The Political Power of Business: Structure and Information in Public Policymaking (Toutledge Research in Comparative Politics)

The Political Power of Business

This book analyses the influence of business in democratic politics. Advice from business actors regularly carries more weight with policymakers than other interests because it refers to the core of the state-market nexus in democratic capitalism: the consequences for voters and policymakers of harming business and the economy. The book examines the resulting informational and structural constraints on public policymaking and their strategic use by business lobbyists. While the role of information is frequently acknowledged in studies on business political influence, very few empirical analyses of its strategic use exist. This book outlines a theoretical model of the role of information and its asymmetric supply for business actors’ ability to influence policy. Focusing on banking regulation and environmental politics, the informational–structural view of business power is evaluated empirically in a cross-national, multi-level research design involving case studies as well as quantitative analyses of elite survey data and policy outcomes in advanced capitalist democracies. Patrick Bernhagen suggests that, while democracy in capitalist society is vulnerable to a pro-business policy bias, better informed policymakers can redress the balance of power with business and improve on bringing policies in line with public preferences. His analysis identifies the institutional and behavioural factors affecting business’ informational power. The Political Power of Business will be of particular interest to students and researchers of political science, policymaking and business studies. Patrick Bernhagen is Lecturer at the Department of Politics and International Relations, University of Aberdeen, UK.

Routledge research in comparative politics 1 Democracy and Post-Communism Political change in the post-communist world Graeme Gill 2 Sub-State Nationalism A comparative analysis of institutional design Edited by Helena Catt and Michael Murphy 3 Reward for High Public Office Asian and Pacific Rim states Edited by Christopher Hood and B. Guy Peters 4 Social Democracy and Labour Market Policy Developments in Britain and Germany Knut Roder 5 Democratic Revolutions Asia and Eastern Europe Mark R. Thompson 6 Democratization A comparative analysis of 170 countries Tatu Vanhanen 7 Determinants of the Death Penalty A comparative study of the world Carsten Anckar 8 How Political Parties Respond to Voters Interest aggregation revisited Edited by Kay Lawson and Thomas Poguntke 9 Women, Quotas and Politics Edited by Drude Dahlerup 10 Citizenship and Ethnic Conflict Challenging the nation-state Haldun Gülalp 11 The Politics of Women’s Interests New comparative and international perspectives Edited by Louise Chappell and Lisa Hill 12 Political Disaffection in Contemporary Democracies Social capital, institutions and politics Edited by Mariano Torcal and José Ramón Montero

13 Representing Women in Parliament A comparative study Edited by Marian Sawer, Manon Tremblay and Linda Trimble 14 Democracy and Political Culture in Eastern Europe Edited by Hans-Dieter Klingemann, Dieter Fuchs and Jan Zielonka 15 Social Capital and Associations in European Democracies A comparative analysis Edited by William A. Maloney and Sigrid Roßteutscher 16 Citizenship and Involvement in European Democracies A comparative analysis Edited by Jan van Deth, José Ramón Montero and Anders Westholm 17 The Politics of Foundations A comparative analysis Edited by Helmut K. Anheier and Siobhan Daly 18 Party Policy in Modern Democracies Kenneth Benoit and Michael Laver 19 Semi-Presidentialism Outside Europe A comparative study Edited by Robert Elgie and Sophia Moestrup 20 Comparative Politics The principal-agent perspective Jan-Erik Lane 21 The Political Power of Business Structure and information in public policymaking Patrick Bernhagen

The Political Power of Business Structure and information in public policymaking

Patrick Bernhagen

First published 2007 by Routledge 2 Park Square, Milton Park, Abingdon, Oxon OX14 4RN Simultaneously published in the USA and Canada by Routledge 270 Madison Avenue, New York, NY 10016 Routledge is an imprint of the Taylor & Francis Group, an informa business This edition published in the Taylor & Francis e-Library, 2008. “To purchase your own copy of this or any of Taylor & Francis or Routledge’s collection of thousands of eBooks please go to www.eBookstore.tandf.co.uk.”

© 2007 Patrick Bernhagen All rights reserved. No part of this book may be reprinted or reproduced or utilised in any form or by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying and recording, or in any information storage or retrieval system, without permission in writing from the publishers. British Library Cataloguing in Publication Data A catalogue record for this book is available from the British Library Library of Congress Cataloging in Publication Data The political power of business : structure and information in public policymaking / [edited by] Patrick Bernhagen. p. cm. – (Routledge research in comparative politics ; 21) Includes bibliographical references and index. ISBN 978-0-415-45105-5 (hbk. : alk. paper) – ISBN 978-0-203-93261-2 (e-book: alk. paper) 1. Business and politics. 2. Corporations–Political activity. 3. Industrial policy. 4. Political planning. I. Bernhagen, Patrick. HD3611.P639 2007 322’.3–dc22 2007029648 ISBN 0-203-93261-7 Master e-book ISBN

ISBN10: 0-415-45105-1 (hbk) ISBN10: 0-203-93261-7 (ebk) ISBN13: 978-0-415-45105-5 (hbk) ISBN13: 978-0-203-93261-2 (ebk)

Contents

List of tables and figures Acknowledgements

ix xi

1

Introduction Linking action and structure: The power of being informed 8 Approach and method 12 Key terms and concepts 18 Structure of the book 20

1

2

Groups, institutions, networks, ideology or structural dependence: What drives business power? Pluralism and the study of business as an interest group 24 Neocorporatist systems and co-ordinated market economies 31 Elite networks and policy communities 34 Ideological dominance 39 Structural explanations of business dominance 43

22

3

An informational–structural model of business power Informational asymmetries between business actors and policymakers 56 Signalling models of lobbying 63 Reputation and lobbying 69

54

4

Two real-world signalling games A note on case selection 76 Regulation of small business banking in England and Wales 79 Phasing out PVC from public construction in Hesse 84 Discussion 88

76

viii Contents 5

Reputation and informativeness in lobbying Data: The 1996 European Members of Parliament Study 97 Lobbying, reputation and ideological proximity 100 Business contacts and influential lobbying 104 Discussion 111

6

Structure, information and environmental regulation Business preferences on environmental regulation 114 Dowsing flooded fields for underground water 117 Measuring structural constraints and information asymmetry 121 Sources and limits of business power: International trade, neocorporatism and eco groups 125 Discussion 131

7

Conclusions: Information, structure and the prospects for democracy under capitalism Information asymmetry in the repertoire of political resources 134 Outlook: Can policymakers’ information problems be alleviated? 139

Appendix A: Interviews and personal communications Appendix B: Coding and summary statistics of variables used in Ch. 5 Appendix C: Coding and summary statistics of variables used in Ch. 6 Appendix D: Jack-knifed estimates of regression models in Ch. 6 Notes References Index

93

113

134

147 149 152 155 158 161 175

Tables and figures

Tables 3.1 3.2 3.3 5.1 5.2 5.3 5.4 5.5 5.6 5.7 5.8 6.1 6.2 B.1 C.1 C.2 D.1 D.2

Payoffs of a signalling game Payoffs of a lobbying game Payoffs of a lobbying game with reputation costs and confidence costs Business contact and party type Business contact and left–right self-placement Business contact and environmental platform Business contact, party type and system of interest intermediation Business contact, left–right self-placement and system of interest intermediation Business contact, environmental platform and system of interest intermediation Lobbyists’ influence and MPs’ contact with different interest groups Determinants of lobbyist influence Effects of business power resources on environmental regime strength Interactive effects of business power resources on environmental regime strength Coding of variables used in Ch. 5 Coding of variables used in Ch. 6 Summary statistics of variables used in Ch. 6 Jack-knifed estimates of effects of business power resources on environmental regime strength Jack-knifed estimates of interactive effects of business power resources on environmental regime strength

65 66 72 100 101 101 102 103 103 107 109 129 131 149 152 154 156 157

Figures 3.1 3.2

A generic signalling model of lobbying A separating equilibrium of business lobbying

67 73

x Tables and figures 3.3 6.1 6.2 6.3

A pooling equilibrium of business lobbying Country scores of economic voting Environmental regime strength and structural sources of influence Environmental regime strength and informational sources of influence

74 123 126 127

Acknowledgements

Over the years during which this book developed from a research proposal to a completed manuscript, I have incurred great debt to many colleagues and friends. Above all, I would like to thank my PhD supervisor, Eddie Hyland, for inspiring conversations and rigorous criticism. I am also grateful to an anonymous reviewer at Routledge for providing very helpful commentary. Shiera Malik, Michael Marsh and Neil Mitchell read and commented on the entire manuscript in its earlier incarnation as a PhD thesis. Thanks are also due to Frank Baumgartner, Ken Benoit, Thomas Bräuninger, Terrence Casey, Raj Chari, Wyn Grant, Grant Jordan, Hyeok Yong Kwon, David Lowery, Gail McElroy, Jonathan Rodden, David Sanders, Mark Smith and Bernhard Wessels for very helpful suggestions on parts of the manuscript or conference papers based thereon. I would further like to thank Bernhard Wessels, Jacques Thomassen and the other scholars involved in the European Representation Study for generously making available to me the dataset from their research project. Some of the theoretical arguments, historical narratives and interview data in Chs 4 and 5 have appeared previously in Bernhagen and Bräuninger (2005) and have been reproduced here by permission of Blackwell Publishing. I am grateful to my interview partners for sharing their insights into business–government relations, whether that was in the context of specific cases or concerning the general logic of the argument of this book. I would also like to thank Heidi Bagtazo and Amelia McLarrin at Routledge for the opportunity to publish this work. Credits are further due to Aravind Kannankara for swift handling of the production process, and to Tony Donegan for excellent copy-editing. Finally, I would like to acknowledge grant support from the German Academic Exchange Service (DAAD), the Irish Research Council for the Humanities and Social Sciences and Trinity College Dublin.

1

Introduction

In August 2006, Forbes magazine published its annual list of the world’s 100 most powerful women. The top three positions are taken up by politicians: the German chancellor, Angela Merkel, the US secretary of state, Condoleezza Rice, and the Chinese vice premier, Wu Yi. The following seven places are occupied by businesswomen in leading positions at transnational corporations, including: the chief executives of PepsiCo, Xerox, Archer Daniels Midland and Sara Lee; the chairwoman of Areva; the chief financial officer of Citigroup; and the co-president of Morgan Stanley. Thus, according to Forbes, the majority of the ten most powerful women are in business, with only a minority being politicians. The picture looks much the same as one scrolls further down the list of top 100 (MacDonald and Schoenberger 2006). This distribution is not the result of the gender bias in public and commercial life. Although the higher echelons of politics are notoriously wanting of female members, this is even more so the case for the commanding heights of private enterprise. What makes these women powerful? How is it possible to compare the relative power of people in different spheres of life such as business and government? Three of the women are involved in the authoritative making of collectively binding decisions that comes with governing countries like China (Yi) or Germany (Merkel), or directing the foreign relations of the world’s supreme military power (Rice). The other seven, by contrast, are concerned with managing technology, agricultural and consumer goods companies and banks. This book seeks to provide answers to the question of how the powers of politicians and businesspeople relate to each other. At closer inspection, the question involves a trio of questions that are frequently raised by citizens, political activists and social scientists: Do the owners and managers of business enterprises have political power significantly above and beyond that enjoyed by other citizens? If the answer to this question is ‘no’, then usually no further questions follow and the discussion ends. If the answer is ‘yes’, two further questions arise: Where do they get their power from? And, are they justified in having it? While these questions have been frequently fought over, they are seldom answered satisfactorily. The first question in particular used to be very vociferously debated in Western industrialized societies, at least until about 1990. But what accounts for the vehemence of the debates is the fact that the

2 Introduction third question – do they deserve such power? – is constantly looming in the background. Critics of the capitalist mode of organizing social and economic affairs tend to hold the view that the power of a businessman or woman that goes beyond their power to cast a vote in an election or referendum or stand as a candidate in an election, is a reason for serious moral concern. If, so the argument goes, we are supposed to live in a society of equals, then surely it cannot be justified that some of us have significantly more power than others (Dahl 1989: 106–18). Unless, that is, this power is conferred on them by some explicit act of empowerment intended to ensure some degree of democratic accountability and due process. It seems that defenders of the capitalist way of organizing economies sensed that the third question would be a difficult one to deal with, and therefore focused on the first one. If they could make a convincing argument that businessmen and women – in short, capitalists – have no greater political power than any other citizen, the two remaining questions, in particular the ethically charged third one, would be irrelevant. Demonstrating that the capitalist mode of production and distribution is in no way inimical to political equality would also enable its proponents to counter most other major defects claimed by critics, typically from the left of the political spectrum, of the capitalist market economy. In particular, its defenders could show that for anyone with a concern for the fundamental values that have informed the organization of Western societies since the Enlightenment – liberty and equality – capitalism was the only game in town. Socialism, the major competitor as politico-economic systems go, appeared to fare much worse than capitalism with regard to individual liberty, but as long as critics from the left were able to claim that socialist systems scored better on equality, a fundamental ideological challenge to capitalist democracy remained. We know now what many have long suspected: that rather than promoting social and economic equality, elites in the planned economies of Eastern Europe and the former Soviet Union were able to enrich themselves at the expense of their fellow citizens. But the resulting inequalities were modest compared to the differences in income and wealth observed in countries like Britain or the US. Thus, with Western societies characterized by indisputable socio-economic inequalities, it was all the more important for advocates of capitalist democracy to refute the notion that capitalist elites wield unequal political power. If this could be achieved, socialism might still be claimed to perform better on social and economic equality, but because of the authoritarian character of their political systems the communist societies of the East were profoundly unequal politically. By contrast, while people’s economic and social circumstances in the West may be characterized by a degree of inequality, at least politically they would all be equal. With the retreat of serious socialist challenges to capitalism in the late 1980s and early 1990s, the erstwhile competition between politico-economic systems ceased to provide a reason for demonstrating the political equality of citizens in capitalist democracies. The concomitantly uncovered extent of the atrocities, human rights abuses, inefficiencies and numerous other failings of the communist regimes in

Introduction 3 the former Soviet Union and Eastern Europe was such that just about any other politico-economic system would be deemed preferable by default. For the same reason, the critical voices that pinned down the shortcomings of capitalist democracy by way of contrasting its performance against utopian or socialist blueprints went silent. Capitalist democracy has factually and almost overnight become the only game in town, before the ideological debates about political equality have been resolved. In fact, the debates themselves seem to have been rendered obsolete by history as the people in the countries undergoing transitions from communism have simultaneously written ‘democracy’ and ‘market’ on their banners (Przeworski 1991: 8). In this situation of easy victory, which prompted some to proclaim that history has come to an end (Fukuyama 1992), the ideological defenders of capitalist democracy have ceased to see the need to demonstrate how the institutions favoured by them – representative democracy, private property and the market – live up to the Enlightenment promises of freedom and equality. Instead, it has now become possible to recognize unashamedly that unequal economic power may imply unequal political power. Excuses no longer seem necessary, as few people would find anything objectionable in the fact that, in terms of power, the CEO of a soft drinks producer (Indra Nooyi of PepsiCo) is ranked just slightly behind the US secretary of state, and before the president of Chile or the prime minister of New Zealand. Now that almost two decades have passed since the contest between the two major politico-economic systems ended with the collapse of one of the competitors, however, it is time for the victor to be judged by the standards of its very own normative foundations inherited from the Enlightenment. Above all, this involves a close look at the political role of economic elites. Politics, Lasswell wrote over 70 years ago, is about who gets what, when and how (Lasswell 1958 [1936]). Today, few would doubt that in this sense business is a formidable political actor enjoying a unique place in the political landscape (Coen and Grant 2006: 13). In a review of the literature on corporate political activity, Hillman, Keim and Schuler (2004: 838) find it ‘indisputable that business firms … are among the most prominent political players not only in Washington, DC but in capitals across the globe’. Thus, the factual question concerning the political power of business has been replaced by a commonly shared understanding that business fares rather well when it comes to securing favourable political outcomes. Concomitantly, the other two questions raised at the outset of this chapter have also become relegated to the background. Consider the third question: Do business people deserve to be as powerful as they are? Put differently, where does the legitimacy of their power derive from? In representative liberal democracies, people hold that procedures and institutions convey legitimate power to individuals and outcomes. For democrats and democratic theorists, election by popular vote constitutes the standard mechanism of conveying legitimate power onto leaders. Looking at the ten most powerful women in the Forbes ranking, it turns out that only Merkel has attained her position by some form of democratic electoral process. Rice comes close, having reached her position by appointment according to constitutionally determined procedures. While not being directly empowered through popular or parliamentary vote, the

4 Introduction office of the US secretary of state derives mediated democratic legitimacy by virtue of the fact that her appointment is made by the president, who himself is held accountable through popular vote. There are, of course, all sorts of powerful social limits to the factual equality of people’s chances to hold these kinds of offices. Despite the enormous differences between countries like, say, China and the US, all ten women have a number of social characteristics in common. All come from relatively wealthy and educated family backgrounds and enjoyed excellent third-level education in the finest institutions of higher learning in their respective countries. None of these women would count as ‘average’ or ‘ordinary’ by any meaning of these terms – although Merkel’s origins are at most middle class. But eight out of the ten most powerful women have come to occupy positions of power through procedures that lack any of the aforementioned criteria of due process and democratic legitimacy. Yi is powerful because of a successful career in the ruling party of a communist dictatorship. Ironically, the mechanisms that brought the remaining seven women – all business executives in liberal democratic countries – to enjoy their reputedly powerful positions are no different than the ones that elevated the viceprime minister of China above the heads of a billion clerks, peasants and workers. They all ‘worked their way up’ by pursuing normal, if highly successful, top-level bureaucratic careers, climbing up strictly hierarchical ladders as they have been deemed reliable, efficient and acceptable by other bureaucrats, who themselves had attained their positions by the same mechanisms before them. At no point in their careers did the adult population of their countries cast a democratic vote over their appointment. Nor were they appointed by a democratically elected official. And nor are there any orderly procedures in place by which the fairness of their ascent to power can be evaluated. As Moene put it, ‘[w]hile democratic governance is occupying new territory in Eastern European politics, most firms in the West are still governed like command economies in miniature’ (Moene 1993: 400). It is possible that the power of business people is of a different kind than that of politicians. After all, unlike Merkel or Rice, Nooyi and Cruz do not wield command over state bureaucracies and armies. But the pollsters at Forbes do not seem to think that these differences matter much. Their power rankings are based on a composite of visibility (measured by press citations) and economic impact. The latter reflects the women’s résumés, the size of the economic sphere over which they hold sway and a weighting designed to render a firm’s size comparable to that of a national economy (MacDonald and Schoenberger 2006). After that, the Forbes investigators make no distinction between the power of a business actor and that of a political actor. As Forbes is a business magazine rather than a social– scientific research institute, the validity of this list and its method of measurement is not beyond doubt. Thus, a more systematic investigation into the nature and sources of the power of capitalists is needed. We can distinguish two dimensions of power wielded by business executives. One dimension concerns their power within firms. Here, as Moene points out, command-style hierarchies continue to prevail, forming islands of authoritarian

Introduction 5 power in otherwise democratically organized societies. Dahl describes this poignantly: Like the government of the state, the government of a firm makes decisions that apply uniformly to all workers or a category of workers: decisions governing the place of work, time of work, product of work, minimally acceptable rate of work, equipment to be used at work, number of workers, number (and identity) of workers laid off in slack times – or whether the plant is to be shut down and there will be no work at all. These decisions are enforced by sanctions, including the ultimate sanction of firing. Dahl 1985: 113–4 The authoritarian power structures within economic enterprises, however, may be justified by the normative underpinnings of capitalist democracy. The claim of owners and managers to rule autocratically within their firms can be backed up by the efficiency-enhancing aspects of that form of governance, through invocation of natural rights to private property or the hint at the fact that few employees might care much to get involved in the running of the firm they work in. Most justifications involve a mixture of these elements (ibid.: 111–35). But while one may agree or disagree with these justifications and their implications for the way economic enterprises are governed, few people would disagree with Hobbes’ succinct formula that ‘[t]o have servants is Power’ (Hobbes 1996 [1651]: 62). Capitalist democracy at any rate makes no pretence to intra-firm democracy. The other dimension of the power of capitalist leaders concerns the possibility that their non-democratic economic power may have significant implications for decisionmaking in the public sphere – a sphere that in capitalist democracies is supposed to be governed democratically. The claim that economic power in some sense means political power is at the heart of disputes concerning the sources of the political power of business. The questions about the sources of business political power and the justification of such power are more relevant today than ever. This is despite the fact that both questions have been around for some time – at least since politics as we understand it today was brought into being with the French Revolution. In the middle of the nineteenth century, for example, Marx and Engels claimed that the power of capitalists is essentially political power when they wrote that ‘[t]he executive of the modern state is but a committee for managing the common affairs of the whole bourgeoisie’ (Marx and Engels 1977 [1848]: 223). One hundred and thirty years later, Lindblom expressed concern about the political power of capitalists in his study Politics and Markets, concluding that ‘[t]he large private corporation fits oddly into democratic theory and vision. Indeed, it does not fit’ (Lindblom 1977: 356). And democratic theorists from different quarters are concerned that, if capitalists wield disproportionate power over political outcomes, then political equality, democratic accountability and the legitimacy of public policy may all be greatly undermined (Dahl 1989: 324–28; Green 1985). To begin with, disproportionate political influence on behalf of some by virtue of their positions as owners and managers of business enterprise is incompatible

6 Introduction with the essential democratic requirement that ‘everybody should count for one and nobody for more than one’ (Green 1985: 13). Furthermore, Macpherson (1973: 8–9) points out that liberal–democratic theory contains an ethical claim to equally maximize human powers. The capitalist distribution of the means of production limits the access to what a person needs in order to use and develop their capacities, thereby forcing a majority of people to pay for access with part of their powers by working for others. In doing so, they subject themselves to the authoritative power of the owners of capital. The result is a ‘continuous net transfer of part of the powers of some men to others, and a diminution of the human essence of those from whom power is being transferred’ (Macpherson 1973: 10). In this way, the capitalist organization of the economy contradicts the moral principle, implicit in the judgement of Western liberal democratic theory, that individual freedom is preferable to authoritative allocation of work and reward (ibid.: 19). The common recourse to a natural or human right to private property may not have sufficient reach to justify such transfer of power. For, as Dahl maintains, ‘[w]e cannot leap from my entitlement to secure possession of the shirt on my back or the cash in my pocket to a fundamental moral right to acquire shares in IBM and therewith the standard rights of ownership that shareholdings legally convey’ (Dahl 1985: 7). But how can we say that capitalists pose a threat to democracy if everywhere we look, capitalism and democracy go hand in hand? Friedman reminds us that, historically, there has been no society ‘that has been marked by a large measure of political freedom and that has not used something comparable to a free market to organize the bulk of economic activity’ (Friedman 1962: 9). Moreover, capitalism has been portrayed as a necessary condition for the development of democracy. According to Lipset, the chances for both the emergence and the survival of democracy increase with the rise of the overall level of socio-economic development of a given society (Lipset 1959: 75), and no other socio-economic system has proved a match for capitalism as far as the production of social wealth is concerned. Indeed, historically, whenever market economies have been successful over a period of time, pressure for democratization has not taken long to ensue (Berger 1992). Schumpeter (1976 [1942]: 297) even characterized modern democracy as ‘a product of the capitalist process’. And for Hayek (1944), capitalism and democracy equally embody the enlightenment foundations of European civilization complete with a commitment to individual freedom. It is important to keep in mind, however, that the modern impetus to democratization came with republican forms of government that were primarily developed by aspiring commercial classes. While broadening access to political power from monarchs to aristocracies and then to propertied citizens was one thing, widening participation in politics to ‘the many’ as a genuine majority of the adult community was a much larger project (Carver 2003: 253–4). In Britain and Europe, it has taken almost two centuries to secure the expansion of civil and political rights for the benefit of substantial groups of the population. All along, capitalists, and many liberals, have watched the processes of democratization with scepticism, and often with outright hostility. And they had good reason to be sceptical.

Introduction 7 Macpherson points out that the gradual extension of the franchise went hand in hand with significant curtailments of capitalist property rights, not least through the emergence of the welfare state (Macpherson 1973: 148). This trade-off reflects a fundamental tension between property rights enjoyed by owners and personal rights enjoyed by citizens (Bowles and Gintis 1986: 32). Furthermore, the first two major waves of democratization in Western Europe cannot be understood without reference to major inter-state conflict. Therborn argues that the two world wars were causally related to democratization in two ways: first, in the form of effective support through external allies that the excluded classes found in their domestic struggle for political rights; and second, by the fact that the ruling elites often had to trade in political concessions to the lower classes in order to gain support for national war mobilization (Therborn 1977: 17–23). Thus, despite the strong historical correlation between capitalism and democracy, the relationship between the two systems is not necessarily a harmonious one. We do not need to invoke natural rights and moral justifications to see that disproportionate political influence by business interests can be highly problematic. As Olson (1982) has shown, the disproportionate political power of producer groups can confound the efficiencies of the market economy, leading to deadweight losses and a retardation of society’s economic development. And in recent years, concern has been voiced that, no matter what disproportionate political power capitalists may have already enjoyed in the past, their privileges may now have been further boosted as a result of economic globalization. Growing numbers of citizens, politicians and scholars suspect that the increasing global integration of markets may further limit the range of policy choices available to democratically elected leaders. A dominant theme in public debate and political and economic research alike is the possibility that, from the 1980s onwards, the increasingly unrestrained movement of capital between nations is reducing the policy autonomy of governments while strengthening the political bargaining power of capitalists. Under these conditions, political parties may find it more difficult to aggregate citizen preferences in a consistent manner; their policies become increasingly similar and the partisan character of government decreases, depriving citizens of an essential tool for making meaningful choices at elections. It is often argued that coping with globalization requires flexible domestic political structures, which in turn entails an orientation towards small, rule-bound government, making social–democratic, Keynesian demand-side management and ‘big government’ less and less feasible. Because large companies and transnational corporations have the option to shift or outsource production to low-wage countries, union bargaining power relative to employers is weakened (Rodrik 1997: 23–5). The need for nation states to create profitable environments for investors may induce them to forgo necessary but costly policies, thus leading to a ‘race to the bottom’ of social, environmental and health-and-safety-related standards between countries. Furthermore, the tax burden shifts from mobile capital to immobile labour, which leads to a redistribution of wealth from the poor to the rich, and to a potential rolling-back of the welfare state due to sharpened constraints on governments’ ability to raise revenue.

