds. Looking at the distribution of funds over different measures, the picture is extremely heterogeneous. For example, Spain, which accounts for almost 20% of the total agricultural area in the EU, received only 6% of EUwide funding for agri-environmental measures; on the other hand, Austria, which accounts for only 2.6% of the EU agricultural area, received more than 15%. This, of course, does not reflect differences in agrienvironmental problems among Member States from an EU perspective; rather, it reflects a difference in national preferences and the success of Austria to attract a disproportional amount of second pillar funds. If, however, the distribution of agri-environmental funding throughout the EU is determined to such a large extent by local preferences, it is unclear why it should be EU co-financed by 50% or more.
204
Harald Grethe
Another drawback of the co-finance system is the resulting ambiguity of the European Commission's motivation in the surveillance of competitiondistorting effects of rural development measures. On the one hand, the Commission is closely involved in the design and funding of rural development programs. On the other hand, the Commission must monitor undue competition-distorting effects of these programs. Agri-environmental measures explicitly allow for compensation payments - exceeding the costs or income foregone involved by a maximum of 20% - to provide an incentive.6 Empirical studies, however, suggest that these maxima are often exceeded and considerable windfall gains for producers exist (Zeddies and Doluschitz 1996; Ahrens et al. 2000; for an overview of the discussion see Osterburg and Stratmann 2002). If a large part of rural development policies was to be overseen exclusively by the Member States/regions, the European Commission could focus more unambiguously on monitoring competition-distorting effects. In conclusion, the current system of co-financing is not adequate for most of the rural development policies. Instead, the clear attribution of responsibility for a large part of these policies to the Member State level and below, with only some responsibilities attributed exclusively to the EU, is more in accordance with the subsidiarity principle. Such a change, however, may require the development of new institutions and mechanisms at local, regional, and national levels to assume control over design as well as funding of policies. In addition, financial flows would need to be adjusted to enable local administrations to finance the provision of public goods within their jurisdictions that were financed previously by the EU (for suggestions in the German context, see Ewringmann and Bergmann 2000).
12.4 Why is the CAP not in accordance with the subsidiarity principle? Political economy considerations and dynamics Today's CAP looks different from the model that would result from the subsidiarity-based perspective discussed above. Specifically, it seems to be overly centralized. Economic efficiency arguments do not support the current allocation of responsibilities. Partially, the current situation can be explained by the historical background of the CAP, as argued above. In addi6
Under the New Rural Development Regulation this provision has been removed and replaced by a reference to allowance for transaction costs, motivated by bringing agri-environmental measures in line with the WTO Green Box provisions (Matthews, 2006).
12 Agriculture Policy
205
tion, political economy arguments can help explain the current design of the CAP and are discussed in this section. 12.4.1 Different interests of Member States Interests of EU Member States concerning the size of the CAP budget and its composition differ widely. These differences result from different endowments and from different national preferences regarding the role of agriculture in the society. Differences in preferences comprise aspects such as the interest to keep agricultural activity in marginal regions and the motivation to move toward efficient farm structures. With respect to endowments, it is mainly the share of the agricultural sector in the economy that determines the national position toward the CAP - it is this share that determines the resulting net budgetary flows. This is because the national contributions to the EU budget are largely determined by the size of the economy as a whole, whereas the receipts resulting from the CAP are related to the size of the agricultural sector. Figure 12.3 displays the average annual net budgetary flow resulting from the CAP for selected Member States over the period 2000-2005. Figure 12.3 shows that Spain, France, and Ireland were substantial net beneficiaries of the CAP, whereas the UK, the Netherlands, Italy, and Germany were net contributors. Figure 12.3 also shows that the CAP is not in conformity with the cohesion objective. Part of the net flows goes from countries with a relatively low GNI per capita such as Italy to wealthier countries such as France and Ireland. Distributional aspects of the CAP have always played a role in reform discussions, an example being that of France's general opposition to the reduction or nationalization of direct payments due, in part, to its strong net beneficiary position resulting from direct payments. In order to assess likely coalitions for a reform of the CAP in the future, Figure 12.4 presents the net budgetary flows projected to result from the CAP in 2012 as a percentage of the GDP per Member State should the British rebate be eliminated.
206
Harald Grethe
3 2 1 UK
Netherlands
Italy
Germany
0 Billion euro
Spain
France
Ireland
Portugal
Sweden
-1 -2 -3 -4 -5
Fig. 12.3. Average annual net budgetary flow resulting from the CAP for selected Member States over the period 2000 to 2005 Sources: Grams (2006); own modifications.7
Figure 12.4 shows that the net budgetary flows resulting from direct payments and rural development measures will be different in 2013 compared with today. Traditional net recipients such as France and Spain will no longer belong to this group but will end up in a balanced situation where receipts from the CAP are similar to their respective contributions. Instead, the new Member States, except Malta and Cyprus, will become significant net beneficiaries of the CAP in 2013 when direct payments are fully phased in for the EU-10 and phased at 70% for Bulgaria and Romania. Among the EU-15 members, only Greece, Ireland, and Portugal will remain substantial net recipients by 0.2% to 0.5% of their domestic GDP. Among the net payers, the UK (in the absence of the rebate), Germany, Italy, and the Netherlands would account for about 90% of net payments.
7
Net budgetary flow calculated as the difference between the contribution to the CAP budget minus the receipts from the CAP budget per Member State.
12 Agriculture Policy
207
3% 2.5 2 1.5 1 0.5 0
rg bu m ds xe lan Lu her et N ium lg Be K U ta al ny M ma er n G de e Sw ly Ita rus yp rk C ma en D tria s Au ce an Fr and nl Fi in a ia S p en ov al Sl tug ic r bl Po nd epu la R I re c h ze C ece re ia G ak ov Sl nia to Es nd la Po ia tv ry La ga un ia H uan th ia Li an om a R ari lg
Bu -0.5
direct payments
rural development measures
Fig. 12.4. Net financial flow resulting from direct payments and rural development measures in 2013 (% of GDP Sources: Direct payments are based on national envelopes in EC Regulation 864/2004 (OJ L206/20-36, 09.06.2004) minus 5% modulation and are adjusted in order to meet commitment appropriations as in European Commission (2006c). Rural development spending is according to European Commission (2006d). Contribution to the EU budget is calculated based on projected shares in GDP (Eurostat 2006; growth rates according to European Commission, 2006c: 3, without British rebate).
The redistributive effects of direct payments and rural development policies alone display a similar pattern in general, although some notable differences exist. France and Denmark, being in a neutral position for the CAP as a whole, are clear net beneficiaries from direct payments but are net payers for rural development measures. For Austria and Finland, the situation is the opposite: they are net beneficiaries from rural development measures but net payers for direct payments. 12.4.2 Common pool budgeting and deference in negotiations Political economy models show that common pool budgeting, which is financed from a highly dispersed tax base but provides public goods with a higher geographical or personal concentration than the tax base, usually
208
Harald Grethe
leads to overspending (Weingast et al. 1981) in the form of "locally beneficial but centrally inefficient government policies" (Inman and Rubinfeld, 2002: 7). For legislators, a tendency for an "I'll-scratch-your-back-if-you-scratchmine" decision rule may result (Inman and Rubinfeld 2002: 9); indeed, decisions in the Council of Agricultural Ministers, which can be described as a repeated game setting, comprise many such examples for what is described as "deference in negotiations" by Ederveen et al. (2008). Yet to overspending, overly detailed systems of coordination that may be at odds with the proportionality principle may result (Ederveen et al. 2007), and the complexity of rural development regulation may serve as an example. 12.4.3 Leviathan and lobby Departing from the assumption of benevolent governments, Pelkmans (2006) distinguishes two cases of government failure: Leviathan and lobby. One may ask, therefore, whether the overly centralized process of the CAP is characterized by weaker public control and allows governments to act more "Leviathan" in pursuing their own rather than the public interest compared to a more decentralized setting. There is little evidence for such a systematic bias. The decision-making process in the Council of Ministers is probably under more public scrutiny than many of the national decisions. In contrast, some authors highlight the relevance of the European coordination process for transparency of the resulting policies, for example, in case of rural development programs (Osterburg and Stratmann 2002: 276). One may also ask whether agricultural policy design at the EU level is more vulnerable to hidden lobbying than at the level of national governments. Tabellini and Wyplosz (2006) argue that centralized policies may be more prone to capture if the foreign and the domestic lobby have the same interests and pull in the same direction. This may be the case with the determination of the overall level of the CAP budget as well as the protection level, which farm lobbies in the EU generally like to see high. With respect to specific policy design, however, interests are highly divergent among farm lobbies in the Member States, and there is little evidence for more successful lobbying at the EU level than in the case of national policy design. In contrast, the fact that the European Commission is not subject to the usual election cycle, and election cycles of national governments are not synchronized, makes lobbying more difficult and results in a rather stable development of agricultural policy. It is especially the gradual and unbro-
12 Agriculture Policy
209
ken transition process from price support policies in the early 1990s toward increasingly decoupled direct payments that may be attributed to the centralized policy process. It is doubtful that agricultural policy would be more liberalized today if Member States had been fully responsible. The EU process often allows agricultural ministers to return home blaming the chosen reforms on the EU as a scapegoat, making the ministers more immune to lobbying. The instability of United States agricultural policy over the same period may be an example of a process more prone to the influence of lobby groups. 12.4.4 Role of the EU Commission The European Commission may have an interest in centralising agricultural policy because of self-selection of EU officials who are prointegration, socialization of its staff during working life, or the administration's interest of expanding or maintaining power (Ederveen et al. 2008). Indeed, it is interesting to note that, in contrast to the trailblazing role of the Commission in shifting from price support to decoupled direct payments and rural development funding, there is silence about any kind of nationalization of core policy fields. One may ask why there is no analysis from the Commission on the appropriate level of centralization of the second pillar while many researchers argue for greater decentralization. Further, why is there no discussion from the European Commission about fully decoupled payments being nationalized? Pollack (2003: 51) provides a list of criteria in order to test whether there is any discretion of the European Commission to promote overly centralist policies. Three of these criteria support the strong potential for the Commission to influence the degree of centralism in agricultural policy: x Information at the Member State level is imperfect, as it is mainly at the EU level where information from all Member States is collected and evaluated systematically. x The uncertainty about future developments is high. The CAP is subject to continuous reform, and there is constant external pressure from international trade integration. x Transaction costs of negotiating alternative policies are high because the policy system is complex and national positions are highly divergent. On the other hand, many potential changes in the CAP have strong distributional consequences among Member States; these consequences contradict the high discretionary power of the Commission because the distribution of power and preferences among Member States dominates the
210
Harald Grethe
decision process.8 The Commission, however, has been very successful in dealing with this problem in several ways, notably through: x The establishment of compensatory payments in case of significant changes in market policies, mainly in order to avoid redistribution of benefits. x The maintenance of financial benefits within Member States in case of reforms through policies such as modulation or decoupled payments still being based on historical endowments. x Balancing of benefits and costs within the CAP. x Balancing with other policy fields.
12.5 Conclusions on agricultural policy - How to get closer to the ideal? The strong involvement of the EU in agricultural policy design and funding is not in accordance with the subsidiarity principle. The evolution of the CAP has resulted in an historical artefact, and the challenge now is to shrink its coverage, administration, and budget. EU involvement should be limited to market policies and agri-environmental policies that address cross-border externalities. For other policies, only competition-distorting effects need to be monitored at the EU level. Furthermore, a uniform implementation of product and product-related process standards should be strengthened. The existence and persistence of EU responsibility for design and funding of direct payments and most of the rural development policies can be explained historically and from a political economy perspective, but they do not pass the functional subsidiarity test. The "health check" in 2008 and the budget review in 2008-2009 are a chance to further develop the CAP in order to better meet efficiency and equity objectives, to address current imbalances in the EU budget, and to bring the CAP more in line with the subsidiarity principle. But how is it possible to get there? Most of the conclusions drawn here are not new (Sapir et al. 2003; Wissenschaftlicher Beirat 1998; Niedersächsische Regierungskommission Padoa-Schioppa et al. 1987). To quote only two sources, Sapir et al. conclude that "there is…a solid argument for decentralising to Member States the distributive function of the Common Agricultural Policy" (2003: 164). Padoa-Schioppa et al. state that the distribu8
A literature review on the relationship between the Member States and the European Commission can be found in Kassim and Menon (2004), who observe a strengthening of the Member States since the 1990s.
12 Agriculture Policy
211
tive character of the CAP "represents a systematic anomaly, since the Community…is not well suited to executing distributive policies at the level of individual persons or small enterprises. Efficient income distribution requires detailed administration at the level of the individual, and coherence with features of income tax and social security systems, and the Community cannot assure this. The Community has thus switched roles with the Member States, counter to the basic principles of subsidiarity and comparative advantage" (1987: 102-103). The policy process, however, shows different dynamics. Three aspects seem especially important when envisaging a fundamental redistribution of responsibilities in designing and financing agricultural policy in the EU. First, changes can only be established gradually. It may be more important to establish principles that point in the right direction than to get results quickly. Second, it must be accepted that trade-offs have to be sought in order to limit the redistributive effects and to encourage buy-in from potential losers. Temporary compensation as well as trade-offs with policies other than the CAP may help. Third, the self-maintaining tendency of centralistic institutions must be taken into account when looking at the potential drivers for reform. It is unlikely that the European Commission would initiate a reduction of the scope of the CAP and its budget. In the remainder of this section, potential tracks for reform are explored for the policy areas of direct payments and second pillar policies, keeping these aspects in mind. For direct payments, the full decoupling, which is currently discussed in the European Commission (Fischer Boel 2006), would be helpful for starting to think about nationalization for two reasons. First, full decoupling allows for different levels of direct payments from a competitive point of view, as fully decoupled direct payments have (almost) no effects on production. Second, full decoupling adds to the transparency of the largely missing justification of these payments - and thus adds to public and political pressure to reduce them. Direct payments tend to be successively reduced, and this also helps open discussion of options for their nationalization. After all, the less their volume, the less the redistributive effects of their nationalization. As direct payments are fixed in nominal terms, inflation leads to an annual reduction in real terms of approximately 1.9%. In addition, chances are that the budget ceiling for the first pillar of the CAP will require the reduction of direct payments from 2008 onwards. Current estimates hint at a reduction by 5% in 2008 and an additional 8% by 2013 (Agra Informa 2006b). Together with an increase of obligatory modulation from 5% to 10%, which is also discussed by the European Commission (Agra Informa 2006c), the real value of direct payments could be reduced by more than 28% in 2013.
212
Harald Grethe
The pace of future reductions, however, is subject to many uncertainties and is rather erratic. A more reliable and transparent phasing out process would be much better and would enable producers to prepare for an environment without such payments. There is a strong interdependency between nationalization and the reduction of direct payments. From both efficiency and equity perspectives, the first best option would be to completely phase out direct payments in their current form. They capitalize into land prices and thus inhibit structural change. Furthermore, they contribute to income inequality in the agricultural sector, as they mainly accrue to large farmers (von Witzke and Noleppa 2006). Also their distribution among Member States is not conforming with the cohesion objective, as part of the net flows go in the direction of countries with relatively higher per capita income (e.g., Italy is a net contributor, whereas France is a net beneficiary of the CAP). If it were possible to phase out direct payments at the EU level over a reasonable period, say 10 years, there would be no need for policy nationalization. If this turns out to be politically infeasible, nationalization may be an option, at the very least, to increase the pace of elimination in many Member States due to budgetary rivalry at the Member State level. Searching for a potential majority in the Council for a reduction or nationalization of direct payments, it is clear that this contradicts the distributional interests of the new Member States, except Malta and Cyprus (see Figure 12.4 above). Traditionally, the transfers involved in the CAP were often an incentive for the relatively poor Southern and Eastern European countries to join the EU - and it is likely that they would oppose the erosion of these benefits. The new Member States, however, will also be the main recipients of structural and cohesion funds, and potential trade-offs could be sought in this area. Also, some trade-offs potentially would have to be sought with Ireland and Greece, which would be negatively affected from a distributional point of view. Turning toward the second pillar of the CAP, the justification of responsibility at the EU level for rural development instruments should be critically reviewed. Except for transitional measures to increase the competitiveness of newly acceding members, it is difficult to see any justification except for environmental measures. But also for those, responsibilities and especially funding should be disentangled to a large extent and assigned to the Member States or the EU along the lines proposed above. It is a relatively small share of today's rural development spending for which a fixed rate of co-financing potentially would be the adequate instrument. For example, because costs at the Member State level can be clearly linked to EU regulation, and benefits are to a significant extent transboundary and local in nature (e.g., costs or revenue foregone resulting from restrictions at
12 Agriculture Policy
213
Natura 2000 sites or from implementation of the Water Framework Directive). How is it possible to start dismantling the system of co-financing? A modified modulation system may be an option to do so. Modulation in its current form is a problematic instrument. The fixed link between the reduction of direct payments and the increase of rural development funds is artificial and is mainly motivated by avoiding distributional effects. Furthermore, the resulting increase of the total volume of agricultural support because of the additional co-finance funds from the Member States is questionable. Instead, a modified modulation mechanism would offer the opportunity to establish a first step out of a fully co-financed rural development policy. Money could be saved in the first pillar of the CAP, and, instead of shifting it to the traditional co-financed funds, part of the money could be transferred to the exclusive responsibility of the European Union for agri-environmental policies described above that address issues of a transnational character and would not necessarily require co-funding from the Member States. The other part of savings could be directly transferred to the Member States (or converted into a reduction of the EU budget, or used for alternative purposes), and it would be the Member States' responsibility whether the money should be spent on rural development and for which measures. For the funding currently allocated to the second pillar of the CAP, such a new type of modulation could be, in the long term, an example for the split of funds into those under exclusive EU-responsibility and those under exclusive Member State responsibility. Funds which would be co-financed by the EU and Member States could be limited to those agri-environmental measures, which are clearly linked to EU regulation. What should be sought, in addition, are mechanisms that would allow reducing the current budget for co-financed rural development funding. Such an initiative, however, realistically would come from the Member States rather than from the European Commission. One supporting element could be a clearly defined, transparent, and objectively rural development motivated key for the distribution of funds among Member States. Such a key could reduce the potential for hidden trade-offs, which are predominantly motivated by distributional motives rather than rural development considerations. In addition, current rural development funding could be shifted to a larger extent to newly acceding Member States with technological and structural gaps (e.g., by increasing the weight of the per capita GDP factor in the formula presented above), and part of funding could be limited to a transition period and would be phased out over a period of 10 years, for example. This would also serve as an incentive for national policy makers in the new Member States to concentrate money on measures
214
Harald Grethe
that make their agricultural sectors independent from such support instead of distributing rents. In addition, this would create a large and powerful coalition of old Member States that would be net payers for such a policy and would, thus, support the limitation of its time horizon. The money saved could then shift to newly acceding countries, or, alternatively, could move into the splitting mechanism discussed above, or be spent on other policy areas that may evolve in the future development of the EU, such as a common foreign policy or a common defence policy.
References Agra Informa (2006a) Rural development suffers as traditional CAP lives on. Agra Europe Weekly 2214, June 30, Analysis: 1–2 Agra Informa (2006b) Balkan enlargement will put CAP under further strain. Agra Europe Weekly, 2227, September 29, Analysis: 1–2 Agra Informa (2006c) Towards 2013 commission paves the way for future CAP reform. Agra Europe Weekly, 2228, October 6, Analysis: 1–2 Ahrens H, Lippert C, Rittershofer M (2000) Überlegungen zu Umwelt- und Einkommenswirkungen von Agrarumweltprogrammen nach VO (EWG) Nr. 2078/92 in der Landwirtschaft. Agrarwirtschaft 49: 99–115 Alesina A, Angeloni I, Schuknecht L (2002) What does the European Union do? European University Institute Working Paper, Robert Schuman Centre for Advanced Studies, RSC no 2002/61 Andreoni J (1993) An experimental test of the public goods crowding-out hypothesis. American Economic Review 83: 1317–1327 Andreoni J (2006) Philanthropy. In: Gerard-Varet L-A, Kolm S-C, Ythier JM (eds) Handbook of giving, reciprocity and altruism, Amsterdam, Elsevier Baland J-M, Platteau J-P (1996) Halting degradation of natural resources, is there a role for rural communities? Oxford, Clarendon Press. Brümmer B Loy J-P (2001) Der Einfluss staatlicher Ausgleichszahlungen auf Landpreise in Schleswig-Holstein. In: Brockmeier M, Cramon-Taubadel S von, Isermeyer V (eds) Liberalisierung des Weltagrarhandels - Strategien und Konsequenzen, Schriftenreihe der GeWiSoLa 37: 389-399 Chatzis A (1997), Flächenbezogene Ausgleichszahlungen der EU-Agrarreform – Pachtmarktwirkungen und Überwälzungseffekte, Bergen, Agrimedia Döring T (1997) Subsidiarität und Umweltpolitik in der Europäischen Union. Ökologie und Wirtschaftsforschung 25, Dissertation, Marburg, MetropolisVerlag Dowding K, John P, Biggs S (1994) Tiebout: A Survey of the Empirical Literature. In: Urban Studies 31: 767–797 Duvivier R, Gaspart F, Henry de Frahan B (2005) A panel data analysis of the determinants of farmland price: an application to the effects of the 1992 CAP re-
12 Agriculture Policy
215
form in Belgium, Paper presented at the XIth EAAE Congress, Copenhagen, Denmark, August 23–27 Ederveen S, Gelauff G, Pelkmans J (2007) Assessing subsidiarity. In Gelauff G, Grilo I, Lejour A (eds) Subsidiarity and economic reform in Europe: 19–40, Springer European Commission (2001) European governance, a white paper, COM (2001) 428, July 27, Brussels European Commission (2002) EU Agriculture and enlargement, fact sheet, Brussels, http://europa.eu.int/comm/agriculture/publi/fact/enlarge/2002_en.pdf. European Commission (2006a) Rural development in the European Union. Statistical and economics information, Report 2006, August, Brussels European Commission (2006b) Simplification of the common agricultural policy action plan, DG agriculture and rural development. Working Paper, Brussels, October European Commission (2006c), Communication from the Commission to the European Parliament and the Council, technical adjustment of the financial framework for 2007 in line with movement in GNI and prices. COM(2006) 327 final, June 22, Brussels European Commission (2006d) EU Support for rural development 2007-2013, Brussels, http://ec.europa.eu/budget/library/documents/multiannual_framework/2007_2 013/tab_rural_devt_2007-2013_en.pdf. European Commission (various issues) The agricultural situation in the European Union, Brussels European Communities (2006) European Union Financial Report 2005, Luxembourg European Communities (various issues) Official Journal Eurostat (2006) Eurostat Database, http://epp.eurostat.ec.europa.eu, Accessed at October 10, 2006 Ewers H-J, Henrichsmeyer W (2000) (eds) (2000) Agrarumweltpolitik nach dem Subsidiaritätsprinzip. In: Denkschrift des Schwäbisch Haller Agrarkolloquiums der Robert Bosch Stiftung, Berlin, Analytica Ewringmann D, Bergmann E (2000) Agrarumweltpolitik nach dem Subsidiaritätsprinzip: Möglichkeiten und Grenzen einer Funktionalisierung des Finanzausgleichs für eine dezentrale Agrarumweltpolitik. Schriften zur Agrarfor-schung und Agrarpolitik, 6, Berlin, Analytica Fischer Boel M (2006) European model of agriculture. Speech at the National Parliaments Conference - European Model of Agriculture, Helsinki, October 12 Flynn B (2000) Is local truly better? Some reflections on sharing environmental policy between local governments and the EU. European Environment 10: 75–84 Fouarge D (2002) Minimum protection and poverty in Europe, An economic analysis of the subsidiarity principle within EU social policy, PhD Thesis, Amsterdam, Thela Thesis
216
Harald Grethe
Frey BS, Eichenberger R (1999) The new democratic federalism for Europe, Functional, overlapping and competing jurisdictions, Cheltenham, UK, Edward Elgar Götz L, Grethe H (2007) The EU import regime for oranges – much ado about nothing? Journal of International Agricultural Trade and Development. vol 1 no 3 Grams M (2006) Finanzierung der Europäischen Union und der Gemeinsamen Agrarpolitik. Presentation at the Humboldt University of Berlin, November 1, Berlin Grethe H (2005a) The perspective of agriculture trade preferences granted by the EU to developing countries. International Agricultural Trade Research Consortium 2005 Summer Symposium, "Pressures for Agricultural Policy Reform: WTO Panels and the Doha Round Negotiations", Seville, Spain, June 19–21 (CD) Grethe H (2005b) Turkey's accession to the EU: what will the common agricultural policy cost? German Journal of Agricultural Economics (Agrarwirtschaft) 2: 128–137 Hagedorn K, Arzt K,. Peters U (2002) Institutional arrangements for environmental cooperatives: a conceptual framework. In: Hagedorn K (ed) Environmental cooperation and institutional change: theories and policies for European agriculture, Cheltenham, UK, Edward Elgar: 3–25 Hampicke U, Müller K, Meyer-Aurich A, Kachel K-U (2000) Agrarumweltpolitik nach dem Subsidiaritätsprinzip: Fallstudie Brandenburg/Uckermark. In: Schriften zur Agrarforschung und Agrarpolitik, 4, Berlin, Analytica Holzinger K, Knill C, Lenschow A (2006) A turn toward soft modes governance in EU environmental policy – viewed from a policy instruments perspective. Paper prepared for the International Conference "Governance and Policy Making in the European Union", Osnabrück, November 2–4 Inman R, Rubinfeld D (2002) Subsidiarity, governance and EU economic policy. CESifoForum, 4/2002, Munich Jordan A, Jeppesen T (2000) EU environmental policy: adapting to the principle of subsidiarity? European Environment 10: 64–74 Kassim H, Menon A (2004) European integration since the 1990s: Member States and the European Commission. Centre for European Studies, University of Oslo Working Paper WP 6/04 Kirwan BE (2004) The incidence of U.S. agricultural subsidies on farmland rental rates. Department of Policy Analysis and Management, Cornell University, Working Paper http://www.human.cornell.edu/pam/vitae/kirwanrentalrates.pdf. Matthews A (2006) Decoupling and the Green Box: international dimensions of the reinstrumentation of agricultural support. Paper presented to the 93rd EAAE seminar "Impacts of Decoupling and Cross Compliance on Agriculture in the Enlarged EU", September 22–23, 2006, Prague Moseley MJ (ed) (2003) Local partnerships for rural development, The European Experience, Wallingford, CABI Publishing
12 Agriculture Policy
217
Moss T (2004) The governance of land use in river basins: prospects for overcoming problems of institutional interplay with the EU water framework directive. Land use policy 21: 85–94 Musgrave RA (1959) The theory of public finance, New York, McGraw-Hill Niedersächsische Regierungskommission "Zukunft der Landwirtschaft – Verbraucher-orientierung" (2001) Endbericht, Hannover Oates WE (1972) Fiscal federalism, New York, Harcourt Brace & Jovanovich OECD (1998) Cooperative approaches to sustainable agriculture, Paris, OECD OECD (2006) 1986–2005 PSE, Paris, OECD Olson M Jr (1969) The principle of “Fiscal Equivalence”: The division of responsibilities among different levels of government. The American Economic Review 59: 479–487 Osterburg B and U Stratmann (2002) Die regionale Agrarumweltpolitik in Deutschland unter dem Einfluß der Förderangebote der Europäischen Union. Agrarwirtschaft vol 51 no 5: 259–279 Padoa-Schioppa T, Emerson M, King M, Milleron JC, Paelinck JHP, Papademos LD, Pastor A, Scharpf FW (1987) Efficiency, stability and equity: a strategy for the evolution of the economic system of the European Community, Oxford, Oxford University Press Pelkmans J (2006) Testing for subsidiarity. BEEP Briefings, 13, College of Europe Petrick M (2006) Why and how should the government finance public goods in rural areas? A review of arguments. Contributed paper at the Gewisola Conference 2006 Pollack MA (2003) The new institutionalisms and European integration In: Wiener A, Diez T (eds) European integration theory, Oxford, Oxford University Press: 137–156 Rudloff B (2002) Agrarumweltpolitik nach dem Subsidiaritätsprinzip: Föderalismustheoretische Grundlagen zur Politikgestaltung. Agrarwirtschaft 51: 239– 248 Sapir A, Aghion P, Bertola G, Hellwig M, Pisani-Ferry J, Rosati D, Viñals J, Wallace H (2003) An agenda for a growing Europe: making the EU economic system deliver. Report of an independent High Level Group established at the initiative of the President of the European Commission, Brussels Tabellini G, Wyplosz C(2006) Supply-side policy coordination in the European Union. Swedish Economic Policy Review 13: 101–156 Tiebout CM (1956) A pure theory of local expenditures. Journal of Political Economy 64: 416–424 Tracy M (1989) Government and agriculture in Western Europe 1880–1988, New York, Harvester Wheatsheaf Tracy M (1997) Agricultural policy in the European Union and other market economies, Brussels, APS – Agricultural Policy Studies, Belgium in association with AGRA FOCUS Urfei G (1999) Agrarumweltpolitik nach den Prinzipien der Ökonomischen Theorie des Föderalismus: Ein Regionalisierungsansatz zur territorialen Abgrenzung effektiver Politikaktionsräume. Schriftenreihe des Rheinisch-
218
Harald Grethe
Westfälischen Instituts für Wirtschaftsforschung, Neue Folge (66), Berlin, Duncker & Humblot Von Witzke H, Noleppa S (2006) Distributive effects of the direct payments in German agriculture under the "New Common Agricultural Policy" of the European Union. A Report to the German Marshall Fund of the United States, Berlin Weingast BR (2005) The performance and stability of federalism: an institutional perspective. In: Ménard C, Shirley MM (eds) Handbook of New Institutional Economics, Dordrecht, Springer: 149–172 Weingast BR, Shepsle KA, Johnsen C (1981) The political economy of benefits and costs: a neoclassical approach to distributive politics. Journal of Political Economy 89: 642–664 Westholm E, Moseley M, Stenlas N (1999) Local partnerships and rural development in Europe, A Literature Review of Practice and Theory, Falun, Dalara Research Institute Wissenschaftlicher Beirat Agrarpolitik, nachhaltige Landbewirtschaftung und Entwicklung ländlicher Räume Beim Bundesministerium für Verbraucherschutz und Landwirtschaft (2005), Stellungnahme zum Vorschlag für die ELER-Verordnung, KOM(2004)490. Berichte über Landwirtschaft, 83 Wissenschaftlicher Beirat beim Bundesministerium für Ernährung, Landwirtschaft und Forsten (1998), Kompetenzverteilung für die Agrarpolitik in der EU. Agewandte Wissenschaft, Heft 468, Bonn Zeddies J Doluschitz R (1996) Marktentlastungs- und Kulturlandschaftsausgleich (MEKA): Wissenschaftliche Begleituntersuchung zu Durchführung und Auswirkungen. Agrarforschung in Baden-Württemberg 25, Stuttgart, Verlag Eugen Ulmer Zeddies J, Baudoux P, Koll H (2000) Agrarumweltpolitik nach dem Subsidiaritätsprinzip: Fallstudie Kraichgau in Baden-Württemberg. In: Schriften zur Agrarforschung und Agrarpolitik, 5, Berlin, Analytica
13 Tax Policy and Subsidiarity in the European Union
Michael Keen and Ruud de Mooij
13.1 Introduction
1
How much tax coordination is desirable in the European Union — and of what type? This chapter seeks to address these issues by taking as guiding principles the notions of subsidiarity and proportionality that have been central in wider discussions of the appropriate architecture of European Union policy making. We start by considering what the literature on fiscal federalism suggests would be an appropriate division of taxing powers within the Union — and then confront this with the realities of current practice. While there have been many achievements, the comparison brings to light several shortcomings of the European experience. Significant constraints have been posed by a variety of institutional realities — such as unanimity voting, the European Court of Justice and the influence of lobby groups — and efficiency arguments that favour tax competition may have had an influence on the progress in tax coordination. We argue too, that the Union has in recent years come to respond to failures of policy coordination by trying to use administrative measures as a substitute. The review here is brief, and far from complete. The focus is on taxes where cross-border spillovers are most apparent, namely taxes on capital, value-added taxes and excises. Thus, in particular, we do not address is1
The views expressed in this paper are those of the authors and should not be attributed to the IMF, its Executive Board, or its management. We thank Sijbren Cnossen, George Gelauff, Isabel Grilo, Arjan Lejour, Matthias Mors, John Norregaard and an anonymous referee for useful comments.
