The Europeanization of Workplace Pensions : Economic Interests, Social Protection, and Credible Signaling 9781107728592, 9781107041059

Citation preview

The Europeanization of Workplace Pensions

Alexandra Hennessy examines an area of Europeanization that has been largely ignored by political analysts: the development of an internal market for workplace pensions. This book offers an analysis of what is at stake in workplace pension reforms, tracing how different states approached them and how national political economy models have shaped actors’ bargaining strategy at the EU level. Employing statistical analysis, formal modeling, and in-depth case study research, Hennessy highlights the role of informal signaling and communication processes in designing a common pension market. This book offers a theoretical framework that accounts for historical institutionalism, informal signaling processes, and discourse in the Europeanization of workplace pensions – a must-read for students of comparative social and public policy, comparative politics, and European politics. alexandra hennessy is Assistant Professor of Political Science at Seton Hall University.

The Europeanization of Workplace Pensions Economic Interests, Social Protection, and Credible Signaling

alexandra hennessy

University Printing House, Cambridge CB2 8BS, United Kingdom Published in the United States of America by Cambridge University Press, New York Cambridge University Press is a part of the University of Cambridge. It furthers the University’s mission by disseminating knowledge in the pursuit of education, learning and research at the highest international levels of excellence. www.cambridge.org Information on this title: www.cambridge.org/9781107041059  C

Alexandra Hennessy 2014

This publication is in copyright. Subject to statutory exception and to the provisions of relevant collective licensing agreements, no reproduction of any part may take place without the written permission of Cambridge University Press. First published 2014 Printed in the United Kingdom by CPI Group Ltd, Croydon CR0 4YY A catalogue record for this publication is available from the British Library ISBN 978-1-107-04105-9 Hardback Cambridge University Press has no responsibility for the persistence or accuracy of URLs for external or third-party internet websites referred to in this publication, and does not guarantee that any content on such websites is, or will remain, accurate or appropriate.

Contents

List of figures

page vii

List of tables

viii

Acknowledgments

ix

List of abbreviations

xi

1 The European dimension of the pension challenge 1.1 Summary of the main argument

1 3

2 National pension regimes, supranational harmonization efforts 2.1 Pressures on public pension systems 2.2 The multiple identities of workplace pensions 2.3 The politics of pension market integration

9 10 14 19

3 The sources of pension reforms in Western Europe 3.1 When and why do countries reform their pension systems? 3.2 The Maastricht Treaty shock: pressure from “above” 3.3 Empirical analysis 3.4 Results 3.5 Conclusion

27

4 Informal signaling and EU-level bargaining 4.1 Decision-making in the European Union: bargaining and procedural approaches 4.2 The costs of adjusting to an EU-wide pension market 4.3 Formal model 4.4 Pressure from “below”: domestic preferences over EU pension regulations 4.5 What makes “domestic constraints” assertions credible?

48

29 32 35 40 46

49 52 54 57 65

v

vi

Contents

4.6 Conclusion 4.7 Equilibrium characterization 5 Agenda setting and the single pension market 5.1 Inefficient agenda setting in the early 1990s 5.2 Efficient agenda setting in the early 2000s 5.3 Taxation, investment rules, and biometric risk coverage 5.4 Country variations 5.5 Alternative explanations 5.6 Conclusion

73 74 76 80 84 87 90 92 94

6 The German position on EU pension policies 6.1 The German workplace pension system 6.2 Germany and EU negotiation failure in the early 1990s 6.3 Domestic discourse – the Kohl era 6.4 Germany and EU negotiation success in the early 2000s 6.5 Domestic discourse – the Schroder era ¨ 6.6 Domestic discourse – the Merkel era 6.7 Conclusion

96 101 103 106 109 112 119 122

7 The British position on EU pension policies 7.1 The British workplace pension system 7.2 Domestic discourse – the Thatcher era 7.3 Britain and EU negotiation failure in the early 1990s 7.4 Britain and EU negotiation success in the early 2000s 7.5 Domestic discourse – the Blair era 7.6 Conclusion

125 131 133 135 140 143 145

8 Conclusions

148

Bibliography

156

Index

173

Figures

3.1 Probability of pension reform: control group and Maastricht signatories page 41 3.2 Difference in probability of reform 43 3.3 Probability of reform, reform in other countries 43 4.1 Utility functions of Bismarckian and Beveridgean member states 54 4.2 Sequence of moves 55 4.3 Representation of preferences over investment rules and biometric risk coverage 62 4.4 Representation of indifference curves 64 6.1 German workplace pension schemes 101 7.1 British workplace pension schemes 132

vii

Tables

3.1 Countries in study, 1980–2002 3.2 Cloglog regression: pension reform 3.3 Comparing Beveridgean to Bismarckian systems,  Pr(reform)

viii

page 36 40 45

Acknowledgments

There are many people without whom I could not have written this book, but I hope you won’t hold this against them. They are all fine people and in no way responsible for the errors that remain. My greatest debt goes to the chair of my dissertation committee, Sofia Perez. Sofia has encouraged this endeavor from its early beginnings and has offered support, counsel, and wisdom throughout its development. I am also very grateful to the other members of my dissertation committee, Cathie Jo Martin and Vivien Schmidt, for their generous time, their unwavering commitment to this project, and invaluable comments. All of them have saved me from many wrong turns along the way. Their own writings on comparative political economy and the European Union continue to inspire me. Many thanks also to Margarita Estevez-Abe for first sparking my interest in the politics of workplace pension provision by hiring me as a research assistant. I would like to thank the Political Science department at the University of Rochester for hosting me when the bulk of this manuscript was written. I am grateful for the opportunity to present my work at workshops and seminars. Special thanks go to Randall Stone for offering very valuable advice on the manuscript. Whatever the shortcomings of the final version may be, they would have been far greater without Randy’s suggestions. Further helpful comments were provided by the participants of the comparative politics working group and the international finance seminar. I am very grateful to Sarah Brooks (Ohio State University) for graciously sharing data with me, of which a subset is used in Chapter 3. During the past years, I have presented various parts of this book at a number of conferences and seminars. I am grateful to Karen Anderson, Christine Arnold, David Coen, Jens Blom-Hansen, Neil Fligstein, Philipp Genschel, Markus Haverland, Alexander Hicks, David Howarth, Torben Iversen, Dan Kelemen, Patrick Leblond, Deborah Mabbett, Moira Nelson, Sebastiaan Princen, Lucia Quaglia, Tal Sadeh, ix

x

Acknowledgments

Amy Verdun, Kent Weaver, Jonathan Westrup, Anthony Zito, and other conference participants for comments and suggestions. My sincere thanks go to the busy officials at the European Commission, European Parliament, the European Federation for Retirement Provision (EFRP), European Trade Union Confederation (ETUC), the German Occupational Pension Association (aba), as well as the government representatives who agreed to be interviewed for this project. Ivo van Es, Jean-Yves Muylle, Martin Merlin, and Raymond Maes at the Commission, MEP Pervenche Beres, Chris Verhaegen at EFRP, Martin Hutsebaut at ETUC, Frank Baumeister and Peter Gorgen at ¨ the German Ministry of Labor, and aba chairman Klaus Stieferman were particularly generous with their time; most interviews lasted one and a half hours or longer. I deeply appreciate their candor and help in sorting through the European dimension of pension policies, and freely acknowledge that the interpretation of the motives of actors involved is my own. Kathleen Kollewe deserves special thanks for hosting me during my field research in Brussels. Financial support was provided by Boston University and the German Academic Exchange Service (DAAD). The department of Political Science and Public Affairs at Seton Hall University provided a highly stimulating environment for the final completion of the manuscript. For inspiring conversations and exceptional collegiality, I thank Kwame Akonor, Heath Brown, Patrick Fisher, Jo-Renee Formicola, Matthew Hale, Annie Hewitt, Nalin Johri, Roseanne Mirabella, King Mott, Robert Pallitto, Mike Taylor, Jeffrey Togman, Wigeby Toussaint, and Naomi Wish. At Cambridge University Press I am indebted to John Haslam, Samantha Richter, Helena Dowson, Carrie Parkinson, Fran Hiller, as well as copy-editor Paul Smith, for shepherding my manuscript through the production process so efficiently. I would also like to acknowledge my two anonymous reviewers who provided unusually constructive comments. I thank my parents for their steadfast moral and financial support over the many years. My deepest gratitude goes to my husband, Martin Steinwand, who read countless drafts and offered incredibly helpful advice at all critical stages. I thank him for his enthusiasm for this project, and for believing in me when I did not. My son Victor Steinwand, who is an endless source of joy, will be glad to know that this project is finally completed. It is my greatest pleasure to dedicate this book to my husband and son.

Abbreviations

aba AGIRC ARRCO BDA CDU CEA CEEP CEEs cloglog CME CONSOB CSU DA DG DWP EET EFRP EIOPA EMU ENEPRI ETT ETUC ETUI

Arbeitsgemeinschaft fur ¨ betriebliche Altersversorgung e.V. Association g´en´erale des institutions de retraite des cadres Association pour le r´egime de retraite compl´ementaire des salari´es Bundesvereinigung der Deutschen Arbeitgeberverbande ¨ Christlich Demokratische Union Deutschlands (Christian Democratic Union) Comit´e europ´een des assurances European Centre of Employers and Enterprises Providing Public Services Central and East European countries complementary log-log regression coordinated market economy Commissione Nazionale per le Societa` e la Borsa Christlich-Soziale Union in Bayern (Christian Social Union) discourse analysis Directorate-General Department for Work and Pensions exempt-exempt-tax European Federation for Retirement Provision European Insurance and Occupational Pensions Authority European Economic and Monetary Union European Network of Economic Policy Research Institutes exempt-tax-tax European Trade Union Confederation European Trade Union Institute xi

xii

FEFSI

List of abbreviations

F´ed´eration Europ´eenne des Fonds et Soci´et´es d’Investissement HI historical institutionalism IAS International Accounting Standards IG Metall Industriegewerkschaft Metall IORP Institution for Occupational Retirement Provision MEP Member of the European Parliament MPIfG Max Planck Institute for the Study of Societies NAPF National Association of Pension Funds OECD Organisation for Economic Co-operation and Development OMC open method of coordination PAYG pay as you go PBE Perfect Bayesian Equilibrium PDS Partei des Demokratischen Sozialismus (Party of Democratic Socialism) PPF Pension Protection Fund PSV Pensionssicherungsverein SERPS state earnings-related pension scheme SPD Sozialdemokratische Partei Deutschlands (Social Democratic Party) TEE tax-exempt-exempt UNICE Union of Industrial and Employers’ Confederations of Europe VOC varieties of capitalism

1

The European dimension of the pension challenge

Workplace pension schemes represent social, labor market, and economic goals. They provide social protection in old age and serve as staff retention device for firms’ most valuable employees. They have also been used as a financial service instrument to complete the single market in the European Union (EU). While much has been written about the social function of workplace pensions, less attention has been devoted to the use of pensions as market-making device in the EU. Yet, studying the role of workplace pensions in the EU’s internal market is intriguing because it highlights tensions between supranational regulations and domestic pension systems. In the European Union, age-related spending represents a large share of public expenditure and is therefore an issue of common concern among the member states. Pensions from public pay-as-you-go (PAYG) schemes are the main source of income for older Europeans. In 2012, the EU as a whole spent more than 10 percent of gross domestic product (GDP) on PAYG pensions. This share is expected to rise to 12.5 percent of GDP in 2060 (European Commission, 2012b: 4). Due to declining fertility rates and increased life expectancy, the population aged sixty-five and above is expected to increase markedly in the coming decades. This group will almost double, rising from 87.5 million in 2010 to 152.6 million in 2060 in the EU (European Commission, 2012a). Unfavorable demographic developments, falling employment rates, and persistent financial instability put enormous pressure on public budgets and make it harder for state pension systems to deliver on benefit promises. To confront these challenges, many countries have provided citizens with economic incentives to build up workplace pensions. However, the expansion of occupational pensions remains less effective as long as they are not transferable across borders. Barriers to workplace pension portability, such as differential tax, investment, and social regulations,

1

2

The European pension challenge

imply that people who move across borders lose a significant portion of their pension entitlements. Thus, the prospect of pension losses has prevented people from exercising their right to live and work in other member states. Segmented pension markets also impose high administrative costs on multinational firms, who have to comply with twenty-seven different tax, investment, and social regulations. About 83 million people in Europe were covered by a workplace pension plan in 20101 and European pension funds managed assets worth 3.5 trillion euros, or 30 percent of the European Union’s GDP (European Commission, 2010). The absence of a level playing field in the area of cross-border pension portability has long inhibited pension funds from using the internal market efficiently. Lower returns on investments entail higher costs for the pension industry, ultimately leading to lower incomes for Europe’s savers. Growing concerns over the incompatibility of funded pension schemes with cross-border labor mobility catapulted this issue onto the European Commission’s agenda.2 After decades of acrimonious negotiations and bargaining failure3 in the 1990s, the EU member states adopted a directive on “the Activities and Supervision of Institutions for Occupational Retirement Provision” (IORPs), wherein member states agreed on the minimum harmonization of investment rules, anti-discrimination measures, and a common supervisory framework.4 This path-breaking agreement constitutes the first step towards a single pension market. It permits the establishment of pan-European pension funds that manage the pension schemes of employees in different member states. This arrangement facilitates the cross-border portability of individuals’ workplace pensions and enables multinational firms to save costs by allowing them to manage all of their employees’ 1 2

3

4

This estimate includes both defined-benefit and defined-contribution schemes. In fact, the EU’s legislative activity in the area of pensions is nothing new. For the past five decades the EU has issued regulations on the coordination of social security systems to protect the social security pension rights of mobile EU citizens. The novel issue discussed in this book is the EU’s activism in the area of workplace pensions. The proposal that resulted in negotiation breakdown was the Proposal for a Council Directive relating to the freedom of management and investment of funds held by institutions for retirement provision, COM (1991) 301 final, OJ C 312, 3 December 1991. Directive 2003/41/EC of 3 June 2003 on the activities, and supervision of institutions for occupational retirement provision (IORPs).

1.1 Summary

3

pension assets in a single country, rather than twenty-seven different states.5 Pension market integration has profound political implications. European legal constraints have greatly reduced the capacity of national governments to influence workplace pension policy, notably in the areas of investment rules, waiting and vesting periods,6 treatment of foreign pension funds, and regulatory supervision. Because a single pension market creates direct links between domestic pension regulations and global finance, pension funds become uncoupled from the national political economies they have traditionally served (Ferrera, 2005). Since this uncoupling disturbs traditional practices and interest constellations, the growing number of EU pension directives produces intense political conflicts. Market integration and the management of cross-border externalities may drive EU legislative activity in this area, but the shaping of national pension policy choices is the outcome.

1.1 Summary of the main argument Despite progress in pension market integration, there is a general unwillingness to acknowledge the European dimension of the pension challenge. If scholars were asked who has control over pension policy in the European Union, most would give the same answer: only the member states! Control over pension regulations is considered one of the hallmarks of the sovereign nation-state. National governments are known fiercely to guard their social policy autonomy against intrusion by Brussels. The European dimension of the pension challenge has been ignored in the literature because the EU is routinely portrayed as a regulatory polity that specializes in apolitical issues of market creation (notable exceptions are Natali, 2008 and Guardiancich and Natali, 2012). According to standard accounts of the European Union, only the 5

6

There are currently eighty-four IORPs in Europe that are operating cross border. Nine countries are home states to IORPs while twenty-three states act as host states (European Insurance and Occupational Pensions Authority, 2011). The waiting period is the elapsed time before an employee earns the right to a corporate pension claim. The vesting period denotes the time from acquiring corporate pension contributions until the benefits are actually owned by the employee.

4

The European pension challenge

member states are in control of highly politicized government functions, such as pensions and other social policies. Voters pay little attention to the EU because legislative activity only concerns issues of low political salience. While the EU’s task is to oversee market-creating regulations, it is considered patently unfit to handle politically charged conflicts of social policy that can only be settled by democratically elected governments. This book disagrees with this notion. We show that the EU influences national pension policy choices, and that national pension regimes have shaped the European single pension market. Pension policy developments at the national and EU levels are linked in two ways: debt and deficit constraining treaties, such as the Maastricht Treaty, exert pressure “from above,” compelling European governments to overhaul their costly public pension systems. In turn, the gradual harmonization of workplace pension regulations arises “from below,” reflecting hard-won compromises between member states with radically different pension systems. In this book, we develop a theoretical framework that accounts for the transformation of pension policies in the member states, and the creation of a single pension market at the EU level. The theory will be introduced in four steps. First, we investigate the sources of pension reforms many European states have implemented in the 1990s and 2000s.7 Given the high risks associated with the overhaul of public pension systems, it is puzzling to observe that so many countries have either introduced funded components to their PAYG pension systems or created incentives for building up workplace pensions. Such reforms are considered high-risk political endeavors because they replace the principle of public social insurance with more individual responsibility for old age income. Organized beneficiaries of public pension benefits are expected to oppose such measures. Existing studies are limited in explaining the reforms. Works attributing pension reform to purely domestic pressures, such as rising longevity and low fertility (Taverne, 2001; Disney, 2003) cannot account for the timing of pension reforms. Studies that have linked pension regime transitions to the diffusion of policy ideas (Orenstein, 2003, 2008; Weyland, 2005; Brooks, 2007; Gilardi, 2010) have convincingly demonstrated that diffusion dynamics have shaped pension 7

Most West European countries opted for the introduction of funded components, rather than wholesale privatization.

1.1 Summary

5

privatization measures in Latin America and Eastern Europe. Yet, no study has found a relationship between learning processes and pension reforms in Western Europe (the countries we are interested in). If domestic factors fail to explain the timing of the observed reforms, and diffusion dynamics do not apply in the case of Western Europe, we must consider other sources of policy change. We propose that the observed pension reforms in Western Europe are causally related to the fiscal constraints of the Maastricht Treaty. Using an event-history set-up, we hypothesize that the Maastricht criteria, which limit the build-up of public sector deficits and debt over time, acted as “common shock” for the implementation of pension reforms in several member states. While EU treaties exert a constraining effect on national pension regimes “from above,” the gradual harmonization of workplace pension regulations arises “from below.” The adoption of the IORP directive, which established the harmonization of investment rules, the non-discrimination of foreign pension institutions, and an EU-wide supervisory regime for pan-European pensions, is genuinely puzzling: what explains member state cooperation in such an “un-European” policy area? Given that workplace pension regulations differ radically across countries in terms of their function, financial design, and levels of social protection, one might have expected workplace pensions to remain trapped in national systems. Adjusting to a single pension market will be relatively painless for countries with a Beveridgean pension system,8 but costly for countries with a Bismarckian insurance culture.9 Thus, our second task is to analyze how the “battle” of pension regimes played out at the EU level. Our theoretical contribution is to show that governments’ capacity to send costly, and therefore credible, signals to one another is the key to successful EU institution-building in this policy area. Although many member states may prefer pension market integration to fragmented pension markets, they will not agree to policy change at any price. Member states with low costs of reform may be willing to accommodate governments whose reform costs are high, but it is difficult for them to distinguish between member states with genuinely high 8 9

Beveridgean countries include the UK, Ireland, the Netherlands, Denmark, Sweden, and Switzerland. For a comprehensive analysis, see Ebbinghaus, 2011. Bismarckian countries include Germany, France, Belgium, Austria, Greece, Spain, and Italy. Ibid.

6

The European pension challenge

costs of reform and governments who are “bluffing” in order to get a more favorable EU agreement. In this situation, the domestic pension reform discourse will serve as a signal about the country’s type: a reform-oriented pension policy discourse serves as a costly and credible signal that a country has genuinely high costs of adapting to a single pension market, whereas a status quo oriented pension policy discourse indicates “cheap talk” and is therefore not credible. Two-level game approaches have convincingly demonstrated that domestic veto points can increase a government’s bargaining power in multi-level negotiations (Schelling, 1960; Putnam, 1988). However, this scholarship has failed to explain why international negotiations may break down despite governments asserting that their domestic win-set is constrained. In the same vein, liberal intergovernmentalists (Moravcsik, 1998) cannot explain why negotiations over a pension fund directive broke down in the early 1990s. At the time, the member states had a strong incentive to cooperate on a single pension market since unfavorable demographic developments, soaring unemployment rates, and financial exigencies had exerted pressure on PAYG pension systems. Although member state preferences had converged, negotiations ended in deadlock in 1994. According to liberal intergovernmentalism, however, bargaining should never break down when there is preference convergence. Our analysis shows that a government’s inability to demonstrate credibly that domestic constraints are a major obstacle to EU harmonization efforts can lead to negotiation breakdown. This argument comports with a growing stream of research that has shown that informal governance tends to be more consequential for the design of international institutions than formal decision-making procedures (Koremenos, 2001; Christiansen and Piattoni, 2003; Farrell and H´eritier, 2003; Achen, 2006; Stacey, 2010; Stone, 2011). Third, informal signaling and communication flows also influence the effectiveness of the European Commission as agenda setter. As the only institution in the EU that can formally propose legislation, the Commission can play a powerful role in securing member states’ commitment to EU covenants. However, failure to act as “efficient” agenda setter may cause the Commission to get defeated in the formal decision-making process. As a result, the Commission has an incentive to assemble strategic coalitions before a proposal is put to a formal vote. Using primary document analysis as well as interviews with EU Commission officials and pension industry representatives in

1.1 Summary

7

Germany and Britain, we find that pension market integration becomes more likely if the Commission can trim a crowded agenda, set aside internal rivalries, and enhance the quality of information flows regarding workplace pension reform options and limits in other countries. In turn, a lopsided representation of supply-side interests, rivalries between Directorates-General (DGs), and failure to restrict the menu of policy options will diminish the Commission’s chances of garnering support for EU pension directives. Fourth, we gauge the real-world relevance of the theoretical considerations with an in-depth analysis of the pension policy discourse in Germany and Britain. The pension systems in both countries reflect the ideal types of Bismarckian insurance cultures and Beveridgean pension fund systems, respectively. Due to ideological disagreements over the proper role of occupational welfare, political incumbents representing Bismarckian and Beveridgean pension regimes are expected to advocate radically different visions of a European single pension market. The two case studies carefully trace the pension policy discourse in Germany and Britain between the early 1980s and 2007, highlighting how domestic pension policy ideas informed actors’ bargaining stance at the EU level. Space constraints prevent the inclusion of more than two case studies in the book. However, we exploit the fact that other EU countries with Bismarckian pension systems sided with the German position, while member states with Beveridgean pension fund cultures supported the British viewpoint. Thus, Germany and Britain represent critical cases. Germany’s position on the single pension market was heavily influenced by its concern over unfunded book reserve pensions, the most popular vehicle of workplace pension provision in the country. Helmut Kohl, Gerhard Schroder, and Angela Merkel all tried to pro¨ tect employer-sponsored book reserve pensions from the scope of EU directives. By contrast, Britain’s stance on the single pension market was shaped by Tony Blair’s commitment to continue Margaret Thatcher’s philosophy of individual risk-taking. As a result, the British pension policy discourse is characterized by bitter disputes between the financial services industry demanding lax EU pension regulations and consumer advocates requesting more efficient protections against missellings and corporate pension losses. This book employs statistical analysis, formal modeling, and case study research. In combination, these modes of analysis offset one

8

The European pension challenge

another’s shortcomings. Quantitative analysis may identify empirical patterns and correlations, but fails to establish nature and direction of causal effects. A formal model allows us to derive hypotheses and illustrates possible results of strategic interaction, but without an indepth analysis of pension policy developments in individual countries we cannot establish validity of the assumptions or gauge the realworld relevance of the findings. Without any of these methodological supports, scholarship becomes unreliable, like a “stool with only two legs” (Stone, 2002: 235). When combined, however, the result is a more thorough picture of economic interests, informal signaling, and EU institution-building in the area of workplace pensions. Our findings contradict the standard account of the EU as a regulatory polity that specializes in apolitical issues of market creation and leaves control of highly politicized government functions to the member states. The gradual emergence of the single pension market suggests that the EU is more active in the area of social policy than standard accounts have allowed for. Market integration and the management of cross-border pension portability may be the drivers of EU pension regulations, but the result is a constraining effect on domestic pension policy choices.

2

National pension regimes, supranational harmonization efforts

In this chapter, we provide essential background information that is necessary to understand what is at stake in the process of pension market integration. National pension systems are under pressure due to unfavorable demographic developments, financial constraints, and new social risks that are not covered by many traditional pension plans. These pressures politicize workplace pensions in new ways and differently across divergent types of pension regimes. This creates dilemmas for would-be reformers in Europe. On the one hand, the EU-wide harmonization of workplace pension regulations is attractive for large corporations in Europe because compliance with a single regulatory framework is easier and more cost-effective than compliance with twenty-seven different regulatory systems. Individuals also stand to gain from pension market integration since they would no longer lose pension benefits and rights if they moved across borders. On the other hand, even a partial harmonization of pension regulations at the EU level produces winners and losers since certain pension regimes are inherently better suited to accommodate EU-mandated change than others. Thus, the creation of a single pension market represents a classic cooperation problem: converging on common rules is desirable in principle, but deep divisions exist over objective and means of harmonization. Countries facing high costs of adjusting to EU directives will seek to keep the pain of reform to a minimum. This makes any common policy hard to adopt. To understand the factors that affect the choice set of would-be reformers, we first lay out how unfavorable demographic developments and new social risks put pressure on first-pillar pension systems. We then identify the multiple functions of workplace pensions in different countries to illustrate why the creation of a single pension market is a particularly difficult endeavor for EU harmonization efforts. The chapter concludes with a discussion of the goals and challenges of pension market integration. 9

10

National pension regimes

2.1 Pressures on public pension systems The pension systems of all EU member states consist of two tiers. The first entails the social security programs established by statutory legislation. The second is made up of employment-based arrangements created by employers, employer associations, and labor unions. In the second tier, these actors – and not the government – are responsible for administering the plans.1 Many pension systems also consist of a third tier, which refers to private pensions. In this case, it is the individual beneficiary – and not the state or the employer – who accumulates savings for his retirement. Different designs of pension schemes have crucial consequences for financial markets, labor relations, and the interaction of these two institutions for the welfare state.

Population aging It is now widely understood how contingent the PAYG pension schemes, financed by payroll taxes or social security contributions, are on favorable demographic developments. After World War II, most advanced industrial nations introduced major measures to extend social citizenship rights to pensioners, whether in the form of the New Deal, the Beveridge Plan, the Nordic Peoples’ Home, or the grosse Rentenreform (Esping-Andersen, 1996). German chancellor Adenauer’s well–known observation that “people will always have children” reflects the widespread belief that the ratio of those paying contributions and those receiving pension benefits would always be auspicious. Indeed, few political leaders had reasons to doubt the long-term sustainability of the so-called generational contract because, in the 1950s and 1960s, approximately eight workers financed the retirement income of one pensioner. Public finances are hugely influenced by pension expenditures. Today these expenditures represent approximately 10 percent of GDP in the twenty-seven member states, and are expected to increase further due to aging and low fertility rates to 12.3 percent in 2060 (European Commission, 2010). In addition to the increase in pension expenditures, member states will also have to finance age-related spending on 1

The terms corporate pension, second-pillar, workplace pension, employer-sponsored pension, and occupational pension all refer to work-related retirement income and are used here interchangeably.

2.1 Pressures on public pension systems

11

health care and long-term care. These costs are expected to increase from 8.1 percent of GDP to 10.6 percent of GDP in 2060. These numbers illustrate the magnitude of the issues at stake in the process of national and supranational pension policy development. The first challenges to the golden age of welfare state expansion arrived with the breakdown of the Bretton Woods systems of fixed but adjustable exchange rates and with the oil shocks of the 1970s, when unemployment rose sharply and wage growth stalled across Europe and the US. Though the deep recessions they induced were temporary, most advanced economies now seemed locked into permanently lower levels of employment and slower productivity growth (Scharpf and Schmidt, 2000). The emergence of massive and persistent unemployment was compounded by the low and declining labor market activity rates of younger and older cohorts. Traditional employment protection measures prevented or postponed labor market entry of younger workers. At the same time, many firms used early retirement options excessively for staff reductions. This means that middle-aged workers were shifted from the contributing population into the inactive group, leading to a dramatic increase of beneficiaries entitled to unemployment assistance at a time when high unemployment rates were already reducing the pool of contributors (Rosenow and Naschold, 1994; Martin and Thelen, 2007: 282). Greater spending demands arose with increasing longevity and declining fertility. Longevity implies that new cohorts of retirees will not only start collecting benefits earlier than their parents had, but they will also collect them until a higher age. In addition, the old have greater health care needs and a higher demand on other social services. Health spending on the age group of over sixty-five is typically four times that on those under sixty-five. It also became apparent by the mid-1970s that the “baby bust” in most European countries signified a fall in fertility rates that would reduce the relative size of future contributions to social security pensions. Both Italy’s and Spain’s birth rate, at 1.3 per woman of childbearing age, is approaching half the replacement rate. Germany’s birth rate is a similar 1.4 per woman. There is some indication, however, that birth rates in Britain are rising again from the present rate of 1.8 per female (Federal Trust for Education and Research, 2010: 10). While aging has plagued the advanced economies since the 1970s, international institutions afforded unfavorable demographic

12

National pension regimes

developments world-wide attention in the early 1990s. In 1994, the World Bank published a prominent report that promoted a multipillar approach to pension reform. Pension privatization was touted as a means to increase domestic savings, enhance growth, and build domestic capital markets. However, the report specified that the choice of radical reform, such as wholesale privatization, or more incremental reforms is contingent upon a country’s starting conditions and constraints in financing the transitional period (World Bank, 1994; Holzmann, 2000). How the World Bank report influenced domestic reform debates in Europe depended on national trajectories of pension policy developments and the influence of veto players on domestic policies. In the UK, Thatcher’s neoliberal ideology of market solutions to policy needs was causally responsible for the dramatic moves in the direction of privatization long before the World Bank had made colorful references to an impending “pension time bomb.” The majoritarian political system enabled the radical nature of the pension reform. A very different case illustrating the importance of national frameworks in shaping reform dynamics is Germany, where the process of reunification in the early 1990s marked a break with the partisan consensus and “smooth consolidation” that had characterized the German welfare state in the postwar period (Offe, 1991). Ideological conflicts flared up when large parts of reunification costs were imposed on the contributors to unemployment insurance and social security pensions, while the larger pool of tax payers was spared. It was in this context that population aging received more political attention. But pension policy developments in Germany are characterized by incremental, rather than radical, reform. More radical policy changes occurred in the Central and East European countries (CEEs). In the 1990s, several CEEs implemented major pension privatization measures that resonated with the World Bank model. By 2002, six out of ten CEEs had either privatized their pension systems wholesale or introduced funded components. The magnitude of pension reform in the new member states surprised many observers as these countries initially struggled with the transition to market economies and democratic politics. However, wholesale pension privatization may have happened precisely because of the rapid pace of institutional reform following the end of the Cold War.

2.1 Pressures on public pension systems

13

Traditional and new social risks Another rationale for the expansion of workplace pensions in Europe arose with the transformation of labor markets from a society based on traditional industry to a service-based economy (Iversen and Wren, 1998). Whereas traditional employment patterns had fueled expectations for lifelong job tenure, labor market biographies in the postindustrial era tend to be unstable. The shift towards flexible labor relations, such as part-time work or short-term work contracts, has greatly reduced social protection for most Europeans in the event of unemployment, sickness, and old age (Taylor-Gooby, 2004; Davidsson and Emmenegger, 2012). Career interruptions, domestic care obligations, single parenthood, and low or obsolete skills make it difficult for individuals to acquire both social security and workplace pension rights (Anderson and Meyer, 2006). The inability to reconcile work and family life can result in pension losses and, ultimately, poverty (Bonoli, 2006). Government responses to economic crises in the 1970s and 1980s were organized around saving the core manufacturing economy, and, as a result, the use of early retirement exploded during this time (Palier and Thelen, 2010). This policy instrument was supposed to make labor market exit for older employees less painful, provide firms with more flexibility, grind down on employment security for labor market insiders, and sugarcoat official unemployment figures. Yet, the early retirement option became discredited, and often subject to budget cuts, when it became clear that employers had used this instrument excessively for staff reductions. While “old social policies” have dealt with traditional risks by means of income protection, new social policies are aimed at covering risks that are more common in a service-based economy, such as atypical employment, long-term unemployment, working poverty, family instability, and lacking opportunities for labor market participation due to care obligations or obsolete skills. New social policies typically entail two different strategies: they bring recipients back into gainful employment while eroding ex ante protection, or they cover previously neglected risk groups, such as labor market outsiders and single parents (Hausermann, 2012). In some cases, governments have imple¨ mented both strategies at the same time: traditional arrangements for labor market insiders are maintained even as protection is extended

14

National pension regimes

for previously neglected risk groups. Such two-pronged strategies are commonly referred to as dualization, and they can take many different forms (Emmenegger, et al., 2012; Thelen, 2012). The transformation to the post-industrial society coincides with growing pressures stemming from traditional social risks (unemployment, sickness, aging) and is particularly problematic for young and middle-aged generations. Unlike their parents, younger generations can no longer expect social security pensions to provide a comfortable transition from work to the retirement phase. However, young cohorts often lack the incentive to build up employer-sponsored pension rights, particularly when the contributions to a pension plan (as opposed to the benefits) are taxed. Likewise, individuals lack the opportunity to build up workplace pensions when restrictive acquisition rules, such as long waiting and vesting periods, impose pension losses on employees who switch jobs. Governments have responded to the challenges of aging populations in distinct ways. While pension fund cultures like Britain have already moved towards funded private schemes in the 1980s, Bismarckian nations took steps to make their existing social security systems more viable, such as tightening eligibility criteria. In many cases the latter did not just cut back on PAYG pension benefits, but re-balanced the insider focus of their social insurance schemes to the needs of labor market outsiders. As a result, some aspects of benefits actually expanded (Hausermann, 2010). The measures have corresponding effects: by ¨ the year 2060, spending on social security pensions (as percentage of GDP) is expected to rise to 14.3 percent in France, 18.4 percent in Germany, and to 15 percent in Austria and Belgium. Only in the UK is the prospect for the taxpayer less bleak – pension spending is forecast to rise to only 5 percent by the year 2060 (Federal Trust for Education and Research, 2010: 43). However, the fiscal rectitude in the UK comes at the expense of the impoverishment of those individuals who rely on social security pensions alone.

2.2 The multiple identities of workplace pensions Workplace pensions have multiple identities and served a range of different purposes in various European countries. It is well known that employer sponsored old age insurance has a longer tradition than the social security pensions Otto von Bismarck introduced in

2.2 Multiple identities of workplace pensions

15

the late nineteenth century in Germany. Early nineteenth-century firm-sponsored insurance programs reflect employers’ sense of responsibility for their workforce at the time, commensurate with the paternalist ethos commonly ascribed to conservative welfare states (EspingAndersen, 1990). Yet only a very small percentage of the population had access to such workplace pension plans. The logic firms acted upon is appropriately described as pre-industrial: “The employer adopted the role of a gratuitous feudal lord, paternalistically dispensing benefits in order to reward or secure unwavering loyalty. Pressure from below was, in this period, mainly notable for its absence.” (EspingAndersen, 1996: 331). The perceived role of corporate pensions and its diffusion changed in the twentieth century, when employers realized that offering corporate pensions was in their own self-interest, and not just a charitable or humanitarian responsibility (Mares, 2003). Vanguard corporations like Krupp, Siemens, American Express, Kodak, and Cadbury introduced workplace pensions as “golden handcuffs” for their most valued employees. The importance of occupational pension plans as a staff retention device is particularly evident in coordinated market economies (CMEs) (Soskice, 1999; Hall and Soskice, 2001). Protective employment security regulations increase the long-term horizon of firms and strengthen the incentives of employers to invest in the skills of their staff members. To protect their investments, employers need institutions that reduce the risk of competitors poaching their skilled workers. A corporate pension that belongs to employees only after long waiting and vesting periods is an effective staff commitment device that encourages employees to stay with their firm for a long time (Manow, 2001; Iversen and Soskice, 2001). Which groups strongly support the expansion of second tier pensions? Political incumbents in Bismarckian pension systems may be willing to expand second-pillar pensions in order to reduce the fiscal overload caused by revenue shortages, population aging, and declining fertility rates. But because governments must uphold commitments to existing social security pension claims by large elderly populations, the introduction of funded components is often preferred to wholesale privatization, as the latter tends to be dangerous for re-election-seeking politicians. Thus, in insurance cultures, workplace pension expansion is more likely to function as a supplement to, but not substitute for, social security pensions.

16

National pension regimes

Unsurprisingly, highly skilled white-collar employees often demand a workplace pension plan to compensate for what they perceive as too modest public benefits. However, labor unions have a more ambiguous attitude towards occupational pension coverage. The social democratic model of the welfare state assumes that labor movements would systematically disfavor occupational plans because of their divisive inegalitarian properties. Workplace pension plans are quite unevenly distributed among the population. The majority of employees with access to employer-sponsored plans work in large corporations concentrated in the unionized sector, whereas occupational pension coverage is rare in smaller firms or within the lower end of the service sector. In contrast to social security pensions, occupational plans do not extend the kind of inviolable social rights that statutory legislation provides and are too easily a potential weapon of managerial power (Esping-Andersen, 1996: 328–329). Furthermore, occupational pensions magnify existing inequalities because employees who have access to a corporate pension already receive a higher social security pension (Schoden, 2003). The labor union perspective defines the relationship between public and occupational social protection in terms of a zero-sum game: public PAYG pensions are inversely related to occupational pension schemes and their growth therefore undesirable. Other scholars have found, however, that labor unions may welcome the expansion of secondpillar pensions when they are invited to share some control over the plans (Hassel, 2007: 24). According to this viewpoint, workplace pensions are not considered a threat to social security pensions, but a desirable supplement. From the perspective of private companies, workplace pension plans are not just an instrument to commit the most valuable employees to the firm for a long time, but they also constitute a cheap source of capital in periods of scarcity. When workplace pensions are provided in the form of unfunded book reserves, the sponsoring employer can use the pension funds to finance current business activities. Unlike funded American 401(k) plans or British trust funds, wherein the sponsoring firm is separate from the trustee, firms sponsoring book reserve pensions administer their employees’ pension assets themselves (Ahrend, 1996). Ultimately, this money will have to be paid out as corporate pension. But until workers have reached retirement age employers can do what they want with the funds they keep. Employer insolvency can

2.2 Multiple identities of workplace pensions

17

lead to plan members, especially those still in work, losing many of their acquired rights (Casey, 2012). In the aftermath of 1945, book reserve pensions were particularly prominent when workers’ contributions to these schemes met urgent post-war requirements for internal investment. Such pension schemes were offered by German firms and, despite all fundamental differences in the pension system, a few French schemes. In Sweden, surplus contributions to a supplementary pension scheme that was controlled by the state were also initially used to fund industrial infrastructure. Book reserve pensions were critical for financing early reconstruction efforts after 1945. Funded German pension schemes, by contrast, were completely wiped out during the hyperinflation of the 1920s, and again after the war was lost since pension funds had been compelled by law to invest in government bonds. After the monetary reforms of 1948 German pension funds again lost value as they converted from Reichsmark into Deutschmark on a ratio of 10 : 1. By contrast, contributors to book reserve pension schemes were less penalized, as their future pension relied not on monetary reserves but on current contributions and future company profits (Whiteside, 2006: 46). In Germany, book reserve pensions are still hugely popular because employers cherish the staff retention function as well as the cheap capital such pensions provide. There are six important dimensions of a pension system: (1) Pay-as-you-go (PAYG) versus funded pensions. PAYG pensions are supported by inter-generational income transfers. Taxes and social security contributions paid by the current working population finance the pension benefits of current retirees. In this case, no accumulation of surplus occurs. Funded pensions, by contrast, are designed to pre-fund future pension benefits. This system accumulates capital, which needs to be invested and administered. The accumulated contributions and returns on capital finance future pension expenditures. (2) Defined-benefit (DB) versus defined-contribution (DC) pension plans. Under a defined-benefit plan, the benefit to which the employee is entitled is determined by a formula which typically links the annual pension to the employee’s years of service and earnings’ history. Such schemes can be unfunded (no separate fund), underfunded (fund is worth less than the present value

18

National pension regimes

of promised benefits), overfunded (fund is worth more than the present value of promised benefits), or simply funded (fund is worth the present value of promised benefits). The employer bears all investment risk and guarantees to make up any shortfall. In defined-contribution schemes, by contrast, the benefit to which the employee is entitled is based on the accumulated contributions made on the employee’s behalf, together with the investment income earned on these contributions. A key feature of a defined-contribution plan is that the beneficiary bears all the investment risk, and by definition the retirement account is always fully funded. (3) Pension acquisition rights. Waiting and vesting periods determine when employees may accrue occupational pension claims. The waiting period is the elapsed period before an employee earns the right to an occupational pension. The vesting period constitutes the time from acquiring occupational pension contributions until the benefits actually belong to the employee. Both waiting and vesting periods vary considerably across EU countries, ranging from a maximum of two years in the UK to five years in Germany.2 In Luxembourg, national law allows waiting periods to last up to ten years, while in Spain waiting periods cannot exceed two years. Employees leaving their pension plan before completion of waiting and vesting periods forfeit their retirement income. (4) Taxes versus social security contributions. Pensions may be financed from general tax revenue or social security contributions. Although the actual distributive content varies depending on the precise nature of the tax structure, a tax-financed system is likely to create a broader base for the pension revenue. (5) External versus internal funding. When a pension scheme accumulates a surplus, the assets may be invested externally or internally. In the former case, an external trust fund or other financial institution that is independent of the sponsoring employer invests the assets on behalf of the beneficiary. Alternatively, pension assets may exist as on balance sheet pensions, or unfunded book reserve pensions, wherein the employer becomes the pension institution and can use the pension assets to finance current business activities.3 2 3

Until the law was changed in 2001, the vesting period in Germany was ten years. We use the terms on-balance pension schemes and book reserve pensions interchangeably.

2.3 Politics of pension market integration

19

(6) Benefit security. Governments may impose portfolio restrictions on the investment of pension assets by setting limits for equity investments, for instance, to safeguard pension money against undue risks. Other safety measures include stringent solvency rules that require pension funds to be funded to a certain extent at all times. In countries where unfunded book reserve pensions are provided, a mutual insurance association guarantees employee pensions in the event a firm goes bankrupt and defaults on pension claims. These dimensions are critical in understanding the impact of pensions on labor relations and capital flows. For example, workplace pension schemes that are externally funded are, by design, much more portable from one employer to another – and across countries – than unfunded book reserve schemes. Such differences systematically influence member states’ preferences over a single European pension market.