8 Introduction Resulting in increased demands on the state to provide social insurance against the calamities of an open economy while simultaneously reducing the ability of the state to perform that role effectively, globalization presents a serious dilemma. And to the extent that in this way globalization leads to the erosion of the power and capacity of the state, democracy is imperilled because the state is the primary locus of democratic arrangements in the modern world (Cerny 1999: 19). Again, it should be pointed out that the problems are all but new. ‘Flight of capital’, for example, has been recognized as a constraint on democratic politics at least since the end of World War I. It has been claimed to play a causal role in the political crises of the interwar period, leading ultimately to the overthrow of the liberal governments of France in 1925, and again in 1938, and has arguably played a role in the development of a successful fascist movement in Germany in the 1930s (Polanyi 2001 [1944]: 25). Indeed, Wilson (2006: 40) reminds us that ‘[t]he argument for the “race to the bottom” as a consequence of globalization is essentially the same as Lindblom’s argument that business enjoys a “privileged position” and can be subjected to the same critique unless (as is plausible to argue) there has been a fundamental change in the character of world capitalism’. It is not the task of this book to argue whether or not such a fundamental change has occurred. Suffice it to say here that there are not many voices in the literature asserting that the political weight of business has decreased as a result of globalization. But while there is little consensus over whether the political position of business really has been strengthened in the process of globalization, there is even less consensus in the literature over what the sources of the political power of business are. And while questions of the political power of economic actors remain among the most important to be asked about politics and government, the degree to which the modern democratic state enjoys ‘autonomy’ under various economic–structural conditions is under-explored. This is particularly significant if we consider that questions of business control over public policy are central to questions of party responsiveness and partisan convergence. Moreover, scholars have come to suspect a structurally induced ‘corporate bias’ at the supranational level of European Union (EU) politics (Chari and Cavatorta 2002), suggesting that the problem of business political power is not confined to national-level politics. Fortunately for the analysis pursued in this book, business political influence varies from one capitalist nation to another. It also varies within the same political system over time (Vogel 1989) and is contingent on attributes of the businesses and policymakers in question (Martin 1989). And because there is variation as well as co-variation with other factors, it is possible to develop causal explanations of these variations which, in turn, may help us to improve our understanding of the sources and limits of business political power.

Linking action and structure: The power of being informed Analyses of the political involvement of business in democratic capitalism tend to follow one of two paradigms. They either view business political power as a matter

Introduction 9 of intentional interference, or as a problem of non-intentional dominance. In the first category, we find studies of special interest politics, lobbying and policy networks, which treat business as a political interest group. Within this paradigm, the subject matter under study is one or another form of corporate political action. By this, scholars usually mean the entering of the political arena by firms to purchase action or inaction from public officials (Getz 1997: 38). Research in this area concerns the resources at the command of business and its interest organizations for political action, encompassing the fields of party and campaign finance (Clawson, Neustadtl and Weller 1998), corruption (Rose-Ackerman 1999), the personal and functional ties between economic and political elites (Domhoff 1998) and sometimes business’ sway over public opinion (Page, Shapiro and Dempsey 1987). To this list, we can add business involvement in major political crises. When other political strategies have failed, business has been known to resort to the ultimate means of contributing to the violent overthrow of governments that enact policies contrary to its interests (Block 1977: 19). This may involve the support of military coups against democratically elected government and their replacement with authoritarian regimes, as happened in 1973 in Chile, or with the direct rule by the business elite, as was unsuccessfully attempted in 2002 in Venezuela. Instrumental-Marxist accounts of business power include all these dimensions of intentional interference in the repertoire of political means available to capitalists (Miliband 1969). While research within this broad area focuses on more or less readily observable political action, the success of this approach in explaining political outcomes is mixed. For example, in a meta-analysis of more than 30 studies on the link between contributions and congressional roll calls in the US, Roscoe and Jenkins (2005) find that only about one-third of roll-call votes exhibit an impact of campaign contributions. This lack of consistent findings is perhaps not surprising, as there are limits to the extent to which policymakers can be responsive to corporate political action. After all, policymakers must take into account that voters would be dissatisfied with sell-outs to organized interests (Denzau and Munger 1986). Avoiding these problems, the second paradigm rests on the powerful, but empirically elusive, claim that the owners and managers of private enterprises enjoy a structural power position that enables them to secure political outcomes even if they abstain from political action. This claim has gained prominence in explaining the failure of the social democratic project of building capitalism with a ‘human face’, to mirror Alexander Dubˇcek’s slogan. It is not very contentious that ‘there are “powerful constraints” on the extent to which “viable” forms of social democracy can stray, if not from dogmatic neoliberal models, then at least from the broader confines of liberal capitalist democracy’ (Whitehead 1992: 156). But the problem of structural constraints concerns more generally the problem of government ‘efficacy’; that is, the government’s ability to respond to its citizens and to solve pressing problems. In an unlikely concurrence of Marxist theories of the state in capitalism and the central tenets of the neoliberal Chicago school of economics, this

10 Introduction paradigm prioritizes the non-intentional, structural constraints on the policymaking capacities of elected officials. Variously known as the ‘structural dependence of the state on capital’ (Przeworski and Wallerstein 1988), the ‘business confidence’ factor (Block 1977) or the ‘structurally privileged position of business’ (Lindblom 1977), the central claim is that any negative effects of redistribution on the share of profits consumed by capitalists lead to disincentives for investment, followed by a whole appendage of sluggish growth, unemployment and decreasing tax revenue. In anticipation of these effects and their electoral and fiscal consequences, policymakers tend to carefully avoid any policy that might affect the revenue prospects of business and instead are concerned to maintain business confidence by all means. As a result, the political preferences and economic interests of business enjoy disproportionate consideration in the formulation and implementation of public policy, even if business abstains from direct political activity (Przeworski and Wallerstein 1988: 12). Because there are different interests of business and factions of capital, even explicitly pro-business governments are likely to produce policies that would harm some business interests, while favouring others. Thus, even where the grip of the business confidence factor on public policy is very strong, some level of business political activity would have to be expected. But to the extent that they exist, these structural constraints clearly pose serious problems for democracy. The essential democratic requirements for an electoral mandate to policymaking are that there is some difference between parties so that electors can make meaningful choices; and that, once elected, politicians will do more or less what they promised to do. Because the outcomes of capitalists’ formally nonpolitical decisions are of importance for both citizens and policymakers, political decisions that are likely to affect capitalists will almost certainly be made with a view to their likely reaction to them. A safe way for policymakers to avoid undesired decisions by corporate leaders is to gear policy so as to please them as much as possible. At the most general level, policymakers have strong incentives to create a climate in which capitalist enterprise flourishes. Plausible as theories of structural dominance may sound, they have their own serious problems. From an empirical vantage, if the structural power thesis holds, we should rarely witness policy change over time that overrides business’ political preferences. However, instances of business failing in the policy struggle are frequently observed (Smith 2000). Moreover, structural dependence theories cannot accommodate the ubiquitous political activity of business. The question inevitably arises: Why, if their interests are automatically taken care of, would capitalists waste time and money on politics in the first place? Moreover, research on the covert sources of business influence over public policy is plagued by fundamental methodological problems. As Mitchell (1997: 170) points out, while we can plausibly argue that the structural power of business exists, it is much more difficult to find direct evidence of structural power than is the case with the more readily observable instruments of power, like making campaign contributions or lobbying. According to Quinn and Shapiro (1991: 868), ‘[s]earching for business structural power is much like using a water-witching stick to dowse a field for underground water’. They add that the complexity of this task is immensely

Introduction 11 exacerbated by the fact that the political field is usually flooded with the overtly manifest exercise of business power. While it may be useful to study the political role of business through the lenses of one of these paradigmatic frameworks, problems arise from the fact that doing so usually hides important insights from the ‘other side’. As Wilson (2006: 34–5) points out, this separation has produced an unfortunate division between studies of business agency in politics on the one hand and structure on the other: ‘By and large, those who study lobbying or the giving of campaign contributions by business do not concern themselves with topics such as the degree to which capital movements constrain governments’. Consequently, Wilson (ibid.: 47) finds that there is a need ‘for work that examines simultaneously both the political behaviour of business and the structural connections between business and the state’. The present book investigates a third alternative that sets out to combine the analysis of agency and structure in the context of business political influence. Specifically, it examines the extent of analysis of political, economic and technical information, the way this information is distributed and how it is used by political and private actors provides avenues for analysing business power. The point of departure is the observation that the predominance of the two separate paradigms has led to a situation in which an important aspect of public policymaking has become obscured, namely the fact that policymakers are confronted with huge informational problems concerning the likely consequences of policies as well as their valuation by citizens, while special interest groups often have privileged access to the pertinent information (van Winden 1999: 3). While interest group scholars have long recognized that lobbyists can gain credibility and influence in policymaking through detailed factual knowledge of the issues (Kollman 1998), they have tended to ignore how the content of this knowledge is applied strategically by lobbyists. Polk (2002) rightfully criticizes the fact that different approaches mainly exist in parallel, which means that lobbying is either conceptualized as the contribution of payments or the transmission of information. Hillman, Keim and Schuler (2004: 851) ask if combinations of financial and informational political strategies can provide higher net political benefits to business than using single strategies. Indeed, lobbying often involves combinations of these activities. Instances of business actors advising elected policymakers on how to best deal with the structural constraints on public policy are ubiquitous in daily political discourse. The president of the German Chamber of Industry and Commerce, Ludwig-Georg Braun, for example, used privileged information to encourage the Social Democrat-led German government to persevere with its otherwise immensely unpopular programme of reducing unemployment benefits. The reforms antagonized the governing party’s core constituency and brought over 40,000 protesters into the streets on a weekly basis for three months in 2004 across hundreds of German cities and towns. But when Braun, who represents 3.4 million companies, pointed out that ‘[t]he country can’t afford these benefits any more [while] [s]taying the course will add wind to the sails of the economy’ (Irish Times, 12 August 2004), the government chose to follow his advice.

12 Introduction But politicians are not naïve. They suspect that such advice is not given free of the pursuit of business’ self-interest and are sceptical toward lobbying from business. For their part, business leaders and lobbyists have an interest in maintaining their reputation as suppliers of reliable information, which is a necessary condition for the ability to exert any influence at all. As a result, they are constrained in their freedom to engage in ‘cheap talk’ with respect to ostensible adverse effects of policy: only firms whose operations or revenue prospects would be severely affected by a pending policy shift and who hold positive expectations that their efforts will succeed have incentives to incur significant expenses for costly political action. Given their informational disadvantage, policymakers can try to infer the credibility of a lobbying message from observing the level of costly political action firms are willing to employ in the lobbying process. In this view, the informational asymmetry between business and policymakers can seriously compromise the capacity of governments to implement comprehensive and long-term objectives in a well co-ordinated manner. But by the same token policymakers are not easily fooled. This is the basic idea expressed in informational accounts of business lobbying: because of policymakers’ capacity constraints and special interest groups’ strong incentives to pool resources and conduct research on issues of concern to their members, interest groups often enjoy crucial informational advantages vis-à-vis policymakers. If the special interest group in question is a business group, the informational asymmetry goes further. Individual firms accumulate knowledge about relevant policy issues in the course of performing their everyday activities. As a result, business has first-hand access to data on the costs of policy. By communicating all or some of this information to policymakers, businesses can utilize the need for governments to preserve business confidence – the essence of the theory of the structural power of business – as an instrument of their lobbying strategies. The remainder of the book will examine this idea further and subject it to a series of empirical tests.

Approach and method The analysis is rooted in the theoretical framework of rational choice. The rational choice approach rests on the principle of methodological individualism; that is, ‘the doctrine that all social phenomena – their structure and their change – are in principal explicable in ways that only involve individuals – their properties, their goals, their beliefs and their actions’ (Elster 1985: 5). The theory of rational choice therefore assumes that individual choices are purposive and influenced by some form of reason. This implies that the actors under study possess cognitive abilities and the ability to apply principles of logic to comprehend the world, to connect ends to means and to use these abilities to make decisions – or choices – between alternative courses of action. Thus, reasoning is not an abstract human ability, but is applied in behaviour, which means that behaviour can be explained and predicted. Operating within this framework involves the acceptance of two further crucial assumptions. First, actors frequently have to make decisions under conditions of imperfect information. Therefore, the individual calculus of choice not only entails

Introduction 13 the consideration of preferences over a set of options, but also consideration of the relative likelihood of an action producing the desired outcome. Second, individuals’ preferences are treated as given and analysed as exogenous causes of actions. While the billions of pounds spent each year on advertising suggest that businesses exert influence over the formation of individuals’ preferences (or at least believe they do), this book does not, for the most part, deal with this problem. However, as a central theme of the book is the information required to translate preferences effectively into outcomes by means of actions, the book deals explicitly with the possibilities of manipulating other actors’ beliefs about the world, specifically about how ends are connected to means. Thus, the analysis is about the way political actors act strategically under conditions of incomplete information, given their preferences.1 The theoretical approach to the study of economic, political and social phenomena that has become known as ‘rational choice’, ‘public choice’, ‘social choice’ or ‘positive political theory’ postulates a fundamental linkage between free-market economics and democratic politics by using the same analytical apparatus to study both. Why then would one want to employ this framework for a critical investigation into the problems for democracy that flow from the capitalist organization of economic life? The answer is that the rational choice approach to political science is superbly equipped for the study of the structural constraints on public policy that arise from the choices of utility-maximizing actors in the institutional framework of capitalist democracy. While the approach has been criticized for its simplistic views on human nature and motivations for action, the theory’s assumption of utility-maximizing actors is immensely plausible in the areas of lobbying and public policymaking, especially where private, commercial interests are involved (Elster 1986; Satz and Ferejohn 1994). The approach is particularly well-suited for the study of the strategic interaction between businesses and policymakers that takes place against the backdrop of structural and informational constraints. At the core of the theory of the structural power of business outlined above is the idea of unintended consequences of behaviour. According to Popper, whose critical rationalism can be seen as providing the philosophical underpinning to the rational choice approach, the study of unintended consequences is the very essence of social science. Popper maintained that social science should analyze the unintended collective outcomes of individual actions, because ‘to believe that much of what happens in society is intended is to engage in conspiracy theory’ (Popper 1966: 95). If explanation is then to be based on individuals and the intended and unintended outcomes of their actions, what room can there be for structural variables? This book analyses the relationship between centralized decisionmaking by policymakers and the profit-seeking choices of business actors, as well as the latter’s strategic political actions. For this purpose, structure is defined as the relational features existing between actors comprising internalized rules, goals and beliefs and the distribution of resources, the totality of which constitutes an essential part of the environment or context which a rational actor will have to take account of in making decisions and devising strategies for action. For any actor, a crucial element of

14 Introduction this contextual environment will be the likely actions of others that are relevant to the success of the actor’s strategies. By examining how strategic actors consciously employ the existence of structural constraints on action in the strategies by which they further their ends in interaction with others, the analysis seeks to move beyond the fruitless dualism of agency and structure. It acknowledges that people ‘make their own history, but they do not make it just as they please; they do not make it under circumstances chosen by themselves, but under circumstances directly encountered, given, and transmitted from the past’ (Marx 1977 [1852]: 300). The strategy of combining structure and agency pursued shares some of the defining features of Jessop’s (1996) strategic–relational approach. Developed out of a critique of Giddens’ (1979) theory of structuration, Jessop’s is perhaps the most promising attempt yet in the world of social theory to combine structural variables with rational and strategic agency. While Giddens’ theory enables us to view structure as both emergent from human action and constraining and determining regardless of the agents and actions subject to constraints, Jessop’s modification allows for an explicitly strategic orientation. Instead of ‘mechanically relating’ agency and structure to each other in the fashion of Giddens’ structuration theory, Jessop (1996: 124) proposes a ‘dialectical duality’ of structure and agency. Action can be analysed in terms of its performance by agents with ‘strategically calculating structural orientation’ (ibid.: 123–4). But the approach underlying the present study deviates from Jessop’s account in one crucial respect: in regarding actors as reflective, strategically calculating subjects oriented to the structural conjunctural complexities of action contexts, Jessop explicitly allows that actors reflect on their identities and interests (ibid.: 125). His strategic–relational view of the structure– agency duality thereby accommodates the active and reflexive remaking by actors of their own identities and interests. In order to analyse business–government interactions in the context of specific policy problems, the present study assumes the exogeneity of actors’ preferences and identities. This requires that stipulations will be made about what businesses and policymakers want to achieve with respect to particular policy issues and that these preferences are then held constant for the purpose of the analysis. Like other deductive theories, rational choice theories must be subjected to empirical tests against observed facts if they want to claim social–scientific relevance. In order to identify the sources and limits of business political power, this book develops a theoretical model that stipulates mechanisms connecting causes and effects. Causation is defined as the processes and intervening variables through which some variables effect changes in the value of others. A causal mechanism consists of a ‘series of social events’ that are themselves causally related to one another in a causal chain (Little 1991: 15–6). A cause can, for example, be a purpose-driven action that employs a useful resource to obtain a certain outcome, where the outcome is the effect. The causal story connecting these events must be internally consistent with respect to both the temporal succession of the events, as well as their spatial relations. This implies that mere covariation between two or more variables does not suffice for a causal relationship to be established empirically. Instead, it must also be shown how this change in the independent variable

Introduction 15 explains variation in the dependent variable. Such explanation involves subjecting the observations in which that correlation appears to more intensive, detailed scrutiny in order to establish whether an intervening process – a causal nexus – exists between the independent variable and the dependent variable (George and Bennett 2005: 127–45). It is therefore necessary to examine the causal pathway in addition to statistical covariation in order to demonstrate the linkage between cause and effect. Rational choice, and above all game-theoretic, models are particularly limiting for empirical work, in spite of the relative specificity of their predictions. The reason for this is that each hypothesis derived from a game-theoretic model is actually a joint hypothesis about the assumed social and information structures, the procedure, a particular form of rationality and usually a specific equilibrium refinement (Kennan and Wilson 1993: 54). Moreover, empirical analyses of games of limited and asymmetric information have to confront the problem that the information known privately by one of the parties is usually inaccessible to the researcher (Bresnahan and Reiss 1991). For these reasons, three different types of empirical assessment are employed successively, following a progression of levels of analysis in increasing abstraction from the level of the individual to the level of aggregate structural data. First, a micro-qualitative analysis provides analytical narratives that demonstrate step-by-step how the hypothesized causal factors work their way through causal processes to eventually bring about the expected outcome – or not. Specifically, the theoretical model is evaluated in the context of qualitative evidence from two case studies of business lobbying on environmental policy and financial services regulation in Germany and the UK. The case-study design seeks out ‘hard cases’ in which the values of theoretically important contextual variables are distinctly unfavourable for the observed value of the dependent variable to occur. This makes it less likely that evidence supporting the hypothesized relationship will be found even though the null-hypothesis is in fact true. If found, such evidence provides stronger support of the hypothesis, since it was found to be true in an unlikely context (Mitchell and Bernauer 1998). The case studies also provide the opportunity to further refine the conditions under which business emerges victoriously from policy struggles. They thereby serve to calibrate the formulation of hypotheses for quantitative tests in subsequent chapters. Second, a micro-quantitative analysis explores observable implications of informational accounts of business influence for the interactions between lobbyists and policymakers. In rational choice analyses with asymmetric information, crucial explanatory factors cannot be directly observed (Kennan and Wilson 1993: 96). The micro-quantitative analysis therefore examines how observed fluctuations in lobbying incidence and the weight put on lobbying messages by policymakers can be explained by individual-level characteristics that are deduced from the theoretical model. Third, a macro-quantitative analysis is carried out to test observable implications of the informational model of business political power in a cross-country analysis of environmental regulation. Specifically, the strength of environmental regulation is related to the strength of structural dominance indicators, the

16 Introduction degree and leverage of business’ informational advantage and a number of control variables. By developing a model of the informational–structural power of business and evaluating it in the context of data on environmental politics and banking regulation, this book hopes to contribute not only to the study of interest group politics and lobbying, but more generally to an improved understanding of how politics works in democratic capitalism. However, with the exception of the case study on financial services regulation, the empirical analyses and other evidence provided in this book are mainly from the area of environmental politics. A few words are therefore in order to justify the choice of this policy area. Environmental policy addresses one of the most pressing problems facing the world today. Households and firms pollute the natural environment too much, mainly because they do not pay a price for the damage they inflict on the environment and thereby on the welfare of other citizens in society. The results are well-known: depletion of the ozone layer, rapidly disappearing rain forests, falling levels of ground water, acidification of the natural environment, global warming, climate volatilities and noise. Environmental policy is intended to contain these problems by establishing standards designed to control the material and energy outputs of consumers and producers to the biophysical environment (Sanchez 1997: 141). It aims to ensure that externalities, or the costs generated by a polluting entity that affect others, are borne by those who produce them. Throughout the world, people have increasingly come to believe that the environment is in serious decline and they support efforts to reverse the trend. Environmental issues acquire salience alongside other issues, and parties and candidates include corresponding pledges in their policy platforms. According to some scholars, these new themes become integrated into the established issue dimensions and party systems. In particular, social democratic parties are seen as having absorbed many of the ‘new politics’ demands, as distributive and redistributive issues become intertwined with so called ‘non-economic’ issues (Kitschelt 1994). For others (e.g. Inglehart 1997), the new politics issues occupy their own dimension. Whichever view one subscribes to, it is likely that an increasing number and saliency of new issues on the political agenda will produce incentives of one kind or another for parties to take a stand on them. Policymakers seem to respond to this demand not only with words but also by action, as environmental regulation has continued to increase worldwide since the early 1970s. For instance, in the US, pollution abatement control expenditures in manufacturing alone have increased by almost 140 per cent between 1979 and 1993 (Berman and Bui 2001: 498). Because environmental policy primarily aims at reducing negative externalities, it has cost implications for business. At the macroeconomic level, environmental policy is often claimed to compromise industrial productivity and competitiveness. Increasing sensitivity to global economic competition and budgetary constraints on the part of governments makes them wary of regulation which might threaten economic growth, foreign investment, export markets and employment creation (Golub 1998: 1).