220
Michael Keen and Ruud de Mooij
sues related to the mobility of labour, likely to become increasingly important in the future and which would raise wider issues concerning access to national benefit systems and sustainability of pension systems (Sinn 2003). The questions of what the European Union has done right in the tax area, and where it has failed, is of interest far beyond Europe. Nascent trading blocs in Central America, Africa and elsewhere naturally look to the European experience for guidance in designing their closer economic cooperation. Given the wider success of European integration it may be natural for them to suppose that in the tax area too the European Union is a good model to follow — but it will be seen that Europe has experienced difficulties that others could perhaps avoid, and that tax development have been shaped largely by circumstances particular to the European Union.
13.2 Principles for the allocation of taxing powers The allocation of powers within the European Union, between the Union itself and its Member States, is guided by principles of subsidiarity and proportionality, discussed at length in Ederveen et al. (2008) in this book. In the economics literature, subsidiarity has been explored under the heading of fiscal federalism. The starting point we take here is that powers should be decentralized unless good reasons for centralized decisionmaking prevail; proportionality then must guide us to the appropriate form of coordination (see also Cnossen 2002). Applying these very broad principles to tax design and implementation, what kind of arrangements would we expect to observe? 13.2.1 Subsidiarity The presumption for decentralization is commonly motivated by preference matching. Clearly, the 27 countries of the European Union vary considerably in terms of per capita income, geography (including between core and peripheral countries) and inherited tastes. The presumption is that national governments know the preferences of their citizens better than could a central authority, and therefore design more efficient tax systems. Only in the presence of decentralization failures will intervention by the Union be desirable. In taxation, decentralization failures can arise from fiscal externalities, distortions in production efficiency, and economies of scope in tax administration and compliance. We discuss each in turn.
13 Tax Policy and Subsidiarity in the European Union
221
Fiscal externalities
Independently operating governments may have an incentive to set taxes on internationally mobile factors at inefficiently low rates.2 This is because they do not take account of the positive welfare implications that setting a higher tax would have on neighbouring countries by inducing movement of these mobile factors. This fiscal externality can cause an underprovision of public goods or, potentially, underprovision of redistributive policies in the adversely affected countries. Such base stealing — or, in Cnossen’s words ‘snatching’ (Cnossen 2002) — is naturally expected to be more aggressive, the more mobile is the tax base. Is there evidence that such tax competition is important in practice? There is certainly evidence suggesting that corporate tax bases are strongly responsive to corporate taxation. This holds both for real capital and paper profits. The real tax base refers to footloose foreign direct investment that is affected by differences in effective tax rates (De Mooij and Ederveen 2003). Paper profits can be shifted across affiliates within multinational companies, either through corporate financial policies or the manipulation of transfer prices (Hines 1999). The impact of the tax system on these aspects of corporate decision making are complex: the incentive to shift profits, for example, depends primarily on the statutory rate of tax compared with that available abroad; the decision where to locate a given investment depends on the comparison between average effective tax rates, taking account too of allowable depreciation and other features of the base; and the marginal effective tax rate — again reflecting both statutory rate and base, but for projects that just earn the minimum required — influence how much to invest in a given country. Trends in statutory corporate tax rates have been as one would expect if tax competition were at work. As shown in Figure 13.1, rates in Western Europe have fallen dramatically during the past decades when the borders to capital movements were removed: from an average of 48% in the early 1980s to around 30% today; in the new Member States from Central and Eastern Europe, tax rates fell from an average of 35% in the mid 1990s, to 18% today.3 Despite this decline in statutory rates, however, revenues from The ‘may’ is because there can be incentives to set inefficiently high taxes: in relation to corporate taxation, for instance, to the extent that profits arising domestically accrue to foreigners, there is an incentive to tax them heavily since the well-being of foreigners is presumably less of a concern than the welfare of domestic citizens. 3 For a sceptical view of the likely importance of tax competition in explaining the fall in statutory rates, see Slemrod (2004). He finds that corporate tax rates are explained better by domestic determinants, such as the balance with the personal 2
222
Michael Keen and Ruud de Mooij
the corporate income tax have held up: indeed Figure 13.2 shows that the average corporate tax-to-GDP ratio in Europe has actually risen during the past decades. To some extent, this reflects the feature that statutory corporate tax cuts were in many cases accompanied by explicit base broadening policies, such as the removal of investment tax credits and less generous fiscal depreciation rules (Devereux et al. 2002). But this seems unlikely to be the full explanation: recent literature also suggests important explanatory roles for the growth of the corporate sector (De Mooij and Nicodeme 2006), growth of the financial sector (Devereux et al. 2004) and increased volatility of corporate profits, combined with imperfect loss offset opportunities (Auerbach 2006). Tax competition and country size Country size matters for tax competition. Kanbur and Keen (1993) show that small countries have incentives to set lower tax rates than large countries. The reason is that a small country that undercuts the rate of the large country attracts a share of the mobile tax base from the large country. While this inflow constitutes a substantial broadening of the tax base in the small country, this flow is relatively small as a percentage of the tax base in the large country. As a result, the small country gains relatively more from setting a low tax rate than the large country. This incentive for small countries to undercut the tax rates in large countries may be offset, however, by the larger risk exposure of citizens in small open economies. Rodrik (1998) argues that more open economies have bigger governments in order to provide more insurance to the people against shocks. Bigger governments require higher tax rates. This offsetting impact may decline to the extent that membership of the European Union will reduce such risks, e.g. because of cross-border insurance mechanisms within a Union. In that case, the accession of several smaller countries from Central and Eastern Europe to the European Union may intensify tax competition within the Union.
Indirect tax bases, such as VAT and excises, can be mobile too. Under the destination principle — with tax levied by the country in which consumption occurs — this can arise from cross-border shopping and smuggling. For the VAT, the evidence suggests that, with some rate exceptions, these considerations are of limited importance. For excises, however, levied on readily transportable, high-value items, cross-border shopping and smuggling are in many places real concerns. Destination-based taxation can also give rise to indirect protection, as countries can differentiate their destination-based excises between predominantly imported and domestically produced goods, favouring the former. For instance, Belgium imposes a low income tax (which illustrates the backstop role of the corporate tax for the personal income tax).
13 Tax Policy and Subsidiarity in the European Union
223
excise on beer (which is typically domestically produced) but a high excise on wine (which is typically imported); Italy does precisely the opposite. Thus, countries try to change the terms of trade in their favour and reap the national welfare gains thereof. Hence, while explicit tariffs on imported goods are prohibited, excise policies can be used as implicit tariffs on imports. % 50 45 40 35 30 25 20 15 10 5 0 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007
EU-15 average
EU-25 average
New Member States
Fig. 13.1. Statutory corporate tax rates in the European Union (incl. local taxes and surcharges) Source: European Commission, Structures of taxation systems - DG TAXUD.
While these developments are suggestive of tax competition, that is not the only possible explanation: it could be, for example, that countries have reduced top rates of personal income tax and in doing so have reduced the corporate tax rate too, so as to avoid excessive distortion of the incorporation decision. Some recent empirical studies look for, and in many cases find, more compelling evidence that tax competition is at work.4 For instance, the correlation between corporate tax rates became greater as capi4
Tax competition between jurisdictions on the same level of government is called ‘horizontal’. ‘Vertical externalities’ arise from over-taxation of a common base by different layers of government. This is less of an immediate concern in the European Union, but potentially relevant to the ultimate choice of a genuine own resource.
224
Michael Keen and Ruud de Mooij
tal controls became weaker, which would not be the case if the correlation were due to common domestic shocks (Devereux et al. 2006; Besley et al. 2001). Strategic tax setting also appears to be important for excises on cigarettes and fuels (Devereux et al. 2007; Evers et al. 2004). These findings provide further evidence that fiscal externalities are important,5 and thus lend some force to the classic argument for some form of tax coordination in corporate taxes and various excises. % 3.5 3 2.5 2 1.5 1 0.5 0 1980
1982
1984
1986
1988
EU-25
1990
1992
EU-15
1994
1996
1998
2000
2002
2004
2006
New Member States
Fig. 13.2. Tax revenues from corporations as percentage of GDP (1980-2004) Source: Structures of taxation systems - DG TAXUD (GDP-Weighted). Production efficiency
A central result in the theory of optimal tax design is Diamond and Mirrlees’ (1971) theorem on production efficiency: in the absence of externalities and non-competitive behaviour, and if pure profits can be fully taxed, any Pareto-efficient tax structure leaves production decisions undistorted. This is because distortions of production reduce aggregate output, placing the economy inside rather than on the production possibility frontier. While Keen and Wildasin (2004) point out that this result does not immediately apply in international settings, with distinct governments facing distinct revenue constraints, it does provide a reasonable benchmark 5
Note that studies on strategic tax responses cannot distinguish between yardstick competition and fiscal externalities due to tax-base stealing (Revelli 2005).
13 Tax Policy and Subsidiarity in the European Union
225
for the European Union given the ability of member countries to arrange transfers (including through common spending policies) among themselves. One important requirement of production efficiency is that marginal productivities of capital are equalized across countries — the condition of capital export neutrality. It is obtained if effective marginal tax rates on capital income are harmonized across countries or, alternatively, if taxes are levied on the basis of the residence country of final investors. If, however, countries levy diverging marginal effective tax rates on investment, the international allocation of capital is distorted: since investors will equate after-tax rates of return, the implication is that pre-tax rates of return differ across countries, and hence that aggregate output in the Union would be increased by relocating real capital from Member States where the pre-tax return is low to members in which it is high.6 How serious are these distortions in welfare terms? In general, this is difficult to assess. If marginal effective tax rates — which measure the impact of the tax system on the cost of capital — are sufficiently close, they keep base movements low. In Europe, however, disparities in marginal effective tax rates are sizeable. Figure 13.3 shows this for 2005 based on computations by the Institute for Fiscal Studies. Marginal effective tax rates on equity-financed investment run between a low 10% in Ireland to a high 29% in Germany. The coefficient of variation equals 0.26. According to the European Commission (2001), the variation in effective tax rates in Europe is driven primarily by differences in statutory tax rates. Production efficiency thus provides a plea for the coordination of these rates. Recently, there have been attempts to estimate the welfare costs of corporate tax competition compared to full harmonization, using computable general equilibrium models. Parry (2003) finds the loss to be relatively small: about 3 percent of capital tax revenues, and even less when a modest degree of ‘Leviathanism’ is present. Sørensen (2004) also reports relatively small costs: less than 1 percent of GDP. In chapter 15 of this volume, Van der Horst (2008) finds that the aggregate welfare impact of harmonisation is almost negligible.
6
Desai and Hines (2003) argue that ownership neutrality should carry at least as much weight in the evaluation of international tax systems as capital export neutrality. This is because the productivity of intangible assets like R&D and marketing units depends critically on who owns and controls them.
226
Michael Keen and Ruud de Mooij
GER BEL NET SPA FRA UK AUT ITA FIN SWE POR GRE IRE 0
5
10
15
20
25
30
% 35
Fig. 13.3. Marginal effective tax rates in a selection of European Union countries, 2005 Source: Institute for Fiscal Studies (The METR refers to investment in plant and equipment, financed by equity or retained earnings. For further assumptions in these calculations, see www.ifs.org.uk).
In indirect taxation, production efficiency requires the destination principle as opposed to the origin principle.7 Significant cross-country tax differentials, however, can erode this through cross-border shopping and smuggling, a potential difficulty made especially challenging in the European Union by the removal of internal fiscal border controls in 1992. Particular difficulties arise, in all parts of the world, in relation to international services, for which border controls in any event are infeasible. Economies of scale/scope
At an administrative level, narrowly interpreted scale economies from European wide tax administration are likely to be modest, there being no evidence that these are less than fully realized at national level. There may, however, plausibly be gains in administration and compliance from better coordination of tax design. Three areas stand out. First, enforcement of the VAT is greatly complicated by the zero-rating of exports between Member States (meaning that no VAT is charged by 7
Note that production efficiency it is not achieved under the destination VAT in the presence of imperfect competition, see Keen and Lahiri (1998).
13 Tax Policy and Subsidiarity in the European Union
227
the exporting country, but refunds due on VAT paid by exporters on their own inputs). The European Commission (2004b) reports that 10 percent of VAT collections are lost in some Member States due to evasion of various kinds, notably in relation to intra-Union exports. Also compliance costs for firms under the current VAT are high, especially due to the delivery of trade data required under Intrastat. Verwaal and Cnossen (2002) report that compliance costs are on average 5 percent of the value of intra-EU transactions, 80% of which are due to Intrastat obligations. Second, in attempting to tax residents on their world-wide capital income, fundamental problems can arise from lack of information about foreign source income. Since investors themselves generally have little incentive to self-report such income, such taxes can only be administered effectively in the presence of cooperation among governments. Third, in taxing companies there would be evident advantages in assessing them on the full range of European activities instead of on a country-by-country basis. A recent Survey of the European Commission suggests that compliance costs associated with corporate tax systems are large: between an average cost-to-tax-revenue ratio of 1.9% for large companies and 30.9% for small and medium sized firms (European Commission, 2004a). A common tax base would be potentially simpler to comply with. This requires similar definitions of the tax base in the various Member States. Yet, the point about compliance costs should not be overestimated as any tax system imposes a cost of compliance. In general, there is a need for coherence (and at a minimum, clarity) in the tax treatment of cross-border operations (e.g. treatment of losses, mergers, etc). 13.2.2 Proportionality There are many possible ways in which tax policies can be coordinated across countries. There could be, for example, coordination only among subsets of EU members, coordination that leaves some room for lowerlevel government discretion, non-binding forms of cooperation, alternative revenue sharing agreements, and so on. Hence, if coordination is desirable, what is its appropriate degree and nature? What strikes the optimal balance between preserving sufficient national discretion while also giving credibility to the agreement reached? Are legally enforceable agreements (perhaps in treaty form) needed, or can ‘name and shame’ methods work? The principle of proportionality requires that coordination be of an extent and nature commensurate with the decentralization failure it seeks to address. But this provides little practical guidance in relation to tax coordination, not least because tax systems are not characterized by a single pa-
228
Michael Keen and Ruud de Mooij
rameter but by complex structures of rates and bases that differ considerably across countries. Hence, even if coordination is desirable, it may not be clear which particular aspects of the tax should be coordinated. For instance, should corporate tax rates or tax bases be coordinated, or both? Even deeper is the question whether coordination should be accompanied by fundamental structural reform of national tax systems in order to better achieve coordination gains. Under the VAT, the question is whether to treat intra-community supplies in the way that exports between fully independent countries are treated, that is by zero-rating. For the corporate income tax, the issue is whether to move from separate accounting to some form of unitary taxation. Under the former system, as at present, profits are attributed to Member State by arms-length accounting methods. Under the latter, multinational profits would be consolidated across national borders within the Union using a common European tax base,8 allocated to individual Member States by the use of some formula reflecting the extent of its activities in each, and countries would then apply their own tax rates to these profits. The European Commission is currently exploring this route for the Union under the heading of a common consolidated corporate tax base. The proposal raises several issues, such as which factors to include in the apportionment formula. This latter approach is not without difficulties. The corporate tax effectively becomes a tax on the factors in the formula, for instance, so that the inclusion of capital could make tax competition worse, at least as long as there is no coordination in tax rates. Moreover, the inclusion of intangible capital would bring back the same problems of valuing assets that create current problems associated with transfer pricing. The Nice Treaty of 2000 opened new opportunities for enhanced cooperation among a subset of Member States in the European Union. This may increase the chances to reach an agreement among countries on tax coordination. It is, in fact, what the European Commission now pursues in its proposal for the common consolidated corporate tax base. However, the question is under which circumstances partial tax coordination among a subset of countries is indeed desirable. For instance, simulation exercises by Sorensen (2004) suggest that, when capital mobility is high, the welfare gains from corporate tax harmonization may indeed be far smaller when only a subset of countries participate.
8
The alternative is to use the tax base of the country where the headquarter of the multinational resides. This so-called home-state taxation runs the risk of intensified tax competition in tax bases by countries trying to attract these headquarters.
13 Tax Policy and Subsidiarity in the European Union
229
13.2.3 Implications If the principles of subsidiarity and proportionality had been given full sway, what type of tax coordination would one expect to observe now in the EU? Clearly, those principles provide much less than complete guidance, but they would lead one to expect: x Considerable coordination in the taxation of corporate income, probably requiring action on both rate and base. x Some degree of coordination on excises with the aim to reduce cross border shopping and rent-shifting. x Limited coordination of VAT, except in dealing with intra-Union trade. How does current tax policy in the European Union measure up to this?
13.3 Tax coordination in the EU The Member States of the European Union do cooperate in some areas of taxation and in a variety of ways. 13.3.1 Indirect taxation European tax coordination has progressed most in the field of VAT under the Sixth Directive of 1977, now recast and consolidated with subsequent directives (Council Directive 2006/112/EC). This defines common concepts for applying the VAT in the Member States and imposes a range of parametric restrictions, including a minimum standard rate of 15 percent and, less commonly remarked, a maximum of 25 percent. Member States may set reduced rates at a minimum of 5 percent on a limited list of products and services. Moreover, they face limitations on the use of exemptions. Many aspects of these rules are widely recognized as not embodying best current practice. The widespread exemptions — for healthcare, education, cultural services, immovable property, finance, insurance and public sector activities — violate the logic and functionality of the VAT and induce unnecessary distortions. They reflect, to a large degree, the age of the EU VAT: more recent VAT-adopters, including Australia, Canada and New Zealand, have learned from these failures and tax (sometimes at zero rates) most of these services that are exempt in Europe. Moreover, consistent with the common rules on rate structures — indeed those rules were largely shaped to permit this outcome — all members except Denmark
230
Michael Keen and Ruud de Mooij
adopt dual or multiple rate VAT structures, increasing costs of administration and compliance in pursuit of distributional advantages that could generally be better achieved by using more closely targeted income taxtransfer instruments. What matters for present purposes, of course, is whether the common EU rules constrain national decision-making significantly, too much or little. As discussed above, fiscal externalities arising from the VAT seem likely to be relatively limited. In this respect the rationale for the minimum standard rate is fairly weak: and it is not in fact binding for any Member State other than Luxembourg (though geography may make that an important case). Indeed there has been some recognition in recent years of greater scope for national autonomy in VAT policy, most notably in allowing the setting (until 2010) of reduced rates on labour-intensive final nontradable goods. Whether or not such differentiation is good policy — and a recent study for the European commission by Copenhagen Economics (2007) suggests not — it could seem to be consistent with subsidiarity. Another serious concern in the European Union is that the break-in-theVAT-chain creates a no man’s land outside VAT control, causing substantial tax evasion. The adoption of minimum rates in the VAT might be considered as a rational response to the threat of tax base stealing in light of cross-border shopping and footloose services that are taxable at the suppliers’ place of location. Yet, empirical evidence on cross-border shopping suggests that this is not very pressing for the VAT. The minimum VAT rate is nonbinding for all countries, except for Luxembourg. Consequently, rate approximation between Member States does not seem a necessary measure. The European Union imposes minimum rates on excises on alcoholic drinks, tobacco and motor fuels. There is a clear rationale for this due to cross-border shopping and smuggling, which is more important for these commodities than for the VAT. Cross-border shopping gives rise to fiscal externalities in the form of revenue losses and time and travel cost of cross-border shoppers. The key issue is that some of the minimum excise rates in the European Union are set at very low levels, with some alcoholic drinks even featuring a zero minimum.9
9
See Cnossen (2006, 2007) for comprehensive reviews of European excise policies regarding tobacco and alcohol, respectively.
13 Tax Policy and Subsidiarity in the European Union
231
13.3.2 Capital income taxation To facilitate the enforcement of residence taxes on interest, in 2005 the European Union has agreed upon a Savings Tax Directive, which requires that countries automatically report the interest paid to European Union residents in another country to the residents’ national tax authorities. Full information sharing among all Member States is the long-term goal of the European Union. During a transition period, however, three European Member States — Austria, Belgium and Luxembourg — plus a number of non-European tax havens were allowed to opt for a withholding tax instead of automatic exchange of information. These withholding taxes will rise to 35% in 2011. Three-quarter of the revenue of these taxes is transferred to the residence country. The residence principle has the desirable feature that it complies with capital export neutrality. It also limits tax competition as investors can no longer take advantage of lower tax rates abroad by reallocating their capital. Residence taxation further involves little sacrifice in national sovereignty: it allows governments to set their own preferred levels of taxation. The current Savings Tax Directive leaves room for various loopholes, however. As interest income can be easily transformed into dividend income and dividends are not covered by the information sharing agreements, it is easy to escape taxation. Moreover, several overseas areas do not participate in the information sharing agreement, which create leakages that so far render the directive ineffective in raising extra revenue. The European Commission has for long been making proposals for coordinating corporate tax systems of European Union Member States. Until today, only a few directives have been implemented, such as the ParentSubsidiary directive and the Merger directive. These aim to avoid international double taxation of dividends, interest and royalties and to facilitate cross-border businesses development. Furthermore, the European Union has agreed upon a non-binding Code of Conduct for business taxation. The code aims to curb business tax measures that discriminate firms within a country in terms of effective levels of taxation. Special schemes which are ‘ring-fenced’ in the sense of being restricted to particular investors or sectors are considered to be ‘harmful’. There is a small literature which suggests that this may not be correct.10 Allowing countries to compete aggressively over particularly mobile aspects of the corporate tax base while maintaining higher rates on the rest may lead to an outcome that is better for all concerned than one in which they are required to set the same tax 10
See Keen (2001) for the basic result, and Janeba and Smart (2003) for a more general analysis.
232
Michael Keen and Ruud de Mooij
rate on all parts of the base. The reason is that it may ultimately be less damaging for countries to compete aggressively over a narrow base than to compete less aggressively over a wider one. The Irish experience illustrates the point: under pressure to move from a system combining a normal corporate tax rate of 36 percent and a special rate of 10 percent for manufacturing and some other activities, Ireland decided to adopt a single rate of 12.5 percent. Whether this represented a restriction on or an intensification of tax competition is by no means clear. Thus, the European Code of Conduct may well have exacerbated, rather than alleviated fiscal externalities. All high profile proposals for corporate tax harmonization — such as those presented in the Ruding report in 1992 for a harmonized tax base and a minimum corporate tax rate — have failed to gain adequate support. This might be considered as the main failure in European tax coordination.
13.4 Why tax coordination fails Coordination failures in European tax policies can be explained by political imperfections as well as alternative views on tax competition that emphasizes its efficiency gains. 13.4.1 Politics: the role of institutions European tax policies are not determined by a benevolent supranational authority, acting in the best interest of European citizens. Instead, various institutions determine the way taxes are designed and coordinated across Member States. Unanimity
The literature suggests that the welfare effects of tax coordination are asymmetric, both between and within countries (Sorensen, 2000; Baldwin and Krugman, 2004; Van der Horst, 2008). For instance, small countries are more likely to win tax competition games than large countries; countries in the periphery of the Union are more likely to lose from tax harmonization than counties in the core of Europe; and the poor are more likely to gain than the rich. These asymmetric effects make it difficult to realize any collective gains from mitigating tax competition, especially as long as unanimity applies in decisions on taxation.
13 Tax Policy and Subsidiarity in the European Union
233
The obstacles induced by unanimity voting can in principle be addressed if coordination is accompanied by compensating transfers to any countries that gain from tax competition (or equivalently, that lose from coordination). This appears unlikely in practice, however, because of the appearance of rewarding those who are gaining at the expense of others. Alternatively, if countries bargain on a range of issues, then — even short of explicit transfers — sensible tax arrangements might emerge via package deals. Finally, it may be possible to design measures that are in themselves Pareto improving. For instance, imposing a minimum tax that forces low tax countries to raise their tax rates can make them better off by inducing high tax countries to raise theirs too. European Court of Justice
Based on notions of non-discrimination, the European Court of Justice has recently become active in banning several tax rules in European Union Member States. In effect, the Court prospectively ‘trumps’ unanimity by facing countries with their Treaty obligations. It is arguable as to whether this has made for good tax policy. For instance, Court decisions have lead to a demise of imputation systems in Member States. The reason is that these systems discriminate between domestic and foreign investments. Yet, there is a strong economic case to be made for imputation systems. Even in its 1975 proposal for harmonizing capital income taxation in the Union, the European Commission argued that each Member State should have a partial imputation system. This example illustrates that a more considered review would have been preferable to the removal of these systems for the fear of adverse legal decisions. More generally — whatever one thinks of the tax policy merits of the decision — these are issues not best left to legal reasoning finding unanticipated implications of past agreements. Rather, it would better to base such decisions on sound policy analysis and be decided in the political arena. Lobby groups
It is hard not to see signs of an influence of national and cross-national lobby groups on tax policy in the European Union. One example is exemplified by the Bolkestein report (European Commission, 2001). The report draws two major conclusions on the corporate tax systems in the European Union. First, distortions induced by corporate tax competition are mainly due to the variation in statutory corporate tax rates. Second, European tax systems create serious obstacles to international business, such as limits to international loss offset and high compliance costs for multinationals.
234
Michael Keen and Ruud de Mooij
What we observe is that the policy response has focused almost entirely on the second issue while the first was ignored completely. This was backed by well-organized interest groups representing the European multinationals, such as UNICE (now called Business Europe). It was certainly not in their interest to harmonize statutory corporate tax rates. The common consolidated corporate tax base proposal is attractive for multinational companies as it reduces compliance costs within the European Union while, at the same time, firms can still use transfer pricing to shift profits to tax havens outside the Union. Another example is excise structures. These reflect a tendency to favour national producers, with low rates on domestic products or tax structures tilted to favour them. This, for instance, is the case with the structure of specific and ad-valorem excises on cigarettes in Member States. Preference biases of this kind would tend to point to high rather than low taxes on these goods, but lobby groups have reversed policies in the opposite direction. Again, the design of tax systems is better not left to interest groups. 13.4.2 Politics again: the efficient side of tax competition11 The literature offers several arguments as to why tax competition may not be all that bad. Yardstick competition
A traditional argument for decentralization is that it permits experimentation from which all can ultimately gain. Recent literature on yardstick competition explores further how this happens, with voters judging policy makers in part by their performance relative to others. There is evidence that this does indeed occur (Besley and Case 2003). However, the efficacy of this process is least clear precisely when the success of the policy rests on fiscal externalities. Mimicking a country that has attracted mobile tax bases by setting low tax rates (e.g. Ireland) may make sense as a best response (by e.g. Central and Eastern European Countries), though coordination would be mutually preferable.
11
Even though governments do not cooperate explicitly under tax competition, they might do so implicitly. Since the tax-setting game is played many times, they may sustain tacitly the cooperative outcome and so avoid inefficiencies.
13 Tax Policy and Subsidiarity in the European Union
235
Taming Leviathan
A different view on tax competition — developed first by Buchanan and others taking the view of government as ‘Leviathans’ — emphasizes that the inefficiencies from non-coordination reduce the welfare of the policymakers themselves, but increase that of the citizenry. Much of the policy debate on this issue quickly becomes a sterile statement of ideology, with tax competition seen as either wholly bad or wholly good. One strand of the literature has sought to better understand the trade-off at issue, between improved efficiency of the tax system as a result of coordination and the possibility that some or all of the additional revenue will be wasted. The trade-off is explained in the textbox. The key difficulty remains, however, of determining what proportion of spending is ‘wasted’ — or indeed what ‘waste’ means in this context. Different people may clearly take different views, for example, on the social value of spending that is essentially redistributional. Fiscal externalities versus fiscal discipline Recognizing the deadweight loss from taxation, governments should take public expenditure to the point at which the marginal social value of public expenditure is 1+MDL, where MDL is the marginal deadweight loss from raising an additional euro of revenue. Suppose now that countries agree to collectively raise the tax rates they set by enough to raise revenue of €1 for each. If aggregate supply of capital is fixed, this is in effect a lump sum tax so that the social cost to each is also €1. On the social benefit side, ‘useful’ public expenditure rises by €(1-Ȝ), where Ȝ is the proportion of public expenditure that is wasted. The social value of these expenditure is €(1-Ȝ)(1+MDL). Comparing the benefits and costs, a coordinated increase in tax rates from the noncooperative equilibrium increases the citizens’ welfare if and only if Ȝ < MDL/(1+MDL). This formula enables one to narrow down the scope of policy disagreement. Suppose, for example, that the marginal deadweight loss from taxation is at least 15 cents per euro. Then all those who believe that government wastes no more than 13 cents per euro of its spending, at the margin, should agree that a small coordinated increase in the corporate tax rate would be beneficial (see Edwards and Keen, 1996).
Time consistency
Tax competition may be beneficial even when policymakers are wholly benevolent. This is because the possibility of capital flight to low tax countries provides a way in which governments can commit not to impose heavy ex-post taxes on savings once they have been made. Thus, tax competition overcomes the basic time consistency problem in taxing capital income (Kehoe 1989). This is probably of limited importance for the Euro-
236
Michael Keen and Ruud de Mooij
pean Union, however, as capital remains mobile relative to the rest of the world. Inefficiencies under tax coordination
This last point refers to a third-country problem of European tax coordination. If countries coordinate towards a higher corporate income tax level amongst themselves, they may simply set themselves up to be undercut by others. Thus, they may end up worse off. It suggests that it can be critical for cooperation to be successful to include third parties in the agreement. The existence of non-participating countries does not necessarily mean that there is no gain from coordination — though it might be less than if all participated. Konrad and Schjelderup (1999) show that coordination will bring welfare gains so long as there is strategic complementarity in tax instruments (others raise the tax if European Union Member States raise theirs). Limits to competition over corporate tax rates may also intensify strategic competition in other policy areas. For instance, Keen and Marchand (1997) show that countries competing for mobile capital tend to modify the composition of public expenditure in favour of public investment in e.g. infrastructure and at the expense of public consumption and redistributive spending (see also Pouget and Steclebout (2008) in Chapter 16).