2.3 The politics of pension market integration Member states are free to organize their pension systems as they wish. However, the lack of a European framework governing the portability of workplace pensions makes it unattainable for citizens to exercise their right to live and work in another member state. Despite the free mobility of persons in principle, the absence of pension portability in Europe ensures the persistent fragmentation of labor markets. Portability barriers exist across countries as well as within any nation due to restrictive tax laws, investment rules, social regulations, and different supervisory approaches. Until 2003, divergent regulatory frameworks hindered funds from using the single market freedoms and the euro effectively. This reduced the return on investment and therefore raised the cost of the benefits. Barriers to workplace pension portability discriminate against workers with discontinuous labor market participation. Many workplace pension plans belong to employees only after long waiting and vesting periods. Unstable labor market biographies will result in irregular contribution payments to pension plans and therefore in lower pension benefits during old age. Career interruptions are particularly problematic for women since care for children and elderly relatives remains highly gendered. The same problem applies to migrant workers when changing country (Frericks, 2012). Due to divergent tax,

20

National pension regimes

social, and investment regulations in the EU, they risk losing their pension rights and benefits when moving across borders. The loss of workplace pension rights is noticeably more ruinous in countries where state-sponsored social security pensions are low and ungenerous, because a large share of people’s income is earmarked for privately building up pension rights (Myles and Pierson, 2001). To create a common pension market, the member states had to address a myriad of questions: which type of retirement institutions should be subject to EU supervisory control; what kind of investment regulations should apply; what are appropriate funding levels for pension liabilities; and what kind of supervisory regime can efficiently regulate pension institutions. As the EU progresses from negative to positive integration (Tinbergen, 1965; Scharpf, 1996), even small technical details can have a significant impact upon peoples’ social risks, employers’ administrative costs, and financial service providers’ legal scope of action. In the EU, member states tend to adhere to an informal norm of reciprocity, wherein governments refrain from imposing economically or politically costly decisions on one another. Domestic opposition to EU directives can, therefore, increase a government’s bargaining leverage at the EU level. Member states with severe domestic constraints should receive more generous terms (such as opt-out clauses). However, informal practices of reciprocity are invariably prone to manipulation, reinterpretation, or misuse (Kleine, 2013). If member states doubt their peers’ claims of domestic opposition to EU harmonization efforts, they may prefer to let negotiations collapse instead of granting concessions to reform-unwilling countries. Thus, how do member states decide whether a country “deserves” more generous terms in an EU agreement or whether the country exaggerates domestic constraints in order to get a better deal? We argue that the domestic pension policy discourse determines the costliness of the signal. A reform-oriented discourse, which generates domestic opposition, serves as a costly signal to other member states. A status quo oriented policy discourse, by contrast, indicates “cheap talk” and is therefore not credible. Governments’ failure to send credible signals increases the likelihood of bargaining breakdown. As Chapter 4 will show in detail, countries with Beveridgean pension systems mistrusted the “narrow domestic win-set” assertions of Bismarckian states in the early 1990s because the latter had repeatedly torpedoed attempts to

2.3 Politics of pension market integration

21

expand workplace pensions. As a result, domestic constraints claims were unconvincing, and negotiations on the single pension market broke down. By 2003, however, several member states had already implemented pension reforms, in many cases despite fierce resistance from organized interest groups (Schludi, 2005; Immergut, et al., 2007). Domestic opposition to pension reforms is highly visible and can be dangerous for re-election-seeking politicians. Consequently, a reformoriented pension policy discourse serves as a credible signal to other member states that an incumbent’s room for maneuver is indeed constrained. A country that can demonstrate credibly that only a narrow range of outcomes will be politically acceptable at home has a higher likelihood of receiving more generous terms in an EU agreement. The implication is that pension market integration cannot be understood by looking only at traditional theories of EU integration or domestic political economy models. Existing scholarship on EU integration misses the critical role of domestic constraints in shaping actors’ bargaining strategy on EU pension regulations, while approaches focusing on the national level miss the way in which the EU offers new institutional solutions to the problem of fragmented labor and pension markets. The neofunctionalist approach (Lindberg and Scheingold, 1971) would attribute efforts to create a single pension market to harmonization measures taken in the banking and insurance sector. According to this viewpoint, banking and insurance market integration “spilled over” to the pension industry, prompting pension funds to demand their own single passport in order to operate EU-wide on the basis of a single license. This perspective views pension market integration as incomplete unless the trillions of capital accumulated in occupational or private pension funds may cross borders and get invested more efficiently. While employers and pension funds tend to focus on competition issues, employees view pension market integration as essential for removing obstacles to labor mobility. Given the prospective benefits of pension market integration – increased labor mobility, deeper capital market integration, lower administrative costs for multinational firms – the neofunctionalist model would predict that member states should have a strong interest in creating a single European pension market. To overcome collective action problems associated with reforming domestic welfarefinance regulations, this perspective expects the most powerful states

22

National pension regimes

in the EU – France and Germany – to make asymmetric contributions to the single pension market, because they are assumed to benefit most from cooperation. Yet, this is contradicted by empirical evidence. Discussions in the European Parliament, in the Parliamentary Financial Services Forum, and position statements by employers and labor unions reveal that representatives of Bismarckian insurance cultures frequently questioned both goals (pension portability, labor mobility) and instruments of harmonization (liberalization of investment rules, cross-border membership, mutual recognition). At times, these nations even tried to torpedo the whole enterprise.4 While a functionalist approach can plausibly show that integration efforts in some internal market areas (banking, stock market, and insurance regulation) can spill over to workplace pensions, it fails to explain the level of integration that was chosen. Intergovernmentalists focus on rational governments and commercial interests as the key players in the integration game (Moravcsik, 1998). This perspective emphasizes divergent national preferences over EU policies and the distribution of power in the union. EU integration is seen as a linear three-stage process of national preference formation, EU bargaining, and choice of EU institution. National governments, not EU institutions, are the most important actors in determining legislative outcomes. Intergovernmentalists would expect the most powerful states, Germany and France, to dictate the design of a common pension market and block integration attempts that are harmful to their own welfare finance arrangements. However, this is also contradicted by empirical evidence. In 2003, Germany and France, along with other Bismarckian nations, implemented far-reaching reforms in order to create the foundation for a single pension market. These changes range from the liberalization of stringent investment regulations to the acceptance of more foreign competition in the financial sector, even though both countries would have favored stricter EU-wide regulations. Evidence shows that both Germany and France accepted an EU pension directive in 2003 they had previously rejected. This indicates that power considerations were not the chief motivation of German or French negotiators. Given the public’s general distrust of second-pillar pensions in both countries, rejecting the directive would have been a politically harmless move. Yet, both countries accepted far-reaching 4

Personal interviews with six EU Commission officials, June 20–July 4, 2006.

2.3 Politics of pension market integration

23

changes that provided the foundation for the Europeanization of workplace pensions. The domestic constraints framework developed by Schelling (1960) and prominently applied by Putnam (1988) shows that governments that face veto points at home tend to have more bargaining leverage in international negotiations. However, this scholarship has failed to explain why international negotiations frequently collapse despite political incumbents claiming to face insurmountable domestic obstacles in their home states. The contribution of this book is an argument about the circumstances that make governments’ “domestic constraints” claims credible. Theories on domestic political economy models have yielded fruitful insights regarding the goodness of fit between the EU and member state institutions (Iversen, et al., 2000; Aoki, 2001; Fioretos, 2001; Thelen, 2004). The “varieties of capitalism” (VOC) works have powerfully refuted globalization theories that predicted institutional convergence in the EU. The VOC framework also helps us understand why Germany insisted on the exclusion of employer-sponsored book reserve pensions – the most popular workplace pension vehicle in Germany – from the purview of EU pension directives. The dual function of book reserve pensions as effective staff retention device and source of patient capital for German companies explains their popularity and durability. Although the European Commission repeatedly tried to include book reserve pensions in an EU directive, the German government successfully blocked all such attempts. Thus, the VOC framework helps explain the resilience of a cornerstone of Rhenish capitalism. A renewed interest in understanding the role of non-state actors in the process of European integration has induced scholars to pool insights of the EU integration literature with the varieties of capitalism approach (V. Schmidt, 2002; Callaghan and Hopner, 2005). Some ¨ have argued that EU integration has entered a new phase in which it systematically clashes with national varieties of capitalism, in particular with the coordinated type (Hopner and Schafer, 2006). Accord¨ ¨ ing to this logic, new incompatibility problems arise with efforts to complete the internal market. While product market integration was essentially about the removal of barriers (negative integration), the liberalization of services requires new institutions and thus a more radical transformation of national political economies (positive integration).

24

National pension regimes

However, concerns that EU policies systematically disadvantage the coordinated type of capitalism remain unsubstantiated unless such hypotheses are steeped in a precise theory of actors’ motives to lobby for the status quo, support reforms, or adjust to EU mandated change. “Goodness of fit” analyses study the congruence between EU legislation and national institutional frameworks. It is generally assumed that political representatives are invariably biased towards the institutional equilibrium underpinning their home state and therefore try to design EU institutions accordingly. Conflict arises because EU regulations that may fit well with the domestic institutions in one member state may clash with the institutions in another, causing social dislocations and a general dissatisfaction with “Europe.” This perspective has been applied to both macro-and microlevel studies. Macro-level examinations focus on the compatibility between EU regulations and member states’ polity (Katzenstein, 1997; V. Schmidt, 2006c), type of capitalism (Brinegar, et al., 2004), and monetary regime (Iversen, 1999). Micro-level investigations study feedback effects between EU mandated policies and the behavior of non-state actors (Sandholtz and Stone Sweet, 1998; Woll, 2006). Goodness of fit arguments are usually embedded within a historical institutionalist (HI) framework. HI assumes that history unfolds through positive feedback mechanisms of reinforcement, which are difficult to change (Stinchcombe, 1968; Mahoney, 2000; Pierson, 2000a). A more refined version of HI departs from the determinism of path dependence and instead considers a variety of indeterminate adjustment processes, including path renewal, modification, or replacement (Streeck and Thelen, 2005). Such accounts have much to say about the basic congruence between EU pension directives and national institutional frameworks. As evidence from pension policy developments in Germany and Britain shows, the fit between EU laws and domestic social policy traditions is an important predictor of a country’s preference for EU mandated change (Cowles, et al., 2001; H´eritier, et al., 2001; Ferrera, 2005). This does not mean, however, that government preferences over EU pension market integration can simply be “read off” from national institutional frameworks. Institutions by themselves do not determine the path of future policy adjustment (V. Schmidt, 2008). The success of EU integration efforts is also contingent upon the way in which political incumbents explain EU mandated change to their domestic audience, and whether this communicative logic resonates with

2.3 Politics of pension market integration

25

cognitive schemes of appropriateness in the member states (March and Olsen, 1989). Failure to conceptualize human agency may lead to the erroneous conclusion that some states are always in favor of integration initiatives whereas others are always opposed. The strength of ideas-based approaches lies in their ability to capture not only the instrumental but also the ideational significance of decision-making (Ruggie, 1998; Blyth, 2003; Abdelal, 2007; Stiller, 2010). Schmidt conceptualizes the normative dimension of policy processes as legitimizing discourse, which entails both a coordinative and communicative sphere of policy construction (V. Schmidt, 2002; Schmidt and Radaelli, 2004; V. Schmidt, 2006b, 2006c). Discourse entails two components: an interactive-procedural dimension (who speaks to whom) and an ideational dimension (what is said). The coordinative dimension of discourse serves political actors to deliberate differing options among them, while the purpose of the communicative sphere is to explain the policy program agreed upon at the coordinative stage to the general public. In the European pension “game,” the coordinative discourse is highly complex, because a multiplicity of political actors need to reconcile differing positions on a vast range of themes. Devising common policies in areas such as social protection, investment regulations, taxation, and supervisory control tends to conflict with national perceptions of what workplace pensions should accomplish. Adding to the intricacy is the number of different actors involved in decision-making: for example, while in Austria the ministry of social affairs was the key player in the pension market negotiations, the main actors in Italy included the Italian central bank, two supervisors (pensions and insurance), the ministry of finance, and CONSOB (the Italian Security and Exchange Commission). The EU Commission officials in charge of formulating draft proposals are affiliated with various departments in the Directorate-General for Employment, Social Affairs and Equal Opportunities, as well as the Directorate-General for the Internal Market. These bureaucrats maintain liaison with the relevant domestic actors, national representatives in the Council, elected officials in the EU Parliament, as well as lobby groups, such as the European Federation for Retirement Provision (EFRP, the EU umbrella organization of pension funds) and Comit´e europ´een des assurances (CEA, the EU umbrella organization of insurers).5 5

Personal interviews with six EU Commission officials, June 20–July 4, 2006.

26

National pension regimes

However, ideas-based approaches also have drawbacks. They have sometimes underestimated the mobilization to which cognitive templates are subject, and failed to specify why politicians favor a certain outcome and how they manage to get their preferences implemented. The usefulness of ideas-based approaches, then, depends on a clear specification of the conditions under which discourse exerts a causal influence on the dependent variable. In sum, theories concentrating on the EU level miss the critical role of domestic economic interests and institutions in shaping actors’ bargaining strategy at the EU level. Approaches focusing on the national level miss the ways in which the EU offers novel institutional solutions to the problem of fragmented pension systems and the lack of pension portability across borders. However, by using a methodologically plural approach it can be shown how different varieties of capitalism condition the preferences and constraints that actors bring to the EU bargaining table, and that credible signaling is necessary to reach a politically efficient agreement. The goal of the book is to show that the EU has more regulatory power over pension regulations than standard models allow for. While the EU imposes considerable constraints on domestic pension policy choices, national pension policy discourses critically influence the design the common pension market. Throughout the book, we will shed light on this interlocking relationship by analyzing the top-down imposition of EU rules on the one hand and the bottom-up projection of member state preferences on the other. We begin with an examination of the sources of pension reforms in Western Europe between 1980 and 2002. Since the purpose of pension market integration is to lower barriers to workplace pension portability, it is important to understand why so many countries in Western Europe decided to improve access to workplace pensions in the first place.

3

The sources of pension reforms in Western Europe

The privatization of public pension systems is commonly viewed as imperative to ease the strain on nations’ fiscal resources. Yet, the overhaul of state pension systems is a high-risk political endeavor. Because privatization replaces the principle of public social insurance with individual responsibility for old age income, organized beneficiaries of state pension benefits will vehemently oppose reform. Structural pension reform is also perilous from a financial cost perspective. By channeling pension contributions away from public to private pension funds, the government accumulates a deficit to cover the current pension liabilities during the reform’s transition period. Since the costs of pension reform accrue immediately but the long-term benefits – lower fiscal outlays – do not materialize until the distant future, it hardly surprising that governments pursuing such a strategy often suffer electoral defeat (Pierson, 2000b; Jacobs, 2011). Given the high risks associated with the reform of state pension systems, it is puzzling to observe that this measure was adopted by so many European countries in the 1990s and early 2000s. During this period, several European governments have either privatized their pension systems or introduced funded components. Existing studies do not offer unequivocal conclusions about the causal pathway of reform because they concentrate on a single impetus instead of probing a variety of possible explanations. Works attributing pension reform to purely domestic pressures, such as rising longevity, low fertility, and concomitant financial exigencies (Taverne, 2001; Disney, 2003) cannot account for the timing of pension reforms. Yet, how governments allocate costs and benefits over time powerfully affects historically grown perceptions of distributive and generational justice. Another influential stream of research locates the source of structural pension reform in processes of policy learning, wherein reforms in one country are linked to similar developments in peer nations (Orenstein, 2003, 2008; Weyland, 2005). However, temporal and geographic 27

28

Pension reforms in Western Europe

correlations in the adoption of policy reform may be due to “umbrella causation,” whereby an external stimulus – like rain – may induce similar, but spuriously correlated, reactions in multiple subjects – like the opening of an umbrella (Weber, 1978: 23). This chapter uses an event history set-up to test whether pension reforms in Europe carried out in the 1990s and early 2000s are the result of domestic factors, policy diffusion, or a common shock. We find that the macroeconomic constraints of the Maastricht Treaty effectively acted as “common shock” that pushed the member states to overhaul their state pension systems. The causal connection lies in member states’ need to put member states’ unfunded pension liabilities on a more solid footing. The growing top-heaviness of the age pyramid has plagued Europeans for decades, but it was the deficit and debt-constraining criteria of the Maastricht Treaty that severely limited governments’ scope to finance pension liabilities by expanding public debt. Furthermore, rating agencies increasingly concentrate on general government obligations, of which public pension liabilities constitute the largest part. Member states retaining large unfunded pension systems in light of unfavorable demographic developments may imply an unwillingness to honor sovereign debt obligations. A loss of credibility, in turn, will increase a country’s risk premium and therefore raise its borrowing costs. Although rating agencies infer meanings from a variety of macroeconomic policies, they regularly implore governments to make their overstretched social security systems more sustainable. Pension reform decisions certainly remain subject to demographic aging and financial cost considerations, but the pressures of the Maastricht criteria account for temporal proximity of pension reform across a set of countries with heterogeneous demographic profiles. We therefore contend that the “common shock” hypothesis comprises more explanatory leverage than the diffusion hypothesis or domestic pressure variables. While several works have referred to linkages between the Maastricht criteria and pension privatization efforts before (Eijffinger and De Haan, 2000; Holzmann, et al., 2003; Holzmann, 2006), this chapter systematically tests the relative importance of all three sources of pension reform. The challenge to distinguish processes of policy diffusion from umbrella causation is commonly referred to as “Galton’s problem” (Jahn, 2006; Franzese and Hays, 2008). Inadequate modeling of interdependence may lead scholars to exaggerate the effect of common

3.1 When and why?

29

shocks, while failure to control for shocks may induce analysts to overestimate the role of diffusion dynamics. Because the underlying causal logic shapes the political room of maneuver, getting the causal story right is important. The remainder of this chapter discusses the possible causes of pension regime transitions. We then perform an event history analysis that tests whether the timing of pension reform during the observed period is the result of policy diffusion, domestic pressures, external shocks, or a combination thereof.

3.1 When and why do countries reform their pension systems? While existing works have overwhelmingly focused on “why” reforms happen, the question of “when” has received only scant attention. This lack of attention is surprising, given that the reform of PAYG pension systems1 requires political incumbents to incur short-term costs in exchange for the promise of long-term benefits (a more sustainable pension system). Yet, the political business cycle approach tells us that re-election-minded incumbents do not have incentives to make far-sighted investments when the benefits do not materialize until the distant future (Alesina, 1989). Why make costly investments that will benefit the political successor? Jacobs (2011) has identified the conditions under which such costly investments in future goods might happen. He argues that, where voters are prospective, politicians relatively insulated from political competition, and organized interests want policy investment, political incumbents may survive the otherwise dangerous act of imposing short-term costs on the voters in exchange for long-term benefits. However, such investments may never happen – even under favorable circumstances – if political institutions grant effective veto power to groups that lose out from the proposed changes. While a burgeoning literature has documented the adjustment pressure that demographic aging poses for the European nations, only a few scholars have analyzed the impact of international factors, such as European Union directives, on domestic pension policies 1

In PAYG systems, pensions are paid from current payroll contributions or other taxes.

30

Pension reforms in Western Europe

(Pochet, 2003; Haverland, 2007). We contend that the adoption of the Maastricht Treaty and concomitant changes in governments’ marginal utility for running high deficits and debt levels – of which unfunded pension liabilities are the largest part – are causally related to the pension reforms in the 1990s and early 2000s. By accounting for the timing of pension reforms we seek to address shortcomings of several influential lines of research. The usual suspects in explaining pension reform are demographic aging and domestic political constraints. Two trends are particularly problematic in this respect: increasing life expectancy and declining fertility rates. Due to both factors, the share of people aged sixtyfive years and over will rise dramatically relative to the working-age population (fifteen to sixty-four years). Within the European Union, the age dependency ratio is projected to increase from a level of 24 percent to 49 percent in 2045, an increase that is much stronger than, for example, in the United States (Schludi, 2005: 15). Another prominent stream of research has linked pension reform to party ideology. According to this view, Left party power matters for the level of income replacement rates of benefits (Korpi and Palme, 2003). Some have argued that Left party power has an impact on social expenditure, indicating that the Left spends more (Garrett, 1998). Others have denied that political parties have any effect on social policy development (Rose, 1984; Pierson, 1994). Martin and Swank obtained inconclusive evidence, finding that government tenure of social democratic parties is only moderately associated with an increase in social spending (Martin and Swank, 2004: 600). A different set of scholars found that Left parties have a bruising effect on the welfare state. From this perspective, Left parties choose unpopular retrenchment policies to position themselves strategically. By moving closer to the center on the ideological Left–Right party spectrum, Left parties hope to gain confidence with the business community and centrist voters more generally (Ross, 2000; Kitschelt, 2001). In the absence of new partisan competition on the Left, this strategy may go without electoral punishment (Kitschelt, 1999). However, purely domestic approaches are incomplete in explaining the temporal clustering of pension reforms. The growing top-heaviness of the age pyramid has plagued European pension systems for decades, yet the majority of pension reforms were carried out in the 1990s and 2000s. To be sure, growing demographic pressures will cause demands

3.1 When and why?

31

on PAYG pension systems eventually to outgrow the fiscal ability of the state. Once a critical threshold is reached, the results are shortterm budget crises and efforts to rein in spending, including cuts in the pension system. However, none of the countries of Western Europe was near such a dramatic tipping point in the 1990s. Demographic pressures by themselves therefore cannot explain the timing of the pension reforms in these countries. The temporal dimension of pension reform poses an even greater puzzle for studies that attribute social policy reform to the partisan composition of government. Changes in partisan complexion have occurred throughout the period we are interested in and across almost all countries of Western Europe. Some authors have identified partisan waves, stating that the electoral fortunes of the Left and Right co-vary across nations (Midtbo, 1998; Kayser, 2009). However, there is no consensus in the literature on the directional effects of party ideology on pension privatization efforts. A different set of scholars has located the impetus for pension reforms in diffusion processes, wherein the decision to introduce reforms is shaped by similar developments in relevant peer nations. The causal mechanism by which the process of diffusion unfolds has been variously specified as one of learning (Heclo, 1974; Gilardi, et al., 2009; Gilardi, 2010), the appeal of certain ideas (Orenstein, 2008) or “soft power” (Keohane and Nye, 1989), competition (Simmons and Elkins, 2004), the dissemination of norms (Keck and Sikkink, 1998), or cue-taking by governments (Brooks, 2005, 2007). Popular perceptions of the EU as a major agent that seeks to harmonize policies in Europe through deregulation might induce analysts to attribute pension reforms to policy diffusion. Furthermore, the existence of extensive trade ties between all twelve EU countries in our sample could be an indicator for diffusion dynamics as integrated economies may also experience synchronization of partisan cycles (Kayser, 2009). Finally, institutions such as the World Bank have been instrumental in disseminating ideas about pension privatization as a strategy to deepen capital markets and increase domestic savings rates (World Bank, 1994; Orenstein, 2003). But, as valuable as the scholarship on diffusion has been, it often suffers from “Galton’s problem,” which denotes the challenge to disentangle peer dynamics from independent government reactions to a common shock. In cases where countries do not adopt reforms

32

Pension reforms in Western Europe

implemented by their peers, these are usually taken to be instances in which policies fail to take hold in other countries. Yet, such accounts are unsatisfactory, because they neglect the possibility that temporal and geographic correlations in the adoption of policy reform are shaped by other factors, such as international pressures. While many scholars have correctly linked pension reforms in Latin America and Eastern Europe to the transmission of social policy ideas (Muller, 1999; Orenstein, 2003, 2008; Weyland, 2005; Brooks, 2007), ¨ nobody found that diffusion mattered for the decision to reform pension systems in Western Europe. Thus, close correlations in timing and geographic proximity of reforms cannot be automatically attributed to an interdependent logic. However, we lack alternative explanations that could account for the decisions of West European governments to reform their pension systems during the observed period. Except for the pieces on policy diffusion, there is no systematic effort to explain the pension reform pattern across countries. Therefore, evidence pointing to common shocks should be considered as well.

3.2 The Maastricht Treaty shock: pressure from “above” The effect of external shocks on governments’ choice to reform PAYG pension systems has received less attention in the literature. However, it forms an important rival explanation to the diffusion hypothesis. We argue that the Maastricht Treaty, which limits the build-up of public sector deficits and debt over time, acted as a powerful external driver for the implementation of pension reform. To meet the Maastricht criteria, member states’ deficits could be no larger than 3 percent of GDP, and debt levels could not exceed 60 percent of GDP. In the view of the treaty’s authors, countries needed a strong disincentive to run excessive deficits in order to preclude monetary and financial instability in the future Eurozone. There are at least three causal pathways through which budget deficits affect the politics of pension reform. First, Brooks (2005: 282) argues that the likelihood of pension reform declines with growing deficits, as domestic financing constraints may preclude governments from shouldering the transitional costs that arise from pension reform. Such costs emerge as individuals begin to divert payroll contributions to privately managed pensions schemes. In the transition period, some governments may offer tax breaks to encourage citizens to set up

3.2 Maastricht Treaty

33

private pension schemes. But because the government must uphold its commitment to existing pension claims by large elderly populations, the resulting fiscal shortfalls may be enormous, contributing to the so-called “double payment” problem. As a result, high deficits may discourage governments from embarking on pension reform. From this perspective, the commitment to a deficit-restraining mechanism, such as the Maastricht Treaty, should increase the likelihood of pension reform. An alternative hypothesis is that the Maastricht criteria have decreased member states’ marginal utility for retaining high deficits and debt levels, of which unfunded PAYG pension liabilities constitute the largest part. Holzmann (2006) has argued that a Maastricht fiscal regime, together with enhanced labor market flexibility, mobility, and labor supply in aging societies, all demand some convergence in the area of pensions. In this context, it is important to note that the Maastricht Treaty alters the intertemporal policy trade-off of governments. Without the Maastricht Treaty, governments had incentives to put off reforming their pension system into the future. Why incur costs in the present when the benefits will not accrue until some distant point in time and might benefit the other guy? However, the deficit criteria altered political incumbents’ time horizon: the longer European governments failed to reduce their future PAYG pension claims, the tougher the actual cuts required to live within the confines of the stability pact would have to be. Therefore, the longer the delay, the higher the likelihood that governments would need to cut entitlements of current beneficiaries (Truglia, 2002: 4). The closer to the present the actual cuts in benefits are, the greater the risk that governments will suffer electoral punishment. In this respect, Maastricht-induced pressure may be seen as a way of overcoming the time inconsistency problem political incumbents invariably confront, namely incurring costs in the present to realize a collective good – in this case, a reduced social security pension deficit – in the future. The final reason why the Maastricht Treaty may be causally related to pension reform is that it changed the level of governments’ implicit and explicit debt obligations and, as such, affected the perception of their solvency (Fiess, 2003: 4). By creating the European Central Bank, national governments lost the option of expanding the money supply to meet debt obligations. Consequently, rating agencies may downgrade countries that fail to make their overstretched

34

Pension reforms in Western Europe

social security pension systems more sustainable for future generations (Eijffinger and De Haan, 2000). Rating agencies provide international financial markets with a common language of risk and carry the “force of law” in many countries around the world (Abdelal, 2007: ch. 7). Lower ratings indicate a decrease in a country’s readiness and willingness to meet debt obligations duly and therefore deter potential investors. This, in turn, may decrease the number of investment funds which will buy government bonds, and increase the interest rate on government debt. Privatizing a pension system can adversely affect sovereign credit risk particularly during the transition period. By channeling pension contributions away from the government and creating a deficit of resources to cover the current pension liabilities during the transition period, the resulting public debt can negatively affect a sovereign’s perceived creditworthiness (Cuevas, et al., 2008). Others argue that unfunded pension claims of nation states play only a minor role in assessing sovereign credit risk. Although calculations of future pension liabilities provide a projection of a given scenario, rating agencies do not necessarily expect the projection actually to materialize: Large future pension claims have not greatly influenced our ratings of government debt in the industrialized world, even where net present value calculations would indicate very substantial claims on government resources over a 20–30 year time horizon. We simply expect that the government will “default” in the future on its pension promises as currently written in law in a way that will favor creditors. (Truglia, 2002: 2–3)

The expectation of default implies, however, that the ratings of the affected countries could come under severe pressure unless appropriate reform steps are introduced. Countries such as Germany, Austria, and France have consistently received the highest scores, but rating agencies strongly emphasize the need for structural reform. According to Standard & Poor’s, “the ratings on Germany will be supported by a consistent long-term approach in addressing the challenges of eliminating structural budget deficits, increasing employment growth, and putting the increasingly overburdened health and pension systems on a more solid footing” (Standard & Poor’s, 2002).

3.3 Empirical analysis

35

The implication is that some pension systems are better placed to face demographic difficulties and external shocks than others. Violations of the Maastricht deficit criteria constitute evidence for “reform fatigue,” indicating that countries will accelerate reform efforts before they enter European monetary union, but discontinue them once they have joined the club (Vamvakidis, 2009). Reforms that modified the excessive deficit procedure in 2005 are considered to have dealt a death blow to the Maastricht Treaty (Buiter, 2006) long before the economic crisis of 2008 caused massive increases in government deficits. However, in the run-up to European Economic and Monetary Union (EMU), countries made visible efforts to meet the Maastricht criteria, including structural pension reform. The following section probes whether structural pension reforms in the 1990s and early 2000s were the result of domestic factors, the diffusion of policy ideas, the reaction to a common shock, or a combination thereof.

3.3 Empirical analysis Our analysis includes sixteen Western European countries from 1980 to 2002. The countries and the year in which initial reform was implemented are listed in Table 3.1. A look at the table shows that four countries implemented major pension reforms prior to the signing of the Maastricht Treaty in 1992. These countries are Austria, Ireland, Switzerland, and the UK. From the group of countries that adopted reform legislation after 1992, only Norway and Sweden refused to sign the Maastricht Treaty.2 Austria and Finland joined the Treaty when they acceded to the Union on January 1, 1995. Austria had applied to join the then European Community in 1989, Finland in 1992. Of the Maastricht signatories, only France and Greece failed to implement any form of major pension reform by 2002. Since we make causal claims about the effects of the Maastricht Treaty, we include a number of countries in the analysis that were not signatories. Unfortunately, the number of potential control cases is small. In order to find cases that are similar in social, political, and economic structures, we concentrate on Western Europe. There are only four states that did not sign the Maastricht Treaty in 1992. They 2

Norway decided in a 1994 referendum to stay out of the EU entirely.

36

Pension reforms in Western Europe

Table 3.1 Countries in study, 1980–2002

Country

Year of reform

Maastricht member

Mature occupational or private pension sector

Austria Belgium Finland France Germany Greece Ireland Italy Luxembourg Netherlands Norway Portugal Spain Sweden Switzerland United Kingdom

1990 1995 1992 – 2001 – 1990 1995 2000 1993 – 2002 1999 1998 1985 1988

yesb yes yesb yes yes yes yes yes yes yes no yes yes no no yesa

no no no no no no yes no no yes no no no no yes yes

a b

The United Kingdom opted out of introducing the euro. Became member upon accession to European Union.

are Norway, Sweden, Switzerland, and, arguably, Britain.3 However, we can exploit temporal variation in the research design. The signatory countries themselves were free of the influence of Maastricht prior to 1992. For these early years, they are therefore part of the control group. The phenomenon we seek to explain is the timing of pension reform, therefore a natural statistical set-up is event history analysis (BoxSteffensmeier and Jones, 2004). The dependent variable is pension reform, which is coded as one for the year in which a government decided to implement full or partial privatization and zero in other years. 3

Despite being a signatory to the Maastricht Treaty, Britain’s initial drive to join the single currency was derailed on ‘Black Wednesday,’ September 16, 1992, when the pound sterling was forced out of the European Exchange Rate Mechanism through speculative attacks.

3.3 Empirical analysis

37

Although it is common to distinguish between pension privatization and notional defined-contribution schemes (Brooks, 2007; Orenstein, 2008), we are more interested in the dichotomy of “reform – no reform,” rather than the scope of privatization. Qualitative differences in reform outcomes are a reflection of varying domestic political realities. What matters for our purposes is the timing of the initial decision to try reform, independent of policy outcomes. Since our sample size is small, we seek to disentangle the causal story by the simplest available means. To capture learning from similar countries, we include a variable that adds up the number of countries that already implemented reform. Learning effects should become larger as the number of countries that have successfully implemented reform increases. To avoid issues of simultaneity, we only count reforms that happened in previous years. To test for the effects of the Maastricht deficit criteria, we include a dummy variable that is coded 0 for all cases prior to 1992, and 1 for member states afterwards. Since the Maastricht Treaty was signed on February 7, 1992, we include this year as the first year in which we expect an effect on the probability of pension reform. For non-Treaty members, the variable is coded as 0 for all years. Since we are using an event history set-up with discrete time, we need to model the baseline hazard. Beck, et al. (1998) propose to model temporal dependence in time-series cross-sectional data using year dummies or cubic splines. We use an alternative that is easier to implement and uses up fewer degrees of freedom. Carter and Signorino (2010) demonstrate that a third-order polynomial of time is more efficient than year dummies and easier to implement than cubic splines. We count the number of years until reform separately for two different groups of countries. For countries that did not join Maastricht, the count starts in the first year of the sample, in 1980. This set of countries belongs to our control group. Once a country implements reform, we drop it from the sample for succeeding years. For those countries that did join Maastricht, the count also starts in 1980. These countries are part of the control group until they sign the Maastricht Treaty in 1992. Since we cannot observe when the countries would have implemented reform without Maastricht, the count stops in 1991, at which time the countries leave the control group through right-censoring.

38

Pension reforms in Western Europe

In the period after 1992, the Maastricht countries represent the “treatment group.” We therefore reset the count variable to zero and count the years since the signing of the Treaty in 1992. To test whether the Treaty has an effect on the probability of reform, we include a dummy variable for the years from 1992 onward, as discussed above. In practice, this allows for the probability of reform to shift for Maastricht countries relative to those in the control group.4 Other independent variables in the analysis are designed to address the influence of domestic political pressures. They are the percentage of the population over sixty-five years of age, government party ideology, and the type of pension system. The age variable captures the reform pressure European societies face as a result of gains in life expectancy and declining fertility rates, raising the specter of massive funding shortfalls with which to finance rising pension liabilities. From the financial cost perspective, one might expect the likelihood of pension reform to rise the higher the demographic pressure weighs on a given nation. Conversely, Pierson (1994) has shown that governments tend to resist cutbacks of mature welfare entitlement programs because retrenchment tends to be unpopular with voters. Thus, a “domestic audience cost” theory would predict that cutbacks of mature entitlement programs may prove too dangerous for re-election-minded politicians. Through this lens, the likelihood of pension reform should decline with rising demographic pressure. The population share of people older than sixty-five varies in the sample between 10.5 percent (Portugal in 1980) and 17.9 percent (Sweden in 1986). Most countries experienced a steady increase of the share starting in the early 1980s and extending throughout the sample period. Two exceptions are Sweden and Norway, which are the only countries that saw the share of over sixty-five-year-olds decline over long time periods, starting in Sweden after 1986, and after 1990 in Norway. The ideology variable captures the influence of party ideology on structural pension reform. It is a dummy variable that takes on a value of 1 for Left-leaning governments, and 0 for Right-leaning governments (Keefer, 2004). In the sample period, Right-leaning governments 4

This research design implies that the baseline probability of reform is the same in the control group and treatment group, conditional on the included covariates. We do not have theoretical reasons that would call for a difference in baseline hazards.

3.3 Empirical analysis

39

slightly outweighed Left-leaning governments, with a share of 59.6 percent. Looking at individual countries, we have cases of great ideological stability and those with frequent turnover. Luxembourg had a Center–Right government from 1980 to 2000, the year it implemented pension reform. In contrast, Italy swung between Left- and Rightleaning governments five times before it implemented reform in 1995. The distinction between Beveridgean and Bismarckian pension systems captures ideological disagreements over the proper role of the state vis-a-vis firms or private schemes in providing retirement income.5 ` Beveridgean pension systems are aimed at poverty prevention. Social security pensions are typically low and ungenerous, providing either universal flat-rate or means-tested benefits. The majority of people do not expect their state pension to secure their living standard during the retirement phase, and hence have to save for their own retirement. Consequently, Beveridgean pension systems tend to have mature occupational and private pension sectors. Bismarckian pension systems, by contrast, are based on the social insurance principle and provide earnings-related benefits aimed at status maintenance during old age (Esping-Andersen, 1996; Bonoli and Shinkawa, 2005; Schludi, 2005). Beneficiaries rely on statutory social security pensions as their primary income during the retirement phase, while occupational and private pensions are underdeveloped. Of the sixteen countries in the sample, only four countries (Ireland, the Netherlands, Switzerland, and the UK) correspond to the Beveridgean model. Austria is the only country with a Bismarckian pension system that implemented reform prior to the signing of Maastricht. The other four early reformers – Ireland, Switzerland, and the UK – all have mature occupational and private pension systems. While pension reform tends to be politically risky, we expect Beveridgean countries to have an easier time implementing reforms vis-a` vis their Bismarckian counterparts. First, the demographic time bomb is ticking less loudly in Beveridgean systems because the low levels of public pensions keep future costs under control. Secondly, moving further down the path of privatization is considerably easier when people 5

We follow Ebbinghaus (2011) in using the distinction between Bismarckian and Beveridgean pension systems. Esping-Andersen’s influential welfare regime typology (1990) is less useful for our purposes because it fails to capture the growing importance of occupational pensions in social democratic and conservative welfare regimes.

40

Pension reforms in Western Europe

Table 3.2 Cloglog regression: pension reform

Maastricht signatory Number of countries with reform Funded system Left government Population over 65 Time Time2 Time3 Constant ∗∗

Coefficient

Standard error

14.2∗∗ −1.34∗∗ 3.61∗∗ −0.34 0.412 0.933 0.0251 −0.000785 −18.8∗∗

(5.70) (0.654) (1.08) (0.787) (0.287) (0.704) (0.0612) (0.00201) (6.14)

p ≤ 0.05, tests two-tailed, n = 270, # of successes = 13.

already expect to save for their own retirement, knowing that the low public pension won’t secure an adequate standard of living during the retirement phase. Pension reform in Bismarckian systems, however, tends to be politically toxic because of the double payment problem – at least one cohort will be asked to pay for the public pensions of their parents’ generation and for their own.

3.4 Results We summarize results of the statistical analysis in Table 3.2. A first glance at the coefficient estimates shows that the Maastricht dummy variable has the expected positive sign and is statistically significant. In contrast, the variable capturing learning effects from other countries shows an unexpected negative sign. Instead of increasing the likelihood of pension reform, the larger the number of early adaptors grows the less likely are late movers to implement reform. Before we look at effect sizes and temporal dynamics, let us examine the control variables. Both the pension system dummy and the demographic aging variable show the expected sign, but only the former reaches statistical significance. In line with our intuition, member states with mature occupational and private pension systems find it easier to implement reform. An increasing share of older people is associated with a higher probability of reform, but this result is not statistically significant. Theoretical considerations produced conflicting expectations about the

41

80 60 40 20 0

Probability pension reform

100

3.4 Results

5

10

15

20

Years to reform without Maastricht

80 60 40 20 0

Probability pension reform

100

(a)

2

4

6

8

10

Years to reform with Maastricht (b)

Figure 3.1 Probability of pension reform: control group and Maastricht signatories

direction of the effect of government ideology on pension reform. In any case, the party ideology variable is not statistically significant. To gauge the substantive effect of the Maastricht treaty on the dynamics of pension reform, we need to look at how treaty membership affects the probability of reform. Figure 3.1 plots predicted

42

Pension reforms in Western Europe

probabilities of reform as a function of time since entering the sample for the control group, and time since joining Maastricht for the treatment group. Dotted lines show 95 percent confidence bands.6 The graphs were produced holding the demographic aging variable and the number of other countries with reform at the sample mean, and setting the left government and pension system dummies to 0. Graph (a) records the probability of pension reform for the control group, while graph (b) shows a similar graph for Maastricht members. For countries in the control group in (a), it takes on average about fourteen years before the probability of reform crosses the 50 percent threshold. The point prediction attains the upper 100 percent bound, but the lower confidence band never gets above a 10 percent probability of reform. In contrast, the predicted probability of reform for signatories of the Maastricht Treaty (b) crosses the 50 percent threshold within three years after signing the treaty. The point prediction also reaches the 100 percent bound, while the lower confidence band increases to about 40 percent for in-sample observations.7 A predicted probability of 100 percent obviously does not correspond with reality. But it is important to keep in mind that the reported predicted probabilities are based on a counterfactual experiment, holding other variables constant. In the sample, the number of countries that have implemented reform increases over time, reducing the probability of reform for longer time periods to below 100 percent. Figure 3.2 shows the difference in the predicted probabilities of reform between Maastricht signatories (the treatment group) and countries prior and outside the Maastricht Treaty (the control group). Upon entering the sample, Maastricht signatory countries are on average 20 percentage points more likely to implement pension reform than countries in the control group. Within four years, this difference increases to 100 percent. Since the lower 95 percent confidence band never includes the zero, this difference is statistically significant throughout. We therefore have strong evidence that Maastricht membership significantly expedited the prospects of pension reform, as predicted by our theory. We now look at the surprising negative effect of pension reforms in other countries. This finding stands in marked contrast to the 6 7

The bands are based on a parametric simulation method (King, et al., 2000). Since the sample ends in 2002, there are no observations in the sample for which the time since Maastricht is greater than eleven years.