Introduction 17 Regulations that require firms to reduce emissions, increase recycling, pay more for energy or switch to more expensive fuels and input materials all raise the final price of their products, with the result that ‘green’ states may lose markets to ‘dirty’ states that lack similar environmental standards (Golub 1998: 4). Governments are therefore concerned that imposing tough environmental policies could damage their domestic industries, either through loss of market share or by ‘flight of capital’ to ‘pollution havens’ abroad (Ulph 1997). Furthermore, there will be less political will to control environmental damage when the adverse effects are perceived as uncertain or when the costs of that damage will be borne by individuals living in a different political jurisdiction or in a different era (Grossman 1993). In the case of global commons problems such as transboundary pollution or global warming, international relocation of capital would not only be economically damaging but could also damage the environment of the countries that lost capital if the relocation of capital did little to reduce aggregate world pollution (Ulph 1997). For example, criticizing the British government’s announced target of a 20 per cent reduction of greenhouse gas emissions by 2010, the then directorgeneral of the Confederation of British Industry (CBI), Digby Jones, complained, ‘if our action is not matched by similar efforts from the rest of the world, we will undermine the competitiveness of British companies for no real environmental gain’ (Financial Times special supplement, ‘Understanding Business and Climate Change’, 15 February 2005). To the extent that governments share this perception of the environment–competitiveness nexus, they are induced to engage in a ‘race to the bottom’ or ‘ecological dumping’ (Golub 1998: 4). This negative link is not undisputed. While environmental pioneers may suffer short-term economic disadvantages in international competition, early movers in the area of environmental protection will be at an advantage in competition for innovative technologies. Furthermore, countries that succeeded in combining structural industrial change with an effective employment policy have been shown for quite some time now to be more successful in environmental protection (Jänicke 1992: 52). At the level of microeconomics, however, no matter what the society-wide benefits and even the long-term benefits to individual businesses may be, environmental policies add considerable compliance costs to firms. This may lead to cutbacks in research and development efforts, limit the innovative efforts of firms or even endanger their general profitability and survival in the market (Cairncross 1992: 108). As a result, firms tend to emphasize the costs of environmental policy, while underestimating the benefits, leading them to oppose environmental policy which they perceive to place them at a competitive disadvantage. This makes environmental politics an ideal area for the study of the extent – alleged or real – to which the political power of business is based in the structural dependence of the state on capital. Moreover, in this policy area, claims about high economic costs of policy and alleged or real trade-offs between environmental protection and economic performance are endemic, rendering it ideal for the study of the informational sources of business political power. Among other things, the analysis that follows will demonstrate how many false and exaggerated claims that are made by business can be exposed as such, while in other instances the informational

18 Introduction privilege of business will ensure that necessary environmental policies will not be implemented.

Key terms and concepts Liberal use has already been made of a number of big and potentially ambiguous terms, including power, democracy, capitalism, business and policymaker. These require definition if confusion is to be avoided at later stages. Other important concepts will be defined as they appear in the course of the exposition. As this book is primarily about power, the term is an obvious one to begin with. Much has been written about power and its various dimensions in the context of political analysis. While some of these conceptual problems will be revisited throughout the chapters that follow, the concept of political power that underlies the entire analysis is the simple, ‘one-dimensional’ concept proposed by Dahl, according to which ‘A has power over B to the extent that he can get B to do something that he would not otherwise do’ (Dahl 1957: 202–3). A crucial ingredient of this concept is the notion of intentionality. To say that A has power over B in this sense assumes that A intends to get B to do something that B would not otherwise do. While this may seem a trivial distinction needed at most to rule out inadvertent and accidental consequences of action that may be of little relevance to political life, the presence or absence of intentionality will become quite important throughout the analysis that follows. The requirement of intentionality can be incorporated in the above definition of power by stipulating that A has power over B to the extent that he can get B to do something that is desired by A that B would not otherwise do. Power in this sense is distinct from ‘influence’. While the two concepts are closely related and Dahl treats them as effectively synonymous, it is helpful to distinguish between the two. Influence is the wider term, denoting the extent to which an actor alters or impinges upon a political outcome in some way, which may or may not involve the exercise of political power. By contrast, power denotes the extent to which an actor brings about or accomplishes an outcome (cf. Morriss 1987: 29–35). Democracy can best be defined as a system for arriving at collectively binding decisions that guarantees equal effective rights of participation in the making of these decisions to each person subject to them (Hyland 1995: 75). The abstract notion of ‘equal effective rights of participation’ requires more exact clarification. Existing delineations range from minimalist definitions in which the role of citizens is largely restricted to electing representatives, to richer ones involving the active participation of citizens in policymaking. A minimalist definition of democracy is provided by Schumpeter, for whom ‘the democratic method is that institutional arrangement for arriving at political decisions in which individuals acquire the power to decide by means of a competitive struggle for the people’s vote’ (Schumpeter 1976 [1942]: 269). Richer conceptions include Pateman’s (1970) notion of participatory democracy with its focus on direct participation by citizens at the local level and the workplace, or Dryzek’s (2000) claim that the essence of democracy is deliberation rather than interest aggregation or self-government. For the present analysis, democracy is defined minimalistically as a political system

Introduction 19 in which the government is held accountable to popular control by means of fair and competitive elections of the rulers. In principle, any alternative definition of democracy would do just as well, but there are at least three advantages to applying a minimalist one. First, while a minimalist conception of democracy neither eliminates the need to think about institutional design nor precludes concerns about the quality of democracy (Przeworski 1999), it can serve as a lowest common denominator, thereby maximizing the domain of the analysis to follow. Second, this book is about the problems that capitalist enterprises can cause for democracy. There are many other factors that compromise the democratic process and its underlying principles – deliberationist and participatory critics of liberal democracy have identified some of them. But while it is possible to combine the critique of representative, polyarchal institutions with that of the capitalist organization of the economy (e.g. Green 1985), the identification of problems caused specifically by capitalism will be facilitated if other (political) institutions are relegated to the background and held constant where possible. Third, in stipulating acceptable minimalist criteria (that rulers be chosen through elections) a minimalist conception of democracy enables us to evaluate political systems as ‘more’ or ‘less’ democratic in proportion to the extent to which business interests are allowed to interfere with democracy’s main task of bringing government action in line with citizen preferences. While defining capitalism is a simpler task, care should be taken here as well to avoid unnecessarily fuzzy terminology. Some definitions place the emphasis on private enterprise, others focus on production for competitive markets, while Marxists emphasize the separation of labour from the means of production. And while all of these elements are important, they all also constitute ideal types: the fact that most existing capitalist economies allow for some role of public enterprises, display restrictions to competition or allow for some form or other of employee ownership does not render them non-capitalist. Lindblom’s (1977: 94–5) notion of ‘private enterprise market-oriented systems’ encapsulates all these elements without insisting on perfection. The same is achieved by Przeworski’s definition of capitalism, but with less awkward terminology. Following Przeworski, capitalism is defined as ‘any economic system in which (1) the optimal division of labor is so advanced that most people produce for the needs of others, (2) the means of production and the capacity to work are owned privately, and (3) there are markets in both’ (Przeworski 1991: 101). The terms business and business actor will be used to denote firms, their owners and senior managers, their public affairs representatives and lobbyists, as well as trade associations and peak organizations. Of course, in reality no two businesses are the same, and any single firm is usually not a unitary actor. But as the question pursued concerns the political power of the average business enterprise vis-à-vis the policymaking organs of the state and relative to citizens and non-business interest groups, indiscriminate use of the term ‘business’ as a singular noun will for the most part be warranted. More differentiated terminology will be used whenever the context of the analysis requires this. Lastly, analogous to the use of the summary term ‘business’, the level of abstraction at which the analysis takes place does, for the most part, not require careful

20 Introduction distinctions between government agencies such as legislators, regulators, executive ministers or civil servants. For the issues dealt with in this book, the crucial common element of all these actors is that they directly partake in the authoritative making of collectively binding decisions or the implementation of such decisions. They will therefore summarily be referred to as policymakers. Again, more differentiated terminology will be applied where the context of the analysis indicates that this is appropriate. For ease of exposition, business actors will often be referred to in the masculine form and policymakers in the female form. The opposite assignment of genders would have done just as well, and in the absence of conventions or other compelling criteria, the chosen assignment was determined by the flip of a coin.

Structure of the book This book analyses how in their political strategies business lobbyists employ privately held information about the effects of public policy in the real world. It examines the claim that capitalists’ privileged access to policy-relevant information is an important power resource alongside the more established factors such as money and organization, membership in elite networks or the prerogative over investment. To this end, a review of the main theoretical strands of scholarly work on business and politics – neo-pluralist group theory, theories of neocorporatism and organized capitalism, network approaches, theories of ideological dominance and structural power theories – will provide an overview of the state of the art in the study of business political power and ask how well the respective theories account for the reach and limits of business power in politics (Ch. 2). The review leads to the conclusion that none of the five explanatory strategies provides a satisfactory account of business political power. Attention is then turned to gametheoretic analyses of strategic information transmission that claim to provide clues about political influence where other theories fail. Using theory-informing data from interviews with lobbyists at German and Irish employers associations, Ch. 3 develops an informational–structural model of business political power in democratic capitalism that takes into account the main characteristic that distinguishes business from other interest groups: the prerogative over investment and allocation, and thereby over some of the most important decisions affecting people’s welfare and the fate of elected political decisionmakers. The chapter then analyses the conditions that lead to the political power position of capital by formulating a signalling game in which business communicates private information about the costs of a pending policy to the policymaker, who must infer the credibility of this signal from the observed levels of costly political action. The analysis shows that business power over public policy is contingent on the reputation costs of business in relation to its material costs of lobbying as well as on the policymaker’s reputation constraints from policy commitments and campaign pledges relative to the costs from adverse policy effects. Ch. 4. evaluates the model in the context of qualitative evidence from two case studies of business lobbying on environmental policy and financial services regulation in Germany and the UK. Ch. 5 analyses

Introduction 21 quantitatively a number of observable implications of the informational–structural model of business political power, using a micro-level dataset on the personal characteristics and interaction with lobbyists of over 1,400 members of the parliaments in 11 European countries. Deriving hypotheses about the behaviour of business lobbyists and parliamentarians, the chapter examines how far the reputation of lobbyists in the eyes of policymakers affects the frequency of lobbying and the willingness of policymakers to take lobbying messages serious. Ch. 6 extends the empirical analysis to the macro-level, testing hypotheses about the informational and economic–structural determinants of environmental regulation in 23 countries using OECD and World Economic Forum data. The concluding chapter (Ch. 7) summarizes the findings of the book, discusses them in the light of existing theories of business–government relations and outlines the scope and limits for addressing the problems of business power in public policymaking.

2

Groups, institutions, networks, ideology or structural dependence What drives business power?

Contemporary debates about the political power of business can be seen as a struggle to reconcile the contradictory relationship between two observations. First, aggregate-level political outcomes appear to be biased in favour of the interests of the owners and managers of capitalist enterprises. Looking at individual wealth and income, for example, this bias is visible in the ‘paradox of redistribution’. As a universal characteristic of democratic capitalist societies, rights in the political realm are distributed equally, while the actual distributions of income and wealth are always such that the median falls short of the mean (Przeworski 2003: 198–9). According to the median voter theorem, for any distribution of single-peaked preferences in a uni-dimensional political space there exists a unique majority rule equilibrium, which is the choice of the voter with the median preference (Black 1948). With respect to income distribution, the voter with the median preference is the one with median income. It would thus be in her interest to support political parties that promise, and carry out, a redistributive shift of the median income toward the mean. The puzzle then is why the poor, unmatched in numbers and equipped with the right to vote, do not take away from the rich. After all, it was the expectation that the masses would use their ballots as ‘paper stones’, which gave Engels such confidence that the peaceful transformation of social institutions towards more egalitarian arrangements was imminent (Przeworski and Sprague 1986: 1). Assuming that above-median incomes are associated with ownership and/or management positions in capitalist enterprises, students of business political power are left to search for answers to the paradox. Second, while business is widely acknowledged to be the most powerful group in society, it does not always get what it wants in politics. In formally democratic systems public officials must respond to a variety of heterogeneous demands. This implies strong political constraints on the ability of business actors to translate economic power into political power. Although the ‘veto power’ of private investment may be significant, this factor alone cannot guarantee a business-friendly policy response, in particular in periods of crisis and political innovation. Mitchell (1997: 7) suggests that the question of why business fails in the policy struggle is complementary to the question of why business has so much influence over policy. This chapter examines the literature on business and politics, widely conceived, for explanations of the sources and limits of business political power.

Groups, institutions, networks, ideology or structural dependence 23 Economic approaches to the study of business political power have been remarkably silent with regard to the sources and limits of business influence. In particular, they have ignored the agendas and incentives of political actors such as parties and policymakers. According to Stigler’s theory of policy formation, for example, policies are ‘acquired’ by industry and designed and operated primarily for industry’s benefit (Stigler 1975: 114). Known as the theory of interest group capture, or agency capture, this approach portrays business interests as seeking beneficial regulation from governments; that is, subsidies, control over market entry and price-fixing arrangements, as well as trying to avert policies that impose costs, such as consumer protection measures, environmental regulation or minimum wage legislation. Stigler expresses little doubt that business can generally expect to get what it wants. This confidence builds on Olson’s (1965) finding that relatively small groups such as business can effectively organize politically. Olson has shown that the relatively small number of group members combined with the concentration of benefits from collective action gives business much stronger incentives to organize for political action than larger groups, such as consumers or taxpayers, over whom both costs and benefits are widely dispersed. With group power being thus unbalanced in favour of business groups, politicians respond by setting up government agencies to look after these groups’ interests. The result is a ‘captured’ bureaucracy that is designed from the outset to promote special interests. Developing this approach further, Peltzman (1976) and Becker (1983) argue that large, diffuse groups can actually exert power, not least through voting at elections. This puts the pressure back on business to become more proactive if it wants to obtain desired policies. This can in turn be met with counter-pressure, leading to political outcomes that are more pluralistically determined than Stigler predicted (Moe 1997: 462). But while Becker and Peltzman pay more explicit attention to the role of lobbying, they too abstract from the individual behaviour of politicians. Becker, for example, specifies an ‘influence function’ that links policy outcomes directly to lobbies’ contribution payments, where the lobbying effort is determined endogenously. He finds, unsurprisingly then, that lobbying increases if the loss of redistribution becomes more important. Applications of this model continue to dominate research on international trade policy, but are also used in the context of environmental policy research. The problem of this approach is that it does not work well empirically. Comparing air quality regulation across the fifty US states, for example, Ringquist (1993) found that rather than being captured by economically advantaged interests, state governments appear to be responding to the real or perceived environmental threats as well as to economic and technological pressures associated with these industries. Approaches in political science, sociology and business and management studies generally assign more explicit roles to the agendas and actions of political parties and elected policymakers in their analysis of the political power of business in capitalist democracies. Five analytically distinct explanatory frameworks can be identified that have produced a variety of often conflicting explanations for the successes and failures of business attempts to prevent popular demands from

24 Groups, institutions, networks, ideology or structural dependence encroaching upon its interests. First, theories in a broadly pluralist approach to the study of interest group politics analyse the political activities of business, focusing on lobbying (Hillman, Keim and Schuler 2004), party and campaign finance (Clawson, Neustadtl and Weller 1998) and corruption (Rose-Ackerman 1999). In explicit contrast to pluralist models, neocorporatist theory has provided a framework for analysing business power that highlights the role of the institutional environment of business–government relations (Schmitter 1981). This, and the related distinction between liberal and co-ordinated market economies (Hall and Soskice 2001), will be scrutinized for its implications for business political power. A third approach emphasizes the commonality of perceptions and preferences as well as social interactions among economic and political elites (Mizruchi 1992; Domhoff 1998). A fourth strand of inquiry concerns the ability of business to influence public opinion (McClosky and Zaller 1984). The fifth type of theory concerns the non-intentional, structural constraints on the policymaking capacities of elected officials that arise from the capitalist mode of production. Studies within this latter field are concerned with the controversial and empirically elusive claim that the owners and managers of private enterprises enjoy a structurally privileged position that enables them to secure favourable political outcomes even if they abstain from political action. The argument has become known, variously, as the ‘structural dependence of the state on capital’ (Przeworski and Wallerstein 1988), the ‘business confidence’ factor (Block 1977) or the ‘privileged position of business’ (Lindblom 1977). The following sections review the five approaches in turn, assessing their respective contribution to explaining why business gets what, when and how in democratic politics.

Pluralism and the study of business as an interest group In the pluralist paradigm, government is an arena in which different social and economic interests compete to be heard by elected officials. In this arena, business and other interest groups aim to secure favourable political outcomes – and avert unfavourable ones – by influencing the formation of public policy, its passage through the legislature and its implementation. In Polsby’s (1980: 118) words, the ideal pluralist commonwealth would present itself as ‘fractured into congeries of hundreds of small special interest groups, with incompletely overlapping memberships, widely differing power bases and a multitude of techniques for exercising influence on decisions salient to them’. The techniques for exercising influence are the various forms of lobbying; that is, contacting and pressurizing legislators, ministers and civil servants. Lobbying is seen as a way of exercising pressure on political decisionmakers. Pressure, in turn, is conceived of in the sense originally identified by Bentley (1993 [1908]: 19) as the ‘push and resistance’ between groups. While this notion of pressure is broad enough to include ‘all forms of group influence upon group, from battle and riot to abstract reasoning and sensitive morality’, it refers primarily to political action, even when, as Bentley is careful to point out, ‘tendencies to activity are pressure as well as the more visible

Groups, institutions, networks, ideology or structural dependence 25 activities’ (ibid.: 20). In this view, political decisionmakers are either subjected to action by interest groups or face the threat of future action. The power of an interest group is then its ability to exert pressure. This ability varies with the size and type of a group’s membership, its financial resources, monopolistic control of expertise and information, status and access to government and the media, the capacity to influence public preferences and its organizational structure. The institutional organization of the government and the state bureaucracy are also relevant for group strength. The pluralist tradition in the study of interest group politics has been firmly committed to this concept of lobbying as pressure politics. The emphasis on political action in Bentley’s notion of pressure politics is exemplified in his description of the political activities of firms and their management: The directors of a company may finish their ordinary business and turn at the same meeting to discuss the part the corporation will take in the next political campaign. Their activity, which a moment before was industrial or economic, then becomes at once political – a part of the governing process of the country – and is to be studied specifically as such. ibid.: 21–2 Thus, in the pluralist perspective, corporate directors and executives undergo a transformation of identity from economic decisionmakers, with little concern for politics, to political actors who pursue the distinctly political aims of their firm or their industry. In this perspective, the immediate objective for interest group leaders and lobbyists is access to political decisionmakers. Interest groups cannot be successful in their endeavours if they lack access to one or more of the key points of decisionmaking in the government (Truman 1963: 264–70). In their study of American interest group politics, Schlozman and Tierney (1986: 104) describe access as ‘at once the most cherished resource and the most pressing objective of the Washington Lobbyist’. For pluralists, the question of business political power therefore turns into the dual question of whether business enjoys privileged access to policymakers and whether it can make more of these encounters than non-business groups. Anecdotal evidence suggests that economic elites find it generally easier to gain access to politicians than other groups. For example, guests at a fundraising dinner with the governor of California, Arnold Schwarzenegger, in February 2004 were expected to donate anything between $50,000 and $500,000 for a seat at the meal (The Guardian, 21 February 2004). While at the time this dinner was billed as the biggest fundraiser in the history of politics, similar events in the US regularly command expected donations of five and six digit sums. Clearly, few people are in a position to attend this kind of event who are not also members of an economically advantaged group – a characteristic that, in Western capitalist countries, is strongly correlated with ownership of, and leadership positions in, business enterprises.

26 Groups, institutions, networks, ideology or structural dependence Yet not all political parties, candidates and policymakers might manage their accessibility in this way. It seems reasonable to assume that left-leaning politicians, for example, prefer to communicate with non-business interests, such as trade union representatives, or indeed ordinary citizens. And research undertaken by Chin, Bond and Geva (2000) on lobbying in Washington DC suggests that politicians of any partisan affiliation value ordinary constituents’ interests no less than business and money givers. Through interviews with Washington staffers, they found evidence that members of Congress give priority to constituent requests over the Political Actions Committees (PACs) of firms and labour unions. According to one of the staffers they interviewed, ‘[f ]olks who are with organizations but are not constituents are very low priority’. In the staffer’s view, this is done for purely electoral reasons: ‘Re-election depends on honoring … constituent matters’ (quoted in Chin, Bond and Geva 2000: 545). The staffer also pointed out that ‘[i]nterest groups may use PAC status to get a foot in the door, but constituents generally find that door open when they call’ (ibid.). In European countries, where local constituency interests are generally less central to the work of national parliamentarians than in the US or the British Isles, comparisons between producer and non-producer groups may be more apposite than the opposition between constituents and non-constituents. To the extent that they do enjoy preferential access to policymakers, can business actors translate these opportunities into political power? While conceding that business has superior political resources such as money, organization and status, pluralist sceptics of business power theories argue that the existence of these resources does not imply that they are effectively employed or actually make a difference for political outcomes (Wilson 1981: 37). Elected politicians may talk to just about everybody in their constituency and listen to their concerns, but that does not tell us much about how they weigh the information they receive. Meeting the governor of California for dinner may be quite a different thing than influencing his politics. Labour unions highlighted that Schwarzenegger’s plans for partial privatization of California’s employee pension system were backed by powerful corporations, including Disney and Citigroup (The Guardian, 5 March 2005). But does that mean that industry backing was the cause of this political move? And even if it was, there is no guaranteeing the success of business strategy: in the event, the industry’s support would turn out insufficient to bring the reform project to completion. Schwarzenegger abandoned the plan in 2005 in the face of stiff opposition from labour unions and community groups. But be it business or labour that carries the day – how exactly do interest groups influence policy? Determining the influence of lobbyists is complicated by the difficulties of identifying when influence has occurred and when there is really only parallelism of purpose and action between the lobbyist and government official (Salisbury 1975: 207). This problem has been addressed early on by Dahl, who highlighted the need for focusing upon observable conflict arising between agents. In a meticulous analysis that has become a classic of the pluralist literature, Dahl (1961) investigated the influence of various interest groups over municipal policymaking in

Groups, institutions, networks, ideology or structural dependence 27 the city of New Haven, Connecticut. In Who Governs?, Dahl identified three political key issue areas in New Haven in the second half of the 1950s: urban redevelopment, public education and nomination for local office. He then tried to find out whose preferences prevailed against what opposition in the respective decisionmaking processes. Based on evidence obtained through lengthy interviews with nearly 50 individuals who had actively participated in one or more of the decisions on these issues, three surveys on several ‘subleaders’, as well as two surveys on voters, and the perusal of minutes of the meetings of the political bodies involved, Dahl concluded that no single interest group in New Haven wielded any disproportionately high degree of power over more than one issue area. At the same time no group of individuals was entirely powerless. This led Dahl to conclude that in New Haven, and, by further inference, elsewhere in the US, power over public policy was widely dispersed and non-centralized. Findings like this form the empirical basis of pluralism’s central claim that a diversity of groups wield non-cumulative power in only one or a few policy areas at a time (Dahl 1971). Because of its focus on observable decisionmaking behaviour, the pluralist approach to the study of group power is also referred to as the ‘decision method’ (Mokken and Stokman 1976: 52). Its core methodological principle is that only observable political action or behaviour in the decisionmaking arena counts as valid evidence about power. Pluralist theory does not recognize the existence of non-overt, non-observable power, and demands an a priori assumption of the absence of power relations. In an essentially Hobbesian manner, all interests are presumed to be fundamentally equal; everyone has some access to policymakers; and, in the end, no one view or interest would become dominant. A major factor underlying this pluralist state of affairs is the observation that business is itself not a monolithic block. Depending on their size and sector, firms may have few interests in common, and there are good reasons to expect that the relationship among different firms in many areas of public policy is characterized as one of competition and disunity. The pressure group view of business political power has received powerful criticism. Early on, Bachrach and Baratz identified two major weaknesses of the pluralist method of studying power. First, it does not allow for the possibility that power is exercised by confining the scope of political decisionmaking to issues that are of little or no relevance to the holder of power (Bachrach and Baratz 1970: 6). Second, and closely related, these authors claim that the decision method provides no objective criteria for determining important – that is, ‘key’ – issues (ibid.: 7). As a remedy to these conceptual shortcomings of the pluralist method, Bachrach and Baratz offer the concept of non-decisionmaking as the ‘second face of power’. According to this notion, ‘to the extent that a person or group – consciously or unconsciously – creates or reinforces barriers to the public airing of political conflicts, that person or group has power’ (ibid.: 8). To identify instances of second-dimensional power, the actual decisionmaking process within the political arena and the resultant outcomes have to be mapped, before remaining grievances on the part of the actors are determined. Finally, the means ‘by which

28 Groups, institutions, networks, ideology or structural dependence some or all of the potential demands for change have been denied an airing’ should be identified (ibid.: 49). With respect to the problem of what constitutes a key issue, Bachrach and Baratz refer to Schattschneider’s (1975 [1960]) notion of the ‘mobilization of bias’. In their words, mobilization of bias is achieved by ‘the dominant values and the political myths, rituals, and institutional practices which tend to favor the vested interests of one or more groups, relative to others’ (Bachrach and Baratz 1970: 8). There are conceptual and methodological problems with this two-dimensional view of power. First, Bachrach and Baratz do not clearly distinguish between nondecisions arrived at by purposive action on the one hand and those evolving from unintentional processes or structural features on the other. Second, their proposed technique for identifying the use of second-dimensional power is likely to lead to inconclusive results or, alternatively, revert to an account of the power resources central to the one-dimensional analysis of power. Apart from the important addition of the agenda-setting dimension, it is difficult to see how their method differs from the resource-oriented, one-dimensional analysis of power without at the same time invoking potentially unobservable structural variables. In the latter case, nondecisions and latent conflicts would be the only ‘evidence’ for the existence of the second dimension of power, which may be persuasive but not conclusive (Barnes 1993: 200).1 Bachrach and Baratz’s is not the only critique of pluralism. Empirically, the trouble is that studies of the effects of interest group activities on political outcomes produce inconsistent and non-robust findings. In particular, researchers have found it difficult to explain legislative and regulatory outcomes by reference to overt political pressure from interest groups, business or other, or to the differential resource endowments of groups as measured by their membership, strategic position, organizational strength or simply the amount of money injected into a lobbying campaign (Baumgartner and Leech 1998: 106–8; Smith 2000: 115–41). The enormous research output generated in relation to the political effects of corporate campaign contributions is a good example of the general failure in pluralist research to connect political inputs (e.g. money) to outcomes (i.e. policy). Material contributions to candidates or legislators generally neither account for election results nor for the policymaking behaviour of the recipients (Roscoe and Jenkins 2005). Given that in most countries such contributions are legally restricted to a few thousand US dollars at a time, while larger sums that may change hands illegally are notoriously difficult to trace, this is perhaps not very surprising. Very few authors deviate from this diagnosis (but see Sabato 1984). As a result, pluralist theories of interest group politics have in general proved unsatisfactory for explaining policy change. From a theoretical vantage, Morton and Cameron (1992) highlight the importance of policymakers’ ideological and electoral incentives. They suggest that one of the reasons for the many weak and contradictory findings in studies on the policy effects of interest group activity is the implicit, unrealistic assumption that voters condone sell-outs by incumbent politicians to organized interests.

Groups, institutions, networks, ideology or structural dependence 29 This qualification applies a fortiori to big business. The objection echoes Truman’s earlier claim that the efficacy of pressure politics by sectional interest groups is ultimately circumscribed and constrained by largely unorganized interests that are so widely held in society that they are ‘taken for granted’ (Truman 1963: 512). Describing these interests variously as a ‘general ideological consensus’, a ‘broad body of attitudes and understandings regarding the nature and limits of authority’ or simply the ‘rules of the game’, Truman holds that the flouting of these rules will result in ‘organized interaction and the assertion of fairly explicit claims for conformity’ on the part of the latent interests, that is, the silent majority of ordinary citizens (ibid.: 512). Thus, in the event that powerful organized interests manage to capture the policymaking process, the otherwise unorganized interests would mobilize in order to claim conformity with the democratic rules of the game. Because of their desire for re-election, political representatives will be responsive to these demands, being thus constantly held accountable by Truman’s ‘taken for granted’ interests. In the pluralist tradition, this mechanism provides the ultimate check against the domination of public policy by organized sectional interests, including business. The claim that the latent interests of the silent majority will become actual and vocal if they find themselves exploited by special interests has in turn met powerful criticism in the work of Olson. In a passage that is worth quoting in full, Olson identifies the freerider problem of collective action as the source of the policy bias toward special interests: Since relatively small groups will frequently be able voluntarily to organize and act in support of their common interests, and since large groups normally will not be able to do so, the outcome of the political struggle among the various groups in society will not be symmetrical. Practical politicians and journalists have long understood that small ‘special interest’ groups, the ‘vested interests’, have disproportionate power … The small oligopolistic industry seeking a tariff or a tax loophole will sometimes obtain its objective even if the vast majority of the population loses as a result. The smaller groups – the privileged and intermediate groups – can often defeat the large groups – the latent groups – which are normally supposed to prevail in a democracy. The privileged and intermediate groups often triumph over the numerically superior forces in the latent or large groups because the former are generally organized and active while the latter are normally unorganized and inactive. Olson 1965: 127–8 Olson further illustrates this argument by invoking V.O. Key’s image that the lobbyists for electrical utilities are ‘eternally on the job’, while the lobbyists for the consumers of this oligopolistic service are ‘conspicuous by their absence’ (Key 1964: 166). Not only is a small group of firms in an oligopolistic industry in a better position to overcome free-rider incentives, but in many cases the entire collective action problem of obtaining public goods does not apply in the first place.