13.5 Administrative measures as a substitute for policy choices One feature of recent developments is that the failure to coordinate taxes has been succeeded by essentially administrative measures directed to the same goal. One example is the focus on improved administrative cooperation to deal with VAT fraud related to intra-community supplies. A second is the focus on information exchange as a substitute for rate coordination, to better enforce residence principle in the taxation of capital income. Seeking administrative solutions can be attractive in terms of proportionality if it is all that is needed. The question is whether these measures will prove enough. Value-added taxes
In the VAT, the universally-acknowledged problem is the scope for fraud created by the VAT treatment of intra-community supplies. A strong case can be made for more than administrative measures to tackle this problem:
13 Tax Policy and Subsidiarity in the European Union
237
as long as intra-community supplies are zero-rated, clever fraudsters will find ways of attacking the system. Experience suggests that administration will always be one step behind. There are well-known policy responses, all intended to fix the break-in-the-VAT-chain at export stage by imposing some form of tax in the exporting country (Keen and Smith 1996). Examples are the clearing house, the compensating VAT (CVAT) and the viable integrated VAT (VIVAT). Under the clearing house, intra-EU exports would be taxed at the rate of the exporting country and importers are entitled to a tax credit for the out-of-state VAT. To restore the destination principle, the administration of the exporting state would remit the VAT collected on exports to the administration of the importing state. Ultimately, only net balances would have to be settled through some central clearing house, e.g. either on the basis of invoices or on the basis of aggregate estimates of consumption in the Member States. Under CVAT, there is a special tax rate for exports operated by a separate European tax authority. This authority both collects VAT on exports and gives the credits for the input taxes paid on imported goods. In principle, the amounts of VAT collections and credits should cancel out. Its only purpose is to fix the break in the VAT chain. The VIVAT would impose an EU-wide uniform rate on cross-border transactions between registered traders. If a Member State would prefer a higher than common rate, it would be permitted to levy a higher rate for sales at retail. These proposals involve no genuine loss of tax sovereignty in that the change to consumer prices could be minimal. It may prove an important test case as to whether the institutions of the union will be adequate in addressing a problem that requires a significant degree of policy coordination. One of the difficulties faced is the incentive problem in the cooperation agreement. For instance, if with the clearing house proposal clearing is done on an invoice basis, national tax administrations may have little incentive to check claims for credit or refund of input VAT charged in another Member State. If clearing is done on the basis of aggregate consumption data, there may be little incentive to collect the VAT itself. In any case, cooperation between national tax administrations will require the provision of proper incentives for such cooperation to be effective. Information sharing
In the taxation of interest, informational difficulties are at the heart of the challenge to enforce the residence principle. Therefore, automatic information exchange is, in principle, the best response. Effective cooperation requires proper incentives for governments to engage in this information sharing. On the face of it, countries benefiting from an inflow of capital
238
Michael Keen and Ruud de Mooij
that is seeking to avoid residence taxes have little to gain by sharing information. A more subtle view qualifies this by noting the potential strategic consequences of such sharing: it will enable the residence country to set a higher tax rate than it otherwise would, which in turn may result in the source country also gaining (Bacchetta and Espinosa 1995). Beyond this, there may be a case for providing a direct incentive to share information in the sense of giving the source country a direct interest in the additional revenue it generates (Keen and Ligthart 2006).
13.6 Conclusions It appears to be widely believed that there has been relatively little purposive coordination of tax policies in the European Union. However, one might also argue that in fact the Union has had considerable successes, and — no doubt reflecting the need to manoeuvre between complex political circumstances — shown considerable ingenuity. Some examples: x Though the Sixth VAT Directive is heavily criticized and probably far from best practice, the fact remains that the European Union has played a key role in developing the VAT into a major and apparently relatively effective source of revenue in its Member States. It has served as a model for the VAT around the world. The European Union has in this respect in itself generated considerable fiscal externalities for the rest of the world. x The adoption of minimum rates on excises, while leaving some important areas unconstrained, was an imaginative and appropriate response to the proportionality issue. x The focus on information sharing is another creative innovation, seeking to rescue the residence principle by addressing the informational problems that are at the heart of the challenges it faces. x The development of a non-binding code of conduct was a creative response to the difficulty of achieving binding agreements — though in a sense its effectiveness remains unclear, given the power that was later found in the state aid rules. The most obvious criticism is the apparent failure to have dealt with core issues of corporate tax coordination — and, not having done so early in the life of the community. The subsequent increase in membership (and, more particularly, increasing diversity of the membership) has made the task only more difficult. A historic mistake was made, one might argue, in focusing so much on the VAT rather than capital income tax coordination.
13 Tax Policy and Subsidiarity in the European Union
239
But here too, criticism can be over-hasty. It is by no means clear that the welfare loss from decentralization failures in this area are substantial. Moreover, the view that tax competition may have beneficial effects — though theoretically uneasy — clearly still has considerable force for many. Nevertheless, a lesson for nascent unions, it would seem, is that capital tax coordination issues may be best addressed earlier on in the process, when lobbies may be at their weakest — given other benefits that they can be expected to realize from closer integration — and membership at its most homogeneous. With agreement on policy coordination proving elusive, the focus of tax coordination has shifted by default to the Court of Justice and administrative cooperation. These, however, raise genuine cause for concern: the former because legalistic interpretations do not necessarily lead to good policy; the latter because, as experience has often shown, even good administration is ultimately no substitute for good policy.
References Auerbach A (2006) Why have corporate tax revenues declined? another look. CESifo Working Paper 1785 Bacchetta P, Espinosa MP (1995) Information sharing and tax competition among governments. Journal of International Economics 39: 103–121 Baldwin R, Krugman P (2004) Agglomeration, integration and tax harmonisation. European Economic Review 48:1–23 Besley T, Case A (2003) Political institutions and policy choices: evidence from the United States. Journal of Economic Literature 41: 7–73 Besley T, Griffith R,. Klemm A (2001) Empirical evidence on fiscal interdependence in OECD countries. Mimeo Besley T, Smart M (2002) Does tax competition raise voter welfare? CEPR Discussion Papers 31 Cnossen S (2002) Tax policy in the European Union: A Review of Issues and Options. FinanzArchiv 58: 466–558 Cnossen S (2006) Tobacco taxation in the European Union. CPB Discussion Paper 67 Copenhagen Economics (2007) Study on reduced VAT applied to goods and services of the Member States of the European Union. Report for DG Taxud 6503 Diamond P, Mirrlees JA (1971) Optimal taxation and public production I: Production efficiency. American Economic Review 61: 8–27 Desai M, Hines J (2003) Evaluating international tax reform. National Tax Journal 56: 487–502 Devereux MP, Griffith R, Klemm A (2002) Corporate income tax reforms and international tax competition. Economic Policy 35: 451–495
240
Michael Keen and Ruud de Mooij
Devereux M, Griffith R, Klemm A (2004) Why has the UK corporation tax raised so much revenue? Fiscal Studies 25: 367–388 Devereux MP, Lockwood B, Redoano M (2006) Do countries compete over corporate tax rates? Oxford University Centre for Business Taxation working paper De Mooij RA, Ederveen S (2003) Taxation of foreign direct investment: a synthesis of empirical research. International Tax and Public Finance 10: 673–693 De Mooij RA, Nicodeme G (2006) Corporate tax policy, entrepreneurship and incorporation in the EU. European Commission Economic Paper 69 Ederveen S, Gelauff G, Pelkmans J (2008) Assessing subsidiarity. In: Gelauff G, Grilo I, Lejour A (eds) Subsidiarity and economic reform in Europe: 19–40, Springer Edwards J, Keen M (1996) Tax competition and the Leviathan. European Economic Review 40: 113–134 European Commission (2001) Company taxation in the Internal Market, COM(2001)582 European Commission (2004a) European Tax Survey, SEC(2004) 1128/2 European Commission (2004b) Report from the commission to the council and the European parliament on the use of administrative cooperation arrangements in the fight against VAT fraud, COM(2004)260 Evers M, De Mooij RA, Vollebergh HRJ 2004, Competition under minimum rates: The case of European diesel excises. CESifo Working Paper 1221 Hines JR (1999) Lessons from behavioral responses to international taxation. National Tax Journal: 305–322 Janeba E, Smart M (2003) Is targeted tax competition less harmful than its remedies? International Tax and Public Finance 10: 259–280 Kanbur R,. Keen M (1993) Jeux sans frontieres: tax competition and tax coordination when countries differ in size. American Economic Review 83: 877–892 Keen M, Smith S (1996) The future of value-added tax in the European Union. Economic Policy 23 Keen M, Marchand M (1997) Fiscal competition and the pattern of public spending. Journal of Public Economics 66: 33–53 Keen M, Wildasin D (2004) Pareto-efficient international taxation. American Economic Review 94: 259–275 Keen M (2001) Preferential tax regimes can make tax competition less harmful. National Tax Journal 54: 757–762 Keen M, Lahiri S (1998) The comparison between destination and origin principles under imperfect competition. Journal of International Economics 45: 323–350 Keen M, Ligthart JE (2006) Incentives and information exchange in international taxation. International Tax and Public Finance 13: 163–180 Kehoe PJ (1989) Policy cooperation among benevolent governments may be undesirable. Review of Economic Studies 56: 289–296 Konrad KA, Schjelderup G (1999) Fortress building in global tax competition. Journal of Urban Economics 46: 156–167
13 Tax Policy and Subsidiarity in the European Union
241
Parry IWH (2003) How large are the welfare costs of tax competition? Journal of Urban Economics 54: 39–60 Pouget F, Steclebout-Orseau E (2008), Corporate tax coordination and differentiated public good provision. In: Gelauff G, Grilo I, Lejour A (eds) Subsidiarity and economic reform in Europe: 273–290, Springer Revelli F (2005) On Spatial Public Finance Empirics. International tax and public finance 12: 475–492 Rodrik D (1998) Why do more open economies have bigger governments? Journal of Political Economy 106: 997–1032 Sinn H-W (2003) The new systems competition, Blackwell Slemrod J (2004) Are corporate tax rates, or countries, converging? Journal of Public Economics 88: 1169–1186 Sorensen, PB (2000) The case for international tax coordination reconsidered. Economic Policy 31 Sorensen PB (2004) International tax coordination: regionalism versus globalism. Journal of Public Economics 88: 1187–1214 Van der Horst A (2008) Who benefits from tax competition in the European Union? In: Gelauff G, Grilo I, Lejour A (eds) Subsidiarity and economic reform in Europe: 259–272, Springer Verwaal E, Cnossen S (2002) Europe’s new border taxes. Journal of Common Market Studies 40: 309–330
14 Coordinating Corporation Taxes in the European Union: Subsidiarity in Action!
Sijbren Cnossen
14.1 Introduction The future of capital income taxation in the European Union (EU) hinges importantly on the future of the corporation tax (CT). Under the EU treaty, the Member States do not have to harmonize their CT rates or bases. Harmonization is to be “approximated” only if required for the functioning of the internal market. But greater approximation of capital income tax systems could promote investment, improve the tax burden distribution and, last but not least, reduce compliance and administrative costs. While the normal return on mobile capital cannot be taxed at the same high rates as labour income, tax coordination should enable the Member States to capture some of that return. After all, capital is less mobile in the EU as a whole than between individual states. Tax coordination should also make it possible to tax firm-specific rents more effectively (although not at the same high rates as location-specific rents, if separately identifiable). Furthermore, there is no reason why foreign share- and bondholders should be completely exempt from tax. Beyond that, the CT is needed as a backstop to the individual income tax (PT). Without a CT, the labour income of the self-employed would be retained in corporate form and largely escape the PT. In short, effective if moderate taxation of capital income seems desirable.1 1
For the rationale of retaining the CT, see Bird (2002) and for the arguments for retaining the CT in a globalized capital market, see Zodrow (2006).
244
Sijbren Cnossen
Although the arguments for coordinating the capital income taxes are persuasive, the difficulties in reaching agreement are daunting. In the spirit of the subsidiarity principle, a gradual, bottom-up and largely reversible “coordination” approach seems preferable to a complex, top-down, all-ornothing “harmonization” approach.2 Also, a broadly based approach encompassing the taxation of all forms of capital income seems preferable to confining the coordination efforts to corporate profits. In search of a common coordinated approach, Section 14.2 reviews the existing CT regimes and attempts to find common features that could form the building blocks for further coordination. Following, Section 14.3 outlines the steps that could be taken to promote CT coordination between the Member States. Section 14.4 concludes.
14.2 Brief Survey of Corporation Taxes 14.2.1 Corporation Tax Regimes Table 14.1 shows the CT systems that are found in the various EU Member States. The statutory CT rates range from 10% in Cyprus to 38.9% in Germany. CT rates have greatly been reduced since the early 1990s when capital markets were liberalized. Generally, the rate reductions have been accompanied by base broadening measures, so that CT revenue contributions changed little as a percentage of GDP. Interestingly, CT rates in the 12 new Member States are on average some 11 percentage points lower than in the 15 old Member States. A partial explanation may be that agglomeration matters: core states appear to have higher CT rates than peripheral states (Garretsen and Peeters 2007). The CT regimes in the Member States can be distinguished depending on whether and to what extent they reduce the double tax on distributed profits – i.e. provide dividend relief – that arises when corporate profits are subjected to the CT and again to the PT when paid out as dividends. Double taxation also occurs when retained profits are subjected to the CT and again to a capital gains tax at the shareholder level on increases in share values – increases that, among others, reflect the corporation’s greater net worth as a result of profit retention.
2
For more on the subsidiarity principle, see Ederveen et al. (2008).
14 Coordinating Corporation Taxes in the European Union 245 Table 14.1. European Union: Corporation Taxes (CTs) and Individual Income Taxes (PTs) in 2007 (rates in %) CT–PT system
CT ratea Tax treatment of dividends PT on capital gainsb at shareholder level Ordinary Substantial shares holdings
Imputation system Malta 35 Spain 35.01 UK 30
Tax credit 35 /65 of dividend 2 /5 of dividend 1 /9 of dividend
— 15 10-40
15 15 10-40
Reduced PT rate Austria Belgium Bulgaria Cyprus Czech Republic Denmark Hungary Italy Lithuania Poland Portugal Romania Slovenia Sweden
PT ratec 25* 25* 7 15* 15* 28/43 10 12.5* 15 19* 20* 16* 20* 30
— — — — 12-32 28/43 20 12.5 15 19 10 1 20 30
½ of gain — 20-24 — 12-32 28/43 20 2 /5 of gain 15 19 10 1 20 30
25 33.99 15 10 24 28 20 33 15 19 27.5 16 25 28
Dividend exemption Finland France Germany Greece Latvia Luxembourg Netherlands Slovak Republic
Size of exemption 26 34.43 38.9 25 15 29.63 29.6 19
/10 of dividend /10 of dividend ½ of dividend Full Full ½ of dividend Full Full
28 27 — — — — — 19
28 27 ½ of gain 5% gross — ½ of gain 25 19
Double taxation Ireland
12.5
PT rate 20/42
20
20
3 4
PT rate No CT 21 Estonia — /79 of net dividend 22 22 a CT rates include (i) a surtax in Hungary (4%), (ii) surcharges in Belgium (3%), France (3.3%), Germany (5.5%), Luxembourg (4%), Portugal (10%) and Spain (0.75%–0.01%), and (iii) local taxes in Germany (effectively 17% in Berlin – de-
246
Sijbren Cnossen
Table 14.1 (Continued) ductible from its own base and from the CT) and Luxembourg (6.75% - not deductible from the CT). b PT rates shown are for long-term capital gains (in France, Slovenia and the UK, the capital gains tax rate declines with the length of the holding period). Shortterm gains (generally, arising in a period of less than 12 months) are taxed at higher (effective) rates in Denmark, Germany, Portugal, Rumania, Spain and the UK. c An asterisk (*) indicates that the PT rate is a final withholding tax, which is optional in Belgium and Portugal. Source: Author’s compilation from Supplementary Service to European Taxation (Amsterdam: IBFD Publications BV, loose-leaf), Vol A and B. The latest update was November 2006, but later rate changes have also been taken into account. Distributed profits
Imputation systems are the most structured form of dividend relief at the shareholder level.3 Under the imputation systems, found in three Member States, shareholders are given a full or partial tax credit against their PT for the CT that can be imputed to the dividends (grossed up by the tax credit) received by them. Accordingly, imputation reduces the excess CT+PT burden on profit distributions in proportion to the marginal PT rates of shareholders.4 Under full imputation, as in Malta, distributed profits are taxed at the marginal PT rate of shareholders. The double tax can also be mitigated at shareholder level by subjecting dividend income to a separate or schedular PT rate lower than the top PT rate. Consequently, the relief is proportionately greater for high-incomebracket PT payers than for low-income-bracket PT payers. This regressive result can be mitigated, but not eliminated, by permitting low-incomebracket PT payers whose marginal ordinary PT rate is lower than the reduced PT rate to opt for full double taxation of their dividend income (with a credit for any PT withholding tax imposed at the corporate level), as is possible in Belgium and Portugal. Fourteen Member States subject dividends to a reduced PT rate. Equivalent relief can be provided at the corporate level under a split-rate or dividend-deduction system. For a discussion of the pros and cons, see Cnossen (1997). 4 More than full relief is possible under the CTs in Member States that permit the payment of dividends out of exempt profits without imposing a compensatory tax at the corporate level. Presumably, for this reason, Malta imposes a 15% tax on dividends paid out of untaxed profits. 3
14 Coordinating Corporation Taxes in the European Union 247
Furthermore, exempting dividend income from the PT, fully or partially, can provide dividend relief. Eight Member States follow this approach. A full exemption is equivalent in effect to a schedular PT rate of 0%. More generally, a partial exemption expressed as a fraction, Į, of the total dividend, is equivalent to Į times the ordinary PT rate under the schedular approach. The exemption approach, however, does not permit the imposition of a final withholding tax at the level of the corporation, because the potential tax liability at shareholder level is not known. Finally, Ireland does not provide dividend relief, while Estonia does not impose a CT, although it taxes profit distributions. Imputation systems, long supported by the European Commission, used to dominate the CT picture in the EU, but in recent years most old Member States have switched to schedular taxes on dividend distributions (as well as other capital income). Perhaps not surprisingly, most new Member States followed this lead. The cross-border implications of imputation were found to be discriminatory and overly complicated. Retained profits
The CT plus the PT on realized capital gains determines the tax treatment of retained profits. Table 14.1 indicates that 17 out of 27 Member States tax capital gains on (widely-held) ordinary shares (e.g. quoted on national stock exchanges). Seven Member States tax only gains realized on the sale of other (non-traded) shares, which often represent a controlling interest (called substantial holding) in (closely-held) corporations. Capital gains on these holdings are more widely taxed than gains on traded shares because they often represent labour income sheltered in the corporate form at a CT rate that is lower than the marginal PT rate on other labour income. The capital gains tax rates shown in Table 14.1 are the nominal rates. Deferral and various tax base preferences result in low effective rates. Furthermore, it should be noted that no Member State makes a systematic attempt to alleviate the double tax on retained profits (as Norway does) by allowing shareholders to increase the acquisition price of shares by the corporation’s retained profits net of CT. 14.2.2 Comparative analysis of CT regimes A thorough review of the corporation tax regimes in the EU, including tax bases, tax incentives and anti-tax avoidance measures (see Cnossen 2005) indicates that the CTs in the EU Member States are levied at widely differing rates applied to widely differing tax bases. No state heeds the norma-
248
Sijbren Cnossen
tive implication of the accretion concept of income that the taxation of corporate earnings (profits and interest) should be integrated with the PT. Generally, dividend income is taxed at reduced PT rates or (partially) exempted, while capital gains on shares are taxed at very low effective rates (or fully exempted). Furthermore, interest is taxed at lower rates than apply to retained profits or dividend income. Overall, capital income is taxed separately from and at much lower rates than labour income that is subject to the PT and various hefty, regressive social security contributions. Corporate profits are determined on the basis of international accounting standards, the European-wide rule from 1 January 2005 for companies listed on EU stock exchanges. The general rules for ascertaining taxable profits are broadly in line with what can be expected, but the extremely generous tax incentives in the new Member States, e.g. tax holidays, reduce the tax base to on average three-fourths of what it otherwise would be. It is difficult to gauge the effectiveness with which the CTs and PTs on capital income are enforced. In all but two Member States, pension and investment funds are not taxed and can hence be used as conduits for not paying tax on the normal return on capital. To some extent, this may be prevented by the use of final source withholding taxes and thin capitalization ratios. Little inbound debt capital appears to be taxed. All Member States are reluctant to impose effective withholding taxes on interest for fear of scaring away foreign investment. Most states attempt to apply appropriate transfer pricing rules, but half of all Member States do not have legislation for controlled foreign corporations. More generally, tax competition forces appear to be at work. Particularly, the 12 new EU Member States seem intend on emulating the Irish economic miracle of promoting economic growth and revenue through low nominal CT rates and generous tax incentives to stimulate domestic and foreign investment. Initially, corporate tax revenues may rise notwithstanding the low rate, because multinational companies channel their income to the low-tax states (without necessarily changing their production locations) through transfer pricing manipulation, thin capitalization, and royalty payments to low-tax states.5 However, as more Member States join the low-tax club, a no-win situation will emerge (see also Van der Horst 2008). Accordingly, some form of tax coordination has to be put in place if the baby is not to be thrown out with the bath water. 5
By following a low rate/large ‘tax base’ philosophy, Ireland has snatched sizable revenues from other Member States. Ireland’s CT/GDP ratio is 3.7 compared with an EU-15 ratio of 2.5, although Ireland’s CT rate is less than one-third of the EU-15’s average rate. Cyprus, the Czech Republic, Luxembourg, Malta and the Netherlands appear to be following a similar strategy.
14 Coordinating Corporation Taxes in the European Union 249
14.2.3 Broad features of CT regimes The actual taxation of corporate earnings and other capital income surveyed above yields a number of insights that have a bearing on the future of the CT in the EU. These insights can be summarized as follows. 1. All Member States tax capital income and labour income separately. Often capital income appears to be taxed jointly with labour income, but in practice no Member State does so. This situation could be recognized more formally by adopting a dual income tax (Cnossen 2000) that would eliminate various ambiguities and tax capital income more effectively (Zee 2002). 2. Capital income is taxed at much lower rates than labour income, by a margin of perhaps as much as one to three. This is due to the greater mobility of capital. If capital would be taxed as high as labour (or, more precisely, at a higher rate than the rate in other countries), the incidence of the excess would almost certainly fall on labour. Also, flat rates seem indicated to limit the countless opportunities for tax arbitrage. 3. With few exceptions, distributed profits are taxed at higher CT+PT rates than retained profits, which may distort dividend payout and investment policies. Equal treatment seems worth pursuing. This would be possible if dividend income were exempted under the dual income tax, whose PT rate on capital income equals the CT rate. 4. Domestic interest income is not taxed if it accrues to tax exempt institutions, such as pension funds. If debt can easily be substituted for equity, this implies that the normal return on capital is not taxed. In the event, the tax on capital income resembles a business cash flow tax, whose tax base is confined to inframarginal profits. Final source withholding taxes (without the possibility of a refund for tax-exempt institutions) or no deduction for interest at the level of the corporation seems the answer if the income tax is to be maintained. This would represent a move toward a comprehensive business income tax (CBIT) under which profits are determined on a normal accrual basis of accounting but interest is not deductible at the corporate level and not taxed at the level of the recipient (US Department of the Treasury 1992). Accordingly, tax-exempt institutions would be taxed implicitly. 5. The tax incentives particularly in the new Member States are so generous that investment costs can often be written off immediately. Again, this converts the CT into a cash-flow tax, because the normal return on capital is not taxed (assuming that interest is actually taxed through, say, (final) source withholding). The abolition of the tax in-
250
Sijbren Cnossen
centives but the retention of the de facto exemption of interest also would make the CT equivalent to a cash flow tax if equity can be fully substituted by debt.6 To the extent that full substitution is not possible, an argument can be made in favour of an allowance for corporate equity (Institute for Fiscal Studies 1990). Under this regime, a deduction is allowed from corporate profits equal to the amount of equity in the balance sheet multiplied by the risk-free rate of interest (normal return on capital). 6. Interest on inbound capital generally is not taxed for fear that debtfinanced investment costs will increase and foreign investment decline. Tax coordination is required if this interest is to be taxed. The third country issue remains, but capital is less mobile in the EU as a whole than with respect to (small) individual Member States. Further coordination could be pursued with the US and Japan. It is difficult to choose between these often conflicting directions for change, but – after allowing for the partiality that may be in the eye of the beholder – the common denominator seems to be that the body politic in most Member States appears to want to tax capital income at positive rates, if some way can be found to temper real or perceived tax competition. This stand is supported by economic arguments, as pointed out by Salanié (2003) and others. First, weak separability is a bad assumption in choosing between current and future consumption. Secondly, it may be optimal to tax capital income if bequests are not properly taxed. Another common strand seems to be that capital income should be taxed separately from labour income and at moderate, flat rates.
14.3 Bottom-up approach: tax coordination by Member States This chapter proposes that an agenda for capital income tax coordination (and perhaps eventually tax harmonization) should comprise five sequential steps:
6
Under a proper cash flow tax, of course, corporations are denied a deduction for interest as well as dividends paid (if not already denied), but they are allowed an immediate write-off of the cost of business assets. As a result, the return on marginal investments, just making a viable economic return, is exempted. For arguments why cash flow taxation has economic and administrative advantages over a conventional income tax, see McLure and Zodrow (1996).
14 Coordinating Corporation Taxes in the European Union 251
1. the introduction of dual income taxes by all Member States under which capital income would be taxed once at a single rate (different for each Member State) to mitigate the distorting effects of the current differential rate CT+PT systems on corporate financial and investment policies; 2. the introduction of interest withholding taxes by the Member States at the CT rate (or, alternatively, the treatment of interest on a par with dividends) to effectively tax the normal return on capital and mitigate incentives for thin capitalization; and 3. the close approximation of the CT rates throughout the EU to eliminate incentives for transfer pricing manipulation and thin capitalization. Following these steps, a fresh review should be made of: 4. the introduction of EU-wide common base taxation with formula apportionment and, subsequently, 5. the adoption of a European CT if and when the EU is given the power to tax. These steps are elaborated below. Dual income tax7
The dual income tax is a pragmatic approach to the uniform taxation of capital income, which, in the early 1990s, was successfully introduced in the Nordic countries, especially Norway, Finland and Sweden. In adopting the dual income tax, these countries argued that, in (small) open economies, any source-based tax on capital income in excess of the real world rate of interest raises the pre-tax return by the full amount of the tax, so that the after-tax return continues to equate to the exogenously given real world rate of interest. Accordingly, caution in setting the CT rate was advisable. Furthermore, capital market innovation in conjunction with tax arbitrage implied that it would not be possible to tax capital income effectively at progressive rates. Since, for revenue and distributional reasons, these countries were not prepared to lower the top PT rate to the level of the lower CT rate, they decided to tax capital income on a schedular basis.
7
For a review and evaluation of the economic and technical aspects of the dual income tax on which this section draws, see Cnossen (2000). For an update on developments in Norway, see Christiansen (2004) and for arguments favouring a dual income tax in Germany, see Spengel and Wiegard (2004).
252
Sijbren Cnossen
The main features of the Nordic dual income tax are the following. 1. Separation of capital and labour income. All income is separated into either capital income or labour income. Capital income includes business profits (representing the return on equity), dividends, capital gains, interest, rents and rental values. Labour income consists of wages and salaries (including the value of labour services performed by the owner in his or her business), fringe benefits, pension income and social security benefits. Royalties are taxed as labour income or as capital income (if know-how is acquired or capitalized). 2. Tax rates. Basically, all capital income is taxed at the proportional CT rate, while labour income is subject to additional, progressive PT rates. To minimize tax arbitrage, the tax rate on labour income applicable to the first income bracket is set at (approximately) the same level as the proportional CT rate. 3. Offset of capital income against labour income. Finland and Sweden tax capital and labour income entirely separately. Alternatively, in Norway, the two forms of income are taxed jointly at the CT rate, while net labour income is subsequently taxed at additional, progressive PT rates. Joint taxation permits the offset of negative capital income against positive labour income. But the same effect is achieved in Finland and Sweden by permitting a tax credit for capital income losses (calculated at the basic rate) against the tax on labour income. Furthermore, joint taxation, as in Norway, permits the application of joint basic allowances. Separate taxation, on the other hand, enables the imposition of flat source taxes, if desired, on various forms of capital income, as is done in Finland. 4. Avoidance of double taxation. In Norway, the double taxation of distributed profits at the corporate level and the shareholder level is avoided through a full imputation system. Alternatively but equivalently, double taxation can be avoided by exempting dividend income at the shareholder level, as Finland does. Under either approach, compensatory taxes guarantee that no dividends are paid out of exempt profits without having borne the CT. The double taxation of retained profits at the corporate level in conjunction with the taxation of realized capital gains at the shareholder level is avoided in Norway by permitting shareholders to write up the basis of their shares by the retained profits net of the CT. Similarly, the basis is written down if losses occur or profits are distributed out of previously accumulated earnings. Appropriate adjustments are also made if capital is paid in or paid out. The first in/first out principle applies if part of the same shareholding is sold. The method deals both with the danger of excessive distributions of retained profits and with the unwarranted exemp-
14 Coordinating Corporation Taxes in the European Union 253
tion of realized gains at the shareholder level due to unrealized gains at the corporate level. The double tax on retained profits is mitigated in Finland (only 70% of capital gains are taxed), but fully maintained in Sweden. 5. Withholding taxes. The single taxation of capital income can be ensured through withholding or source taxes at the corporate level or at the level of other entities paying interest, royalties or other capital income. In principle, withholding or source rates should be set at the level of the CT rate. Consequently, these rates could represent the final tax liability if capital income is taxed separately from labour income and no basic allowance applies. This is the case in Finland and Sweden with respect to interest income. No country, however, imposes a withholding tax on interest or royalties paid to non-residents in treaty countries. Withholding taxes are imposed only on dividends paid to non-resident (portfolio) shareholders. 6. Proprietorships and closely-held corporations.8 In Finland and Norway, the taxable profits of proprietorships and closely-held corporations, conventionally computed, are split into a capital income component and a labour income component, and these are taxed on a current basis. The capital income component is calculated by applying a presumptive return (the sum of the nominal interest rate plus an entrepreneurial risk premium) to the value of the gross assets of the business (Norway) or to equity (Finland). Residual profits are considered as labour income. Interest withholding taxes
The goal of ensuring single taxation under the current dual income taxes, however, is mostly honoured in the breach with respect to interest (and royalty) payments to exempt entities, such as pension funds, and foreign debt holders (or suppliers of know-how). This hole in the capital income tax bucket can only be plugged by imposing a withholding tax at the CT rate on all interest – in effect, treating interest on a par with dividend income, which is taxed only at the corporate level. Arrangements could then be made under which the tax withheld at the business level would be cred8
This scheme avoids most of the deferral and lock-in effects of the tax that various EU Member States impose on capital gains on substantial shareholdings. Also, the profit-splitting rules of the dual income tax seem easier to administer than some of the tortuous and arbitrary provisions for preventing the undertaxation of the self-employed currently on the statute books in countries without a dual income tax.
254
Sijbren Cnossen
itable in the residence Member States (hence, capital income could be taxed at different rates by these Member States). Alternatively, the tax withheld would not be creditable but would constitute the final liability in the source state (which would require approximation of tax rates if investment location decisions are not to be distorted).9 Final, source-based, withholding taxes on interest would make the dual income tax equivalent to a comprehensive business income tax. This tax, proposed by the US Department of the Treasury (1992), proceeds from the fundamental equivalence between a CT levied at source and an equalrate PT on corporate earnings with a full credit for the underlying CT. Accordingly, no deductions are allowed at the corporate level for dividends and interest paid to shareholders and debt holders, but these income items are not taxed at the level of the recipients, be they individuals, corporations, exempt entities or non-residents. This makes the debt-equity distinction irrelevant and greatly reduces the distinction between retained and distributed earnings (depending on the treatment of capital gains).10 The comprehensive business income tax can be introduced while largely maintaining the present rules for determining taxable profits, including those applicable to depreciation and inventory accounting. Exempt entities and non-residents would be treated the same as resident individuals or corporations. They would not be eligible for a refund of the tax, nor would they have to pay any additional tax in the form of a withholding tax or otherwise. Corporations receiving income that had been subjected to the comprehensive business income tax as dividends or interest would also not be taxed on such income. To ensure that dividends and interest are not paid out of exempt earnings, a compensatory tax should be levied on exempt income (made available for distribution as dividends or interest). Capital gains on shares would be taxed only to the extent that they exceed the acquisition cost stepped up by the corporation's retained profits net of the CT. The main problem of the (final) withholding tax on interest under the dual income tax and the comprehensive business income tax is that they Slemrod (1995) states that “although it is not desirable to tax capital income on a source basis [because source-based taxes are distortionary], it is not administratively feasible to tax capital on a residence basis.” Furthermore, paraphrasing Slemrod (1995a), a EU featuring (equal-rate) source-based capital income taxes would be more efficient than an EU featuring fully enforced residence-based taxes (if feasible of implementation) only because the cost of enforcement is lower for the system of source-based taxes. 10 For a probing comparative analysis and evaluation of the dual income tax, the comprehensive business income, as well an allowance for corporate equity, see Sørensen (2007) and Radulescu and Stimmelmayr (2007). 9
14 Coordinating Corporation Taxes in the European Union 255
would raise capital costs and dampen (debt-financed) investment, because the normal return on capital (i.e. interest), even if received by exempt entities and non-residents, would be implicitly taxed. Although the introduction of interest withholding taxes would seem a goal worth pursuing, gradual and concerted action is called for. Approximation of CT rates
The exemption of dividend income at the personal level and the taxation of interest income at source should reduce the need for concerted tax harmonization at the central EU level. The problem of thin capitalization would be solved and the schemes for CT-PT integration would become redundant. Manipulation of transfer prices, however, could still affect the allocation of the corporate tax base across the Member States. To limit this form of tax arbitrage, a minimum rate, as proposed by the Ruding Committee (1992), would have to be agreed to. Presumably, rate approximation would be easier to achieve following the introduction of dual income taxes and interest withholding taxes. Common base taxation?