43

80 60 40 20 0

Pr(reform | Maastricht) − Pr(reform | control)

100

3.4 Results

2

4

6

8

10

Years to reform

80 60 40 20 0

Probability pension reform

100

Figure 3.2 Difference in probability of reform

0

2

4

6

8

10

12

Number of countries with reform

Figure 3.3 Probability of reform, reform in Other Countries

policy learning hypothesis put forward by the diffusion scholarship. We find that observing other countries implement reform actually retards the prospects of pension reform. To gauge the size of this “negative learning” effect, consider Figure 3.3. The graph shows predicted probabilities of reform as a function of the count of other countries that previously implemented reform. The predicted probabilities are

44

Pension reforms in Western Europe

calculated for Maastricht countries, setting the time since the signing of the treaty to three years, the percentage of people over sixty-five years old to the variable’s mean, and the Left ideology and pension system dummies to 0. In this example, pension reform in other countries causes the probability of reform to drop all the way to zero if seven or more countries already have reformed. For one or two countries with reform, no delay is experienced. The biggest marginal decline in the probability of reform occurs when three to five countries have implemented prior reform. As was the case when we described the effects of time, when the predicted probability of reform attains its upper or lower bound (100 percent or 0 percent), we are faced with an unrealistic situation in which reform occurs deterministically. Again, this is the result of holding time constant, which in the sample varies alongside with the number of countries that have implemented reform. How can we make sense of the negative effect of reforms in other countries? It is possible that this is indicative of a negative form of learning. Observing policy reform in other countries allows politicians better to gauge the risks and unanticipated consequences of such a policy move. Anecdotal evidence suggests that implementing reform indeed can cause electoral backlash. During the 1990s, coalition governments in Germany, France, and Italy have lost elections because voters resisted their controversial pension reform plans. In Germany, chancellor Kohl was dethroned after sixteen years in office largely because his government had introduced a demographic factor into the pension formula that was to reduce the replacement rate from 70 percent in 1999 to 64 percent in 2030. Fulfilling a central campaign promise, the Red–Green successor government initially revoked this reform but later reinstated the demographic factor after realizing the extent to which generous PAYG pensions placed a strain on fiscal outlays (Streeck and Trampusch, 2005). In France, prime minister Jupp´e lost the election in 1997 after he tried to force pension cuts on private-sector employees without any negotiation. In Italy, a general strike organized by the labor unions in 1994 against the government’s pension reform plans resulted in the fall of prime minister Berlusconi’s Center–Right government (Bonoli and Palier, 2008). In all of these cases, the reform plans were limited to retrenchment measures, while wholesale pension privatization was considered too risky. Consequently, widespread resistance to pension cutbacks should

3.4 Results

45

Table 3.3 Comparing Beveridgean to Bismarckian systems,  Pr(reform) Percentile Variable

10th

50th

90th

Population over 65 Number of countries with reform Time in sample

0.0846 10.7 0.00130

0.240 2.92 0.103

0.500 0.000248 69.3

make politicians contemplating more ambitious reforms particularly cautious. However, it is possible that the observed negative relationship is the consequence of un-modeled heterogeneity between cases. If there are factors that drive some countries to reform earlier than others that we do not account for in the analysis (and that are not picked up in the baseline hazard), those cases would drive up the count of countries that already implemented reform. Countries that tend to implement reform later for unobserved reasons therefore would appear to be influenced by the early movers. The diffusion literature does not provide good guidance how to disentangle learning from such a statistical artifact. Accordingly, we leave it to future research to address this problem. Our remaining task is to take a look at the substantive effect size of pension systems. Beveridgean systems are more likely to reform than Bismarckian systems, but the effect size depends on the values of other variables in the equation. Table 3.3 summarizes the change in the predicted probability of reform when comparing the two systems. Each cell of the table shows the outcome of this comparison for a different profile of regressors. For each cell, we set the variable recorded in the row (population over sixty-five years of age, number of countries with reform, or time) to the percentile listed in the column (10th, 50th, or 90th sample percentile). All other variables are held at their respective means, and the Maastricht and Left government dummies are set to zero. We then calculate the predicted probability of reform for Beveridgean and Bismarckian systems, and report the difference between these numbers. The table shows that the pension systems can have substantive effects, but only if the overall probability of reform is relatively high. This is the case if the number of countries with reform is relatively low (10th percentile of number of countries with reform variable), or after

46

Pension reforms in Western Europe

a relatively long period of time (90th percentile of time variable). In the former case, Beveridgean systems have a 10.7 percentage points greater probability of reforming than Bismarckian systems. In the latter case this difference is 69.3 percentage points. Overall, our analysis has produced strong support for the hypothesis that pension reforms in Western Europe in the 1990s and early 2000s were driven by external factors. Joining Maastricht and preparing for EMU clearly accelerated the drive towards reform. We were able to distinguish this external influence from a possible learning effect. We find that countries did not simply copy each other. On the contrary, observing early reformers appears to have made other countries more reluctant to follow suit. It is not clear if this can be attributed to a form of “negative learning,” or if this is due to un-modeled heterogeneity. From a range of domestic level influences on reform pressures, only the type of pension system exhibited a statistically significant effect. Unsurprisingly, governments with Beveridgean pension systems find it easier to implement reform. Political ideology or demographic pressures appear to have no effect on the actual timing of pension reform.

3.5 Conclusion Close correlations in the timing and location of pension reforms have been variously explained by either domestic pressures (such as demographic aging or overburdened social security systems), or policy diffusion (learning, emulation, or cue-taking). Yet, extant analyses are limited because they ignore the possibility that the temporal and spatial clustering of reforms may be the result of common shocks. Using event history analysis, this chapter has systematically tested whether pension privatization efforts in Europe are the result of domestic factors, the diffusion of pension reform ideas, or the external constraints of the Maastricht Treaty stipulations. Our findings indicate that the restraining effects of the Maastricht Treaty on national deficits and debt levels are causally linked to structural pension reform in Europe in the 1990s and early 2000s. In order to meet the treaty criteria, many European governments regarded structural pension reform as a necessary instrument to reduce budget deficits, of which unfunded PAYG pension liabilities constitute the largest part. We found no evidence that policy diffusion had a positive effect on governments’ decision to reform the pension system.

3.5 Conclusion

47

We do not dispute the importance of domestic pressure variables in explaining pension reforms, in particular demographic aging and financial cost of social security systems. However, these variables by themselves cannot explain the timing of reform efforts. Research on welfare state development has systematically neglected this temporal dimension. This is the first study to present evidence that the common shock of the Maastricht Treaty is causally linked to the clustering of pension reforms in Western Europe throughout the 1990s and early 2000s. Identifying the sources of policy reform is necessary to understand the conditions under which member states are capable of absorbing EU policy shocks that have distributive consequences. In light of Europe’s sovereign debt crisis and controversial discussions on excluding member states from the union, interest in measures to sustain policy coordination in times of crises is likely to grow in the years to come. The first step towards understanding such questions, however, is to establish the causal pathway of policy change. While this chapter has demonstrated how EU treaties can impose constraints on national pension policy choices, our next task is to analyze how the bottom-up projection of preferences influenced the design of EU-wide pension directives. The following chapter investigates the role of informal signaling and information flows between governments in the creation of the common pension market.

4

Informal signaling and EU-level bargaining

At first glance, the governance of pension funds across borders may seem like a simple target for EU harmonization efforts. The potential benefits seem sizeable: more integrated capital markets, fewer barriers to labor mobility and substantial savings in administrative costs for corporations. However, a single pension market requires institutional changes in sensitive policy areas. Member states must agree on the harmonization of investment, social, and supervisory regulations. Since the mature pension fund culture and liberal investment regulations in Beveridgean states fits well with EU pension directives, these countries face fewer adjustment costs and are therefore expected to support the creation of a single pension market. Countries with a Bismarckian pension system, however, face higher adjustment costs because a single pension market requires drastic changes in investment regulations, risk coverage, as well as waiting and vesting periods. These regulations, in turn, have a major impact on economic interactions between governments, workplace pension plan sponsors, and beneficiaries. Concerns that EU harmonization efforts might destabilize established patterns of labor relations makes cooperation in this policy area problematic. In this chapter, we ask why negotiations over pension market integration failed in the early 1990s, but succeeded in 2003. Any theory should explain both bargaining breakdown and success. We argue that the key to understanding negotiation failure and success is the process of informal signaling between the member states. In order to receive more generous terms in an EU agreement (such as opt-out clauses or more discretion in the implementation of EU directives), political incumbents must signal convincingly that their domestic win-set permits them to consider only a narrow range of EU settlements. However, merely claiming that their room for maneuver is constrained may not persuade other member states. 48

4.1 Decision-making in the European Union

49

The credibility of constrained win-set claims is linked to the domestic pension policy discourse in the member states. Reform-minded political incumbents are more likely to receive more generous terms in an EU agreement than officials pursuing a status quo oriented policy. Unpopular pension reforms that are associated with the political survival of the incumbent are considered risky and therefore costly. Thus, narrow win-set assertions made by reformers are considered credible, whereas the same avowal made by incumbents pursuing a status quo policy are not. Requests for accommodation at the EU level are more likely to be granted to member states that can demonstrate credibly the existence of domestic veto points. We find that the member states managed to adopt a politically efficient – albeit economically incomplete – pension fund directive in 2003 after credible signals had been exchanged. The null hypothesis is that the Bismarckian countries exaggerated their true costs of reform. Thus, an empirical test showing that the Bismarckian states were bluffing and could have implemented an EU agreement with only marginal costs would undermine our argument. The remainder of this chapter is organized as follows. The following section makes the case that we need to pay more attention to the informal government interactions before formal voting procedures take place. The subsequent part analyzes why governments with a Bismarckian pension system faced higher costs of reform. We then formulate the basic problem of pension market integration as a dynamic game of incomplete information, explicating the role of credible and non-credible signals in pension market negotiations. The subsequent part builds an account of the preferences of governments and domestic actors, relying on insights available from the VOC approach and discursive institutionalism.

4.1 Decision-making in the European Union: bargaining and procedural approaches In the majority of formal decision-making models, the reasons for and implications of negotiation breakdown are rarely explored. Proponents of procedural approaches, for example, have modeled the formal legal sequencing of the various EU decision-making processes. Their goal is to assess whether veto or gate-keeping rights empower an institutional actor, such as the European Parliament or the

50

Informal signaling

Commission (Tsebelis, et al., 1996; Kreppel and Tsebelis, 1999; Kreppel, 2002; Sbragia, 2002). The hidden assumption in these studies is that informal deliberations are unimportant compared to legal procedures. While not every legal rule is unequivocal, the clarity and traceability of legal procedures compares favorably to the obscure informality of pre-decision consultations. For example, in 1991 (the initial but unsuccessful attempt to adopt a pension fund directive) and in 2003 (when the IORP directive was adopted by both the EU Council and European Parliament),1 our two cases of interest, qualified majority voting was applicable. The introduction of the co-decision procedure with the Maastricht Treaty altered the decision-making process, although the direction of this change is contested in the literature. Some critics claim that the codecision procedure only strengthens the EU Parliament on paper, but effectively tips the balance of power towards the Council (Curtin, 1993). Others consider the Parliament a genuine co-legislator that is on a par with the Council (Crombez, 1997). The latter position seems more plausible. The emergence of a sizeable coalition in the European Parliament advocating a certain course of action can indeed translate into strong political pressure, which the Commission may exploit to win over reluctant Council members. Before the inauguration of the co-decision rule, a blocking majority in the Council was sufficient to torpedo any integration effort. However, we disagree with the notion that only formal decisionmaking procedures explain legislative outcomes. The outcomes of specific EU decision processes are not only affected by formal rules but also by informal government interactions that take place before voting takes place. Formal rules may set the boundaries within which action takes place, but they do not determine politicians’ behavior (Achen, 2006: 124). Informal deliberations prior to the legal steps are critical to political outcomes in the European Union, just as they are in every democratic decision-making body around the world. Cognitive and normative orientations define the “game” that is being played (Hall, 1993; Sabatier, 1998; Risse, 2001; Blyth, 2003). The deliberations preceding the formal decision-making process are often more consequential than the narrow legalities. We do not wish to 1

European Parliament/Council Directive 2003/41/EC, of 3 June 2003. The deadline for transposing the directive was September 23, 2005.

4.1 Decision-making in the European Union

51

diminish the importance of legal structures in shaping policy outcomes or the contributions of such work to our understanding of how certain decisions come about. After all, the passage of any EU regulation or directive is dependent on the formal ratification process. But an exclusive focus on legalities cannot explain why governments sometimes fail to reach an agreement despite tangible benefits from cooperation. The European Commission or Presidency frequently tests the waters to gauge which proposals may have a chance of getting formally adopted: “They do not vote on various proposals until one alternative wends its way through a set of legal procedures and emerges as law. Instead, it is the prior exploration of feasible outcomes, bargaining, and political power that dominate” (Achen, 2006: 126). Thus, understanding the nature of decision-making in the EU requires analysis of why informal negotiations sometimes result in deadlock, and sometimes in politically successful agreements. Although bargaining approaches have been successfully applied to address this question (Arregui, et al., 2004; Thomson, et al., 2006), they are unsatisfactory for studying feedback effects between the international and domestic levels. Historical institutionalists have shown on numerous occasions that domestic political parties, coalitions, and other modes of interest representation tend to be rather sticky. Therefore, we not only need to explore when and why politicians attempt to shift established policy patterns by brokering agreements at the European level, but also how EU directives might feed back on national elites and the public to sharpen certain cleavages and stifle others. We contend that a domestic constraints model, framed as a signaling game, is better suited for capturing the dynamics underlying the deliberations to the single pension market than standard procedural or bargaining approaches, because the actor-centered perspective allows us to understand more precisely how EU integration is linked to domestic politics. Modeling contestation over the IORP pension fund directive as a signaling game between European governments demonstrates how political success in one area of integration gives rise to success in another area. In contrast to simplistic “spill-over” arguments that are impossible to falsify, we offer a testable argument about how functional necessities arising from integration in a particular policy field (EMU) influence negotiation processes in another policy area (pension market integration). We also disagree with the notion that EU integration follows a linear three-stage process, ranging from

52

Informal signaling

preference formation at the national level to EU bargaining and institutional choice (Moravcsik, 1998). Instead, our model accounts for feedback effects that occur between those stages, including the updating of member states’ beliefs about each other, and the possibility that a European agreement may not reflect member states’ most preferred alternative, but second-best preferences. The focus of our model is on informal bargaining taking place before the legislative proposals are adopted as laws.

4.2 The costs of adjusting to an EU-wide pension market We hypothesize that the relative weight each government assigned to the costs and benefits of a single pension market determined its bargaining type at the European negotiation table. Governments can be of two kinds: those representing a Beveridgean pension fund culture and those a Bismarckian insurance culture. A single pension market affects the two types differently because of their distinctive welfare–capitalism arrangements. Because the Bismarckian states harbored occupational pension sponsors and beneficiaries most likely to suffer from integration, they found the creation of a single pension market politically more costly than the Beveridgean countries. Pension funds in insurance cultures are typically subject to heavy regulatory restrictions and investment limits, making it difficult for the pensions industry to attract the pension management skills it needs to be competitive (Iversen and Soskice, 2001; Deutsch, 2002). Most pension schemes offer biometric risk coverage, which denotes relatively expensive social protection against disability, survivor dependence, and longevity. Crucially, many employers in Bismarckian states offer corporate pensions that only exist on the books and are not funded.2 The method of internal financing is preferred by many employers because deferred taxation rules make these assets available for financing current business activities (Ahrend, 1996; Schoden, 2003). Furthermore, long waiting and vesting periods attached to book reserve pensions commit workers to the firm for many years and therefore 2

Such pensions are also referred to as on-balance sheet pensions. In Germany, where unfunded book reserve pensions make up two-thirds of all occupational pensions, the solvency of any not fully funded pension scheme has to be guaranteed by membership in the German Pension Security Fund Pensionssicherungsverein.

4.2 Costs of an EU-wide pension market

53

discourage poaching. Such on-balance sheet pensions provide firms with an effective staff retention device. Any move towards a single pension market posed a threat to this cheap source of capital, because the Beveridgean states intended to introduce funding requirements for all pension schemes (including book reserve pensions). This is because funding is commonly thought of as vital for safeguarding corporate pension entitlements.3 However, there is variation in adjustment costs across Bismarckian member states. Structural differences, such as the strength of labor unions (Hassel, 2007; Davidsson and Emmenegger, 2012), the extent of public sector employment (Martin and Thelen, 2007), state capacity to impose unpopular reform packages on social partners (Hassel and Ebbinghaus, 2000; van Wijnbergen, 2002), as well as the dominant workplace pension vehicle, all influence the precise costs of reform. While some states have successfully incurred short-term costs to introduce funded components to their PAYG pension systems, other governments may have refrained from doing so in order to establish, or maintain, a reputation as the guardian of the welfare state. Since pension reform is connected to a wide variety of welfare finance institutions in the member states, an objective quantification of the political costs involved in this process is hardly possible. The costs of pension reform may fall disproportionally on a single player (employers, consumers, the state, the financial industry, labor unions), or any combination thereof. Because the costliness of adjusting domestic pension systems to EU-mandated regulations is not directly observable from outside, governments have an informational advantage concerning their true cost types. This is particularly true for a government’s readiness to honor or default on debt obligations, of which pension liabilities constitute the largest part. Although EU negotiators meet in a variety of formal and informal settings and swap information about each other, they may have an incentive not to disclose their information fully or accurately. Furthermore, if there is no underlying consensus on the goals and design of the single pension market, member state representatives 3

There is no consensus in the literature whether book reserves or funded pensions are better at safeguarding beneficiaries’ occupational pension rights. Although the German Pension Security Fund reported record payments in 2009, the market turmoil following the global financial crisis has damaged public trust in funded pension schemes. See Casey, 2012 for an overview of this debate.

54

Informal signaling

Figure 4.1 Utility functions of Bismarckian and Beveridgean member states

might disagree on the best alternative despite all of the relevant information being shared. Such systemic biases may induce negotiators not to disclose fully the severity of constraints they face at home (AustenSmith and Feddersen, 2006: 209). We therefore presume that the true cost of reform is private information of individual member states. The following section develops a formal model that captures the capacity of the member states to send and receive signals, accounting for the discrepancy between bargaining inefficiencies in 1991 and informative exchanges in 2003.

4.3 Formal model Consider a simple game of incomplete information played between a single Bismarckian state C and a single Beveridgean state L in the European Union. Both are concerned with the extent to which they have to reform their own social and economic institutions following any agreement at the EU level. They bargain over a convex policy space defined by the closed interval [0, 1]. C’s ideal point lies at 0 and L’s ideal point lies at 1. The players are assumed to have single peaked preferences. As Figure 4.1 shows, the utility functions are monotonically decreasing when moving away from their respective ideal points. The player with private information (C) can be of the high-cost type (CH ) or of the low-cost type (CL). The two types differ with respect to their utility for an EU agreement, with the high-cost type incurring more adjustment costs resulting from integration. Only C knows his true cost type, whereas L needs to make inferences about C based on C’s actions.

4.3 Formal model

55

Figure 4.2 Sequence of moves

Sequence of moves As illustrated in Figure 4.2, nature moves first and determines whether Bismarckian state C is of the high-cost or low-cost type. The Bismarckian state moves next and offers a level of integration x, with x ∈ [0, 1]. The game is interesting because the Beveridgean state does not know which type of C it is negotiating with. L prefers to deal with a low-cost Bismarckian state to implement a single pension market that mirrors its own market-based social and economic institutions, but if L believes with certainty that C is of the high-cost type, L prefers to move further away from its ideal point to accommodate C. This means L accepts a less favorable point in the policy space to avoid the worst possible outcome (no integration, or maintenance of status quo). However, L has a prior expectation q ∈ [0, 1] about the distribution of types. Depending on the observed offer by C, the Beveridgean state L updates his belief about C. The updated belief μ denotes the probability L attaches to C being of the low-cost type. L takes the next step and decides whether to accept or reject C’s proposed point in the policy space. If L accepts, the game ends, and the pay-offs resulting from an agreement are distributed. If L rejects, the negotiations collapse and the players receive the status quo utility minus a penalty, if any, for bargaining failure. The pay-offs for INTEGRATION and bargaining BREAKDOWN are determined by the costs involved in implementing any negotiated agreement. They are as follows:

56

Informal signaling

BREAKDOWN U(CH ) = U(CL) = 0, U(L|CH ) = −k, U(L|CL) = 0. INTEGRATION U(CH ) = −r L x + I C , U(CL) = −r H x + I C , U(L) = −s(1 − x) + I L, where r L < rh and r L = 1, s = 1, I C , I L ∈ (0, 1), and x ∈ [0, 1]. Notation: r L and r H denote the costs of reform C incurs to reach an agreement with L (r L normalized to 1); [0, 1] is the policy space (or level of integration) both parties bargain over; I C and I L signify the respective benefits the Bismarckian and Beveridgean states derive from integration; and s represents L’s cost of accepting C’s offer (normalized to 1).

Equilibria Recall that information about the true cost type is private information of the Bismarckian government. This means the Beveridgean state’s decision about whether to accept or reject C’s integration offer is shaped by its evolving assessment of C’s true cost type. The appropriate solution concept in this game of incomplete information is Perfect Bayesian Equilibrium (PBE).4 Note that no separating equilibrium exists. In other words, the Beveridgean government knows that a lowcost Bismarckian government will always propose a point in the policy space that is more favorable to L and a high cost Bismarckian government will always propose a point less favorable to L. However, such a course of action is not consistent with the incentives of a low-cost Bismarckian government, which always prefers to pretend to be of the high-cost type in the first phase, inducing its preferred outcome, “low level of integration.” There are two interesting equilibria. One is semi-pooling, in which both BREAKDOWN and INTEGRATION can occur. Consider the 4

See section 4.7. For an overview of Perfect Bayesian Equilibrium, see, for example, Gibbons, 1992 or Ordeshook, 1986.

4.4 Domestic preferences

57

first case: if L observes a low integration offer it might mistake a high-cost Bismarckian government for a low-cost government that is bluffing to get a more favorable deal. Although the Beveridgean state knows it might be making this mistake, the probability of dealing with a low-cost C conveying cheap talk may be perceived to be sufficiently high that the Beveridgean state prefers to take its chances, reject the offer, and find out C’s true cost type. Under this outcome, L randomly challenges C, because claims about domestic constraints are cheap and thus cannot be informative. Since in this case L is trying to catch cheaters, bargaining can break down. In the second scenario, the semi-pooling equilibrium ceases to exist if the prior belief of the Beveridgean state about facing a bluffing low-cost type Bismarckian government is sufficiently low. In this case, there is a pooling equilibrium, where L simply agrees to accept C’s offer if it is better than the status quo and reject otherwise. No updating of beliefs occurs. L will always accept C’s offer, regardless of C’s true cost type, because in this case it does not pay to catch cheaters. Furthermore, C should offer the minimum necessary to get L’s acceptance. Although the Beveridgean state prefers to deal with a low-cost C to realize a European framework that is closer to its own ideal point, it clearly wants to avoid negotiation breakdown with a high-cost type C. This is because bargaining collapse with a high-cost Bismarckian state would signify a big victory for all risk-averse groups in Bismarckian states who are anxious to preserve the status quo, such as employers sponsoring book reserve pensions, risk-averse beneficiaries, or pension funds balking at foreign competition. It would make L look like a loser and considerably diminish the prospect of reaching an agreement in the future. Thus, if L is sufficiently certain that it is dealing with a high-cost C, it will always prefer to accommodate C and move further away from its ideal point to avoid bargaining breakdown, a worse outcome for L. Yet, as the above discussion has shown, errors can be made when member states’ signals are costless.

4.4 Pressure from “below”: domestic preferences over EU pension regulations The previous section analyzed the conditions under which bargaining may break down or lead to a politically successful agreement. The following part will examine the sources of domestic political and

58

Informal signaling

economic actors’ preferences over a single market for pension funds. Examining actors’ policy positions is crucial, because domestic contestation over social and financial regulations is part of the signal states send and receive at the European level. Because a study of all European states is beyond the scope of this chapter, the following section will be limited to an analysis of the pension policy discourse in Germany and Britain. There are good reasons for this selection. Germany represents the prototype of a Bismarckian insurance culture, while Britain corresponds to the ideal type of a Beveridgean pension fund culture. We also exploit the fact that countries with an insurance culture sided with Germany, while countries with a pension fund culture supported the British stance on pension market integration.5 Thus, both countries constitute critical cases. When it comes to EU-level negotiations on pension market integration, we assume that both Bismarckian and Beveridgean pension regimes want to keep the costs of reform as low as possible. Governments perceive costs and benefits of a single pension market differently, depending on the magnitude of changes they have to make to their distinctive welfare finance arrangements. Coordinated market economies are characterized by long-term relations between firm and workers, access to patient capital, generous biometric risk coverage, and a society lacking an equity culture (Aoki, 2001; Hall and Soskice, 2001; Coates, 2005). Long waiting and vesting periods reward loyal employees but punish frequent job changers, especially women who interrupt their careers for domestic care responsibilities. Because the goals of the single pension market (more efficient management of occupational pensions across borders, deeper capital market integration, and labor mobility) seem diametrically opposed to the coordinated market model, we expect governments representing this group of states to oppose the single pension market. Conversely, governments representing liberal market economies (and mature pension fund cultures) should embrace the Commission proposal, because the goals corresponded to their comparative institutional advantage. The general discourse in these countries largely reflects citizens’ acceptance that statutory pensions will not provide 5

This is documented in the European Parliamentary Debates (1999–2004) and in the minutes of the European Parliamentary Financial Services Forum (2000–2004).

4.4 Domestic preferences

59

for old age. Because the basic pension is low and ungenerous, occupational pension schemes are well established and characterized by substantial funding and significant voluntary provision, both of which are facilitated by flexible investment rules (Bodie, et al., 1996; Lynes, 1997; Davis, 2000). Key features of asset managers in liberal market economies are a strong international orientation and specific expertise in equity investment. In contrast to coordinated market economies, biometric risk coverage is rarely offered. Given the significant consequences of social policy and investment choices for the future operation of the single pension market, we expect to find strong disagreements both among and within countries over its regulatory design. The basic problem is the trade-off each government faces between creating the collective good of a single pension market and maintaining its own supervisory, investment, and social regulations. Official communications, position statements, and minutes by the European Council, European Parliament, and European Financial Services Forum indicate that the politically most contested issues revolved around (1) investment regulations, (2) biometric risk coverage, and (3) unfunded book reserve pensions. The subsequent section explores the factors that influence the sensitivity of governments and pension institutions to these trade-offs. Cross-national variation in investment regulations essentially mirrors the trade-off between a high-risk/high-return strategy and a low-risk/low-return approach. British and Dutch funds, for example, are free to invest more than 5 percent of the fund in any sponsoring company. German funds, however, have to meet no fewer than six separate limits: they cannot invest more than 30 percent in EU equities; not more than 25 percent in EU property; not more than 6 percent in non-EU equity; not more than 6 percent in non-EU bonds; and not more than 20 percent in overall foreign assets. In Denmark and Austria, pension funds may invest up to 40 percent of the funds they manage in equities; in France the limit is 25 percent (Deutsch, 2002). Surveys by aba,6 the German Occupational Pension Association, show that many member states still discriminate against foreign investment funds. Together with onerous administrative requirements, such as registration obligations, this constitutes a major obstacle to a single pension market. 6

www.aba-online.de.

60

Informal signaling

Restrictive investment regulations are generally a reflection of a riskaverse society that lacks an equity culture. They are routinely justified on the grounds that investors need protection against market risk, although the effectiveness of quantitative investment limits for reducing risk has been contested in the literature (Nurk ¨ and Schrader, 1995). In Germany, such a risk-averse attitude is especially pronounced where pension providers are subject to codetermination. Joint control over pension assets essentially affects the composition of a company’s executive board and implies that the investment behavior reflects the attitude of the insured. This is mirrored in the small proportion of shares in the monetary assets of German households. Thus, pension funds and their managers generally prefer steady profit development. Irregular developments would result in an irregular distribution of profits to the insured over time and is perceived as unfair (Deutsch, 2002). The British financial system, by contrast, imposes a relatively shortterm horizon on companies, but at the same time allows high risktaking. Flexible investment rules fit with the comparative institutional advantage Britain enjoys in the areas of investment services, asset management, and financial product innovation. Such investment rules facilitate the substantial funding of pensions and the development of the pension industry as one of the largest in the world. Key features of the international expertise of British asset managers are a strong international orientation and expertise in equity investment (Bodie, et al., 1996; Davis, 1999). While investor protection concerns dominate the discourse in Bismarckian nations, Britain is known for very liberal investment and supervisory regulations. Even after scandals like the Maxwell affair or the failure of Equitable Life, wherein large numbers of employees lost their future retirement income, British politicians and the Bank of England continue to defend the liberal regulatory regime because the majority of occupational pension plans are provided as final salary schemes, with the employer bearing all risk (Blake, 2003). British politicians defend flexible rules on the grounds that employers’ duty to honor final-salary plans requires them to pursue a highrisk/high return strategy, including equity investment. Even though many final-salary schemes have been closed for new hires and definedcontribution plans become more prominent (since the employee bears the risk), previous final-salary commitments still constitute a burden to many employers. Because the Commission advocated the adoption of the relatively liberal prudent person rule (which is already used in the UK), the British financial industry stood to gain most from a

4.4 Domestic preferences

61

single pension market. Since the City of London had a clear competitive advantage in the business of pension management vis-a-vis finan` cial centers in Bismarckian pension cultures (Talani, 2000), the British pension industry strongly lobbied for the adoption of the Commission proposal. The second controversial issue surrounding the IORP pension fund directive concerned the extent of biometric risk coverage, which denotes relatively expensive longevity, invalidity, and survivor insurance. The percentage of the population that has access to biometric risk benefits varies considerably across countries, and, within the same country, across different pension schemes. While German pension plans routinely offer biometric risk coverage, British plan sponsors offer this coverage to a very limited extent, or not at all. What are the institutional benefits for firms to provide relatively expensive social risk protection? It has been shown that the design of corporate benefits affects not only the labor market strategies of firms, but also the nature of competition in product markets (Swenson, 2002). If firms want to engage in radical innovation, they have to have easy access to risk-willing capital, they must be able to hire flexibly and to fire workers on the external market, and they must be able easily to reorganize work (Hall and Soskice, 2001). In Britain, social policies enter the utility of employers with a negative sign because they do not fit with the high-return/high-risk strategy British companies pursue. Biometric risk coverage raises the labor costs of companies, decreasing overall profitability. Other economic analyses have pointed to social insurance as an impediment on the ability of firms to deploy their labor market resources flexibly and, thus, as a source of “welfare losses” for the firm. The limited supply of biometric risk insurance demonstrates that the costs of social policies usually outweigh the benefits of social policies to British employers. In contrast to this, firms pursuing incremental innovation strategies need access to patient capital, a workforce with industry-specific skills, which in turn requires long-term relations between firm and workers, and a wage-setting system that prevents poaching of skilled workers by competing firms (Hall and Soskice, 2001). Book reserve pensions constitute an important part of this type of capitalism. The reason for their popularity is that they provide companies with a cheap source of patient capital while at the same time rewarding loyal core workers. Because such pensions belong to employees only after unusually long waiting and vesting periods, they provide an effective mechanism

62

Informal signaling

Investment restrictions

Status quo Ideal points of countries with an “insurance culture”

Ideal points of countries with a “pension fund culture”

Contract curve

Biometric risk coverage

Figure 4.3 Representation of preferences over investment rules and biometric risk coverage

against poaching. Scholars of the German welfare state have demonstrated that important segments of the business community have on numerous occasions supported institutions of social risk coverage (Mares, 2003). The nature of welfare finance arrangements in coordinated and liberal market economies allows us to map the preferences of Bismarckian and Beveridgean member states over this policy trade-off in a two-dimensional policy space. Given the significant consequences of social policy and investment choices for the future operation of a single pension market, one expects to encounter strong conflict and disagreement among different countries over these details at EU level negotiations. This is based on the assumption that representatives of both Bismarckian and Beveridgean states have an interest in keeping adjustment costs low. Therefore, one should expect governments to push for policies favorable to the regulatory framework in their own country. These considerations allow us to determine countries’ ideal points regarding a European pension directive in the spatial model. Figure 4.3 shows Germany’s ideal point in the upper right corner of the “high biometric risk coverage/high investment restriction” policy

4.4 Domestic preferences

63

space. Such an outcome would compel the European financial industry to abide by restrictive investment rules and force them to offer a certain set of expensive pension products. Although the formal agreement on EU level cooperation of the supervisory authorities would have made cross-border pension activity legally possible, de facto it would have done little more than perpetuate the status quo. Most firms would have abstained from setting up European IORPs in other countries, because the costly absence of a single pension market would have been merely replaced with the costs of navigating an overly restrictive one. The British ideal point lies diametrically opposed to Germany’s, in the lower left corner of the “low biometric risk coverage/low investment restriction” policy space. This outcome would create the most flexible regulatory framework and therefore the most cross-border activity. The European financial industry would follow the relatively liberal prudent person rule and offer only minimal biometric risk coverage, enticing many firms to join the lucrative pension market and offer services in other countries. Because companies could manage all of their employees’ pension entitlements in a single country, real savings in administrative costs would materialize. This version emphasizes the financial aspects of the single pension market and tones down the social concerns. Figure 4.4 illustrates Bismarckian and Beveridgean states’ preferences over investment rules and biometric risk coverage as indifference contours that are centered about their respective ideal points. The status quo is represented on the high end of the vertical axis and signifies the absence of a European regulatory framework. The shaded area is the intersection of Bismarckian and Beveridgean indifference curves and represents the set of all outcomes that are Pareto-superior to the status quo. Thus, moving from outside the status quo inside the lens always leads to an improvement. Note that the indifference curves are not presented as concentric circles, but ellipses. This means that, if a country moves away from its ideal point, the direction matters. This reflects the assumption about countries’ preference intensity. The preceding sections suggest that Germany felt more intensely about the issue of biometric risk coverage than investment rules, while the UK felt more strongly about investment rules than biometric risk coverage. What factors affect the sensitivity of governments to biometric risk coverage? As mentioned earlier, important factors that influence the sign of the utility along these dimensions are countries’ welfare finance

64

Informal signaling

Contract curve

Investment restrictions

Status quo

Ideal point of countries with an “insurance culture”

Set of all outcomes that are Paretosuperior to the status quo

Biometric risk coverage Ideal point of countries with a “pension fund culture”

Figure 4.4 Representation of indifference curves

arrangements. Empirical measures of the welfare finance nexus include the size and generosity of first-pillar statutory pensions, the maturity of second-pillar pensions, and their financial design. The generosity of statutory pensions in Bismarckian countries, the underdevelopment of externally funded pension schemes, and the importance of social policies for the employment practices of firms indicate that German negotiators should exhibit a strong preference for the inclusion of mandatory biometric risk coverage into the pension fund directive. Given the small size of statutory pensions in Beveridgean states, the maturity of its second-pillar pension system, and the fact that social policies do not play an important role in employment strategies of companies, we expect only a weak preference over biometric risk coverage. What factors affect the sensitivity of governments to investment rules? Germany shows a pattern for risk-averseness, the lack of an equity culture, and a preference for quantitative investment restrictions. However, since the majority of occupational pensions are provided in the form of book reserves and externally managed pension funds (Pensionskasse, support funds, and pension funds play only a minor role in Germany), we expect only a weak preference over

4.5 Domestic constraints

65

investment rules. By contrast, the long experience and international expertise of UK fund managers and consultants suggests that the British financial industry would gain disproportionally from a single pension market. This implies that, for Britain, any move away from the status quo into the shaded area would be an improvement. The contract curve in Figures 4.2 and 4.3 represents all Paretooptimal outcomes, where no side can deviate from its position without making the other side worse off. Due to the elliptical shape of the countries’ indifference curves, the contract curve is actually a curve, instead of a straight line. Pareto-optimal outcomes on the contract curve include the two countries’ ideal points as well as the universe of possible negotiated policy compromises between the two.

4.5 What makes “domestic constraints” assertions credible? The previous section has examined the sources of Beveridgean and Bismarckian countries’ preferences over a single pension market, contending that historical trajectories of welfare finance institutions determine governments’ willingness and capacity to adjust to EU pension policies. Yet, goodness of fit considerations only go so far in explaining EU integration efforts. We therefore need to examine the domestic pension policy discourse in those countries that sought accommodation at the EU level, and how their demands were perceived by their peers. Discourse analysis, which includes a coordinative and a communicative dimension, highlights both the substantive content of ideas and the interactive processes by which ideas are conveyed (Blyth, 2003; Schmidt and Radaelli, 2004; V. Schmidt, 2006c; Abdelal, 2007). The coordinative sphere comprises the key policy actors who talk to each other in order to reach agreement on reforms. Particularly in the Rhenish capitalism model, it is important to trace the interaction between employers, unions, and political parties to understand how cognitive schemes of legitimacy militate in favor or against policy reform. The communicative dimension of discourse denotes the processes by which government officials, opposition party leaders, organized interests, and the media incorporate new ideas into political programs, and communicate them to the general public (V. Schmidt, 2006c). “Ideational leadership” occurs when policy-seeking officials manage to expose the downsides of the status quo and make consistent efforts to redefine

66

Informal signaling

the values underlying well-established policies and institutions (Stiller, 2010). Discourse analysis avoids historical determinism by showing how political actors may or may not alter historical trajectories by acting differently within them as a result of new ideas conveyed by new legitimating discourses (V. Schmidt, 2006b). Discourse analysis serves to gauge the empirical validity of the gametheoretic model. Our model implies that Bismarckian states are not able to negotiate more discretion at the EU level without being able to send credible signals. Tracing the actual pension policy debate in Bismarckian nations will allow us explore the logic of the model and evaluate which signals were deemed credible and which ones were not. The following section analyzes the pension policy discourse in Germany, the most powerful representative of Bismarckian insurance cultures. Under chancellor Helmut Kohl, pension policy was largely status quo oriented. If the Christian Democratic pension policy in the late 1980s can be summed up in a single sentence, it is the one that Kohl’s long-time social affairs minister, Norbert Blum, famously plastered ¨ on a billboard in 1986: “One thing is certain: the [social security] pensions are safe.” This claim was supposed to counter Social Democratic accusations that the strain on fiscal outlays was growing due to demographic aging, and the state was not doing enough to put the overstretched social security system on a more solid footing. According to Blum, historical experiences with externally funded ¨ pensions strongly militated against an expansion of the second tier. In the 1930s and 1940s, the Nazi regime had compelled pension funds to invest in the war effort by buying government bonds. When the war was lost, employees’ pension assets were gone, depriving many people of their retirement income. Unfunded book reserve pensions, by contrast, had not been lost. Such pensions only exist on firms’ balance sheets, and the company can do what it likes with the funds it keeps. Eventually, this money will have to be paid out as a corporate pension. But, until then, these funds constitute a cherished source of capital for the firm. After 1945, book reserve pensions were badly needed for reconstruction efforts (Whiteside, 2006). Such experiences, Blum ¨ argued, militate against capital funding of corporate pensions and suggest that governments should give more weight to social security pensions, as well as tax breaks to companies sponsoring book reserve pensions.

4.5 Domestic constraints

67

Yet, the status quo oriented rhetoric contradicted German claims of being the “high-cost type.”7 There was no mechanism allowing the Beveridgean states to determine whether Germany was a sincere high-cost type whose domestic win-set was indeed constrained or whether it was bluffing to walk away with a more favorable agreement. Germany’s unwillingness to reform its pension system in times of demographic aging made it look more like a bluffing low-cost type abusing the “domestic constraint” logic to secure a better deal.8 After the embarrassing breakdown of negotiations, which Commission officials referred to as the “nuclear option,” pension market integration reappeared on the agenda in the context of EMU. As Chapter 3 has shown, the Maastricht Treaty prompted several member states to introduce reforms that strengthened second-pillar pensions. The creation of the European Central Bank deprived member states of their instrument to increase the money supply to offset government debt. The largest parts of government debt constitute unfunded health and pension liabilities. In aging societies, such future obligations put mounting strain on fiscal outlays. In this environment, many member states took measures to increase the retirement age or reduce pension benefits – or introduced a funded component to the traditional PAYG pension system. The implementation of these reforms, however, should not be taken to indicate that a compromise on the single pension market was a foregone conclusion. Although economic change in the run-up to EMU may have altered the costliness of creating a single pension market, both Bismarckian and Beveridgean countries still endeavored to design the pension directive in accordance with their own social and financial regulations. Adopting the IORP directive was not the only option member states had in responding to EMU pressures. Some member states had already taken considerable electoral risks by reforming their national pension systems between 1999 and 2002. They could have simply left it at that and allow subsequent EU pension negotiations to fail, just as they had in 1994. Failure of EU directives is much more embarrassing for the Commission than for national governments. Consequently, neither side was keen on making concessions at the EU level just to return home with an agreement. 7 8

Personal interviews with six EU Commission officials, June 20–23, 2006. Personal interviews with officials at the German Ministry of Labor, Bonn, July 2006.

68

Informal signaling

The adoption of the IORP directive in 2003 was possible because the member states exchanged costly and therefore credible signals. The pension policy discourse by Social Democratic chancellor Schroder ¨ represented a sharp break with the pension policy of his predecessor. Schroder introduced measures such as funded occupational pension ¨ schemes, along with cuts in benefits and social security pensions, exposing citizens to greater risk. These measures reflected the emphasis of his administration on personal responsibility, generational justice, and sustainability of the welfare state. One of the most painful decisions of Schroder’s tenure was reneging on his campaign promise to elimi¨ nate the so-called “demographic factor,” a mathematical formula his predecessor had introduced in 1998 into the PAYG system in order to account for rising life expectancy. The demographic factor would have reduced employees’ pension benefits from 70 percent to 64 percent. Whereas the demographic factor was eliminated as soon as Schroder ¨ took office in 1998, fiscal constraints forced him to reintroduce it in 2003, prompting the chancellor to admit that the initial elimination “was a mistake, no question about that.”9 Overall, Schroder’s pension policy was widely perceived as one that ¨ exposed both consumers of first- and second-pillar pensions to more risk.10 It met with a howl of criticism from the Left wing of his own party, labor unions, and the constituency that had voted for him. The ¨ Left-wing media vilified him as a “social benefit thief” (Sozialrauber) whose policies reflected a sell-out of Social Democratic principles. Likewise, Klaus Zwickel, the former head of Germany’s biggest labor union IG Metall, was reviled as a “strike-breaker” for failing to fight the chancellor’s social policy agenda with more union strikes.11 Opposition to Schroder’s pension reform was highly visible in the ¨ form of angry protests and vituperations in the media. The public outcry changed the beliefs of the Beveridgean states about Germany’s cost type. Vocal objections to Schroder’s social and pension reforms ¨ demonstrated that Germany was indeed a high-cost type. German officials feared that exposing beneficiaries to additional risks by agreeing to EU pension policies would only further fan the flames. Against this background, claims that Germany could not support a more liberal 9 10 11

Frankfurter Allgemeine Zeitung, October 8, 2003, “Blums Kritik an Merkel: ¨ Sein letzter Kampf.” Die Zeit, October 9, 1999, “Arme Junge, reiche Alte.” ¨ Neue Internationale, September 2003, Agenda 2010: Stoppt die Sozialrauber.

4.5 Domestic constraints

69

regulatory framework for pension funds were perceived as costly and therefore credible.12 As a result of this informative exchange, Germany was able to extract concessions from their Beveridgean counterparts, including the maintenance of quantitative investment limits (the Beveridgean states had originally sought a more radical liberalization), the principle of home country control regarding biometric risk coverage (the Beveridgean states originally wanted to leave the decision to provide biometric risk insurance entirely to the market), and the exclusion of book reserve pensions from the scope of the directive (which the Beveridgean states had sought to include). These two episodes explicate the circumstances under which informal signaling is causally relevant in influencing legislative outcomes at the EU level. Assertions of “domestic constraints” are convincing only if the domestic policy discourse is perceived as costly. Since the status quo oriented pension policy by Helmut Kohl contradicted German claims of being the high-cost type, German demands for more generous terms were not met. By contrast, Schroder’s demands for more ¨ discretion in the implementation of EU pension directives were deemed costly because his reform agenda was inevitably connected to his reelection prospects. For this reason, the Beveridgean states perceived Schroder’s demands as costly and therefore credible. ¨ The following section examines the level of integration – or point on the contract curve – that was chosen when the member states eventually agreed on the 2003 pension fund directive, and analyzes the extent to which the Beveridgean states accommodated their Bismarckian counterparts after credible signaling had paved the way to a consensus.