30 Groups, institutions, networks, ideology or structural dependence This is because much political activity of firms is directed at obtaining ‘private goods’ (Godwin and Seldon 2002). Rather than collectively opposing a policy wholesale, individual businesses engage in ‘loophole lobbying’, concentrating on very specific aspects of the policy with the aim of minimizing the cost implications for themselves (Polk and Schmutzler 2005). For instance, an economy-wide energy tax can be designed in many different ways, with exemptions for various sectors. This leaves considerable scope for loophole lobbying of specific sectors or individual firms. Godwin and Seldon (2002: 216) report, for example, that airline lobbyists in Washington DC spend ‘75–95 per cent of their time on issues affecting only their firm or their firm and one other’. And yet, in a capitalist democracy characterized by regular free elections and at least some degree of competition in most markets, there remain strong constraints on the ability of elected politicians to cater to special interests. The main reason is that rational voters can be expected to see through political attempts to curry favour with special interests at their expense (Lohmann 1998: 810). A disaffected majority aware of such a policy bias could punish its political representatives by voting them out of office. Therefore, the idea of small, efficient groups hijacking the policy process at the expense of latent groups has to rely on some degree of voter ignorance. Political incumbents have to be able to systematically fool a majority of voters by favouring special interests while keeping the costs imposed on the majority below some awareness threshold (Becker 1983: 391–4). This might be plausible to some degree: because information gathering is costly, voter rationality implies incentives for the electorate to be ill informed (Downs 1957). But because citizens avail of low-cost information short-cuts as by-products of engaging in various social and economic activities in daily life, voter ignorance can only be relied upon to a certain extent. At a minimum voters accumulate low-cost information about their personal wealth and well-being. ‘Even if they are ill-informed about policy detail, they nonetheless know how well off they are, and they can base their vote on this knowledge’ (Lohmann 1998: 810). Indeed, one robust empirical regularity observed across industrialized democracies is the phenomenon of retrospective voting. Incumbents who preside over good economic performance tend to get re-elected (Fiorina 1981). Furthermore, Lupia (1992) has shown that in an election year otherwise ill-informed voters vote by and large as if they had full information. All it takes for this to happen is that a few fully informed voters exist, that parties and candidates know that they exist and that candidates know the platforms of their competitors. It then suffices for candidates to devise median platforms and voters to vote for the most preferred candidate. This, of course, brings us back full circle to Morton and Cameron’s (1992) aforementioned critique that policymakers and business interests should not be assumed to get away easily with circumventing the democratic process. Should we therefore agree with the pluralists that business does at least not wield extraordinarily large amounts of political power in comparison to ordinary citizens, thus posing no real threat to democracy? While the preceding arguments provide strong reasons for such a conclusion, the paradox of distribution remains as a strong reminder that,

Groups, institutions, networks, ideology or structural dependence 31 at least in this important policy area, business interests fare remarkably well as far as political outcomes are concerned. Critics of the pluralist perspective remind us of business’ ideological influence, its structural dominance deriving from control over society’s economic assets and the superior resources business can commit to the usual channels of interest group politics. It is also important to note that neo-pluralist accounts by McFarland, Vogel and Wilson have paid much greater attention to business power than did earlier pluralists. In their view, business may not be the ‘dominant power group’, but it is seen as ‘the most important power group’ (McFarland 1991: 271). Neo-pluralists have also alerted us to the cyclical fluctuations involving a succession of periods of business political control during times of political reform, when countervailing groups successfully challenge business power (ibid.: 263). In Vogel’s (1989) view, the performance of the economy and the amount of unified business political action jointly determine the political effectiveness of business. Good economic performance is said to decrease business political power as the public raise their expectations about what business can afford. Furthermore, Salisbury (1984) finds that business has a systematic advantage over others in politics because business corporations and trade associations are ‘institutional groups’, who tend to be able to sustain a more permanent presence in the policymaking process. This is particularly the case in politico-economic systems that grant formal deliberative and consultative rights to business in exchange for information and implementation of policies on the ground. The institutional mechanisms that achieve this integration will be looked at next.

Neocorporatist systems and co-ordinated market economies Neocorporatism is a political system in which organizations representing major economic interests, such as employers, trade unions and, less centrally, agricultural interests, are granted privileged opportunities of participation in the formulation and negotiation of public policy, in particular economic policy (Wilson 2003: 102–3). This is seen in contrast to pluralism, which represents the more loosely structured competitive systems of interest intermediation that prevails when no corporatist structures are in place (Siaroff 1999: 177). Past analyses of neocorporatism have tended to involve two different uses of the term. The first refers to neocorporatism as a system of interest representation. According to Schmitter, [c]orporatism can be defined as a system of interest representation in which the constituent units are organized into a limited number of singular, compulsory, non-competitive, hierarchically ordered and functionally differentiated categories, recognized or licensed (if not created) and granted a deliberate representational monopoly within their respective categories in exchange for observing certain controls on their selection of leaders and articulation of demands and supports. Schmitter 1974: 93–4

32 Groups, institutions, networks, ideology or structural dependence The second, and perhaps more common usage, is that of neocorporatism as an ‘institutionalized pattern of policy-formation’, above all ‘in the shaping of economic policy’ (Lehmbruch 1979: 150). While it has been argued that the first usage refers to the formal organization of the interest group system, while the second one is about the effective role of interest groups in the process of policy formation, and thus about ‘concertation’ (Schmitter 1982: 263–4), empirically the two strongly overlap. For this reason, Siaroff suggests an encompassing view of neocorporatism. He identifies the main features of such a unified concept of neocorporatism at the national level as reflecting, ‘within an advanced industrial society and democratic polity, the co-ordinated, co-operative, and systematic management of the national economy by the state, centralized unions, and employers (these latter two co-operating directly in industry), presumably to the relative benefit of all three actors’ (Siaroff 1999: 177). The reference to co-ordinated management suggests the proximity of neocorporatism to the conceptual apparatus of the ‘varieties of capitalism’ school (Hall and Soskice 2001). This approach argues that different national economic institutions offer different opportunities to firms, who then adjust their production strategies to these opportunities. Two broad groups of national economies can be distinguished: co-ordinated market economies, such as Germany or Sweden, and liberal market economies, such as the UK and the US. In co-ordinated market economies, labour-market institutions such as works councils and industry-wide collective agreements in conjunction with expectations of low and long-term returns to investment promote the provision of firm-specific skills and facilitate a long-term outlook among companies (ibid.: 22–7). In liberal market economies, by contrast, firms are relatively free to hire and fire workers as they see fit while being under pressure to deliver fast and high returns as a result of finance with a strong role of equity capital. This discourages firms from pursuing production strategies based on promises of long-term employment (ibid.: 27–33). The former systems tend to be observed in countries with neocorporatist systems of interest intermediation, while the latter tend to concur with political systems characterized by strong, centralized executive power embodying significant discretionary strength vis-à-vis business (ibid.: 49). According to the varieties of capitalism school, this conjuncture is no coincidence. In neocorporatist systems, governments are highly constrained. In Germany, for example, responsibility for many policy areas of key interest lies outside the boundaries of the state and is instead reserved for ‘para-public institutions’ such as an independent central bank and chambers of commerce vested with significant public power (Wood 2001: 254). These features, combined with consensus–democratic political institutions and a federal constitution make the German state a truly ‘semi-sovereign’ one (ibid.: 254). The resulting constraints on government provide firms with the expectation that the para-public institutions for self-government will remain in place, which in turn reduces the risks otherwise attached to sharing information with other firms and the government, and with co-operating in collective action to further common long-term goals. In short, neocorporatist institutions, in conjunction with consensus–democratic political institutions, facilitate the development and maintenance of co-ordinated

Groups, institutions, networks, ideology or structural dependence 33 market economies by empowering business and putting intuitional constraints on government capacity (ibid.: 249–60). A pluralist and majoritarian system such as Britain, by contrast, does not have these constraints, leading to powerful singleparty governments (ibid.: 255). As a result, governments in pluralist systems are unable to credibly commit to self-restraint. This provides companies with strong incentives to withhold rather than share information, and to resist other behavioural requirements of co-ordination (ibid.: 259). These accounts of the institutional environment in which business and policymakers interact have three implications for the political power of business. First, neocorporatist institutions and practices are relevant not only to concerns of collective wage-bargaining and industrial peace. They are associated with consensus – as opposed to majoritarian – democracy (Lijphart and Crepaz 1991), and jointly these systems shape government–business relations in any area of public policymaking where business interest are concerned. Second, neither the corporatist–pluralist dimension nor the dimension of co-ordinated-versus-liberal market economies fully captures the institutional features that are relevant for co-ordination and political influence. Several attempts have been made in the literature to summarize the relevant features into a more encompassing dimension of integration of policymaking and government–business relations. Hicks (1988: 149–52) suggests a framework of national collective economic action which involves grouping national political economies into five types, including the ‘capitalist pluralism’ of the US and the UK; the ‘capitalist statism’ of Japan, France and Italy; the ‘social democratic corporatism’ of Austria, Denmark, Finland, Norway and Sweden; the ‘consociationalism’ (or ‘liberal corporatism’) of Switzerland, Belgium and the Netherlands; and the ‘quadripartite corporatism’ of Germany. Building on these refinements, Siaroff (1999: 189) proposes a general concept of integration defined as ‘a long-term co-operative pattern of shared economic management involving the social partners and existing at various levels such as plant-level management, sectoral wage bargaining, and joint shaping of national policies in competitiveness-related matters (education, social policy, etc.)’. Because of its encompassing and yet well-defined delineation, Siaroff’s concept of economic integration solves many of the problems that have tainted previous measures of the concept by combining all important aspects of the pluralism–corporatism contrast (Lijphart 1999: 176). It will therefore be used throughout the remainder of this book. Third, the degree of political–economic integration tells us something about the extent to which business is powerful in the context of public policymaking, but it is not exactly clear what. Recall, on the one hand, the definition of neocorporatism given at the beginning of this section. Clearly, being granted privileged opportunities of participation in the formulation and negotiation of public policy means being given considerable power and influence over public policy. While these privileges will usually have to be shared with organized labour, they are also confined to producer interests – at the expense of other societal interests (Offe 1981: 128–9). Thus, neocorporatist arrangements can lead to a marginalization of interests that are not efficiently represented by the major interest organizations. But the gain in business

34 Groups, institutions, networks, ideology or structural dependence power thus achieved should not necessarily be understood in a zero-sum redistribution of power at the expense of government policymakers or interests outside the neocorporatist institutions. While neocorporatism empowers firms vis-à-vis government, this very empowerment also facilitates policy co-ordination and concertation. It biases the policy options of governments towards policies that are incentive-compatible with existing institutional complementarities, and thereby towards policies that are more likely to succeed (Wood 2001: 258). By contrast, the attributes normally associated with the ‘strength’ of a state may turn out to denote weakness, preventing governments in pluralist systems from implementing many kinds of policies effectively (Hall and Soskice 2001: 48). This ambiguity of neocorporatist institutions and co-ordinated market economies for business political power will remain unsolved for the moment but shall be revisited later in this study. In the meantime, another less formally institutionalized dimension of integration among business actors and policymakers must be examined.

Elite networks and policy communities While pluralists hold that the basis for polyarchal pluralism in advanced capitalist countries is the existence of conflict and fragmentation among elites, they acknowledge the risk of elites acquiring political power and achieving dominant positions if these elites manage to co-operate or even unite. According to Dahl, ‘[t]he actual political effectiveness of a group is a function of its potential for control and its potential for unity’ (Dahl 1958: 465, original emphasis). And Vogel (1989: 12) finds that ‘[w]hen business has been unified, its political power has often been extremely impressive’. Pluralists generally seek solace in the fact that business unity is the exception rather than the norm, not least because of the structural properties of the market, which lead to constant competition among businesses. Potential sources of conflict include relations between financial and manufacturing firms, manufacturing versus retail, utilities versus corporate consumers, importing versus exporting industries, etc., to which regional conflicts and competition within industries should be added (Mizruchi and König 1986: 482–3). For neo-pluralists, business unity (and, by implication, power) varies cyclically, thus allowing non-business interests to advance their goals more effectively in periods of weak business unity. According to Mizruchi and König (ibid.: 482), business unity concerning political strategies is contingent on structural factors like market dominance and inter-firm transaction volumes. Elite theorists acknowledge these tendencies toward fragmentation but emphasize numerous mechanisms that facilitate cohesion and unity among the business community and, by further inference, its mobilization and power. Interlocking directorates and overlapping board memberships, business associations and private club memberships enable businesswomen and men to recognize their common, class-wide and long-term interest. Furthermore, these networks facilitate and promote collective political and social activity of corporations. These mechanisms, it is argued, help the business community as a whole to overcome the divisive implications of intra-class competition and sectorally determined

Groups, institutions, networks, ideology or structural dependence 35 conflicting interests (Bearden and Mintz 1987). Consequently, much empirical elite research documents the historical mobilization of dominant segments of the capitalist class to transcend their own individual interests in order to act politically for broader class-wide goals, especially in periods of crisis (Domhoff 1998). Sociometric network analyses have repeatedly found that American business elites constitute highly integrated social groups. Bearden and Mintz (1987) and Mizruchi (1982) have shown how firms are tied together by economic interdependence, common ownership ties and interlocking directorates. This integration, Mizruchi concedes, is partially due to a few very large firms within the most highly concentrated industries. However, Mizruchi is keen to emphasize that it is not so much firm size or market concentration per se, but rather corporations’ economic, organizational and social interactions that unify and empower them. A coherent and tightly organized business community might of its own not pose very serious problems for democracy – as long as economic elites can be assumed to constitute groups and encompassing individuals that are largely distinct from public decisionmaking bodies and their members. Of course, personal unions occur from time to time between political and business empires, as in the case of the former Italian prime minister, Silvio Berlusconi. Perhaps more frequently are direct family ties between political and business leaders. One of many examples is the marriage between the former Irish deputy prime minister and minister for enterprise, trade and employment, Mary Harney, and the former economic affairs director of the Irish Business and Employers Confederation (IBEC), Brian Geoghegan. And lastly, business leaders are frequently appointed to ministerial posts – an example is the appointment of the former director-general of the CBI, Digby Jones, as minister of state in the newly created Department for Business, Enterprise a Regulatory Reform in June 2007. Overall, however, personal unions or intimate personal ties are the exception, and there are usually separate career paths for business and political leaders. But the same sociological mechanisms that are found to forge bonds among the members of economic elites are also claimed to integrate economic elites with politicians and civil servants. As a result of these processes, the captains of industry share common social and sometimes professional, educational and recreational experiences with politicians and leading civil servants. This theme has of course been around since the elite theories of Hunter (1953) and Mills (1956). In The Power Elite Mills insisted that American politics was run by three small groups of elites residing in the Pentagon, the executive branch of the federal government and the boardrooms of America’s largest corporations. According to Mills (1956: 11), these elites derive their power from the fact that they ‘form a more or less compact social and psychological entity; they have become self-conscious members of a social class’. Mills’ thesis was later canonized by US president Eisenhower, who in his 1961 farewell address voiced concern about the power of the ‘military–industrial complex’ undermining American democracy. The notion of a power elite and its unity rests upon fundamental commonalities of interest among economic and political elites. It also implies a similarity of origins and outlook, all due to the social and personal intermingling of the top strata of

36 Groups, institutions, networks, ideology or structural dependence the constituent elites. Consequently, much empirical power elite research focuses on the aspects of education and socialization that induce similar outlook and sentiments among future members of both the political and economic elites. Cookson and Persell (1984) document the historical role performed by elite preparatory schools in the formation and maintenance of the American upper class. Historically, the collective identity forged in these prep schools would become the basis of upper-class solidarity and consciousness – a process of elite integration that is continued in the Ivy League universities. In the UK, similar functions are performed by ‘public’ schools and the Oxbridge colleges, while in France the École Nationale d’Administration is the unique cadre production facility that no aspiring politician, civil servant or manager can afford to bypass (Miliband 1969: 55–67). In an empirical assessment of the central claims of elite network theory, Moore (1979) gauges the degree of integration among economic and political elites in the US. Using data from interviews with over 500 leaders of major political, economic and social institutions, Moore uses formal network analysis to investigate the structure of an elite interaction network. Her results demonstrate the existence of largely unfragmented elites in different institutional and issue-related contexts. The structure of the networks indicates that at least a strong potential for unity among elites exists, and that almost all of the individuals involved have a high potential for political influence due to their incumbency in high-level positions. Thus, the structure exists through which elites in various institutions could become unified in the pursuit of common goals. Crucially, however, Moore also finds that these networks are formed around specific issues, and that membership in a network is not so much the result of privilege from birth or upbringing, but derives from the issue-related interests shared by the members of the network. Contradicting the central claims in the work of Domhoff, Hunter and Mills, Moore’s findings suggest that the integrated network in her data is not based on similarities in the social origins and affiliations of its members, to say nothing of upper class origins (ibid.: 689). Rather, the results correspond to the earlier political science literature on policy communities. In the 1950s and 1960s, this literature has identified the ability of groups to shape the particular part of government of concern to their interests as an important point of leverage for interest group pressure. It is argued that economic elites proceed by developing relationships and entering into alliances with the relevant branches of government such as ministerial civil servants and parliamentary committees, creating what became known as ‘islands of functional power’, ‘policy subsystems’ or ‘iron triangles’ (Maass 1951; Sayre and Kaufman 1965; Redford 1969). These functional, rather than social, ties between the members of economic and political elites are reinforced by the politics of the ‘revolving door’ (Salisbury et al., 1989). Private firms find it attractive to hire former senior civil servants, regulators and even cabinet ministers for their technical expertise and social connections. Once in their new private sector positions, the former policymakers use their connections and influence to further their new employers’ political interests, as well as using their factual expertise to advise them on government relations and regulatory matters. Similar passage in the opposite direction also occurs. A prominent

Groups, institutions, networks, ideology or structural dependence 37 example of someone who has travelled the revolving door full-circle is the US vice president, Richard Cheney. A former defence secretary in the George H.W. Bush administration from 1989 to 1993, Cheney became chief executive of the transnational energy company Halliburton in 1995, only to find his way back into the federal administration in 2000 under his former boss’ son, George W. Bush. This mechanism can be found in virtually all capitalist democracies. In the UK, for example, there is an established revolving door between the Foreign Office and petroleum firms such as Shell: four of the last five permanent heads of the Foreign Office have moved on to become directors of oil and gas companies; two of them at Shell (The Guardian, 15 May 2007). According to Heclo (1978), the functionally more important phenomenon are socalled ‘issue networks’; that is, shifting coalitions involving policymakers, interest groups and professional lobbyists. Working within this approach, Carpenter, Esterling and Lazer (1998) investigate the working of ‘weak ties’ between and across networks, as opposed to ‘strong ties’ between people that know each other well, as implied in the accounts of ‘old boys’ networks by Domhoff and others (e.g. Domhoff 1975). Using network analysis, they found that lobbyists’ access to political decisionmakers depends crucially on these weak contacts. In a similar vein, the ‘policy advocacy coalition’ approach of Sabatier and Jenkins-Smith (1993) analyses how different interest groups temporarily join forces to advance their common interests. By contrast, Richardson and Jordan’s (1979) analysis of elite integration in British ‘policy communities’ encompassing civil servants and business actors suggests something altogether more permanent than the relatively ephemeral issue networks. These policy communities are seen as clusters of special interests and policymakers within a given policy area that tend to be quite institutionalized and effectively exclude certain groups from the policy process. In a similar vein, ‘policy network analysis’ views the political involvement of business primarily as the developing and maintaining of close relationships with policymakers and other actors with common stakes in a policy area (Marsh and Rhodes 1992). Researchers ascribe to a given network a set of relatively stable characteristics, such as types of exchanges and sets of values that differ according to the policy area in question (John 2000). Common to all policy areas is the notion that ‘[t]he existence of a policy network, or more particularly a policy community, constrains the policy agenda and shapes the policy outcomes’ (Rhodes and Marsh 1992: 197). Policy change takes place if and when the relevant policy community agreed it was necessary and a consensus exists on the direction of change (Richardson 2000: 1006). While many empirical studies carried out within this framework remain at the level of anecdotal evidence, some scholars working within a newer policy network paradigm undertake systematic network analyses of political and economic elites. Laumann and Knoke (1987), for example, analyse how inter-organizational communication networks shape collective action in the US energy and health sectors. Studying the policy areas of agriculture, energy, health and labour in the US, Heinz et al. (1993) uncover a variety of network configurations. And an examination of the similarities and differences in labour policy networks between the US,

38 Groups, institutions, networks, ideology or structural dependence Germany and Japan, leads Knoke et al. (1996) to find that different state traditions and historical patterns of government intervention in the economy are important sources of variation in network structure. But what does network analysis tell us about business political power? As with work carried out in the pluralist paradigm, a common theme emerging from the policy network studies is the absence of a systematic correlation between similarity of behaviour, political–economic elite integration and political success of business actors. This is puzzling as these studies generally find that co-operation among businesses is rather common and that business actors enjoy privileged access to policymakers (e.g. Knoke et al., 1996; Su, Neustadtl and Clawson 1995). So, if their findings generally support the notion that business constitutes a powerful force in public policy, why do these studies not provide us with clearer explanations of the sources and limits of business power over political outcomes? One answer involves the acknowledgement that the empirical finding of business’ privileged access is beyond the explanatory scope of the policy network approach. According to Dowding (1995: 137), attempts to build a network theory of the policy process fail because the independent variables ‘are not network characteristics per se but rather characteristics of components within networks. These components both explain the nature of the network and the nature of the policy process’. Rhodes and Marsh (1994) admit that the policy network approach is really more a research tool than a theory. The concept is perhaps primarily descriptive rather than definitional or explanatory (Richardson 2000: 1006). Börzel (1998) reinforces this view, arguing that the policy network approach is ill-suited to formulate hypotheses that connect the nature and structure of networks to the policy process and political outcomes. To this, we should add problems of conceptualization and measurement. Policy network researchers disagree on central elements such as what constitutes a network or a community, and where similar definitions are used there is often no agreement on the concepts’ distinctive dimensions (Pijnenburg and Thomas 2004: 289). And John (2000) points out that it is not clear what exactly the network is supposed to measure. By counting contacts it is not possible to know whether the researcher has picked up symbolic, operational or implementation matters as well as policy choices (ibid.). It is therefore possible that networks fail to represent important relationships and illuminate political power. By contrast, the policy network framework appears to be more fruitful when networks are explicitly treated as the dependent variable. Research on policy network formation suggests that the same factors that endow actors with influence over a particular area of public policy also account for why these actors are included in the relevant issue network or policy community (König and Bräuninger 1998). Similarly, Maloney, Jordan and McLaughlin (1994) emphasize the exchange-based nature of the relationship between interest groups and policymakers. ‘Insider groups’ have preferential access to policymakers because of their perceived legitimacy from claiming to speak for a segment of the population. Policymakers offer organized interests the opportunity to shape policy while groups provide policymakers with useful resources including knowledge, advice and expertise, membership compliance and implementation guarantees

Groups, institutions, networks, ideology or structural dependence 39 (ibid.: 20–7). Thus, it is what groups bring to the consultation process that determines the extent to which they acquire influential status in a policy community or network. To take this line of thought a little further, the unity that makes business powerful can be treated as an objective characteristic of social structure, rather than being produced by a shared set of normative sentiments. For Mizruchi, the most important phenomenon to examine in order to understand business political power is the similarity of behaviour. Consequently, he suggests replacing the concept of class cohesion with that of ‘structural equivalence’ (Mizruchi 1992: 33–52): ‘Structurally equivalent actors’, in this view, are ‘those with identical relations with other actors in the social structure’ (ibid.: 44). Usually, it is assumed to suffice that actors have a high degree of similar ties to others in order to qualify for being placed in a structurally equivalent position. Furthermore, the existence or absence of a direct link among structurally equivalent actors is irrelevant (Knoke 1990: 239). Thus, even where no interpersonal network integration exists, a structurally equivalent position of business elites in democratic capitalist societies can be identified to provide firms with incentives to pursue common political strategies with pooled force and good prospects for success. For example, both the plastics and automobile industries have a strong interest in policies that maintain a low price of oil. As a result, both industries might support policymakers or election candidates who favour oil price controls without there being any need for these industries to communicate with one another (Mizruchi and König 1986: 484). Therefore, it may not be necessary to demonstrate the existence of business cohesion in order to demonstrate that the corporate community in the aggregate is a powerful force capable of acting in a politically unified manner. Rather, business is a formidable political actor to the extent that individual businesses react identically to a change, or anticipated change, in their operational or regulatory environment. Regardless of any intention on the part of business to act with one voice, the structurally equivalent position of any number of individual enterprises with respect to a particular policy induces them to act in a similar way. Overall, the main contribution of network approaches to the question of business political power seems to be in the area of network formation. Here, the bulk of research hints at two characteristics that boost an actor’s chances of being a member of a policy community or of acquiring insider status in a consultation network: their structurally important position with respect to the successful implementation of policy, and their ability to provide useful information and expertise to policymakers. The first factor will be dealt with in the last section of this chapter, while the second factor will be analysed in detail in the next chapter. But first, another possible source of business political power needs to be discussed: the ability of business to shape public opinion in its favour.