The dual income tax and the comprehensive business income tax would still proceed from the separate-accounting approach in determining the taxable profits of affiliated corporations in different Member States. Provisions for the removal of cross-border obstacles to economic activity and business restructuring, therefore, would still be needed. As pointed out by the European Commission (2001), a comprehensive solution to these problems, if desired, can only be achieved through common base taxation, i.e. the joint determination of the profits of firms with cross-border operations on the basis of consolidated accounts and, subsequently, the assignment of those profits to each of the Member States in which the firms carry on business on the basis of the weighted share in various economic activities of the corporation, represented by such factors as its sales, payroll and property (in other words, formulary apportionment – widely practiced in the United States and Canada). The advantages of common base taxation with formula apportionment are fewer distortions, less tax arbitrage and lower compliance costs. Crossborder loss offset would occur automatically. But the path to common base taxation would not be easy, as pointed out by McLure (2004) in a cogent assessment of the European Commission proposals. According to McLure, under common base taxation, firstly, there would be the problem of the diversity of existing definitions of profits (see Cnossen 2005) and the lack of
256
Sijbren Cnossen
an objective standard against which to judge those definitions. Secondly, there is no clearly best way to define groups of firms for purposes of consolidation. Thirdly, no apportionment formula is conceptually and theoretically superior to others. And finally the administration of the common base tax would require unprecedented cooperation among participating Member States.11 Agreement would probably be easier to reach, however, following the introduction of dual income taxes, the taxation of interest accruing to foreign bondholders, and the approximation of CT/PT rates on capital income. A European CT?
EU-wide unitary taxation would fully reduce interstate distortions and compliance costs only if applied by a joint administration under a common code uniformly interpreted by the European Court of Justice. Indeed, common base taxation would probably not be possible without these conditions. Accordingly, the logical conclusion of the tax coordination and tax harmonization steps outlined above would be a European CT whose revenue would either be shared by the Member States on the basis of some formula or flow into the EU's budget. A truly European CT, however, would require fundamental changes in the EU's constitution moving it in the direction of a federal (tax) system. For the time being, this seems a bridge too far.
14.4 Concluding comment This chapter has argued for tax coordination in a form that relinquishes tax subsidiarity gradually but is also reversibly. It has not come out in favour of unbridled tax competition, although it should be acknowledged, particularly in the EU, that tax competition can serve as a discipline on the “prof11
McLure (2004) is even more apprehensive about another proposal of the European Commission, i.e. home state taxation (HST) under which participating Member States would maintain their own rules for determining taxable profits, but firms with cross-border operations would be taxed by the Member State in which their headquarters are located. Subsequently, the consolidated profits would be assigned to each of the participating Member States on the basis of formulary apportionment. According to McLure, HST has no counterpart in the real world and might impede further evolution toward a harmonized CT system. Also, substantial cooperation would be required in the choice of an apportionment formula and perhaps in the rules for consolidation and cross-border loss offsets. Moreover, competition for headquarter’s locations would increase.
14 Coordinating Corporation Taxes in the European Union 257
ligacy of Princes” (Adam Smith) and present-day governments in the EU (Edwards and Keen 1996). Neither has this chapter advocated the exemption of the normal return on capital by confining the corporate tax base to business cash flow or by introducing a personal consumption tax for which strong arguments can be brought to the fore. It has not taken either of these routes in the belief that economic arguments favour the taxation of the return on capital, although at a lower rate perhaps than the tax on labour income. In sum, tax coordination reconciles the requirement of fiscal efficiency with the desire to tax capital income more effectively.
References Bird RM (2002) Why tax corporations? Bulletin for International Fiscal Documentation, May: 194–203 Christiansen V (2004) Norwegian income tax reforms. CESifo Dice Report, 2/3. Journal for Institutional Comparisons: 9–14 Cnossen S (1997) The role of the corporation tax in OECD Member Countries. In: Head J, Krever R (eds) Company Tax Systems, Melbourne: Fiscal Publications: 49–84 Cnossen S (2000) Taxing capital income in the nordic countries: a model for the European Union? Chapter 8. In: Cnossen S (ed) Taxing capital income in the European Union: issues and options for reform, Oxford, Oxford University Press Cnossen S (2005) The future of corporate income taxation in the European Union. capital taxation after EU enlargement, Proceedings of OeNB Workshops, Oesterreichische Nationalbank, January 21, 2005 Ederveen S, Gelauff G, Pelkmans J (2008) Assessing subsidiarity. In: Gelauff G, Grilo I, Lejour A (eds) Subsidiarity and economic reform in Europe: 19–40, Springer Edwards JSS, Keen MJ (1996) Tax competition and leviathan. European Economic Review 40: 600–638 Garretsen H, Peeters J (2007) Capital mobility, agglomeration and corporate tax rates: is the race to the bottom for real? CESifo Economic Studies Advanced Access, published online on May 22, 2007 Institute for Fiscal Studies (1991) Equity for companies: a corporation tax for the 1990s. IFS Commentary no 26 by the IFS Capital Taxes Group McLure Jr CE (2004) Corporate tax harmonization in the European Union: the commission’s proposals. Tax Notes International, November 29: 775–801 McLure Jr CE, Zodrow GR (1996) A hybrid consumption–based direct tax for Bolivia. International Tax and Public Finance 3: 97–112 Radulescu DM, Stimmelmayr M (2007) ACE versus CBIT: Which is better for investment and welfare? CESifo Economic Studies Advanced Access published online on May 22, 2007
258
Sijbren Cnossen
Ruding Committee (1992) Company taxation in the Internal Market, COM(2001)582 final, Brussels Salanié B (2003) The economics of taxation, MIT Press, Cambridge, Mass. Slemrod JB (1995) Comment on V. Tanzi, Taxation in an integrating world. In: Tanzi V (ed) Taxation in an integrating world, Washington, DC, Brookings Institution: 141–146 Slemrod JB (1995a) Free trade taxation and protectionist taxation. International Tax and Public Finance 2: 471–489 Sørensen PB (2007) Can capital income taxes survive? and should they? CESifo Economic Studies Advanced Access published online on May 22, 2007 Spengel C, Wiegard W (2004) Dual income tax: a pragmatic tax reform proposal for Germany, CESifo Dice Report. Journal for Institutional Comparisons 2/3: 9–14 US Department of the Treasury (1992) Integration of the individual and corporate tax systems: Taxing Business Income Once, Washington, DC: US Government Printing Office Van der Horst A (2008), Code of coordination for corporate taxation. In: Gelauff G, Grilo I, Lejour A (eds) Subsidiarity and economic reform in Europe: 259– 272, Springer Zee HH (2002) World trends in tax policy: an economic perspective. Intertax 32/8–9: 352–364 Zodrow GR (2006) Capital mobility and source-based taxation of capital income in small open economies. International Tax and Public Finance 13: 269–294
15 Code of Coordination for Corporate Taxation
Albert van der Horst
15.1 Introduction Should EU-Member States give up their right to design their corporate income tax? Currently, Member States may freely set their tax rates and are allowed to design their tax base as long as it does not constitute harmful tax competition. This is regulated in the Code of Conduct, which is not a legally binding instrument but does have political force. By adopting this Code, the Member States have undertaken to roll back existing tax measures that constitute harmful tax competition and refrain from introducing any such measures in the future.1 Should the European Union go beyond this minimum and coordinate the taxes on corporate income? In the policy debate, a distinction is made between coordination of tax rates and of tax bases. The stance on tax rates is a clear “hands off”, because tax rates are and should be the sole responsibility of the Member States. The stance on the tax base is less clear cut as the European Commission (2002, 2006) aims at consolidation. This chapter questions this current stance on both the rate and the base of the corporate income tax by investigating the economic aspects of both. Is tax rate harmonization, or alternatively a minimum rate, justified on economic grounds? Will tax base consolidation be a step forward, in improving efficiency in the European Union?
1
European Commission on harmful tax competition, quoted from: http:// ec.europa.eu/taxation_customs/taxation/company_tax/harmful_tax_practices/ index_en.htm.
260
Albert van der Horst
The subsidiarity principle may provide an answer to both questions once the economic effects of tax rate harmonization and tax base consolidation are known. The first argument in assessing subsidiarity is the economies of scale: is coordinated policy cheaper? These scale economies exist in tax base consolidation by reducing the administrative burden for multinational enterprises (MNEs), but are hardly present in tax-rate policies. The second argument is externalities: does the tax policy in one country affect others? These externalities clearly exist in tax rate policy, where countries underbid each others tax rate in order to attract foreign direct investment (FDI) and stimulate profit shifting. Somewhat surprisingly is the result that these externalities are quantitatively unimportant in Europe, on average. The situations where negative spillovers really matter are rare and limited to Member States with intensive bilateral investment positions. Externalities also exist in consolidation policies, in particular due to the allocation of the consolidated profits to the Member States. Each country will try to get a large pie of the cake, the taxable profits, at the expense of others. The heterogeneity within the European Union matters for both tax-rate and tax-base reforms. For tax-rate reforms this heterogeneity likely outweigh the ‘limited’ spillovers, implying that this kind of policy should not be coordinated. For tax base consolidation, the situation is less clear cut, but strongly depends on a proper design. Finally, corporate tax base cum rate coordination, though strongly overruling national tax sovereignty, will likely improve welfare in the EU. In the remaining of this chapter, we investigate the economic effects of uncoordinated and coordinated changes in the tax rate and of consolidation of the tax base. Our tool in the investigation, a general equilibrium model for corporate tax policy in the European Union, is discussed in the textbox on CORTAX. In the concluding section, we return to the subsidiarity principle in somewhat more detail. Our line of reasoning: CORTAX The key economic mechanisms on the behaviour of enterprises, households and the government and on the functioning of markets are assembled in a general equilibrium model, named CORTAX. This model allows for a numerical assessment of the economic effects of tax reforms in the European Union. We summarize the main features of the model.2 CORTAX distinguishes 17 European countries (the EU15, with Belgium and Luxembourg combined and Poland, Hungary and the Czech Republic)
2
See Bettendorf and Van der Horst (2006) for more details on the model and its calibration.
15 Code of Coordination for Corporate Taxation
261
and the United States. It is calibrated for 2002. With the model, we investigate the long-run implications of corporate tax reforms, and abstract from the transitional dynamics. The model investigates the investment response of firms to changes in the tax rate or tax base. It includes the statutory tax rate and a measure for the tax deductions. Together, they determine the marginal effective tax rate (METR), which are taken from Devereux et al. (2002). The model distinguishes between domestic and multinational firms, which may be affected differently by tax reforms. Foreign direct investment by multinational firms is sensitive to differences in the effective tax rates between Member States. This is one important channel by which tax policies in one Member State affects the others. Profit shifting is a second channel: differences in the statutory rates induce multinational enterprises to shift profits to low-tax countries via transfer pricing. The third channel is foreign ownership, as firms are partly owned by households in other Member States. Corporate taxation affects the labour market as investment and labour demand are closely linked (the substitution elasticity between capital and labour is 0.7). A higher corporate tax therefore reduces the demand for labour and depresses wages. Households split their time between employment and leisure. They spend their after tax labour and profit income on consumption. The main sources of income for the government are the taxes on labour income, consumption and corporate income. The main expenditures are government consumption and income transfers to households. We assume in the simulations that budget deficits (or surpluses) are financed with a change in the labour income tax rate.3
15.2 Competition in tax rates Capital market integration within the European Union has been successful. It brings about a superior allocation of capital over Member States by linking capital markets. But capital market integration also links national capital income taxes. In response, Member States have reduced their statutory corporate income tax rates in order to attract highly mobile paper profits of multinational firms. Should the European Union respond by coordinating tax rates? There might be scope for European coordination of CIT-rates if countries harm each other by unilateral tax policies. We investigate this ‘if’ by looking at the spillovers from unilateral reductions in the CIT-rate. In addition, we show how strong the incentives are for Member States to unilaterally reduce their CIT rate. Finally, we switch the focus from national poli3
The results in this chapter are qualitatively similar if the budget is closed with a change in the consumption tax rate. Outcomes are different with lump-sum financing which is a nondistortive source of financing.
262
Albert van der Horst
cies to European coordination, by investigating whether the EU benefits from a minimum CIT-rate, or from harmonization.4 Does unilateral tax rate policy harm or benefit other Member States? A tax rate reduction may benefit the foreign owners of domestically operating firms, whose after-tax profits increase. And it may benefit capitalexporting firms, if the tax rate reduction drives up the return to capital. This ‘if’ holds, however, only for countries which are large enough to unilaterally affect the world interest rate, which is quite unlikely even for large European countries. On the other hand, countries may attract foreign capital at the expense of others, as investments flow towards locations where it yields the highest net rate of return. Again, the size of the effect on other countries depends on the relative size of both countries: a tax rate reduction in say Finland may boost investment at home, but will hardly harm large countries like Germany and France. A second harmful spillover is profit shifting, where countries reduce their CIT-rate in order to attract highly mobile paper profits of multinational firms. Important means of profit shifting are transfer pricing, where MNEs manipulate the price for their intra-firm deliveries, and thin capitalization, where MNEs have excessive debt in high-tax countries in order to benefit from interest deductions. In order to see whether spillovers are on average harmful or beneficial, we simulate in CORTAX an unilateral reduction in the CIT rate. We run this simulation for each Member State separately and each time the question is whether other countries gain or lose. Figure 15.1 shows that the harmful spillovers dominate the beneficial spillovers, as the unilateral reduction in one Member State reduces welfare in the others (rest EU). The size of the effect is limited, though, up to 0.02% of European GDP for taxrate reductions in the United Kingdom and Germany, which are large countries with a highly distortive corporate tax system (as measured by the METR).
4
This section is based on Bettendorf et al. (2006), to which we also refer for references to the literature.
15 Code of Coordination for Corporate Taxation
263
Fig. 15.1. Welfare effects of unilateral CIT-rate reduction
Why do we find that spillovers are so small? It is not because profit shifting is unimportant: it is, according to the empirical literature on which we have calibrated our model. It is not because FDI does not respond to tax rates: it does.5 However, an increase in FDI in one country hardly goes at the expense of others, given the large supply of foreign capital from other countries, both European and abroad. Only if all EU Member States jointly reduce their tax rates, a serious repercussion on FDI in non-EU countries might obtain. The final reason for the limited spillovers is the relatively small share of foreign direct investment (less that 10% on average) in total investments. The gains from the CIT-reduction primarily accrues to domestic firms, that do not have the possibility to transfer profits or expand foreign direct investment. Do Member States themselves gain from a tax-rate reduction? Although one might expect the answer to be affirmative, the case is not that clear cut. On the benefit side, the losses for the other Member States reflect a gain for the tax-reducing country: it benefits more from profit shifting than it loses from capital exporting. In addition, the tax distortion on domestic and foreign investment is reduced. On the other hand, if a Member State decides to engage in tax competition and reduces its CIT-rate, then it must mend the resulting budgetary hole, either by cutting public expenditure or by increasing the burden of alternative taxes. Alternative means of financ5
See also De Mooij (2005) for a survey of the literature on the responsiveness of profits and FDI to corporate taxation.
264
Albert van der Horst
ing, like raising taxes on labour income or consumption, likely dampen the gains from unilateral CIT-reforms. Again, we rely on simulations to quantify which effect dominates in the domestic effects of cutting corporate tax rates. As before, we concentrate on the situation where a single Member State reduces it tax rate, as if other Member States do not respond. Now the answer is: it depends, and it depends primarily on the alternative measures to close the budget of the government. If the government has the option to save on lump sum income transfers to households, where lump sum means that the transfer is unrelated to households’ income or wealth, then each Member State benefits from its own reduction of the CIT-rate. In a more realistic scenario governments have to raise the labour tax rate to compensate the revenue loss of the corporate income tax. This tax reform benefits only countries with a highly distortive corporate tax system, i.e. with a high tax rate and/or a broad tax base (see ‘home country’ in Figure 15.1). Examples are Germany, the Netherlands (before the recent tax reforms) and Belgium. Only these countries have an incentive to reduce their CIT-rate. Countries like Ireland and several Eastern-European Member States with low tax rates have nothing to gain from a CIT reduction. For them it is better to cut the labour tax rate than the tax on corporate income. Will there be a race to the bottom? The previous discussion shows that, starting at the current tax rates, not all countries benefit from cutting tax rates. Supplementary simulations (Bettendorf et al. 2006) reveal that even the Member States which do benefit from a tax-rate reduction will not completely abandon the tax on corporate income. At lower CIT-rates, the distortions in the alternative taxes on consumption and labour exceed the distortive effects of the corporate income tax on investment and profit shifting. So, no Member State will unilaterally abandon its tax on corporate income. Things might change, however, if countries respond to each other. Suppose that all other countries cut their corporate tax rate, should the remaining country respond by cutting its tax rate too? The answer is yes, but only slightly. Yes, because the negative spillovers from tax-reforms in other countries can be undone by cutting ones own tax rate too. Slightly, because domestic reasons for taxing corporate income (i.e. generating revenues which should otherwise be raised differently) prevent a strong response. The answer is therefore, there will be a race, but not to the bottom. Given that the spillovers are limited and the incentives for unilateral tax rate reduction are small, is there any room for coordination? Table 15.1 shows the welfare effects for the EU and a selection of countries of four coordination policies. The first row recalls the welfare effects of unilateral tax rate reductions (in each Member State separately).
15 Code of Coordination for Corporate Taxation
265
Table 15.1. Welfare effects (%GDP) of tax rate coordination (in a selection of countries) IRE Unilateral tax reform (-5%-point) Multilateral tax reform (-5%-point) Minimum tax rate (30%) Harmonized tax rate (33%) Harmonized tax rate (20%)
-0.10 -0.22 -0.28 -0.46 -0.22
UK NETH FRA GER 0.05 -0.02 0.02 -0.08 -0.11
0.30 0.07 0.04 0.14 0.06
-0.07 -0.15 0.01 -0.02 -0.69
0.21 0.15 0.01 0.25 0.13
EU -0.04 0.00 0.00 -0.29
The second row shows the welfare effects of a coordinated reduction of the tax rate with 5%-point. The potential gains from a CIT-reform are significantly reduced if Member States take into account that other states may respond. In this case, the inability to attract foreign profit income reduces the benefits from favourable tax planning by multinational enterprises. This implies that for most countries a reduction in the tax rate creates a welfare loss. Only for countries with a highly distortive CIT-tax, like Germany, a reduction in the tax rate is beneficial even if other countries respond. Most countries will therefore restrain from tax competition if they realize that other countries likely follow their tax cut. The next three rows show the welfare effects of three coordination policies: introducing a minimum tax rate and harmonization of the tax rate at the current average (33%) or a much lower level (20%). The simulations clearly show that the EU will not benefit from tax rate coordination, but even lose if the harmonized tax rate is set ‘too low’. From an economic point of view, competition in tax rates is hardly worth pursuing at current levels of corporate-income taxation, and even less so at a lower level of taxation. Moreover, the spillovers are harmful but limited. Policies to remedy tax competition, like setting a minimum tax rate or even harmonising the CIT-rates, therefore hardly enhance growth and welfare in the European Union: the winners just gain enough to compensate the losers.
15.3 Tax base Companies operating across the internal market are hampered by tax obstacles such as high compliance costs for cross-border operations, transfer pricing and the lack of cross-border loss compensation. These obstacles are inherent in the current system of separate accounting (SA), where the corporate income of foreign subsidiaries is treated separately for tax purposes.
266
Albert van der Horst
In its 2002 Tax Communication, the European Commission (2002) proposed consolidation of the tax base as an answer to the inherent difficulties of separate accounting and large compliance costs. Consolidation implies that all taxable profits of multinationals enterprises (MNEs) are added up into a single base. The Member States have to make key choices on the definition of the single base (either according to the tax rules of the parent country or according to new European rules) and on the question whether the consolidated European base will be introduced instead of, or in addition to national tax rules. We shed some light on both issues from an economist’s point of view.6 Consolidation implies that the subsidiaries of an European multinational are treated as a single entity for tax purposes. This brings several gains, both for multinational enterprises and for governments. First, multinationals save on compliance costs, as they have to file only one (consolidated) corporate income tax return, where all affiliates are included. Second, cross-border losses are offset automatically with tax base consolidation. Currently this is not the case, as Member States may prevent a parent company from deducting from its taxable profits losses incurred by a subsidiary in another Member State. This differs from the treatment of resident subsidiaries whose losses may generally be deducted from the companies profits. Finally, consolidation makes profit shifting for tax purposes obsolete, as all profits are added up into a single tax base. This implies that transfer pricing, charging different prices for intra-firm exports than for regular exports, becomes redundant. Governments in high-tax countries are the main beneficiaries of the latter effect, as firms are unable to minimize tax payments by shifting profits to low-tax Member States. Consolidation also has several drawbacks, which crucially depends on its design. The largest gains from consolidation might be expected if all enterprises, both domestic and multinational, are treated equally. Proposals for consolidation which do not treat all firms equally may create an unlevel playing field which induces substantial restructuring both within and between Member States. An example of this is the EU-proposal of home state taxation, where firms have to make their tax declaration according to the rules of their home country. Clearly, domestic firms and multinational headquarters are treated equally, but unevenness is introduced between subsidiaries with different home states. This proposal gives preferential treatment to subsidiaries originating from Member States with a narrow tax base.
6
See Van der Horst et al. (2007) for an extensive analysis of the economic effects of consolidation.
15 Code of Coordination for Corporate Taxation
267
A second issue is whether consolidation is optional or compulsory for multinationals. Multinationals likely prefer the first option, where they may choose the tax system, either national or European, which suits them best. The disadvantages for governments are: they have to deal with two tax systems and their tax revenues will decline as firms will exploit the opportunity to minimize their tax payments. The third issue is the definition of the common base including the amount of tax deductions. If consolidation is compulsory, then a broad base benefits governments, whose CIT-revenues increase, but implies a higher tax burden for MNEs which have to cut down production. Simulations with CORTAX reveal that the latter effect dominates, such that welfare in the EU declines if the tax base is broadened. The combination of a voluntary system and a broad base, however, likely implies that very few MNEs will shift towards the common consolidated base, which would make the reform superfluous. One of the most important issues is that a common consolidated base limits the tax authority of Member States, in particular from a subsidiarity point of view. As a solution, Member States are allowed to tax a fraction of the consolidated base at their own rate. This requires the apportioning of the consolidated base to the Member States, presumably with some kind of apportionment formula. Formula apportionment is a way to distribute the tax base between the Member States. Some measure of economic activity is used to determine which fraction of the consolidated base is generated in each jurisdiction and may therefore be taxed by each jurisdiction. Formula apportionment creates new tax planning possibilities for MNEs. Tax planning is the ability of firms to minimize their tax obligations by shifting profits or economic activity across jurisdictions. Transfer pricing, the most common means of tax planning in the current system of separate accounting, will become meaningless with the consolidation of the tax base. However, with formula apportionment, the share of the tax base apportioned to each jurisdiction can be influenced by shifting economic activity from one jurisdiction to another. Even though real economic activity, like production or FDI, can be shifted less easily than paper profits to other Member States, its economic impact is larger. The change in the tax planning strategy of MNEs, by reallocation instead of transfer pricing, therefore reduces welfare in the EU. Governments are likely to respond to the tax-planning strategies of MNEs by cutting their tax rates. In the current system of separate accounting, countries may thus attract paper profits and FDI. In the consolidated system with formula apportionment, the possibilities for (paper) profit shifting are limited, but the incentives for FDI are enforced.
268
Albert van der Horst
Adopting a system of consolidated base taxation with formula apportionment allows the Commission to achieve its goals in a single stroke (Martens-Weiner 2006): corporate taxation is simplified for multinationals and autonomy in fiscal policy is guaranteed for governments. Is consolidation a good policy option from a subsidiarity point of view? For an answer to this question we again rely on simulations with CORTAX. We first investigate a scenario where MNEs adopt a common consolidated base whereas domestic firms stick to the national tax rules. The tax allowances in the common base are fixed at the current EU average. The apportionment formula is defined by three factors, namely employment, capital and production of MNEs in each Member State, with equal weight. A summary of the economic effects in a selection of countries is shown in Table 15.2. Table 15.2. Economic effects of a common consolidated tax base (reported for a selection of countries) IRE Corporate tax revenue (%GDP) -0.37 Labour tax rate (%-point) 0.20 Wage rate (%) 0.79 Employment (%) 0.28 Capital (%) 0.52 GDP (%) 0.44 Welfare (%GDP) 0.06
UK -0.14 0.01 0.36 0.06 1.04 0.27 0.18
NETH FRA
GER
EU
-0.26 0.23 0.51 -0.09 1.45 0.33 0.21
-0.56 0.48 0.95 0.03 2.62 0.76 0.12
-0.07 0.11 0.18 -0.01 0.51 0.10 0.02
0.01 0.19 -0.08 -0.17 -0.29 -0.19 -0.11
Economies of scale are clearly present in the compliance costs. It is cheaper for multinationals to fill in one tax return for the EU than many different tax returns for each Member State. The size of this scale effect is, however, hardly investigated. The European Commission (2004) reports evidence on perceived compliance costs (these include costs required for company taxation and VAT, next to costs voluntarily incurred to minimize taxes). Compliance costs are estimated at 1.9% and 30.9% of taxes paid by large firms and SMEs, respectively. Costs are larger for firms with subsidiaries. In the simulations with CORTAX, we assume that the compliance costs of subsidiaries are eliminated by tax-base consolidation, which incurs a welfare gain of 0.04% of GDP in the EU. Tax planning has shown to be a key externality in tax rate competition, though quantitatively not very large. Consolidation eliminates tax planning via transfer pricing, but creates new opportunities via formula apportionment. Firms will expand production or sales in Member States with low statutory rates. For example, FDI will increase with 25% in Ireland, but will decline with 5% in Belgium & Luxembourg and the Netherlands. This
15 Code of Coordination for Corporate Taxation
269
reallocation aggravates tax competition, which is stronger in a consolidated tax system than in the current system of separate accounting. The economic effects of consolidation with formula apportionment are unevenly distributed, both between and within countries. With separate accounting, low tax countries are attractive for the location of paper profits. With formula apportionment, however, low tax countries are attractive for the location of production (and production factors): higher production in low-tax countries enlarges the apportioned share of the tax base in these jurisdictions and thus reduces the average tax payments of MNEs. This expansion of MNEs implies an increase in GDP, employment and capital in low-tax countries. In contrast, production in high-tax countries declines. This uneven distribution of gains and losses due to formula apportionment adds up to the unbalanced impact of a common consolidated base. Currently, the tax deductions differ significantly between Member States. For countries with generous tax deductions, like Italy and Greece, a common European base is likely to imply a broadening of the tax base. This raises their effective tax rate and suppresses investments. The opposite likely holds for Member States with limited tax allowances, like Germany and Ireland. Both the change to formula apportionment and the asymmetry in the tax bases imply that the change in welfare ranges between a reduction of 0.4% of GDP (in Greece) and an increase of 0.4% of GDP (in Belgium & Luxembourg). From the point of view of subsidiarity, we have three affirmative answers: there are economies of scale, externalities exist and Member States are heterogeneous. The latter might possibly be overcome by redistribution if the total gains for the EU are positive. However, Europe hardly benefits on average from the common consolidated base taxation. The gains from a reduction in compliance costs and the elimination of transfer pricing are offset by the efficiency losses from reallocation. Corporate tax revenues decline on average by about 2% due to the expansion of firms in Member States with low tax rates and/or narrow tax bases. Alternative means of financing have to be found in order to balance the government budget. The resulting gains in GDP and welfare are small, respectively 0.05% and 0.01% of GDP. This shows that the gains from consolidation depend on the details of its design.
270
Albert van der Horst 1.5 %
1.0
0.5
0.0
-0.5
-1.0 low
high
small
broad
EU
tax rate
tax rate
tax base
tax base
average
Fig. 15.2. Impact of EUCIT on GDP (black bar, %) and welfare (grey bar, %GDP)
The full benefits from consolidation can only be reaped if all firms participate and apply to a common tax base. Moreover, the incentives for reallocating production are minimized if the apportionment formula resembles the distribution of corporate income of MNEs. Much better, however, is to pull the tax-planning issue out by the roots, by harmonising the tax rates in addition to consolidating the tax base. In this far-reaching scenario, known as the European Union Corporate Income Tax (EUCIT), a welfare gain of 0.14% GDP can be obtained, as shown in last column of Figure 15.2.7 Still, the variation in economic effects is large, as is indicated by the other columns of Figure 15.2 In particular, countries which switch from a small base to the common base (indicated by ‘small tax base’) will lose, up to 1% GDP in Greece, but the welfare loss is much smaller. On the other hand, countries with a broad base, like Germany with 0.8% GDP, tend to gain. The distribution of winners and losers does not depend clearly on the initial tax rate: GDP and welfare increase in both low-tax and high-tax countries. Consolidation in combination with tax-rate harmonization might pass the subsidiarity test if it is complemented with a proper redistribution scheme. Still, Member States would have to agree on the design of the common base and would be prepared to pass over their ability to tax corporate income to the European Union. 7
Figure 15.2 presents the GDP- and welfare effects of harmonising the tax rate and tax base at the EU-average. The low-tax countries have tax rates below the EU-average and the small-base countries have a narrower-than-average tax base. The results in Figure 15.2 are the weighted averages of the low tax countries, etc.
15 Code of Coordination for Corporate Taxation
271
15.4 Subsidiarity in corporate taxation Should Member States give up their right to design their corporate income tax? We pick up this question, by confronting corporate taxation with the subsidiarity test (Ederveen et al. 2008). The first question is: are there economies of scale in corporate taxation? Indeed there are, in particular in the consolidation of the tax base. With separate accounting, firms have to file tax returns in each country in which they have subsidiaries. With consolidation, firms save on these fixed costs by filing one return for the European Union. How important are the externalities from corporate taxation? The section on tax-rate reforms extensively discusses this issue, showing that the beggar-thy-neighbour externalities of a tax rate reduction dominate. By cutting taxes, Member States are able to attract profits and foreign direct investment (partly) at the expense of other countries. However, the size of these spillovers is limited, and do not justify coordination of CIT-rates. Quantitatively more important are the spillovers from tax planning with formula apportionment. Countries have a stronger incentive to underbid each others tax rates, in an attempt to attract investment, production and (possibly) employment from other Member States. The third element in the subsidiarity test is the heterogeneity to which national policies might be adapted. Several elements of this heterogeneity are included in the analysis of this chapter. First, the analysis of tax base consolidation clearly shows that its gains depend on the heterogeneity in tax rates: low-tax countries tend to gain more. Second, the gains from an unilateral reduction in the tax rate depends on the openness of a country: Member States with strong foreign investment linkages have more to gain, but also more to lose, from tax rate reforms. Openness affects the size, rather than the direction, of the economic effects of tax reforms. The final element of heterogeneity that proves to be important is the different starting situation.8 For example, a common base generally benefits Member States which now have a broad base, but harms Member States with a narrow tax base. This difference in the initial situation not only determines the distribution of the economic effects, but might as well reflect national preferences for corporate tax policy or might depend on local circumstances and economic distortions. Probably, an egalitarian society may prefer a broad tax base and/or a high tax rate, whereas a liberal society may choose the opposite in an attempt to boost economic growth. 8
This list of heterogeneity is not complete. The contribution by Pouget and Stéclebout-Orseau (2007) points at differences in public infrastructure, which may be (partly) financed by corporate income taxes.