The 2003 IORP pension fund directive: who wins and who loses We now illustrate the ways in which the Beveridgean states – although initially reluctant to compromise – departed from their ideal points to accommodate the preferences of the Bismarckian states. To understand how EU decisions translate into domestic policies, we use Schmidt’s 2002 analytical framework, which distinguishes between different EU adjustment pressures (decisions accompanied by more or less highly 12

Personal interviews with six EU Commission officials, June 20–23, 2006.

70

Informal signaling

specified rules for compliance, suggested rules, or no rules) and potential adjustment mechanisms (coercion to a greater or lesser degree, mimesis, or regulatory competition) to predict how EU regulations affect national policies (inertia, absorption, or transformation) (V. Schmidt, 2002). Applied to the IORP directive, the member states agreed on coercion in the area of investment rules; on the principle of home country control regarding quantitative investment limits and provision of biometric risk coverage; and on the exclusion of certain occupational pension schemes from the purview of the directive. Pension institutions will be able to select investment managers established anywhere in the EU. They will also need to comply with the newly established supervisory authorities in Brussels, Frankfurt, and Paris. Thus, the result is a pension fund directive that strikes a careful balance between coercion, mimesis, and regulatory competition. Article 18(5) represents a sacrifice on the part of the Beveridgean states by allowing member states to “lay down more detailed rules, including quantitative rules,” as long as they do not prevent institutions from investing 70 percent of assets in equity, 30 percent of assets in foreign currency, and investing in risk capital markets.13 The Beveridgean states were initially opposed to this stipulation, because it provides pension funds in Beveridgean countries with a disincentive to gain a foothold in those countries. The Bismarckian member states insisted on its inclusion to make clear to their domestic constituents that they were protecting beneficiaries against market risk. The general risk averseness and lack of an equity culture in Bismarckian member states demanded nothing less. “Now they can at least protect their own nationals against unsound investments.”14 Although the Beveridgean states conceded the article on quantitative investment limits to grant their Bismarckian counterparts a “safety net,” the overall investment policy put forward by the directive indicates the liberalization of investment rules. In the spatial representation, this agreement denotes a point on the contract curve that is closer to the Beveridgean states’ ideal point. However, paragraph 5 qualifies this ostensible victory for the Beveridgean states by specifying that member states may apply more 13 14

Directive 2003/41/EC, article 18(5). Personal interview with six EU Commission officials, June 21, 2006 and June 23, 2006.

4.5 Domestic constraints

71

stringent investment rules in their respective home countries as long as they are prudentially justified.15 By using the formulation “prudentially justified,” this paragraph accommodates all member states uncomfortable with the prudent person rule, which they perceive as too relaxed. “Since it is unclear even to Commission members involved in drafting the directive text what prudentially justified really means, it was relatively easy to agree on it.”16 While coercion was limited to investment regulations, the social aspects of the directive were defined by the principle of home country control. This is because the divergent welfare-finance arrangements in the member states militated against a “one size fits all” solution. Although a majority of MEPs originally wanted to force pension institutions to offer biometric risk coverage if desired by employers or employees, the Council opposed any framework that would compel pension institutions to offer a certain product. Council representatives reasoned that, aside from restricting competition, this provision would also violate the principle of subsidiarity (Karas, 2003). Yet, where insurance against biometrical risks are provided, the directive requires pension institutions to have sufficient provisions to cover these benefits.17 Such pension schemes engaged in cross-border activity are also required to be fully funded at all times. This means that the Bismarckian states were able to maintain their ideal point on biometric risk coverage. At the same time, the principle of home country control allowed the Beveridgean states to retain their ideal point position on the same policy dimension, thus reflecting a Pareto improvement for both types of states. Finally, the inclusion of article 2, section 2(e) was very important to all Bismarckian states where book reserve pensions play a prominent role. Germany in particular insisted on removing unfunded book reserve pensions from the purview of the directive. This is because book reserve pensions would directly conflict with article 18(e) (“assets shall be properly diversified”). Although the Beveridgean states countered 15

16 17

“Paragraph 5 shall not preclude the right for member states to require the application to institutions located in their territory of more stringent investment rules also on an individual basis provided they are prudentially justified, in particular in the light of the liabilities entered into by the institution.” Directive 2003/41/EC, para. 5. Personal interview with six EU Commission officials, June 21, 2006. Directive 2003/41/EC, para. 30; article 15(2); and article 17(1).

72

Informal signaling

that an exemption for book reserve pensions would give an unfair advantage to employers using them to finance current business activities, it soon became clear that no compromise would be reached if employers were forced to surrender their cherished source of cheap capital and means for staff retention.18 The Bismarckian states were able to signal credibly that such far-reaching regulatory change would never be accepted by their domestic constituents. Consequently, only funded pension schemes will be subject to the definition of “institutions for occupational retirement provision”, while on-balance sheet pensions are exempt from the directive. This was acceptable to the Bismarckian states since externally funded pension schemes play only a minor role compared to book reserve pensions. The Bismarckian states, on the other hand, reluctantly agreed to article 16, section 2, which specifies that member states may allow institutions “for a limited period of time, to have insufficient assets to cover the technical provisions.” This provision clearly reflects the position of the powerful British pension funds, trumping the concerns of the European insurance industry.

Significance of the 2003 compromise The nature of the compromise the European member states reached in 2003 regarding the contested issue areas of investment regulations, biometric risk coverage, and book reserve pensions reveals two things: the Beveridgean states sacrificed more than their Bismarckian counterparts, and the Bismarckian states made concessions in areas they cared only marginally about. Although the overall regulatory approach to investment regulations was indeed “coercion,” the restrictions recorded in article 18(5) and paragraph 5 considerably softened the pain of liberalizing investment regulations for the Bismarckian states. Furthermore, by agreeing on the principle of home country control regarding biometric risk coverage and permitting the exclusion of book reserve pensions from the purview of the directive, the Beveridgean states allowed the Bismarckian countries to maintain their ideal points in two crucial policy areas. This outcome supports our hypothesis that the 2003 directive is the result of credible signaling on

18

Personal interviews with six EU Commission officials, June 20–23, 2006.

4.6 Conclusion

73

part of the Bismarckian states, while preferences over the design of the single pension market remained unchanged between 1991 and 2003. Although the IORP directive constitutes a first step towards a single pension market, it is unlikely that there will be an upsurge in crossborder pension portability anytime soon. Sensitive issue areas such as tax treatment of workplace pensions have yet to be sorted out before we can speak of a truly integrated pension market. Nonetheless, given the manifold political and economic obstacles to integration, the 2003 compromise must be considered a bargaining success.

4.6 Conclusion This chapter asked how the member states with their divergent pension and financial regulations negotiate solutions to the problem of governing pension funds across borders. Although governments with Bismarckian and Beveridgean pension systems stood to gain from integration in this area, they could not reach an agreement when the Commission had first proposed pension market integration in 1991. In 2003, however, an amended directive was passed by the Council of Ministers. We argued that the success of EU harmonization efforts in the area of pensions depends on the ability of governments to signal credibly that their domestic win-set is constrained. Signals sent by a government pursuing a status quo oriented pension policy are less credible than signals sent by a government embarking on ambitious pension reforms. Furthermore, this chapter showed that the Beveridgean states, although more enthusiastic about creating a single pension market, would not agree to EU harmonization at all costs. In 1991, failure to distinguish between sincere Bismarckian governments facing high adjustment costs and deceitful “low-cost” governments trying to mimic the high-cost type lead to bargaining breakdown. More than a decade later, however, the Council of ministers approved a pension directive that constitutes the first step towards an integrated pension market. We argued that the ambitious pension policy conducted by Social Democratic chancellor Schroder changed ¨ the Beveridgean states’ prior beliefs about the Bismarckian governments’ bargaining type. Requests for concessions by a reform-oriented Social Democratic government were considered costly and therefore received an accommodating, separating response from the Beveridgean

74

Informal signaling

countries. As a result, the member states adopted a pension fund directive authorizing high harmonization of investment regulations, low harmonization of funding requirements, and mutual recognition of biometric risk coverage. We do not, however, pretend that the model presented here is an exhaustive depiction of deliberations between member state representatives. The purpose of the signaling game is to provide a baseline model highlighting the role of information and informal signaling between the member states, an often overlooked aspect in the literature that has predominantly focused on formal decision-making processes. Future research may be guided by the baseline model, because it is important to understand the structure underlying information flows and the mechanisms that make credible signaling possible. While this chapter has focused on the negotiation dynamics between governments, our next task is to analyze the role of the Commission in pension market integration.

4.7 Equilibrium characterization A PBE consists of the players’ optimal actions, given the other players’ equilibrium actions and beliefs about types. To simplify construction of the equilibrium, assume CH has a dominant strategy of always offering x = 0. Let x∗ be the offer for which L is indifferent between accepting and rejecting, i.e. x∗ satisfies −(1 − x) + I L ≥ 0 ⇒ x∗ = 1 − I L. Next, we define off-equilibrium path beliefs. For any x ∈ (0, 1], let L believe it faces CL, i.e. it sets its posterior belief to μ = 1. This assumption has intuitive appeal, since CH has a dominant strategy of always offering x = 0, and thus not stands to gain from deviating to x > 0. With these off-equilibrium path beliefs (and with beliefs on the equilibrium path obtained by Bayesian updating), optimal strategies in a mixing equilibrium (corresponding to BREAKDOWN) are given by the probability that CL offers x0 = 0, p=

(π − 1)(I L + k − 1) , π (I L − 1)

4.7 Equilibrium characterization

75

where π is L’s prior belief about C’s type, and the probability that L rejects an offer x0 , q=

1 − IL x∗ = , C I IC

such that (existence conditions) k>

IL − 1 π −1

and

I L + I C > 1.

Proof of existence (omitted) follows directly from the construction of the equilibrium.

How do changes in the valuation of agreement affect the probability of bargaining breakdown? The probability of bargaining breakdown is given by q ((1 − π ) + π p) . Substituting in for q and p, and taking the first derivative with respect to I C gives   (π − 1)(I L + k − 1) ≤ 0. −x∗ 1 − π + IL − 1 In order to ensure that higher valuations of agreement lead to a lower probability of bargaining breakdown, this expression must be less or equal to 0. Rearranging and simplifying gives (π − 1)k > 0. IL − 1 Since π, I L < 1, this condition is always met.

5

Agenda setting and the single pension market

The previous chapter has focused on the informal communication flows between governments in shaping the common pension market. However, informal signaling and communication processes also influence the effectiveness of the European Commission as agenda setter. As the only institution in the EU that can formally propose legislation, the Commission can play a powerful role in securing member states’ commitment to EU covenants. However, failure to act as “efficient” agenda setter may cause the Commission to get defeated in the formal decision-making process. As a result, the Commission has an incentive to assemble strategic coalitions before a proposal is put to a formal vote. Yet, which factors constitute efficient agenda setting in the area of pension market integration? What was the Commission’s role in the bargaining fiasco of the early 1990s and the successful creation of the single pension market in 2003? This chapter analyzes the agenda setting strategy employed by the European Commission. We argue that negotiations had failed in the early 1990s because of the Commission’s inability to restrict the menu of policy options and frame pension market integration in a way that resonated with key domestic actors. The Commission’s lopsided representation of supply-side interests and emphasis on the financial aspects of demographic aging eroded its reputation as honest broker. Furthermore, rivalries between the two Directorates-General involved in the directive prevented efficient problem solving. Consequently, negotiations ended in gridlock, although all countries had initially signalled an interest in creating a single pension market. The impasse compelled the Commission to withdraw its directive proposal in 1994. In 2003, by contrast, the Commission’s role as educator and mediator contributed to the successful agreement on the IORP directive. By trimming an overcrowded agenda, accommodating demand-side interests, educating governments about pension reform options and limits in other member states, and setting aside internal rivalries, 76

Single pension market

77

the Commission was able to break a decade-long deadlock in the Council. Of course, pension market integration would not have happened without the support of domestic actors. Commission officials nonetheless performed a pivotal part as agenda setters. Intergovernmentalist theorists argue that member state collusion drives EU legislation, whereas supranationalists contend that the Commission constitutes the motor of the integration process. Our account comports with a growing stream of research that goes beyond this either-or dichotomy (Cram, 1997; Beach, 2005; Princen, 2009). EU officials do matter vis-a-vis governments in intergovernmental decision-making, but their ` influence varies according to the agenda setting strategy they employ. The following section reviews existing works of domestic politics and supranational agenda setting that will be used to derive hypotheses about the Commission’s ability to shape EU agreements. The empirical part compares the Commission’s strategic maneuvering between 1991 and 2003. Liberal intergovernmental approaches have powerfully demonstrated that government officials play a dominant role in major EU treaty negotiations (Moravcsik, 1998). Obviously, pension market integration would not have got off the ground without the support of domestic actors who were in charge of making formal decisions. Yet, pension market integration is not just a product of collusion between powerful governments. An exclusive focus on intergovernmental negotiations is blind to instances where the Commission plays an exceptionally strong role in defining the agenda and shaping EU agreements (Wallace, et al., 1999). Furthermore, a case can be made that intergovernmentalism fails to explain the breakdown of the negotiations over a pension fund directive in the early 1990s. At this time, demographic aging, stretched social security budgets, and rising unemployment had provided all EU states with a strong incentive to cooperate on a single pension market. Failure to strengthen pension portability at the European level imperilled the legitimacy of cutting back on social security pensions at home. Although member state preferences had converged, negotiations ended in deadlock in 1994. According to liberal intergovernmentalism, however, bargaining should never break down when there is preference convergence. Another reason why the adoption of the IORP pension directive is counterintuitive is that the diversity of welfare-finance regimes in

78

Single pension market

liberal and coordinated market economies militates against the development of common EU pension policies. The design of a “one size fits all” EU pension market is hampered by different perceptions of risk, which are inextricably linked to the way pensions are regulated. Beveridgean pension systems, such as Britain, are characterized by the existence of a mature pension fund culture. The majority of occupational pensions are provided as defined benefit schemes, wherein the employer guarantees his employees a certain sum at retirement. Since the employer assumes all investment risk, British politicians defend flexible investment rules on the grounds that employers’ duty to honour defined benefit plans requires them to pursue a high-risk/high-return strategy. Thus, British employers view risk as investment underperformance. Bismarckian pension systems such as Germany, however, have a strong insurance culture, with strict solvency rules protecting beneficiaries from volatility. In those countries, exposure to investment risk is seen as the problem, not investment underperformance. EU-wide investment regulations benefit either nations applying a high-risk/high-return strategy, or countries using a low-risk/low-return strategy, but not both. Such national differences often translate into acrimonious Council negotiations. To avoid the embarrassing scenario of getting defeated in the formal decision-making process, the Commission has an incentive to assemble strategic coalitions (Cram, 1997) before a proposal is put to a formal vote. Yet, severe distributive conflict over the choice of EU institutions provides member state representatives in the Council with incentives to represent all of their objections to integration as non-negotiable. This strategy conceals underlying zones of agreement (Tallberg, 2006: 25). What tools, if any, does the EU Commission possess to break such impasse? As the only institution in the EU that can formally propose legislation, the Commission can play a powerful role in securing member states’ commitment to an agreement (Pollack, 2003: 84). A crucial element of successful agenda control is issue framing. If the Commission can redefine issues in a way that appeals to relevant participants in the policy-making process, the line between proponents and opponents of a proposal may be drawn differently (Princen, 2009). The key to successful framing is to highlight some dimensions of an issue and downplay others, even if all of them are relevant in principle (Jones and Baumgartner, 2005). Another mechanism by which the Commission can exert influence is through the dissemination of

Single pension market

79

technical knowledge. Commission officials are particularly skilled at “inventing institutional options” to problems that previously had no supranational dimension (Moravcsik, 1999). A single pension market requires not just the removal of existing barriers but the design of wholly new institutions, such as cross-national pension funds and a common supervisory authority. Although the member states possess the technical knowledge and experience to regulate the movement of workplace pensions within their borders, they may be less familiar with the variety of problems that can arise when benefits are moved across borders. Thus, coordinating the movement of workplace pensions in the European Union constitutes a novel issue, demanding solutions to previously non-existent problems that are highly complex. Novel integration efforts also mobilize previously unrepresented constituencies, such as multinational corporations and European pension and insurance associations that will try to influence the design of new institutions. Princen (2009) refers to this process as “conflict expansion” – the mobilization of different groups to build momentum for policy change. It is likely that the absence of efficient supranational agenda control will leave national incumbents unaware of institutional options they would otherwise support. As agenda setter, the Commission can powerfully influence the menu of alternatives the member states consider in the first place. The capability to trim down a crowded agenda can provide a “structure-induced equilibrium” in an otherwise unstable setting of rational member states trying to maximize their utilities (Shepsle, 1986). Agenda control allows the Commission to broker compromises between the Council of Ministers and the European Parliament (Tsebelis and Garrett, 2001; Hooghe, 2005). Getting the Council presidency to support a directive proposal can also expedite the bargaining process. Since governments holding the Council presidency have routinely used procedural resources to steer negotiations toward solutions they most prefer (Tallberg, 2004; Thomson, 2008), the presidency constitutes a crucial ally the Commission needs to win over. However, since the presidency does not possess the kind of formal agenda setting powers the Commission has, this chapter focuses on the latter. Yet, precisely what role the Commission plays in agenda setting is in dispute. Kingdon (1995) has famously defined agenda control as an “idea whose time has come,” but this notion does not tell us how political actors choose from a given set of agenda setting tools.

80

Single pension market

Powerful agenda setting does not only manifest itself in key political decisions, but also in the ability to prevent contests from arising in the first place (Bachrach and Baratz, 1962: 949). When the agenda is overcrowded, negotiators tend to be overextended with the number of issues they need to sort out. This comes at the expense of forming a compromise around particular items. However, effectual mediation will only work if member states view the Commission as unbiased. If the Commission’s neutrality is in doubt, the Commission will not be able to act as honest broker. The Commission’s ability in effectively employing such strategies depends on the willingness and capacity of the different DGs to work cooperatively with one another. Yet, the Commission can hardly be characterized as a unitary actor. Institutional fragmentation often hinders problem solving, and “local fiefdoms” may jealously guard their spheres of influence (Hooghe, 2005: 39). Intra-Commission dynamics may differ, however, depending on member state positions in the Council. A divided Council and discord between DGs can translate into deadlock, while a unified Council may induce divided DGs to cooperate in preparing legislative proposals (Hartlapp, 2008). The following sections probe how these theoretical considerations apply to European pension market negotiations between 1991 and 2003. Comparing the Commission’s agenda setting strategy in both instances, we will examine the factors that led from bargaining breakdown to successful agreement.

5.1 Inefficient agenda setting in the early 1990s British Commissioner Sir Leon Brittan was the first to present a proposal on the creation of pan-European pension funds in 1991. British occupational pension institutions were considered as a legitimate blueprint for an EU-wide pension market. In light of the excellent reputation of British occupational pensions, Brittan assumed that the adoption of Anglo-Saxon-style pension regulations would be a matter of course. Strong stock market performance in the 1980s had contributed to great wealth accumulation in British pension funds, providing many beneficiaries with a comfortable retirement income. The high rates of return prompted former Labour minister Frank Field to describe British occupational pension funds as the envy of the world. This attitude was shared by Brittan’s co-authors of the pension

5.1 Early 1990s

81

directive proposal, who considered British welfare finance institutions as best practices. Convinced that the member states would gladly adopt British-style investment and social regulations, a co-author of the pension directive proposal remarked: “Our framework of law [. . . ] has ensured the growth of private occupational pensions in this country, which has served scheme members extremely well. Therefore we believe that the directive should be warmly welcomed” (Baroness Turner of Camden, Hansard, HL Deb. (14 Dec. 1992)). Although the speaker acknowledged flaws in the occupational pension system, she did not doubt the superiority of British welfare finance institutions as such: “Trust law has often been blamed for problems that have arisen in the Maxwell affair; but there is little wrong with the principles upon which trust law is based” (ibid.). These statements indicate that British pension regulations were considered as a reasonable blueprint for an EU-wide pension directive because British pension funds tended to outperform funds in Bismarckian pension systems. Another reason why British officials supported the most liberal version of EU pension regulations was the prospect of channeling rents to the financial service industry, an important constituency. The British pension industry already enjoyed a comparative advantage in assetliability management vis-a-vis their competitors in Bismarckian pen` sion cultures. As a result, a largely unrestricted single pension market would have allowed British firms to exploit economies of scale by tapping previously closed European markets. Although the Bismarckian member states supported pension market integration in principle, they had misgivings about the Commission’s financial market led approach, which ignored the social dimension of pension regulations (Esposito and Mum, 2004: 9–11). While countries with a Beveridgean pension fund culture view investment underperformance as risky, Bismarckian insurance cultures consider exposure to investment risk as the problem. Accordingly, Bismarckian nations suggested the creation of an EU-wide pension protection fund to guarantee employees’ pensions in case of employer bankruptcy. They also wanted to force the EU pension industry to offer biometric risk coverage, which denotes relatively expensive insurance against the risks of longevity, disability, and survivor dependence. Yet, such proposals were rejected by Beveridgean member states who view the provision of biometric risk coverage as too costly for the industry.

82

Single pension market

While Commissioner Brittan was responsive to the strongest advocates of a single pension market, he ignored proponents of a more cautious approach. The European Trade Union Confederation (ETUC) expressed disappointment at having been snubbed in the consultation process (Esposito and Mum, 2004: 18). The European association of public enterprises and employers’ organizations, the European Centre of Employers and Enterprises Providing Public Services (CEEP), criticized the “low rung in the EU hierarchy” they occupied during Brittan’s tenure, and the Commissioner’s “resolutely free-market doctrine” (European Association of Public Enterprises and Employers, 2007: 19). The most ardent supporters of an unrestricted pension market included the European Federation for Retirement Provision (EFRP), the European investment fund industry, F´ed´eration Europ´eenne des Fords et Soci´et´es d’Investissement (FEFSI), and the Union of Industrial and Employers Confederations of Europe (UNICE). These organizations strongly opposed all benefit-related requirements, such as the introduction of compulsory biometric risk coverage, which the Bismarckian nations had sought. The EFRP in particular was afraid that including compulsory biometric risk coverage meant losing EU-wide business to insurance companies, who would be able to decide whether they want to operate such business in a separate legal entity or alternatively ring-fence the pension assets and have a selective set of rules applied to those ring-fenced assets. Leon Brittan turned out to be responsive to these concerns by keeping the directive free of any benefit-related requirements. In a letter thanking the EFRP for its assistance, Brittan noted that the EFRP had become the principal source for consultation on pension matters (EFRP, 2001: 47). A crucial problem arising from Brittan’s one-sided commitment to EFRP preferences was that governments with an insurance culture did not regard the Commission as honest broker in this game. While EFRP chairman Alan Broxson repeatedly highlighted the pan-European profile of the federation in his speeches, Bismarckian member states strongly doubted that he represented the European pension industry as a whole. As he himself noted, When I started to visit the Commission some of the key staff were Italian, Danish, French, and German. In fact, there was only one Briton. So mostly I was dealing with people from many different pension backgrounds, and one could not expect many of these individuals to understand the real issues affecting pensions in a pan-European sense. (EFRP, 2001: 67)

5.1 Early 1990s

83

German officials took offense at the suggestion that their rejection of Brittan’s vision of a single pension market constituted a “lack of understanding.” Far from acting as impartial mediator, Commission officials were perceived as a mouthpiece of the British pension industry. Other member states with an insurance culture, particularly Spain, Italy, and France, also doubted the Commission’s ability and willingness to act as honest broker (personal interviews with officials at the German Ministry of Labor, Bonn, July 2006). Furthermore, the Commission’s decision to ignore the European insurance lobby, the Comit´e europ´een des assurances (CEA) in the consultation process did not bode well for subsequent negotiations. In Bismarckian pension cultures, the insurance industry plays a more prominent role than pension funds and therefore has more influence in pension policy developments. According to the EFRP, the CEA “had reacted violently to the original [1991 pension directive] draft” (EFRP, 2001: 46). The insurance industry argued that retirement benefits should be subject to strict solvency rules. Prior consultation of the insurers might have revealed some room for compromise and perhaps provoked a less “violent” reaction. A greater willingness to discuss the social aspects of pension market integration might have swayed Bismarckian nations to consider the directive in a more positive light. The discord between proponents of a financial market led approach and supporters of a “social” pension directive is furthermore reflected in intra-Commission rivalries. Brittan’s pension fund directive was drafted in DG Internal Market and Services (MARKET), which emphasized the financial aspects of a single pension market. DG Employment and Social Affairs (EMPL), however, took issue with this version and presented a draft that highlighted the social protection function of workplace pensions (Hartlapp, 2008). These competitive dynamics between key DGs are another reason the Commission refused to consider alternative approaches to pension market integration. For officials representing Bismarckian insurance cultures, the British vision for a single pension market was unacceptable as it ignored the social dimension of occupational pension provision. Brittan’s refusal to consider alternative approaches to pension market regulation had created an atmosphere of distrust (personal interviews, with officials at the German Ministry of Labor, Bonn, July 2006). Without consulting other governments, the Commission was unable to decide which policy issues to consider in the short term, which ones in the long term, and

84

Single pension market

which ones to remove from the agenda altogether. Failure to trim down a crowded agenda induced each member state to hold out for a better deal, and negotiations resembled a war of attrition.

5.2 Efficient agenda setting in the early 2000s After the Commission had withdrawn its directive proposal in 1994, the topic reappeared on the agenda in the context of European Economic and Monetary Union (EMU). As Chapter 3 has shown, Maastricht and the run-up to EMU prompted many member states to introduce reforms that strengthened second-pillar pensions. The creation of the European Central Bank deprived member states of their instrument to increase the money supply to offset government debt, of which unfunded pension liabilities constitute the largest part. In aging societies, such future obligations put mounting strain on fiscal outlays. In this environment, many member states took measures to increase the retirement age, reduce pension benefits, or introduced a funded component to the traditional PAYG pension system. Moreover, by the end of 1994, the single European insurance market was complete with the implementation of the second generation of life and non-life insurance directives (S. Schmidt, 2002). This implies that actors with a continental regulation background and no exposure to liberal regulations had gained significant experience with a more liberal regulatory approach by 2003. The implementation of these reforms, however, should not be taken to indicate that compromise on a European pension directive was a foregone conclusion. A top spot on the agenda neither guarantees legislative passage nor the implementation of EU laws according to original intent (Peters, 2001). Although economic change in the context of EMU may have altered the costliness of creating a single pension market, both Bismarckian and Beveridgean countries still wanted to design the pension directive in accordance with their own social and financial regulations. Moreover, member states had several options in responding to EMU pressures. Many Bismarckian states had already taken considerable electoral risks by reforming their national pension systems between 1999 and 2002. They could have simply left it at that and allowed subsequent EU pension directives to fail, just as they had in 1994. Failure of EU directives is much more embarrassing for the Commission than for national governments. Thus, neither side was

5.2 Early 2000s

85

keen on making concessions at the EU level just to return home with an agreement. Furthermore, opinion polls show that opposing the single pension market would not have hurt politicians. A Eurobarometer (2002: 51) survey found that almost 60 percent of Europeans prefer social policies to be decided at the national, not European, level. Liberalizing investment rules within the realm of a European directive was perilous for officials in Bismarckian countries because exposing pension beneficiaries to greater investment risk tends to be politically toxic. While 57 percent of Europeans think that pensions should be provided mainly by PAYG pension schemes, only 7 percent support the dominance of solutions based on private insurance (Janky and Gal, 2007: 5). Thus, staunch support for the IORP pension directive might have hurt political incumbents. No deal on pension market integration was still the favored outcome for many re-election-minded incumbents, although the issue had found its way back on the Commission’s agenda (personal interviews with officials at the German Ministry of Labor, Bonn, July 2006). During the run-up to the IORP directive, the Commission was able to broker a compromise on the most contested issues: taxation, investment rules, and biometric risk coverage. The instruments the Commission employed between 2002 and 2003 – but had failed to apply during previous negotiations – include: broad consultation of interest groups, holding long educational sessions in the Council, elimination of the most controversial items from the agenda (taxation, certain pension vehicles), application of an institutional solution to biometric risk coverage (home country control), and building of a strategic coalition regarding investment regulations and scope of the directive. The implementation of this strategy was possible because intra-Commission rivalries gave way to cooperation between key DGs. Let us consider each in turn. In 2002, the Commission carried out a consultation effort that differed markedly from Leon Brittan’s unilateral leanings. All major European labor, business, pension, and insurance associations were invited to submit their viewpoints before a directive proposal was put to a formal vote. Through these deliberations, the Commission learned what kind of directive proposals were politically feasible and which ones would so embarrass national governments as to be unproductive. This information was not systematically sought in 1991. Although

86

Single pension market

Commissioner Brittan had set up a network of experts studying the financial and social impact of the directive, he did so too late – one year after the formal directive proposal had already hit a brick wall of opposition from almost all member states (Esposito and Mum, 2004: 8). This consultation process allowed the Commission to overcome information asymmetries. The design of a supranational pension regulator requires not only a high level of legal and financial education. Decision-makers also need to understand the distributional consequences of supranational choices for each pension regime in Europe. While finance and social affairs ministries in each country collect such information, government representatives are more likely to have a comprehensive understanding of their own country’s preferences toward a common regulatory regime, but only limited knowledge of other systems’ capacity and willingness to adjust. Pension market integration is a novel issue that requires not just the pooling of existing resources but the creation of wholly new regulatory institutions. In this context, the Commission performs the essential role of educator and mediator by providing a forum for the exchange of ideas. It can decisively influence agreements by proposing institutional options, such as mutual recognition, that national governments might have otherwise overlooked. It is likely that learning about pension reform options and limits may also have taken place within the open method of coordination (OMC) and thus under a more intergovernmental mode (Nanz and de la Porte, 2004; Eckardt, 2005). However, the OMC fosters voluntary cooperation, while the IORP pension directive uses binding instruments to ensure compliance. Governments that fail to enforce the rules recorded in the IORP directive will face sanctions by the European Court of Justice. How well the Commission had fulfilled its educational role was evident in the knowledge government officials displayed regarding the cross-national distribution of preferences over pension market integration. Staff members in the German ministry of Labor knew exactly what reservations each of the fifteen member states harbored against the creation of a single pension market. They were also keenly aware of translation issues that tend to arise in the Council and can lead to legal adversarialism (Kelemen, 2009). Although it would be unrealistic to attribute their understanding of regulatory preferences entirely to the Commission’s initiatives, the officials interviewed for this project acknowledged that the educational sessions held in the Council helped them recognize which regulatory areas were considered legitimate for

5.3 Biometric risk coverage

87

harmonization efforts, and which spheres represented sacred cows for certain governments (personal interviews with officials at the German Ministry of Labor, Bonn, July 2006). Although it would be unrealistic to attribute their understanding of regulatory preferences entirely to the Commission’s initiatives, the officials interviewed for this project acknowledged that the educational sessions held in the Council helped them recognize which regulatory areas were considered legitimate for harmonization efforts, and which spheres represented sacred cows for certain governments (personal interviews with officials at the German Ministry of Labor, Bonn, July 2006). An important consequence of these educational sessions was that the social aspects of a single pension market were no longer ignored. As a result, intra-Commission dynamics also changed. While previous interactions between the key DGs were characterized by competition, DG Economic and Financial Affairs (ECFIN) and EMPL now presented a joint communication on “supporting national strategies for safe and sustainable pensions through an integrated approach” (COM (2001) 362). This document is considered a compromise between different factions within the Commission (Hartlapp, 2008). With cooperating DGs, the Commission was subsequently better able to strike a compromise between supply- and demand-side interests. The following paragraphs provide an overview of the most contested issue areas in pension market integration and analyze the Commission’s activism in restricting the agenda.

5.3 Taxation, investment rules, and biometric risk coverage Without a common taxation framework that prevents mobile employees from being double-taxed – or from paying no taxes on retirement income at all – the movement of pension rights across borders will remain a chimera. Taxation of pension assets can kick in at three points: when contributions are made, investment income is paid, or pensions are disbursed. The international standard is the exemptexempt-tax (EET) approach: contributions exempt, investment income exempt, benefits taxed. The EU Commission initially recommended this standard for EU-wide adoption. However, major opposition to tax harmonization came from countries that do not tax pension funds on the basis of EET. Denmark, Italy, and Sweden have the

88

Single pension market

contributions exempt, investment income taxed, benefits taxed (ETT) system and Germany and Luxembourg operate a contributions taxed, investment income exempt, benefits exempt (TEE) system. While the immediate elimination of taxes on pension fund contributions seems easy to implement, re-election-seeking incumbents fear the transitional costs associated with initial revenue losses involved in revamping the tax structure. The second change the EU Commission proposed was the EU-wide adoption of the prudent person rule that prohibits pension funds from making risky investments. Pension funds in most Beveridgean nations already invest pension assets according to the prudent person rule, thus no adjustment costs accrue in this area. Yet, member states with Bismarckian pension systems found the creation of a single pension market more costly than their Beveridgean counterparts. Pension funds in these countries are subject to heavy regulatory restrictions and investment limits. Thus, while Beveridgean nations lobbied for the EU-wide adoption of the prudent person rule, the Bismarckian nations wanted the opposite: casting tough investment restrictions in stone. Another obstacle to cross-border pension movements included divergent regulations of biometric risk coverage. While German pension plans routinely offer biometric risk coverage, British plan sponsors offer this coverage to a very limited extent. The CEA supported a directive that made biometric risk coverage compulsory. In addition, the CEA requested supervisory regulations that required full funding of technical provisions at all times to protect beneficiaries from unsound investments (Commission of the European Communities, 2005: 35). ETUC echoed these requests (ETUC, 2003: 60). The supporters of mandatory biometric risk coverage had an influential advocate in the rapporteur Othmar Karas, who had worked in the Austrian insurance industry before becoming a member of the European Parliament. Given the informational advantage of EU “insiders,” the influence of the rapporteur on EU agreements can be substantial (Meyer, 1999). Evidence indicates that the Commission broke the deadlock by proposing to apply the principle of home country control to biometric risk coverage (i.e., member states decide the extent to which pension institutions will have to provide biometric risk coverage). This solution allows the Bismarckian insurance states to retain their ideal point on an issue area they cared strongly about. The Beveridgean member states, on the other hand, appreciated the home

5.3 Biometric risk coverage

89

country control principle because it lets employers unwilling to provide such expensive risk coverage off the hook. This compromise represents a second-best solution for both Bismarckian and Beveridgean member states: between the extreme points of forcing all European pension institutions to provide biometric risk coverage and leaving this choice to individual pension providers. Mutual recognition was not considered in the early 1990s; only the two extreme viewpoints were discussed. It is likely that the 2003 compromise would not have emerged without the Commission’s consultation and mediation efforts. Given that the EU-wide coordination of biometric risk coverage constitutes a novel issue, Commission officials who spend most of their waking hours thinking about the technical implications of EU regulations were probably better suited to working out an agreement than a single actor designated by national governments. While the Commission found a creative institutional solution to the question of biometric risk coverage, it employed an altogether different tool with regard to taxation by eliminating this controversial subject matter from negotiations altogether. Taxation issues were met with more political opposition than other factors because tax harmonization represents a classic time-inconsistency problem. Taxation regulations cannot be changed overnight from a national to a European dimension, because political choices entailing high upfront costs and long maturation of benefits are politically risky. In addition, the adoption of a taxation directive would have required unanimity in the Council, thus requiring a much higher decision threshold. Agenda setting is more successful the lower the decision threshold. Consequently, the Commission decided to eliminate the tax issue from the deliberations on the single pension market (Haverland, 2007). While taxation constituted the toughest bone of contention, it was not the only issue that was “organized out.” National governments had a strong interest in protecting additional “sacred cows” from Brussels, such as German book reserve pensions, French 1408/71 pensions, and legal personalities in Italy and Spain. A strong commitment to a particular issue matters since governments will devote more scarce resources, negotiate with greater care, and stay longer at the negotiation table (Tallberg, 2008: 692). Thus, the challenge for the Commission is to get national governments to move away from the position of principled opposition to any EU agreement toward a more productive discussion of their strongest reservations.

90

Single pension market

5.4 Country variations Germany’s main hesitation to support pension market integration stems from the pervasiveness of on-balance sheet book reserve pensions. With 60 percent of all occupational pensions being book reserve pensions, it constitutes the most popular form of employer-sponsored pension in Germany. Firms prefer book reserve pensions to externally funded pension institutions because it provides them with a cheap source of finance as well as an effective staff retention device. The creation of a single European pension market posed a threat to this cheap capital as well as to the long-term nature of German labor relations. Thus, Germany fought hard for the exemption of book reserve from the scope of the directive. As a result, the 2003 directive text specifies that the directive shall not apply to companies using book reserve schemes. In return for exempting book reserve pensions from the directive, Germany decided to drop its resistance to the liberalization of investment regulations and no longer insisted on making the provision of biometric risk coverage mandatory for the European pension industry. Germany’s eventual acceptance of the directive sent a strong signal to other reluctant member states. As one Commission official recalls: “We finally understood that Germany’s biggest concern were the book reserve pensions. As soon as we got the Germans on board, it was much easier to persuade other hesitant member states.” France shrewdly negotiated its way out of the directive, declaring that France simply does not have occupational pensions. The secondpillar in France is essentially provided by Association pour le r´egime de retraite compl´ementaire des salari´es (ARRCO) for blue-collar workers and Association g´en´erale des institutions de retraite des cadres (AGIRC) for white-collar employees. Of these, the majority take the form of book reserves, as in Germany. Consequently, France shared many of Germany’s doubts regarding the IORP directive. An EU regulation from the 1970s was the solution in this case (Council Regulation (EEC) No. 1408/71 of 14 June 1971). This regulation excludes social security schemes from its scope. By declaring that all French occupational pension schemes fall under the 1408/71 Council Regulation, France managed to exclude all of its occupational pension schemes from the scope of the 2003 directive. However, the directive also specifies that no country must discriminate against foreign pension

5.4 Country variations

91

funds, indicating that France does not escape the specter of foreign competition. Britain, by contrast, stood to gain the most from the liberalization of financial regulations under the IORP directive. The occupational pension sector is mature and the pension fund industry competitive. The prudent person rule is already used to invest pension assets, relieving the country of any adjustment costs in this area. However, Britain successfully opposed German and French attempts to force the European pension industry to offer biometric risk coverage. Britain was initially opposed to exempting book reserve pensions from the scope of the directive, arguing that lack of oversight by EU institutions would give employers sponsoring such pensions an unfair advantage. However, given that the Bismarckian nations eventually accepted a liberalization of investment regulations, as well as the principle of mutual recognition regarding biometric risk coverage, Britain conceded. The Bismarckian states, on the other hand, reluctantly agreed to article 16(2), which specifies that member states may allow pension funds for a limited period of time to have “insufficient assets to cover technical provisions.” This stipulation clearly reflects the position of the British pension funds, trumping the security concerns of the Bismarckian insurance industry. Another important ally the Commission managed to win over includes the Council presidency. Each member state, on taking up the presidency, has tended to give emphasis to directives that most closely affect its own nationals. Given the underdeveloped secondpillar pension sector in these countries, neither the Spanish nor the Italian presidency was expected to push the issue of pension market integration. And yet, the impasse in the Council was broken under the Spanish presidency in 2002, while the directive itself was signed into law under the Italian presidency in 2003. What explains this counterintuitive result? For the Spanish and Italian governments, it was most important that the pension fund directive also applied to pension institutions that do not have legal personality. This is because in Italy and Spain pension funds do not have legal personality or capacity, but the plan members have a legal title to the pension fund assets. The pension fund assets are held in an account managed by a financial company for plan affiliates, where the account is legally separated from the balance sheet of the managing entity. These concerns were accommodated by article 2(1) of the 2003

92

Single pension market

directive text by specifically including pension institutions without legal personality. In the end, the Council presidency was swayed by the prestige associated with shepherding a directive through the legislative process, and the reputation for being effectual. At the same time, political incumbents did not incur audience costs because the most sensitive issues had been excluded from the purview of the directive. Thus, the Commission succeeded in persuading the presidency to make pension market integration a priority (personal interviews with six EU Commission officials, June 20–23, 2006). The compromise the European member states reached in 2003 reveals a mix of coercive and voluntary elements. While Bismarckian member states had to accept liberalization of investment regulations, the principle of home country control allowed them to retain their ideal point on biometric risk coverage. This permits them to protect at least their own nationals against excessive risk. For the Beveridgean member states, on the other hand, the principle of home country control means that their own employers escaped a stifling clause forcing them to provide expensive risk coverage. These institutional solutions – home country control for biometric risk coverage, elimination of certain pension vehicles and taxation from the agenda altogether – were an important step toward agreement in 2003. None of these options had been considered in the early 1990s.

5.5 Alternative explanations So far, this chapter has provided evidence to support the hypothesis that supranational agenda setting was a decisive precondition for agreement, while member state activism alone is insufficient to explain the 2003 pension directive. Yet, critics might challenge our argument, given that it is rarely clear or agreed upon what constitutes efficient agenda setting. To assess the plausibility of our theory, the following section will probe alternative explanations for the adoption of the IORP directive. One alternative explanation might credit Left party control of government with the adoption of the IORP pension directive. The largest and most influential member states in the European Union – Germany, France, and Britain – all had conservative government leaders in 1994. In 2003, by contrast, two of the three conservative

5.5 Alternative explanations

93

administrations had been replaced by Center–Left coalition governments. Does this mean that Left parties are more supportive of EU pension directives? This proposition contradicts scholarship that views Left parties as opposed to an expansion of occupational pensions. According to this perspective, Left parties disapprove of the inegalitarian properties of workplace pensions because such plans do not extend the kind of inviolable social rights that statutory legislation provides (Esping-Andersen, 1996: 328–329). Yet, one might claim that Left parties are known for moving to the center of the Left–Right continuum in order to maximize popular vote share. Tony Blair and Gerhard Schroder’s “new middle” party ¨ manifestos constitute the prime examples of this strategy. However, the past two decades of occupational pension policy development in Europe do not reveal any reform pattern suggesting that Left parties are more likely to support competition and economic freedoms in order to create a single pension market. If the German Social Democrats constituted a classic example of a Left party moving towards the center, they have abandoned this path in light of rising electoral pressures from the radical Left Linkspartei. Moreover, partisan theory cannot explain variation in pension policy development by Left parties within member states. Given that adjustment costs and political risks involved in transposing EU pension directives are distributed unevenly across Bismarckian and Beveridgean member states, it is difficult to discern a distinct Leftist or Conservative pension policy strategy. Therefore, partisan theory does not constitute a convincing alternative explanation. Another alternative hypothesis might attribute the legislative passage of the IORP directive to the impending EU enlargement, rather than the Commission’s agenda control. The prospect of adding ten Central and East European nations in 2004 may have increased pressure on the member states to come to an agreement before the new entrants could delay negotiations. Such pressure was obviously absent during the 1990s. The new member states might have been expected to oppose the IORP directive because of their continuous struggle with the transition to market economies and democratic politics. However, a glance at the institutional structure of the Central and East European pension systems does not support such an interpretation. Precisely because of the rapid pace of institutional reform following the fall of the Berlin wall, many Central and East European

94

Single pension market

countries implemented major pension reforms in the 1990s that resonated with the World Bank model of privatizing state pension systems. By 2002, six out of the ten accession candidates had either privatized their pension systems wholesale or introduced funded components.1 Since the IORP directive saddled mostly Bismarckian pension systems with adjustment costs, but not countries with a pension fund culture, there is no reason to believe that the Central and East European governments would have opposed or even delayed negotiations. Given the growth of private pension funds in these countries, it is not unreasonable to speculate that the domestic pension industry might have supported the directive. Consequently, there is no evidence to suggest that impending accession should have influenced the 2003 negotiations.