Ideological dominance In the ideal world of business, policymakers and the public will already want exactly what business wants (Mitchell 1997: 41). This, it could be argued, would

40 Groups, institutions, networks, ideology or structural dependence also be an ideal state of affairs for democracy for, if the public wants what business wants, business interests would hardly pose a problem for democracy (ibid.). And indeed, business fares relatively well in public opinion in modern capitalist societies, where it tends to be viewed ‘as a repository and exemplification of values that are seen as central to society as a whole’ (Coen and Grant 2006: 17). According to survey research undertaken in the US by McClosky and Zaller (1984), support for democratic values and support for capitalist values go hand in hand. Similar findings have been reported by Lipset and Schneider (1987). These results are for the US, however, and one might well expect different attitudinal patterns in a European context. There is, for example, a considerable difference between US and European citizens in the way inequality and redistribution are viewed and valued (Svallfors 1997). Furthermore, critics claim that, far from indicating a harmonious commonality of interests, positive attitudes toward business may be the result of business control over the means of knowledge and preference production in society. Lindblom (1977: 204–5) points out that public political debate in capitalist countries is overwhelmingly silent with respect to the ‘grand issues’ such as the legitimacy of private enterprise or the basic patterns of income distribution. Indeed, there appears to be wide consensus on the fairness or legitimacy of social and economic inequality. For Lindblom, ‘[i]t is one of the world’s most extraordinary social phenomena that masses of voters very much like their elites. They demand very little for themselves’ (ibid.: 209). He goes on to suggest that the prevailing consent is largely artificially manufactured. Such manufacturing is said to be possible because, of all the factors of socialization, the mass media are the ones most immediately accessible to the business message (Miliband 1969: 227–38). Mitchell demonstrates that the mass media are also generally very sympathetic to business, as judged by the perceptions of business representatives compared to those of trade unionists (Mitchell 1997: 50–1). This finding is supported by content analysis of media coverage of economics and politics (Glasgow University Media Group 1980). And Page, Shapiro and Dempsey (1987) have shown that it is possible to influence public policy preferences if the position of those doing the influencing is backed up by expert opinion or framed as being in the public interest. The required expertise and scholarly underpinning of business-friendly messages, it is argued, is regularly provided by conservative think tanks such as the Heritage Foundation and the American Enterprise Institute in the US, or the Centre for Policy Studies and the Adam Smith Institute in the UK (Smith 2000: 189–96). Thus, the prerequisites are there for business to be able to manipulate citizens’ preferences and public opinion in such a way that business interests will seldom be vitally challenged in politics. Business, in this sense, is powerful to the extent that it can exercise hegemonic dominance to prevent people from conceptualizing challenges to its interests (Wilson 2006: 35). The implied dominance of elite interests in the formation of public opinion was famously theorized by Lukes’ (2005) critique of Bachrach and Baratz’s ‘second face’ of power. Lukes argues that Bachrach and Baratz’s focus on the observation of grievances led them to overlook a third crucial dimension of the exercise

Groups, institutions, networks, ideology or structural dependence 41 of power, namely the power of shaping other people’s preferences. In Lukes’ (2005: 28) words, ‘[t]o assume that the absence of grievance equals genuine consensus is simply to rule out the possibility of false or manipulated consensus by definitional fiat’. Lukes’ alternative, three-dimensional concept of power involves the capability of actors to induce in others beliefs, preferences or wants that run counter to their real interests. He argues that, instead of conflicts of preferences, the social and political world is formostly characterized by latent conflicts of real interests. What prevents a latent conflict from erupting into an open clash of wills is the fact that people whose preferences diverge from their interests are unable to realize this. Third-dimensional power, for Lukes, thus consists in the ability of some agents to intentionally manipulate other people’s beliefs and preferences: ‘A may exercise power over B … by influencing, shaping or determining his very wants’ (ibid.: 27). Effectively, this is the ability to distort people’s beliefs through indoctrination and propaganda to prevent conflict from arising in the first place – considered by Lukes to be ‘the most effective and insidious use of power’ (ibid.: 27). The resulting beliefs, preferences and wants that contradict a person’s or group’s ‘real interests’ can best be described with the Marxist term of ‘false consciousness’ (Hyland 1995: 201). Just like latent class conflict, the underlying conflict of interests is a ‘latent conflict, which consists in a contradiction between the interests of those exercising power and the real interests of those they exclude’ (Lukes 2005: 28, original emphasis). The methodological difficulties implied by this concept of power are obvious, and Lukes is by no means unaware of them: ‘[H]ow can one study, let alone explain, what does not happen?’ (ibid.: 40). Moreover, if a ruling elite were at work to manipulate people’s minds, would it not be naïve to assume this elite to be active in some observable manner (Mokken and Stokman 1976: 45)? Third-dimensional power by definition operates invisibly. But there are more fundamental problems with this notion of power. The first problem concerns the ontological status of ‘objective’ or ‘real’ interests independent of subjective preferences. Any overt conflict of preferences between two parties A and B that results in the prevailing of one party, let us say A, over B can be interpreted as evidence for A’s greater power. As Hyland (1995: 202) points out, however, according to Lukes’ concept of power the precise opposite can just as well be true, since A’s preferences could be contrary to his real interests, with the same being true for B. The defaulting claim that it is by definition against B’s interest to be subjected to the power of A renders the entire argument circular, since the very existence of a power relationship depends, in turn, on inflicting harm on someone’s interests (ibid.: 202). Second, the necessity of giving a coherent account of a person’s objective interests independent from subjective preferences inevitably has to employ a normative theory of human nature, which will not be open to straightforward proof and disproof (ibid.: 203). Again, Lukes is aware of this problem. He admits that different conceptions of interests are based on different moral and political positions (Lukes 2005: 37) and that his entire view of power is ‘ineradicably

42 Groups, institutions, networks, ideology or structural dependence evaluative’ and ‘essentially contested’ (ibid.: 14). Far from solving the problem of objective interests, however, this merely illustrates the ‘epistemological relativism’ of the concept of third-dimensional power (Hyland 1995: 203). While it is probably true that universal agreement is unlikely to be reached on any basic theoretical concept in the social sciences, this does not suspend the social scientific imperative to maximize the area of intersubjective comparability and replicability. Third, any defensible theory of human nature will inevitably be based on human experience (ibid.). According to the concept of third-dimensional power, however, human experience is itself distorted, since it is exposed to the working of third-dimensional power in the form of ‘the control of information, through the mass media and through the process of socialization’ (Lukes 2005: 27). According to Hyland, this leads to another circularity of the argument: If we base our theory of human nature on distorted experience we will be unable to detect third-dimensional power; but we will only be able to determine which experiences and forms of consciousness are the products of distortion after we have formulated a defensible theory of human nature. Hyland 1995: 203 To be sure, these problems should not cause us to rule out altogether the possibility of misconstrual of interest and a misunderstanding of their situation by social agents. Nor should we discard the possibility of intentional disinformation and actual attempts at the manufacturing of consent. However, we should reject the idea that a significant part of most people’s preferences can be brought about by intentional manipulation (ibid.: 207). The problem is not so much the notion that individuals or groups hold beliefs that prevent them from appropriately interpreting their situation of subordination and subsequently taking steps to transform their grievances into political issues. Furthermore, such misguided beliefs may even be shown to serve the interests of some dominant group of people. The problem consists rather in the notion of the intentional manufacturing of these beliefs by the wielders of third-dimensional power. According to Elster, preferences tend to be co-determined by the feasible set of options. Elster calls this adaptive preference formation, which he defines as the adjustment of wants to possibilities, … a causal process taking place ‘behind the back’ of the individual concerned. The driving force behind such adaptation is the often intolerable tension and frustration, or ‘cognitive dissonance’, of having wants that one cannot possibly satisfy. Elster 1982: 126–7 Rather than being the product of some external manufacturing, however, adjusted preferences are developed endogenously. The fact that adjusted preferences on part of the subjects are beneficial to the interests of other agents does not by itself

Groups, institutions, networks, ideology or structural dependence 43 warrant the conclusion that the latter play an active role in the formation of these beliefs: There are good reasons for thinking that the oppressed classes often will be victims of a kind of myopia that prevent them from seeing the injustice of their situation – and it is clear that this is good for the ruling class; but even a systematic correlation would not by itself warrant an explanatory statement. ibid.: 137 Instead, the explanatory key to such adjusted preferences lies, not in the benefits they bring about for some ruling elite, but in their benefits for the subjects, namely in the reduction of cognitive dissonance (Elster 1983: 110–6). Once again, this is not to say that elites are entirely inactive with respect to beliefs and desires created in others. Indeed, ideological indoctrination has always been part of what can be perceived as political culture in a broader sense, and most likely always will be. Quoting the importance of methodism for the acceptance of the capitalist division of labour and wealth at the dawn of English capitalism, Elster insists, however, that such instances of indoctrination are as much the ideology of the rulers as they shape the beliefs of the ruled (ibid.: 116). To the extent that indoctrination takes place, it can be successful only because the rulers – and, of course, the actual preachers – themselves believe in it (ibid.: 117). Attempts of purposively producing mental states in other persons are bound to fail, unless there is little doubt on part of the recipient that the propagated beliefs are genuinely shared by the sender (ibid.: 66–71). This makes sense when applied to normative beliefs about the benefits of free enterprise and the rewards of ownership and leadership of firms. Capitalists and managers by and large genuinely believe that market capitalism is a fair and efficient system and that their incomes are justified rewards for their investment of capital, skills and time. This means that any beneficial effects of certain ideologies and beliefs for the elites are mere by-products of the elite’s own convictions. As such side effects, however, they have no explanatory power, because we cannot conclusively explain the existence of a phenomenon by its effects (ibid.: 117). But unless clear evidence can be presented for demonstrating the conscious and successful utilization of ideology on the part of a powerful elite, to interpret widespread popular support for capitalist institutions and values as products of third-dimensional power would be little more that an implausible conspiracy theory (Hyland 1995: 207).2 What remains of claims about business’ control over citizens’ preferences and beliefs is the adaptation of preferences to what is perceived to be a feasible set of options. This is the subject of the next section.

Structural explanations of business dominance During the 1992 US presidential election campaign, Bill Clinton’s adviser, James Carville, placed a sign over his desk in the Little Rock headquarters displaying a simple four-word phrase: ‘It’s the economy, Stupid!’ (Safire 1993: 376).

44 Groups, institutions, networks, ideology or structural dependence While its mantraic use as Clinton’s central campaign theme introduced the phrase into the popular lexicon, the gist of the message – that economic performance is of overarching importance for the popularity and electoral success of politicians – has long been established among practitioners and academics alike. Since Key’s (1966) study, The Responsible Electorate, popularity functions assume a simple reward–punishment model of political popularity and electoral success: If economic conditions are good the electorate will reward incumbents, while it will punish them if the economy is less than satisfactory. Yet, while recent comparative research suggests that the economy tends to weigh more heavily than other factors in the vote decision (e.g. Alvarez, Nagler and Willette 2000; Lewis-Beck and Stegmaier 2000; Duch and Stevenson 2006), this does not necessarily mean that voters value the economy more than other political goods. The underlying claim is rather that the economy represents an informative and easily interpretable clue about the competence of incumbent politicians. It is primarily for this reason that voters are likely to rely on economic evaluations more than on other signals in making a voting decision (Lewis-Beck and Stegmaier 2000). Crucially for the present context, there is evidence that policymakers anticipate the effects of economic voting in their policy decisions and strategic behaviour (Kiewiet 2000; Palmer and Whitten 2000). As a special case of Friedrich’s (1963) ‘law of anticipated reactions’, economic voting is the mechanism through which the structural power of business unfolds. The phenomenon of economic voting and its anticipation by policymakers is at the heart of theories of the structural power of business. Essentially, these theories claim that even if capitalists did not engage in any political action at all, public policy would nonetheless be systematically biased in favour of their interests. The prime exponent of this strand of explanation of business political power is the theory of the structural dependence of the state on capital. This theory asserts that democratic state action is ultimately constrained by macroeconomic parameters that are in turn largely determined by allocative decisions made under private – that is, capitalist and consumer – prerogative. Traditionally applied in the context of redistributive taxation and social-democratic welfare politics, the theory asserts that the negative effects of redistribution on the share of profits consumed by capitalists lead to disincentives for investment followed by negative effects on growth, employment and tax revenues. In anticipation of any negative economic effects of policy and their electoral and fiscal consequences, policymakers therefore tend to carefully avoid any policy that affects the revenue prospects of business, prioritizing business confidence instead. Contrary to this commonplace notion, Przeworski and Wallerstein have shown that under certain circumstances, a pro-labour government can redistribute income in an egalitarian manner and simultaneously keep the rate of investment unchanged, if not increase it (Przeworski and Wallerstein 1988: 18). Thus, no strict structural dependence of the state on capital exists. However, things look different once Przeworski and Wallerstein take into account that a redistributive tax measure will involve changing the status quo. In order to preserve their existing consumption share of profits, capitalists who attribute some positive probability to a future

Groups, institutions, networks, ideology or structural dependence 45 tax increase on their private income will reduce investment in the period prior to the change coming into effect. Unless the government is willing to incur the consequences of disinvestment, it will refrain from redistributive reform. Thus, through firms’ anticipation of future threats to shareholders’ consumption, the state is dependent on capital in the selection and pursuit of its policies. Przeworski and Wallerstein’s model therefore predicts that egalitarian changes to the wealth and income distribution in capitalist democracies are impossible without negatively affecting investment. As a result, all it takes for business preferences to prevail in politics is that capitalists go about their business as usual. In the words of Przeworski and Wallerstein, The effective capacity of any government to attain whatever are its goals is circumscribed by the public power of capital. … It does not matter who the state managers are, what they want, and whom they represent. … Capitalists do not even have to organize and act collectively: it suffices that they blindly pursue narrow, private self-interest to sharply restrict the options of all governments. ibid.: 12 This theme is in stark contrast Bentley’s (1993 [1908]: 21–2) aforementioned account, according to which the activity of the directors of a capitalist enterprise only becomes politically relevant when they ‘finish their ordinary business and turn … to discuss the part the corporation will take in the next political campaign’. In the structural dependence thesis, it is precisely their ‘industrial or economic’ activity that exerts important effects on governments’ ability to do their job. Capitalists’ power over investment and employment may thus lead to their de facto domination of the political process. The partisan composition of governments does not matter, as government agents are constrained to act, not at the behest, but effectively on behalf of a ruling class (Miliband 1977: 74). The economy affects the utility of elected policymakers not only through economically oriented voting behaviour. At its most fundamental level, the very economic existence of the state ultimately depends on the economy and its performance. Less dramatically, but nevertheless immensely importantly, every governmental policy programme requires funds, and is thus in principle affected by the availability or otherwise of sufficient tax receipts. Policies that lead to negative incentives for economic activity curtail the tax base and may thereby undermine the policymaker’s chances to achieve their various other policy goals (Block 1977: 15). In addition, reduced surplus generation in the economy shrinks the resource pool from which monetary contributions to policymakers and candidates can be made. The history of business political action across the world, however, does not suggest that the latter aspect constitutes an important link between the fortunes of capital and policymakers. Even during times of extremely bad economic performance and profitability, such as the Great Depression from 1929 to 1933, business has always managed to come up with the funds required for ambitious political action.

46 Groups, institutions, networks, ideology or structural dependence This, then, is the classic theory of business’ structural dominance. It is central to structural Marxist theories of the state in capitalist society (e.g. Poulantzas 1973), but it is also a central tenet of neoclassical political economy (e.g. Becker 1983). The social and political implications are considerable. If the theory is true, democratic politics is a mere sham. The familiar remedies for curbing the political power of business in the more overt arenas of pressure politics, such as tighter controls on corruption, or the regulation of party and campaign contributions or lobbying, would be largely futile. Furthermore, reformist policies designed to alleviate the undesirable side effects of capitalist market economies are severely constrained. Social inequalities might be impossible to overcome by any political strategy other than the wholesale overthrow of the capitalist system of ownership and production (Miliband 1969: 152). At the very minimum, the pluralist premise that business is just another interest group alongside others such as trade unions or environmental groups can no longer be upheld. In the literature on business and politics, the theory of structural state dependence is often invoked when other explanations of business power over public policy fail. For example, in their detailed study on corporate PACs in the US, Clawson, Neustadtl and Weller regularly refer to the importance of ‘the wider power and influence of business’ (Clawson, Neustadtl and Weller 1998: 23) or a ‘constraining field of power’ (ibid.: 196) for corporate political action to work. Yet they do not clarify what exactly these factors are or how they work. Similar references to systematic biases and structural constraints of various kinds have played an important role in the community power debate of the 1960s and 1970s. While that debate has left the conceptual apparatus of democratic theory in a state of considerable ontological disarray, Lukes’ aforementioned ‘third dimension’ of power has widened the scope of democratic theory’s concern with political power towards the possible inclusion of unobservable phenomena. On closer inspection we find that Lukes’ account of third-dimensional power is actually made up of two distinct theoretical claims, each of which being somewhat precarious. Lukes’ first claim denotes the capability of powerful elites to induce in others beliefs, preferences or wants that run counter to their ‘real’ interests. The previous section has already dealt with the problems attached to this line of reasoning. Aware of the substantial difficulties of an account of third-dimensional power as intentional ideological indoctrination, Lukes offers an alternative, non-intentional foundation of third-dimensional power which he refers to as ‘the bias of the system’ and ‘the socially structured … practices of institutions’ (Lukes 2005: 26). Instead of explaining what exactly is meant by such a structurally and institutionally produced bias, however, Lukes redirects his readers first to Schattschneider’s notion of the mobilization of bias and later to the insights and limits of structural Marxism (ibid.: 20, 54–8). Structural Marxism, in turn, approaches the question from the opposite end, starting from the premise that there is a ruling class and that the state in capitalism fulfils functionally beneficial tasks on its behalf. The challenge for structural Marxists, therefore, has been to account for instances in which the state appears to act independently or autonomously from the immediate preferences of the owners

Groups, institutions, networks, ideology or structural dependence 47 and managers of capital. Taken to its logical consequence by Poulantzas, this functionalist version of the Marxist theory of the state asserts that the state fulfils a global role as the political organizer of the bourgeoisie in the social formation that is capitalism (Poulantzas 1973: 279–95). In this view, whatever the state does is by definition functionally beneficial to the interests of the capitalist class, even if the state’s actions contravene the expressed preferences of business. But Poulantzas offers no actual explanation for why policies imposed on the capitalist class against its will are nevertheless in that class’s long-term interest (van den Berg 1988: 356–7). He is therefore unable to explain why and how the state, though it has the ability to override the opposition and resistance of the capitalist class, is nevertheless constrained to serve the long-term interests of that class only. Poulantzas’ account rests on the determinate character of the historical formation of capitalism. For Poulantzas, the relation between the bourgeoisie and the state is an objective one. This means that if the function of the state in a determinate social formation and the interests of the dominant class in this formation coincide, it is by reason of the system itself. One feels compelled to agree with van den Berg (1988) that this is a somewhat tautological argument, which at any rate is wholly unfalsifiable. The analytical response to Poulantzas’ structuralist theory of relative autonomy also presumes a harmonic relationship between the interests of the managers of the state and those of capital, before proceeding to look for reasons why the bourgeoisie has abdicated from directly exercising political power itself (Elster 1985). Yet, far from providing grounds as to what should reassure the bourgeoisie that government policy will, on average, be ‘good enough’ for its interests, Elster (1985: 422) suggests that the state, out of self-interest, is constrained to very closely follow the policies that are optimal from the point of view of capitalists. This, of course, brings us back full-circle to the problem of the structural dependence of the state on capital. It is because of this structural dependence that, in Block’s (1977) words, ‘the ruling class does not rule’. It does not have to, because, constrained by their structural dependence, relatively autonomous state managers can be sufficiently relied upon to make decisions largely in line with the needs of capital. The autonomy of state managers, as the theory of structural state dependence tells us, is only relative because their own political survival depends on the accumulation process, with a healthy economy guaranteeing adequate state revenues and political legitimacy and support. Because major investment decisions in a capitalist system are usually private, public officials cannot risk a capital strike by proposing policies that threaten the owners of capital (Lindblom 1977: 170–88). In its focus on redistributive politics and the class antagonism between the owners of the means of production and those that operate them, the theory of the structural dependence of the state on capital also features as a general theory of the structural power of capital. The structural constraints on political action apply not only to the state, but can also be shown to rigidly constrain labour unions’ bargaining power. Przeworski and Wallerstein (1988: 15–6) demonstrate theoretically that labour unions cannot push their share of the product beyond a certain point at which capitalists will withdraw capital from investment and reallocate

48 Groups, institutions, networks, ideology or structural dependence it toward their consumption. As with the constraints on state action, while no prohibitive constraints are present in the static setting, the problem is a dynamic one. And again, in one form or another, this familiar argument is frequently used as part of the rhetorical repertoire of political and economic actors. In times of global capitalism, the claim has even been stretched to mean that trade unions as we know them represent outmoded forces that obstruct welfare generation. Thus, the former director-general of the CBI, Digby Jones, warned that unless the unions became more flexible, ‘the very people they are paid to protect are the people who will suffer most’. According to Jones, [t]he unions face stark choice. Either recognize the competitive demands of the 21st century, reform, put training as your top priority for employees – and, by encouraging flexibility – help the nation with the battle for competitiveness. Or wither on the vine of growing irrelevancy. The Guardian, 3 September 2004 Empirical applications of the classic theory of business’ structural power are found mostly in the context of redistributive taxation and social-democratic welfare politics (e.g. Quinn and Shapiro 1991; Swank 1992; Williams and Collins 1997). But the same kinds of constraints apply to other policy areas as well. Dryzek (2000: 142), for example, asserts rather generally that ‘policies that damage business profitability – or are even perceived as likely to damage that profitability – are automatically punished by the recoil of the market. But while the theory of the structural dependence of the state on capital predicts constraints on all policies that could adversely affect business’ revenue prospects, the constraints on nondistributive policies that result from the interaction of centralized, authoritative decisionmaking by policymakers with the decentralized, allocative decisions of private market actors are often overlooked. An exception is Sunstein’s (1998) study of the structural constraints on deliberative decisionmaking. Sunstein analyses how the aggregate outcomes of the decentralized responses to deliberative decisionmaking can have unintended consequences that render measures of public policy inefficient, reverse their intended effects or even avert their very formulation in the first place should the policymaker anticipate such adverse reactions. Elaborating this problem in the context of risk regulation, Sunstein (1998: 235–6) distinguishes between risks that the government is trying to control by means of regulation (‘regulated risks’) and the risks that are increased by the same regulation by way of unintended side-effects (‘ancillary risks’). Examples of this pervasive problem abound: fuel economy standards, designed to reduce environmental and health risks, can make automobiles less safe and in that way increase risks to life and health; regulations designed to control the spread of AIDS and hepatitis among healthcare providers can increase the costs of healthcare, making healthcare less widely available and thus costing lives; government bans on the manufacture and use of asbestos can lead companies to switch to even more dangerous substitutes; and bio fuels designed to curb the damaging effects of fossil fuel consumption can lead to increases in the price of corn, depriving poor people

Groups, institutions, networks, ideology or structural dependence 49 of affordable food. A quarter of a century earlier, Peltzman (1973) showed how safety and efficacy regulation on drugs can backfire by thwarting innovation and research on new and improved drugs. The US Office of Management and Budget regularly estimates the cost of regulations and calculates the cost per life saved that would result from their adoption. Using such cost effectiveness measures, some programmes such as seatbelt standards in cars or benzene exposure limits for workers can be shown to be cheaper than others, such as banning land disposal of hazardous waste or setting formaldehyde exposure limits for workers (Kraft 1996: 167). In the presence of budget constrains (which is almost always a given), spending money on some programmes may be at the expense of others, ideally those where lives saved would come at a higher price. The essential commonality at the core of structural state dependence and the complex of regulated-cum-ancillary risks identified by Sunstein and others is that the policy in question has unintended, though in principle foreseeable, negative consequences that may outweigh the benefits. Furthermore, both phenomena occur irrespective of any intention on the part of the regulated to undermine the desired effect of policy. Rather they result from the structure of the capitalist organization of the economy via the decentralized and uncoordinated responses to policy by consumers and producers who have no particular agenda other than the blind pursuit of their legitimate cost-minimizing strategies (Przeworski and Wallerstein 1988: 12). It is this interaction of market and non-market modes of decisionmaking, ‘the fact that individuals are simultaneously market agents and citizens’ (Przeworski 1991: 111), that makes capitalist democracy appear to be an irrational political– economic system. To deal with these constraints in a unified manner, whether they occur in distributive, re-distributive or regulatory areas of public policy, the structural constraints on policymaking that arise from either structural state dependence or ancillary risks are henceforth referred to as negative inducement effects.3 Negative inducement effects concern the ability of policies to achieve their intended effect without creating too many or too severe undesirable side-effects. Similarly to the economic dependence of the state on capital, policymakers’ popularity may suffer dents if the unwanted side-effects of their policies outweigh the benefits. The extent, likelihood and severity of negative inducement effects depend on the responses from private actors. Because businesses have control over the allocation of society’s capital, their responses are particularly important. Policymakers depend on their policies to work, and the responses of businesses to these policies are major determinants of their success (Castles 1982). Thus, in capitalist democracies, the ‘wrong’ political or regulatory measure can upset the conditions under which private enterprises are expected to deliver their contribution to the factors determining the fate of elected policymakers. Refusal to deliver can, but does not have to, take on the heavy-handed form of ‘capital strikes’ envisaged by structural dependence theorists. More subtle responses can include, for example, changes to product prices (Hart 2002: 15). Because of the possibility of such negative inducement effects, the interests of the state and its incumbent rulers are inextricably linked to business interests. The consequences policymakers have to suffer depend on various factors such as the gravity of the effects. They can range