272
Albert van der Horst
Therefore, there are insufficient reasons to coordinate the CIT-rate at the European level, unless tax bases are consolidated in the first place. Whether or not consolidation should be introduced is a matter of choice between potential economic benefits and delegation of political autonomy: we show that compulsory consolidation in combination with tax-rate harmonization is beneficial, but overrules the primacy of Member States in corporate tax policy. An attempt to run with the hare and hunt with the hound by consolidating the base but leaving tax-rate policy at the Member States, is unlikely to boost welfare in the EU.
References Bettendorf L, Gorter J, Van der Horst A (2006) Who benefits from tax competition in the European Union? CPB Document 125 Bettendorf L, Van der Horst A (2006) Documentation of CORTAX . CPB Memorandum 161 De Mooij RA (2005) Will corporate income taxation survive? In: De Economist vol 153: 277–301 Devereux M, Griffith R, Klemm A (2002) Corporate income tax: reform and tax competition. Economic Policy 35: 450–495 Ederveen S, Gelauff G, Pelkmans J (2008) Assessing subsidiarity. In: Gelauff G, Grilo I, Lejour A (eds) Subsidiarity and economic reform in Europe: 19–40, Springer European Commission (2002) Company taxation in the Internal Market, Communication COM(2001) 582 final European Commission (2004) European Tax Survey, Taxation Papers 3 European Commission (2006) Implementing the community Lisbon programme: progress to date and next steps towards a Common Consolidated Corporate Tax Base (CCCTB), Communication COM(2006) 157 final Martens-Weiner J (2006) Company tax reform in the European Union: guidance from the United States and Canada on implementing formula apportionment in the EU, Springer, New York Pouget F, Stéclebout-Orseau E (2008) Corporate income taxation and the subsidiarity principle. In: Gelauff G, Grilo I, Lejour A (eds) Subsidiarity and economic reform in Europe: 273–290, Springer Van der Horst A, Bettendorf L, Rojas-Romagosa H (2007) Will corporate tax consolidation improve efficiency in the EU? CPB Document 141
16 Corporate Income Taxation and the Subsidiarity Principle
François Pouget and Eloïse Stéclebout-Orseau
16.1 Introduction
1
According to the political economy literature, the allocation of tasks between different levels of government should follow the basic principle of fiscal federalism (Musgrave 1959, Oates 1972). This principle also applies to “international unions” whereby some countries provide together certain public goods or policies. In particular, the assignment of responsibilities to the supranational level should reflect a trade-off between efficiency gains induced by delegation and the costs resulting from the implementation of a single policy between heterogeneous countries (Alesina et al. 2005 for a theoretical analysis; Alesina et al. 2005 for a more applied analysis). This theoretical framework is embodied in the European concept of subsidiarity since European governance is characterized by a complex system of overlapping jurisdictions whose design constantly requires improvements. Because fiscal policy remains under the responsibility of the national decision makers, discussions on the potential application of the subsidiarity principle are restricted to the issue of coordination. Under these circumstances, coordination is needed if and only if decentralized decisions produce significant spillover effects. Nonetheless, this basic guideline is incomplete, as a large range of alternative arrangements is available. On 1
The opinions expressed herein are those of the authors and do not necessarily reflect those of the European Central Bank. We would like to thank Alessandro Turrini, Jean Pierre Vidal, Hubert Kempf and participants of the Subsidiarity and Economic Reform conference for their useful comments.
274
François Pouget and Eloïse Stéclebout-Orseau
the one hand, coordination could take the form of strong restrictions on national actions (in the most extreme cases, a one-size-fits-all harmonization policy). On the other hand, coordination could be softer and mostly informal. Furthermore, the assessment of interactions raises additional questions. When decentralized policies create large negative spillovers, strategic interactions between national policies have to be controlled, but this may be hindered by uncertainties in the measurement of these interactions and difficulties in determining their underlying nature. This is especially problematic as the characteristics of the spillover effects may also affect the form of coordination required. Therefore, both a clear identification of the forms of coordination available as well as a good understanding of negative spillovers at stake is required. It may therefore come as no surprise that answers might differ according to the type of fiscal issue considered. Focusing on corporate taxation, the conclusions of a functional subsidiarity test (Ederveen et al. 2008) could at first sight appear non-ambiguous. The liberalization of international capital flows, fostered by the completion of the Single Market, has dramatically increased pressure on European governments to compete for a highly mobile tax base. Not surprisingly, a race to the bottom has occurred: Figure 16.1 displays a striking downward trend in statutory corporate tax rates in the last twelve years in Europe. top statuory corporate tax rate, % (simple average)
40
35
30
25
20
15 1995
1996
1997
1998
1999
NMS-10
2000
2001
EU-25
2002
2003
2004
2005
2006
EU-15
Fig. 16.1. Statutory corporate tax rates in the European Union between 1995-2006 Source: Eurostat.
16 Corporate Income Taxation and the Subsidiarity Principle
275
Negative spillovers can be easily identified: in order to maintain or even attract larger corporate tax bases, countries engage in a non-cooperative game leading to lower tax rates. The underlying mechanism is well described by the traditional tax competition literature initiated by Oates (1972) and developed more formally by Zodrow & Mieszkowski (1986) and Wilson (1986; 1999). The main harmful implication of this mechanism would be a decline in corporate tax revenues, which in turn would lead to an under-provision of public goods, and in this sense low tax rates are considered sub-optimal. In order to restrain these strategic interactions, it has been argued that some form of coordination is desirable, which could for instance consist in harmonizing corporate tax rates between EU countries. However, there is no broad consensus on the impact of the negative spillovers involved in this race to the bottom phenomenon. Particularly the impact on public policy is unclear. As Cnossen (2008) points out, despite the fact that tax competition exists,2 the decline in statutory corporate tax rates has been accompanied by a base broadening strategy in most Member States, leaving the ratio of corporate tax receipts as a percentage of GDP almost unchanged. Besides, since corporate tax receipts represent around 8% of total tax revenues in EU countries, their impact on the level of total public spending may be insignificant. By contrast, these spillover effects are likely to affect the composition of public spending. According to Keen & Marchand (1997), tax competition leads to a systematic bias in favour of “productive” spending. This can be explained by the dual nature of such competition. Indeed, Bénassy-Quéré et al. (2005) show that foreign direct investment flows do not only depend on tax rates but also on a certain category of public spending that enhances capital productivity. As a result, tax competition may be more than one-dimensional (i.e. only focus on tax rates). The starting point of our analysis is therefore to consider that the theoretical framework provided by the traditional theory of tax competition may prove irrelevant particularly when it is applied to corporate taxation. In other words, our analysis suggests that corporate tax competition could be a more complex phenomenon than a race to the bottom in terms of corporate tax rates and its direct negative consequences on total public good provision. As a result, the strategic interactions could turn out to be more difficult to analyse than expected. In the next section, we give an overview of the European Commission’s past interventions on these issues. We then present the main findings of 2
For empirical evidence, see Krogstrup (2006) who finds a robust negative relationship between the degree of capital mobility and corporate tax burdens in thirteen European countries. See also Devereux et al. (2004).
276
François Pouget and Eloïse Stéclebout-Orseau
our model of corporate tax competition. This model will help us to understand how countries compete in order to attract a larger corporate tax base. Following these results, the last section provides some policy conclusions and areas for further work.
16.2 Corporate taxation coordination in the European Union Corporate tax competition has been a long-standing policy issue in Europe. The first report on this topic, the Neumark Report, was published in 1962 at a time when statutory tax rates harmonization was at the core of the debate. This question remained topical for decades and again in 1992, the Ruding Report advocated a minimum statutory tax rate of 30% in the EU. Partly because of the reluctance among Member States to give away their fiscal sovereignty, all the attempts to harmonize corporate tax rates at the European level have failed so far. It is indeed a politically highly sensitive issue between high and low tax countries.
Fig. 16.2. Top statutory corporate income tax rates in EU countries in 2006 Source: KPMG 2006.
However, the debate is not closed and may re-emerge through the intergovernmental political game. Figure 16.2 gives us a clear picture of the current situation: the gap between ‘old’ and ‘new’ Member States is quite striking and can be a potential source of conflict. A good illustration of such tensions was a television interview in 2004 in which the French fi-
16 Corporate Income Taxation and the Subsidiarity Principle
277
nance minister strongly criticized the level of corporate tax rates in the new Member States and threatened to propose that countries below the European average should not be eligible for EU Structural Funds. In this context, one cannot exclude that pressure for some harmonization of corporate tax rates may arise from the high tax countries in the future. Since the 1990s, the European Commission has modified its strategy and no longer advocates corporate tax rate harmonization. The EU took an initiative in 1997 against “harmful tax competition”, which concerns tax privileges for foreign investors and specific industries. The Code of conduct adopted in 1999 aims at foreclosing potentially harmful features in tax regimes (Primarolo 1999). In 2001, another initiative was launched by the Commission who now seems to favour less binding and constraining methods (European Commission 2001). Several attempts have been made to progress on harmonising the corporate tax base in Europe. Among the options proposed, the ambitious project of the adoption of a common consolidated corporate tax base was launched. It was followed by the creation of a Working Group set up after an informal ECOFIN meeting in September 2004. This Group provides a useful expertise for the Commission on this highly technical and political issue. It was initially set up for three years and the Commission is expected to make a proposition after this period. In a recent communication,3 the Commission provides a short evaluation of the different points discussed so far. It appears however that no common agreement has been reached concerning some key issues, including the sharing mechanism determining the allocation of the consolidated tax base between Member States. This new strategy of the European Commission is justified by the need to achieve the full completion of the Single Market. The EU remains characterized by a patchwork of separated national corporate tax systems (Cnossen 2008 provides an extensive description). Although this complex system is organized by tax treaties in order to avoid double taxation (based on OECD and EU guidelines), firms still need to comply with each particular domestic tax code in the countries where they operate. The associated compliance costs are an obstacle to the establishment of a wellfunctioning single capital market. But this new strategy is not free from tax competition. Indeed, under the current regime, a multinational firm is required to establish separate accounts for its activity in each country. This obviously provides strong incentives for firms to adopt strategies between the different corporate tax systems in order to reduce their overall tax burden. In theory, international transfer prices for a given service or product between the national entities 3
For the recent developments see European Commission (2007b).
278
François Pouget and Eloïse Stéclebout-Orseau
of a multinational company must be set at arm's length (i.e. the price that an independent party would pay), but separate accounting systems facilitate their manipulation. In particular, profit can be shifted away from high tax countries by attributing higher expenses to the respective local entities. Management fees or royalties between the parent and the subsidiaries are other ways to shift profit for tax evasion purposes. Such profit shifting operations undoubtedly foster tax competition between Member States. For obvious reasons, no comprehensive measurement of profit shifting is available but it is likely to play a significant role in international tax base mobility. For instance, Bartelsman & Beetsma (2003) have calculated that more than 65% of the additional tax revenue from a unilateral corporate tax increase is lost because of profit shifting. Besides, 60% of international trade takes place among multinational firms, which represents a huge potential volume of profit mobility. Some progress has been made with the creation of a code of conduct on transfer pricing4 and a move towards increased coordination of tax systems between Member States. However, we are still far from a full tax base consolidation at the EU level. The recent focus on tax bases rather than tax rates indicates a change in strategy for the European Commission. This tells us that no clear agreement on a method of coordination has emerged so far. This uncertainty may be linked to the complex nature of the strategic interactions involved: as profits can be easily shifted "on paper" from one country to another, the geographical distribution of the multinational activity does not by itself fully explain the race-to-the-bottom mechanism.
16.3 Analysing strategic interactions We believe that a good understanding of the way countries behave in an environment of tax competition is needed in order to shed light on the spillover effects involved. We present below the main intuitions of our model of tax competition (Pouget and Stéclebout-Orseau 2008). In this section, we will first introduce the major hypotheses that we consider to be the key elements of corporate tax competition in Europe. We then expose the expected outcomes derived from these assumptions. The elements included in our analysis are not new in essence and they can be found in the rich literature quoted below, but we consider that their combination is a step forward for a better understanding of this issue.
4
See European Commission (2007a and 2005).
16 Corporate Income Taxation and the Subsidiarity Principle
279
The theoretical framework consists of a single representative firm operating in two competing countries. The firm can freely decide where to allocate its capital and its profits between the two jurisdictions. In order to finance their public policies, governments in the two countries rely on the taxable profits of the representative multinational firm, along with other tax revenues considered exogenous in the model. The environment in which tax competition takes place is simplified, as in any model, but it hopefully contributes to providing a clearer insight of the current situation in Europe. It is based on three major assumptions that we consider to be critical for the analysis of corporate tax competition. Assumption 1 (tax base mobility): Capital is perfectly mobile whereas it is costly to shift profits from one country to another. This assumption gives a description of corporate tax base mobility, which includes profit shifting, in line with Van der Horst (2008). As pointed out by Devereux et al. (2004), the vast majority of the tax competition literature focuses on the role of capital mobility, thereby ignoring the role of paper profit in the determination of tax rates. Since profit mobility is likely to affect the determination of tax rates, we put more emphasis on this aspect. The geographical distribution of the multinational firm activity is simply described by capital allocation in the two countries (we disregard sales and payroll for simplicity). Perfect capital mobility is assumed to be ensured by a well-functioning single market among both countries. By contrast, national accounting systems remain separated. Moreover, as in Kolmar & Wagener (2006) and Sorensen (2004), a profit shifting variable is included. Tax evasion through profit shifting is costly: distorting transfer prices have to be justified to tax authorities, which entails operational costs such as the cost of hiring accountants. We assume that the cost of profit shifting is convex: the higher the amount of profit shifted, the more tax evasion becomes unaffordable. For instance, we could assume that taxpayers would put stronger pressure on governments to control profit movements more tightly. The double assumption of perfect capital mobility and imperfect profit mobility is fairly realistic: although economic integration has a direct impact on capital flows, this is not so much the case for accounting systems, which are mostly defined within national boundaries. In our analysis, this double assumption means that profit mobility, as measured by an inverted index of the cost of profit shifting, can be considered as an index of tax competition and will have a direct influence on the strategic interactions.
280
François Pouget and Eloïse Stéclebout-Orseau
Assumption 2 (tax competition): Tax competition is not onedimensional. Corporate tax competition may be a more much more complex phenomenon than a simple race to the bottom in terms of corporate tax rates with its consequences for public input provision. This idea draws on a study by Bénassy-Quéré et al. (2005) on the determinants of foreign direct investment (FDI) flows. They show that a so-called ‘public input’, defined roughly as a stock of public infrastructure that enhances capital productivity, has a significant impact on FDI flows. As a result, competition for the corporate tax base is likely to reflect this dimension as well. We assume, therefore, that the multinational firm's output in each country depends not only on private capital but also on a domestic public input or ‘productive’ public good provided by government (through public investment spending for instance). It can be interpreted as all the categories of public expenditure that increase the marginal productivity of capital, including for instance infrastructures such as highways and bridges. Under these circumstances, competition between countries turns out to be twodimensional, with both the level of tax rates and the amount of public investment being used to attract the mobile corporate tax base. Assumption 3 (society’s preferences): Public consumption depends on two “types” of public spending. In its vast majority, the tax competition literature assumes that the decision maker is a Leviathan whose objective is simply to maximize tax revenue. In such models, public resources collection and public good provision are two aspects of the same objective. In line with Keen & Marchand (1997), we clearly make a distinction between these two public finance issues. We assume that public resources are collected for two different “types” of public spending. We consider first public investment, which is not firm-specific and also concerns households. For instance, public infrastructures primary built as amenities for the corporate sector are also largely used by households (think of roads or IT infrastructures for instance). The second category of expenditure concerns “non-productive” or general public spending, which can imperfectly refers to public consumption. It covers all public spending with no direct productive purposes.
16 Corporate Income Taxation and the Subsidiarity Principle
281
% 65 60 55 50 45 40
Latvia
Cyprus
Estonia
Lithuania
Czech Republic
Malta
Hungary
Slovakia
Ireland*
Netherlands
Spain
Belgium**
Poland
Portugal
Greece
Slovenia
Italy
United Kingdom
Sweden
Luxembourg
Austria
Denmark
France
Finland
Germany
35
Fig. 16.3. Social security and health as a share of total public expenditure in 2006 (*: data for 2005; **: data for 2004) Source Eurostat.
We assume that governments’ relative preferences for these two types of public spending reflect the preferences of their constituencies. Figure 16.3 displays for instance the share of social security and health spending in total public spending for each Member State. These categories might have a positive effect on productivity in the long run but they cannot be used directly by firms. They can therefore be interpreted as a rough estimation of non-productive spending. We can see that the asymmetry is striking. It reflects a large set of economic and political factors, such as the level of GDP per capita, ageing-related costs and historical trends. The public finance issue we want to address is twofold: - Collection of public resources: each government decides on a statutory corporate tax rate. Other categories of taxes are included in total public revenue but are treated here as exogenous. - Allocation of public spending: each government also decides on the composition of public expenditure between the ‘productive’ and the ‘general’ spending. We have therefore two intertwined but well-identified public finance issues. Analysing how they combine sheds light on the nature of corporate tax competition. We assume that the governments set their public policy anticipating the multinational firm’s decision. Following backward induction, we first analyse the firm’s decision on capital and profit allocation. Under perfect mobility, capital is allocated so as to equalize its after-tax
282
François Pouget and Eloïse Stéclebout-Orseau
marginal productivity in the two countries. Consistent with the twodimensional nature of corporate tax competition, capital inflows in a particular country are decreasing with the tax rate and increasing with the stock of public input. As for the allocation of profits between the two countries, it depends on the tax rate differential (independently from public input provision) and profit flows decrease with the cost parameter of profit shifting. Anticipating the corporate sector’s decision, governments simultaneously and non-cooperatively choose their statutory tax rates so as to maximize their own tax revenue. Assumption 1 leads to equilibrium tax rates that are increasing with the cost parameter of profit shifting: the intensity of strategic interactions only depends on profit mobility. This outcome can be explained by the fact that governments anticipate the free-riding behaviour of the multinational firm that seeks to benefit from the public input provided in a given country without paying for it (by shifting its taxable revenue into the other country). This triggers a traditional race to the bottom in tax rates fostered by tax base mobility. In the extreme case where profit shifting operations are costless, tax evasion would force both countries to simply give up corporate taxation. Assumption 3 tells us that governments decide on the allocation of public spending according to the society’s relative preferences. Nevertheless, this public finance issue is not solved independently from the environment of tax competition (households want their relative preferences to be taken on board by the decision maker but their utility also depends on the level of public goods provided). More precisely, as in Keen & Marchand (1997), we expect a systematic bias in the composition of public spending in favour of the public input. This comes from the fact that distorting the society’s preferences “pays off” because, by providing a larger share of spending to public input provision, each country can attract a larger corporate tax base and thereby increase total public resources. This result holds as long as the public input is ‘productive’ and has the ability to attract capital inflows (Assumption 2). Understanding the consequences of tighter tax competition on the composition of public spending tells us how the two public finance issues are intertwined. In their theoretical analysis, Keen & Marchand (1997) have shown that tighter tax competition leads countries to spend even more on the productive public good. Our analysis suggests an opposite effect (provided that countries are perfectly symmetric).
16 Corporate Income Taxation and the Subsidiarity Principle
283
% 3.7
3.2
2.7
2.2
1.7
6
5
20 0 5
4
20 0 4
3
20 0 3
2
20 0 2
1
20 0
0
Italy
20 0
9
20 0
8
Germany
19 9
7
19 9
6
19 9
5
United Kingdom
19 9
4
19 9
3
19 9
2
19 9
1
19 9
0
19 9
9
19 9
8
19 8
7
19 8
19 8
19 8
6
1.2
France
7.3% 6.8 6.3 5.8 5.3 4.8 4.3 3.8 3.3
United Kingdom
Germany
Italy
6 20 0
20 0
20 0
20 0
20 0
1 20 0
20 0
0
9 19 9
8 19 9
7 19 9
6 19 9
5 19 9
4 19 9
3 19 9
2 19 9
1 19 9
0 19 9
9 19 8
8 19 8
7 19 8
19 8
6
2.8
France
Fig 16.4a. and 16.4b. The downward trend of public investment as share of GDP and general government spending Source: Eurostat.
The key for understanding this outcome is Assumption 2 which gives an important indication on the nature of strategic interactions. As profit mobility increases (i.e when profit shifting becomes more affordable), it mat-
284
François Pouget and Eloïse Stéclebout-Orseau
ters gradually more than capital mobility in determining the tax base. The relative ability of the public input to raise corporate tax revenue therefore declines, whereas paper profits can be easily attracted through tax cuts. We would therefore expect that tighter tax competition might entail a reallocation of public spending in favour of the ‘general’ public good. Such a mechanism can be easily understood in the extreme case of perfect profit mobility. The final location of the tax base then depends almost entirely on the location of profits and the public input has no impact on it. Therefore, a permanent bias in favour of the public input would simply not “pay off” and the government has no reason to deviate from society’s preferences. Surprisingly, an environment of tighter tax competition does not lead governments to rely more on productive spending in order to attract the mobile tax base. This seems consistent with the current pattern of public spending in the largest European countries and particularly the downward trend of public investment over the last decade. % 6.2 5.7 5.2 4.7 4.2 3.7 3.2 2.7 2.2 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005
United Kingdom
Germany
Italy
France
Fig. 16.5. Public spending on economic affairs as a % of GDP Source: Eurostat.
Since public investment may be an imperfect proxy for productive spending, we use the COFOG classification to isolate the share of public spend-
16 Corporate Income Taxation and the Subsidiarity Principle
285
ing on economic affairs in each of the four countries.5 We also clearly observe a downward trend. By contrast, during the same period, social security and health spending have significantly increased, as we can observe on Figure 16.6. Although Figures 16.5 and 16.6 give a rough estimate of the distinction between productive and non-productive public spending, the comparison with Figure 16.1 corroborates the intuition of a positive correlation between corporate tax rates and the share of productive public spending. % 31
29
27
25
23
21
19 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005
United Kingdom
Germany
Italy
France
Fig. 16.6. Public spending on social security and health as a % of GDP Source: Eurostat.
16.4 Policy conclusions and areas for further work We have seen in the previous section that corporate tax competition is likely to produce some spillover effects, not only on tax rates but also on the composition of public spending. We believe that this element should be taken into account when analysing the potential effects of tax rate harmonization, to the extent that the main objective of harmonization is to internal5
This ratio covers support programs, subsidies and spending on public infrastructures in industry (mining, manufacturing, agricultural, energy, construction, transport, communication and other services).
286
François Pouget and Eloïse Stéclebout-Orseau
ize the strategic interactions at the origin of a race-to-the-bottom mechanism. Corporate tax rate harmonization would lead most countries to increase their statutory corporate tax rates. Simultaneously, in line with the positive relationship we have exhibited in the previous section, we find that it would foster a re-allocation of public spending in favour of productive spending. As a result, tax rate harmonization might prevent a race to the bottom in tax rates, yet it would not rule out competition but simply change its focus. In other words, the reduction in paper profit shifting once tax rate differentials have disappeared would imply that the location of the corporate tax base would depend first and foremost on the location of capital rather than profits. This would lead countries to rely more on public input provision in order to attract the mobile tax base. Households could benefit from such a change in the pattern of public spending to the extent that productive public goods are not exclusively provided to firms. However, in the case of a high preference for general/non-productive spending (which could occur for instance within an ageing population or if there is a strong social demand for more redistribution), shifting the focus of corporate tax competition onto public input provision could produce ambiguous welfare effects. This question therefore requires further investigation: in order to capture the overall welfare effect of tax rate coordination, a general equilibrium framework would be required. Interestingly, this outcome – higher corporate tax rates and a larger share of productive public spending – holds when tax competition takes place among symmetric countries. Introducing asymmetry between countries would bring forth other issues. Asymmetry may stem from differences in the productivity of public investment or in stocks of public capital, which would affect the ability of governments to attract the corporate tax base through public spending. It is well known that larger countries can usually afford to set higher tax rates (Baldwin and Krugman 2004). Assuming that tax rate harmonization is politically feasible, we cannot exclude the case of a harmonized tax rate set at a level that would not be consistent with the structural characteristics of currently low-tax countries. Indeed, keeping in mind that tax competition is two-dimensional, tax rate harmonization would force countries to rely on public input provision in order to maintain their corporate tax base. As a result, harmonization could undermine the process of competition by preventing low-tax countries from using tax cuts as part of their fiscal strategy. This problem is well reflected in two conflicting views in Europe. The first view, expressed by high-tax countries, considers that tax competition is harmful per se and advocates tax rate harmonization. The second view puts forward compara-
16 Corporate Income Taxation and the Subsidiarity Principle
287
tive advantage arguments to justify the strategy adopted by low-tax countries. The scope of harmonization is also at stake. Due to the difficulties in implementing some form of tax rate harmonization, the countries willing to co-operate in this area could consider tax base harmonization, or even a tax base consolidation, as a second best option. Figure 16.7 below illustrates the two systems. On the left hand side, the traditional system of separate accounting is displayed: each country decides on the statutory tax rate (IJ) to be applied to its corporate tax base. The two accounting systems are connected through profit shifting operations (S) which have been identified as the major source of strategic interactions in the previous section. On the right hand side, we consider the extreme case of a common tax base. Under this system, the taxable profit of the multinational firm (Bu) is "allocated" according to an apportionment formula. In large federal states where this system is implemented (such as the US and Canada), this formula depends on one or several reliable indicators describing the geographical distribution of the activity of the firm (usually sales, payroll and the stock of capital invested in each jurisdiction). Here, the apportionment is simply denoted by Ș. Note also that the adoption of a common tax base does not necessarily require any tax rate coordination (although this issue can also be raised). Therefore each country independently decides on a tax rate and applies it on the fraction of the consolidated firm's profit that it receives.
IJA
corporate tax base country A
IJB
S
Multinational activity
corporate tax base country B
IJA
Ș
IJB
BU
1-Ș
Multinational activity
Fig. 16.7. Separate accounting and a common consolidated corporate tax base
288
François Pouget and Eloïse Stéclebout-Orseau
The common tax base appears to be an appropriate way to rule out profit shifting, and we could therefore expect that the strategic interactions displayed in the previous section would disappear. Nevertheless, we find that new types of strategic interactions could arise. Provided that capital flows affect the definition of the apportionment formula, a shift to a common tax base system could paradoxically increase the strategic use of capital movements, which in turn could result in a race to the bottom in tax rates. While paper profits cannot be shifted, firms may indeed arbitrage ex ante on the location of their activity. They would allocate capital not only on the basis of its marginal productivity but also in order to influence the apportionment formula and thus reduce the overall tax burden. The more emphasis the apportionment formula puts on capital flows, the more this strategic behaviour becomes crucial and the more capital allocation responds to tax rate differentials. As Van der Horst (2008) points out, the main drawback of consolidation arises from its design, and particularly the characteristics of the apportionment formula. In line with Gerard (2006), our model advocates a formula that focuses on factors that are not easily manipulated by the firm (for instance, sales depend broadly on demand and, as such, could be considered as a fairly exogenous factor). Besides, the shift towards tax base consolidation does not leave the composition of public spending unchanged. Our model exhibits an allocation pattern that directly responds to the partner country's tax policy decision. As in the case of tax rate harmonization, tax base consolidation therefore leads countries to compete more through productive public spending. One major public policy implication we can draw from our model is that any coordination scheme, either based on tax rates or tax bases, is likely to affect the composition of public spending. However, the implementation of a coordinated policy is subject to a number of political tensions that could impede its implementation, and corporate tax competition may get even tighter until an agreement is reached. Under these circumstances, the potential negative effects of tax competition on productive public spending have to be taken into account since this could represent a crucial issue for the Member States’ commitment to the Lisbon Strategy. If the spillover effects involved by tax competition cannot be controlled, some form of coordination of public investment policies might be required. In fact, the Lisbon agenda and the open method of coordination represent a step in this direction.
16 Corporate Income Taxation and the Subsidiarity Principle
289
References Alesina A, Angeloni I, Etro F (2005a) International unions. American Economic Review 95: 602–615 Alesina A, Angeloni I, Schuknecht L (2005b) What does the European Union do? Public Choice 123: 275–319 Baldwin R, Krugman P (2004) Agglomeration, integration and tax harmonisation. European Economic Review 48: 1–23 Bartelsman EJ, Beetsma R (2003) Why pay more? Corporate tax avoidance through transfer pricing in OECD countries. Journal of Public Economics 87: 2225–2252 Bénassy-Quéré A, Gobalraja N, Trannoy A (2005) Tax competition and public input, mimeo. CEPII working paper 2005–08 Cnossen S (2008) Coordinating corporation taxes in the European Union: subsidiarity in action!. In: Gelauff G, Grilo I, Lejour A (eds) Subsidiarity and Economic Reform in Europe: 243–258, Springer De Mooij R, Ederveen S (2003) Taxation and foreign direct investment: a synthesis of empirical research. International Tax and Public Finance 10: 673–693 Devereux MP, Lockwood RB, Redoano M (2004) Do Countries Compete over Corporate Tax Rates? Mimeo, University of Warwick Ederveen S, Gelauff G, Pelkmans J (2008) Assessing subsidiarity. In: Gelauff G, Grilo I, Lejour A (eds) Subsidiarity and Economic Reform in Europe: 19–40, Springer European Commission (2007a) Implementing the community programme for improved growth and employment and the enhanced competitiveness of EU business: further progress during 2006 and next steps towards a proposal on the common consolidated corporate tax base. Communication COM(2007) 223 European Commission (2007b) On the work of the EU joint transfer pricing forum in the field of dispute avoidance and resolution procedures and on guidelines for advance pricing agreements with the EU, Communication COM(2007) 71 European Commission (2005) Proposal for a code of conduct on transfer pricing documentation for associated enterprises in the EU Communication, COM(2005) 543 final European Commission (2001) Company taxation in the internal market, COM(2001) 582 final Gerard M (2006) Reforming the taxation of multijuridictional enterprises in Europe: A tentative appraisal. European Commission Economic Papers 265, Directorate General Economic and Financial affairs Keen M, Marchand M (1997) Fiscal competition and the pattern of public spending. Journal of Public Economics 66: 33–53 Kolmar M, Wagener A (2006) The role of the tax base in tax competition with formula apportionment. Mimeo, University of Vienna Krogstrup S (2006) Are corporate taxes racing to the bottom in the European Union? Mimeo, Graduate Institute of International Studies, Geneva
290
François Pouget and Eloïse Stéclebout-Orseau
Musgrave, RA 1959 The Theory of Public Finance: A Study in Political Economy, New York: Mcgraw–Hill Oates W (1972) Fiscal federalism, Harcourt Brace, New York Pisani-Ferry J, Von Hagen J (2003) Pourquoi l’Europe ne ressemble-t-elle pas à ce que voudraient les économistes? Revue Economique vol 54 no 3 Pouget F, Stéclebout-Orseau E (2008) Corporate tax competition and differentiated public goods, forthcoming Primarolo, D (1999) Code of conduct (Business taxation) http://europa.eu.int/comm/taxation_customs/resources/documents/primarolo_ en.pdf Sorensen PB (2004) Company tax reform in the European Union. International Tax and Public Finance 11: 91–115 Van der Horst A (2008) Code of coordination for corporate taxation. In: Gelauff G, Grilo I, Lejour A (eds) Subsidiarity and Economic Reform in Europe: 259– 272, Springer Wilson JD (1986) A theory of interregional tax competition. Journal of Urban Economics 19: 296–315 Wilson JD (1999) Theories of tax competition. National Tax Journal 52: 269–304 Zodrow GR, Mieszkowski P (1986) Pigou, Tiebout (eds) Property taxation and the under-provision of local public goods. Journal of Urban Economics 19: 356– 370
17 Subsidiarity in Regional Policy
Iain Begg
17.1 Introduction
1
Cohesion is one of only two areas, alongside the Common Agricultural Policy (CAP), in which the EU level of government/public administration plays a substantial role in formulating and delivering policy that requires significant public expenditure. Yet, although it accounts for around a third of the EU budget, cohesion policy expenditure is still a relatively small component of overall public expenditure in the Union at just under 1% of the aggregate public spending of all Member States. Moreover, the term ‘cohesion’ is open to differing interpretations. Manifestly, the EU level does not have exclusive competence in the policy areas which underlie cohesion, which comprise a combination of those aimed at supporting economic development in regions which are under-performing or less competitive, and mechanisms for the redistribution of income. In fact many of the major spending programmes of the national tier of government (and, depending on the fiscal constitution of the Member State, sub-national tiers) are intended to reduce disparities, including regional disparities in access to public services and real incomes. Formally, EU cohesion policy is not intended to redistribute current incomes, but to enhance the long-run competitive position of the regions and social groups it supports. Although the term cohesion encompasses more 1
This chapter draws on research financially supported by the European Union under the 6th Framework Programme through the DIME and EU-CONSENT Networks of Excellence. The author is grateful for this funding and also wishes to thank Arjan Lejour for helpful comments.