5.6 Conclusion Informal information flows and “behind the scenes” deliberations not only shaped negotiations in the Council, but also the effectiveness of the European Commission as agenda setter. This chapter has traced the successive steps of pension market integration back to a specific agenda setting strategy employed by the European Commission. Liberal intergovernmentalism – although a powerful theory – is limited in explaining pension market integration because it cannot account for bargaining breakdown when there is a convergence of national preferences. We showed that, without the Commission’s ability to restrict a crowded agenda, reframe policy issues, build coalitions, and apply certain institutional solutions to cross-border problems, the same gridlock that prevented member states to reach an agreement in 1994 would have torpedoed legislative passage in 2003 as well. Thus, supranational agenda control was just as decisive for pension market integration as the informal signaling dynamics between the member states. However, the results should not be understood as an automatic mechanism predicting that the Commission will always rely on educational, framing, and agenda restricting strategies. Nonetheless, the analysis has broader implications for democratic accountability in the EU. Given the distributional impact of pension policies on the life 1

Only Cyprus, Malta, the Czech Republic, and Slovenia had not introduced pension reform by 2002.

5.6 Conclusion

95

course risks of citizens, political incumbents may not escape electoral backlash, no matter whether pension reforms reflect domestic political preferences or a reluctant response to EU regulations. Thus, interest in policy instruments that can enhance the efficiency of supranational agenda setting is likely to continue. The following two chapters will examine the German and British positions on EU pension regulations. The choice of these two countries is advisable for at least three reasons. First, the purpose of indepth case study research is to overcome the limitations of the formal model, which cannot establish validity of the theoretical assumptions. Secondly, both countries represent the respective ideal types of Bismarckian insurance cultures and Beveridgean pension fund systems, as outlined in Chapter 2. Thirdly, we exploit the fact that the demands of other member states with Bismarckian insurance cultures coincided with the German position, while the majority of Beveridgean pension fund cultures supported the British viewpoint. Thus, Germany and Britain represent critical cases. We analyze three decades of pension policy developments in the two countries to determine how the interests and strategies of different political actors were informed by the prevalent national pension policy discourse. Our findings confirm the basic intuition of our formal model, which posits that a government’s credibility translates into bargaining power at the EU level.

6

The German position on EU pension policies

To eliminate obstacles to pension portability across borders, the European Commission put forward three key directive proposals between 1991 and 2005. Germany torpedoed the first pension directive proposal in 1991, offered contingent support for the second directive in 2003, and vociferously opposed the pension portability proposal of 2005. Why would German incumbents support some aspects of the single pension market, but not others? This chapter argues that the German position towards EU pension policies needs to be explained by a combination of historical institutionalism (HI) and domestic discourse analysis (DA). Each approach by itself is insufficient to account for the articulation of preferences between 1991 and 2007. HI explains why the Kohl, Schroder, and ¨ Merkel governments – despite all fundamental differences in pensionrelated values – protected employer-sponsored book reserve pensions, a cornerstone of Germany’s coordinated market economy, from the scope of EU directives. However, national trajectories of policy development are not destiny and do not predict variation in German preferences. We therefore need to include a theoretical framework that illustrates how interests were conceptualized and reframed during this period. Discourse analysis, which consists of a coordinative dimension (who said what to whom) and a communicative sphere (how political officials justify and legitimize change), supplies this crucial piece that is missing from HI explanations. While the status quo oriented policy stance promoted by the Kohl government succeeded in delegitimizing pro-reform voices, the Schroder administration imposed an eco¨ nomically efficient pension reform without much public support. The grand coalition, in turn, abandoned chancellor Merkel’s initial plan to expand second-tier pensions in light of rising pressure the Left party posed for the Social Democratic coalition partner. Shortcomings of DA accounts include after-the-fact justification and legitimization, or poor 96

German position

97

specification of the conditions under which ideas exert a causal influence on public policy developments. When combined, however, HI and DA supply each other’s shortcomings. Thus, to understand variation in the German position towards EU pension policies, a methodologically pluralist approach accounting for both structure and agency is necessary to get the full picture. This chapter is organized as follows. The second section probes how different theoretical frameworks fare in explaining variation in German preferences towards EU pension directives. The subsequent part provides an overview of the German corporate pension system. The following sections examine the linkages between EU harmonization efforts and the social policy discourse communicated by the Kohl, Schroder, and Merkel governments between 1991 and 2007. The final ¨ section concludes with a summary of the main findings and suggestions for future research. Since this chapter seeks to explain the policy positions of three governments, each consisting of a different set of coalition partners, partisan theory is an natural starting point. According to this school of thought, the Left–Right ideological position of governments explains preferences better than other conflict dimensions, such as national varieties of capitalism (Hix and Lord, 1997; Hooghe and Marks, 2001; Hix, et al., 2007). However, any political issue may pitch ideological preferences against the interests of important national groups (Callaghan and Hopner, 2005; Ringe, 2005). Therefore, a gov¨ ernment’s public stance towards EU policies cannot be automatically attributed to its ideological proclivities, but must be analyzed with regard to historical trajectories of path dependence and the concerns of key constituencies. Although MEPs are accountable to two masters – the political party in the European Parliament and their domestic constituency – EU pension politics tends to be focused more on interest-based compromise than on party-based competition. Because policies are made without the kind of political competition, partisanship, and party politics typical of the national political arena, EU-level decision-making has been characterized as “policy without politics” (V. Schmidt, 2005: 769). Careful examination of German preference articulation between 1991 and 2007 supports the theory that national interests, and particularly the Rhenish model of capitalism, explains the German position more than Left–Right ideological proclivities of the Kohl, Schroder, or Merkel governments. Even though all ¨

98

German position

administrations implemented fundamentally different pension policies at the domestic level, their position towards European pension directives was remarkably similar. Moreover, partisan theory cannot explain why a Social Democratic chancellor implemented far-reaching pension reforms and supported EU pension directives, while his Conservative counterparts did not. We argue that historical institutionalism and domestic discourse approaches fare better in explaining the German position towards EU pension policies. Historical institutionalism assumes that history unfolds through positive feedback mechanisms of reinforcement, which are difficult to change (Stinchcombe, 1968; Mahoney, 2000; Pierson, 2000a). A more refined version of HI departs from the determinism of path dependence and instead considers a variety of indeterminate adjustment processes, including path renewal, modification, or replacement (Streeck and Thelen, 2005). Historical institutionalist theories have much to say about the congruence between EU directives and national institutional frameworks. The “goodness of fit” between EU laws and domestic social policy traditions is an important predictor of a country’s preference for EU-mandated change. As the following sections will illustrate in greater detail, this approach helps us understand Germany’s intense interest in protecting employer-sponsored book reserve pensions from the purview of EU pension directives. Unlike funded American 401(k) plans or British trust funds, wherein the sponsoring firm is separate from the trustee, German firms administer their employees’ book reserve pensions themselves (Ahrend, 1996; Estevez-Abe, 2002). Ultimately, this money will have to be paid out as corporate pension. But, until workers have reached retirement age, employers can do what they like with the funds they keep.1 In the aftermath of World War II, book reserve pensions provided companies with cheap capital, thereby playing an important role in economic reconstruction efforts (Whiteside, 2006). Long waiting and vesting periods tie the most cherished employees to the firm for a long time, making book reserve pensions an effective staff 1

Because this practice could give rise to liquidity risks as future pension obligations might be inadequately reflected in company balance sheets, the EU Commission has repeatedly sought to make external funding compulsory under a European directive. However, there is no consensus in the literature whether book reserve pensions or externally funded pensions are superior in terms of pension security.

German position

99

retention device. EU efforts to create a single pension market threatened to deprive German firms of this convenient source of corporate finance and staff retention. Thus, the traditional role of book reserve pensions in sustaining Germany’s characteristically patient capital flows and long-term labor relations is causally linked to the government’s intense lobbying to exclude book reserve pensions from the scope of EU directives. The resilience of employer-sponsored book reserve pensions is counterintuitive to scholarship that has pronounced the erosion of Modell Deutschland. Strong liberalization tendencies in the areas of corporate governance (Hopner and Jackson, 2002; Beyer and Hopner, 2003), ¨ ¨ industrial relations (Hassel, 2004, 2007; Rehder, 2006), the financial system (Deeg and Lutz, 2000; Jackson, 2003; Deeg, 2005), and eco¨ nomic policy (Allen, 2004) have supposedly spelled the end of Rhenish capitalism. Yet, the removal of particular institutions does not automatically spell the death of complementarities – the presence of two or more institutions that enhance the functioning of each (Aoki, 2001; Hall and Soskice, 2001). Employer-sponsored book reserve pensions are a case in point. Although the existence of this pension vehicle was repeatedly challenged by the EU Commission, governments of different partisan complexion have fought tooth and nail to preserve this cornerstone of Germany’s coordinated market economy. This implies that national institutional frameworks are not destiny and therefore do not predict behavior. While the long-standing attractiveness of unfunded book reserve pensions explains much of Germany’s opposition to their elimination, historical institutionalism fails to account for variation in the German position towards a European single pension market. HI by itself cannot explain why Germany torpedoed the European Commission’s first proposal for a pension fund directive in 1991, offered contingent support for the IORP directive in 2003,2 and fiercely opposed the pension portability proposal of 2005. To explain variation in policy preferences, we need to add an analysis of how interests were conceptualized and justified between 1991 and 2007. Whether EU-mandated pension reforms are perceived as legitimate or not depends on how government leaders communicate the need to revamp existing practices and policies. An investigation of the public discourse provides this missing link. 2

“Institutions for Occupational Retirement Provision.”

100

German position

Discourse analysis, which includes a coordinative and a communicative dimension, highlights both the substantive content of ideas and the interactive processes by which ideas are conveyed (Blyth, 2003; Schmidt and Radaelli, 2004; V. Schmidt, 2006c; Abdelal, 2007). The coordinative sphere comprises the key policy actors who talk to each other in order to reach agreement on reforms. Particularly in the Rhenish capitalism model, it is important to trace the interaction between employers, unions, and political parties to understand how cognitive schemes of legitimacy militate in favor or against policy reform. The communicative dimension of discourse denotes the processes by which government officials, opposition party leaders, organized interests, the media, and social movements incorporate new ideas into political programs, and communicate them to the general public (V. Schmidt, 2006c). “Ideational leadership” is said to occur when policy-seeking officials manage to expose the downsides of the status quo and make consistent efforts to redefine the values underlying well-established policies and institutions (Stiller, 2010). In contrast to historical institutionalism, discourse analysis avoids historical determinism by showing how political actors may or may not alter historical trajectories by acting differently within them as a result of new ideas conveyed by new legitimating discourses (V. Schmidt, 2006b). However, ideas-based approaches also have drawbacks. They have sometimes underestimated the mobilization to which cognitive templates are subject, or failed to specify why politicians favor a certain outcome and how they manage to get their preferences implemented. The usefulness of ideas-based approaches, then, depends on a clear specification of the conditions under which discourse exerts a causal influence on the dependent variable. Existing streams of research have fruitfully shown that the causal influence of ideas on policy outcomes is best demonstrated when combined with rational choice or historical institutional accounts of policy development, e.g. Stiller’s research on ideational leadership in German social policy reforms (Stiller, 2010), Zohlnhofer’s examination of German economic policy in the Kohl era ¨ (Zohlnhofer, 2003), or Cox’s analysis of why welfare reforms hap¨ pened in Denmark and the Netherlands, but not in Germany (Cox, 2001). We adopt the same approach to the study of workplace pension policy, a so far neglected area of comparative public policy. To understand the sources of German preferences on EU pension policies,

6.1 Workplace pension system

101

Pensionskasse (21%) support funds (8%) direct insurance (12%)

book reserve pensions (59%)

Figure 6.1 German workplace pension schemes

the next section provides a brief outline of the German workplace pension system.

6.1 The German workplace pension system It is well known that employer-sponsored old age insurance has a longer tradition in Germany than the social security pensions Bismarck introduced in the late nineteenth century. Early firm-sponsored insurance programs reflect employers’ sense of responsibility for their workforce at the time, commensurate with the paternalist ethos commonly ascribed to conservative welfare states (Esping-Andersen, 1990). The perceived role of corporate pensions and its diffusion changed in the twentieth century, when employers realized that offering corporate pensions was in their own self-interest, and not just a charitable or humanitarian responsibility. Occupational pensions can take on a variety of institutional forms. Some pension claims are externally managed by insurance companies or trust funds, others only exist on the books and are not funded. Some corporate pensions are exclusively controlled by the employer; others are jointly managed by employer and labor unions. The criterion that distinguishes occupational pensions from state-sponsored or private pension schemes is that they are provided by individual employers or a whole industry. The most popular form of occupational pension provision in Germany is the book reserve pension, which is hidden within the company’s own corporate assets.3 Figure 6.1 shows that 3

Book reserve pensions are also found in France, Greece, Ireland, Latvia, Luxembourg, the Netherlands, Norway, Poland, Portugal and Britain. However, in some countries workplace pensions are not a source of cheap capital for companies. For example, in the Netherlands and in Denmark regulatory rules require that workplace pension assets are held outside the firm (Anderson, 2007).

102

German position

book reserves account for almost 60 percent of German pension liabilities out of a total of 331 billion euros, while funded pension vehicles play only a minor role in corporate pension provision. German employers favor book reserve pensions over alternative schemes for two reasons: first, book reserve pensions provide firms with access to capital independent of current profitability. This allows companies to finance their core businesses activities cheaply. Instead of having pension funds invested in a broad range of assets – stocks, bonds, and real estate – they are all embedded within one company. The pension entitlements are retained and used to finance business activities. Book reserve pensions were particularly prominent after 1945, when workers’ contributions to these schemes met urgent postwar requirements for internal investment (Whiteside, 2006: 46). To protect employees’ pension entitlements in the event of corporate bankruptcies, German firms have to be a member of the Pensionssicherungsverein (PSV), a mutually owned association that reinsures employers’ pension obligations against the risk of insolvency. Another favorable feature of book reserve pensions is their role as effective staff retention device. A productive stream of research has shown that protective employment security regulations increase the long-term horizon of firms and strengthen the incentives of employers to invest in the skills of their staff members (Hall and Soskice, 2001; Iversen and Soskice, 2001; Mares, 2003; Thelen, 2004). To protect their investments, employers need institutions that reduce the risk of competitors poaching their skilled workers. A corporate pension that belongs to employees only after long waiting and vesting periods constitutes an effective commitment device that encourages employees to stay with their firm for a long time. Occupational pension plans are quite unevenly distributed among the population. The majority of employees with access to employersponsored plans works in large corporations concentrated in the unionized sector, whereas occupational pension coverage is rare in smaller firms or within the lower end of the service sector. In contrast to social security pensions, occupational plans do not extend the kind of inviolable social rights that statutory legislation provides and may be abused as weapon of managerial power (Esping-Andersen, 1996: 328– 329). Furthermore, occupational pensions magnify existing inequalities because employees who have access to a corporate pension already receive a higher PAYG pension (Schoden, 2003). For this reason,

6.2 Negotiation failure in the early 1990s

103

occupational pensions are often a target of labor movements who see their extension as undesirable. In contrast to book reserves, exter¨ nally funded pensions (Pensionskasse, Unterstutzungskasse, and pension funds) play only a minor role in Germany, because only a small part of the population is covered by such plans.

6.2 Germany and EU negotiation failure in the early 1990s Given this institutional setting, how might we expect German political and economic actors to react to EU pension policies? To create a European-wide framework for pension portability, the member states need to find answers to a complex set of problems: whether the social laws of the sending or the receiving country should apply; what kind of pension institutions should be subject to European supervisory control; and whether the sending or receiving employer should bear the administrative costs of calculating the amount of transferred assets. Naturally, member states will try to keep adjustment costs as low as possible. This is because corporate pension institutions evolved along member states’ distinct social policy trajectories and locked in particular types of labor relations, capital flows, and benefit design. EU-mandated regulatory changes may have favorable distributional consequences in some nations, but not in others. Thus, a “one size fits all” framework governing the movement of pensions across borders requires adjustments in sensitive policy areas – social and financial regulations – that the majority of member states fiercely guards against intrusion from Brussels. The first attempt to create a single pension market occurred in 1985, when it became clear that pension funds had been forgotten in Jacques Delors’ White Paper on the single market (Delors, 1985). Yet it was not until 1991 that the Commission began to take action in this field.4 To create pan-European pension funds, Commissioner Brittan proposed reforms in three critical areas: first, investment managers should be able to operate in all EU countries without having to adhere to twelve different investment regulations. The way to circumvent national investment rules was to adopt the prudent person rule. Secondly, EU citizens 4

Proposal for a Council Directive relating to the freedom of management and investment of funds held by institutions for retirement provision, COM (1991) 301 final, OJ C 312, 3 December 1991.

104

German position

should be free to join any cross-border pension scheme of their choice. Thirdly, a single administrative body was going to be in charge of supervising the operation of the affected pension schemes. Germany soon joined the vocal drumbeat of member states opposed to Brittan’s directive proposal. Suspecting that a single European pension market would only serve the interests of countries with a mature pension fund culture (i.e. Britain, the Netherlands, and Ireland), the Bismarckian insurance states (all other member states) feared that they were going to bear the brunt of adjustment costs. Germany rejected the Commission proposal for three reasons. The first is linked to Germany’s unwillingness to adopt the prudent person rule, the equivalent of liberalization of investment regulations. The existence of stringent investment limits has been traditionally justified with the need to protect investors against market risk. Given the divergent investment regulations in the member states, no common definition of “prudent” could be reached at the EU-level. Thus, along with like-minded Bismarckian states, Germany tried to turn the 1991 directive proposal into its antithesis: instead of agreeing to liberalize investment rules across Europe, Germany demanded that quantitative investment limitations be cast in stone.5 While the struggle over investment regulations was settled in 2003 – all member states finally agreed on a definition of the prudent person rule – contestation over crossborder membership and the scope of EU pension directives continues unabated. The second reason for Germany’s opposition has to do with the EU Commission’s concept of cross-border membership. The proposal would have allowed EU citizens to join occupational pension schemes that operated on the basis of PAYG. Since PAYG systems are based on the assumption that the previous generation pays for their parents, they are inherently ill-suited for accommodating late entrants. Permitting people to enjoy the benefits of the system without having contributed to it would have dealt a blow to common perceptions of generational justice. Anticipating lengthy delay, the European pension fund lobby EFRP had worked hard, via the EC Committee, to separate two freedoms – cross-border investment and cross-border pension management – from cross-border membership. However, the attempt to make cross-border membership a separate issue was rejected by the EU Commission (EFRP, 2001: 45). 5

Personal interviews with six EU Commission officials, June 20–23, 2006.

6.2 Negotiation failure in the early 1990s

105

The third reason why the EU plans were unacceptable for Germany concerns the costs of providing corporate pension coverage, in particular book reserve pensions. Employers feared that permission to transfer pension fund assets across borders would require the capitalization of pensions that have hitherto only existed on the books. The federation of German employers’ associations, Bundesvereinigung der Deutschen Arbeitgeberverbande (BDA), has argued that if book reserve pensions ¨ are no longer available to finance current business activities employers lose both an important source of capital as well as a staff retention device (Bundesvereinigung der Deutschen Arbeitgeberverbande, 2006; ¨ aba, 2006). In addition, improved pension transferability rights may increase job turnover, which may impose greater liquidity constraints on the firm sponsoring occupational pension coverage. Greater liquidity requirements, in turn, will increase both the costs of sponsoring occupational pensions and decrease the return on invested pension assets. “Such losses would be borne by everyone, including loyal workers, which is not fair. Why should workers with long job tenure be required to pay a loyalty tax?” (Interview with aba chairman Klaus Stiefermann, 29 June 2006). Echoing this viewpoint, the German government questioned both the desirability of a single pension market as well as the appropriateness of promoting labor mobility at the expense of occupational pension coverage: European-wide regulations to encourage labor mobility must not lead to a deterioration of national structures governing occupational pension provision. [. . . ] The costs associated with pension portability may discourage employers to offer occupational pensions. However, the build-up and expansion of occupational pension coverage is no less important than the goal of increasing worker mobility.6

This means that the German government considered the risks associated with implementing EU pension policies to be greater than the potential repercussions of failing to create a single pension market. Negotiators at both domestic and EU levels now agree that the 1991 directive proposal was too ambitious in its attempt to tackle all obstacles to cross-border pension movements at once. “The member states simply had too much on their plate.”7 6 7

German Bundestag, Ausschussdrucksache 16(11)1999, Entschliessungsantrag der Fraktionen CDU/CSU and SPD, pp. 1–2. Personal interview with EU official, June 21, 2006.

106

German position

Consistent with earlier works analyzing the nature of German labor relations (Thelen, 2000), this episode demonstrates that the major players in the corporate pension game – employers, the government, and the occupational pension association – preferred the known stability of the status quo to unknown measures that might have made labor markets more dynamic. In other words, the 1991 directive proposal was a poor fit with Germany’s patient capital flows and long-term labor relations. However, institutional misfit explains only partly why Germany rejected the proposal. How the pension-related values of the government were communicated and legitimized is every bit as important in understanding Germany’s preference articulation at the EU level. The following section will analyze how the 1991 EU pension fund proposal was perceived by political actors in Germany.

6.3 Domestic discourse – the Kohl era Where the majority of the population depends on social security pensions as the sole income during the retirement phase, few topics are as emotionally loaded as the long-term sustainability of pensions. As former CDU deputy chairman Kurt Biedenkopf put it: “If a hard to comprehend system, which is critical for retirees’ financial security, suddenly causes tongues to wag, fear and emotions run high. Both provide poor guidance” (Biedenkopf, 1981). Thus, successful pension reforms are either reflected in a pension system capable of keeping citizens’ angst to a minimum, or in the extent to which political incumbents are able to exploit fears of competing pension systems. The Kohl government managed to keep pension reform off the table because it successfully played on public fears of alternative pension security concepts. In 1991, when the EU Commission first presented a proposal to create cross-border pension funds, the ideas communicated by the Kohl government strongly militated against expanding second-pillar pensions. Public discourse was famously dominated by the notorious slogan of former social affairs minister Norbert Blum: “One thing ¨ is certain: the [social security] pensions are safe.” This was Blum’s ¨ response to Social Democratic accusations of fiscal recklessness regarding his pension policies. In 1984, during the first hearing on budgetary procedures in the Bundestag, he insisted that social security pensions are guaranteed by the state and urged citizens not worry about Social

6.3 Kohl era

107

Democratic Cassandra prophecies. Headlines in the Westdeutsche Allgemeine Zeitung read: “Blum: pensions are safe – opposition: state ¨ coffers are empty.”8 The Labor minister regurgitated his motto during the second reading in the Bundestag, reacting to allegations that the state had to borrow in order to finance social security pensions.9 In the same vein, the coordinative dimension of discourse was not favorable to the expansion of workplace pensions, either. Although the Social Democrats accused the Conservative–Liberal government of jeopardizing first tier pensions by what they deemed an irresponsible fiscal policy, the SPD had no plans to strengthen second- or third-tier pensions. Despite these quarrels, pension policy was largely dominated by a “grand coalition of social policy experts” and was based on a highly valued bipartisan pension consensus (Hering, 2007: 4). The labor unions were even more unlikely supporters of alternative pension policy concepts, particularly regarding plans to strengthen secondpillar pensions. They are traditionally opposed to the inegalitarian properties occupational pensions entail and worry that any expansion of second-tier pensions will be accompanied by cuts in social security pensions (Esping-Andersen, 1996). Significant social security pension reforms in the Kohl era did not happen until 1996, when contribution rates surpassed the psychological barrier of 20 percent (Hering, 2007: 4). The government reacted by passing a law that would have gradually reduced benefits from 70 percent to 64 percent through the inclusion of a “demographic factor” in the pension adjustment formula. Zohlnhoefer argues that this unpopular move, along with further economic reforms, shortly before a major election, was the government’s desperate reaction to an economic predicament that put the incumbent’s re-election at risk and at the same time could not be solved by moderate reforms (Zohlnhofer, ¨ 2003). In 1991, however, pension reform was not high on the political agenda. The status quo oriented pension policy discourse of the Kohl administration and concomitant delegitimization of opposing viewpoints are a consequence of agenda setter effects resulting from the high degree of autonomy German ministers enjoy. In Germany’s big 8 9

Westdeutsche Allgemeine Zeitung, September 14, 1984 (“Blum: Renten sicher – ¨ Opposition: Kassen leer”). Frankfurter Allgemeine Zeitung, November 29, 1984.

108

German position

People’s parties (Volksparteien), certain factions are responsible for particular policy fields. Blum’s reluctance to change the existing pen¨ sion policy discourse crowded out reformist voices because the labor wing of his Christian Democratic party not only dominated the ministry of Labor and corresponding Bundestag committees, but also the chairmanship of the Social Committees, which Blum ¨ had held until 1987 (Zohlnhofer, 2003: 129). In addition, the divisions within the ¨ Christian Democratic party, particularly discord between the business and labor wings, reinforced the status quo bias in the area of pension policy. Today, Blum’s assertion of the safety of social security pensions ¨ is rarely cited without cynicism or mockery. His assurances notwithstanding, pensioners eventually had to contribute their part to fiscal consolidation by succumbing to the introduction of health insurance contributions, relinquishment of dynamic increases, and cutbacks on disability pensions. In fact, the biggest cost reductions between 1982 and 1990 resulted from cutbacks in social security and survivor pensions, decreasing their share of GDP by 1 percent (M. G. Schmidt, 2005: 101). However, curtailments of social security pensions were not flanked by efforts to expand occupational pension coverage. On the contrary, the negligence of workplace pensions is mirrored in the tax policies of the Conservative–Liberal coalition. The government decided to raise the tax on employee contributions to direct insurance, an occupational pension vehicle that is particularly popular among small and medium-sized enterprises, from 10 percent to 20 percent (Schoden, 2003: 25). Such measures destroy trust in the funding principle and discourage employees to save for their own retirement. Trust in externally funded pension institutions was further undermined by Blum’s public ¨ reminders of past experiences with pension funding. Seeking to deflect skepticism about the soundness of social security pensions, the Labor minister argued that the German experience with economic crises in the first half of the twentieth century strongly militated against the funding of pensions. In the aftermath of 1945 and the monetary reforms of 1948, German pension funds had lost value as they converted from Reichsmark into Deutschmark on a ratio of 10 : 1. By contrast, contributors to book reserve pension schemes were less penalized, as their future pension relied not on monetary reserves but on current contributions and future company profits (Whiteside, 2006: 46).

6.4 Negotiation success in the early 2000s

109

Additional disincentives for expanding occupational pension coverage included vast increases in non wage labor costs, especially after reunification. German reunification posed unprecedented challenges for social and economic policy. The decision to pay for reconstruction programs in Eastern Germany (Aufbau Ost) with employees’ social security contributions – as opposed to tax increases – significantly increased the cost of labor for employers.10 To make matters worse, the German government tried to hide the ensuing unemployment upsurge with early retirement measures, exacerbating the extant strain on state social security pensions. The resulting increase in non wage labor costs reduced the room of maneuver for unions to negotiate wage restraint and maintain the virtuous impact autonomous wage-bargaining previously had on the economy (Mares, 2006). This new environment eroded labor unions’ wage discipline, discouraged employees from using part of their wage to build up retirement savings, and induced them to renew their previous demands for compulsory occupational pension coverage. This section has shown that not only national institutional frameworks but also cognitive schemes of legitimate pension security designs militated against the adoption of an EU pension directive. The popularity of Germany’s generous first-pillar pensions, along with the status quo oriented pension ideology communicated by the Kohl government, delegitimized any debate about alternative (more-risky) pension concepts. That this fight was not just a matter of harmonizing different institutional structures but also different value systems, is also reflected at the EU level. Whereas countries with a Beveridgean pension tradition framed the debate in terms of unequal conditions of competition that the EU needed to rectify, member states with a Bismarckian pension culture categorized it as a “social” issue that strictly belonged into the purview of the member states.

6.4 Germany and EU negotiation success in the early 2000s These divergent philosophies continued to shape national discourse when European Economic and Monetary Union heralded another 10

Some tax increases did occur, such as the solidarity surcharge and an increase of the value added tax, but they played a minor role compared to the hike in social security contributions.

110

German position

phase of EU pension policy development. Given that banks and insurance companies had long been able to operate EU-wide on the basis of a single license, pension market integration was gradually viewed as a missing link in completing financial market integration. Although representatives in the Council of ministers and European Parliament still had competing ideas about the role of the private sector and about the role of capital markets in taking care of old age provision, they were able to agree on several welfare finance reforms in 2002, paving the way for the IORP directive. Although the level of integration varied across issue areas, and many legal uncertainties make it still difficult for employers to take full advantage of the directive, this first step toward a single pension market must be considered a success. EU Commission officials interviewed for this project spoke of a “mentality change” they had noticed when discussing pension policies with members of the EU Council and Parliament.11 While representatives from Bismarkcian insurance nations had categorically rejected pension funding during the 1991 negotiations, they now acknowledged that it was no longer feasible to depend on the first pillar alone. Although the more favorable attitude toward pension reform in Germany is partly due to demographic change, a more important reason is the changing significance of unfunded PAYG pension liabilities due to the Maastricht criteria. Since political incumbents can no longer postpone pension reform into the future without being compelled at some point to cut entitlements of current, rather than future, retirees, failure to strengthen second-pillar pensions became more costly for national governments. Yet the existence of additional incentives for cooperation must not belie the friction the member states had to overcome in order to achieve the adoption of the IORP directive in 2003. Protracted struggles over the application of the prudent person rule, biometric risk coverage, and the location of the supervisory authority delayed negotiations. Commensurate with its risk-averse insurance culture, Germany demanded that cross-border pension funds should offer a high level of biometric risk coverage and operate according to restrictive investment regulations. Germany also insisted on installing the European supervisory authority in Frankfurt. 11

Personal interviews with six EU Commission officials, June 20–23, 2006.

6.4 Negotiation success in the early 2000s

111

However, due to the weakly developed second-pillar pension sector, unfavorable demographic developments, and the constraints posed by Maastricht, German negotiators found themselves in a weak bargaining position vis-a-vis member states representing a pension fund ` culture. In the end, Germany made concessions in a number of policy areas that were previously regarded as non-negotiable (investment rules, biometric risk coverage) in exchange for excluding Germany’s most sacred cow – unfunded book reserve pensions – from the scope of the directive. Although Germany considered the liberalization of investment rules as too far-reaching and the application of the home country control principle to biometric risk coverage as too lax, Germany was able to retain its ideal point in the policy area that mattered most to employer associations, the occupational pension association aba, and workers with skills that do not travel well (Hennessy, 2008). Germany fought hard for the exemption of book reserves and support funds by making the case that they are on-balance sheet pensions. EU Commission officials point out that long educational sessions held in the Council eventually helped Germany get what it wanted.12 In contrast to the hurriedly adopted directive text the EU Commission had proposed in 1991, the 2003 directive was the result of a meticulously prepared consultation process with the member states. Once we understood Germany’s biggest concern and supported the request to exclude book reserve pensions from the scope of the directive, the Germans were in favor. This was two months before we reached an agreement in the Council [in 2002, under the Spanish presidency]. The support of the Germans made things a lot easier for us, because they helped us get other hesitant members on board.13

This explains in part why Germany agreed to support the 2003 pension fund directive, but not the 1991 directive. First, the Maastricht criteria, and later the stipulations of the Growth and Stability Pact, imposed constraints on government debt – of which unfunded pension liabilities are the largest part – that had previously played no role in the formulation of Germany’s pension policy preferences. While the Maastricht criteria exacerbated existing pressures to improve access 12 13

Personal interviews with six EU Commission officials, June 20–23, 2006. Personal interview with EU Commission official, June 21, 2006.

112

German position

to second-tier pensions, Germany’s willingness to make concessions and support the 2003 IORP directive is the result of credible signaling between the member states. Public discourse during the Schroder era ¨ was also more favorable toward the expansion of workplace pensions, as the following part will show.

¨ 6.5 Domestic discourse – the Schroder era Contrary to the unwillingness of the Kohl government to expand second pillar pensions, the sociopolitical catchwords of the Schroder ¨ administration – “personal responsibility,” “generational responsibility,” and “sustainability of the welfare state” – suggested a radical break with Germany’s traditional social policy path.14 While pension policies had played no role in previous election campaigns, they became one of the most hotly contested issues in 1998 after the Kohl government had passed a series of cutbacks shortly before the election (Hering, 2007). The introduction of a new capital-funded occupational and private pension tier (Riesterrente), named after his Labor minister Walter Riester, signified the most visible break with Germany’s past focus on social security pensions. The government expected that a prefunded pension sector would affect the mind of workers, making them more “shareholder value-oriented.” Although the Riesterrente is unlikely radically to change the institutional architecture of the German market economy (Estevez-Abe, 2002), it marks an important change compared to the pension policies of the Kohl era. A major innovation the Riesterrente entails is that employees are now entitled to an occupational pension by their employer whereas in the past firms have provided occupational pension coverage only voluntarily. Employees may now request their employers to set up externally funded pension plans and convert a certain percentage of their wage into a future pension claim (Schoden, 2003: 26). To support uptake of so-called “Riester-pensions” the government generously subsidized the conversion of up to 4 percent of net wages into future supplementary pension coverage.15 The biggest resistance to Schroder’s pension reform did not come ¨ from the opposition but from the Left wing of his own party and 14 15

“Bundesregierung legt Sozialbericht vor.” Ministry of Labor, March 6, 2002. Those subsidies were scheduled to run out in 2008.

¨ 6.5 Schroder era

113

the labor unions. They viewed the introduction of new capital-funded occupational and private pension plans, along with tax and benefit cuts, as evidence of an unacceptable neoliberal turn in Social Democratic policy-making (M. G. Schmidt, 2005: 113). One aspect the labor unions found particularly provocative was the departure from the principle of parity financing. Employees were obliged to pay 2.5 percent of their gross wages into private pension plans, whereas employers were not required to make any contributions. Klaus Zwickel, the chairman of the German metalworkers’ union IG Metall, called Riester’s pension reform plans “a shame for a social democratic government” (Hering, 2007: 12). Opposition to Schroder’s pension reform was highly visible in the ¨ form of angry protests and vituperations in the media. The public outcry changed the beliefs of the Beveridgean states about Germany’s true costs of reform. Vocal objections to Schroder’s social and pension ¨ reforms demonstrated that Germany was, indeed, a high-cost type. German officials feared that exposing consumers to additional risks by agreeing to EU pension policies would only further fan the flames. Against this background, claims that Germany could not support a more liberal regulatory framework for pension funds were perceived as costly and therefore credible.16 After the exchange of informative signals at the EU level, Germany was able to extract concessions from their Beveridgean counterparts, including the maintenance of quantitative investment limits (the Beveridgean states had originally sought a more radical liberalization), the principle of home country control regarding biometric risk coverage (the Beveridgean states originally wanted to leave the decision to provide biometric risk insurance entirely to the market), and the exclusion of book reserve pensions from the scope of the directive (which the Beveridgean states had sought to include). A Social Democratic chancellor whose pension policies provoke the very constituency who has just helped elect him is highly counterintuitive to partisan theory. Schroder’s pension policies show that the ¨ coordinative discourse among policy actors may not be as elaborate in compound polities as the DA literature assumes. Schroder simply ¨ imposed social policy reforms from above without much support from either labor unions or the Left wing of his own party. When a leader 16

Personal interviews with six EU Commission officials, June 20–23, 2006.

114

German position

imposes reforms without much support from his core constituency the communicative discourse becomes increasingly important. Walter Riester framed his reform as desirable in terms of intergenerational solidarity, and rebuffed critics as irresponsible supporters of a system that was doomed to collapse under the growing fiscal and demographic strains (Stiller, 2010). The changes heralded by the Red–Green coalition, along with previous cuts in PAYG pensions and selective deregulations of private pension investments, have gradually increased the awareness of future beneficiaries that social security pensions alone will not provide financial security in old age. A 2005 survey by the Allensbach institute shows that 73 percent17 of Germans believe that their state pension will not suffice to maintain their current standard of living, compared to 45 percent of West Germans and 57 percent18 of East Germans in 1996. However, greater awareness of the need to supplement social security pensions is not accompanied by a rush to set up new occupational pension schemes. The main reason is that Germans are ill-informed about state assistance in creating pre-funded pension schemes: 80 percent of Germans know very little or nothing at all about state subsidies they may be entitled to. Even highly trained executives fare no better: only 36 percent feel that they are sufficiently informed about state subsidies for building up a funded occupational pension.19 This episode shows that the move of the Social Democratic–Green coalition toward pension funding made it easier for the German government to compromise on certain provisions (investment liberalization, biometric risk coverage) specified in the EU pension fund directive of 2003. Yet, despite the gradual acceptance of future beneficiaries that pension reform is necessary, employers, pension associations, and the government continued to defend book reserve pensions from interference by Brussels. The reasons for doing so were the same as in 1991 – subjecting book reserve pensions to EU regulations would have stripped employers of their cherished source of capital and staff retention device. As soon as the reform requirements of the 2003 pension directive were limited to off-balance sheet pensions, Germany was in favor of the directive. 17 18 19

Institut fur ¨ Demoskopie Allensbach, press release, September 19, 2005. Associated Press Worldstream (German), October 23, 1996. Institut fur ¨ Demoskopie Allensbach, press release, December 5, 2005.

¨ 6.5 Schroder era

115

While the 2003 pension fund directive accomplished the EU Commission’s goal of liberalizing investment regulations, barriers to individual pension portability – and therefore labor mobility – remained firmly in place. Therefore, in 2005, the Commission embarked on the next step towards pension market integration: the improvement of individual pension portability across and within any country. Contending that opportunities for cross-border mobility are of little use when within-country mobility barriers continue to persist, the Commission’s 2005 directive proposal targets three regulatory areas that play a critical role in obstructing labor mobility: (1) the transferability of acquired pension assets, (2) the acquisition of pension rights, and (3) the preservation of dormant pension rights.20 Unlike previous EU pension proposals, this endeavor encountered noticeably more resistance from governments and labor unions in Europe. The reason is that the 2005 portability proposal drills much deeper into national institutional frameworks than was the case for the rather technical provisions recorded in the 2003 pension fund directive. Instead of focusing on the removal of barriers to the transfer of pension assets (negative integration), the EU Commission has gone much further and proposed to regulate dynamic adjustments and vesting and waiting periods (positive integration). However, delegating control over the latter to Brussels would profoundly alter the nature of labor relations and capital flows in coordinated market economies. Consequently, the occupational pension association Arbeitsgemeinschaft fur ¨ betriebliche Altersversorgung e.V. (aba) has dubbed the EU Commission’s proposal a “Trojan horse” that could alter Germany’s institutional architecture in ways nobody desires – not even the EU Commission (aba, 2006). The following section explains how the EU portability directive – in its original version – would have increased employers’ costs of offering occupational pensions and transformed German labor relations and capital flows. Ostensibly, the adoption of pension reforms in Germany and the objective of the EU to promote pension portability do not seem irreconcilable. In anticipation of developments at the EU level, the German government had already adopted the retirement income act in ¨ 2005 (Alterseinkunftegesetz), granting citizens the right to transfer pension assets. However, the EU proposal differed from Germany’s 20

Articles 6, 2, and 5 of the pension portability proposal.

116

German position

¨ law in important ways. While the German Alterseinkunftegesetz limited the transferability of pensions to assets and therefore to off-balance schemes, the EU directive proposal envisioned the transferability of pension rights, which includes unfunded book reserve pensions. Predictably, the reaction of the key domestic actors was essentially the same as in 1991, when pension market integration was initially proposed. BDA, aba, and the governing parties (CDU/CSU/FDP) again thundered that capitalization requirements of book reserve pensions would endanger the long-term investment horizon of German companies and concurrent changes in corporate liquidity planning (aba, 2006; BDA, 2006; Deutscher Bundestag, 2006), depriving workers of valuable corporate social protection and firms of their cheap source of capital. If adopted, the portability directive would force employers to look for external capital sources, such as international financial markets. Borrowing on financial markets, however, is both more expensive (and therefore less attractive) as well as onerous, since this step requires the adoption of international accounting standards (IAS). Although many large firms listed on foreign stock exchanges have already assumed IAS, many smaller firms with a domestic focus may consider this option as too costly. An additional burden employers would face is an increase in administrative costs, resulting from the need to calculate the amount of employees’ pension assets that are being transferred from and to the firm (aba, 2006). In the same vein, German officials fear that EU portability demands will make it unattractive for employers to offer occupational pension coverage, which in turn could lead to an underprovision of corporate pensions. This prospect is particularly troublesome in light of past efforts to persuade people to build up supplementary pension coverage. As a German MEP pointed out, “If pension portability translates into higher costs for employers, this directive will constitute a death blow for occupational pension coverage. After all, we are talking about supplementary pension rights, not compulsory pension rights.”21 German actors have made it clear that they would support an EU agreement regulating the transfer of pension assets – but not pension rights – if the mobile employee will bear the administrative costs of transferability. 21

German Christian Democratic Party member, during a hearing of the Employment and Social Protection Committee, June 21, 2006.