50 Groups, institutions, networks, ideology or structural dependence from a temporary downward blip in approval ratings to losing votes at the next election, being kicked out of office and, in extremis, provoke social unrest and upheaval. The structural explanation of business political privilege is a powerful theoretical argument, but it has a number of problems. First, if the structural power thesis holds, we should rarely witness policy change over time that overrides business complaints regarding the detrimental effects of a policy. However, instances of business failing in the policy struggle are frequently observed. Business sometimes does lose: the victory of Greenpeace over Shell in the Brent Spar episode (Jordan 2001) is only one of the more high-profile examples of many instances of business defeat in the policy struggle. Students of public policy have often found political outputs discordant with industry preferences in, for example, consumer protection (Pertschuk 1982), environmental protection (Berman and Bui 2001), occupational health and safety regulation (Scholtz and Wei 1986), financial regulation (Worsham 1997), public lands policy (Culhane 1981) and insurance regulation (Meier 1988). Thus, unless one adopts the structural–functionalist view that even policies that clash with business preferences are still in business’ objective interest, there clearly are explanatory limits to the mechanisms of structural dependence and privilege. The existence of structural constraints can provide business with enormous clout over public policymaking. But while the anticipation of negative inducement effects on the part of policymakers leads to the systematically preferential consideration of business interests in public policymaking, the intentional view of power applied here does not warrant the claim that these effects are themselves instances of the exercise of power. In a similar vein, Dowding argues that the fact that constraints upon what is feasible in public policymaking are systematically beneficial for capitalists does not mean that capitalists are powerful. Rather, they are ‘systematically lucky’ (Dowding 1996: 71–5). If only because there is a certain tension between the notion of systematic outcomes and that of outcomes involving luck, the systematic benefits to business accruing from these constraints are perhaps better thought of as structural dominance (Hyland 1995: 207–9). Second, the implication that the partisan composition of governments does not matter for political outcomes does not survive in the light of empirical evidence. Some authors detect a process of convergence of party platforms. Thus, Caul and Gray (2000) find that the ideological distance between left and right-leaning parties decreases over time in ten out of fifteen nations in their analysis. While being more pronounced in countries with majoritarian electoral system such as Britain, Canada or Australia, ideological convergence also takes place in proportional representation systems such as the Netherlands, Belgium or Germany. But convergence of positions is not the same as convergence of output. Quinn and Shapiro (1991), for example, find that partisan change in the US has noticeable effects on the country’s tax policies. And in a cross-national comparative study, Garrett (1998) finds that the partisan composition of government continues to affect levels of government spending even under conditions of considerable integration into global markets. He also notes that left-leaning governments generally outspend right-wing

Groups, institutions, networks, ideology or structural dependence 51 governments. Furthermore, and contrary to much lore about the political effects of globalization, Garrett finds that the spending lead of left-leaning governments increases with global economic integration (Garrett 1998: 74–105).4 Finally, the notion that parties matter has, at least until recently, been shared by business representatives themselves. Whether warranted by political realities or not, business leaders overwhelmingly believe that social-democratic or Labour governments are bad for the economy, while right-leaning and business-friendly parties are expected to do a better job in that regard (Grant 1993: 137–8). Third, with their emphasis on structural determinants, theories of structural dependence find it difficult to account for the ubiquitous political activity of business. The question inevitably arises why, if their interests are automatically (i.e. ‘structurally’) taken care of, would capitalists waste time and money on politics in the first place? After all, the principle goal of business is economic success, not participation in politics (Mitchell 1997: 11). The obvious answer is that there are instances of policymaking in which at least some firms or sectors favour policies that are deemed undesirable by others (cf. Smith 2000: 13–7). This may be because the policy bestows on them an advantage vis-à-vis competitors, or because it benefits a particular firm or sector but not others. In the context of environmental politics, for example, manufacturers of environmental technology have a vested interest in increased regulation while polluters prefer the status quo. And any policy that leads to increased costs for established players may be deemed desirable by potential market entrants for the simple reason that they harm incumbents. But it is important to keep in mind that the presence of countervailing business preferences about policy does not lead to an overall relaxation of the constraints on public policymaking, nor does it render the system of market capitalism more accountable or democratic. Today’s challengers and market entrants are tomorrow’s incumbents. They will then enjoy the same structural advantages that they now help to undermine. Unlike elite theories, therefore, structural accounts of business political power do not have to rely on business unity. More problematic for the structural account of business dominance is the fact that we witness a great deal of business political action directed against policies claimed to harm business per se, such as environmental protection measures or social-democratic redistribution programmes. In an analysis of 137 legislative issues lobbied by US interest groups in 1995–6, Baumgartner and Leech (2001) found that the majority of cases involved only a small number of lobbyists and that the most frequently active lobbyists represented business. Thus the ‘veto power’ of capital based on the private control of investment does not seem to ensure that the state will act in the interests of the capitalist class (Domhoff 1998: 63–5). Business leaders themselves certainly seem to believe that their structural power as such does not suffice to induce politicians to treat their interests favourably. When a researcher suggested to him that his firm probably had enough economic power to dispense with its efforts to influence elected officials, a top corporate leader replied, ‘I’m not sure, but I’m not willing to find out’ (quoted in Clawson, Neustadtl and Scott 1992: 121). Structural constraints and privileges notwithstanding, it appears that capitalists often have to work hard to achieve their political goals (Akard 1992: 599).

52 Groups, institutions, networks, ideology or structural dependence Fourth, there are powerful constraints on the behaviour of policymakers other than those transmitted by the electoral effects of macroeconomic performance. Politicians use ideology as an efficient way to communicate their goals and platforms to voters (Downs 1957: 100–12). This compels them to display some minimum level of ideological consistency. One of the fundamental problems faced by voters is to determine which candidate will best serve their interests in an environment where much of the necessary information is unavailable (Ferejohn 1990). According to Downs (1957: 39), voters evaluate the policy outcomes that a candidate would likely produce in office rather than merely the policy platform that a candidate announces: a voter ‘knows that no party will be able to do everything it says it will do. Hence, he cannot merely compare platforms; instead he must estimate in his own mind what the parties would actually do were they in power’. In trying to estimate what a candidate will do if elected, voters must consider veracity, ambiguity and feasibility. To this end, they can obtain cost-efficient clues about the quality and position of candidates by observing, first, the announcement, pledges and promises that political candidates make, and second, their track record with respect to honouring past pledges. The fact that voters may thus be able to estimate what politicians will do from their party and election platforms provides elected policymakers with strong incentives to be predictable and reliable – in other words, do what they promise. If politicians defect on their promises, voters will not trust their platform in the future because it offers little basis for predicting their performance in office. Voter cues from party platforms are of course often fuzzy, and the resulting incentives to redeem election promises can be compromised by institutional factors. Differences between what candidates announce and what they are likely to do have many sources, but one of the most important ones is the institutional environment, such as a separation of powers (Lacey and Paolino 1998). Moreover, politicians may hope that information asymmetries between them and voters enable politicians to explain away any inconsistencies between campaign promises and performance (Downs 1957: 135–7). Indeed, Page (1978) and Shepsle (1972) have shown early on that candidates have incentives to make ambiguous statements about what they will do in power. But for the sake of their credibility and reputation, elected officials have strong incentives to adhere to their ideological self-labelling and deliver on their pre-election pledges at least in broad terms (Hinich and Munger 1994). Politicians, therefore, seek to enact their declared policy preferences. During the 1997, 2001 and 2005 UK election campaigns, the Labour party even formalized this incentive by introducing ‘pledge cards’, inviting voters to check the party’s pledge redemption record in detail. Business lobbyists are aware that these incentives may lead policymakers to enact policies that harm their interests. In the words of an industry spokesman, ‘This is why it is so important for an organization like ours to keep our foot in the door no matter who is in government; because they would all have their own agendas otherwise’ (interview with Marc Schnerr). This suggests a fifth problem for the structural power argument – the possibility that policymakers may have political interests and agendas of their own. The revival of state-centred theory

Groups, institutions, networks, ideology or structural dependence 53 in the early 1980s has led some theorists to reject approaches that see political power as deriving from class location or that explain state policies with reference to economic or other societal-level phenomena (Skocpol 1985). Instead, the state and its officials are viewed as having their own interests as well as the means required to pursue them. From a Marxist perspective, Block (1977) argues that the state managers could gain a consciousness all their own and take actions that are independent of the immediate preferences of capitalists. In the spirit of Weber’s (1964) idea that the organizational apparatus of the state is the original site and primary field of the constitution of power relations, these critics attack pluralist and structural–functionalist conceptions of politics alike for an overly determined view of the state. Instead, they stress the independent influences of state and party structures and state managers (Skocpol 1980). Finally, besides being a powerful source of political action, possibly with preferences of its own, the state is the creator and ultimate guarantor of private property rights. The ‘invisible hand’ of market capitalism depends on the very visible hand of law enforcement as an ally. But the alliance works both ways, and governments produce property rights and markets with the expectation of receiving political and economic benefits in return. Every government depends on some level of popular support and on economic resources in the form of tax revenues. It is the anticipation of inefficient production in an environment without property rights that leads governments to grant these rights to induce more efficiency and productivity, in the hope of making tangible benefits from excess productivity through increased tax revenues and enhanced political support. This is why non-state actors (and, in capitalist democracies, business actors in particular) matter to an extent that is insufficiently acknowledged by state-centred theory.

3

An informational–structural model of business power

The classification of research on business political power into the five categories of pluralist group theory, neocorporatism, network approaches, ideological dominance and structural theories undertaken in the previous chapter is an ideal-typical one. Not all research done to date can easily be assigned to one or other of these categories. For example, scholars working within a group framework of reference are by no means unaware of the fact that policy struggles are broadly constrained by the economic and political structures in which they take place (McFarland 1991). Conversely, identifying these institutional and structural constraints does not prevent scholars working within a structural framework from carefully studying the political processes through which resources are mobilized and coalitions formed over specific issues (Mizruchi 1992). As ideal types go, however, these analytic distinctions are important and useful, and the most promising attempts at explaining the scope and limits of business political power have as their point of departure the explicit acknowledgement of these conflicting modes of explanation. One such strategy has been formulated by Mitchell (1997: 69), who suggests overcoming the ‘artificial separation’ of the non-intentional, structural power of capital from business political activity by acknowledging that the structural link is variable. Combining the claim that the political–economic system of democratic capitalism is structurally loaded in favour of business with the claim that business can still lose, Mitchell formulates a theory of ‘calculated heroism’ to explain why, given their structurally privileged political position, formidable political resources and extensive control over the media, business interests nevertheless at times incur political loss (ibid.: 167–89). Facilitated by external events such as scandals or catastrophes, non-business interests and policymakers can sometimes succeed in creating an environment that is conducive for policymakers to embark on a ‘fit’ of heroic policymaking (ibid.: 10). In these situations, policymakers are able to temporarily override the business confidence factor and the negative inducement effects commonly attributed to such disregard. Unlike in Vogel’s (1989) theory of cyclical patterns of public support for business, business legitimacy in this view varies not so much cyclically, but rather as a function of irregular, shocklike perturbations to the political climate within which business operates. Mitchell (1997: 78) suspects that interruptions to the structurally induced routine of acceding

An informational–structural model of business power 55 to business demands are more likely to occur either when policymakers discount business interests’ impact on public support against other factors affecting support, or because they have reason to interpret economic performance or other policy outcomes differently. But while this means that business confidence is a conditional rather than constant factor in politics, Mitchell’s ‘post-structural view of business confidence’ relies on exogenous shocks that upset the political routine to temporarily offset an otherwise firm grip of business structural power over public policy. This chapter examines a theoretical perspective that has the potential of predicting fluctuations of business power by reference to more systematic factors. Informational accounts of business influence rely on factors such as reputation and expectations about policy costs instead of shocks as the major factors that determine when policymakers can overrule business warnings about policy costs. To introduce the central idea of informational models of special interest politics, it is helpful to clarify what happens when business representatives lobby policymakers. Caldeira, Hojnacki and Wright (2000) distinguish three broad categories of lobbying activities: position taking, grassroots mobilization and gathering and transmitting information (i.e. purchasing advertisements, conducting research, providing information to the mass media, distributing leaflets and sending special mailings to organizational members). They find evidence that groups’ choices of lobbying tactics have less to do with organizational structure, resources or the wider political context of the campaign. Instead, according to these scholars, lobbying ‘is about winning politically and doing whatever is necessary to do so’ (Caldeira, Hojnacki and Wright 2000: 52). Often, this involves the strategic dissemination of information. Caldeira, Hojnacki and Wright (2000: 53) ‘expect lobbyists to provide whatever information will be most effective in influencing legislators’ decisions, and that legislative success frequently requires that groups provide both electoral and policy relevant information’. Even political decisionmakers who enjoy a degree of autonomy from societal interests nevertheless depend on some minimum level of economic and political support from these groups. They also need information to translate their political intentions effectively into political reality. While information on the size and likelihood of a policy’s consequences is not readily available to policymakers, business lobbyists often have private information; that is, information known only to them, on this. Thus, information and expertise are part of business actors’ ‘political potential’; that is, of those characteristics that allow them to convert whatever economic power they may have into political power (Coleman 1988). These resources enable actors to gain access and develop legitimacy with policymakers by providing briefings, technical expertise and analysis to politicians who lack the resources or the time required to collect, analyse and interpret vast amounts of policy-relevant information on their own (Grant 2000: 78). Through this, special interest groups can derive influence from their control of information and their acceptance as a ‘policy insider’ (ibid.). Of course, this mechanism works for business actors only to the extent that they really do enjoy an informational advantage concerning the likely consequences of policy. How plausible,

56 An informational–structural model of business power then, is the assumption of asymmetric information in the context of business lobbying?

Informational asymmetries between business actors and policymakers An important observation concerning the interaction between centralized authoritative decisionmaking and decentralized responses by private actors is that some, but not all policies will lead to negative inducement effects. From a business perspective, however, even low-cost policies entail costs. Typical examples are compliance costs attached to newly imposed environmental standards, health and safety regulations at the workplace or policies aimed at curbing discrimination against the disabled. It is in the interest of cost-minimizing actors such as firms to avoid these costs. Of course, firms often do not know themselves with sufficient degree of precision just what costs and benefits would be associated with a new policy and how they will affect their bottom line (Hart 2002: 21). Therefore, part of business’ overall efforts to minimize costs is to reduce the regulatory uncertainty permeating their operational environment (Birnbaum 1984). In the face of considerable complexity firms apply informational shortcuts to understand their environments and make decisions. One available device is ideology. Ideology serves as a set of cognitive filters and predispositions that reduce uncertainty and facilitate decisionmaking (Hart 2002: 26–7). While business ideology is certainly not a constant, it does on the whole tend to involve scepticism toward government action (Vogel 1996: 29–72). Another device is to ‘play it safe’ and locate estimates of expected costs on the high end. Referring to the European Working Time Directive and to a number of European and Irish employment equality laws, one industry representative said: I think it is fair to say that some of those regulations might not have quite the negative impact that we felt they would have, but at the same time there are other side effects that are unforeseen, like this record-keeping provision that might be looked upon in seven or eight years as being a big problem for companies. Whereas in one issue it’s maybe not as bad as feared, in others it’s that we really didn’t anticipate what might later become a big problem. Interview with Ciaran Higgins Faced with this uncertainty, rational business actors should tend to estimate policy costs on the high end. This in turn provides business lobbyists with incentives to exaggerate the size and likelihood of negative inducement effects when communicating their views on the effects of policy to policymakers. In environmental politics, for example, exaggerated predictions of negative inducement effects are a common occurrence. And while business representatives tend to characterize many a new piece of environmental policy as ‘economically unsound’, many of the doomsday scenarios painted by industry in retrospect turn out to be exaggerated and sometimes wholly unfounded (e.g. Haq et al., 2001). For example,

An informational–structural model of business power 57 the CBI stated in February 2005 that ‘[c]ompanies will go abroad if extra costs are imposed here that are not imposed abroad’. But when pressed to verify these claims before the House of Commons Environmental Audit Committee, the CBI director-general proved unable to name a single company that had relocated abroad as a consequence of environmental regulations (The Guardian, 1 February 2005). The problem for policymakers is that their information problems are often larger than those of business. When this is the case, policymakers lack the information necessary to assess the veracity of business’ assessment of a policy’s negative inducement effects. This information problem can be thought of in terms of incomplete knowledge concerning the state of the world. The members of society are affected by policy and by the state of the world on which policy decisions impact. Therefore, individuals’ preferences over policies are a function of the state of the world (Lohmann 1993: 319). A policymaker who is perfectly informed about both the state of the world and voter preferences will set policy in an unbiased and efficient way in order to maximize the welfare of her constituents and thereby maximize her re-election chances (Lohmann 1998: 810). In the real world, however, the targets of policy are often better informed on issues that affect them than policymakers are. Policymakers are imperfectly informed and depend on the representations of highly interested constituents to ascertain how their actions will affect their support among voters (Downs 1957: 90–2). Structural explanations of business power, by contrast, rest on the crucial but largely implicit assumption that political decisionmakers, recognizing that the prosperity of their constituents, the tax revenue required for various government programmes and their own re-election prospects depend on giving business what it needs, actually know what business needs. In the perspective of structural dependence theory, it therefore suffices for business to merely hint at the undesired effects of a policy in order to persuade policymakers to rethink their ambitions. In the words of Lindblom, ‘[o]rdinarily, [businessmen] need only point to the costs of doing business, the state of the economy, the dependence of the economy’s stability and growth on their profits and sales prospects – and simply predict, not threaten, that adverse consequences will follow on a refusal of their demands’ (Lindblom 1977: 185). But policymakers operate in highly uncertain environments. They have general ideas of the positions they should take to maximize their political support, but they do not avail of all the information necessary to make the right choices. While political decisionmakers may generally be well aware of the fact that their interests are linked to the well-being of business (‘it’s the economy, Stupid!’), they will often be less clear about what exactly it is that business needs in a given situation and with respect to a particular policy area. In Mitchell’s (1997: 69) words, ‘[t]he needs of business are not self-evident to government officials. They require articulating – by the governor of the Bank of England, the Business Roundtable, or the CEO of RJR Nabisco’. Thus, even if we assume that policymakers care about policy only insofar as they care about its consequences, information becomes valuable, and those who possess it are accordingly in a position to influence policy (AustenSmith 1993: 799–800). These consequences may be political (‘How are reelection chances affected?’), or they may be technical (‘How will a revised clean air act

58 An informational–structural model of business power hurt employment in the car industry?’) (Austen-Smith 1993: 799). Oftentimes, they will be both (‘How are re-election chances affected if the revised clean air act hurts employment in the car industry?’) In this situation, lobbyists offer help in the form of political intelligence about preferences of constituents or about the way in which a policy affects the welfare of voters. This is the main reason why, in spite of any structurally privileged position, lobbying and other forms of political action are indispensable for business groups to have their interests observed and safeguarded. Furthermore, policymakers balance political support from one group against other likely sources of contributions and votes (Denzau and Munger 1986). The prominence of such balancing acts ensures that, even when business concerns are in harmony with the goals of policymakers, business has an incentive to overstate the importance of its position by exaggerating the electoral salience of its cause so that a policymaker places even more weight on it. Misrepresentations to the extent of outright falsehood are therefore quite frequent. Ainsworth (1993: 51) cites numerous examples, including an episode from the 1930s in which utility companies had thousands of postcards sent to members of the US Congress by high school students who were paid to sign names taken randomly from the phone book. Lobbying aims above all at changing policymakers’ perspectives, widening the picture, introducing new evidence or obscuring unfavourable information. Legislators readily acknowledge that lobbyists’ influence and legitimacy stem from their private information. This view of lobbying depends on the presumption that some essential pieces of information can be observed by one party, but not the other. Ainsworth (1993: 44) quotes a US Senator’s view of lobbyists as ‘performing extremely useful functions in the national interest. They can be tapped for expert information on problems, [and] they can analyze the impact of proposed legislation of their area of concern’. Furthermore, for lobbying to have an effect on the behaviour of the less informed, common knowledge among both parties about the existence of information asymmetry must be assumed. As Austen-Smith (1992: 47) put it, ‘[F]or a speaker to be able to persuade the listener to act in a particular way […] it is clearly essential that the listener believe the speaker knows something that the listener does not’. How does business get the information that others do not, or cannot, have? Much of the information required for the assessment of negative inducement effects is routinely available to business and its interest organizations. Special interest groups in general enjoy some informational advantages vis-à-vis policymakers due to the latter’s capacity constraints and because of interest groups’ own strong incentives to pool resources and routinely conduct research on issues of concern to their members. To some extent, being well-informed is a function of an organization’s internal capacity or resources to gather information first-hand through research or direct observation (Carpenter, Esterling and Lazer 1998: 428–9). Special interests are, in Lohmann’s words, ‘naturally’ better informed: [C]ompared to the general public, they get costless information as a by-product of their specialized activities, and they have stronger incentives to invest in costly

An informational–structural model of business power 59 information gathering, to pay costly attention to complex information, and to invest in costly expertise that allows them to understand such information. Lohmann 1998: 825 In the case of business the informational asymmetry goes even further. Individual firms accumulate knowledge about relevant policy issues in the course of performing their everyday activities (Carpenter, Esterling and Lazer 1998: 420–1). Firms automatically possess data as private information that is important for political decisionmaking; for example, data about costs and demand, their intrinsic productivity or efficiency (Laffont and Tirole 1986) or technological expertise (Polk 2002: 22). Corporations are also more likely to engage in research than other actors (Schlozman and Tierney 1986). And because political decisions are often very important for specific sectors, relevant industry trade associations have strong incentives to concentrate forces. Policymakers, by contrast, have to spread their forces across a diverse range of policy areas at the same time, which puts them at a disadvantage. In the words of one industry representative: I am actually surprised how often [ministerial staff] ring me up looking for information I would have assumed they would have at their fingertips. Maybe it’s just a matter of us having access to several thousand members, and [the members] trust us, so we survey them. I think we are a good source of reliable data. Interview with Ciaran Higgins Much of business’ private information is legally, and in many countries even constitutionally, protected. It is privately owned in the most literal sense of that phrase. Information is part of capitalist property rights over productive assets, and it is widely assumed that capitalism cannot function without this protected informational realm. The legal protection of information about production and production costs is a constituent feature of capitalism. Information that is openly accessible for purposes of taxation and auditing, however, is often insufficiently informative for policymakers. Everything else is ‘corporate secret’. Kennan and Wilson describe the implications of business’ legally protected information realm and the resulting asymmetric information between business and trade unions in the context of collective bargaining and industrial disputes: A firm may open its books along with publicly accessible accounting and financial statements … to the union, but even so there are substantial limitations on the relevance if this kind of information. Partly this is because the past only partly reveals the future: A firm’s predictions of production demand and prices of factors and output can be manipulated. Kennan and Wilson 1993: 49 Furthermore, actual production costs may be less important than opportunity costs. Policymakers or labour unions will often be unable to assess the opportunities the

60 An informational–structural model of business power firm has to redeploy its capital in other locations, product lines or industries, or substitute capital for labour. Ultimately, it is the firms themselves who decide whether or not the alleged negative inducement effects become reality. Business’ informationally privileged position thus stems from the same source as its structurally privileged position in Lindblom’s sense – the control over allocation and investment. The same private property rights that provide business with privileged access to information about the costs of policy is also what, in principle, puts business actors in the position to cause the predictions based on that information to materialize. A capitalist firm, by definition, makes its own business decisions. Intentions are internal states of mind. Unlike in Steven Spielberg’s science fiction film Minority Report, in reality intentions are inaccessible to others until they are communicated by the person that holds them. Thus, a business actor’s lobbying message is a prediction about his own future behaviour, contingent on the actions of a policymaker. What the business actor intends to do is, logically, private information: others can only guess what his intentions might be. As has already been suggested, the notion of business’ informational advantage does not rely on an assumption that business is perfectly informed about the real-world impact of policy. All that is needed is the assumption that business is frequently better informed than the policymaker. However, the existing uncertainty of business does imply that, if in doubt, business lobbyists should hedge their bets and assess policy from a primarily conservative; that is, negative, perspective. According to an executive with the Irish Software Association, [i]t’s not about being negative for the sake of it all the time. I am sure there are plenty of cases where IBEC would be very cautious. Sometimes, it may be that what ultimately came out of the whole process mightn’t have been as bad as first was foreseen, but I guess it’s because people overestimated it genuinely. I think it’s human nature to look at what’s negative. Interview with Ciaran Higgins How do lobbyists bring their messages across in a credible manner? Access to the government is facilitated by having the reputation of being a trustworthy supplier of good information. But access is not perfectly conditional on informedness, reliability and truthfulness. Business lobbyists have incentives to misrepresent the likelihood or size of negative inducement effects, and these incentives are reinforced by the fact that, just like other lobbyists, they tend to frame their preferences in such a way as to be congruent with the wider interest of society. Typically, messages sent by business lobbyists to policymakers already invoke the link between the policy’s effects on business, the general interests of society and the policymakers’ electoral interests. In the words of a German public affairs representative, Of course, private interest representation is private interest representation; there is no denying that. However, we do frame our goals in such a way that they are congruent with the common good, and by and large they are,

An informational–structural model of business power 61 not least in the sense that a tunnel vision approach to interest representation would backfire. Interview with Marc Schnerr As Mitchell observes, ‘[j]ust as the boundary between expertise and self-interest becomes uncertain, so does the line between the firm’s interest and the community interest’ (Mitchell 1997: 48–9). It is therefore possible for businesses to make exaggerated or otherwise inaccurate claims about a policy’s negative inducement effects. If these messages achieve their desired effect of averting the policy, any lies will go undetected. Thus, there is a possibility that policymakers are misled into believing that their policies would have considerable negative inducement effects even if the true costs of policy are much lower. As one Irish industry representative put it: ‘We can tell them pretty much anything – how would they know?’ (Interview with Ciaran Higgins). But while information provided by business is frequently seen as ‘possessing a particular validity’ (Coen and Grant 2006: 18), policymakers have reasons to be wary of the cost-predictions of self-interested firms. One and the same message can come from two different types of lobbyists – those severely affected by the policy and those on whom the policy imposes only minor costs that they would nonetheless prefer to avoid. Both types of lobbyist thus share an interest, each for their own good reasons, in having the policy altered or removed from the agenda and may therefore send identical messages to the policymaker. The problem for the policymaker is that she does not know which type of lobbyist she is dealing with. Policymakers are not naïve and know that lobbyists are self-interested. Simple self-serving assertions that do not directly affect players’ payoffs would therefore be dismissed as ‘cheap talk’; that is, as non-verifiable claims.1 Unsubstantiated speech acts, therefore, have little effect on the beliefs of the policymaker (Farrell and Gibbons (1989: 1215), leading policymakers to dismiss obviously self-serving assertions. Referring to his time at the US Office of Price Administration during the Second World War, the late John Kenneth Galbraith recalled how he and his colleagues communicated their conviction to each other that an industry representative tried to misinform them. Having earlier observed a colony of hungry ants, antennae quivering with excitement at the sight of a pile of horse manure in the street near their offices, they simply imitated the insects’ display: When an industrialist, already making more money than in the previous decade, came to see us to ask permission to put his prices up, my staff and I would quiver our index finger and the next one to it on the desk, as a sign to each other that the guy was talking horse shit. The Guardian, 14 August 2004 This illustrates that it is simply not sufficient for business lobbyists to reiterate in one version or another the old mantra that ‘what’s good for the United States is good for General Motors, and vice versa’, as was famously stated by the former CEO of General Motors, Charles E. Wilson, during the 1953 Senate hearing that

62 An informational–structural model of business power confirmed him as the US secretary of defence. A lobbyist who has something to say that is actually of interest to the policymaker must therefore find a way to differentiate himself from the babbler that tries to mimic him. If words are not sufficient to convince a policymaker that the prediction of negative inducement effects is true, action is required to underscore the seriousness and credibility of the claim. Such differentiating, or ‘separating’, can be achieved by supplementing the lobby message with some costly action that the ‘other type’ of lobbyist would not be expected to find a sacrifice worth making. Kennan and Wilson (1993) illustrate the logic of information transmission through costly political action in the context of wage bargaining and industrial relations. Bargaining is substantially a process of communication necessitated by initial differences in information known to the parties separately. Delay or endurance tactics are often required to convey private information credibly. For instance, willingness to endure a strike might be the only convincing evidence that a firm is unable to pay a higher wage. Likewise, the costly pursuit of a damage claim might signal that the plaintiff can muster sufficient evidence to convince a court of the magnitude of her injury. In Kennan and Wilson’s words, [c]redibility is the important ingredient of this hypothesis. The firm can always assert that a high wage would be ruinous, but it may be profitable for the union to verify this assertion by striking long enough: the firm would profit by capitulating early if it could afford a high wage. Kennan and Wilson 1993: 46 Thus, in the context of bargaining games, delays and muscle flexing constitute costly actions with the aim to deliver convincing clues about the accompanying claim’s truthfulness. The same logic applies to the lobbying of policymakers by business interests.2 In the language of game theory, the situation in which one type of business, but not the other, shares an interest with the policymaker, while both types of business have the same preference with respect to the policy outcome, is characterized by a partial conflict of interests. The interaction between business and the policymaker in this setting becomes a signalling game in which a business lobbyist ‘reminds’ the policymaker that she is structurally dependent on business as far as her electoral fortunes or the success of her policy are concerned. At the same time, the policymaker is aware that the relevance of this structural dependence varies according to a parameter that is known to the lobbyist but unknown to her: whether or not the pending proposal would really produce negative inducement effects. But just what constitutes costly political action in the context of political lobbying, and how plausible is the notion that costly campaigns convey credibility on the lobbying message? Even the larger sums afforded by business for political purposes are generally dwarfed by the sums that change hands on a daily basis in the normal market transactions of commercial life. A possibly apocryphal anecdote of a ‘well-known entrepreneur’ quoted by Hart (2002: 15) illustrates the point: ‘When I put $250,000 on the table in a business meeting, people laugh.