292
Iain Begg
than regional policy, narrowly defined (for a discussion of regional policy, see Armstrong and Taylor 2000), this chapter concentrates on the economic development component of cohesion policy and thus largely equates it with regional policy, understood as interventions designed to raise the level and/or rate of growth of economic activity and employment in regions assisted by the policy. It is important to recognize that regional policy so defined is a long-term policy and that its function is to create the conditions for improved future economic performance rather than to boost current incomes. A frequently-articulated argument against EU policy cohesion is that it is ineffective, with the implication that it ought to be re-thought and, conceivably, decentralized or renationalized (Tarschys 2003). Empirical findings on cohesion policy tend, however, to be inconclusive. Moreover, the ineffectiveness argument is as much (or more) one about whether spatially-targeted economic development policies per se are worthwhile, than about whether the EU is the appropriate level to offer them. Part of the subsidiarity question, though, is whether the approach adopted in EU cohesion policy is qualitatively worse than if it were left to national or regional/local levels of government. The chapter is structured as follows. Section 17.2 explores the rationale for having an EU regional policy, distinguishing between constitutional, economic and political arguments. In the next section, the effectiveness of policy is reviewed, then the wider context of cohesion policy is discussed in Section 17.4. Section 17.5 brings together the arguments for and against subsidiarity in regional policy, and concluding comments complete the chapter.
17.2 The rationale for EU regional policy In any economic space, there will be weaker and stronger territorial units in terms of economic performance and dynamism. Even the most prosperous cities typically contain economically weaker neighbourhoods characterized by high levels of social exclusion and deprivation, while other parts of the city thrive. Regions in relative decline or ascendancy are found in all Member States, with the disparities in performance often persisting over decades rather than being temporary aberrations. In the EU as a whole, there have been long-run shifts in the relative positions of Member States, on the one hand, and of classes of region, on the other (see the analyses in successive cohesion reports and periodic reports published by the Commission, as well as the literature on convergence, such as Tondl 2001). If
17 Subsidiarity in Regional Policy
293
there is then political agreement that some attempt should be made to limit the consequences of uneven economic performance, the policy will tend to be a top-down one in which the polity as a whole assumes responsibility. Such policy is justified normatively on grounds of solidarity, but may also reflect a search for spatial balance in economic development to ensure that the potential of the economy as a whole is realized (Kaldor 1970). The question then is how to define the encompassing polity and the challenge for subsidiarity is to determine where policy responsibility should ultimately lie for responding to these disparities. In turn, it is important to identify the differing sources of disparities and to relate them to determinants of the competitive positions of territories.2 To determine which level of government should be responsible for a policy response, it is necessary to ascertain what the source of the problem is. European integration, generally, will affect territorial performance, with gains for certain locations and losses for others occurring for diverse reasons, ranging from the impact of loss of internal borders to the much more profound economic processes flowing from shifting specializations. The new economic geography approach tries to model these issues by looking at factors such as the economies of agglomeration, the diseconomies of congestion or transport costs (for a survey, see Puga 2002). Top-down influences may be dominant in some cases. If the EU as a whole gains or loses certain sectors of economic activity in global markets (be it textiles and clothing, pharmaceuticals, tourism or financial services) or because of the expansion or attrition of domestic demand (personal services or coal mining), the territories relatively heavily specialized in the activities in question will tend, similarly, to gain or lose. Over time, there will be a response, as the economies adjust, but the evidence (for example from successive EU ‘Periodic Reports’ – for example: Commission 1995 – on regional development, which show that many of the ‘problem’ regions remain problematic for decades) is strong that such adjustment occurs only slowly. How various macroeconomic indicators evolve, such as the interest rate, the exchange rate or the fiscal position will also have uneven effects on different localities. With the advent of EMU, such changes are increasingly mediated at EU level. The level of economic development a territory has attained manifestly bears on performance, often because it tends to be associated with resilience to shocks. Relative advantage may be attributable to purely geographical features, such as central location, remoteness or physical separa2
The deliberately vague term ‘territories’ is employed here to cover different ways of delineating the economy in spatial terms. It can, therefore, embrace urban/rural, regional and national boundaries.
294
Iain Begg
tion from trading partners (islands, mountain areas). Finally there are effects on territorial economies from governance, including other policy impacts. 17.2.1 Assigning policy responsibility With so many different determinants of relative territorial advantage, coupled with the fact that there are challenging normative factors behind whether there should be a policy response, the choice of level of government to which to assign responsibility for regional policy in the EU is not easy. There is, first, an unambiguous constitutional case for locating at least part of regional policy at the EU level, on the grounds that cohesion is a fundamental EU-wide objective (set out in Art. 2, TEU) on a par – in constitutional terms - with the single market and the economic and monetary union. The mandate for a European policy is fleshed out in Art. 158, TEC which emphasizes the need to reduce ‘disparities between the levels of development of the various regions and the backwardness of the least favoured regions’ and Art. 159-161 which establish the legal base for the Structural and Cohesion Funds. The bulk of Structural Fund (SF) spending is targeted at low GDP regions, and is thus explicitly targeted at promoting convergence. In previous rounds of the SFs, this was known as Objective 1, and since 2007, explicitly labelled as a convergence objective. But a chunk of the money has consistently been reserved for other spatial aims. What was, for the last three rounds of the SFs called Objective 2, has been targeted at regeneration of economies afflicted by specific forms of decline, especially consequent on the loss of basic industries such as coal mining, textiles or other ‘old’ manufacturing. Resources have also been devoted to various kinds of territorial co-operation, most prominently among border regions. However, it could be argued that the constitutional case is not necessarily sound economics and it has been evident that EU regional policy has occurred for different reasons. The argument can be made in terms of fiscal and institutional capacity, as well as in terms of political will and choices. If the locus of problems is an entire (or much of a) Member State, the argument for the wider polity (the EU) to have some responsibility for finding remedies is more persuasive than if it is just a smaller territory. By contrast, if the Member State chooses not to implement an internal cohesion policy or to articulate its own priorities for such a policy, that is its prerogative, but it should not expect the EU level to intervene. Instead, given the overall EU commitment to cohesion, whether and how a Mem-
17 Subsidiarity in Regional Policy
295
ber State should be obliged to adopt some form of cohesion policy should be open to discussion. Thus, it can be argued that a city such as London, which is prosperous overall, should be able to deal with the acknowledged problems of deprivation in its worst affected boroughs3. France and Germany have the capacity to deal with their different sorts of regional disparities, whereas Portugal or Greece have, over the years, been less well-placed to do so. Because lesser fiscal resources are available domestically for economic development, a more compelling case can be made for the lower income Member States that acceded to the Union in 2004 and 2007. They tend to have lesser institutional capacity and, while the presumption is that lower income countries will have more rapid growth as they catch-up with richer Member States, this process is often accompanied by a widening of regional inequalities in GDP per head. Since cohesion policy tries to assure territorial balance in economic development, there are grounds for an EU policy that tries to mitigate the widening of disparities. There is however, a dilemma to be confronted. Aggregate Polish or Hungarian growth that achieves the best ‘catch-up’ trajectory may be best promoted by concentration of activity in the more favoured parts of those countries, with ‘trickledown’ relied upon to reach the rest of the country. But such an approach may result in the sort of enduring spatial imbalance evident – in extreme form – in Italy and – in possibly more transitional form – in other Member States such as Belgium or Germany. 17.2.2 Added value from EU interventions A key question is whether there is added value from EU regional policy. There cannot be an unambiguous answer, because in some cases what the SFs offer fills gaps in national policies, while in others it overlaps. Bachtler and Taylor (2003) argue that added value has both quantitative and qualitative dimensions. Assessing the added value of locating regional policy at EU level is further complicated because, de facto, the policy has multiple objectives, with the implication that the rationale for the policy is also diffuse. In principle, the purpose of cohesion policy is to promote the real convergence of GDP per head, an aim that implies that it is primarily allocative in intent. Yet there is also a redistributive element to the policy 3
London as a unit of government does not, in fact, command resources that would enable it to fulfil this role, but subsidiarity principles suggest that it should be up to UK electors or interests to make the case for change, rather than relying on the EU level.
296
Iain Begg
because it steers fiscal resources to lower income regions and thus represents a transfer of income. Public investment in the Member States and regions which benefit is underpinned by the Structural Funds, making it easier for levels of current consumption to be higher. Added value can be envisaged in different facets of the policy process, with ramifications for the discussion of subsidiarity. A distinction has to be drawn between the formulation of policy, its financing and its implementation, including the choice of policy instruments. In parallel, it is important to take account of externalities and spillovers: for example, decentralized policy may result in unfair competition between regions or in incompatibilities between aggregate and spatial objectives. Administrative costs are another facet of added value. It could be that the need to obtain support for the economic development of a region through the Structural Funds diminishes the impact because of the inevitable transactions costs of going through different tiers of bureaucracy, with ‘Brussels’ imposing conditions inappropriate for the region. An alternative standpoint is that an external body such as the Commission improves the quality of policy and obliges national or regional governments that might be susceptible to ‘capture’ by domestic interests to adopt more enlightened policies. The question of the time span over which regional policy is applied has a number of ramifications. Development is acknowledged to be a longterm aim subject to dislocations along the way, with the implication that consistent policy is required. Equally, policy interventions need to occur in the right sequence and have cumulative effects. In considering subsidiarity, it may be that different steps in the sequence should be undertaken by different levels of government, but there is also an argument that the EU level will be better able to engage in long-term coherent policies than national governments which are subject to shorter-term pressures. Endogeneities and learning-by-doing also have to be taken into account. Part of the reasoning of the UK Government (2003) for repatriating regional policy is that while the approach adopted in the 1988 reform of the SFs did improve national policy-making in the Member States that receive the bulk of Objective 2 funding, they have now assimilated the lessons and no longer need the external impetus. 17.2.3 Political economy From a political economy perspective there are various arguments behind an EU role in regional policy, starting with inertia. Although at least some of what is now EU regional policy might not be justified if policy were be-
17 Subsidiarity in Regional Policy
297
ing assigned afresh, it is difficult to shift policy assignments because there is path dependency reinforced by the presence of bureaucratic and other structures that take on a life of their own. As Alesina et al. (2002: 40) put it: ‘policy agendas are the result of bargaining processes, typically delegated to elected representatives or bureaucracies. Competing bureaucracies may not be willing to easily give up responsibilities, from which they derive prestige and influence’. An economic integration argument for regional policy is that because economic integration creates adjustment problems, it is at least partly incumbent on the EU level to assist regions which lose (relatively, if not absolutely) from integration. The irony in this position is that the most potent adjustment mechanisms in an economic space (labour and capital mobility, or induced fiscal flows) are much weaker in the EU than in most Member States or in comparable economies such as the US. Hence, because the economies of the EU are not that well integrated a regional policy may well be justified to counter the effects of integration. Put this way, the proposition seems almost contradictory, but if normative considerations such as a preference to curb intra-EU migration or a reluctance to countenance cross-border fiscal flows for equalization purposes are taken into account, regional policy as a means of facilitating adjustment could be seen as a plausible second-best policy. Political economy factors are also evident in the machinations around the EU budget, with the net contribution becoming of greater concern than whether or not there is a compelling case for EU regional policy being applied in each Member State. Richer net contributors to the EU budget, such as Germany and the UK, demand and obtain SF funding that helps them to contain the size of their net contributions, while poorer Member States – as Poland did in the last budget negotiations – hold out for bigger SF allocations for the opposite reason. A similar political economy argument is that the visibility of cohesion policy plays a valuable role in fostering support for EU regional policy and, indeed, the EU generally. If, say, Austrians are to be expected to support a policy that transfers sizeable resources to poorer neighbouring Member States, the fact that some resources flow to regions such as Burgenland can help to sell the policy in Austria, even though an ability to pay test might suggest that the policy should be funded domestically. Poll data also lend some support to the idea that a regional policy should exist at EU level, although there is plainly a risk that such polls will find support for what is already in place. Alesina et al. (2002) report a Eurobarometer study which shows that in all Member States, there is a balance of opinion in favour of EU involvement in regional policy. A more recent study by Ahrens et al. (2008) using Eurobarometer data up to 2004 also
298
Iain Begg
shows that the EU15 and ten new Member States have a unanimous preference for regional policy at the EU level.
17.3 Regional policy effectiveness A different perspective on whether or not there should be a regional policy comes from posing the simple question of whether it works. EU cohesion policy has been assessed from diverse standpoints and the range of theoretical approaches and empirical techniques used in work on evaluating cohesion policy is extensive (Bachtler and Wren 2006). 17.3.1 Empirical assessments Many of the empirical assessments of the SFs focus on the degree to which cohesion policy alters trajectories of convergence in the sense pioneered by Barro and Sala-I-Martin (1991, 1992). A distinct, but related approach is growth regressions: econometric work which, essentially, take some measure of economic development (usually either the level or the growth of GDP) as the dependent variable and includes SF expenditure among the explanatory variables. A third class of studies uses simulation models to try to capture the full range of effects of the SFs, recognizing that spillover effects of policies can be crucial. By contrast, much of the evaluation effort – including large-scale studies of SF programmes - has concerned how cohesion policy has been implemented and whether or not policies have been well-conceived. Although some evaluation studies arrive at tolerably favourable conclusions about the SFs, they interpret effectiveness in different terms from more narrowly based econometric studies. Not surprisingly, the Commission argues that cohesion policy has been a success, for example in the 3rd Cohesion report (Commission 2004) which notes that growth per capita in the cohesion countries (even excluding Ireland, an extreme outlier) exceeded the Community average by one percentage point over the period 1994-2001. The OECD (2004: 153) observes, however, that ‘it is difficult to assess empirically the impact of the EU regional funds, and the evidence that the Union’s regional policy succeeds in its aims is mixed’. Molle (2006) points out that EU policy can be adjudged, on the whole, to have been appropriate, insofar as it has focused on infrastructure and human resources, the two elements of backward regions that he considers need most attention. He is more dubious about whether policy has been effective in promoting convergence and believes
17 Subsidiarity in Regional Policy
299
that resources have often been poorly used, citing eligibility criteria that have distorted project choices and resulted in sub-optimal investments. Econometric studies
Efforts to appraise the SFs using econometrics have emanated from a number of different theoretical perspectives and, not surprisingly, arrive at different empirical results (see, for example, Tondl 2001; Beugelsdijk 2002; Magrini, 2003). Boldrin and Canova (2001, 2003) assess the impact of transfers under the Structural and Cohesion Funds and find hardly any effects. An obvious problem with any econometric approach, though, is that the lags involved in economic development are uncertain and it may even be the case that some initial impacts are perverse. A second strand of work draws on new economic geography (NEG), with several recent studies appraising the effectiveness of the SFs from this perspective. The question posed tends to be not just whether there has been an effect, but also whether the changes induced by the policy support help to steer the recipient economy in an appropriate direction and whether the effects on the Member State and EU economy as a whole are positive. Some authors have argued that EU policy should concentrate on convergence at the national level, leaving domestic cohesion policies to deal with intra Member State disparities; thus, de la Fuente (2004) finds that Spain would have had greater overall welfare gains if SF investment had been targeted at dynamic regions, rather than the least prosperous ones. Midelfart-Knarvik and Overman (2002), in a study of the effects of structural support for less-advantaged regions, find that the regions that have adapted most successfully have increased their specialization within manufacturing. They find, however, that policy ‘is encouraging relocation counter to comparative advantage’, but also that ‘regional comparative advantage…is severely restricted by the fact that factor price returns tend to be equalized within nation-states due to the centralized nature of wage-setting’ (Midelfart-Knarvik and Overman 2002: 351). This raises a central issue about what the SFs are intended to do that also bears on subsidiarity. If it is accepted that an aim of policy is to facilitate specialization along NEG lines, these findings could be interpreted as evidence that the SFs are misconceived and would be better as national policy. The trouble with this conclusion is that it offers no concessions to the ambition of structural policy to transform the competitive (as opposed to comparative) advantage of assisted regions.
300
Iain Begg
Simulation models
Modelling exercises have been extensively used to assess the impact of the SFs, notably in work undertaken by John Bradley and his various collaborators using the Hermin model. By focusing on the supply-side, the SFs work in a different manner from more conventional income transfers that simply boost demand and need to be analysed as such. This is a point stressed by Bradley and Morgenroth (2004) and which they argue bears on how the SFs are assessed, because the implication is that it is not just demand-side elasticities that matter, but also the induced increases in stocks of human and physical capital. Some of the studies of individual Member States or regions using model-based assessments show that the EU policy did contribute significantly to growth. For example, Honohan (1997) suggests that it added one percentage point to Irish growth (which attained nine to ten percent per annum at its peak during the 1990s), but he stresses that the contribution has to be placed in perspective, since the growth also reflects several other factors. Sosvilla-Rivero et al. (2006) find that the SFs increased the growth of a case study region (Castilla La Mancha) by 0.64 percentage points over the period 1988-99. Evaluations
For accountability reasons, the SFs are now very extensively evaluated, and often receive quite positive appraisals. As an illustration, a comprehensive evaluation of the Objective 1 programmes (ECOTEC 2003) finds much to applaud in the functioning of the SFs and identifies several quantitative gains in terms of jobs created, business supported and so on. The study has an extended discussion of the effectiveness of policy, although it emphasizes the difficulty of arriving at a convincing assessment because of the diversity of qualitative and quantitative targets adopted by programmes. Similarly, an ex-post evaluation of the 1994-99 Objective 2 programmes by Centre for Strategy and Evaluation Services (2003) reached an overall conclusion that the ‘programmes made a significant contribution to regional developments’. While there might be some temptation to say that a study for DG Regio is bound to reach such a conclusion, the report does provide more detailed justifications for the assertion. However, in a passage that sums up the dilemma of how to assess the effectiveness of the SFs, the report states that ‘because target-setting practices remained poor, it is not possible to fully assess the extent to which specific Objective 2 aims were achieved (‘effectiveness’).’
17 Subsidiarity in Regional Policy
301
17.3.2 Conclusions on policy effectiveness Where cohesion policy correlates with improved performance, there may be problems in attributing observed results to the policy itself. Moreover, even if there is a plausible case that policy intervention has had positive effects, the added value from carrying out the policy at EU level may not be proven. Consequently, it is difficult to judge whether EU cohesion policy is optimal or a distantly second-best or whether the same resources would have achieved more if deployed at another level of government. Cumulative effects, especially, are hard to capture and may be one explanation for the generally ambivalent econometric results. The different techniques for assessing policy effectiveness produce diverse conclusions and none in isolation offers a wholly convincing answer. Distinguishing between the three broad classes of assessments, an (admittedly crude) characterization is that econometric studies – broadly – offer little support for the contention that cohesion policy is effective. Modelling exercises tend to be more positive, but it is the survey based assessments that seem to be most sanguine. Interpreting these findings from a subsidiarity perspective is far from easy.
17.4 The Structural Funds in economic governance Any assessment of the SFs has to take account of the effectiveness of other adjustment mechanisms,4 and if cohesion policy is not integrated with, or at least consistent with, wider national structural policies, there is a risk that the policies will pull in opposite directions (Midelfart-Knarvik and Overman 2002). This bears on subsidiarity in that what the SFs do may be distinctive from policies implemented by Member States. The OECD (2004) places a strong emphasis on the wider policy context in promoting growth in lower income regions, citing not just other supply-side measures, such as a pro-business environment and efficient public administration), but also stable macroeconomic policies (see also, Barry and Begg 2003). The OECD (2004) also argues that a high level of educational attainment may be a means of attracting investment. Analyses of the Irish experience suggest that while the SFs have made a difference, a long-term strategy of investing in education may, indeed, have been pivotal (Barry 1999).
4
Low labour mobility in the EU has long been a characteristic, although the 2004 enlargement has altered the pattern, at least for the time being.
302
Iain Begg
In practice, EU cohesion policies are complemented by a wide range of other policies at Member State or sub-national level that also bear on economic development and living standards. Quantitatively the most important is the net fiscal transfer that arises from the spatially unbalanced incidence of national tax and public expenditure flows (Begg 2004). In many Member States, various public services (notably health and education) are organized so as to ensure comparable per capita entitlements, irrespective of the capacity of the spatial unit to raise taxes, and pension systems, too, affect net flows across spatial boundaries, especially if pensioners tend to locate in specific parts of the country in preference to others. Similarly, social assistance and unemployment benefits flow to regions with higher than average numbers of people affected by poverty or unemployment. The financing of these net flows varies according to national traditions, with some countries having explicit inter-regional equalization mechanisms to redistribute taxes,5 while in others central government picks up the bill and redistributes via block grants or the direct financing of public services. Regional economic development is also affected by a range of supplyside policies, some of which have an explicit spatial component, while for others any spatial bias in their incidence is the less direct result of the focus of the policy. The policies with explicit spatial objectives include domestic regional policy, urban policy and various forms of spatial planning. Some countries, such as France, Italy and the UK, have a long tradition of territorial development policies administered by central government, albeit employing very different approaches which have, in addition, fluctuated over time. Although the designated regions are often aligned with those eligible for SF support, there are also noteworthy exceptions pointing to a further challenge related to subsidiarity, namely how to avoid clashes of priorities in a multi-level governance system. Plainly, a solution would be for the power to designate territories to be reserved to the national level. Active labour market policies (ALMP) have substantial budgets in some countries, reaching 1.52% of GDP in Denmark where they are most intensively used and averaging 0.64% (Eurostat 2006), approximately double what is spent on the SFs. Because these policies support unemployed persons, their spatial incidence will reflect – but not exactly follow - disparities in unemployment rates. Such policies may dovetail with the support to Member States from the Structural Funds via the European Social Fund (ESF) and it is certainly true that many local providers of training and related schemes rely on ESF support. Equally, there is no guarantee that the 5
Germany and Austria also have systems which transfer resources directly between sub-national units of government via the Finanzausgleich.
17 Subsidiarity in Regional Policy
303
sorts of project that obtain ESF funding are necessarily those that the Member State would choose. 17.4.1 State aids An awkward issue in decentralizing regional policy is how to regulate competition between regions. Such competition arises in a variety of ways, with regions competing to attract different investment flows, to retain taxpayers or to stimulate enhancement of factors of production. There is an evident danger that if many regions adopt similar approaches to public intervention there will be a zero-sum (or possibly negative-sum) game, but political economy arguments often over-ride pure market logic. State aids offer a good illustration. They are widely criticized for their distortionary effect on markets and for inhibiting restructuring and regeneration. Yet any government that is confronted with a possible loss of a major employer or the prospect of luring one will be sorely tempted to offer such aids. Davies and Hallet (2001) draw attention to a contrast in effectiveness between the SFs and indigenous regional policy. They argue that, despite relatively small budgets, the former can have a significant impact on income and employment, but that state aids (especially) are subject to very high levels of deadweight that render them relatively ineffective. The usual rationale for regulation of state aids is to ensure that the internal market is not at risk (Wishlade 2003), but it can also be argued that a cohesion rationale should be adopted. Community policies, by contrast, are specifically geared towards altering the economic development trajectory of eligible regions. For the future, this poses a number of challenges about what should be expected of national policies and, subsidiarity notwithstanding, the degree to which Community policies are needed to reinforce the overall commitment to cohesion. Member State policies that involve state aids may undermine, rather than support cohesion in the Union as a whole. 17.4.2 Confluence with the Lisbon strategy With the 2005 re-launch of the Lisbon strategy, there has been a growing expectation that the SFs will contribute to wider economic reform aims. With such a linkage, the discussion of which level of government should carry out regional policy has to be viewed differently. Spatial balance in economic development is, however, a dimension of economic reform that does not feature strongly in the Lisbon strategy, but arguably should do. Consistency in policy is also a pre-condition for effectiveness. Davies and
304
Iain Begg
Hallet (2001) find that effectiveness is diminished if regional policy is too diffused and tries to hit too many targets. In addition, poor implementation can often weaken impacts. At issue is not just whether the expenditure on the SFs stimulates growth, the employment rate, research potential and other Lisbon targets in assisted regions, but also the much stiffer test of whether it adds to the collective national and EU-level performance. The SFs do plainly have spatial objectives that the Lisbon strategy does not, and their focus is on decentralized operations, while the National Reform Programmes that are at the centre of the re-launched Lisbon strategy are about integrated strategy. Nevertheless, it is important to recognize that there is substantial congruence between Lisbon and cohesion objectives. Policies that promote new technologies, investment in R&D and innovation have been highlighted as part of the Lisbon strategy. Expenditure on R&D is regionally very concentrated in the EU, as is patenting activity. A policy response has been to try to diffuse this activity more widely, but the danger with such an approach is twofold: first, it risks dissipating research efforts when concentration may be a more effective strategy; and, second, there may be little advantage for relatively backward economies in trying to stimulate research activity. On the other hand, it could be argued that unless more effort is put in to transforming an economy’s underlying characteristics, it will be unable to make a leap forward. Midelfart (2004) argues that intra-industry knowledge spillovers in the EU are probably more significant than inter-industry ones and that this tends to militate against policies which lean against the wind in this way, implying that it would be more efficient to deal with inequalities by direct income transfers, thereby allowing (efficient) spatial concentrations - that increase intra-industry spillover - in favoured areas. This reasoning is bound to be controversial since it invites a dependency culture in the less-favoured areas, implies the possible wasting of assets in such regions, and can also result in spatial imbalances that have damaging macroeconomic consequences
17.5 A regional policy subsidiarity test? The forgoing discussion leads to a question about whether a subsidiarity test can be devised to determine when regional policy should be at the EU level one and when it should remain a national (or, conceivably, subnational) competence. Several factors need to be taken into account. The first is whether there ought to be regional policy at all, although the Treaty commitment to promote cohesion is germane. Here, the ambiguity about
17 Subsidiarity in Regional Policy
305
the rationale for the policy is problematic. If it is genuinely to promote economic development, as advertised, and thus an allocative mechanism, then the argumentation should be partly about whether recipient regions could obtain the necessary resources in the absence of the SFs and partly about whether the ‘Brussels’ policy model offers better prospects than what Member States would implement if left to their own devices. If the rationale for the EU policy is, in practice, redistribution, then EU regional policy is a clumsy way of achieving it and an explicit equalization transfer (such as the finanzausgleich used in Germany and Austria) would be a better solution. Second, so long as the EU budget remains capped at around 1% of GDP, the resources available for cohesion policy will be sorely limited. Consequently, even if there were a case for a policy that extends to all Member States, financial imperatives might dictate limits to the geographical spread of cohesion policy. The concentration principle for the SFs is supposed to achieve this aim, yet the fact that the Funds are seen as one element in the equation of net contributions to the EU budget means that even the richest Member States obtain some receipts from them. It follows that the budget wrangles that result in a diffusion of EU regional policy weaken the case for having the policy in the first place. A threshold of GDP per head of 75% of the EU average, applied at the regional level6 has long been used to concentrate spending for convergence purposes, while the test for the Cohesion Fund is 90% of the national GDP per head. If fiscal capacity is accepted as a first factor determining whether there should be EU policy, a subsidiarity test would emphasize national prosperity by making a Member State above a certain level of GDP per head responsible for funding its own regional policy, while envisaging transfers from the EU for those below a threshold. However, an abrupt cut-off point would lead to contestation and create the sort of ‘poverty-trap’ familiar to social policy-makers, so that a tapered or banded support would be more congenial. A test for subsidiarity in policy design raises different issues. The approach to regional development espoused by the EU is to promote longterm structural change by altering the productive potential of a region. Indeed, in a background report for the 3rd Cohesion Report on factors of regional competitiveness, Cambridge Econometrics and Ecorys (2004) argue that regions have to avoid the temptation of ‘transient’ Keynesian transfers, and to focus instead on policies that steer the regional economy in a direction that makes best use of its current and prospective assets. The report, in putting forward typologies of regional competitiveness, does ac6
Although some Member States have been designated in their entirety.
306
Iain Begg
knowledge that part of the art of regional policy is to enable a region to shift to a different competitive position, often implying an attempt to break-out from a structure of activity that is overly constraining or traps the region in its own history. The question then is whether subsidiarity allows this to be achieved more easily. At one level the answer is straightforward and, ostensibly, clear-cut. Policies that are formulated by policy-makers close to the regional problem would be expected to be better attuned to the region’s needs. But there are counter-arguments, centering especially on the risks that dominant local interests will ‘capture’ local policy-makers, leading to policies that serve these dominant interests rather than the wider regional interest. Extreme forms of the latter are where organized crime or systemic corruption results in the resources being diverted to favoured recipients with little discernible impact on regional development. More common is weak institutional capacity, which tends to diminish the effectiveness of any economic development policy, whether because of poor project choices, inadequate implementation and monitoring, a lack of absorption capacity or shortcomings in complementary policies. This can even lead to the almost paradoxical finding (as suggested, for example, by Ederveen et al 2006) that regions with stronger institutional capacity make the most effective use of the SFs, even though it could be argued that they need them least. Failure to spend money allocated has been another problem that has beset the SFs, especially in areas of weak institutional capacity. The argument for an EU policy then becomes one of ensuring acceptable standards of policy implementation. A frequently-heard assertion is that the sheer bureaucracy of the Funds is an obstacle and that this is a predictable effect of having too many tiers of government involved, with the implication that decentralization would improve matters. The other side of the coin is that the controls serve the purpose of avoiding poor choices and that the failure to spend may even show that quality of spending is enhanced by EU oversight. In this regard the evaluation evidence does suggest that the ‘Brussels model’ has been helpful, even if firm econometric evidence of enhanced growth is lacking. A possible reconciliation is that the SF approach has enabled practitioners in institutionally weak environments to learn by doing, but that they can subsequently regain full control over the policy. Assignment of the policy area to the EU would, therefore, be a transitional arrangement aimed expressly at capacity building in institutional terms, as well as investment in physical and human capital. As in so much of the subsidiarity debate it is hard to arrive at firm conclusions.