¨ 6.5 Schroder era

117

“Why should companies bear the costs of mobile employees? Loyal workers must not be penalized. The costs associated with portability should be borne by those who cause them, that is to say, by mobile employees.” (aba, 2006: 8). According to aba, the prospect of frequent pension transfers will force firms to hold more liquid assets, resulting in immobile employees paying a “loyalty tax” due to a diminished actuarial rate of return. Aba considers this unacceptable. Whereas the issue of transferring pension assets left room for a middle ground between the German and the EU positions, Germany categorically rejected the Commission’s proposal on the acquisition of pension rights, in particular qualifying ages and waiting and vesting periods. The reason is that the proposed measures would critically alter Germany’s characteristically long-term labor relations and patient capital flows.22 As is well known, long waiting and vesting periods penalize frequent job changers, in particular women who interrupt their career for domestic care responsibilities. The right to a pension materializes only after a long period of uninterrupted job tenure. In 2001, the vesting period in Germany was lowered from ten years to five years, but the EU Commission still regards this reduction as insufficient. To improve corporate social protection for people with intermittent career paths, the directive proposal suggests an EU-wide vesting period of two years, and a waiting period of one year.23 This implies that the EU Commission attempts to shift the function of occupational pension provision away from “golden handcuffs” toward ownership of supplementary pension rights. The purpose is that employees should not have to pay a pension penalty in the event of career interruptions. The implementation of the original portability proposal would have eliminated the staff retention function of book reserve pensions. Easier pension portability might have reduced fears of occupational mobility and increased labor turnover in German companies. This development, however, would have changed employers’ incentives to train their workers. As Hall and Soskice have pointed out, the willingness of firms to invest in their employees’ education depends on their belief that 22

23

The waiting period (Vorschaltzeit) is the elapsed time before an employee earns the right to a corporate pension claim. The vesting period (Unverfallbarkeitsfrist) is the time from acquiring corporate pension contributions until the benefits belong to the employee. Waiting periods in Germany are often two years or longer.

118

German position

employers who do not invest in workers’ training will not poach extensively from those who do (Hall and Soskice, 2001: 10). Unfunded book reserve pensions that belong to employees only after long waiting and vesting periods effectively prevented poaching. Therefore, the disappearance of this staff-retention device would have rendered employers reluctant to invest in the education of their workers or offer occupational pensions in the first place.24 Thus, providing citizens with better pension portability rights at the EU level might have had the perverse result of leaving more citizens without any workplace pension coverage at all. Similar to long waiting and vesting periods, high minimum age requirements attached to workplace pensions also discourage labor mobility. A departure before reaching a certain age would result in the loss of pension rights (Commission of the European Communities, 2005: 7). In 2001, the qualifying age to draw an employer-sponsored pension in Germany was reduced from thirty-five to thirty years. The EU Commission proposal, however, proposed to lower the minimum age further, to twenty-one years. A reduction of the qualifying age is another reason employers and political actors opposed the EU portability proposal. Given that labor turnover in the group of employees aged twenty-one to thirty is much higher than is the case for older cohorts, employers feared that a soaring turnover rate would result in more pension claims being transferred, which in turn would increase firms’ administrative costs (aba, 2006; BDA, 2006; Deutscher Bundestag, 2006). Because the EU Commission proposals on acquisition rights would have considerably increased firms’ costs of providing occupational pension coverage, they were rejected by the German government. Another controversial point in German–EU deliberations concerned the regulation of dormant pension rights. Article 5 of the 2005 directive proposal states that “member states shall adopt measures they deem necessary in order to ensure a fair adjustment of dormant pension rights so as to avoid that outgoing workers are penalized.” The EU Commission justified legislative activity in this area by the need to protect outgoing employees from a depreciation of their dormant pension rights due to inflation or other factors.25 The Commission 24 25

Interview with officials at the German Ministry of Labor, July 4, 2006. EU pension portability proposal, article 5, section 7.

6.6 Merkel era

119

suggested adjusting dormant pension rights to a variety of different pension portability parameters, including inflation, wage levels, social security benefit levels, and employers’ capital gains. Although it remains somewhat unclear to what extent dynamic increases of dormant pension rights constitute a “fair” adjustment, aba estimates that this provision would have increased employers’ costs by over 30 percent (aba, 2006: 7). Echoing aba, the BDA complained that the compulsory dynamic increases would essentially penalize the sponsoring firm as well as the loyal workforce, “who can expect lower and qualitatively inferior pension claims due to the costs imposed by their mobile colleagues” (BDA, 2006: 2). Since the additional costs of dynamic increases can only be kept under control with sufficiently long transition periods, employers requested that the right to dynamic increases should be limited to new hires to preserve trust between employers and the workforce. Employers also feared that compulsory dynamic increases would lead to a complete overhaul of the German workplace pension system, affecting in particular tax breaks they have traditionally enjoyed (aba, 2006: 7). The general misfit between the EU pension portability proposal and the institutional framework sustaining the German capitalism model explains why it will be a long time until obstacles to labor mobility will be removed in Germany. The following section illustrates why the pension policy ideas communicated by the Merkel government are similarly unlikely to spark enthusiasm for future EU integration efforts in the pension sector.

6.6 Domestic discourse – the Merkel era How does the trajectory of Angela Merkel’s policies toward EU pension policies contrast with the policies of her predecessors? During the first one-and-a-half years of her tenure, several pointers suggested she would carry on her predecessor’s policy of increasing incentives to build up workplace pension rights. The tenor of the CDU/CSU/SPD coalition agreement resembled the catchwords of the Schroder admin¨ istration – more personal responsibility, more generational justice, and making the social security system “fit for the future.” In the same vein, the grand coalition raised the retirement age from sixty-five to sixty-seven years and undertook steps to encourage young families to save for supplementary pensions. These developments suggested

120

German position

that Merkel would continue with Schroder’s novel social policy course ¨ of strengthening supplementary pensions at the expense of the first pillar. However, the rise of the Left party has robbed the Social Democrats of their eagerness to implement any pension reforms that prioritize economic efficiency over redistribution. Between 2004 and 2006, the Linkspartei (or its precursor, the party of Democratic Socialism (PDS)) was represented in six state legislatures. In May 2007, it accomplished the unprecedented feat of moving into a West German state parliament after attaining more than 8 percent of the vote in Bremen. In the previous year, the Left party consistently scored between 10 and 14 percent in national opinion polls, a significant rise compared to previous years.26 The same survey put the number for the SPD at a crushing 28 percent. Initially, the numbers for the Christian Democrats were put at an equally bleak 30 percent, but in the second half of 2006 and in 2007, they consistently enjoyed a ten- to fifteen-point lead over their Social Democratic coalition partner.27 The SPD now finds itself in the uncomfortable position of fighting on two fronts: on the one hand, it needs to prevent disaffected Social Democratic voters from defecting to the Linkspartei. Hoping to woo voters who regard Schroder’s social policy reforms as a ¨ sell-out of Social Democratic principles, the Left party has vowed to fight for the removal of many agenda 2010 reforms. The Christian Democrats, on the other hand, work hard to portray themselves as the Volkspartei committed to maintaining Germany’s lavish welfare state. This is reflected in a number of initiatives by state prime ministers, such as North Rhine-Westphalia’s Jurgen Ruttgers, to advocate ¨ ¨ more generous social policies. In 2006, Ruttgers proposed to extend ¨ unemployment benefits for elderly citizens from eighteen to twentyfour months and cut this period for younger unemployment recipients. Challenging the insurance principle, he suggested that unemployment benefits should be based on the number of contribution years.28 Although this initiative initially positioned the Christian Democratic premier Ruttgers to the left of the Social Democrats, SPD chairman ¨ Kurt Beck quickly made this proposition his own. Under electoral 26 28

27 Forsa Institut. www.wahlrecht.de/umfragen/forsa.htm. Ibid. Under the current law, beneficiaries receive unemployment assistance regardless of the number of years they have contributed to it.

6.6 Merkel era

121

pressure from both the Left and the Right, the Social Democrats have decided to abandon the “new middle” and return to their left-wing, blue-collar roots. This move entails the retraction of many hard-won accomplishments of Schroder’s reform program. Although the Chris¨ tian Democrats initially sought to rescue Schroder’s reform agenda, the ¨ CDU increasingly succumbed to Social Democratic efforts to retract key measures. As Angela Merkel put it: “Because the SPD is moving to the left, we have to follow [. . . ] We will have to move to the left frontier of the middle.”29 Efforts to water down the agenda for the 2010 program, which had gained support in both the CDU and SPD, were harshly criticized by the independent Council of Economic Advisors.30 In the area of pensions, an extension of unemployment benefits for older workers would open the door to the pernicious early retirement policies of the 1990s. Employers exploited the early-retirement option excessively for staff reductions, leading to a dramatic increase of beneficiaries entitled to unemployment assistance at a time when the pool of contributors to unemployment insurance continued to shrink (Rosenow and Naschold, 1994). The ensuing strain on the welfare state demonstrated that the most important rationale of early-retirement policies – sugarcoating unemployment figures – had backfired badly. In an evaluation of the state of the economy, the Council does not see any need for further reforms in the area of pension policy, because the agenda 2010 reforms seem to have fulfilled their purpose (Sachverstandigenrat ¨ zur Begutachtung der gesamtwirtschaftlichen Entwicklung, 2007: 16). In the past two years, more than 250,000 unemployed people over fifty years of age have found jobs.31 In light of these positive employment effects, the Council prefers no further activity in this area to pension reforms that aim at turning back the wheel (ibid. 17). Public debates of “turning back the wheel” are commensurate with the notion that successful pension policies are reflected in the extent to which politicians are able to exploit voters’ fears of alternative pension insurance concepts. A survey which found that 82 percent of Germans support an extension of the period older workers can receive 29 30 31

Quoted in Der Spiegel, December 3, 2007, p. 36. Sachverstandigenrat zur Begutachtung der gesamtwirtschaftlichen ¨ Entwicklung. Der Spiegel, October 8, 2007.

122

German position

unemployment benefits corroborates this hypothesis.32 This implies that mistrust in the soundness of statutory pensions is growing, while lack of information about building up supplementary pensions is still common, as surveys of the Allensbach institute found.33 Given the current pension policy discourse, Germany is unlikely to support future attempts to create a single pension market.

6.7 Conclusion This chapter asked why Germany supported some EU pension policies, but not others. The question was motivated by the observation that Germany torpedoed the first attempt to create a single pension market in 1991, offered contingent support for a pension fund directive in 2003, and fiercely opposed the pension portability proposal in 2005. We found that partisan theory fails to account for the observed variation. This school of thought cannot explain why a Social Democratic– Green government implemented a major domestic pension reform and supported EU pension market integration while Conservative administrations did not. We argued that variation in the German position toward EU pension directives is better explained by a combination of historical institutionalism (HI) and discourse analysis (DA). HI accounts for the resilience of employer-sponsored book reserve pensions, which only exist on firms’ balance sheets and are not funded. Book reserve pensions are central in making Germany a coordinated market economy: first, they provide employers with cheap and patient capital to finance current business activities. Secondly, long vesting and waiting periods attached to book reserve pensions commit workers to the firm for a long period of time, making it an effective mechanism against poaching. As a staff retention device, book reserve pensions ensure that employers’ investment in the education of their employees is protected against freeriding by firms who do not train. Because of their role in sustaining long-term labor relations and capital flows, book reserve pensions are diametrically opposed to the European Commission’s goal of improving pension portability and labor mobility. EU efforts to eliminate obstacles to pension portability jeopardized firms’ convenient source of capital and their staff retention device, because 32 33

¨ Suddeutsche Zeitung, November 11, 2006. Frankfurter Rundschau, March 10, 2006.

6.7 Conclusion .

123

the Commission sought to include as many pension vehicles as possible under a common European framework. Depriving employers of book reserve pensions would have disrupted a key pillar of Germany’s coordinated market economy. Germany’s support for EU pension policies has been consistently contingent upon the exclusion of book reserves from the scope of EU directives. Whenever the Commission tried to apply regulatory control to book reserve pensions (as in 1991 and 2005), Germany was opposed. By contrast, when the Commission limited EU directives to externally funded pension schemes (as in 2003), Germany supported EU efforts to create a common pension market. Thus, the misfit between EU-mandated measures and German pension institutions explains the different governments’ intense lobbying on part of German companies. However, historical trajectories of pension policy development explain only part of the story. Institutional accounts fail to explain variation in the German position towards EU pension directives. For this reason, we also included an analysis of the domestic pension policy discourse. An ideas-based account helps us understand how national political economy interests informed member states’ negotiation strategy at the EU level, and provides insights into the factors that enabled governments to exchange credible signals. In the early 1990s, the pension policy discourse communicated by the Kohl government succeeded in delegitimizing supporters of alternative pension security concepts. Consequently, Germany’s requests for accommodation at the EU level failed in light of a domestic social policy agenda that was geared towards preserving the status quo. In 2001, however, the Schroder administration imposed an innovative pension reform with¨ out much public support. Therefore, German requests for concessions at the EU level (in 2003) were credible because chancellor Schroder had ¨ explicitly connected his electoral survival to the success of his ambitious and politically risky domestic reform agenda (introduction of the capital-funded Riesterrente and the notorious Hartz I–IV reforms). The Merkel government, in turn, abandoned its initial plan to expand second-tier pensions in light of rising pressures the Left party posed for the Social Democratic coalition partner. Thus, German preferences towards EU pension policies are not just a function of institutional fit, but also of cognitive templates about generational justice and individual responsibility in building up pension rights. While ideas-based accounts are often criticized for failing

124

German position

to specify the conditions under which ideas exert a causal influence on the dependent variable, this chapter has embedded discourse analysis within a historical institutionalist framework to show how a certain pension policy discourse informed German preference articulation at the EU level.

7

The British position on EU pension policies

What are British interests in a single pension market, and how does the domestic social policy discourse correspond to the EU’s effort to develop common pension policies? This chapter will argue that, contrary to conventional works that depict Britain as the European Union’s “awkward partner” (George, 1990), Britain was quite enthusiastic about EU pension policies and frequently pushed for their adoption. While Germany had to send credible signals in order to get its preferences heard at the EU level, the maturity of the British occupational pension sector made it easy for Britain to call the shots on formulating EU pension policy. This is because both goal (pension portability) and instrument (liberalization of investment rules) of EU pension policies fit well with the existing British welfare finance nexus. On the other hand, the British pension policy discourse is characterized by bitter disputes between the financial industry demanding a liberal pension regulator and consumer advocates requesting more efficient protection against corporate pension losses. Thus, in contrast to Germany, what needs to be explained in the British case is not the mechanisms that make signaling credible, but how the push-and-pull dynamics of British enthusiasm over pension market integration on the one hand and the inadequate representation of domestic consumer demands for a stringent pension regulator on the other fed into legislative outcomes at the EU level. The following sections on domestic discourse suggest that the majoritarian British party system fails to represent consumer concerns adequately. Yet, we also find that this failure does not have electoral repercussions, which is consistent with scholarship that downplays the importance of economic pocketbook voting (Kayser, 2007). British enthusiasm over a European pension market was strictly limited to pension directives that were primarily internal market measures (such as the 2003 IORP directive) while Britain opposes policies that

125

126

British position

penetrate too deeply into national pension law (such as the 1980 EC directive or the 2005 directive proposal). The reason Britain supports pension policies that are internal market measures is that the European Commission’s goals in this area – deeper financial market integration, improved access to second-pillar pensions, expansion of pension portability and labor mobility – have already been realized in Britain and therefore require only little adjustment costs. Together with Ireland and the Netherlands, Britain comprises the most developed occupational pension system in the EU. Because the 2003 IORP directive was for the most part congruent with the British national institutional framework, Britain never had to deal with the kind of credibility issues at the EU level that the Bismarckian nations faced. However, Britain opposes those EU stipulations that conflict with the complicated national wind-up procedures and minimum-funding regulations. In particular, Britain has raised concerns over additional red tape that free movement measures would place on firms sponsoring occupational pension plans. In this regard, British interests are not so different from Germany’s, despite all the fundamental differences in their pension regimes. The basic interest of both countries in this respect is to avoid too much intrusion by Brussels into national institutional frameworks. Disagreement between the UK and Germany revolves around the security of occupational pension rights. Policy-makers in both Bismarckian and Beveridgean nations profess to care about protecting beneficiaries’ rights, but deep divisions over both definition and means of protecting consumers’ pension rights abound. While Bismarckian member states typically view investor protection as most effective in protecting pension rights, the Beveridgean approach to protecting consumers manifests itself in the defense of a most liberal regulatory framework for a European single pension market. Two reasons are causally related to Britain’s intense preference in this area. First, the pervasiveness of defined-benefit pension plans requires a high-risk/high-return investment strategy. In a defined-benefit scheme, the employer has promised the staff member a certain benefit level, regardless of investment return. Such a costly commitment – wherein the employer bears all risk – can only be met by liberal investment regulations. To enable British pension plan sponsors to pursue such a high-risk/high-return strategy, the state has traditionally adopted a hands-off supervisory approach. This is reflected in the absence

British position

127

of regulations pertaining to biometric risk coverage, as well as the application of liberal investment rules permitting high levels of equity and other investments considered risky. This implies that the European debate on “pension security” is not about risk-aversion, but about expected pay-offs. The British state maximizes its utility by shifting a large part of retirement income provision to the private sector. This comes at the expense of adequate protection of consumers’ accrued pension rights. The German state, by contrast, maximizes its utility by protecting investors against market risk. Regulations prohibiting risky investments come at the expense of high returns and an underdeveloped occupational pension sector. Second, causal explanations need to attend not just to what happens but also when events happen relative to other events or ongoing processes (Hacker, 2002: 26). Thus, we argue further that the failure to establish an effective supervisory system when the pension system was privatized in 1986 renders all government attempts belatedly to introduce regulations a matter of negotiation with the powerful pension industry and therefore uncommonly difficult. This omission lies at the heart of Britain’s domestic pension policy discourse, which is characterized by bitter contestation between the state and advocates of occupational pensioners. Questions of pension security did not figure prominently in the Thatcher administration. When the Pension Act of 1986 dramatically shifted responsibility for retirement income from the state to the private sector, strong stock market performance had contributed to great wealth accumulation in British occupational and private pension funds. According to former Labour minister Frank Field, British corporate pension schemes were considered the envy of the world. Why burden the system with cumbersome rules when it seemed to work just fine? In the end, the Thatcher government refrained from installing a tough regulatory regime at the time of privatization, partly as a result of Conservative ideology that sought to minimize state influence in the economy, to some extent because the majority of occupational plans were defined-benefit schemes, and because the system as a whole was thought to function effectively. As a result of this foregone opportunity, the British government changed or amended existing pension laws only after instances of pension fraud created domestic audience costs (Fearon, 1995) and subsequent expectations for a socially acceptable remedy.

128

British position

Although the strong reliance on the private sector in providing retirement income alleviates the burden on fiscal outlays, it also constitutes a predicament for the government: on the one hand, a laissez-faire regulatory approach was deemed necessary to enable the private sector to honor their occupational pension liabilities. On the other hand, the perpetual question British authorities grapple with concerns the extent to which financial institutions should be subject to market mechanisms while avoiding negative externalities, such as corporate pension fraud, mis-sellings, excessive administrative charges for consumers, or the winding up of pension schemes with enormous deficits. It is particularly the termination of corporate pension schemes that regularly results in a showdown between the state and consumer advocacy groups demanding adequate compensation for victims of pension fraud. Responding to such negative externalities confronts the government with a dilemma: compensating the victims of pension fraud seems unjust from the taxpayers’ viewpoint, because the schemes are either private or occupational ones and as such not underwritten by the government. In addition, a bailout raises the specter of moral hazard: employers may not invest pension assets responsibly because in case of insolvency they expect the state to take care of pension fraud victims. On the other hand, failure to compensate the victims is certain to generate voter resentment since the government has aggressively encouraged the uptake of occupational and private schemes in the past. Placing the burden of compensation on the taxpayer, in turn, should theoretically create demands for a more effective state regulator. The downside of more pension regulations, however, is that mountains of legislation aimed at protecting consumers’ pension rights either get blocked by the powerful pension industry, turn out to be woefully inadequate, or are so costly that many employers have decided to close their most generous occupational pension plans to new entrants (Fawcett, 2002; Blake, 2005). Given the state’s laissez-faire approach towards pension fund regulation, British support for a liberal European framework governing cross-border pension transfers is unsurprising. However, in the two decades prior to the adoption of the European IORP directive, largescale pension scandals have greatly undermined public confidence in occupational and private pension plans. From this perspective, one

British position

129

might have expected the UK to take a somewhat different stance – one that is closer to the tough regulatory approach of Bismarckian governments. The mis-selling of personal pension plans in the late 1980s and early 1990s, as well as Robert Maxwell’s infamous theft of his employees’ pension assets, all revealed weaknesses in a self-regulatory system governing pension products. Between 1991 and 2004, well-known firms such Kalamazoo, United Engineering Forgings, Allied Steel and Wire, Blyth and Blyth, Dexion, among others, all wound up their pension schemes with deficits of several million pounds. Thousands of pension fund members lost large portions of their occupational pensions after their sponsoring employer became insolvent, even though the Pension Protection Fund offered some compensation to individuals who lost their benefits. In 2004, the Department of Work and Pensions (DWP) estimated that 65,000 people had lost more than 20 percent of their expected pension benefits from underfunded schemes wound up by insolvent employers (Department for Work and Pensions, 2004). In 2007, estimates have put the number of pension victims at 125,000 people.1 As a result, trust in private sector pension provision is at an all time low. In light of the regulator’s inefficiency in protecting pensions, and subsequent public outcry, one might have expected the British government to seize upon European pension directives to respond to demands for a more efficient pension regulator in the UK. The adoption of strict consumer protection laws, as envisioned by original drafts of the 2003 IORP and 2005 portability directives, may have averted several pension scandals in the first place. Tight regulatory oversight may have also signaled to British voters the government’s commitment to safeguarding accrued pension rights and the capability to revamp an ineffectual regulator. Such a position would have also facilitated negotiations with Bismarckian governments who have traditionally precluded pension sponsors and administrators from pursuing high-risk/high-return investment strategies. However, the British government never lobbied for the creation of a restrictive pension framework during the examined time period (1991– 2005). On the contrary, together with the Irish, Dutch, and Swedish governments, Britain vigorously pushed for the application of liberal 1

Daily Telegraph, January 25, 2007

130

British position

regulations. The most prominent negotiation success Britain achieved constitutes the EU-wide adoption of the prudent person rule in the 2003 IORP directive, which required no adjustment costs in Britain and other mature pension fund cultures but constituted the liberalization of investment rules in Bismarckian nations. The reason why the government did not seize the IORP directive to implement more-stringent pension fund regulations at home is that an ineffective regulatory regime does not seem to pose a political danger to political incumbents. This goes against conventional wisdom postulating that voters regularly punish governments for poor economic outcomes, particularly in systems where accountability is high (Powell and Whitten, 1993; Powell, 2000). In the same vein, we should observe individuals suffering significant losses of pension assets to engage in “pocketbook voting.” Yet, survey research shows that neither Thatcher’s radical pension reform, nor Tony Blair’s highly unpopular occupational pension policies had electoral repercussions. In both cases, citizens’ assessments of the government’s economic policy competence remained favorable, even when corporate pension losses dominated daily news (Clarke, et al., 2004: 86 ff.). Favorable assessments of the government’s economic policy in times of corporate pension scandals suggests that victims of pension fraud tend to blame their individual employer for their plight, not the government for flawed regulatory oversight. As the following sections will show, the notion that a few reckless employers are to blame for widespread pension fraud has been actively promoted and reinforced by politicians. For such an unambiguous message to stick, it is not necessary to assume that voters are uninformed fools. However, it is reasonable to presume that they are not as capable as highly trained pension experts to evaluate the menu of regulatory supervision choices governments have at their disposal. Thus, the biggest challenge that remains for the British pension system is not EU-induced, but domestic in origin. Yet – and in sharp contrast to Germany – the failure to represent consumer interests never jeopardized Britain’s resolve or credibility at the EU level. No matter how bitterly the DWP was attacked by opposition parties, pensioner advocacy groups, or employers – British negotiators continued to influence EU pension policy design and sacrificed only when the Bismarckian member states were able to signal credibly that their domestic win-set was indeed constrained.

7.1 Workplace pension system

131

7.1 The British workplace pension system The flaws in the British pension system diverge sharply from the ones plaguing the German system. While Germany’s generous social security pensions are highly vulnerable to demographic aging, the demographic time bomb is ticking less loudly in Britain because the low level of the state earnings-related pension scheme (SERPS) keeps future entitlements under control. SERPS is considered effective, targeted, nondistortive for the labor market, and projections for future pension expenditures are estimated at 4 percent of GDP in 2050 (Schulze and Moran, 2007: 49). As a result, SERPS places less strain on government finances than social security pensions in Germany. The picture looks less rosy, however, when we consider that state projections often do not account for the increasing costs for means-tested benefits. In addition, constant decline in the real value of pension benefits implies that SERPS alone does not provide for an adequate standard of living (ibid. 50). To offset the low level of SERPS, Britain’s occupational and private pension system is well developed. The National Association of Pension Funds (NAPF) estimates that UK pension fund assets totalled almost 1.6 trillion pounds at the end of 2006.2 In Europe, 82 percent of all assets of pension funds are managed by three countries: the UK administers 57 percent, or 1.366 billion euros, the Netherlands 27 percent, or 635 billion euros, and Switzerland 16 percent, or 378 billion euros (EFRP, 2007: 28), making the UK the dominant player in the European supplementary pension landscape. Yet these impressive numbers mask that the British occupational and private pension system is fraught with problems: workers are exposed to high levels of biometric and investment risk, regulatory oversight by the government is of poor quality, employers increasingly close finalsalary schemes to new hires, and beneficiaries find pension plans non transparent and difficult to understand. British politicians frequently refer to their regime as “one of the most complicated pension systems in the world” (Department for Work and Pensions, 2002: 79; see also Lord Leitch (Labour), Hansard, HL Deb. (6 June 2007)). Furthermore, frequent tampering with pension regulations – such as taxation of pension fund surpluses – causes uncertainty among plan sponsors regarding future returns and liabilities (Weaver, 2006). 2

See online data by NAPF. www.napf.co.uk

132

British position

defined-contribution deferred/pensioner (6%) defined-contribution active (12%)

defined-benefit active (21%)

defined-benefit deferred/pensioner (61%)

Figure 7.1 British workplace pension schemes

In Germany, the first-pillar PAYG pension provides extensive coverage for the biometric risks of longevity, disability, and dependent survivors. The flat-rate pension in the UK, by contrast, does not provide such coverage. Occupational pensions, which tend to take the form of individual savings plans, are often paid out as a lump sum and therefore do not cover the biometric risk of longevity. This implies that the retiree may outlive the benefit. Coverage of other biometric risks, such as disability and benefits for surviving dependants, is also often lacking (Haverland, 2004: 6). Discrimination against disabled individuals is particularly pronounced since 1995, when the Pension Act gave plan sponsors the right to refuse pension fund membership to individuals if the cost of providing benefits was “substantially greater than it would be for a comparable person without the disability” (Blake, 2005: 9). The most common retirement benefit in the UK, the final-salary scheme, is similar to the German defined-benefit pension (Leistungszusage). Final-salary schemes are based on a proportion of the final salary, that is, a graduated benefit, which increases with the relevant salary measure (Blake, 2003: 168). Figure 7.1 shows that, in 2006, 61 percent of all 13 million workplace pension schemes were calculated according to the final-salary formula.3 If the earnings profile of the employee is such that the terminal salary is less than the salary earned in earlier years, then the method of calculation might be based 3

Ibid.

7.2 Thatcher era

133

on average salary or some variant such as the best three years in the ten years preceding retirement (ibid. 169.) At the other extreme, more than 30 percent of workplace pensioners are part of a money-purchase scheme, or defined-contribution pension, in which the pension is a flatrate benefit depending on the amount of contributions paid and the interest earned on these contributions. In this case, the employee carries the investment risk.4 British pension fund structures are derived from trust law, which has been used for wealth preservation within families since medieval times. Most occupational pension schemes are set up as a trust fund, wherein the sponsoring company and the pension plan administrator (the trustee) are two separate entities. This stands in contrast to the widespread German book reserve pension, where the employer himself is the pension-plan administrator. The pension trust serves three functions: it is the primary source of payment of pension entitlements; it is a security for payment; and it is a vehicle for the collective protection and enforcement of the rights of individual scheme members. In the past, the trustee’s role and training have been increasingly contested: supporters of the traditional lay trustee system agree with Sir Roy Goode, the Oxford University professor who led the pensions law review following the death of Robert Maxwell. Goode argued that trustees should not just be “techies,” but lay people who can create confidence among plan members.5 Others defend the governmentsponsored report by Paul Myners, former chairman of Gartmore, who identified trustees’ lack of investment knowledge as the main problem. Past reforms, such as 1995 Pension Act, the 2000 Trustee Act, and the 2004 Pension Bill, have repeatedly redefined the relationship between trustees and sponsoring employers. As the following sections will discuss in more detail, the sources of these reforms were politically motivated attempts to react to domestic corporate pension scandals as well as European Union legislation.

7.2 Domestic discourse – the Thatcher era During prime minister Thatcher’s reign, discontent with the pension regime focused on questions over collective or individual control over 4 5

See ibid. Financial Times, March 15, 2004, “The big picture: pension fund trustees’ role is facing overhaul.”

134

British position

the pension funds. Following Thatcher’s radical but largely hidden attack on the earnings-related state pension (Pierson, 1994), occupational and personal pension plans now represent the old age savings of millions of workers. Striving to afford citizens more choice in retirement arrangements and more personal control over the financial resources on which entitlements are based, the Conservatives reduced vesting periods for scheme leavers and for the personalization of pensions (Hannah, 1986: 139). The Tories succeeded in changing both discourse and options in the pension policy debate, enhancing the attractiveness of contracting out of SERPS and purchasing occupational and private schemes. In line with the median-voter theorem (Downs, 1957), Labour therefore departed from its opposition to second- and thirdtier pension provision as a matter of principle and focused instead on regaining control over the wealth of pension funds. Employers were powerful in pension trusts from the beginning and gradually consolidated their grip by using increased funding of their employees’ plans as a managerial tool to squelch union demands for more control over the investment strategy of the funds. Yet, in his report A Review of Investor Protection, which became the blueprint for the 1986 Financial Services Act 1986, Jim Gower recognized that a balance had to be struck between market freedom and investor protection. In his view, “regulation should be no greater than is necessary to protect reasonable people from being made fools of.”6 Gower proposed a model of regulation focused on self-regulatory bodies, which would encompass all securities and investments. The industry would appoint its own regulatory authority, accountable to an overall regulator. Sponsoring companies were contractually obliged to comply with the rules of their self-regulatory organizations. However, the laudable ambition to balance investor freedom and consumer protection was not served well by the unworkable mass of regulations and rules, which rendered compliance with the regime difficult. Infringements of the rules carried a right to compensation and so exposed those concerned with arranging investments to the risk of claims from private investors.7 The compliance burden, together with adviser fears, had enormous cost implications that were ultimately borne by the consumer. In short, the measures introduced by

6

Quoted in Financial Adviser, March 8, 2007.

7

Ibid.

7.3 Negotiation failure in the early 1990s

135

the 1986 Act proved ineffectual in preventing large-scale corporate pension fraud. The failure of the government to implement a strong disclosure, licensing, and supervision regime amplified the risk that employers and private pension scheme sponsors might abuse corporate pensions by diverting them from their original purpose. However, this risk was obfuscated by political incumbents’ ideological standpoints. According to Conservative ideology, a laissez-faire regulatory approach was seen as the most efficient way to ensure fair competition in the financial industry. Since individuals would make financially prudent choices for themselves, burdening the pension industry with additional regulations was perceived as detrimental to the self-regulatory market. It is tempting to ascribe the light regulation of the pension industry and concomitant lack of an effective regulator to the Conservative Party’s ties to City of London interests. After all, it is conceivable that the 1986 reform was supposed to channel rents to an important constituency. But the financial sector’s response to Thatcher’s plans shows that the pension industry was highly skeptical of compulsory private pension schemes (Jacobs, 2011). The industry’s biggest concern was that the administration of many small private pension plans would be exceedingly costly. In addition, pension funds objected to the prospect of future government interference in the private pension market, which the management of compulsory schemes might bring (Pierson, 1994: 62). It is therefore unlikely that the 1986 reform was motivated by politicians’ desire to cater to financial interests – unless they dramatically misjudged the preferences of the group they were trying to serve. Instead, a set of ideological factors seemed to carry more weight in the decision to privatize the system (Jacobs, 2011).

7.3 Britain and EU negotiation failure in the early 1990s In light of this background, what were the international reactions to Britain’s first proposal for designing a single pension market? When British Commissioner Sir Leon Brittan, Thatcher’s former home secretary, first presented his proposal for cross-border pension funds in 1991, the European member states had a mixed opinion of the British pension system. On the one hand, the UK enjoyed a unique balance of pension provision. The basic state pension was topped up with extensive and lucrative occupational and private pensions. During the

136

British position

golden years of strong stock market performance in the 1980s, with pension assets averaging double-digit returns, occupational pension plans had provided a comfortable retirement for many people. Scheme members generally considered themselves to be in “secure and safe” schemes (Ring, 2005: 353). On the other hand, the strong stock market performance led pension plan trustees to focus on pension asset management to the detriment of their liabilities. They took advantage of high asset values and returns by taking so-called contribution holidays, enabling benefit increases, introducing cost-of-living adjustments, and increasing net income by reporting pension asset income as corporate income. All this was offered without triggering increased contributions (Clark and Monk, 2006: 45). According to Inland Revenue figures, employers collectively saved almost 18 billion pounds during the 1990s pension contribution holidays. Ian Wright, Labour MP for Hartlepool, recounts the failure of Unilever to honor their future pension obligations: In the 1990s, Unilever took seven years of pension holidays. It also took 270 million pounds out of the then pension fund surplus and put it into its profit and loss account. In the decade after 1992, almost 1.5 billion pounds were taken from its pension fund and almost two thirds given back to shareholders in the form of higher profits and larger dividends. I can envisage people saying, “nothing wrong with that”. However, last month, the company announced that it would close its final salary scheme to new entrants. It has failed to provide for liabilities in the long term. (Hansard, HC Deb. (17 Apr. 2007)).

This statement reflects a broader trend of MPs to put the blame for firms’ failure to honor their pension liabilities squarely with the relevant employer. The government’s role in permitting contribution holidays and payout of pension surplus to shareholders in the first place is rarely mentioned in House of Commons debates. After the stock market crash and subsequent years of stock market underperformance, the problems associated with contribution holidays and underfunding became readily apparent. In addition, the confluence of low asset returns and interest rates in 2001 and 2002 was so unexpected that few sponsors had run stress test scenarios of such an outcome. This had devastating effects on pension plans: as lower discount rates increased the value of expected liabilities, the assets intended to cover these liabilities fell dramatically in value. The

7.3 Negotiation failure in the early 1990s

137

confluence of declining interest rates and asset returns took Fortune 1000 plan sponsors from an average funding level of 122 percent in 1999 to 76 percent in 2002 (Clark and Monk, 2006: 47). As a result, occupational pensions fell out of favor. In addition, stock market fluctuations have emphasized the risks alternative forms of pension vehicles, such as money purchase provision. Yet, regarding EU negotiations, Leon Brittan sought to capitalize on those aspects of the British pension system that during his tenure were portrayed as the “crown jewel.”8 Indeed, in the 1980s and 1990s, employer-sponsored final-salary schemes enjoyed vast surpluses and the self-employed and non-pensioned employed accepted that they needed to take responsibility for a large part of their retirement income. Seeking to portray pension market integration as a natural outgrowth of the single market for goods and services, Brittan argued in his speech of July 2, 1990 that “the London stock exchange was dominated by the British pension funds and that in order to achieve a common financial market it was no longer possible to avoid integrating supplementary pensions and pension funds” (Esposito and Mum, 2004: 7). In an earlier draft, Brittan had outlined the three freedoms (investment, provision of services, and cross-border membership) necessary to create a single market for pension funds, of which the last was the most controversial part. Cross-border membership of pension funds would have lifted an important barrier on labor mobility, but would have involved much longer negotiations on minimum vesting and funding requirements, employee representation in funds, and a certain harmonization of the taxation of pension funds, to avoid major distortions. Thus, to minimize opposition, Brittan ultimately decided against pushing for a liberalization of cross-border membership and instead advocated all but freedom of investment and provision of services.9 Relative to the demographic problems and budgetary strains of Bismarckian nations, Britain compared favorably in terms of aging, fertility, and government spending on state social security pensions. The Federal Trust calculated in 1991 that failure to reform the current pension system will cause public expenditure on pensions to rise by more than 30 percent in Belgium and the UK, 40 percent in Germany and Italy, and a staggering 70 percent in France over the next forty 8 9

Financial Times Business Limited, May 1, 2007. Financial Regulation Report, 18 November 1991.

138

British position

years.10 Based on this outlook and the strong performance of British final-salary schemes in the 1980s, other member states, notably Ireland and the Netherlands (as well as the respective national pension association) supported Leon Brittan’s claim that the mature pension fund cultures in Europe should be protected while the Bismarckian member states, under great pressure to increase second-pillar pensions, should be the ones to reform. But the overall framing of supplementary pensions as a financial product – as opposed to insurance against longevity and other social risks – did not go down well with the Bismarckian member states. First, politicians in Bismarckian nations were afraid of voters’ response to a directive that seemed to emphasize workplace pensions, possibly at the expense of social security pensions, and import ill-suited Anglo-Saxon regulations to a well-functioning welfare state. In terms of EU-wide competition, the idea of British pension fund managers dominating the European asset-liability business and freely handling continental pension funds where codetermination has been the norm, was too much to swallow from a Bismarckian perspective. While the Beveridgean pension fund cultures focused on the financial market aspects of the directive, the Bismarckian governments feared voters’ reaction to a directive that they perceived as a social, redistributory measure. Therefore, a broad alliance of MEPs representing Bismarckian nations demanded the adoption of more rigorous rules for a single pension market that underlined pension security. Their principled opposition against pension market harmonization precluded the Bismarckian member states from sending any costly signals that could have provided the Beveridgean member states with a clue as to their “true” type. Thus, failure to distinguish between cheap talk and sincere requests for concessions regarding EU pension market harmonization induced the Beveridgean member states to refuse any accommodation. From their point of view, Bismarckian nations were under much greater pressure to reform and therefore had to give in eventually. The Bismarckian governments’ opposition to Leon Brittan’s pension fund proposal derives not only from the British framing of pension market integration as an internal market measure, but also from the unfortunate timing of the Maxwell affair, hitherto the biggest pension scandal in British history. Only three weeks after the Commission 10

Financial Times, November 8, 1994.

7.3 Negotiation failure in the early 1990s

139

had formally presented Brittan’s directive proposal, the Serious Fraud Office launched an investigation into how the pension schemes of Maxwell’s public companies, Mirror Group Newspapers and Maxwell Communication Corporation (MCC), had incurred potential losses of 400 million pounds on loans to his main private companies, Headington Investments and Robert Maxwell Group. It eventually emerged that during the six months prior to his death Maxwell had stolen 700 million pounds in cash and securities from his public companies and their pension funds. Maxwell had failed to disclose either to the Stock Exchange or to shareholders that his effective ownership of MCC shares had increased from the publicly disclosed level of 68 percent to 79 percent. With 79 percent of MCC shares owned by Maxwell, the Stock Exchange would have obliged him to take his company private (Blake, 2003: 340–343). At the EU level, British lobbying for the adoption of Anglo-Saxon investment and oversight regulations when those same rules had just failed to prevent the biggest pension fraud in UK history seemed a futile task. Shortly after the Maxwell affair had begun to make headlines, critics blamed Britain’s previous failure to implement the 1980 EC Insolvency directive (80/987) to protect the pay and pensions of employees in the event of insolvency. Under this directive, Britain’s government should already have set up a compensation scheme for pensioners who fell prey to the bankruptcy of the firm sponsoring their fund. The government denied accountability, but prominent consumer advocates, such as pension lawyers Ros Altmann and Robin Ellison, have on numerous occasions thundered that the proper promulgation of the directive could have alleviated corporate pension fraud in the first place. Implementation of the 1980 EU directive occurred only with the 2004 Pension Act, when the Blair government responded belatedly to several high-profile firms winding up their pension schemes with million-pound deficits with the establishment of a Pension Protection Fund (PPF). This instance reveals once again Britain’s willingness to implement EU pension directives that are internal market measures as opposed to measures like the 1980 insolvency directive that affect national pension scheme termination and minimum funding regulations wherein the member states have priority. Disagreements over who was to blame in the Maxwell affair provided opponents of Leon Brittan’s proposal with further ammunition to torpedo the adoption

140

British position

of Anglo-Saxon welfare finance regulations, and therefore prevented an informative exchange between Bismarckian and Beveridgean governments.

7.4 Britain and EU negotiation success in the early 2000s After banks, investment companies, and insurance firms had received permission to operate EU-wide on the basis of a single license in the early 1990s, pension funds finally requested the same privilege for themselves in the early 2000s. The potential benefits of the IORP directive for the UK have long been clear: the international expertise of British fund managers was seen as an excellent prerequisite for entering foreign markets. Furthermore, since the City of London had always held a clear competitive advantage in the business of pension management vis-a-vis continental financial centers (Talani, 2000), ` British IORP sponsors expected to exploit regulatory arbitrage. However, the IORP directive was not without problems for Britain. Even though most of the stipulations set forth in the IORP directive were already part of British pension law, the biggest problem for Britain was how to implement article paragraph (c) of Article 16(2). The article, which has attracted the most criticism in Britain, applies where a fund that has a recovery plan begins to wind up. The transposition of this stipulation in the UK has been a lengthy process because of the complicated interrelationship between the country’s scheme funding, PPF provisions, and existing winding-up provisions.11 Article 16(2) requires an occupational pension scheme that has started to wind up during a recovery period (the period covered by a “recovery plan”), which sets out how a shortfall in the scheme’s funding is to be made good, to do the following: r inform the competent authorities in the home member state; r establish a procedure to transfer the assets and the corresponding liabilities to another financial institution or similar body; r disclose that procedure to the competent authorities; and r make available to members and, where appropriate, their representatives, a general outline of the procedure, in accordance with “the principle of confidentiality.” 11

European Pensions and Investment News, May 8, 2006.