An informational–structural model of business power 63 In a political meeting, they bow’. Business funds for political action tend to be peanuts, and while politicians might find the absolute sums at play impressive, they also know that they are not so impressive when considered in the context of the day-to-day turnover of the firm that dishes them out. But it is also common knowledge that public affairs departments and government affairs offices do not enjoy the highest priority in the eyes of senior management. Companies of all sizes face budget constraints, and in times of economic misfortune or recession government affairs departments are usually the first to be cut back (Coen and Grant 2006: 24). This is exactly what provides expenditure on business political action with its persuasive property, and more so than in the case of non-business groups. The typical non-business interest group is a more or less professionally run, voluntary association of citizens whose primary goal it is to influence public policy. The ideal-typical business corporation, by contrast, is an organization that seeks to maximize profits. It does so mainly through commercial market activities; investment in political capabilities is merely one among many less central means for doing so (Hart 2002: 10–1). Any firm’s government affairs functions usually receive only a tiny fraction of corporate revenues, but in a crisis a firm’s central leadership can divert funds to match virtually any challenge (Hart 2002: 14–5). When a policymaker observes costly activity that is in this sense out of the ordinary, she has reasons to infer that the business lobbyist in question either really does mean it seriously – or, alternatively, is mimicking a serious one. To illustrate how different types of lobbyist try to mimic each other and how the recipient can go about finding out whose message is genuine, the next section describes the generic structure of a signalling game.

Signalling models of lobbying A signalling game captures a situation of asymmetric information between two players. The choice of the two players game-theoretical framework is independent of the degree of competition or oligopolistic distortions of the market in which the businesses in question operate. In perfectly competitive settings, a decisiontheoretical framework might be thought more appropriate for analysing the choices made by firms. In that case, the implications of the decisions made by other actors confront the actor anonymously, as ‘market forces’. Thus, if the number of firms is large enough and competition sufficiently close to perfect, large-N games converge to N players’ non-strategic decision problems. In the present non-market context, however, the case for game theory is strong even if competition among firms in the sector to be regulated or taxed was assumed to be perfect. Sectorally defined interest groups, peak associations and trade and industry organizations often lobby the government quasi-monopolistically on behalf of their member firms and rarely in a decentralized manner. Lastly, the government itself is a monopolistic actor confronting business.3 While the game itself is of a finite and non-cooperative nature, the long-term importance of both lobbyists’ and politicians’ reputations can be seen as part of a wider societal and political game of electoral competition, as well as of a co-operative game of collective action required for the provision

64 An informational–structural model of business power of public goods such as a pollution-free environment. In repeat-play versions of this game, business will care about its reputation as an informative source of policy advice because it wants to retain the chance for an informational impact on future decisions. In the one-shot version outlined here, the cost of damaging one’s reputation is included directly in the payoffs. One player has private information that is relevant for a decision made by the other player. The second player’s decision affects both players’ payoffs. Before the second player decides upon an action, the first player can send some sort of signal. This signal can be a verbal statement (e.g. ‘If you, government, increase the fuel tax we will be forced to close down a plant and shed workers’) or a meaningful action that reveals information indirectly (e.g. starting to fire workers after the government announced a rise in fuel tax). By sending appropriate signals the first player can influence the second player’s beliefs and in this way affect that player’s choice. The first player may have an incentive to reveal false information or to avoid revealing any information at all. This type of game is well-suited for studying the informational role of interest groups and has often been applied in that context (e.g. Potters and van Winden 1992; Ainsworth 1993; Austen-Smith 1993). The following outline of a signalling game is informed by overviews in Sloof (1998) and Polk (2002). Common to all of these is the notion that special interest groups are better informed on issues that affect them than policymakers, which enables them to play a role in the policy process by sharing (some of) their private information with the relevant branches of government. Policymakers have incentives to listen to special interest groups when they are imperfectly informed and special interest groups possess better information about the policy area in question. In this way, lobbying gives policymakers a chance to improve their decisions as well as special interests a chance to influence these decisions in their favour. As interest group’s preferences over outcomes are rarely identical to those of policymakers, they have incentives to behave strategically when transmitting information to policymakers (AustenSmith 1993). Lobbyists are self-interested and policymakers cannot simply trust their speech acts. They have to extract the informative part of the received messages, based on their a priori information; that is, on the hunches, experience and other information they bring to the game before observing the actions or listening to the words of the lobbyist. Special interest groups, on the other hand, know that policymakers discount their messages on these grounds. They adjust their strategies accordingly to appear trustworthy. The general setup of a signalling game is as follows: There are two states of the world, t1 with probability θ1 , and t2 with probability θ2 = 1 − θ1 , whose realization is not known to the policymaker, but is known to the special interest group. There are also two actors, a lobbyist B and a policymaker P. Technically, a third player, ‘nature’, chooses between t1 and t2 . In this way, problems of incomplete information can be analysed as sequential games of imperfect information. The policymaker has prior beliefs about the probabilities with which each state of the world occurs and implements one of two policy alternatives, a1 and a2 . The payoff to the policymaker is high (0) if she implements the ‘right’ policy in each state of the world and low (−u) if she implements the

An informational–structural model of business power 65 wrong one. The policymaker prefers a1 if the state is t2 , and a2 if the state is t1 . Thus, the task of the policymaker is to choose a correct policy availing only of incomplete information. The payoffs to the special interest group are arbitrary. They will be denoted by 0, α and β, and their specification determines the kind of game that is played. Some common games will be outlined in turn. Their payoff matrix is shown in Table 3.1. The matrix and the prior probabilities are common knowledge; that is, they are known to all players. Suppose that α > 0, and β = 0. In this case, there is no conflict of interest between the policymaker and the special interest group. They share the same interest, which is to find the policy suitable for each state of the world. There may or may not be an incentive for the special interest group to send a message. This depends on the – as yet unspecified – lobbying costs and on the policymaker’s a priori beliefs. The important property of this parameter specification is that the policymaker can always trust the signal she receives from the special interest group. Knowing that the lobbyist has no incentive to mislead her, the policymaker is perfectly informed by the time she chooses between a1 and a2 . Without a signal, she knows that the special interest group has no incentive to send one, which indicates that the choice based on her a priori belief is the right one. Suppose next that α > 0 > β. This parameter range represents a full conflict of interest between the special interest group and the policymaker. In his situation, the policymaker knows that she is better off mistrusting any signal she receives. Therefore, she will always ignore any signal. Facing such an unresponsive policymaker, a lobbyist knows that whatever he does will have no effect. Hence, he will never send a signal if doing so involves any costs, no matter how minimal. In this setup, with the interests of the two players fully opposed and any signal falling on deaf ears, there is no scope for meaningful communication. The parameter range in which lobbying becomes interesting is α > β > 0. This is the situation of partial conflict of interest between the special interest group and the policymaker: whether or not the two players have the same interest depends on the state of the world. The lobbyist always prefers policy a2 over a1 , regardless of the state of the world. If the state is t1 , the policymaker shares the lobbyist’s interest. In this situation, the lobbyist, who observes the state of the world and has perfect information, offers to guide the policymaker in her choice. He can inform the policymaker about the realized state by sending a signal. The policymaker, who observes only the signal but not the true state of the world, anticipates that

Table 3.1 Payoffs of a signalling game (lobbyist, policymaker) α > β, u > 0

a1 a2

t1

t2

−α, −u 0, 0

−β, 0 0, −u

66 An informational–structural model of business power sending a signal is only worthwhile for certain types of interest groups and updates her beliefs accordingly. In contrast to the case involving a full conflict of interest, the policymaker will not outrightly ignore any message because she knows that in some cases truth-telling is the lobbyist’s preferred strategy. Hence, she can learn something from the signal she receives. If the state of the world is t1 , any lobbyist will be referred to as a type-1 lobbyist; if the state of the world is t2 , any lobbyist will be called type-2. Recall that unsubstantiated assertions – cheap talk – will not be deemed informative by the recipient if she believes that these assertions are self-serving at least some of the time. In this situation, a so-called ‘sorting condition’ must be introduced which ensures that the type-2 lobbyist cannot always mimic type-1. If the sorting condition does not hold, the payoff to type-2 is always equal to or higher than the payoff to type-1. To achieve this, a type-2 interest group only needs to mimic the behaviour of type-1, which would make sense for him to do whenever β > 0. Anticipating this, a sceptical policymaker will mistrust any signal, and information revelation will again not be possible. If, however, the sorting condition is satisfied, imitation is only sometimes possible. To enable the sorting condition, an additional element – the costs of lobbying (l) – must be introduced to the payoff matrix which ultimately affects the type-1 lobbyist differently from type-2. With lobbying costs in place, the type-1 lobbyist has an incentive to send a signal that type-2 does not have. The policymaker knows the decision problem of the lobbyist and is able to make inferences about the state of the world from the observed signal. The expanded payoff matrix in Table 3.2 includes the costs of lobbying. To accommodate the additional information, the table is unfolded into two panels, one in which the special interest group sends a costly signal (s1 ) and another in which he does not (s2 ). The solutions to this game rest on the concept of ‘sequential equilibrium’ (Kreps and Wilson 1982). This is a refinement of Nash equilibrium, characterized by strategies for each player and probabilistic beliefs about the state of the world at the time a player is called upon to act. A strategy is Nash equilibrium if each player’s strategy is an optimal response to the other players’ strategies (Nash 1951). For sequential games such as signalling games, the additional criterion of subgame perfection is required. A strategy is subgame perfect if, for every proper subgame, the strategy restricted to the subgame constitutes a Nash equilibrium for the subgame (Selten 1965). Table 3.2 Payoffs of a lobbying game (lobbyist, policymaker) α > β > 0 and u, l > 0 s1

a1 a2

s2

t1

t2

t1

t2

−α − l, −u −l, 0

−β − l, 0 −l, −u

−α, −u 0,0

−β, 0 0, −u

An informational–structural model of business power 67 Matters become slightly more complicated due to the fact that the signalling game is a sequential game involving asymmetric information (i.e. imperfect information on the side of one player). For a strategy to be Nash equilibrium in a sequential game of imperfect information it must also meet the requirement of sequential rationality. Every decision must be part of an optimal strategy for the remainder of the game. In other words, at every juncture a player’s subsequent strategy must be optimal with respect to some assessment of the probabilities of all uncertain events, including any preceding but unobserved choices made by other players, including the unobserved choices made by ‘nature’ between different states of the world. This requires that an equilibrium specify beliefs as well as strategies (Kreps and Wilson 1982: 872): ‘The point to be stressed here is that an assessment (and not simply a strategy) will or will not be a sequential equilibrium’. The equilibrium strategies are chosen such that each player’s expected payoff is maximized, given the other players’ equilibrium strategies and their beliefs about the state of the world. Beliefs are derived from a priori knowledge and updated after observing the other player’s actions according to Bayes’ rule.4 Figure 3.1 shows the tree of the generic form of this type of signalling game. For the game described in Table 3.2, the resulting equilibria depend on the signalling costs l, and on the policymaker’s a priori belief. If signalling costs are high

B a1 s1 B t1

−a

a2 a1

s2

q1

a2

q2

a1

−l

P −u

−l

0

−a

−u 0

0

Nature t2

s1 B

P

a2 a1

s2 P

−l

−b

a2

−l −b

0 −u 0

0

−u

Lobbyist’s (B) payoffs

Policymaker’s (P) payoffs

a

Severe policy costs

u

b

Mild policy costs

l

Lobbying costs

Figure 3.1 A generic signalling model of lobbying

Policy costs

68 An informational–structural model of business power (l > α), a unique pooling equilibrium exists in which the special interest group never lobbies, and the policymaker makes a decision based on her a priori beliefs. By contrast, if lobbying costs are intermediate (α > l > β), two types of equilibria exist, which are sensitive to the a priori belief of the policymaker. A type-2 lobbyist will never send a signal, since this would result in a negative payoff even if the policymaker is persuaded to enact a2 . The focus is therefore on type-1 and his incentive to send a signal, which in turn depends on the policymaker’s prior beliefs concerning θ1 . If the policymaker is inclined to choose the ‘wrong’ policy based on her a priori beliefs, the unique equilibrium is separating, and the type-1 interest group always lobbies. In the language of signalling games, a type-1 lobbyist ‘separates’ if his chosen strategy enables the policymaker to distinguish t1 from all other types of lobbyist. Conversely, ‘pooling’ means that all types of lobbyist chose the same strategy (Banks 1991: 7). Anticipating this, the policymaker trusts the signal and chooses policy a2 after receiving a message and a1 otherwise. A type-1 lobbyist has no incentive to deviate, because not sending a signal would lead to the unfavourable policy and −α payoff. However, if the policymaker’s prior beliefs induce her to enact policy a2 without signalling, two equilibria exist. The first is a pooling equilibrium in which no signal is sent. In this situation the policymaker chooses the policy based on her prior beliefs, which is the appropriate one. But there is also a separating equilibrium: If the policymaker assumes that the type-1 lobbyist always indicates t1 when it occurs, she only selects policy a2 if t1 is affirmed by the lobbyist. In this case, signalling is optimal only for the type-1 interest group. Finally, if lobbying costs are low (β > l), type-2 lobbyists have incentives to mislead the policymaker. But they take into account that the policymaker knows their incentive to do so. Lobbyists, therefore, need to provide an incentive for the policymaker to listen to the signal. This incentive is not given if the type-2 lobbyist always signals, because the policymaker would anticipate this strategy and ignore any signal. Thus, for a type-2 interest group lobbying is only a good strategy if it still allows the policymaker to extract some information from the signal. This is the case if the probability of a signal being sent by type-1 is higher than the respective probability for type-2. This condition, in turn, will hold under the following circumstances. If the policymaker’s prior belief about the state of the world leads her to enact a1 in t2 , the type-2 lobbyist has an incentive to always send a signal. Otherwise, type-1 could mimic type-2. A unique semi-separating equilibrium exists in which type-2 always sends a signal, and type-1 plays a mixed strategy and mimics type-2 every now and then. In this case, the policymaker is able to extract some limited information from an observed signal. After having observed a signal, it is not optimal for the policymaker to choose a2 in any case, because doing so would give the type-1 lobbyist an incentive to send a signal in any case. Hence, in equilibrium, the policymaker will sometimes choose a2 if a signal is received, and always choose a1 if no signal is forthcoming. Accordingly, the type-1 lobbyist will send a signal every now and then. Next, consider the case where the policymaker chooses a2 in t2 if she has to make a choice based on her a priori beliefs. In this case, no separating equilibrium exists,

An informational–structural model of business power 69 since this would require that the type-2 lobbyist always sends a message, whereas type-1 never does. But as argued above, this is not a good strategy for type-1 if signalling costs are sufficiently low. Thus, with β > l, there are only pooling and semi-separating equilibria. However, two different pooling equilibria exist. In the first one, neither type of lobbyist has an incentive to send a costly signal, and the policymaker chooses the best policy based on her a priori beliefs. In the present case, this means that neither type of lobbyist sends a signal and the policymaker always chooses a2 . In the second pooling equilibrium, both types always lobby and the policymaker chooses a1 after observing the signal. Not sending a signal cannot be optimal for either lobbyist, because this would induce a2 , which yields a strictly lower payoff for any type of lobbyist. Finally, in the semi-separating equilibrium, the type-1 interest group never sends a message and type-2 lobbies every now and then. The policymaker anticipates these signalling strategies. She knows that a signal is a sure indicator of the lobbyist being of type-2, but that the state of the world is not necessarily t1 if no signal is sent. Accordingly, she chooses a2 if she observes a signal and randomizes between a1 and a2 if she does not observe one, because she knows that the type-2 lobbyist sometimes remains silent. In situations like these where multiple equilibria exist, a slight change in any variable can lead to a jump from one equilibrium to another. This makes it difficult to obtain general comparative static results. But because lobbyists have to weigh the expected benefits against the costs of lobbying, we can generally say that the likelihood of lobbying increases in the payoffs to the lobbyist and decreases in the signalling costs. This result seems intuitively plausible and is valid in most cases (Polk 2002: 22). However, these generalizations can break down under certain circumstances, as the lobbyist must take into account the policymaker’s reaction in the assessment of his lobbying strategies. If, for example, the benefit from lobbying increases in state t2 , the policymaker anticipates the interest group’s higher incentive to lobby. This, in turn, makes her less willing to react to lobbying, which again decreases the incentives for dishonest lobbying. The general likelihood of lobbying depends on the magnitude of these effects, but tends to increase in the payoffs (Polk 2002: 22).

Reputation and lobbying In the generic lobbying games outlined above, the policymaker knows the decision problem of the lobbyist, and is therefore able to infer the best content from any observed signal. This means also that only the decision to send a signal or not is important, but not the contents of that signal. All that matters for lobbying to acquire informativeness, therefore, are the signalling costs but not the contents of the signal itself (Polk 2002: 18). This implies that the costs of lobbying are the same irrespective of whether or not the lobbyist tells the truth about the state of the world. Furthermore, whether a special interest group is lobbying a longtime ally or an adversarial politician, the policymaker is never fully certain of the salience of the lobbyist’s cause (Ainsworth 1993: 41). In the real world of politics,

70 An informational–structural model of business power however, whether the policymaker is an ally or foe might matter a great deal for the lobbyist’s payoffs and therefore for the way the policymaker interprets the message. This leads to variation in the countervailing constraints on the making of exaggerated claims in addition to the material costs of lobbying. Pluralists have long argued that there are costs in taking political action other than those related to personal time and effort involved in political mobilization, to becoming informed or to the actual ‘physical’ activities of lobbying and campaigning. These other costs refer to the possibility of losing and of making political enemies (Polsby 1960). Above all else, they refer to the importance of an actor’s reputation. According to Kennan and Wilson (1993: 46), reputational payoffs provide ‘strong motives for steady adherence to forthright revelation of private information, based on building and maintaining a valuable reputation for honesty over many repeated bargaining relationships’. These reputational considerations explain the reliability of intermediate agents, such as brokers and attorneys, whose longer-term interests are served by adherence to professional standards of conduct. Employers and unions that repeatedly negotiate short-term contracts may have similar incentives to be forthright in their negotiations (Kennan and Wilson 1993: 46). Special interest groups in general have strong incentives to maintain their reputation in the eyes of policymakers. Berry (1997: 98) finds that credibility is a lobbyist’s greatest asset: ‘Lobbyists simply cannot do the job if there are any doubts about their credibility’. As a corporate representative in Washington DC put it: ‘All you have is your word’ (quoted in Berry 1997: 98). Business groups in particular have a strong, long-term interest in maintaining their reputation as suppliers of good information. Hart (2002: 22) suggests that US firms place a value on their political reputation of being ‘a good citizen’. According to a German lobbyist, that holds not only in the US: If we did some scaremongering with figures that could in retrospect be dismantled by other research institutes, we would manoeuvre ourselves off the pitch because nobody would believe us anymore. Where I work, I don’t think this bargaining scenario of ‘ask for 100, take home 50’ really applies. Interview with Marc Schnerr Thus, it is reasonable to assume that when devising a lobbying strategy, business will be careful to avoid situations in which wildly exaggerated predictions can be exposed as such. This also implies that the likelihood of business political efforts being successful will become a major determinant of action (Bonardi, Hillman and Keim 2005). Thus, not only do business lobbyists have to bestow credibility onto their messages by underlining them with costly campaigning,5 they also have to consider the reputational costs of ‘lying’ as well as the material costs of lobbying. The following section describes a signalling model proposed by Bernhagen and Bräuninger (2005) that incorporates the reputational costs of lobbying in the payoff structure while integrating political lobbying with the implications of business’ structurally privileged position in capitalist democracy.