17 Subsidiarity in Regional Policy
307
17.6 Concluding comments Cohesion in the EU is advanced by policies implemented both by the Member States and the Community. While what Member State policies do has a large and positive impact on aspects of cohesion, Community policies contribute to promoting convergence, especially for the least competitive regions and the poorer Member States. In principle, there is a qualitative difference between the main instruments of national policy, which have explicit re-distributive aims, and the more specific interventions from the SFs aimed at economic development. This sort of regional policy is best organized in a top-down manner for three main reasons. First, to the extent that it is intended to alter development trajectories, the inevitable competition between regions needs to be recognized and mediated. If lower tiers of government are the principal actors, this will be much harder to achieve since use of state aids or other forms of support may undermine efforts in competitor regions. Second, the net transfer of resources has to be orchestrated by a higher tier of government. Third, in transferring resources, the higher tier is able to impose governance conditions that ensure that the support is used appropriately. This could also offset (national or regional) deficiencies in institutional capacity and promotes capacity building. On this basis, for the Objective1/convergence element of current EU regional policy, the fact that there is a net transfer of resources does warrant retention of the policy at the Community level. It is also noteworthy that in the Member States that receive substantial resources from Community cohesion policy, Community and national policies towards economic development have become quite closely inter-twined. While the evaluation evidence is far from conclusive, there are at least signs that the SF model for fostering economic development has facilitated a more strategic approach than past national models. In short: SFs, properly supported can be effective, but they may often appear to have a disappointing effect, above all in the short-term. For policy purposes, this is a message that is difficult to interpret, the more so when associated with the issue of subsidiarity. Whether the EU regional policy should continue to be offered in other regions or in richer Member States is more problematic. Unless juste retour arguments are adduced, there is no resource transfer argument to justify EU spending in richer areas, since the recipients are generally net contributors to the EU budget. In that case closer matching of regional preferences and lower transaction costs would also be motives for regional policy design at the Member State level. However, there may still be a case for the EU level to oversee the orientation of policy and there may also be
308
Iain Begg
advantages in being obliged to follow the EU policy model, despite the reservations expressed by the likes of the UK Government (2003). For Leonardi (2005: 1), one of the distinguishing features of the EU’s cohesion policy is the ‘extensive involvement of different administrative levels and socio-economic groups in the formulation and implementation of policy’. Leonardi asserts that cohesion policy has had a major influence in the governance of economic development policy by broadening the participation of a wider range of actors, including the empowerment of sub-national levels of government. He notes that evaluation evidence suggests that an outcome has been economic development policies better attuned to local circumstances and also that cohesion policy has encouraged institutional innovation and reform. Moreover centralization at the EU level might overcome capture of national and regional authorities by dominant local interests. This does not mean that the SFs, as presently constituted, should be a permanent feature of EU policy, although the political economy arguments cited above suggest that policies are hard to shift once established. Indeed, the history of the CAP demonstrates the power of inertia in EU policy: as the EU embarks on its second half-century, it is still the most costly EU policy. For constitutional and political economy reasons, EU cohesion policy is also likely to prove its resilience, even if some of the more theoretical economic arguments for subsidiarity are more persuasive, at least in relation to some Member States. At the same time, cohesion could be retained as an aim while adopting different methods of co-ordinating policy (Begg 2003). The principle of subsidiarity is that a policy area should be located at the lowest level of government able to conduct it efficiently. It is tempting to reduce subsidiarity to a simple, even simplistic cost-benefit calculation, based on fiscal federalism principles, of whether the benefits of tailoring policy to local circumstances outweigh the gains from internalizing externalities through situating the policy at the EU level. But the issues raised in this chapter suggest that a more nuanced approach is needed. In regional policy the EU level is better placed than the Member State level in some respects and worse in others. Consequently, there can be no easy formula to apportion regional policy in the EU among the tiers of government, especially with what is already in place as a result of past political compromises and current constitutional commitments. In fact, the most straightforward answer will often be that there are sound arguments for either or both of the EU and Member State levels having a role.
17 Subsidiarity in Regional Policy
309
References Ahrens J, Meurers M, Renner C (2008) Who shall decide what? Citizens’attitudes towards political decision-making in the EU. In: Gelauff G, Grilo I, Lejour A (eds) Subsidiarity and Economic Reform in Europe: 41–58, Springer Alesina A, Angeloni. I, Schuknecht L (2002) What does the European Union do? Harvard Institute of Economic Research Working Paper Armstrong HA, Taylor J (2000) Regional Economics and Policy, Oxford: Blackwell Bachtler J, Taylor S (2003) The added value of the Structural Funds: a regional perspective. European Policies Research Centre, University of Strathclyde Bachtler J, Wren C (2006) Evaluation of EU cohesion policy: research questions and policy challenges (editorial introduction). Regional Studies 40 vol 2 special issue: 143–153 Barro RJ, Sala-I-Martin X (1992) Public Finance in Economic Models of Growth. Review of Economic studies 59: 645–661 Barro RJ, Sala-I-Martin X (1991) Convergence Across States and Regions. Brookings Papers on Economic Activity 1: 107–181 Barry F (1999) Understanding Ireland’s Economic Growth, London: Macmillan Barry F, Begg I (2003). EMU and Cohesion. Journal of Common Market Studies 41: 781–796 Begg I (2003) Complementing EMU: rethinking cohesion policy. Oxford Review of Economic Policy 19: 161–179 Begg I [directed by] (2004) The impact of Member State policies on cohesion. Background study for the 3rd Cohesion Report, for DG Regio Beugelsdijk M (2002) Should structural policy be discontinued? The macroeconomic impact of structural policy on the EU-15 and the main candidate countries. De Nederlandsche Bank Research Memorandum WO 693 Boldrin M, Canova F (2001) Inequality and convergence in Europe’s regions: Reconsidering European Regional Policies. Economic Policy 32: 205–253 Boldrin M, Canova F (2003) Regional Policies and EU Enlargement. CEPR Discussion Paper 3744 Bradley J, Morgenroth E, with Gacs J, Untiedt G (2004) A Study of the Macroeconomic Impact of the Reform of EU Cohesion Policy, Dublin: ESRI Cambridge Econometrics and Ecorys (2004) A study on the factors of regional competitiveness. Report for the European Commission, Directorate-General Regional Policy Centre for Strategy and Evaluation Services (2003) Ex-Post Evaluation of 1994– 99 Objective 2 Programmes. Report for DG Regio of the European Commission Commission of the European Communities (1994) Competitiveness and Cohesion: Trends in the Regions, Fifth Periodic Report on the Social and Economic Situation and Development of the Regions in the Community, Luxembourg, Office for Official Publications of the European Communities
310
Iain Begg
Commission (2004) A New Partnership for Cohesion: Convergence, Competitiveness, Co-operation Third Report on Economic and Social Cohesion, Luxembourg, OOPEC Davies S, Hallet M (2001) Policy responses to regional unemployment: lessons from Germany, Spain and Italy. DG Ecfin Economic Paper 161 ECOTEC (2003a) Ex-Post Evaluation of Objective 1 1994-99. Report for DG Regio of the European Commission Ederveen S, De Groot HLF, Nahuis R (2006) Fertile soil for Structural Funds? A panel data analysis of the conditional effectiveness of European cohesion policy. Kyklos 59: 17–42 Eurostat (2006) Expenditure on labour market policies in 2004. Statistics in Focus: Population and Social Conditions 12 Fuente A de la (2004) Second-best redistribution through public investment: a characterization, an empirical test and an application to the case of Spain. Regional Science and Urban Economics 34: 489–503 Honohan P (1997) EU structural funds in Ireland: a mid-term evaluation of the CSF 1994-99. Policy Research Series Paper 31: Dublin: ESRI Kaldor N (1970) The case for regional policies. Scottish Journal of Political Economy 17: 337–348 Leonardi R (2005) Cohesion Policy in the European Union: The Building of Europe, Basingstoke: Palgrave Magrini S (2003) Regional (Di)convergence. Working paper, Universita Ca’ Foscari, Venice Midelfart KH (2004) Regional policy design: an analysis of relocation, efficiency and equity. CEPR Discussion Paper 4321 Midelfart-Knarvik KH, Overman H (2002) Delocation and European integration: is structural spending justified? Economic Policy 17: 321–359 Molle W (2006) Evaluating the EU cohesion policy. Paper presented at the Regional Studies Association conference, Leuven, May 2006 OECD (2004) Chapter V: Regions at work. In Survey of the Euro Area 2004, Paris, OECD Puga D (2002) European regional policy in the light of recent location theories. Journal of Economic Geography 2: 372–406 Sosvilla-Rivero S, Bajo-Rubio O, Díaz-Roldán C (2006) Assessing the effectiveness of the EU's regional policies on real convergence: An analysis based on the HERMIN model. European Planning Studies 14: 383–396 Tarschys D (2003) Reinventing cohesion: the future of European structural policy. SIEPS Report 17, Stockholm, SIEPS Tondl G (2001) Regional Policy. In: The Economics of the European Union: Policy and Analysis, 3rd edition, Oxford University Press UK Government (2003) A modern regional policy for the United Kingdom, Norwich, HMSO Wishlade F (2003) Regional State Aid and Competition Policy in the European Union. European Monographs 43 London: Kluwer Law International
18 Regional Policy as a Means to Curb Immigration
Robert Fenge and Volker Meier
18.1 Why does a EU regional policy exist? Regional policy accounts for approximately 35% of the EU budget (European Commission 2005). This share is even likely to increase further in the next few years, following the accession of twelve Member States in 2004 and 2007. The regional policy budget of the EU is mainly spent on subsidies for infrastructure investment in poor regions. The so-called structural funds subsidize infrastructure spending in regions with a GDP per capita of less than 75% of the EU average. The cohesion fund supports environmental and traffic infrastructure in poorer Member States, where the GDP per capita threshold lies at a level of 90% of the EU average. The existence and relative importance of the EU regional policy is still puzzling. It is not self-evident why rich countries spend money to support investment in infrastructure in poorer countries by means of regional transfers. It is neither obvious why the transfers should exist nor why spending the money is tied to conditions on public investment. The principles of the EU regional policy are spelled out in the EC and EU treaties, stressing the goal of reducing economic disparities (see Begg 2008, for details). This suggests a common-interest motivation. Solidarity or altruism between the Member States by means of regional policy can raise economic interdependence and reduce political tensions. But some countries can expect to be net payers for long periods of time. Therefore, an abstract commoninterest explanation is not sufficient from an economist’s point of view. Noting that still no central government of the European Union exists, the
312
Robert Fenge and Volker Meier
policies of the EU are determined in multinational negotiations. Hence, it is worthwhile to explore the presumption that the volume and the structure of the EU regional policy lies in the self-interest of the net payer countries. Otherwise, the rich countries should be influential enough to cut or abolish the regional policy budget. Conventional economic wisdom suggests some answers to the puzzle, which are not fully convincing, however. First, distributional considerations can provide an explanation for spending money on regional policies. Rich countries may simply be altruistic, and investment subsidies to poor countries can speed up income convergence. Tying aid to this specific use can then be explained by mistrust against the recipient countries. As the EU has a supranational structure, it seems doubtful that the transfer rules were enacted for altruistic reasons. Moreover, EU regional policy might have created a trade-off between overall economic growth and convergence. The message of the literature that studies the effects of the EU regional policy on convergence and growth is somewhat mixed (Martin 1998, 1999; Martin and Ottaviano 1999; Boldrin and Canova 2001; von Ehrlich et al. 2007). In general, regional convergence within a country is accelerated, while the impact on national GDP growth is small, zero or even negative. Such an outcome can be explained as follows. EU transfers reduce the economic pressure on disadvantaged regions to put more effort on structural reforms. In the short run, transfers facilitate regional investment projects. At the same time, public investment policy is distorted toward less effective projects. Second, regional policy can be used to enhance efficiency in the presence of spillover effects. Positive international spillovers may lead regional and national governments to invest too little. This consideration can indeed explain why federal matching grants are spent on such investment projects. If the internalization of production externalities is the driving force behind regional policy, it seems rather odd to restrict the access to the funds to poor regions. Again, it is not obvious that GDP growth of the EU as a whole is promoted by supporting infrastructure investment in poor countries. Growth of the EU GDP can be reduced through a reduction in specialization if public investment makes transactions within poor regions easier (Martin 1999). The literature also mentions other types of externalities that may justify the intervention of the EU. Regional policy can correct distorted location decisions of firms when these firms cannot capture the full social benefit of the investment (Fuest and Huber 2006). Further, the transfer rules between the different layers of the government may give rise to so-called vertical fiscal externalities. Thus, the federal taxation and transfer rules imply that regions do not capture the full return to their investment because the additional GDP implies higher contributions to the EU. This
18 Regional Policy as a Means to Curb Immigration
313
situation may lead to underinvestment in public infrastructure, calling for corrective subsidies. However, total annual public investment of a country is associated with vertical fiscal externalities that lie around 0.8 percent of the respective contribution to the EU, being too small to justify the observed high EU matching grant rates (Fenge and Wrede 2007). Third, regional policy may in principle be used to sustain a personal income redistribution between rich and poor individuals within the Member States. It has often been noted that increasing labour mobility across regions puts pressure on intraregional redistribution. By earmarking federal subsidies to the purpose of redistribution, socially desirable levels of redistribution can be maintained (Wildasin 1991; Wellisch and Wildasin 1996). As the sustainability of redistribution is mainly a problem of the emigration countries, it is doubtful that the net payers support the EU regional policy to pursue such a goal. Fourth, if immigrants represent a burden to the welfare state rather than being net tax payers in the host countries, a strong motive for reducing migration by regional policy transfers is present. Bribing potential immigrants to stay at home can be cheaper than transferring resources to them through the welfare scheme (Myers and Papageorgiou 2000; Hatzipanayotou and Michael 2005). Finally, it cannot be excluded that transfers implied by the EU regional policy rules have to be perceived as compensation payments to buy the votes of the poor countries in multi-national negotiations on a package of topics. Hence, the existence of the EU regional policy can be explained in a positive approach as political side payments that take account of the complicated majority rules in EU decision-making. In the following we give a new argument for the existence and the structure of regional policy, building on the fourth line of reasoning. The transfers that are given can reduce immigration into unemployment, and subsidies for infrastructure investment are particularly successful in achieving this aim. Our story bears some similarity to Myers and Papageorgiou (2000) and Hatzipanayotou and Michael (2005), dealing with foreign aid, where transfers are used to reduce immigration. Unlike these papers, immigrants do not receive net transfers from the natives through the welfare state, and decreasing government expenditure on border control to reduce illegal immigration is not an issue in the EU context. Our argument may be particularly relevant in the current accession period with twelve new Member States, most of them Central and Eastern European low-wage countries. Some countries in the old EU-15 with tightly regulated labour markets try to defend their high wages that are supported by generous unemployment benefits. In fact, free migration can then be expected to lead to higher unemployment among natives in the rich
314
Robert Fenge and Volker Meier
countries. Free mobility of labour within the EU, one of the basic liberties in the EU that has to be granted to all citizens in Member States, can yield welfare losses for the EU as a whole. As the losses arise exclusively in the immigration countries, the rich countries have a motive to induce workers from the new Member States to stay in their home countries. One option to achieve this goal is to provide regional transfers that improve the lot of workers in Central and Eastern Europe. The next section motivates our analysis by reviewing the empirical literature on the migration potential from Central and Eastern Europe following the EU accession and the impacts of migration on unemployment. Afterwards we discuss the mechanisms that are relevant for a regional policy combating immigration into unemployment in detail. Finally, we address the question of subsidiarity - whether regional policy is appropriately placed at the EU level in view of the different possible motives behind enacting this policy.
18.2 Potential migration and its impact on unemployment Large differences in wages of countries belonging to a common economic area with free labour mobility will induce migration of workers from lowwage to high-wage countries. This can generate considerable welfare gains in the country of origin as well as in the destination country. With full employment in the source and destination countries, the output gain from international migration is almost perfectly described by the wage increase of the migrant. In addition to this main impact of migration, general equilibrium effects may arise as the relative scarcity of capital and labour changes. While wages tend to fall in the immigration country and to rise in the emigration country, interest rates are predicted to move in the opposite direction. The message of the empirical literature is that these general equilibrium effects are generally negligible or very small even if substantial migration waves are considered (e.g., De New and Zimmermann 1994; Gang and Rivera-Batiz 1994; Boeri and Brücker 2005). Matters are quite different if migration occurs in the presence of unemployment in the immigration country. For example, the destination country may provide a minimum wage – legally or implicitly by social assistance regulations – that causes unemployment. As the wage is prevented from falling, immigration then induces additional unemployment unless additional capital moves along with the immigrants. Provided that output in the source country falls due to a reduction in employment, migration can decrease rather than increase aggregate GDP in the economic area as a
18 Regional Policy as a Means to Curb Immigration
315
whole. This conclusion holds irrespective of whether or not the migration decision of the worker is distorted by the unemployment benefit. Hence, immigration can be considered as harmful if the increasing labour supply forces immigrating or domestic workers into unemployment in imperfect labour markets. Motivating our story of regional policy as a means to reduce immigration into unemployment requires that free migration from the new Member States would be substantial, and indeed lead to increasing unemployment rates. A number of studies have projected migration from the new Member States to the EU-15. These studies rely on extrapolations of South-North migration in Europe during the 1960s and 1970s, surveys among the population of the new Member States, or econometric estimates considering differences in income levels and labour market conditions across countries as explanatory variables. The majority of these studies predicted a long-run migration potential of 3-5 percent of the population from the new Member States, and a short-run net inflow of about 300,000 – 400,000 persons per annum (Bauer and Zimmermann 1999; Alvarez-Plata et al. 2003; Krieger 2003; Boeri and Brücker 2005). Other studies show significantly lower (Fertig 2001; Dustmann et al. 2003) or higher projections (Sinn et al. 2003). Summarising 12 studies, De Mooij and Tang (2003) report a median of migration estimates from the Central and Eastern European countries of 2.9 millions of people after 15 years, with 1.9 million migrating to Germany and 350 thousand to Austria. These figures indicate a substantial amount of immigration under a hypothetical free migration regime. The obvious question is whether the fear is justified that immigration from low-wage countries increases unemployment among native workers. The empirical literature on this topic arrives at mixed results, ranging from insignificant or even slightly negative effects to substantially positive impacts on the unemployment rate of natives. This variety of outcomes can be explained by differentiated impacts on different types of labour and changing effects in different stages of the business cycle. In fact, immigration is expected to increase the demand for types of labour that represent complements to the immigrants, and immigration can be helpful to overcome shortages in an upswing period. On the other hand, natives who directly compete with the immigrants can reasonably expect a rising unemployment risk. In a recent simulation study dealing with the migration impact in the enlarged EU with rigid labour markets, a 1 percent increase in the labour force by immigration raises the unemployment rate in the immigration countries by 0.1 to 0.2 percentage points if the share of blue collar workers among the additional immigrants lies at 70 percent, where the impact on blue collar workers is stronger (Brücker 2007). Similar figures turn out in
316
Robert Fenge and Volker Meier
Velling’s (1995) study on the effects of immigration on local labour markets in Germany and Hunt’s (1992) paper on the impact of the inflow of Algerian repatriates to France in 1962 on the 1968 unemployment of nonrepatriates. Substantially stronger effects are estimated in the macroeconometric simulation of Franz et al. (1994) on immigration to Germany. Lowering free migration of ethnic Germans between 1989 and 1995 by 75% would reduce the German unemployment rate by 6 percentage points. Other studies dealing with different periods of time and using different econometric techniques find substantial impacts of the share of foreigners on the unemployment frequency of domestic workers only in some years (Winkelmann and Zimmermann 1993; Mühleisen and Zimmermann 1994), tiny effects of immigration on the unemployment rate of native workers (Winter-Ebmer and Zweimüller, 1999, on Austria), or even ambiguous impacts of immigration on the unemployment rate (Pischke and Velling, 1997). In sum, the empirical literature tends to predict substantial migration flows to Western Europe upon allowing for free mobility of labour in the enlarged EU, with a moderate increase of unemployment in the immigration countries. At the same time, fears of strong impacts of free migration on unemployment rates also receive some support from this literature.
18.3 Regional policy against undesired immigration Having observed that free migration may indeed lead to welfare losses in the EU immigration countries if these seek to sustain high wages, the question arises if regional policy is a useful tool to tackle this problem. It is obvious that transfers from the rich to the poor countries directly hurt voters in the rich countries, who have to pay more taxes to finance these transfers. Moreover, it has to be taken into account that the EU regional policy will stimulate capital flows from rich to poor Member States, which shifts jobs out of the rich states. The counteracting benefit from a reduced migration pressure arises when unemployment among natives goes down, increasing gross income per capita. The question is whether the burden through higher taxation and the capital outflow is more than offset by the employment gain. While this can ultimately only be decided by empirical research, it is not even theoretically clear that regional policy can yield a net gain to the natives in the rich countries. Obviously, regional policy can be advantageous for the rich countries only if the reaction of potential migrants to changing financial incentives is sufficiently strong.
18 Regional Policy as a Means to Curb Immigration
317
In a companion paper (Fenge and Meier 2006) we explore this question by analysing a model with a federation consisting of two representative countries that differ in their endowment with private capital and infrastructure. Production requires the input of the three factors labour, private capital, and public infrastructure. Increasing public infrastructure raises the productivity of both capital and labour. As capital is perfectly mobile, the interest rate does not differ across countries. Investing in public infrastructure will attract private capital from abroad. The capital import reinforces the positive impact on the productivity of labour. This increases the number of jobs at given wages and drives wages up at given levels of employment. At the outset, a higher level of public infrastructure in the rich country amounts to a technological advantage and implies a higher wage rate. Such a wage differential due to a technological difference can be substantial and persistent even if international trade in goods drives up the demand for labour in labour-abundant countries. Workers are interested in a high level of net income and differ in terms of their home attachment. At a given level of net income, every worker would prefer to live in his home country. Under free migration, some workers from the poor country with relatively low costs of moving abroad migrate to the rich country to capture a gain in expected income. As a consequence, a differential in expected wages between the two countries will always remain. In the rich country, unemployment prevails due to a wage being fixed above the market clearing level. Such a minimum wage may be set by the government or agreed upon in negotiations. Alternatively, it can simply arise when a benefit granted to the unemployed represents a lower bound for the wage. Jobs are randomly allocated among all individuals living in the rich country, without any discrimination of foreign workers. When making their decision to move abroad, the potential migrants expect to face some weeks of unemployment during the year according to the unemployment rate of the rich country. Both countries have national budgets to finance their share of investment in infrastructure. While the rich country has to finance its investment in full, the poor country may receive supplementing matching grants from the regional policy budget of the federation. The federation finances wage subsidies and investment subsidies that are exclusively spent in the poor country. The rich country makes a take-it-or-leave-it offer to the poor country by proposing the levels of the federal subsidies. It can use two instruments to avoid the attraction of workers, namely, wage subsidies and matching grants for infrastructure investment in the poor country. In our setup, wage subsidies are equivalent to unconditional transfers. When the government of the poor country decides to invest in public infrastructure, it is supposed to maximize net income of a representative
318
Robert Fenge and Volker Meier
voter living in this country. It has to balance a higher resulting wage of its workers against the sum of a higher tax burden to finance the national share of public investment and a higher contribution to the EU budget. It is obvious that wage subsidies directly reduce the migration incentive, as the wage differential becomes smaller. Similarly, an unconditional grant would drive net income up by reducing the tax burden for any given public investment plan. Subsidies in the form of EU co-payments for infrastructure investment also serve to reduce the migration incentive by bringing a tax relief to poor countries. In either case, more workers will stay in the poor country. This reduces the tax that has to be levied for any additional unit of public infrastructure. As a consequence, the government of the poor country will decide to invest more. This in turn will again reduce the migration incentive to the rich country. Even more workers will not move to the rich country, and an inflow of private capital occurs. Up to this point, the two instruments may be recognized as equivalent, where the same total amount of subsidies yields the same effects. On top of this, subsidies for investment in infrastructure will directly reduce the price of an additional unit of infrastructure. Therefore, the government of the poor country will decide to invest more. This additional investment in infrastructure again attracts capital and drives up the marginal product of labour. The objective of the government of the rich country lies in maximising the net income of its citizens. Apart from its own investment decision in public infrastructure, it sets the parameters of regional policy. The income of the natives in the rich country can be increased by reducing the unemployment rate, and by increasing their capital income through a higher amount of infrastructure. Investing in infrastructure will attract private capital from the poor country. Initially, this reduces the unemployment rate. Since potential migrants care about the unemployment rate in their destination country, the resulting gain to natives is counteracted by more immigration from the poor country. It cannot even be excluded that additional public investment finally yields an increasing unemployment rate in the rich country. As the effectiveness of own investment in infrastructure is limited through undesired immigration, implementing some kind of regional policy turns out to be attractive. Although both the financial transfer and the outflow of capital harm workers in the rich economy, the losses may be more than offset. The shrinking international wage gap helps the rich country to overcome its unemployment problem. Indeed, with given levels of public investment in infrastructure, migration and capital movements after transferring money from the rich to the poor country will bring down the unemployment rate in the rich country. Further, some additional infrastructure investment in the poor country finally also reduces the unem-
18 Regional Policy as a Means to Curb Immigration
319
ployment rate in the rich country. As long as the size of the federal budget is small, the gains to the natives in the rich country from increasing the employment rate outweigh the additional tax burden. Therefore, introducing either wage subsidies or co-payments for investment in infrastructure will have a positive impact on the welfare of voters in the rich country. Matching grants for infrastructure investment are preferred because they accelerate public investment in the poor country, and the overall impact of public investment in the poor country on the net income of natives in the rich countries is positive. A sufficient condition for this result is that the level of the federal budget for regional policy is sufficiently small. As the total regional policy budget of the EU is approximately 0.35% of GDP, this condition is assumed to be satisfied. If the federal budget is already high, increases of the tax burden dominate gains associated with higher employment rates, where additional investment in infrastructure in the poor country has even a negative impact on aggregate net income of natives in the rich country. In sum, our analysis may serve as an explanation for the exclusive use of matching grants in the EU. The motive to use regional policy to reduce immigration arises from substantial losses in the income of natives that come about when immigration increases unemployment. In a scenario with full employment, we may have native working poor receiving in-work benefits instead. The benefits of rich economies from encouraging potential immigrants to stay behind will typically be negative then. From a political economy perspective, matters may be different when the uneven distribution of income is taken into account. Regional policy can then be implemented in order to deter migration irrespective of a potential unemployment problem even if total income of the natives falls. Among the natives, workers without capital income tend to be hit most by immigration. At the same time, they only finance a small share of regional policy. Therefore, a net gain for them from implementing regional policy may go along with a loss of aggregate native income. Unions representing these workers may use their influence on policy-makers to enact a policy which ultimately avoids wage cuts. The government may then suggest to employ regional policy with matching grants for investment in infrastructure. The goal would be to achieve some target wage or unemployment rate in the rich country at the lowest cost for capital owners. Their interests may be partially represented as they have to bear the vast majority of the tax load. Since the migration pressure is reduced with more infrastructure investment in the poor country, the targets can be achieved at a lower tax burden. Hence, similar arguments as in our approach may justify the existence and structure of the EU regional policy in a public choice perspective.
320
Robert Fenge and Volker Meier
The most obvious objection against our analysis comes from the fact that a substantial amount of mobility of workers is needed to arrive at the predicted results. Moreover, the impacts we have illustrated in our approach with a single type of labour may be viewed as too strong. It would certainly be more realistic to assume different skill groups in the labour market, where only the low-skilled are confronted with minimum wage unemployment. As migration incentives between the new and the old EU exist for all skill groups, the problem of immigration into unemployment is often mitigated by skilled migration. This will hold if skilled and unskilled labour are complements in production. In such a situation, skilled immigration raises the marginal product of unskilled labour. This in turn reduces unemployment among the unskilled. However, the mere fact that several old Member States in the EU have chosen a transition period regime with restricted labour mobility indicates a widespread fear of strong impacts of free migration on unemployment.
18.4 Regional policy and subsidiarity Regional policy is a field where the argument in favour of coordination at the EU level is not always strong. Applying the principle of subsidiarity, regional policy should be treated at the EU level if this is preferred to both assigning the policy to national governments and organising it by bilateral agreements. Justifying regional policy at the EU level requires either substantial international externalities of the projects supported by regional policy or economies of scale from policy coordination at the EU. International externalities are relevant only if at least three Member States are concerned, as otherwise bilateral treaties can be used for internalization. The economies of scale criterion will be met if given policy goals can be achieved at a lowest cost via the EU. Finally, a decentralized solution would be preferable if the preferences of the national governments did not match each other. This last condition does not create problems when substantial externalities or economies of scale are involved. Efficiency gains can then be supplemented by side payments to losers to reach unanimity. When regional policy is used to reduce migration into unemployment, the arguments in favour of a centralized coordination are comparatively strong. Several old Member States with tightly regulated labour markets like Germany, Austria, France, and Italy are potentially hurt by immigration into unemployment. Bilateral investment subsidies or unconditional transfers to the poor new Member States in Central and Eastern Europe would reduce the migration pressure for any of these immigration coun-
18 Regional Policy as a Means to Curb Immigration
321
tries. As this implies positive externalities of international transfers, the resulting level of aid would be too small without a supranational agreement. Such a coordinated action is provided by the EU regional policy, which makes it superior to bilateral agreements. Moreover, economies of scale may be realized through lower costs of negotiation when moving from bilateral agreements to a single EU rule. Similar arguments can be given to justify organising regional policy at the EU level due to altruistic concerns or an insurance motive. The altruistic concern suffers from the Samaritan’s dilemma, consisting in the fact that the altruists in the rich country are also better off if somebody else supports the poor Member States. Organising international aid at the bilateral level would result in an amount of international redistribution that is too low. If regional policy is perceived as income insurance, increasing the degree of centralization is an effective means of risk spreading. Providing the transfers as matching grants for infrastructure investment then can be justified by paternalistic considerations. The donators want to ensure that the receiving countries will become richer and independent of foreign aid soon. Economies of scale are relevant insofar as some costs of negotiation may be saved when organising aid at the central level. By contrast, the EU regional policy will generally not pass the subsidiarity test if the infrastructure investment projects are considered in itself. At first sight, many of the investment projects supported by the European Union carry no international dimension. Further, many transnational traffic infrastructure projects have a significant economic impact only for the neighbouring Member State. Considering the argument that regional policy is used to internalize international production or consumption spillovers, the rationale in favour of an EU involvement (see de Borger and Proost 2008, for a detailed discussion) is convincing only for a small fraction of projects being co-financed by the EU. Economies of scale in providing infrastructure are often present. However, they do not play an important role when moving the decision of the provision from the national to the EU level.