7.4 Negotiation success in the early 2000s

141

These stipulations were bothersome for the British government because – until the EU initiated infraction proceedings against the UK in 2006 – British pension legislation did not expressly enforce the stipulation requiring scheme administrators to inform the competent authorities and did not require all schemes that begin windingup to have a recovery plan in place. The British government balked at the requirement to cover all schemes because it meant imposing additional administrative costs and burdens on employers. Given the strong reliance on the private sector in topping off the inadequate state pension, the British government was intent on sparing employers the pains of adhering to yet more onerous legislation by Brussels. Britain’s failure to transpose this part of the directive prompted the European Commission to send a reasoned opinion, the first stage of the formal infringement process, to the British Department for Work and Pensions. Acknowledging the reprimand, the DWP opened a public consultation with the Association of Consulting Actuaries, the Association of Pension Lawyers, the Financial Services Authority, and the Government Actuary’s Department. These deliberations resulted in an explanatory memorandum (Department for Work and Pensions, 2006b) detailing the ways in which the IORP stipulations and existing wind-up schemes would be reconciled. The government argued that two regulations in the Pension Act of 2004 would implement article 16.2(c) as stipulated in the IORP directive. One is embodied in regulation 7(1)(e) of the Occupational Pension Schemes Regulations of 1996. This requires the trustees or managers to make available, on request, an outline of the winding-up procedure to scheme members and prospective members; their spouses and civil partners; scheme beneficiaries; and recognized trade unions (Department for Work and Pensions, 2006a). The second stipulation, spelled out in section 231A, requires the trustees or managers of a scheme that starts to wind up during the period specified in the scheme’s recovery plan prepared under section 226 of the Act to prepare a winding-up procedure “as soon as practicable”; prescribes the content of the procedure; requires trustees to send a copy of the procedure to the Pensions Regulator “as soon as practicable”; and provides that a civil penalty applies when trustees or managers fail to take reasonable steps to comply with the requirements (ibid.).

142

British position

However, it still remains unclear how formulations such as “as soon as practicable” will be enforced. It is therefore likely that the conflict between the UK and the Commission on the proper transposition of Article 16.2(c) will continue for some time. Such details do not figure prominently in public pension policy discourse, but may have potentially far-reaching implications for wind-up procedures and minimumfunding regulations when the European Court of Justice will set a precedent. For now, the British government seems to be content with its reputation as a major force promoting further financial market integration. Unpopular rulings by the Court can always be blamed on “Europe.” Even though Britain, as holder of the Council presidency in 2005, has helped push the portability directive onto the EU agenda, the government’s stance on the directive’s key provisions (overall goal, transferability of acquired pension assets, acquisition of pension rights, and preservation of dormant pension rights) was not too different from the viewpoint of its German counterpart. The DWP supported the overarching principle of increasing labor market flexibility behind the directive, but – like its German counterpart – has maintained the line that it is important to achieve a balance between improving rights for members of pension schemes and any additional burden on schemes, remaining schemes members, and employers. “We believe the UK has already legislated to ensure pension portability” (Department for Work and Pensions, 2006b: 5). Similar to the German government was also the British point of view that the overlap between the IORP and portability directives was not absolute in scope. “This is necessitated by the difference in objectives of the two directives, the former being primarily an internal market measure which regulates schemes, and the latter, a free movement measure focussed on increasing the rights of workers” (ibid. 9). On the transferability of assets, the DWP echoed the German government by emphasizing that only pension assets should be refunded or transferred, rather than the actual value of the contributions. UK employers essentially offered the same reasons for their rejection as their German counterparts: the calculation of actual pension values would be extremely costly for pension trusts, particularly those managing many smaller schemes.12 12

Interview with six EU Commission officials, June 21–23, 2006.

7.5 Blair era

143

Strikingly, even the EFRP – the UK-dominated European lobby group for pension funds – was not an unconditional supporter of the 2005 portability directive. While the EFRP welcomed the overarching goal of increasing labor mobility, it vehemently opposed the Commission’s proposal to index dormant rights and eliminate vesting periods. These proposals was deemed unacceptable, because it would have required EFRP member institutions to administer very small pension schemes and thus incur disproportionate administrative costs. Currently, pension funds are allowed to impose certain thresholds under which schemes may decline to manage pension entitlements. To maintain the profitability of managing large pension schemes, the EFRP continues to oppose the indexation of dormant rights and the elimination of vesting periods. Although the EFRP is usually one of the most prominent groups lobbying for a single pension market, their stance on the 2005 directive proposal is consistent with the position of the UK government, which seeks to strike the right balance between providing the pension industry with new business opportunities while keeping administrative costs as low as possible.

7.5 Domestic discourse – the Blair era Tony Blair came to power in 1997 ideologically committed to establishing a “third way.” Recognizing the declining significance of bluecollar workers for winning elections, Blair’s “New Labour” party departed from traditional Left-wing ideas which explicitly aim to advance the interests of the working class and instead worked hard to win the business community’s trust in his economic policy. In the realm of pension policy, the main goal of the Labour party was the continuation of Thatcher’s philosophy of individual risk-taking, “to help people help themselves” (Department of Social Security, 1998a). In 1998, the government lauded workplace pension schemes as one of the “great welfare success stories of this century” (Department of Social Security, 1998b: 18). However, only a few years later abuse of workplace pension schemes as well as the government’s inability to protect pension rights had contributed to consumers’ loss of confidence in both the regulatory capacity of the state and the expertise of the financial industry. Despite laudable goals to enhance pension security, the mountains of legislation produced under the Labour government has not only failed to protect accrued pension rights, but

144

British position

in some cases actually undermined trust in the state as protector of rights. In 2006, the Parliamentary Ombudsman, Ann Abrahams, concluded in her report “Trusting the Pensions Promise” that between 1995 and April 2004 successive governments have published information about schemes that was “sometimes inaccurate, often incomplete, largely inconsistent, and therefore potentially misleading and [. . . ] this constitutes maladministration” (Department for Work and Pensions, 2006c: 6). Public outcry over Labour’s occupational pension policy first occurred in 1997, when former chancellor Gordon Brown abolished dividend tax credits for corporate pensions. The main criticism of this policy was that it applied without a transition period, making it impossible for fund managers to offset the resultant shortfalls. Labour MP Ed Balls defended the scrapping of the tax credit by arguing that the withdrawal of tax credits was part of a wider package of measures designed to improve the climate for long-term investment in the UK. “This measure removed a distortion in the tax system, which encouraged companies to pay out their profits in dividends, rather than retain them for reinvestment in the business” (Hansard, HC Deb. (23 Nov. 2006)). But those representing consumer interests condemned the government’s decision to favor industry interests at the expense of pensioners: “So because the CBI [Confederation of British Industry] wanted a more favourable tax system for UK companies, the new Labour Government decided to sacrifice the interests of unsuspecting pension scheme members and personal pension investors.”13 Gordon Brown’s decision to remove tax relief on pension funds continued to haunt the chancellor until 2007, when the Tories confronted him with a vote of no confidence in 1997 – a highly unusual move against an individual minister. Following several high-profile pension scheme wind-ups, where a total of 125,000 workers had lost all or part of their corporate pensions, the Tories charged that “this house has no confidence in the chancellor of the exchequer’s handling of occupational pensions.”14 Brown defended his position by arguing that his pension decisions had raised 5 billion pounds a year and had been a key reason the UK economy had been so strong. Shadow chancellor 13 14

Ros Altmann, “Release on Gordon Brown’s Removal of ACT Relief.” www.rosaltmann.com/ACTDynamite.htm. Guardian, April 17, 2007.

7.6 Conclusion

145

George Osborne, who tabled the motion, accused Brown of “acting with stealth” and destroying the lives of those whose funds had been raided. However, with Labour’s majority in the House of Commons, the motion was defeated by sixty-five votes.15 This episode demonstrates that frequent tampering with occupational pension regulations destroys trust in the security of workplace pensions. Brown’s decision to abolish the tax relief on pension funds has highlighted the contradictions between official Labour ideology, encouraging individuals to contribute to corporate pensions, and government policy, reneging on incentives designed to realize this very vision. While the problem of such short-termism might be particularly pronounced in majoritarian systems, the previous chapter on Germany has shown that proportional representation systems are not immune to this quandary. A good number of Germans felt just as betrayed by German chancellor Kohl’s decision to increase tax on a particularly popular pension vehicle (direct insurance) as many British did by Gordon Brown’s “tax grab”: they would not have chosen this kind of pension scheme had they known in advance how their contribution would be taxed.

7.6 Conclusion The maturity of the British occupational pension system as well as the congruence between EU pension policy proposals and the British welfare finance nexus make it relatively easy for the British to get their preferences written into EU pension directives. European Commission officials interviewed for this project all agreed that “nobody can force anything on the British in terms of pension policy” (personal interviews, June 20–28, 2006). When British officials did offer concessions to their Bismarckian counterparts (for example, in the form of quantitative investment limits), it was after the latter had credibly demonstrated that domestic constraint demanded a different course of action. As a consequence of the 2003 pension fund directive, the already competitive British pension industry will be able to enter previously closed markets, offer consumers more choice in occupational pension products, and engage in regulatory arbitrage at home.

15

BBC News, April 17, 2007.

146

British position

However, not all EU decisions on pensions are coveted in Britain. The main defect of the British pension system is domestic in origin: public trust in the security and viability of employer-sponsored pensions is at an all-time low following a series of corporate pension defaults wherein thousands of employees have lost part or all of their company pension. It is precisely in this area – regulatory oversight – where British officials are wary of EU rulings. Yet, where EU stipulations conflict with British interests, the UK has managed to formulate certain EU directive paragraphs so vaguely that eventually the European Court of Justice will have to interpret their precise meaning. The most prominent examples are the wind-up provisions and minimum-funding regulations recorded in the IORP and pension portability directives. Leaving the interpretation of important directive paragraphs to the Court raises the specter of an ever-wider democratic deficit. However, the government’s lobbying for pension market integration is strategic in the sense that it hopes to be remembered as the main force driving financial market integration, whereas unfavorable ECJ rulings in the future can be blamed on “Brussels.” Even though the 2003 IORP directive has harmonized investment regulations in the EU, future pension legislation is unlikely to lead to convergence between Beveridgean and Bismarckian pension systems. This is because national social policy still has priority in determining the weight each pension pillar plays in providing retirement income. Whereas the German state has increasingly provided incentives for individuals to take up occupational pension plans, the British state has so far successfully fended off pressure to adopt a more stringent regulatory approach by framing instances of pension fraud as the result of a few firms’ recklessness. Consequently, government reactions to large-scale pension scandals – i.e. tax-financed compensation for victims of pension theft – are ill-suited for combating corporate pension fraud. Strikingly, this failure does not seem to translate into a political threat for the government as victims tend to blame their individual employers for pension losses, not inadequate oversight by the government (Clarke, et al., 2004: 86 ff.). How well do British politicians fare in communicating and explaining EU pension legislation to domestic audiences? In the existing literature, Britain’s general problems in adapting to EU decisions have been correctly attributed to the lack of a discourse capable of persuading voters that change is necessary and legitimate (V. Schmidt, 2006a).

7.6 Conclusion

147

However, as the above analysis has shown, we find the opposite in the area of occupational pension regulation: Britain has almost no problems justifying EU legislation to domestic audiences, but immense difficulties legitimizing domestic pension policies because they tend to be onerous for the pension industry. Given the low levels of take-up of SERPS and the strong reliance of British pensioners on private-sector schemes, politicians are desperately trying to discourage employers from closing generous final-salary schemes to new hires. An important part of this strategy is to deflect Bismarckian attempts to design a single pension market with too many restrictions. Future research may productively address if and how ideas about “good” pension policies change over time, at both the EU and national levels. Existing research has shown that the diffusion of best practices tends to take place via institutionalized mechanisms, such as European expert groups (e.g. the European Federation for Retirement Provision), parliamentary forums (e.g. European Parliamentary Financial Services Forum), or soft law mechanisms (e.g. the Open Method of Coordination). Yet informal mechanisms, such as observing pension policy in other member states, might also be instructive. In this context, the regulatory path the newly sovereign Central European member states have chosen is particularly interesting. Contrary to widespread expectations that these transition economies would follow laissez-faire Anglo-Saxon rules to exploit regulatory arbitrage, new member states like Poland, the Czech Republic, and Hungary have instead opted for stricter regulations, mostly from considerations of best practice.16 A shifting EU-level discourse on best practice could well have an impact on British ideas of “good” pension fund supervision. 16

European Pensions and Investment News, “Anglo-Saxon Hands-Off Approach Losing out to Stricter Regime”, August 13, 2007.

8

Conclusions

Our analysis implies that standard accounts portraying the EU as a regulatory polity with little discretion over pension policy issues are wrong. The constraining impact of EU treaties on national pension policy choices and the creation of a single European pension market demonstrate that the scope of EU regulations is much broader than traditional accounts on the EU have acknowledged. Far from being limited to apolitical areas of market creation and maintenance, the European Union crucially shapes workplace pension regulations, with highly controversial implications for national labor relations and capital flows. European legal constraints have greatly reduced the capacity of national governments to influence workplace pension regulations, notably in the areas of investment rules, waiting and vesting periods, treatment of foreign pension funds, and regulatory supervision. Market integration and the management of cross-border externalities may drive EU pension legislation, but the reshaping of national pension policy choices is the outcome. The findings of our inquiry have been discussed at length in the book. We provided evidence that pension reforms in the member states are not just a consequence of domestic factors, such as demographic aging, but also the result of international factors, in particular the constraining effects of the Maastricht Treaty. Using an event history set-up, we systematically tested whether the timing of pension reforms in Western Europe ensued from the diffusion of policy ideas, domestic pressures, or common shocks. Standard accounts have concentrated only on individual explanations for pension reform. Our findings suggest that the macroeconomic constraints of the Maastricht Treaty effectively acted as an external shock that pushed the member states to overhaul their costly public pension systems. While the Maastricht Treaty prompted European governments to reform their pension systems, national pension policy discourses, in turn, affected member states’ willingness to cooperate on pension 148

Conclusions

149

legislation at the EU level. Employing a formal model, we have posited a multi-level polity with strategic political actors who are intent on benefiting from a European single pension market while keeping the necessary adjustment costs to a minimum. In such a model, EU institutions emerge from the strategic interaction of those agents, shaped by the information they share about their domestic reform capabilities. Whether the released information is considered credible, however, critically depends on the pension policy discourse in the member states. A reform-oriented pension policy discourse, which generates domestic opposition, serves as a costly signal to other member states. A status quo oriented policy discourse, by contrast, indicates “cheap talk” and is therefore not credible. Governments’ failure to send credible signals at the EU level increases the likelihood of bargaining breakdown. Thus, our theoretical contribution is an argument about the circumstances under which EU negotiations may break down. Because existing studies on EU integration efforts have overwhelmingly focused on formal rules and decision-making processes, bargaining breakdown is never observed in the stages before the official voting takes place. Consequently, the literature on formal rules and legalities is systematically biased towards “successful” cases of international institution building. Our analysis, however, suggests that informal communication and signaling processes between member states are more consequential than the narrow legalities. In addition to informal signaling processes, the design of the single pension market was also heavily influenced by the behavior of the Commission. As the only institution in the EU that can formally propose legislation, the Commission can play a powerful role in securing member states’ commitment to EU covenants. Yet, it is rarely clear or agreed upon what constitutes efficient agenda setting. Using primary document analysis as well as interviews with Commission officials and pension industry representatives in Germany and Britain, we find that the Commission’s ability to act as honest broker between the member states increases when it can trim a crowded agenda, set aside internal rivalries, and enhance the quality of information flows regarding workplace pension reform options and limits in other countries. In turn, a lopsided representation of supply-side interests, rivalries between Directorates General, and failure to restrict the menu of policy options will diminish the Commission’s chances of garnering support for novel EU institutions.

150

Conclusions

To strengthen the case for the theoretical hypotheses, we have accumulated several layers of empirical evidence on pension policy developments in German and Britain. Both cases are critical for a number of reasons. First, Germany and Britain represent the ideal types of Bismarckian insurance cultures and Beveridgean pension fund systems, respectively. Secondly, they are among the politically and economically most powerful countries in the union. Consequently, pension market integration is unlikely to occur without the support of these governments. On the one hand, we examined the pension policy discourse in Germany and Britain between the early 1980s and 2007, highlighting economic interests, institutional frameworks, and the framing of pension policy ideas. On the other hand, we compared national pension policy debates with actors’ stance on EU pension legislation. Drawing on interviews with government officials in Germany and Britain as well as Commission officials in Brussels, we explained variation in German and British preferences toward a single pension market with the nature of the prevalent pension policy discourse in each country. In terms of research design, we argued that pension market integration needs to be explained by a methodologically plural framework that accounts for economic interests, historical trajectories, and national templates of generational and distributive justice. Each approach by itself would not be sufficient for explaining different levels of pension market integration over time. Interest-based rational-choice approaches are useful for identifying actors’ stakes in pension market integration, but they do not predict which of the competing interests should be expected to prevail. Historical institutional accounts tell us how past policy trajectories have influenced pension institutions, but structures alone do not determine behavior. Discourse analysis explicates how pension policy interests were defined and legitimized during the observed period, but ideas-based approaches sometimes underspecify the circumstances under which political officials manage to get their preferences implemented. When combined, however, these approaches offset each others’ limitations. Embedding an ideas-based discourse analysis within a rationalchoice/historical-institutionalist framework has two advantages. First, this link strengthens our institutionalist account by showing how informal information exchanges shape formal pension policy outcomes. In the same vein, the investigation of historical trajectories of workplace pension institutions allows us to assess the compatibility of

Conclusions

151

EU-mandated change and national institutional frameworks. In contrast to conventional “goodness of fit” studies that portray Germany as being always in favor of EU integration while Britain is always opposed, we found that EU pension policies fit well with Britain’s pension fund culture, whereas the long-term nature of German capital flows and labor relations was diametrically opposed to the Commission’s goal of promoting pension portability. We therefore need to be cautious regarding the generalizability of “goodness of fit” studies and instead examine the individual policy areas that are subject to market-making or market-correcting initiatives. Furthermore, we avoid the risk of overstating the importance of ideas. This framework allows us to formulate testable hypotheses about the conditions under which domestic discourse is causally related to preference articulation at the EU level and when it is not. The analysis of the German stance on EU pension policies over three decades has demonstrated that German preferences were more likely to be accommodated in EU pension directives when Germany could credibly demonstrate that its domestic win-set was indeed narrow. While all German governments tried to revise EU pension directives in their favor by invoking “domestic constraints” arguments, only the reformoriented Schroder administration was able to twist EU agreements ¨ (i.e. the IORP directive) in its favor. In 2001, the Social Democratic chancellor imposed an economically efficient pension reform without much support from the Left wing of his party and the labor unions, his core constituency that had elected him in 1998. Since he pushed through a major pension reform without paying much attention to the coordinative dimension of discourse, which is particularly important in Germany, this course of action constituted a costly, and therefore credible, signal at the EU level. Consequently, Schroder’s requests for ¨ modifications of EU pension legislation were accommodated. Helmut Kohl, by contrast, who had previously tried to exploit the domestic constraint logic to get a more favorable EU directive, was unsuccessful. This is because Kohl’s status quo oriented pension policy rendered his narrow win-set argument unconvincing. While Germany had to send credible signals in order to get accommodated at the EU level, the congruence of the British welfare finance regime with EU pension policies enabled Britain to influence EU directives in its favor. Therefore, what needed to be explained in this case was not a mechanism that made signaling credible, but how the push

152

Conclusions

and pull dynamics of British support for pension market integration on the one hand and the failure of the party system adequately to represent consumer interests on the other influenced EU pension directives. Unsurprisingly, the biggest pension policy challenges that remain in Britain include the chronic underfunding of pension funds, the lack of transparency for consumers, the closing of generous final-salary schemes to new hires, and the tendency of firms to wind up their pension schemes with several million pound deficits. In contrast to Germany, the British communicative sphere of discourse is more important than the coordinative one. In a majoritarian democracy, only one set of interests wins out, regardless of the government’s partisan complexion. Even though each year dozens of British employers wind up their pension schemes with million pound deficits, leaving thousands of workers with significantly reduced occupational pensions, Britain still serves as a role model for countries with an underdeveloped occupational pension sector. The goal of the Commission is to emulate those aspects of the British pension system that in the 1980s had earned it the label “envy of the world,” while at the same time safeguarding against the drawbacks – in particular, the inefficient pension regulator. The purpose of the case studies was to show that a clear specification of the circumstances under which domestic discourse has a causal impact on EU pension regulations is necessary to understand pension market integration over time. Our analysis has important implications for democratic accountability in the EU and for how we study European integration. Given the distributional impact of pension policies on the life-course risks of domestic constituencies, political incumbents may not escape electoral backlash, no matter whether pension reforms reflect political preferences or a reluctant response to external shocks. This means that domestic discourse matters greatly for the adoption of EU legislation, more than traditional studies on European Union politics have recognized. While liberal intergovernmentalist approaches have argued that EU institutions mirror the national interests of the most powerful countries in the EU, our analysis has shown that national discourse, which consists of a coordinative and a communicative dimension, is analytically more useful than empty conceptualizations of a fixed “national interest.” While this book has concentrated on member states’ pension policy responses to the Maastricht Treaty “shock,” the dynamics

Conclusions

153

outlined here can be applied to other policy areas. The Eurozone crisis is another case that highlights the importance of credible signaling for the adoption of new EU institutions, such as the European Stability Mechanism and banking union. Undoubtedly, the Eurozone crisis has reminded Europeans of the interconnected nature of their economies, and has dramatically underlined the importance of EU legislation for distributive outcomes in the member states. The German government’s contribution to bailouts of, and financial guarantees to, troubled member states raised doubts about the fairness of Germany’s own austerity program. Domestic dissatisfaction with Germany’s role as creditor-in-chief was highly visible and politically risky for political incumbents. Domestic dissatisfaction, however, enabled German chancellor Merkel to signal credibly to other member states that her room for maneuver was severely constrained. Although many Eurozone states and EU officials criticized her long hesitation before offering any assistance to troubled states, as well as her insistence on stringent conditions of financial guarantees, Merkel’s ability credibly to communicate the magnitude of domestic hostilities to financial assistance programs enabled her to dictate the terms of the European Stability Mechanism and banking union. These developments are consistent with our finding that a government’s credibility translates into bargaining power at the EU level. External shocks, such as the global financial crisis, have also affected the reputation of workplace pensions more generally. Although there is no doubt that governments’ over-reliance on public pension systems in the past was not a viable strategy, defined-contribution workplace pensions do not look so attractive any more in the aftermath of the financial turmoil. Individuals who rely on their workplace pensions as the primary source of retirement income have awoken to the fact that inadequate financial regulations can have devastating consequences. Many who have worked hard for a lifetime, and had looked forward to imminent retirement, face the remaining days of their lives in destitution and insecurity. Germany’s unfunded book reserve pensions, by contrast, have remained safe. Thus, the current era of economic insecurity may prompt a revision of national pension policy debates. As Chapter 6 has shown, Germany fought hard for the exemption of book reserve pensions from the scope of EU directives. Commission officials had insisted on including book reserve pensions under the scope of the directive better to “protect” consumers from corporate bankruptcies.

154

Conclusions

However, the financial turmoil has primarily exposed the vulnerability of defined-contribution workplace pension systems to external shocks. In any case, to gauge which pension regimes are more robust to economic shocks, we first need to understand the institutions, ideas, and political coalitions that underpin them, as well as the causal pathway of pension reform. While the single pension market has increased the cross-border activity of pension funds, it is far from complete. National disparities in the fiscal treatment of pension products, as well as their complexity and specificity, are still serious obstacles to pension portability and therefore labor mobility in the EU. Reviews of the IORP directive (European Commission, 2010; European Insurance and Occupational Pensions Authority, 2011, 2012) indicate that the development of IORPs operating cross-border is still in its infancy. The slow takeoff of pan-European pension funds suggests that more must be done to eliminate tax discrimination and to improve the efficiency and safety of pension schemes. The set-up procedures for IORPs need to be simplified, and portability for selected occupational categories must be further encouraged. The Council’s decision to designate 2012 as European Year for Active Ageing reflects the EU’s commitment to deepening pension market integration. But the completion of the single pension market is likely to be an uphill battle. While nobody denies that European citizens need an adequate and sustainable retirement income, there is still considerable disagreement concerning the type of regulations that can ensure stability, growth, and protection. The future shape of Europe’s single pension market will depend on member states’ national pension policy discourses, governments’ ability to send credible signals, and the Commission’s agenda setting strategy, as explicated in this book. However, it also depends on how the European Court of Justice will interpret the transposition of EU pension directives in the member states, which we have not explored. According to Max Weber, “a fundamental fact of all history” is that the results of an action regularly stand in a completely inadequate and even paradoxical relation to its original meaning (Weber, 1978: 148–149). Nowhere are unintended consequences more prevalent than in the European Union, where the multi-layered system of decision-making occasionally produces outcomes that stand in opposition to member states’ original intentions. Even former Commission president Jos´e Manuel Barroso

Conclusions

155

acknowledged that “it is almost a miracle when we arrive at a final outcome or resolution that is exactly as it was originally planned.”1 Discretionary scope embedded in EU pension directives could lead to the transposition of pension policies that nobody desired. The directives that created the single pension market leave substantial room for national autonomy. For example, there is neither a clear definition of the “limited time” that EU pension funds are allowed to be underfunded, nor is it obvious what precisely constitutes a “prudential” investment, and how these rules will be enforced. Disagreements over the precise meanings of these stipulations are bound to reach the European Court of Justice, who may exploit ambiguous wording in order to push through far-reaching changes that could be diametrically opposed to governments’ original predilections. This raises the specter of judicial activism. Conversely, “ministerial drift” implies that national governments may implement EU directives in ways that differ from the EU compromise position. This risk emerges when the preferences of the national minister responsible for implementing a directive diverge from those of her peers in government (Franchino and Hoyland, 2009). Whether judicial activism or ministerial drift poses a bigger problem for democratic accountability in the EU is an open question. It remains, however, that the way in which pension directives will be transposed in the member states constitutes a critical test for the EU’s desire to create deeper capital markets and improve labor mobility. Because the member states have different ideas about the appropriate degree of state involvement and consumer protection levels in regulating workplace pensions, a divergent implementation of EU pension legislation may undermine the effectiveness of the single pension market. If the EU chooses protectionism, it will be Europe’s pension savers who ultimately pay the cost. 1

European Union Studies Association, Newsletter, 20(4) (2007): 4.

Bibliography

¨ betriebliche aba. 2006. Stellungnahme der Arbeitsgemeinschaft fur ¨ eine Richtlinie des Altersversorgung e.V. zum Vorschlag fur ¨ Europaischen Parlaments und des Rates zur Verbesserung der Porta¨ von Zusatzrentenanspruchen. ¨ bilitat Abdelal, Rawi. 2007. Capital Rules: The Construction of Global Finance. Cambridge, MA: Harvard University Press. Achen, Christopher. 2006. “Institutional Realism and Bargaining Models.” In The European Union Decides, ed. Robert Thomson, Frans N. Stokman, Christopher H. Achen, and Thomas Konig. Cambridge University ¨ Press, pp. 86–123. Ahrend, Peter. 1996. “Pension Financial Security in Germany.” In Securing Employer-Based Pensions: An International Perspective, ed. Zvi Bodie, Olivia S. Mitchell, and John A. Turner. Philadelphia: University of Pennsylvania Press. Alesina, Alberto. 1989. “Politics and Business Cycles in Industrial Democracies.” Economic Policy 8: 55–98. Allen, Christopher S. 2004. “Ideas, Institutions, and the Exhaustion of Modell Deutschland?” German Law Journal 5: 1133–1154. Anderson, Karen M. 2007. “The Netherlands: Political Competition in a Proportional System.” In The Handbook of West European Pension Politics, ed. Ellen M. Immergut, Karen M. Anderson, and Isabelle Schulze. Oxford University Press. Anderson, Karen M., and Traute Meyer. 2006. “New Social Risks and Pension Reform in Germany and Sweden: The Politics of Pension Rights for Childcare.” In The Politics of Postindustrial Welfare States, ed. Klaus Armingeon and Giuliano Bonoli. Oxford: Routledge, pp. 171– 191. Aoki, Masahiko. 2001. Toward a Comparative Institutional Analysis. Comparative Institutional Analysis, 2. Cambridge, MA: MIT Press. Arregui, Javier, Frans Stokman, and Robert Thomson. 2004. “Bargaining in the European Union and Shifts in Actors’ Policy Positions.” European Union Politics 5: 47–72.

156

Bibliography

157

Austen-Smith, David, and Timothy J. Feddersen. 2006. “Deliberation, Preference Uncertainty, and Voting Rules.” American Political Science Review 100: 209–217. Bachrach, Peter, and Morton Baratz. 1962. “Two Faces of Power.” American Political Science Review 56: 947–952. BDA. 2006. Stellungnahme der Bundesvereinigung der Deutschen Arbeitge¨ ¨ eine Richtlinie des Europaischen ¨ berverbande zum Vorschlag fur Parla¨ von Zusatzrentements und des Rates zur Verbesserung der Portabilitat ¨ nanspruchen. Beach, Derek. 2005. The Dynamics of European Integration: Why and When EU Institutions Matter. London: Palgrave Macmillan. Beck, Nathaniel, Jonathan N. Katz, and Richard Tucker. 1998. “Taking Time-Series Seriously: Time-Series Cross-Section Analysis with a Binary Dependent Variable.” American Journal of Political Science 42: 1260– 1288. Beyer, Jurgen, and Martin Hopner. 2003. “The Disintegration of Organ¨ ¨ ised Capitalism: German Corporate Governance in the 1990s.” West European Politics 26: 179–198. ¨ ¨ die Biedenkopf, Kurt. 1981. “Vorwort.” In Sicherheit im Alter: Pladoyer fur Weiterentwicklung des Rentensystems, ed. Meinhard Miegel. Stuttgart: Verlag Bonn Aktuell, pp. 247–274. Blake, David. 2003. Pension Schemes and Pension Funds in the United Kingdom. 2nd edn. Oxford University Press. 2005. “Overregulating your Pension Out of Existence: The Long-Term Consequences of British Pension Policy over the Last 30 Years.” Discussion Paper PI-0616, Pensions Institute, Cass Business School, London, June 15, 2005. Blyth, Marc. 2003. “Structures Do Not Come with an Instruction Sheet: Interests, Ideas, and Progress in Political Science.” Perspectives on Politics 1: 695–706. Bodie, Zvi, Olivia S. Mitchell, and Andrew J. Turner. 1996. Securing Employer-Based Pensions: An International Perspective. Philadelphia: University of Pennsylvania Press. Bonoli, Giuliano. 2006. “New Social Risks and the Politics of PostIndustrial Social Policies.” In The Politics of Postindustrial Welfare States, ed. Klaus Armingeon and Giuliano Bonoli. Oxford: Routledge, pp. 3–26. Bonoli, Giuliano, and Bruno Palier. 2008. “When Past Reforms Open New Opportunities: Comparing Old Age Insurance Reforms in Bismarckian Welfare Systems.” In Reforming the Bismarckian Welfare Systems, ed. Bruno Palier and Claude Martin. Oxford: Blackwell Publishing, pp. 21–39.

158

Bibliography

Bonoli, Giuliano, and Toshimitsu Shinkawa. 2005. Ageing and Pension reform around the World: Evidence from Eleven Countries. Cheltenham: Edward Elgar. Box-Steffensmeier, Janet M., and Bradford S. Jones. 2004. Event History Modeling: A Guide for Social Scientists. Cambridge University Press. Brinegar, Adam, Seth K. Jolly, and Herbert Kitschelt. 2004. “Varieties of Capitalism and Political Divides over European Integration.” In European Integration and Political Conflict, ed. Gary Marks and Marco Steenbergen. Cambridge University Press. Brooks, Sarah M. 2005. “Interdependent and Domestic Foundations of Policy Change: The Diffusion of Pension Privatization Around the World.” International Studies Quarterly 49: 273–294. 2007. “When Does Diffusion Matter? Explaining the Spread of Structural Pension Reforms across Nations.” Journal of Politics 69: 701–715. Buiter, Willem H. 2006. “The ‘Sense and Nonsense of Maastricht’ Revisited: What Have we Learnt about Stabilization in EMU?” Journal of Common Market Studies 44: 687–710. Callaghan, Helen, and Martin Hopner. 2005. “European Integration and ¨ the Clash of Capitalisms: Political Cleavages over Takeover Liberalization.” Comparative European Politics 3: 307–332. Carter, David, and Curtis S. Signorino. 2010. “Back to the Future: Modeling Time Dependence in Binary Data.” Political Analysis 18: 271–292. Casey, Bernhard H. 2012. “The Implications of the Economic Crisis for Pensions and Pension Policy in Europe.” Global Social Policy 12: 246– 265. Christiansen, Thomas, and Simona Piattoni. 2003. Informal Governance in the European Union. Cheltenham: Edward Elgar. Clark, Gordon L., and Ashby H. B. Monk. 2006. “The ‘Crisis’ in Defined Benefit Corporate Pension Liabilities.” Pensions 12: 43–54. Clarke, Harold D., David Sanders, Marianne C. Stewart, and Paul F. Whiteley. 2004. Political Choice in Britain. Oxford University Press. Coates, David (ed.) 2005. Varieties of Capitalism, Varieties of Approaches. New York: Palgrave Macmillan. Commission of the European Communities. 2005. Proposal for a Directive of the European Parliament and the Council on the Improvement of Portability of Supplementary Pension Rights. SEC (2005) 1293. 20 October 2005. Brussels. Cowles, Maria Green, James A. Caporaso, and Thomas Risse-Kappen. 2001. Transforming Europe: Europeanization and Domestic Change. Cornell Studies in Political Economy. Ithaca, NY: Cornell University Press. Cox, Robert Henry. 2001. “Why Welfare Reform Happened in Denmark and the Netherlands, but Not in Germany.” World Politics 53: 463–498.

Bibliography

159

Cram, Laura. 1997. Policymaking in the European Union: Conceptual Lenses and the Integration Process. London: Routledge. Crombez, Christophe. 1997. “The Co-Decision Procedure in the European Union.” Legislative Studies Quarterly 22: 97–119. Cuevas, Alfredo, Maria Gonzalez, Davide Lombardo, and Arnoldo LopezMarmolejo. 2008. “Pension Privatization and Country Risk.” IMF Working Paper WP-08-195. Curtin, Deirdre. 1993. “The Constitutional Structure of the Union: A Europe of Bits and Pieces.” Common Market Law Review 30: 17– 69. Davidsson, Johan B., and Patrick Emmenegger. 2012. “Insider–Outsider Dynamics and the Reform of Job Security Legislation.” In The Politics of the New Welfare State, ed. Giuliano Bonoli and David Natali. Oxford University Press, pp. 206–232. Davis, E. Philip. 1999. “Institutionalisation and EMU: Implications for European Financial Markets.” International Finance 2: 33–61. 2000. “The Regulation of Private Pensions: A Case Study of the UK.” Pensions Institute, Birkbeck College, University of London. Deeg, Richard. 2005. “Change from Within: German and Italian Finance in the 1990s.” In Beyond Continuity: Institutional Change in Advanced Political Economies, ed. Wolfgang Streeck and Kathleen Thelen. Oxford University Press. Deeg, Richard, and Susanne Lutz. 2000. “Internationalization and Fiscal ¨ Federalism: The United States and Germany at Crossroads?” Comparative Political Studies 33: 374–405. Delors, Jacques. 1985. “Completing the Internal Market: White Paper from the Commission to the European Council (Milan, 28–29 June), COM(85)310, June 1985.” Technical report. Commission of the European Communities, Brussels. Department for Work and Pensions. 2002. “Simplicity, Security and Choice: Working and Saving for Retirement.” 2004. “Insolvent Pension Wind-Ups: New Research into Numbers Affected.” June 30, 2004. 2006a. “Explanatory Memorandum to the Occupational Pension Schemes (Winding-Up Procedure Requirement) Regulations.” No. 1733. 2006b. “Government Response to the Consultation on a Draft European Directive.” July 2006. 2006c. “Response to the Report by the Parliamentary Ombudsman ‘Trusting in the Pensions Promise’.” June 2006. Department of Social Security. 1998a. “New Ambitions for our Country: A New Contract Welfare.” 1998b. “A New Contract for Welfare: Partnership in Pensions.”

160

Bibliography

Deutsch, Klaus Gunter. 2002. “A Single Market for Occupational Pensions.” ¨ Technical report. Deutsche Bank Research, Frankfurt am Main. Deutscher Bundestag. 2006. Stellungnahme der Fraktionen CDU/CSU und ¨ eine Richtlinie des Europaischen ¨ SPD zum Vorschlag fur Parlaments ¨ von Zusatzrenteund des Rates zur Verbesserung der Portabilitat ¨ nanspruchen. Disney, Richard. 2003. “Public Pension Reform in Europe: Policies, Prospects and Evaluation.” World Economy 26: 1425–1445. Downs, Anthony. 1957. An Economic Theory of Democracy. New York: Harper. Ebbinghaus, Bernhard. 2011. The Varities of Pension Governance: Pension Privatization in Europe. Oxford University Press. Eckardt, Martina. 2005. “The Open Method of Coordination on Pensions: An Economic Analysis of its Effects on Pension Reforms.” Journal of European Social Policy 15: 247–267. EFRP. 2001. The European Federation for Retirement Provision at 20 Years. London: IPE International Publishers Ltd. 2007. The European Federation for Retirement Provision: Annual Report 2006. London: IPE International Publishers Ltd. Eijffinger, Sylvester C. W., and Jakob De Haan. 2000. European Monetary and Fiscal Policy. Oxford University Press. Emmenegger, Patrick, Silja Hausermann, Bruno Palier, and Martin Seeleib¨ Kaiser. 2012. The Age of Dualization. New York: Oxford University Press. Esping-Andersen, Gosta. 1990. The Three Worlds of Welfare Capitalism. ¨ Cambridge: Polity Press. 1996. “Conclusion: Occupational Welfare in the Social Policy Nexus.” In The Privatization of Social Policy?: Occupational Welfare and the Welfare State in America, Scandinavia, and Japan, ed. Michael Shalev. Houndmills and New York: Macmillan/St. Martin’s Press, pp. 327– 338. Esposito, Mariachiara, and David Mum. 2004. “Pension Funds in the European Debate and Within the Multilevel Decision-Making of the European Union.” European Trade Union Institute (ETUI). DWP 2004.02.02. Estevez-Abe, Margarita. 2002. “Reforms of German and Japanese Pension Systems in Comparative Perspective.” Paper presented at the annual meeting of the American Political Science Association, Boston, August 31–September 2, 2002. ETUC. 2003. “Trade Union Statement to the World Social Forum (Porto Allegre) and the World Economic Forum (Davos).” In Our Priorities: ETUC Resolutions, ed. European Trade Union Confederation. Brussels.

Bibliography

161

Eurobarometer. 2002. Standard Eurobarometer 56. European Commission. European Association of Public Enterprises and Employers. 2007. “Intervention of Mr. Jacques Fournier.” In 20 Years of European Social Dialogue, ed. European Commission. Brussels: DG Employment, Social Affairs, and Equal Opportunities, pp. 19–20. European Commission. 2010. Green Paper on the Future of Pensions in the EU. COM(2010)365/3. Brussels. 2012a. “The 2012 Ageing Report.” 2012b. White Paper: Agenda for Adequate, Safe, and Sustainable Pensions. COM(2012) 55 final. Brussels. European Insurance and Occupational Pensions Authority. 2011. “2011 Report on Market Developments.” EIOPA-BoS-11/023. July 14, 2011. 2012. “EIOPA’s Advice to the European Commission on the Review of the IORP Directive 2003/41/EC.” EIOPA-BoS-12/015. February 15, 2012. European Parliament/Council Directive 2003/41/EC, of 3 June 2003. “Directive on the Activities and Supervision of Institutions for Occupational Retirement Provision (IORP).” Farrell, Henry, and Adrienne H´eritier. 2003. “Formal and Informal Institutions under Codecision: Continuous Constitution-Building in Europe.” Governance 16: 577–600. Fawcett, Helen. 2002. “Pension Reform in the UK: Recasting the Public/Private Mix in Pension Provision 1997–2000.” European Consortium for Political Research 21: 1–26. Fearon, James D. 1995. “Rationalist Explanations for War.” International Organization 49: 379–414. Federal Trust for Education and Research. 2010. Pension Compendium: Pension Reform in Europe. London: Federal Trust for Education and Research. Ferrera, Maurizio. 2005. The Boundaries of Welfare: European Integration and the New Spatial Politics of Social Protection. Oxford University Press. Fiess, Norbert M. 2003. “Pension Reform or Pension Default? A Note on Pension Reform and Country Risk.” Technical report. Washington, DC: World Bank. Fioretos, Orfeo. 2001. “The Domestic Sources of Multilateral Preferences: Varieties of Capitalism in the European Community.” In Varieties of Capitalism: The Institutional Foundations of Comparative Advantage, ed. Peter A. Hall and David Soskice. Oxford University Press. Franchino, Fabio, and Bjorn ¨ Hoyland. 2009. “Legislative Involvement in Parliamentary Systems: Opportunities, Conflict, and Institutional Constraints.” American Political Science Review 103: 607–621.

162

Bibliography

Franzese, Robert J., and Jude C. Hays. 2008. “Interdependence in Comparative Politics.” Comparative Political Studies 41: 742–780. Frericks, Patricia. 2012. “Funded Pensions and their Implications for Women and Migrant Workers.” Global Social Policy 12: 342–344. Garrett, Geoffrey. 1998. Partisan Politics in the Global Economy. Cambridge Studies in Comparative Politics. Cambridge University Press. George, Stephen. 1990. An Awkward Partner: Britain in the European Community. Oxford University Press. Gibbons, Robert. 1992. Game Theory for Applied Economists. Princeton, NJ: Princeton University Press. Gilardi, Fabrizio. 2010. “Who Learns from What in Policy Diffusion Processes.” American Journal of Political Science 54: 650–666. Gilardi, Fabrizio, Katharina Fuglister, and St´ephane Luyet. 2009. “Learning ¨ from Others: The Diffusion of Hospital Financing Reforms in OECD Countries.” Comparative Political Studies 42: 549–573. Guardiancich, Igor, and David Natali. 2012. “The Cross-Border Portability of Supplementary Pensions: Lessons from the European Union.” Global Social Policy 12: 300–315. Hacker, Jacob S. 2002. The Divided Welfare State: The Battle over Public and Private Social Benefits in the United States. New York: Cambridge University Press. Hall, Peter A. 1993. “Policy Paradigms, Social Learning and the State: The Case of Economic Policy-Making in Britain.” Comparative Political Studies 25: 275–296. Hall, Peter A., and David W. Soskice. 2001. Varieties of Capitalism: The Institutional Foundations of Comparative Advantage. Oxford University Press. Hannah, Leslie. 1986. Inventing Retirement: The Development of Occupational Pensions in Britain. Cambridge University Press. Hartlapp, Miriam. 2008. “Intra-Kommissionsdynamik im Policy-Making: EU-Politiken angesichts des demographischen Wandels.” In Die ¨ Europaische Union: Governance und Policy-Making, ed. Ingeborg Tommel. Wiesbaden: VS Verlag fur ¨ ¨ Sozialwissenschaften. Hassel, Anke. 2004. “The Crumbling Pillars of Social Partnership.” West European Politics 26: 103–124. ¨ 2007. “Gewerkschaften.” In Interessenverbande in Deutschland, ed. Ulrich Willems and Thomas von Winter. Opladen: Leske & Budrich. Hassel, Anke, and Bernhard Ebbinghaus. 2000. “Striking Deals: Concertation in the Reform of Continental European Welfare States.” Journal of European Public Policy 7: 44–62. Hausermann, Silja. 2010. The Politics of Welfare State Reform in Conti¨ nental Europe: Modernization in Hard Times. Cambridge University Press.