References Alvarez-Plata P, Brücker H, Siliverstovs B (2003) The Impact of Eastern Enlargement on Migration — An Update. Report for the European Commission, DG Employment, Social Affairs and Equal Opportunities, Brussels Bauer T, Zimmermann KF (1999) Assessment of Possible Migration Pressure and Its Labour Market Impact Following EU Enlargement to Central and Eastern Europe. IZA Research Report 3, Bonn
322
Robert Fenge and Volker Meier
Begg I. (2008) Subsidiarity in regional policy. In: Gelauff G, Grilo I, Lejour A (eds) Subsidiarity and Economic Reform in Europe: 291–310, Springer Boeri T, Brücker H (2005) Why are Europeans so tough on migrants? Economic Policy 20: 629–703 Boldrin M, Canova F (2001) Inequality and convergence in Europe's regions: reconsidering European regional policies. Economic Policy 16: 207–253 Brücker, H. (2007). Labor Mobility after the European Union’s Eastern Enlargement: Who Wins, Who Loses? Report to the German Marshall Fund of the United States, Washington De Borger B, Proost S (2008). Subsidiarity and transport policy in Europe: what EU subsidies do we need for the TEN? In: Gelauff G, Grilo I, Lejour A (eds) Subsidiarity and Economic Reform in Europe: 325–341, Springer De Mooij R, Tang P (2003) Four Futures of Europe, CPB, The Hague De New JP, Zimmermann KF (1994) Native wage impacts of foreign labor: a random effects panel analysis. Journal of Population Economics 7: 177–192 Dustmann C, Casanova M, Fertig M, Preston I, Schmidt CM (2003) The impact of EU enlargement on migration flows. Report commissioned by the Immigration and Nationality Directorate of the UK Home Office, Home Office Online Report 25/03 European Commission (2005) General Budget of the European Union for the Financial Year 2005, Brussels and Luxembourg Fenge R, Meier V (2006) Subsidies for wages and infrastructure: how to restrain undesired immigration. CESifo Working Paper 1741, Munich Fenge R, Wrede M (2007) EU financing and regional policy: vertical fiscal externalities when capital is mobile. Finanzarchiv 63: 457–476 Fertig M (2001) The economic impact of EU enlargement: assessing the migration potential. Empirical Economics 6: 707–720 Franz W, Oser U, Winker P (1994) A macroeconometric disequilibrium analysis of current and future migration from Eastern Europe into West Germany. Journal of Population Economics 7: 217–234 Fuest C, Huber B (2006) Can regional policy in a federation improve economic efficiency? Journal of Public Economics 90: 499–511 Gang IN, Rivera-Batiz FL (1994). Labor market effects of immigration in the United States and Europe: substitution vs. complementarity. Journal of Population Economics 7: 157–175 Hatzipanayotou P, Michael MS (2005) Migration, tied foreign aid and the welfare state. CESifo Working Paper 1497, Munich Hunt JC (1992) The impact of the 1962 repatriates from Algeria on the French labor market. Industrial and Labor Relations Review 43: 556–572 Krieger T (2003) Migration trends in an enlarged EU, European Foundation for the Improvement of Working and Living Conditions, Dublin Martin P (1998) Can regional policies affect growth and geography in Europe? World Economy 21: 757–774 Martin P (1999) Public policies, regional inequalities and growth. Journal of Public Economics 73: 85–105
18 Regional Policy as a Means to Curb Immigration
323
Martin P, Ottaviano GIP (1999) Growing locations: Industry location in a model of endogenous growth. European Economic Review 43: 281–302 Mühleisen M, Zimmermann KF (1994) A panel analysis of job changes and unemployment. European Economic Review 38: 793–801 Myers GM, Papageorgiou YY (2000) Immigration control and the welfare state. Journal of Public Economics 75: 183–207 Pischke J-S, Velling J (1997) Employment effects of immigration to Germany: an analysis based on local labor markets. Review of Economics and Statistics 79: 594–604 Sinn H-W, Flaig G, Werding M, Munz S, Düll N, Hofmann H (2003) EU Enlargement and Labour Mobility - Consequences for Labour Markets and Redistribution by the State in Germany. CESifo Research Reports 2, Ifo Institute for Economic Research, Munich Velling J (1995) Immigration und Arbeitsmarkt: Eine empirische Analyse für die Bundesrepublik Deutschland, Nomos, Baden–Baden Von Ehrlich M, Eggert W, Fenge R, König G (2007) Konvergenz- und Wachstumseffekte europäischer Regionalpolitik in Deutschland. Perspektiven der Wirtschaftspolitik 8: 130–146 Wellisch D, Wildasin D (1996) Decentralized income redistribution and immigration. European Economic Review 40: 187–217 Wildasin D. (1991) Income redistribution in a common labor market. American Economic Review 81: 757–774 Winkelmann R, Zimmermann KF (1993) Ageing, migration and labour mobility. In: Johnson P, Zimmermann KF (eds) Labour Markets in an Ageing Europe. Cambridge University Press, Cambridge, 255–283 Winter-Ebmer R, Zweimüller J (1999) Do immigrants displace young native workers: the Austrian experience. Journal of Population Economics 12: 327– 340
19 Subsidiarity and Transport Policy in Europe: What EU-Subsidies Do We Need for the TEN?
Bruno de Borger and Stef Proost
19.1 Introduction
1
Why and how a supra-national government should intervene in the local supply of public goods, such as transport infrastructure, is a question that has received a lot of attention in the economic literature. This chapter studies the advantages and disadvantages of supra-national intervention in the provision of transport infrastructure, and how this intervention interacts with the use of pricing instruments for the use of the infrastructure. We review the basic economics underlying the pricing and investment decisions by Member States and the arguments for supra-national intervention. This allows us to draw some policy conclusion for EU intervention in the capacity provision and pricing of the Trans-European Networks (TEN-T). We interpret these networks mainly as serial transport corridors, and we study the outcomes of pricing and investment decisions for the network when different Member States make individual decisions on capacity and user charges on the links of the network they control. These outcomes are compared with efficient policies from the European perspective. Finally, we show the desirability of EU-intervention in the financing of the TEN, and we propose simple subsidy rules the EU can follow in deciding on subsidy levels for particular projects. The chapter is structured as follows. First, using simple graphical examples, we analyse the motives for supra-national intervention in the provision of transport infrastructure in Member States, and we describe the likely effects of these interventions. In the third section, we survey the re1
We acknowledge the support of the FUNDING consortium (EU 6th framework program). We thank two anonymous reviewers and the editors for useful comments.
326
Bruno de Borger and Stef Proost
cent literature on pricing and capacity choices on simple transport networks in the absence of cooperation and government intervention. Here we mainly draw upon De Borger et al. (2005) and De Borger et al. (2006, 2007). In Section 19.4, we build upon the insights obtained to suggest new subsidy rules the EU might want to consider for the TEN-T infrastructure projects. A final section offers some policy conclusions.
19.2 The economics of infrastructure provision by Member States in the presence of spillovers In this section we review the basic economics of infrastructure provision and pricing by Member States, and the implications for supra-national intervention.2 We start by briefly describing the principles underlying subsidiarity. Then we analyse how Member States make capacity decisions, and explain why this may necessitate supra-national intervention. We first look at the case without congestion; next we consider the case where congestion spillovers are present and Member States use pricing instruments. 19.2.1 The economics of subsidiarity and decentralization: a quick review In a supra-national system one broadly distinguishes three motives for supra-national aid in the provision of transport infrastructure for member states: spillover effects, issues of fiscal and financial capacity, and paternalism. First, spillover effects exist whenever the infrastructure in a Member State is also used by citizens or firms of other Member States. Whenever these “foreign” users do not pay for the use of the infrastructure, Member States will have lower incentives to supply the infrastructure. In this chapter, this phenomenon will be one of the key motivations for supra-national subsidies for transport infrastructure. Second, whenever a Member State has insufficient fiscal and financial capacity, the supra-national government could intervene by proving financial support to the Member States. However, this argument is probably more valid for local public finance problems (at the city or regional level) than at the level of the Member States. A third argument for supra-national aid is paternalism. This motive 2
One can consult the OECD-ECMT report on “Transport and decentralization” for a more extensive review of the decentralization situation in different countries. Sikow-Magny (2006) discusses in more detail the EU institutions.
19 Subsidiarity and Transport Policy in Europe 327
for intervention is sometimes used in supra-national contexts to protect minorities in some Member States. However, this argument is again difficult to defend in the context of decisions on transport infrastructure in the European Union: Member States are all democratic regimes, and subsidiarity is one of the founding principles of the Union in the first place. The first argument for supranational aid will be developed below. 19.2.2 Infrastructure provision by Member States: the case without congestion and without pricing We will discuss the inefficiencies in the provision of infrastructure by a Member State using a graphical example. We start with the simplest case where the use of the infrastructure cannot be priced and where the only decision is whether or not to build the infrastructure. Later we will introduce the option to price the use of the infrastructure and the choice of the capacity level. Take a member country that can decide whether or not to build a given infrastructure (a bridge, railway line, etc.). Once constructed, the infrastructure is used by local transport (inhabitants of the country, local freight transport) and by foreign transport (non-inhabitants, freight by foreign firms), but initially assume that no congestion occurs and that use of the infrastructure cannot be priced. To simplify, we further assume that the use of the infrastructure is constant over time, and that its construction and maintenance cost can be translated into a yearly equivalent cost Z. Whether or not it makes sense to construct the infrastructure can be analysed with the help of Figure 19.1. There we represent the total willingness to pay for trips of the local users (area under the MWPlocal curve) and the total willingness to pay of local and foreign users (MWPlocal+foreign). The more beneficial the infrastructure is (i.e., the higher the time and cost savings), the higher is the willingness to pay. As no user price is charged, the use of the infrastructure by local users will be given by the intersection of the MWPlocal -curve and the horizontal axis. Similarly, overall use by local and foreign demand is given by the intersection of MWPlocal+foreign and the horizontal axis. The total benefit is given by the area under the MWPcurves: it can be seen as the sum of the total willingness to pay for all trips. What is important for our case is the difference between the shaded (grey) area that represents the benefits for the inhabitants of the Member State and the area under the MWPlocal+foreign curve, that represents the total benefits for the federation. This includes the inhabitants of the Member State plus other foreign users that belong to the federation.
328
Bruno de Borger and Stef Proost
Member state will use : Total benefit local (grey)> total cost Z Instead of federal criterion: Total benefit local + foreign > total cost Z
price
MWPlocal+foreign
Federal intervention: subsidy rate s Total benefit local (grey) > total cost (1-s) with s= share of foreign use
MWPlocal
trips
Fig. 19.1. Benefits of building an uncongested and unpriced local infrastructure
A Member State will be interested in the benefits for its own citizens, but not in the benefits for foreign users. It follows that it will only build the infrastructure if the local benefit exceeds the cost of the infrastructure Z. On the contrary, a supra-national government will be interested in the benefits for all citizens. If Z is sufficiently large, it is clear that there are cases where it will be worthwhile for the federation to build the infrastructure, but where the Member State will not deliberately do so. Some worthwhile investments will not be undertaken, due to the spillovers of benefits to other Member States. These spillovers are one of the main motivations of supra-national aid for infrastructure projects of Member States. Indeed, an obvious way for the supra-national level to overcome the inefficiency described above is to use a supra-national subsidy. Such a subsidy (cfr. subsidy s in Figure 19.1) should satisfy some obvious conditions. It has to be conditional on the construction of the infrastructure; otherwise Member States will take the money and not build the infrastructure. The conditional subsidy should also be a function of the level of the spillovers. In our simple case, a percentage subsidy rate equal to the share of foreign use would make sure that the Member States perceive an infrastructure cost that induces them to take the supra-nationally optimal decision.
19 Subsidiarity and Transport Policy in Europe 329
Unfortunately, in practice, things are somewhat more complicated than this simple subsidy scheme, because the supra-national government typically does not have perfect information on the costs and the benefits of the proposed infrastructure. This asymmetry of information leads to two extra problems. One is that, if the supra-national level overestimates the cost or underestimates the local benefits to the Member State, then some subsidies will be given for infrastructure that would be build anyway, as it is beneficial also from the viewpoint of a Member State. The other problem is that the costs of infrastructure will be higher than necessary whenever the locals benefit from oversupply of some characteristics (say noise insulation etc.), as they do not have to pay the full cost. The design of an optimal subsidy scheme is therefore not a simple matter, and incentives for cost minimization and for a truthful revelation of transport needs have to be included. This is one of the main motivations to leave the subsidy task to an agency with sufficient technical expertise (see Meunier and Quinet 2007). 19.2.3 Infrastructure provision by Member States: introducing congestion and pricing When we introduce congestion on the infrastructure, the net benefit computation has to be reconsidered. The presence of transit traffic by foreign users actually decreases the local benefits in a Member State, because the extra congestion created by transit increases the average travel time. It follows that, ceteris paribus, higher subsidies are needed to motivate the Member States to build the new infrastructure. Moreover, introducing the possibility to price the use of infrastructure changes the cost-benefit computation quite drastically. In Figure 19.2 we represent the case of an infrastructure that is subject to congestion but where the use of the infrastructure is priced by the Member State. When the use of the infrastructure is not priced, one obtains an equilibrium use given by Qno-toll: traffic increases up to the point where the willingness to pay of local and transit traffic equals the average user cost in terms of time and other costs. The average user cost increases with the volume of traffic because loading more transport on the same infrastructure decreases the average speed and so increases the time cost. We immediately see that the absence of pricing leads to an excessive use of the infrastructure, because the willingness to pay of the last user is lower than the marginal social cost (MSC). The marginal social cost is the full cost of a trip for society as a whole. It equals the change in total resource and time cost for the whole economy of an extra trip on the given infrastructure. It will consist of the own resource cost plus the own time cost plus the extra
330
Bruno de Borger and Stef Proost
time costs imposed upon all other users because of the decrease in speed. The difference between the MSC curve and the average cost is the marginal external cost. Users are aware of their own time and resource cost (the average cost) but not of the extra costs imposed upon the others. For that reason the difference between the MSC and the average cost needs to be included in the user price via a toll. This would be the case if the optimal federal toll is implemented. In this case the sum of overall user benefits (ACD) and the toll revenues (DCEF) is maxim ized. A Member State has a different objective and goes for the maximum of the user benefits of the locals (ANL) and the total toll revenues (LNST). A Member State has an interest to charge a toll in excess of the marginal external cost and this leads to a significantly higher toll. Based on Figure 19.2, we report in Table 19.1 the benefit that is taken into account by a Member State and by a supra-national government when it considers investment decisions and this under different toll regimes. The benefit areas reported need to be compared with the lump sum investment and maintenance cost of the project: if the benefit is larger, then there is a larger probability that the project will actually be accepted by the decision maker. In the first line of Table 19.1 we see that the benefits for a supranational government are always larger than the benefits for a Member State because the former cares for the user benefits that accrue to foreign users while the state government does not. This property will hold for all the pricing regimes we consider. In the second line, we report the benefits when the infrastructure would be priced optimally from a supra-national point of view. The optimal toll charges the difference between the marginal social cost and the average user cost, taking into account the willingness to pay of all users. This leads to the transport volume Qopt toll in Figure 19.2. We assume that the transaction costs of pricing are zero. In the third line of Table 19.1, we assume that the Member State maximizes the welfare of its own citizens plus the total revenues from user pricing. It can be shown that a Member State government would actually opt for a toll in excess of the marginal congestion cost, i.e., it would engage in tax exporting behaviour (see, e.g., De Borger et al.2005). The corresponding transport volume is therefore to the left of Qopt toll; it is indicated as Qm state toll in Figure 19.2.3
3
The toll preferred by a Member State is actually a complex expression involving the relative share of local and transit traffic, the slope of the cost function and the elasticities of demand. It can be shown that it has to be to the left of Qopt toll.
19 Subsidiarity and Transport Policy in Europe 331
Member state compares: Total benefit local + tax revenue > total cost X Compared to MSC pricing, larger chance that project is realised BUT smaller total surplus (mainly for the foreign users)
price A MWPlocal+foreign L
M
D
Federal intervention: investment subsidy + Pricing contract
N
MSC C
B Toll Member state
Optimal federal toll
G F T
Average User Cost
K
H E S
Q m state toll
MWPlocal Q opt toll
Q no toll
trips
Fig. 19.2. Benefits of building a congested and priced infrastructure Table 19.1. Benefits at supra-national and Member State level for different user pricing regimes
No user pricing Optimal supranational toll Toll chosen by Member State
Benefits Supranational level
Benefits state level Comments
AKG > User benefits ACD+DCEF > User benefits + toll revenues ANL+LNST > User benefits + toll revenues
AHG User benefits ABD+DCEF User benefits + toll revenues AML+LNST User benefits + toll revenues
Best outcome for the supra-national level Best outcome for the Member State
The simple analysis summarized by Figure 19.2 and Table 19.1 leads to the following insights: 1. When the implementation of user pricing is not very costly for a Member State, the possibility of pricing the infrastructure increases the likelihood that an infrastructure project is actually realized without supra-national help. To see this, observe that the benefits at the
332
Bruno de Borger and Stef Proost
Member State level are always higher with pricing than without pricing, or AML+LNST >AHG. 2. The user prices that Member States would like to charge are generally too high, in the sense that the supra-national level will prefer lower tolls (in Figure 19.2, note that ACD+DCEF>ANL+LNST). However, setting the toll at the level preferred by the supra-national authority decreases the probability that the project is realized by the Member State without supra-national help. This can be seen in Figure 19.2., where AML+LNST>ABD+DCEF. 3. As a consequence, whenever supra-national aid is needed to make a project happen, there is an interest to try to link the subsidy to the pricing policies pursued by the Member States. For example, supranational support is much more justified if member countries agree to charge the supra-nationally optimal toll (equal to marginal social cost MSC) than when they charge above MSC. These important insights can be further elaborated in more realistic settings. In fact, most TEN-T’s are not isolated projects, but they are part of a network where competing or preceding links are controlled by other Member States. This means that pricing policies and investment projects of one Member State become complements or substitutes for other states, and there will be strategic behaviour by each of the Member State governments. When successive links of the same corridor are controlled by different Member States they become complements, when two links compete for the same transit traffic they become substitutes. This is explored in the next section.
19.3 Pricing and capacity choices by Member States or regions in the absence of EU intervention We extend the analysis of Section 19.2 by explicitly analysing how the economic efficiency effects depend on the structure of the international network on which transport takes place. This is important for transport infrastructure, because investments by Member States can be seen as capacity provision on particular links in an international road, rail or inland waterway network. It is useful, therefore, to study the outcomes of pricing and investment decisions for the network, when different Member States make individual decisions on capacity and user charges on the links of the network they control, and to see how these outcomes differ from supranationally efficient policies.
19 Subsidiarity and Transport Policy in Europe 333
The theoretical background for this section is provided by the recent literature on pricing and capacity decisions in transport networks. Both parallel and serial network structures have been studied. In the parallel case, long distance transit traffic is assumed to have a choice between different jurisdictions’ networks (see De Borger et al. (2005) and De Borger et al. (2006) for a review of the theory). An example is the transalpine crossing between Germany and Italy, where the main links pass either through Austria or through Switzerland. Both local and transit traffic contribute to congestion, and the two countries compete for revenue from transit. However, networks such as the Trans-European Networks TEN-T are more plausibly described in a serial network setting. Long distance traffic passes through a series of countries, each of which makes capacity and pricing decisions related to the part of the network it controls. This setting has recently been studied by Levinson (2001) and De Borger et al. (2007). This simple model set up is used to derive information on the likely behaviour of member countries that make non-cooperative decisions on capacity and pricing of transport infrastructure. We are interested in the following questions. First, for given capacity, how will member countries set user charges, given that their decisions affect users on other links in the network, and that other member countries will react to price changes? How does the outcome depend on the network structure? Second, what capacities will be selected by the member countries? Finally, knowing the capacity decisions and the tolling decisions of the Member States, does EUwelfare increase or decrease when tolling by Member States is allowed? In other words, is it a good idea for the EU to allow Member States to freely decide on tolls on the infrastructure, or is it better to restrict tolls or to transfer authority over tolls to the supra-national level? To discuss the answer to these questions, we will restrict the attention to the case of identical Member States in which transit transport is an important share of total traffic. 19.3.1 What user prices will be set by member countries when capacity is given? The literature referred to above shows that, whatever the network structure (parallel or serial), when member countries can decide on user prices, they will set prices above the marginal social cost. The Nash equilibrium that captures the interaction between member countries suggests that the intuition, shown graphically in Section 19.2 for an isolated country, carries over to a network setting. Member countries have an interest in tax exporting.
334
Bruno de Borger and Stef Proost
However, the network structure matters a lot for the extent to which tax exporting takes place. In the parallel case, the excess of tolls over the marginal social cost is limited, because price competition between the Member States keeps these margins down. Too high tolls imply that transit transport changes route, making demand very elastic. In the serial network case, the problem is much more severe. As transit passes through the two countries, each country sets user charges that maximize its local user surplus plus its own revenues. Because each country disregards the (negative) effect of its higher user charges on the revenues of the other country, the serial setting gives rise to excessively high charges4. This is known as double marginalization When both countries could cooperate5 and maximize total user surplus plus total revenues, the charges would be much lower and overall welfare would improve. This phenomenon offers an important justification for the promotion of multi-modal trips and for cooperation between countries for long distance rail trips. In fact, it may be much more important than the standard justification typically given, viz. that cooperation saves on transaction costs. 19.3.2 What capacity decisions will be made by Member States? What capacity decisions can we expect from Member States in the simple network settings described before? It is instructive to first consider the case where tolling is not possible, because in that case the answer is easy and highly intuitive. Consider the parallel case first. Whenever one country extends its capacity, it will attract more transit traffic. However, this mainly benefits the other country, because this country will face less transit traffic and less congestion. The implication is that each of the countries will be reluctant to expand capacity, as to a large extent the benefits flow away to the other country. Therefore, capacity levels offered by countries will be too low. In the serial case, the situation is different. If a country invests in capacity then overall transit demand through the whole transport corridor increases. This generates more congestion in the other country, which is a negative effect ignored by the investing country. The consequence is that each country tends to offer too much capacity from the viewpoint of the federation as a whole. Nash (2005), reports very high rail rates in Switzerland and Slovakia that may be due to this double marginalization. 5 Cooperation could take the form of a joint venture, such as the Thalys in the case of long distance rail trips. 4
19 Subsidiarity and Transport Policy in Europe 335
Before dealing with the general case where tolling is possible, let us point out another interesting characteristic of the interaction of investment in capacities between countries. One can easily show that capacity levels are strategic substitutes in the case of parallel links (the capacity increase by one country leads the other country to decrease its capacity, both reaction functions are downward sloping), but they are complements in the serial case. Interpreting the TEN as serial transport corridors, this latter finding has important implications. It means that, if the supra-national EU-level did not invest in the TEN then, whenever one country increases capacity of the link under its jurisdiction, other countries have an interest to do the same. They would also increase capacity on the links passing through their jurisdictions in order to cope with the increased transit flow. When tolling is possible, the simple story presented above needs to be slightly adapted, and the results are much less transparent. In fact, one needs a numerical model to reach clear conclusions (see De Borger et al. (2006, 2007) for details). Table 19.2 summarizes the findings based on such a model. In a parallel setting, it is found that the inefficiencies from non-cooperative behaviour are quite small. If countries can choose capacity levels and tolls, they will set tolls too high and capacity too low. As argued before, however, competition for transit traffic will keep the profit margins low. Consequently, it turns out that the non-cooperative equilibrium tends to be rather efficient, and that it approaches the supra-national solution. The welfare effects of toll and capacity competition are therefore very limited: allowing tolling, even if it leads to higher than optimal tolls from the European perspective, produces quite reasonable capacity levels (although smaller than optimal), and the welfare cost of non-cooperative behaviour is limited. In the serial network case, which is more relevant for most of the TENprojects, we find that the possibility for countries to toll the use of their infrastructure has dramatic effects. Tolls will be extremely high and because of this, capacity investment will be much below the supra-nationally desired levels. The welfare consequences of non-cooperative behaviour are in this case very large. The policy implication for the TEN that follows from this analysis is threefold. First, to avoid high tolls, insufficient investment and large welfare losses, it is in the serial case, better for the EU not to allow individual countries to freely decide on tolls at all. This is clear from Table 2, serial network case. Second, there is a case for European intervention in the provision of infrastructure in serial transport corridors. Whether or not tolling by member countries is allowed, capacity is under-provided compared to the optimal level from a European perspective. Even if the supra national level fixes the tolls, the Member States will choose too low capacities as
336
Bruno de Borger and Stef Proost
they are not interested in the benefits for the non local users. Third, there is a case for the EU to link the provision of financial support to the introduction of marginal social cost pricing by the Member States. This will come closest to the supra-national optimum, which requires the optimal capacity combined with marginal social cost pricing. In the next section, we therefore turn to the development of economically sensible subsidy rules for serial corridors. Table 19.2. Summary of outcomes when member countries set tolls and capacities non-cooperatively and cooperatively PARALLEL NETWORK
Toll
No tolling possible Toll=0 (1)
Toll determined by Toll>Marginal member countries External Cost (2) (MEC) EU supra-national Toll= MEC optimum (3)
Capacity
Qualification
Low capacities
Worse than case (2) with tolls. Hence, allowing countries to toll improves welfare. Higher capacity but EU welfare not so still lower than supra- much lower than the national optimum supra-national optimal Higher capacity Best welfare level
SERIAL NETWORK No tolling possible Toll=0 (1)
High capacity
Toll determined by Toll>>MEC Low capacities member countries (2) EU supra-national Toll=MEC Highest capacity optimum (3) Definitions: MEC = Marginal External Cost.
Better than case (2) with tolling. Hence, allowing countries to freely toll reduces welfare. Worst case of all Best welfare level
19 Subsidiarity and Transport Policy in Europe 337
19.4 What subsidy rules make sense for TEN investments? In this section,6 we analyse the type of subsidy rules the EU might use, starting from marginal social cost pricing in all Member States. There are three important considerations that should be taken care of in the design of such a subsidy scheme. We first describe them one by one, and then present a proposal for a simple subsidy scheme based on these principles. 19.4.1 Cost recovery considerations Once marginal social cost pricing is imposed, member countries will not be able to reach a break even situation when there are decreasing average costs of capacity expansion. Indeed, according to the cost recovery theorem (Mohring and Harwitz 1962; de Palma and Lindsey 2007), marginal cost pricing implies that the cost recovery ratio will be equal to the elasticity of the total capacity cost with respect to capacity (denote this by İ). If there are constant returns in capacity expansion (constant average cost of capacity), then the cost recovery ratio is one. However, if there are increasing returns due to, e.g., large fixed costs of capacity provision, then and capacity costs cannot be fully financed by the revenues from marginal social cost pricing for the use of the infrastructure. In order to grasp the intuition, compare the supply of transport trips with the supply of any good by a competitive industry: a decreasing long run marginal cost does not allow cost recovery when prices equal the marginal cost. For railroads and canals, the empirical literature suggests that there are high fixed capacity costs, so that the cost recovery ratio will be rather low (e.g., ); for roads, this elasticity is much closer to one. This implies that imposing marginal cost pricing for the use of rail and inland waterways requires public subsidies that are equal to a share (1- İ) of the total investment cost. Subsidies for road investment are much less needed at least if there is user cost pricing. 19.4.2 Introducing the marginal cost of public funds It may be necessary to deviate from marginal social cost pricing and the simple subsidy rule just presented, because raising public revenue to fi6
This section uses material of the FUNDING-project, as reported in Dunkerley et al. (2006).
338
Bruno de Borger and Stef Proost
nance the subsidies can itself have important efficiency costs. For example, if the subsidies require higher labour taxes then it increases the existing (and already large) distortion on the labour market. Economists express the welfare cost of raising an extra euro of public revenue as the marginal cost of public funds. Recent research shows that in most countries it is situated between 1 and 2 (Kleven and Kreiner 2003). If this cost of funds exceeds one there is an extra welfare loss associated with revenue generation. In that case it may be beneficial to charge more than the marginal social cost and offer somewhat lower capacity. This will raise the cost recovery ratio and reduce the public subsidy that is needed. To illustrate the implications note that one can show, under a particular set of assumptions, that the required subsidy becomes (1- İȽ), where Ƚ is the marginal cost of public funds (Van der Loo and Proost 2006). 19.4.3 Restricting the share of the EU Finally, we know from Section 19.3 that, in the absence of cross border traffic, there is no point in handing out EU subsidies. Doing so would indeed generate strong incentives for common pool type of lobbying (cf. Ederveen et al. 2008). Moreover, from Sections 19.2 and 19.3 we know that inefficiencies in pricing and capacity provision are related to the use of local infrastructure by foreign users. It seems plausible, therefore, to relate the EU-share in funding infrastructure to the importance of transit traffic through the country. A crude way of doing so is to limit this share to the share of transit traffic in overall infrastructure use. Countries with a lot of transit have less incentive to provide the correct capacity and therefore receive a higher share of EU-funding. 19.4.4 A new possible subsidy rule for the TEN Combining the three different considerations, we suggest that the EU subsidy share in total investments would be given by the following equation:
EU subsidy share (1 H *) X where X represents the share of transit in total traffic, and other parameters are as before. As an example, take a rail line where transit represents 50%, the elasticity of the cost of capacity is of the order of , and the marginal cost of public funds equals 1.2. This means that the subsidy share of the EU in total investment would be [1–(0.3)(1.2)]0.5 = 0.32, or 32 %. The
19 Subsidiarity and Transport Policy in Europe 339
rest would have to come from funding by the Member State itself and from user fees. The coverage of part of the capacity costs by user fees takes the form of charging for external congestion costs. For example, one could cover a share İ Ƚ (36%) of capacity costs by user fees. Another 32% would then ideally be financed by the Member State or region from other sources. Note that the cost recovery theorem (Mohring and Harwitz 1962) implies that user fees increase over time with the external congestion costs; they do not take the form of an average infrastructure cost. The user fees also cover the maintenance and operation costs as well as the external costs other than congestion. Note that the proposed rule does not contain any equity component: it used subsidies to help enforce efficient pricing rules and efficient investments. Although transport investments in Europe are often a very indirect way of redistributing income over regions or Member States, from an economic viewpoint direct subsidies towards these regions may be more efficient to achieve these goals. Finally, we have restricted our attention to the case where Member States behave non-cooperatively. If member countries are able to make binding cooperative agreements, no intervention of the supra national authority is needed (see subsidiarity test of Ederveen et al. 2008). This is certainly an interesting solution if there are only two neighbouring countries involved.
19.5 Conclusions Conclusions include the following. First, we showed that the desirability of European funding for infrastructure projects crucially depends on the importance of cross-border transport. Indeed, if there is no transit transport, subsidies are not needed from an efficiency viewpoint. Second, we interpreted Trans-European Networks as serial transport corridors and showed that, if countries are allowed to determine tolls for the use of their infrastructure, they will charge excessively high tolls and strongly under-invest in transport capacity. To avoid high tolls, insufficient investment and large welfare losses it is better for the EU not to allow individual countries to freely decide on tolls at all in the serial case. Third, because capacity is underprovided, there is a case for European intervention in the provision of infrastructure in serial transport corridors. We suggested linking the provision of financial support to the introduction of marginal social cost pricing by the Member States. A subsidy scheme was proposed that positively depends on the share of transit through the country and on the degree of scale
340
Bruno de Borger and Stef Proost
economies in capacity provision, and negatively on the cost of public funds.
References De Borger B, Proost S, Van Dender K (2005) Congestion and tax competition on a parallel network. European Economic Review 49: 2013–2040 De Borger B, Dunkerley F, Proost S (2006) Strategic investment and pricing decisions in a congested transport corridor, forthcoming. Review of Network Economics De Borger B, Dunkerley F, Proost S (2007) The interaction between tolls and capacity investment in serial and parallel transport networks. Journal of Urban Economics De Palma A, Lindsey R (2007) Financing needs and cost structure, Chapter 2. In: De Palma A, Lindsey R, Proost S (eds) Investment and the use of tax and toll revenues in the transport sector, Elsevier Dunkerley F, Mackie P, Proost S (2006) Consortium Funding Scenarios for EU Infrastructure Fund and Mark-ups, Deliverable 2 of FUNDING consortium financed by DG-TREN, http://www.econ.kuleuven.be/funding/ Ederveen S, Gelauff G, Pelkmans J (2008) Assessing subsidiarity. In: Gelauff G, Grilo I, Lejour A (eds) Subsidiarity and economic reform in Europe: 19–40, Springer European Commission, DG-TREN (2005). TEN–T priority axes and projects 2005 Fabian T (2005) Grenzüberschreitende Infrastrukturplanung in Europa, Interna tionales Verkehrswesen 57: 192–195 Kleven HJ, Kreiner CT (2003) The marginal cost of public funds in OECD countries: hours of work versus labor force participation. CESifo Working Paper 935 Meunier D, Quinet E (2007) The contracting of operation and the management of infrastructure funding bodies, Chapter 4. In: De Palma A, Lindsey R, Proost S (eds) Investment and the use of tax and toll revenues in the transport sector, Elsevier Mohring H, Harwitz M (1962) Highway benefits: an analytical framework, Evanston, Illinois, Northwestern University Press Nash C (2005) Rail infrastructure charges in Europe. Journal of Transport Economics and Policy 39: 259–278 OECD-ECMT (2006) Transport and decentralisation, Report of the Round Table 131 held in Paris September 2004 Persson T, Tabellini G (1996) Supra-national fiscal constitutions: risk sharing and moral hazard? Econometrica 63: 623–646 Sikow-Magny C (2006) Subsidiarity and transport policy coordination in the European Union: 65–75. In: OECD-ECMT (2006) Transport and decentralisation. Report of the Round Table 131 held in Paris September 2004
19 Subsidiarity and Transport Policy in Europe 341 Turró M (1999) Going trans-European: planning and financing transport networks for Europe. Pergamon Press Van der Loo, Proost S (2006) Mark up Rules for funding investments, Task report for deliverable 1, FUNDING project, 6th Framework Programme of the EC