Bibliography

163

2012. “The Politics of Old and New Social Policies.” In The Politics of the New Welfare State, ed. Giuliano Bonoli and David Natali. Oxford University Press, pp. 111–134. Haverland, Markus. 2004. “European Single Market vs. National Social policy: The Crucial Case of Occupational Pensions.” Paper presented at the annual conference of the Dutch and Flemish Political Science Associations, Antwerp, May 27–28, 2004. 2007. “When the Welfare State Meets the Regulatory State: EU Occupational Pension Policy.” Journal of European Public Policy 14: 886–904. Heclo, Hugh. 1974. Modern Social Politics in Britain and Sweden: From Relief to Income Maintenance. Yale Studies in Political Science, 25. New Haven, CT: Yale University Press. Hennessy, Alexandra. 2008. “Economic Interests and the Construction of a European Single Pension Market.” British Journal of Politics and International Relations 10: 105–128. Hering, Martin. 2007. “The New Approach to Welfare State Reform in German Social Democracy.” Paper prepared for the “Third Ways in Europe” workshop, European Consortium for Political Research, Grenoble, April 6–11, 2001. H´eritier, Adrienne, Dieter Kerwer, Christoph Knill, Dirk Lehmkuhl, Michael Teutsch, and Anne-Cecile Douillet. 2001. Differential Europe: The European Union Impact on National Policymaking. Lanham, MD: Rowman & Littlefield. Hix, Simon, and Christopher Lord. 1997. Political Parties in the European Union. European Union Series. New York: St. Martin’s Press. Hix, Simon, Abdul G. Noury, and Gerard Roland. 2007. Democratic Politics in the European Parliament. Themes in European Governance. Cambridge University Press. Holzmann, Robert. 2000. “The World Bank Approach to Pension Reform.” International Security Review 53: 11–34. 2006. “Toward a Coordinated Pension System in Europe: Rationale and Potential Structure.” In Pension Reform: Issues and Prospects for NonFinancial Defined Contribution (NDC) Schemes, ed. Robert Holzmann and Edward Palmer. Washington, DC: World Bank, pp. 225–292. Holzmann, Robert, Landis MacKellar, and Michal Rutkowski. 2003. “Accelerating the European Pension Reform Agenda: Need, Progress, and Conceptual Underpinnings.” In Pension Reform in Europe: Process and Progress, ed. Robert Holzmann, Mitchell Orenstein, and Michal Rutkowski. Washington, DC: World Bank, pp. 1–46. Hooghe, Liesbeth. 2001. The European Commission and the Integration of Europe: Images of Government. Cambridge University Press. 2005. The European Commission and the Integration of Europe. Cambridge University Press.

164

Bibliography

Hooghe, Liesbeth, and Gary Marks. 2001. Multi-Level Governance and European Integration. Lanham, MD: Rowman & Littlefield. Hopner, Martin, and Gregory Jackson. 2002. “The German System of Cor¨ porate Governance between Persistence and Convergence. Reply to ¨ ¨ Soziologie Thomas Heinze in KZfSS 4/2001.” Kolner Zeitschrift fur und Sozialpsychologie 54: 362–368. Hopner, Martin, and Armin Schafer. 2006. “A New Phase of European ¨ ¨ Integration: How the EU Threatens Organized Capitalism.” In Paper presented at the American Political Science Association, August 31– September 3, 2006. Philadelphia, PA. Immergut, Ellen M., Karen M. Anderson, and Isabelle Schulze. 2007. The Handbook of West European Pension Politics. Oxford University Press. Iversen, Torben. 1999. Contested Economic Institutions: The Politics of Macroeconomics and Wage Bargaining in Advanced Economics. Cambridge University Press. Iversen, Torben, Jonas Pontusson, and David W. Soskice. 2000. Unions, Employers, and Central Banks: Macroeconomic Coordination and Institutional Change in Social Market Economies. Cambridge Studies in Comparative Politics. Cambridge University Press. Iversen, Torben, and David Soskice. 2001. “An Asset Theory of Social Policy Preferences.” American Political Science Review 95: 875–893. Iversen, Torben, and Anne Wren. 1998. “Equality, Employment, and Budgetary Restraint: The Trilemma of the Service Economy.” World Politics 50: 507–546. Jackson, Gregory. 2003. “Corporate Governance in Germany and Japan: Liberalization Pressures and Responses during the 1990s.” In The End of Diversity? Prospects for German and Japanese Capitalism, ed. Wolfgang Streeck and Kozo Yamamura. Ithaca, NY: Cornell University Press. Jacobs, Alan M. 2011. Governing for the Long Term: Democracy and the Politics of Investment. Cambridge University Press. Jahn, Detlef. 2006. “Globalization as ‘Galton’s Problem’: The Missing Link in the Analysis of Diffusion Patterns in Welfare State Development.” International Organization 60: 401–431. Janky, Bela, and Robert I. Gal. 2007. “Public Opinion on Pension Systems in Europe.” ENEPRI Research Report No. 36. Brussels. Jones, Bryan D., and Frank R. Baumgartner. 2005. The Politics of Attention: How Government Prioritizes Problems. University of Chicago Press. ¨ ¨ Karas, Othmar. 2003. Die Richtlinie uber die Tatigkeiten und die Beaufsichtigung von Einrichtungen der betrieblichen Altersversorgung. IORP Directive.

Bibliography

165

Katzenstein, Peter J. 1997. Tamed Power: Germany in Europe. Ithaca, NY: Cornell University Press. Kayser, Mark A. 2007. “How Domestic is Domestic Politics? Globalization and Elections.” Annual Review of Political Science 10: 341–362. 2009. “Partisan Waves: International Business Cycles and Electoral Choice.” American Journal of Political Science 53: 950–970. Keck, Margaret E., and Kathryn Sikkink. 1998. Activists Beyond Borders: Advocacy Networks in International Politics. Ithaca, NY: Cornell University Press. Keefer, Philip. 2004. “Database of Political Institutions.” Washington, DC: World Bank. www.econ.worldbank.org Kelemen, Daniel R. 2009. “The Strength of Weak States: Adversarial Legalism in the US and the EU.” Paper presented at the EUSA Conference, Los Angeles, April 23–25, 2009. Keohane, Robert O., and Joseph S. Nye. 1989. Power and Interdependence. 2nd edn. Glenview, IL: Scott, Foresman. King, Gary, Michael R. Tomz, and Jason Wittenberg. 2000. “Making the Most of Statistical Analyses: Improving Interpretation and Presentation.” American Journal of Political Science 44: 347–361. Kingdon, John W. 1995. Agendas, Alternatives, and Public Policies. New York: HarperCollins College. Kitschelt, Herbert. 1999. “European Social Democracy between Political Economy and Electoral Competition.” In Continuity and Change in Contemporary Capitalism, ed. Herbert Kitschelt, Peter Lange, Gary Marks, and John D. Stephens. Cambridge University Press, pp. 317– 345. 2001. “Partisan Competition and Welfare Retrenchment: When do Politicians Choose Unpopular Policies?” In The New Politics of the Welfare State, ed. Paul Pierson. Oxford University Press, pp. 265–302. Kleine, Mareike. 2013. “Knowing your Limits: Informal Governance and Judgment in the European Union.” Review of International Organizations 8(2): 245–264. Koremenos, Barbara. 2001. “Loosening the Ties that Bind: A Learning Model of Agreement Flexibility.” International Organization 55: 289– 325. Korpi, Walter, and Joakim Palme. 2003. “New Politics and Class Politics in the Context of Austerity and Globalization: Welfare State Regress in 18 Countries, 1975–95.” American Political Science Review 97: 425–446. Kreppel, Amie. 2002. The European Parliament and Supranational Party System. Cambridge University Press. Kreppel, Amie, and George Tsebelis. 1999. “Coalition Formation in the European Parliament.” Comparative Political Studies 32: 933–966.

166

Bibliography

Lindberg, Leon N., and Stuart A. Scheingold. 1971. Regional Integration: Theory and Research. Cambridge, MA: Harvard University Press. Lynes, Tony. 1997. “The British Case.” In Enterprise and the Welfare State, ed. Martin Rein and Eskil Wadensjo. ¨ Cheltenham: Edward Elgar, pp. 309–351. Mahoney, James. 2000. “Path Dependence in Historical Sociology.” Theory and Society 29: 507–548. Manow, Philip. 2001. “Globalization, Corporate Finance, and Coordinated Capitalism: Pension Finance in Germany and Japan.” MPIfG working paper 01/5. March, James G., and Johan P. Olsen. 1989. Rediscovering Institutions: The Organizational Basis of Politics. New York: Free Press. Mares, Isabela. 2003. The Politics of Social Risk: Business and Welfare State Development. Cambridge Studies in Comparative Politics. Cambridge University Press. 2006. Taxation, Wage Bargaining and Unemployment. Cambridge Studies in Comparative Politics. Cambridge University Press. Martin, Cathie Jo, and Duane Swank. 2004. “Does the Organization of Capital Matter? Employers and Active Labor Market Policy at the National and Firm Levels.” American Political Science Review 98: 593–611. Martin, Cathie Jo, and Kathleen Thelen. 2007. “The State and Coordinated Capitalism: Contributions of the Public Sector to Social Solidarity in Postindustrial Societies.” World Politics 60: 1–36. Meyer, Christoph. 1999. “Political Legitimacy and the Invisibility of Politics: Exploring the European Union’s Communication Deficit.” Journal of Common Market Studies 37: 617–639. Midtbo, Tor. 1998. “The Open Politiconomy: A Dynamic Analysis of Social Democratic Popularity and Economic Policies in Scandinavia.” British Journal of Political Science 28: 93–112. Moravcsik, Andrew. 1998. The Choice for Europe: Social Purpose and State Power from Messina to Maastricht. Cornell Studies in Political Economy. Ithaca, NY: Cornell University Press. 1999. “A New Statecraft? Supranational Entrepreneurs and International Cooperation.” International Organization 53: 267–306. Muller, Katharina. 1999. “Beyond Privatization: Pension Reform in the ¨ Czech Republic and Slovenia.” Journal of European Social Policy 12: 293–306. Myles, John, and Paul Pierson. 2001. “The Comparative Political Economy of Pension Reform.” In The New Politics of the Welfare State, ed. Paul Pierson. Oxford University Press, pp. 305–333. Nanz, Patrizia, and Caroline de la Porte. 2004. “The OMC: A Deliberative– Democratic Mode of Governance? The Cases of Employment and Pensions.” Journal of European Public Policy 11: 267–288.

Bibliography

167

Natali, David. 2008. Pensions in Europe, European Pensions: The Evolution of Pension Policy at National and Supranational Levels. Brussels: Peter Lang. ¨ Nurk, Bettina, and Alexander Schrader. 1995. Von der Pensionsruckstellung ¨ ¨ den deutschen Finanzmarkt. zum Pensionsfonds: Eine Chance fur Frankfurt am Main: DB Research. Offe, Claus. 1991. “Smooth Consolidation in the West German Welfare State.” In Labor Parties in Postindustrial Societies, ed. Frances Fox Piven. Cambridge: Polity Press, pp. 124–146. Ordeshook, Peter C. 1986. Game Theory and Political Theory: An Introduction. Cambridge University Press. Orenstein, Mitchell. 2003. “Mapping the Diffusion of Pension Innovation.” In Pension Reform in Europe: Process and Progress, ed. Robert Holzmann, Mitchell Orenstein, and Michal Rutkowski. Washington, DC: World Bank. 2008. Privatizing Pensions: The Transnational Campaign for Social Security Reform. Princeton, NJ: Princeton University Press. Palier, Bruno, and Kathleen Thelen. 2010. “Institutionalizing Dualism: Complementarities and Change in France and Germany.” Politics and Society 38: 119–148. Peters, Guy. 2001. “Agenda-Setting in the European Union.” In The European Union: Power and Policy-Making, ed. Jeremy Richardson. London and New York: Routledge, pp. 75–94. Pierson, Paul. 1994. Dismantling the Welfare State? Reagan, Thatcher, and the Politics of Retrenchment. Cambridge Studies in Comparative Politics. Cambridge University Press. 2000a. “Increasing Returns, Path Dependence, and the Study of Politics.” American Political Science Review 94: 251–267. 2000b. “Not Just What, but When: Timing and Sequence in Political Processes.” Studies in American Political Development 14: 72–92. Pochet, Philippe. 2003. “Pensions: The European Debate.” In Pension security in the 21st Century: Redrawing the Public–Private Debate, ed. Gordon L. Clark and Noel Whiteside. Oxford University Press, pp. 44–63. Pollack, Mark A. 2003. The Engines of European Integration: Delegation, Agency, and Agenda-Setting in the EU. Oxford University Press. Powell, G. Bingham. 2000. Elections as Instruments of Democracy: Majoritarian and Proportional Visions. New Haven, CT: Yale University Press. Powell, G. Bingham, and Guy Whitten. 1993. “A Cross-National Analysis of Economic Voting: Taking Account of the Political Context.” American Political Science Review 37: 391–414. Princen, Sebastiaan. 2009. Agenda-Setting in the European Union. London: Palgrave Macmillan.

168

Bibliography

Putnam, Robert D. 1988. “Diplomacy and Domestic Politics: The Logic of Two-Level Games.” International Organization 42: 427–460. Rehder, Britta. 2006. “Missing Legitimacy – Company Pacts for Employment and Competitiveness as an Issue of Conflict among Works Coun¨ Soziologie 35: 227–242. cils and Employees.” Zeitschrift fur Ring, Patrick John. 2005. “Security in Pension Provision: A Critical Analysis of UK Government Policy.” Journal of Social Policy 3: 343–363. Ringe, Nils. 2005. “Policy Preference Formation in Legislative Politics: Structures, Actors, and Focal Points.” American Journal of Political Science 49: 731–745. Risse, Thomas. 2001. “A European Identity? Europeanization and the Evolution of Nation State Identities.” In Europeanization and Domestic Change, ed. Maria Green Cowles, James Caporaso, and Thomas Risse. Ithaca, NY: Cornell University Press, pp. 198–216. Rose, Richard. 1984. Do Parties Make a Difference? London: Macmillan. Rosenow, Joachim, and Frieder Naschold. 1994. Die Regulierung von Altersgrenzen: Strategien von Unternehmen und die Politik des Staates. Berlin: Edition Sigma. Ross, Fiona. 2000. “Beyond Left and Right: The New Partisan Politics of Welfare.” Governance 13: 155–183. Ruggie, John G. 1998. “What Makes the World Hang Together? Neoutilitarianism and the Social Constructivist Challenge.” International Organization 52: 855–885. Sabatier, Paul. 1998. “The Advocacy Coalition Framework: Revisions and Relevance for Europe.” Journal of European Public Policy 5: 98– 130. Sachverstandigenrat zur Begutachtung der gesamtwirtschaftlichen Entwick¨ lung. 2007. Jahresgutachten 2007/08: “Das Erreichte nicht verspielen.” Sandholtz, Wayne, and Alec Stone Sweet. 1998. European Integration and Supranational Governance. Oxford University Press. Sbragia, Alberta. 2002. “Conclusion to the Special Issue on the Institutional Balance and the Future of EU Governance: The Treaty of Nice, Institutional Balance, and Uncertainty.” Governance: An International Journal of Policy, Administration, and Institutions 15: 393–411. Scharpf, Fritz W. 1996. “Negative and Positive Integration in the Political Economy of European Welfare States.” In Governance in the European Union, ed. Gary Marks, Fritz W. Scharpf, Philippe C. Schmitter, and Wolfgang Streeck. London: Sage, pp. 15–39. Scharpf, Fritz W., and Vivien A. Schmidt. 2000. Welfare and Work in the Open Economy. Oxford University Press. Schelling, Thomas C. 1960. The Strategy of Conflict. Cambridge, MA: Harvard University Press.

Bibliography

169

Schludi, Martin. 2005. The Reform of Bismarckian Pension Systems: A Comparison of Pension Politics in Austria, France, Germany, Italy and Sweden. Amsterdam University Press. Schmidt, Manfred G. 2005. Sozialpolitik in Deutschland: historische Entwicklung und internationaler Vergleich, vol. 2. Opladen: Leske & Budrich. Schmidt, Susanne K. 2002. “The Impact of Mutual Recognition: Inbuilt Limits and Domestic Responses to the Single Market.” Journal of European Public Policy 9: 935–953. Schmidt, Vivien A. 2002. The Futures of European Capitalism. Oxford University Press. 2005. “Democracy in Europe: The Impact of European Integration.” Perspectives on Politics 3: 761–779. 2006a. “Adapting to Europe: Is it Harder for Britain?” British Journal of Politics and International Relations 8: 15–33. 2006b. “Bringing the State Back into the Varieties of Capitalism and Discourse Back into the Explanation of Change.” Paper presented at the American Political Science Association, August 31–September 3, 2006, Philadelphia, PA. 2006c. Democracy in Europe: The EU and National Polities. Oxford University Press. 2008. “Representative Democracy in Europe: EU ‘Policy without Politics’ and Member State ‘Politics without Policy’.” Working Paper. Schmidt, Vivien A., and Claudio M. Radaelli. 2004. “Policy Change and Discourse in Europe: Conceptual and Methodological Issues.” West European Politics 27: 183–210. Schoden, Michael. 2003. Betriebliche Altersversorgung (BetrAVG): Kom¨ die Praxis mit arbeitsrechtlicher Einfuhrung. ¨ mentar fur 2nd edn. Frankfurt am Main: Bund-Verlag. Schulze, Isabelle, and Michael Moran. 2007. “United Kingdom: Pension Politics in an Adversarial System.” In The Handbook of West European Pension Politics, ed. Ellen M. Immergut, Karen M. Anderson, and Isabelle Schulze. Oxford University Press, pp. 49–96. Shepsle, Kenneth A. 1986. “Institutional Equilibrium and Equilibrium Institutions.” In Political Science: The Science of Politics, ed. Herbert Weisberg. New York: Agathon, pp. 51–82. Simmons, Beth, and Zachary Elkins. 2004. “The Globalization of Liberalization: Policy Diffusion in the International Political Economy.” American Political Science Review 98: 171–189. Soskice, David. 1999. “Divergent Production Regimes: Coordinated and Uncoordinated Market Economies in the 1980s and 1990s.” In Continuity and Change in Contemporary Capitalism, ed. H. Kitschelt,

170

Bibliography

P. Lange, G. Marks, and J. D. Stephens. Cambridge University Press, pp. 101–134. Stacey, Jeffrey. 2010. Integrating Europe: Informal Politics and Institutional Change. New York: Oxford University Press. Standard & Poor’s. 2002. “Fiscal Slippage in Germany: A Coordination Problem?” Technical report. Standard & Poor’s. Stiller, Sabina. 2010. Ideational Leadership in German Welfare State Reform. Amsterdam University Press. Stinchcombe, Arthur L. 1968. Constructing Social Theories. New York: Harcourt. Stone, Randall W. 2002. Lending Credibility: The International Monetary Fund and the Post-Communist Transition. Princeton, NJ: Princeton University Press. 2011. Controlling Institutions: International Organizations and the Global Economy. Cambridge University Press. Streeck, Wolfgang, and Kathleen Thelen. 2005. Beyond Continuity: Institutional Change in Advanced Political Economies, ed. Wolfgang Streeck and Kathleen Thelen. Oxford University Press. Streeck, Wolfgang, and Christine Trampusch. 2005. “Economic Reform and the Political Economy of the German Welfare State.” German Politics 14: 174–195. Swenson, Peter. 2002. Capitalists Against Markets: The Making of Labor Markets and Welfare States in the United States and Sweden. Oxford University Press. Talani, Leila. 2000. “Who Wins and Who Loses in the City of London from the Establishment of European Monetary Union.” In After the Euro, ed. Colin Crouch. Oxford University Press, pp. 109–139. Tallberg, Jonas. 2004. “The Power of the Presidency: Brokerage, Efficiency, and Distribution in EU Negotiations.” Journal of Common Market Studies 42: 999–1022. 2006. Leadership and Negotiation in the European Union. Themes in European Governance. Cambridge University Press. 2008. “Bargaining Power in the European Council.” Journal of Common Market Studies 46: 685–708. Taverne, Dick. 2001. “Can Europe Pay for its Pensions?” In Pension Compendium: Pension Reform in Europe, ed. Federal Trust for Education and Research. London: Federal Trust for Education and Research. Taylor-Gooby, Peter. 2004. “New Social Risks in Postindustrial Society: Some Evidence on Responses to Active Labour Market Policies from Eurobarometer.” International Social Security Review 57: 45–64. Thelen, Kathleen. 2000. “Why German Employers Cannot Bring Themselves to Dismantle the German Model.” In Unions, Employers, and

Bibliography

171

Central Banks: Macroeconomic Coordination and Institutional Change in Social Market Economies, ed. Torben Iversen, Jonas Pontusson, and David Soskice. Cambridge University Press. 2004. How Institutions Evolve: The Political Economy of Skills in Germany, Britain, the United States, and Japan. Cambridge Studies in Comparative Politics. Cambridge University Press. 2012. “Varieties of Capitalism: Trajectories of Liberalization and the New Politics of Social Solidarity.” Annual Review of Political Science 15: 137–159. Thomson, Robert. 2008. “The Council Presidency in the European Union: Responsibility with Power.” Journal of Common Market Studies 46: 593–617. Thomson, Robert, Frans N. Stokman, Christopher H. Achen, and Thomas Konig. 2006. The European Union Decides: Testing Theories of Euro¨ pean Decision-Making. New York: Cambridge University Press. Tinbergen, Jan. 1965. International Economic Integration. Amsterdam: Elsevier. Truglia, Vincent J. 2002. “Sovereign Ratings and Aging Societies.” Paper presented at the “Financing Global Aging” conference, Brandeis University, Rosenberg Institute of Global Finance. Tsebelis, George, and Geoffrey Garrett. 1996. “Agenda Setting Power, Power Indices, and Decision Making in the European Union.” International Review of Law and Economics 16: 345–361. 2001. “The Institutional Foundations of Intergovernmentalism and Supranationalism in the European Union.” International Organization 55: 357–390. Vamvakidis, Athanasios. 2009. “Is There a ‘Reform Fatigue’ in the Euro Area?” Economic Modelling 26: 767–777. van Wijnbergen, Christa. 2002. “Imposing Consensus: State Steering of Welfare and Labor Market Reforms in Continental Europe.” Unpublished dissertation, Department of Political Science, Northwestern University. Wallace, Helen, James Caporaso, Fritz W. Scharpf, and Andrew Moravcsik. 1999. “Review Section Symposium: The Choice for Europe.” Journal of European Public Policy 6: 155–179. Weaver, R. Kent. 2006. “The Politics of Parallel Pensions: Lessons from the United Kingdom for the United States.” Center for Retirement Research at Boston College. CRR WP 2006–7. Weber, Max. 1978 [1921–1922]. Economy and Society: An Outline of Interpretive Sociology. 2 vols., Berkeley: University of California Press. Weyland, Kurt G. 2005. “Theories of Policy Diffusion: Lessons from Latin American Pension Reform.” World Politics 57: 262–295.

172

Bibliography

Whiteside, Noel. 2006. “Adapting Private Pensions to Public Purposes: Historical Perspectives on the Politics of Reform.” Journal of European Social Policy 16: 43–54. Woll, Cornelia. 2006. “Lobbying in the European Union: From Sui Generis to a Comparative Perspective.” Journal of European Public Policy 13: 456–469. World Bank. 1994. Averting the Old Age Crisis: Policies to Protect the Old and Promote Growth. Washington, DC: International Bank of Reconstruction and Development. Zohlnhofer, Reimut. 2003. “Partisan Politics, Party Competition and Veto ¨ Players: German Economic Policy in the Kohl Era.” Journal of Public Policy 23: 123–156.

Index

1408/71 Council Regulation, 89, 90 401(k) pension plan, 16, 98 aba, see also German Occupational Pension Association, 59 Abdelal, Rawi, 25, 100 Achen, Christopher, 6, 50 Adenauer, Konrad, 10 agenda setting, 6, 76, 79, 84, 86, 92, 94 aging population, 1, 9–11, 15, 27, 28, 30, 31, 38, 40, 66, 131, 137 Ahrend, Peter, 16, 52 Alesina, Alberto, 29 ¨ Alterseinkunftegesetz, 115, 116 Anderson, Karen, 13, 101 Aoki, Masahiko, 23 Arregui, Javier, 51 Austen-Smith, David, 54 Austria, 25, 35 Bachrach, Peter, 80 Banks, Jeffrey S., 54 Baratz, Morton, 80 bargaining breakdown, 23, 67, 73, 77, 83, 94, 135, 137, 138 bargaining success, 69–73, 84, 94, 109, 140, 141 barriers to workplace pension portability, 1, 19, 23, 48, 79, 115, 137 Barroso, Manuel, 155 Baumgartner, Frank R., 78 Beach, Derek, 77 Beck, Nathaniel, 37 Belgium, 35 Berlusconi, Silvio, 44 Beveridgean pension system, 5, 7, 14, 39, 40, 45, 48, 52, 54, 55, 57, 58,

60, 63, 64, 67, 69–72, 80, 88, 93, 109, 113, 126, 129, 138, 140 Biedenkopf, Kurt, 106 biometric risk coverage, 52, 58, 59, 61, 63, 71, 72, 74, 81, 85, 88, 90, 92, 111, 114, 127, 131, 132 Bismarck, Otto von, 14 Bismarckian pension system, 5, 7, 14, 15, 22, 39, 40, 45, 48, 49, 52–55, 57, 58, 61, 63, 64, 67, 69–72, 88, 93, 109, 113, 126, 129, 138, 140 Blair, Tony, 7, 93, 130, 139, 143 Blake, David, 60, 128, 132, 133, 139 Blum, Norbert, 66, 68, 106–108 ¨ Blyth, Marc, 25, 50, 65, 100 Bodie, Zvi, 60 Bonoli, Giuliano, 13, 39, 44 book reserve pensions, 7, 16–18, 23, 52, 57, 59, 61, 66, 69, 72, 90, 98, 99, 101, 102, 111, 113, 118, 122, 133 Box-Steffensmeier, Janet M., 36 Brinegar, Adam, 24 Britain, 35, 39, 58–61, 63, 80, 81, 126 Brittan, Leon, 80, 82, 83, 85, 103, 135, 137–139 Brooks, Sarah M., 4, 32, 37 Broxson, Alan, 82 Callaghan, Helen, 23 Caporaso, James, 77 Carter, David, 37 Casey, Bernhard H., 17, 53 Central and East European countries, 12 Christian Democracy, 66, 98, 122 Christian Democratic party, Germany, 108, 119–121 Christiansen, Thomas, 6 Clark, Gordon L., 136, 137

173

174 Clarke, Harold D., 130, 146 co-decision procedure, 50 coding of pension reforms, 36 COM (1991) 301 final, OJ C 312, 3 December 1991, 2, 80, 85, 103 Comit´e europ´een des assurances, 25, 83, 88 common shock, 28, 32, 46 communicative discourse, 25, 65, 100, 106 comparative institutional advantage, 58, 60 confidence bands, 42 Conservative governments, 38, 98, 107, 122, 127, 134 contribution holidays, 136 control variables, 40 cooperation problems, 9 coordinated market economies, 15, 58, 59, 61, 78, 99, 122 coordinative discourse, 25, 65, 100, 107 cost types, 6, 52–54 costs of pension reform, 6, 52, 53, 67 Council of Ministers, 50, 59, 71, 73, 77, 79, 80, 85, 110 Council Presidency, 51, 79, 91, 92, 142 Cowles, Maria Green, 24 Cram, Laura, 77, 78 credibility of signals, 20, 48, 66, 68, 69, 72, 73, 107, 109, 113, 123, 126, 130, 138, 139 Crombez, Christophe, 50 Cuevas, Alfredo, 34 Curtin, Deirdre, 50 Davidsson, Johan B., 13 De Haan, Jakob, 28 de la Porte, Caroline, 86 Deeg, Richard, 99 defined-benefit pension plans, 17, 60, 127, 128, 131–133, 137, 138 defined-contribution pension plans, 17, 60, 133 Delors, Jacques, 103 demographic aging see also aging population, 29 Department of Work and Pensions, Britain, 129–131, 141, 142 dependent variable, 36

Index Deutsch, Klaus Gunter, 59, 60 ¨ diffusion of policy ideas, 27, 28, 31, 32, 37, 43, 45 Directive 2003/41/EC see also IORP directive, 2 Directorates-General, 25, 76, 80, 83, 87 disability, 52, 81, 132 discourse analysis, 7, 21, 25, 49, 50, 58, 65, 69, 96, 97, 100, 106–108, 125, 127, 133, 142–144 Disney, Richard, 4, 27 domestic audience costs, 38, 127 domestic constraints argument, 20, 21, 66, 67, 69, 73, 107, 109, 113, 138, 139, 151 domestic win-set, 48, 67, 130 double-payment problem, 33 Downs, Anthony, 134 early retirement, 11, 13, 109, 121 Ebbinghaus, Bernhard, 5, 39, 53 EC Insolvency directive 80/987, 139 Eijffinger, Sylvester C. W., 28, 34 EIOPA, 3, 154 Elkins, Zachary, 31 Emmenegger, Patrick, 13 employers, 52, 57, 60, 61, 71, 99, 101, 102, 106, 127, 130, 131, 133, 136, 137 employment security, 15 equity culture, 58, 60 Esping-Andersen, Gosta, 10, 15, 16, ¨ 39, 93, 107 Esposito, Mariachiara, 81, 86, 137 European association of public enterprises and employers’ organizations, 82 European Central Bank, 67 European Commission, 6, 23, 25, 49, 51, 58, 60, 67, 70–72, 76–78, 80, 83, 84, 86, 88, 89, 94, 103, 110, 115, 117, 119, 142, 154 European Economic and Monetary Union, 35, 46, 67, 84, 109 European Federation for Retirement Provision, 25, 82, 83, 104, 143 European Financial Services Forum, 59 European Parliament, 22, 49, 50, 59, 79, 110

Index European Parliamentary Debates, 58 European Parliamentary Financial Services Forum, 58 European Trade Union Confederation, 82, 88 event history analysis, 37 Fuglister, Katharina, 31 ¨ Farrell, Henry, 6 Fawcett, Helen, 128 Fearon, James, 127 federation of German employers’ associations, 105, 116, 118, 119 feedback effects between EU and domestic levels, 51, 52, 67, 69, 103, 104, 106, 111, 112, 114, 115, 117 FEFSI, 122 Ferrera, Maurizio, 3, 24 Field, Frank, 80, 127 final-salary schemes see also defined-benefit pension plans, 60 Financial Services Act of 1986, Britain, 134 Finland, 35 Fioretos, Orfeo, 23 flat-rate benefits, 39, 133 formal model actor preferences, 58, 59, 62–64, 69 bargaining breakdown, 57 beliefs, 55, 57, 74, 75 cost types, 54, 56, 57, 67, 68 equilibrium, 56, 57 ideal points, 57, 62, 63, 71 information structure, 54, 56, 57, 67, 69 policy space, 54, 56, 62, 63, 65 sequence of moves, 55, 57, 63 utility functions, 54 France, 35, 44, 59, 89, 90 Franchino, Fabio, 155 Franzese, Robert J., 28 Frericks, Patricia, 19 funded pensions, 17, 19, 40, 66, 72, 101, 108, 128, 133 ‘Galton’s problem,’ 28, 31 Garrett, Geoffrey, 30

175 George, Stephen, 125 German Occupational Pension Association, 59, 105, 115–117, 119 German reunification, 12, 109 Germany, 35, 52, 58, 60, 62, 63, 66, 90, 97, 104–106, 109, 111–113, 115–117, 119–123 Gibbons, Robert, 56 Gilardi, Fabrizio, 4, 31 Gonzalez, Maria, 34 goodness of fit analyses, 24 government debt, 32, 33, 53, 67, 84, 111 government deficit, 32, 33, 53 Greece, 35 Guardiancich, Igor, 4 Hacker, Jacob, 127 Hall, Peter A., 15, 50, 61 Hannah, Leslie, 134 Hartlapp, Miriam, 83, 87 Hassel, Anke, 16, 53, 99 Hausermann, Silja, 13, 14 ¨ Haverland, Markus, 30, 89, 132 Hays, Jude C., 28 Heclo, Hugh, 31 Hering, Martin, 107, 112, 113 H´eritier, Adrienne, 6, 24 historical institutionalism, 24, 51, 96, 97, 122, 123 Hix, Simon, 97 Holzmann, Robert, 12, 28, 33 Hooghe, Liesbeth, 79, 80 Hopner, Martin, 23 ¨ Hoyland, Bjorn, 155 ¨ ideas-based approaches, 26 IG Metall, 68 Immergut, Ellen M., 21 incomplete information, 54, 57 independent variables, 38 informal governance, 6, 20, 48, 50, 54 insurance culture see also Bismarckian pension system, 15 interdependence see also diffusion of policy ideas, 32 intertemporal policy trade-off, 33 Intra-Commission dynamics, 80

176 investment regulations, 19, 48, 52, 59, 60, 63, 70–72, 74, 85, 90, 91, 103, 104, 111, 114, 125, 127, 130, 131, 139 IORP directive, 2, 5, 49–51, 61, 68, 69, 71, 73, 77, 85, 86, 90–92, 110, 115, 125, 128–130 IORPs, 2, 3, 63 Ireland, 35, 39, 129 Italy, 25, 35, 39, 44, 89, 91 Iversen, Torben, 13, 15, 23, 24 Jackson, Gregory, 99 Jacobs, Alan M., 27, 29, 135 Jahn, Detlef, 28 Jones, Bradford S., 36 Jones, Bryan D., 78 Jupp´e, Alain, 44 Karas, Othmar, 71 Katz, Jonathan N., 37 Katzenstein, Peter J., 24 Kayser, Mark A., 31, 125 Keefer, Philip, 38 Kelemen, Daniel R., 86 Kent, Weaver, 131 Keohane, Robert O., 31 King, Gary, 42 Kingdon, John W., 79 Kitschelt, Herbert, 30 Kleine, Mareike, 20 Kohl, Helmut, 7, 44, 66, 69, 96, 98, 100, 106, 107, 109, 112, 123, 145, 151 Konig, Thomas, 51 ¨ Koremenos, Barbara, 6 Korpi, Walter, 30 Kreppel, Amie, 50 labor unions, 53, 68, 101, 102, 109, 113 attitude towards workplace pensions, 16 Labour party, Britain, 134, 143–145 Left parties, see also social democracy, 30 legal personality, 89, 91, 92 Leistungszusage, 132 Liberal Intergovernmentalism Theory, 22, 52 77, 94

Index Liberal Market Economies, 58, 59, 61, 63, 78 Lindberg, Leon N., 21 Linkspartei, 93, 120 Lombardo, Davide, 34 longevity, 11, 52, 61, 81, 132 Lopez-Marmolejo, Arnoldo, 34 Lord, Christopher, 97 Luxembourg, 18, 35, 39 Luyet, St´ephane, 31 Maastricht Treaty, 5, 28, 30, 32, 33, 35, 37, 41, 46, 47, 50, 67, 84, 110, 111 Maastricht Treaty signatories, 35, 40 MacKellar, Landis, 28 Mahoney, James, 24, 98 Manow, Philip, 15 March, James G., 25 Mares, Isabela, 15, 62 Martin, Cathie Jo, 11, 30, 53 Maxwell affair, 60, 81, 129, 138, 139 means-tested benefits, 39 Merkel, Angela, 7, 68, 96, 98, 119, 123, 153 Meyer, Christoph, 88 Meyer, Traute, 13 Midtbo, Tor, 31 Mitchell, Olivia S., 60 money supply, 33 Monk, Ashby H. B., 136, 137 Moran, Michael, 131 Moravcsik, Andrew, 6, 52, 77, 79 Muller, Katharina, 32 ¨ Mum, David, 81, 86, 137 mutual recognition, 22, 74, 86, 89, 91 Myles, John, 20 Nanz, Patrizia, 86 Naschold, Frieder, 11, 121 Natali, David, 3 National Association for Pension Funds, Britain, 131–133 negative learning, 44, 46 neofunctionalism, 21 Netherlands, 35, 39, 59, 129 new social risks, 9, 13, 14 Norway, 35, 38 notional defined-contribution schemes, 37

Index Nurk, Bettina, 60 ¨ Nye, Joseph S., 31 Offe, Claus, 12 Olsen, Johan P., 25 on-balance pension schemes see also book reserve pensions, 18 open method of coordination, 86 Ordeshook, Peter C., 56 Orenstein, Mitchell, 4, 27, 31, 32, 37 Palier, Bruno, 13, 44 Palme, Joakim, 30 Parliamentary Financial Services Forum, 22 partisan waves, 31 party ideology, 30, 31, 38, 41, 44, 46, 93 PAYG pensions, 17, 28, 29, 32, 39, 44, 53, 66, 85, 104, 106, 114, 131, 132 pension acquisition rights, 18, 115, 117, 118, 143 Pension Act of 1986, Britain, 127 Pension Act of 1995, Britain, 132, 133 Pension Bill of 2004, Britain, 133 pension fund culture see also Beveridgean pension system, 7, 14, 48, 52, 104, 111 pension losses, 2, 7, 13, 14, 125, 128–130, 136, 139 pension market integration, 9, 19–21, 48, 51, 58, 59, 63, 67, 69, 76, 77, 86, 94, 99, 103, 125, 138 pension plan sponsors, 52, 60 pension policy discourse see also discourse analysis, 7 pension portability, 115–119, 129, 142 pension privatization, 127, 135 Pension Protection Fund, 81, 129, 139, 140 Pensionskasse, 64, 154 Pensionssicherungsverein, 52, 102 Peters, Guy, 84 Pierson, Paul, 20, 24, 27, 30, 38, 98, 135 Pochet, Philippe, 30 policy learning see also diffusion of policy ideas, 27 Pollack, Mark A., 78

177 Pontusson, Jonas, 23 Portugal, 35, 38 Powell, Bingham, 130 Princen, Sebastiaan, 77–79 probability of pension reform, 42, 43 prudent person rule, 60, 63, 71, 88, 91, 103, 104, 110, 130 Putnam, Robert D., 6, 23 qualified majority voting, 50 Radaelli, Claudio, 65, 100 rating agencies, 28, 34 reform fatigue, 35 regression analysis, 40 Rehder, Britta, 99 replacement rate, 44 Riester, Walter, 112 Riesterrente, 112, 114, 123 Ring, Patrick J., 136 Risse, Thomas, 50 Rose, Richard, 30 Ross, Fiona, 30 Ruggie, John G., 25 Rutkowski, Michal, 28 Sabatier, Paul, 50 Sanders, David, 130, 146 Sandholtz, Wayne, 24 Sbragia, Alberta, 50 Schafer, Armin, 23 ¨ Scharpf, Fritz W., 11, 20, 77 Scheingold, Stuart A., 21 Schelling, Thomas C., 6, 23 Schludi, Martin, 21, 30, 39 Schmidt, Manfred G., 108, 113 Schmidt, Susanne K., 84 Schmidt, Vivien A., 11, 24, 25, 65, 69, 100, 146 Schoden, Michael, 16, 52 Schrader, Alexander, 60 Schroder, Gerhard, 7, 68, 69, 73, 93, ¨ 96, 98, 112, 113, 119–121, 123, 151 Schulze, Isabelle, 131 segmented pension markets, 2, 5, 19 Shinkawa, Toshimitsu, 39 signaling, 5, 6, 26, 48, 51, 57, 69, 73, 74, 104, 107, 109, 113, 123, 130, 138, 139

178 Signorino, Curtis S., 37 Simmons, Beth, 31 skills, 52 social democracy, 16, 30, 39, 66, 93, 96, 122 Social Democratic party, Germany, 112, 113, 120, 121 social regulations, 48, 59 social security pensions, 14 solvency rules, 19, 52, 83 Soskice, David, 15, 61 sovereign credit risk, 34 Spain, 18, 35, 89, 91, 111 Stacey, Jeffrey, 6 state earnings-related pensions, Britain, 131, 134 Stewart, Marianne C., 130, 146 Stiller, Sabina, 66, 114 Stinchcombe, Arthur L., 24, 98 Stokman, Frans, 51 Stone, Randall W., 6, 8 Stone Sweet, Alec, 24 Streeck, Wolfgang, 24, 44, 98 supervisory regulations, 48, 59, 60, 110, 127 support funds, 64 survivor dependence, 52, 61, 81, 132 Swank, Duane, 30 Sweden, 35, 38, 129 Swenson, Peter, 61 Switzerland, 35, 39 Talani, Leila, 61, 140 Tallberg, Jonas, 78, 79, 89 Taverne, Dick, 4, 27 tax treatment of PAYG pensions, 18 tax treatment of workplace pensions, 73, 87–89, 92, 108, 119, 144 Taylor-Gooby, Peter, 13 Thatcher, Margaret, 7, 12, 127, 130, 133, 135, 143 Thelen, Kathleen, 11, 13, 23, 24, 53, 98 Thomson, Robert, 51, 79 time-series cross-sectional data, 37

Index timing of pension reform, 29, 30, 36, 37 Tinbergen, Jan, 20 Tomz, Michael R., 42 Tories, Britain, 134, 144 traditional social risks, 13 Trampusch, Christine, 44 treatment group, 38, 42 Truglia, Vincent J., 33, 34 Trustee Act of 2000, Britain, 133 Tsebelis, George, 50, 79 Tucker, Richard, 37 Turner, Andrew J., 60 umbrella causation, 28 Union of Industrial and Employers Confederations of Europe, 82 ¨ Unterstutzungskasse, 103 van Wijnbergen, Christa, 53 varieties of capitalism, 15, 23, 58, 59, 61, 78, 99, 122, 126 vesting period, 3, 14, 15, 18, 52, 58, 61, 115, 117, 118 veto points, 23, 49 Volkspartei, 120, waiting period, 3, 14, 15, 18, 52, 58, 61, 115, 117, 118 Wallace, Helen, 77 Weaver, Kent, 131 Weber, Max, 28, 154 Weyland, Kurt G., 4, 27, 32 Whiteley, Paul F., 130, 146 Whiteside, Noel, 17, 66 Whitten, Guy, 130 wind-up of pension schemes, 128 Wittenberg, Jason, 42 Woll, Cornelia, 24 women and workplace pensions, 19, 58 World Bank reports, 12, 31 Wren, Anne, 13 Zohlnhofer, Reimut, 100, 107, 108 ¨ Zwickel, Klaus, 68, 113