An informational–structural model of business power 71 In the model, there are two actors – a business lobbyist B and a policymaker P, and a pending policy proposal that may or may not have significant negative implications for the economy or the success of the policy. Because policymakers sometimes have incentives to cater for non-business interests, the proposal has found its way onto the political agenda despite the fact that it is disliked by both types (type-1 and type-2) of affected businesses. For the ease of exposition, the present analysis makes the simplifying assumption that the policy in question has either no, or only negative, effects on business and, after the distinction of type-1 and type-2 business actors, affects all firms and industries equally. The game then proceeds as follows. Nature chooses whether a pending policy will lead to negative inducement effects (t1 ) with probability θ1 , or not (t2 ) with probability θ2 = 1 − θ1 , and reveals its move to the business lobbyist but not the policymaker. The business lobbyist is, from then on, privately informed about the state of the world. He can choose to send a costly message (s1 ) to the policymaker to inform her decision, or refrain from doing so (s2 ). The policymaker can take one of two actions: pass the policy in question (a1 ), or retain the status quo (a2 ). The resulting payoffs for the policymaker and for business depend on which of the two states of the world is realized. In t1 the policy has severe negative effects of size α on business. Also in t1 the policy has negative inducement effects, either because it negatively affects the economy and thereby the policymaker’s popularity or her ability to achieve various other policy goals, or by leading to other undesired effects of the policy. The negative inducement effects for the policymaker are represented by u, which can be interpreted as the extent to which she will suffer electoral and/or fiscal punishment for enacting a policy with undesirable side effects. It was argued in the previous chapter that policymakers have incentives to keep election promises and honour pledges. If the policymaker chooses a2 , she incurs a knock to her credibility and reputation and will be punished at the next election for defaulting on her policy pledges and selling out to special interests. These pledge costs are denoted by p and apply regardless of the state of the world. They thereby affect the policymaker’s payoffs differently than in the standard signalling models outlined above. In the classic models, u occupies the diagonal cells of the payoff matrix, thereby denoting the cost incurred by the policymaker for picking the ‘wrong’ policy in a symmetrical decision dilemma. The decision problem for the policymaker comes down to picking the right policy given the state of the world; that is, choosing a1 if the state of the world is t2 , and a2 if it is t1 . In the context of the signalling game between business and policymakers specified here, this asymmetry no longer holds. Instead, the policymaker only incurs u if she enacts a1 in t1 , but she incurs p every time she chooses a2 . Business incurs minor policy costs of size β in situation t2 , but is relieved of all costs net of any material lobbying costs l if the lobbying strategy is successful; that is, if the policymaker complies with the lobbyist’s demands (a2 ). However, if it turns out that business sent an untruthful signal, it will incur an exogenous penalty for lying k. This can be interpreted as being denied access to the policymaker for future representations. Misinformation can only be detected once the policymaker

72 An informational–structural model of business power enacts the policy (a1 ). Note that business will also be penalized it it fails to warn the policymaker about the realization of t1 are ignored here. Finally, while the policymaker can to some extent infer the credibility of the claims from the observed levels of costly political action business is willing to employ in the lobbying process, there are limits to playing tough and enticing business into showing its teeth. Research on the economic effects of regulation has shown that expectations and uncertainty concerning regulatory change may be just as important for the investment behaviour of firms as the actual regulatory change itself (Binder 1985; Granato 1996). Recall also that according to Przeworski and Wallerstein’s (1988: 22–4) analysis of the structural dependence of the state on capital it is not so much the actual policy that has negative inducement effects, but rather the anticipation of a policy in a dynamic setting that may undermine business’ willingness to invest and thereby harm the policymaker. Open disputes over contested legislation can have negative effects on investment behaviour, conferring on the policymaker a confidence cost v whenever business decides to lobby. The payoffs of this more complex game are summarized in Table 3.3. To investigate the interaction between the policymaker and the business lobbyist, the game is analysed using the concept of sequential equilibrium. The conditions of sequential equilibrium require that for any type t of business its strategies must be optimal given the strategy of the policymaker. Furthermore, the policymaker’s strategy must maximize her expected payoff given her posterior beliefs about the type of business. Finally, the policymaker’s posterior beliefs must satisfy Bayes’ rule whenever possible. The following paragraphs describe sample equilibria for the two most interesting scenarios.6 In the first scenario, the reputation costs of business are large compared to the material costs of lobbying (l < k), and the negative inducement costs for the policymaker are small compared to her pledge costs ( p > u). In the second scenario, the costs of negative inducement effects are larger than the pledge costs ( p < u), while both business’ expected costs from the policy in t2 and material lobbying expenses are considerable compared to business’ reputation costs (l < β). There is a separating equilibrium if p > u and l < k. In a separating equilibrium the two possible types of lobbyists send different messages and the policymaker, having observed the signal, can infer the state of nature with certainty. Consequently, her updated belief coincides with the true type of business. Table 3.3 Payoffs of a lobbying game with reputation costs and confidence costs (business lobbyist, policymaker) α > β > 0 and u, v, p, k, l > 0 s1

a1 a2

s2

t1

t2

t1

t2

−α − l, −u − v −l, −v − p

−β − l − k, −v −l, −v − p

−α − k, −u 0, −p

−β, 0 0, −p

An informational–structural model of business power 73 The policymaker expects relatively high pledge costs from retaining the status quo and consequently always passes the policy whatever the signal. The lobbyist still separates, because the costs for lying exceed the material lobbying costs he would be willing to pay. Thus, if pledge costs are high, the separating behaviour on the side of the lobbyist can be in equilibrium even if the costs for lobbying are low. The game tree in Figure 3.2 illustrates this equilibrium; the equilibrium paths are highlighted in bold print. Next, there is a pooling equilibrium if l < β and p < u. In this scenario, the lobbyist always claims to be severely affected. In a pooling equilibrium, both types of lobbyist send the same message s1 predicting negative inducement effects of the pending policy. This equilibrium is illustrated in the game tree in Figure 3.3. This only happens if both cost ratios are small; that is, when there is little incentive to misrepresent on the side of the lobbyist and no incentive for the policymaker to push through with the policy regardless of any lobbying. In this equilibrium, type-2 lobbyists can get away with telling a lie. In this situation, the policymaker strictly speaking cannot learn anything from the message and her beliefs do not change after receiving the signal. Thus, the lobbying act should be interpreted as

B a1

Pass

−a

−l

P −u−v

Lobby

Negative effect t1

B

s1

a2

s2

a1

−a−k

a2

0

a1

−b−k−l

Silence

q1

−l

Drop

−v−p −u −p

Nature t2

q2 s1

No effect B

a2

P

a1

s2

−l −b

a2

P

−v−p 0

0

Business’ (B) payoffs

Policymaker’s (P) payoffs

a

Severe policy costs

u

Severe policy costs

b

Mild policy costs

p

Pledge costs

k

Reputation costs

v

Lobbying costs

l

Material lobbying costs

Figure 3.2 A separating equilibrium of business lobbying

−v

−p

74 An informational–structural model of business power

a1 s1 Negative effect t1

B

Pass

B −a −l

Drop

−l

P −u−v

Lobby a2

−a−k

a1

s2

−v−p −u

Silence

q1

a2

0

a1

−b−k−l

a2

−l

−p

Nature t2

q2 s1

No effect B

P

P

Business’ (B) payoffs

−b

a1

s2

a2

−v −v−p 0

0

−p

Policymaker’s (P) payoffs

a

Severe policy costs

u

Severe policy costs

b

Mild policy costs

p

Pledge costs

k

Reputation costs

v

Lobbying costs

l

Material lobbying costs

Figure 3.3 A pooling equilibrium of business lobbying

the ‘friendly reminder’ implied in Lindblom’s account of the way policymakers become aware of negative inducement effects. But note that if the policymaker had more reason to be resolute (e.g. if p increased), a type-1 lobbyist would do well to play a mixed strategy and pool only some of the time. This would lead to a semiseparating equilibrium in which the policymaker can extract some information from the observed behaviour of the lobbyist. The game-theoretic analysis in this chapter suggests that the behaviour of the lobbyist and the policymaker’s response both depend on the ratio of business’ material lobbying cost in relation to the costs of being caught lying, and on the size of the policymaker’s pledge costs relative to her costs from negative inducement effects. As a result, in some circumstances business can achieve its political goals by virtue of its privileged information about the consequences of policy. This means that the political power of business is at least partly due to business’ informational advantage vis-à-vis policymakers with respect to the effects of policy. ‘Knowledge’, as Hobbes pointed out over 350 years ago, emerges as one of ‘severall sorts of Power’ (Hobbes 1996 [1651]: 53). The degree to which policymakers depend on business for the supply of information, impact assessments and

An informational–structural model of business power 75 predictions should therefore be a key variable in the study of business political power. But is the ability of achieving one’s goals by instilling false beliefs in political decisionmakers a form of political power? It could be argued that the analysis of business lobbying as a signalling game is really about gullibility rather than power. Of course, no firm would forgo future profits solely in order to ‘punish’ the policymaker for implementing an undesired policy. In the case of truthful lobbying in the presence of negative inducement effects – the classic case of structural state dependence – businesses might indeed have little choice but to disinvest, close down plants and so on, were it not for the fact that a timely signal and persuasive lobbying might convince the policymaker to change her mind. In this case, business gets what it wants because its rational responses to unwanted policy – disinvestment – would lead to outcomes that are also harmful to the policymaker. We could follow Sloof and van Winden (2000: 86) here and call this an instance of ‘structural coercion.’ These authors define ‘structural coercion’ in opposition to ‘pressure’ when talking about the enforcement of a threat by a strong actor for whom this enforcement is in itself profitable action. But note that the same scenario is also equivalent to what Dowding (1996) would call ‘systematic luck’ – with the difference that in the informational account it takes informative action on the part of business to kick-start the lucky circumstances. More importantly, the existence of pooling and semi-separating equilibria suggests that sometimes business can get policymakers to comply with its preferences even if the negative inducement effects central to notions of structural dominance or ‘systematic luck’ do not in fact apply. The power wielded through the private information held by business about crucial parameters for policymaking is then power in the sense of Dahl’s intentional, action-based concept. The informational privileges of business in capitalist democracies enable business actors to get policymakers to enact policy that they would otherwise not enact, or vice versa. In the informational perspective business actors do not exert power by directly coercing policymakers. Instead, they can achieve their goals by virtue of instilling beliefs about the world (though not preferences, as in Lukes) in political decisionmakers. In the following chapter, outcomes of two real-world lobbying games will be explained by reference to the values of the variables that determine the two lobbying equilibria outlined above: the relative size of the reputation constraints of business in relation to its costs of lobbying, and the ratio of the policymaker’s reputation costs from pre-election policy commitments to the costs from adverse policy effects.

4

Two real-world signalling games

This chapter applies the previous theoretical analysis of business power in the context of qualitative data from case studies of business lobbying on environmental and financial services regulation in Britain and Germany. Empirical studies of signalling games have to confront the problem that the information known privately by some players is also inaccessible to the researcher. In this situation, micro-level case studies can serve as useful quasi-experiments in which the central deductive claims of the model are put to a test (Morton 1999: 133–5). For this purpose, rich empirical information on the respective policy and lobbying processes was collected from newspaper and other media reports, official documents including parliamentary minutes, press releases and expert reports by political, financial and technical consultants. Evidence from six lengthy communications with policymakers, parliamentarians, lobbyists and industry experts serves to provide further insights into the historical episodes. Details of the interviews and email communications are reported in Appendix A.

A note on case selection In analogy to the scenarios of the signalling games in Ch. 3, the focus is on critical cases in which elected politicians must consider calling off already-made policy pledges in order to cater to business interests. The cases concern scenarios in which a policy proposal that is disliked by business has entered the agenda. The strategy of studying critical cases provides valuable insights into the underlying mechanisms determining the strength and scope of business structural power that would be black-boxed by a purely correlational analysis. Studying such critical cases is also of substantive interest. As Caul and Gray (2000: 208) point out, ‘[a]fter elections, the winning parties seek to translate their campaign programs into policies and implement them. In practice, this move from policy presentation to policy implementation provides an integral bridge in the representation process’. Critical-case studies can therefore provide litmus tests for what democratic elections are supposed to accomplish: having voters embrace a set of policies that elected officials will pursue. In the present context, critical-case studies can illustrate how the sequentially rational decisions identified in the previous chapter manifest themselves

Two real-world signalling games 77 in practice. The cases were therefore selected on the values of the key explanatory variables such that the crucial parameter constellations of the two lobbying equilibria identified in the previous chapter are represented. Case selection in small-N research in the social sciences is often guided by a desire to emulate Mill’s method of difference or Przeworski and Teune’s (1982) ‘most similar systems design’. This strategy seeks to compare cases whose system properties are similar with respect to as many features as possible in an effort to control for theoretically irrelevant differences, thereby isolating those differences that are relevant. In this way, the method seeks to identify those variables that are different among otherwise similar cases that account for the observation of the dependent variable. In comparative studies of interest group politics, a plethora of institutional and structural variables are assumed to be summarized in countries and policy areas. Thus considered to be ‘contained’ in a black box of policy- and country-specific categories, the entire set of contextual variables is then represented by ‘proper names’ (e.g. ‘Britain’, ‘taxation’ etc). This is why scholars conducting small-N comparisons of interest group lobbying of policymakers usually draw their cases from either the same policy area, the same type of policy (e.g. distributive, redistributive or regulatory) or the same country, or indeed all three of these, in an effort to hold systemic variables constant. For the two case studies in this chapter, the variables that are assumed to be captured by country or policy area in traditional studies of interest group politics are of substantive theoretical importance and are therefore measured directly. For example, the reputation costs associated with the detection of untruthful lobbying will generally be higher in countries where ‘insider’ representation or corporatist patterns of interest intermediation are key characteristics of public policymaking. The size of negative inducement effects can be expected to vary with the policy area concerned. Finally, the probability of negative inducement effects can be interpreted as capturing firm or sector-level performance, the vulnerabilities of particular industries or the more general economic situation. With the study thus involving more variables than cases, the model that is fitted to the cases is of course overidentified. As the case studies serve to trace the causal processes implied by the model rather than providing substitutes for correlational analysis, this is a price worth paying. The two policy areas from which the cases are drawn, environmental politics and financial services regulation, highlight the conceptual limitations of Lowi’s (1964) famous typology of policies as being either regulatory, redistributive or distributive in nature. Both cases involve explicitly regulatory policies that impose rules for behaviour which are specific and individual in their direct impact, but cannot be completely disaggregated to the individual or firm level. Yet the case of banking regulation constitutes simultaneously an instance of redistributive policy, where losers and winners are rather easily distinguishable. The second case, from the world of environmental politics, shares all the characteristics of a regulatory policy in intent, yet it also has a clear and simultaneous distributive impact that is individual and disaggregated in character. In the context of environmental politics, policymakers often make conscious choices between

78 Two real-world signalling games redistributive and non-redistributive means of regulation. Consider the following example from Kraft’s (1996) textbook on environmental politics. Given that motorists drive too much and waste fuel, thereby causing urban congestion and pollution as well as contributing to undesirable increases of carbon dioxide in the atmosphere, two mechanisms are available by which such behaviour can be discouraged. Regulatory schemes summarized under the term ‘command-andcontrol’ policies include the setting of high fuel standards for automobiles or restricting the use of private vehicles through restricting the lanes that may be used by single-occupancy vehicles, to name but two. An alternative, so-called ‘incentives-based’ policy, works by redistributive means involving ‘economic incentives’ instruments. These ‘market alternatives’ to ‘command and control’ might involve raising petrol taxes sufficiently to achieve the same goals (Kraft 1996: 173). It is only at a superficial level that one of these ‘regulatory styles’ appears distribution-neutral, while the other is redistributive. In both cases, undesired behaviour – driving too much and wasting fuel – is made less attractive by authoritative and enforceable means. The difference is that in one case the opportunities to drive fuel-inefficient vehicles become fewer by means of a diminution of the availability of fuel-inefficient vehicles or road lanes, while in the case of a petrol tax hike the price that has to be paid for continuing to drive fuel-inefficient vehicles increases. In both cases, people are affected differentially depending on their financial and socio-economic position, thus adding a redistributive dimension even to ‘command-and-control’ regulation. Effectively, the main difference between the regulatory modi is that rich people will find it difficult – but not impossible – to buy their way out of the imposed restrictions, while they, unlike the poor, can afford to continue engaging ad libitum in wasteful behaviour if the goal is to be achieved by fuel tax increases. Of course, the environmental problems supposed to be tackled by such policies are themselves not distribution-neutral. The burden of environmental pollution falls inequitably on the population, with the poor and otherwise disadvantaged groups generally worse affected than others. Beck was simply wrong when he asserted a fundamental distinction between socioeconomic and environmental issues by saying that ‘poverty is hierarchic’, while ‘smog is democratic’ (Beck 1992: 36). Like most regulatory areas, environmental policy has tangible distributive and redistributive implications. In the case of financial services regulation, where the regulated behaviour involves the handling of financial assets and the distribution of returns, the redistributive implications of policy do not require further explanation. Informational disadvantages of policymakers vis-à-vis business are widely reported in the context of environmental policy. Some degree of industry ‘buy-in’ to policymaking is often deemed inevitable in this area, as manufacturers have essential information about product characteristics, processing technologies and consumer demand (Harrison 1999: 127). Informational asymmetries can also be observed in the finance and financial services sectors. Historically, the interest groups most harmed by inflation, led by the financial sector, have often sought protection by lobbying for the independence of the central bank (Goodman 1991).

Two real-world signalling games 79 The financial sector benefits disproportionately from inflation control, while also having the knowledge to persuade policymakers to adopt this policy. Other interest groups such as trade unions or exporters may oppose this reform, but are ineffective due to their lack of knowledge of this technical issue area and its likely impact on their members (Posen 1995). The control of information relevant to monetary policy thus puts the actors in this sector in a powerful position vis-à-vis policymakers (ibid.: 256). Thus, the case studies promise to provide interesting clues about the role of information asymmetry in government–business relations.

Regulation of small business banking in England and Wales In 1998, the British chancellor of the exchequer, Gordon Brown, asked the director at the Office of Telecommunications, Don Cruickshank, to conduct an independent review of competition in banking services in the UK. Cruickshank’s banking review found, inter alia, that there was a lack of competition in retail services to small and medium-sized enterprises (SMEs). The findings, published on 20 March 2000, were deemed sufficiently serious by both the chancellor and the secretary of state for trade and industry to make a joint reference of the matter to the Competition Commission which, after a two-year in-depth investigation, agreed with the Cruickshank findings on most aspects. In its report, the commission found that a ‘complex monopoly’ existed among the four largest British high street banks, Barclays, HSBC, Lloyds TSB and RBSG, enabling them to secure almost 90 per cent of the market in small business banking in England and Wales and to realize ‘excess profits’ of £725 million a year between them (Competition Commission 2002, Ch. 1: 4).1 Based on these reports, and in close collaboration with the secretary of state for trade and industry, Brown set out regulatory measures designed to encourage competition and curtail the banks’ excess profits. At the core of these was the requirement that the four largest banks pay interest on current accounts in England and Wales at least at Bank of England base rate less 2.5 per cent; alternatively offer SMEs accounts that are free of money transmission charges, as applies in the personal sector; or offer their SME customers a choice between the two options. As a result of these remedies, the banks would sacrifice an annual slice of their profits of over £100 million each. Not surprisingly, they responded negatively to these policy ideas. The banks argued that the measures would fail to encourage competition. According to the head of policy of the British Bankers’ Association (BBA), the measures could undermine the potential competitors’ attempts to break into the market, since part of their unique selling point was paying interest on small business current accounts: ‘If the other banks are forced to do that, they lose that selling point’ (Financial Times, 21 March 2002). Moreover, it was suggested that the policy might harm rather than help SMEs. Some banks suggested that price controls might result in the banks raising the price of loans in an attempt to recoup losses by increasing charges elsewhere, while Barclays argued that it would seek to increase interest rates or fees charged on lending and, at the margin,

80 Two real-world signalling games refuse to lend altogether to less creditworthy SMEs (Competition Commission 2002, Ch. 2: 150). Pledge costs The issue of small business banking enjoyed a considerable amount of publicity, with major British newspapers covering the developments extensively. The angle and tone taken by the media was not very favourable toward the banks, indicating that public opinion was unsympathetic to their current pricing policies. From the publication of the Cruickshank report onward, the government committed itself explicitly to address the problems identified therein (HM Treasury 2004: 68–9). In his 2000 Budget speech, Brown announced ‘major reforms … to open up competition in banking’ (HM Treasury 2000). On occasion of the publication of the Competition Commission’s report on 14 March 2002, Brown stated in a press release that his ‘goal is to create a fully competitive environment where new entrants can compete with existing banks on a fair basis and can offer more competitive banking services for small businesses’ (Department of Trade and Industry 2002). On the same day, the secretary of state for trade and industry, Patricia Hewitt, released a press statement that further committed the government to taking action on the issue of small business banking: We cannot allow small businesses to continue paying a higher price for banking services than is right. SMEs are a vital part of our economy, employing over half the workforce. … The recommendations of the Competition Commission, which the Government has accepted in full, will bring competition into the market and fundamentally change the way it works … SMEs should find it easier to switch banks, charges should be lower and more transparent and SMEs should have greater choice both within and between banks. Department of Trade and Industry 2002 According to a senior manager at HSBC who worked on the bank’s strategy concerning the regulation, ‘some banks felt that the process arguably was unnecessarily influenced by certain political goals, supporting a desire to be popularist with voters, as opposed to being a truly independent exercise’ (interview with Martin Lord). By the same logic, the manager suggested, the Treasury had a strong interest in keeping the drive for regulation up, as any retreat from its current stance would be seen as a ‘climb-down’ and an ‘embarrassment’ in front of the public. By contrast, making good on the promise to improve the fate of millions of small and micro-businesses across Britain would earn Brown and the Labour government immense credit among the electorate. Apart from the positive image conferred with the posture of standing up to the big banks, the proprietors and employees of SMEs themselves constitute a large constituency. Of Britain’s 3.5 million SMEs, 90 per cent are very small; that is, they have an annual turnover of less than £1 million and typically employ less than a handful of staff. Archetypes referred to in the NFO’s 2002 Small Business Banking Survey include small property agents,

Two real-world signalling games 81 family-run restaurants, freelance consultants and bicycle shop (quoted in Bank of Ireland et al. 2003: 94–7). Thus, the Labour government could expect considerable reputation costs if they abandoned the policy. Negative inducement effects Despite the negative publicity, the banks’ cause was not lost all along. After hearings in May and June 2002, the Treasury Select Committee echoed the banks’ view that the policy ‘could damage rather than improve SME financing, and could well deter entry of the smaller banks into SME lending’ (House of Commons Treasury Committee 2002: 22). However, the Competition Commission did its own calculations based on primary data about the banks’ capital base, their allocation of capital to services to SMEs in relation to their operating profits and the cost of equity. It found that the returns generated by the banks’ SME business were at least seven percentage points above the average cost of equity of 15 per cent (Competition Commission 2002: 4). These findings decisively contradicted the banks’ claims of negative inducement effects. According to the Commission’s report, The remedy … in no way adversely affects the terms on which banks lend to their customers. Hence, we see no justification for the clearing banks, in response, to increase money transmission charges or interest rates on loans, or reduce lending to SMEs as some clearing banks said they would do. We recognize the risk that the clearing banks will seek to negate the effect of paying interest on SME current accounts by increasing money transmission charges, but we are conscious that regulating money transmission charges – the obvious response to this risk – would represent a substantial burden. Competition Commission 2002: 6 Thus, facing considerable pledge costs combined with rather small expected policy costs, the government had sufficient grounds to call the bluff and show itself determined not to buy the banks’ predictions of negative inducement effects. Any possible negative implications of the policy will not be dramatic. Moreover, the government could expect low costs from adverse policy effects, should they arise at all, because any such effects could be blamed on the banks, who in the public eye had been cast in an unfavourable light all along. Therefore, negative inducement effects were small, while pledge costs would appear to have been considerable. Material lobbying costs How much effort and money did the banks spend on fighting the policy? At the height of the controversy in June 2002 the policy chairman at the Federation of Small Businesses (FSB) claimed, ‘[t]he big four banks have huge resources at their

82 Two real-world signalling games disposal and we are disappointed that the resources have been directed towards opposing the findings of the Competition Commission Report, rather than carrying out its recommendations’ (ePolitix, 14 June 2002). The FSB feared that the banks’ combined lobbying power may bring ‘the whole process off the rails’ (ibid.). Retrospectively, however, it appears that this potential power was not actually mobilized at the time. The senior manager at HSBC quoted above, for example, estimated that the actual lobbying costs expended by HSBC amounted to less than £20,000 during the 18-month period of the Competition Commission’s investigation (interview with Martin Lord). This sum breaks down primarily into management time and work afforded for preparing media representations and press interviews. Lord characterized the banks’ lobbying effort as ‘overall muted’ and a ‘mere sideshow’ and expressed the view that the Competition Commission would have been rather unresponsive to any lobbying, given the ‘highly academic’, ‘legal’ and ‘technical nature’ of its investigation. In his opinion, the Competition Commission ‘immunized itself against lobbying’ by circumscribing its investigation with a high level of confidentiality which it enjoyed due to its ‘significant legal powers’. According to the HSBC manager, the bank concluded at an early stage that any further lobbying would ultimately not be worth the effort: ‘Did we really think we would influence the outcome? Not really’ (interview with Martin Lord). None of the other three banks went any further in their efforts to affect the process. Reputation costs According to the literature on financial service regulation in Britain, an unwritten contract exists between successive governments and the banks, designed to maintain public confidence in the banking system (Grant 1993: 69–70; Le Fanu 2004). This relationship of trust has shaped communication and government–industry interaction within the ‘City–Treasury–Bank nexus’ (Grant 1993: 69) since the Department of Economic Affairs’ effective incorporation into the Exchequer in 1947. While it may have been somewhat weakened by the Cruickshank report, this relationship would not easily be sacrificed even for the short-term gains of about £100 million for each of the four leading banks. The HSBC manager believes that more avid lobbying against the regulation would have been futile as well as having the potential to damage the relationship between the banks and the Treasury. He emphasizes the government’s obstinacy, pointing out that HSBC would have been prepared to jeopardize the unwritten contract if it had felt that would have helped its case. This, however, was not perceived to be the case: ‘But we didn’t think that we could influence the outcome by lobbying. … In any case, the so-called contract has in recent years disappeared, largely as a result of the Cruickshank Report plus the regulatory actions taken against the banks’ (Interview with Martin Lord). Thus, relative to their low material lobbying costs, the banks’ reputation costs implied in jeopardizing the time-honoured and carefully nurtured relationship of trust with the Treasury were considerable. Given these relations, a separating

Two real-world signalling games 83 equilibrium like that in Figure 3.2 can be expected, in which the policymaker is resolute and presses ahead with the regulation, while business is unwilling to bear the costs of lobbying and separates. Outcome The banks reacted to the chancellor’s plans by arguing that the regulation might end up harming rather than helping SMEs, suggesting that the measures could undermine potential competitors’ attempts to break into the market and result in price increases for loans to SMEs. This view has been partially reflected in the conclusions of a Treasury Select Committee hearing on small business banking, which stated: The Committee was not convinced that price controls were the appropriate solution, and shares the misgivings of the Bank of England about their effectiveness in securing competition. However, we accept that in the short term the reduced charges will have a beneficial effect on SMEs. We are concerned that crude controls of this kind could damage rather than improve SME financing, and could well deter entry of the smaller banks into SME lending. The Committee is also concerned that regulation of charges could hinder competition, by reducing or eliminating a differentiating feature of new entrants’ products. House of Commons Treasury Committee 2002: 22 By stressing these points, the banking lobby managed to slow down the regulatory process and realize excess profits while they lasted. But the Treasury Committee’s view turned out to be inconsequential for the government’s actions. As one MP serving on the committee at the time said, ‘the Treasury will listen with interest, but not necessarily more than that’ (personal communication with Nick Palmer, MP). Around the time of the Competition Commission investigation, the BBA launched various press notices and pamphlets, pointing out that providers of financial services were already subject to heavy regulation. The BBA’s chief executive, Ian Mullen, argued that any additional ‘crippling regulation’ may ultimately lead to a shift of focus of foreign and domestic investment from UK-based financial services to other shores (Mullen 2001). But with the exception of this rhetoric, the banks ultimately abstained from engaging in a more costly political campaign that would underscore the severity and seriousness of the policy’s asserted negative inducement effects. Furthermore, only scant concertation of strategy took place among the banks. Ultimately they backed down rather quietly. The banks accepted a binding agreement with the government on 19 July 2002, under which they can choose between providing free money transmission services to their SME customers and paying interest on current accounts. In a further agreement reached on 31 October 2002, they committed to undertake measures to ease account switching and increase transparency of charges (HM Treasury 2002). Compliance with the announced undertakings is monitored

84 Two real-world signalling games by the Office of Fair Trading (2003), which has since sent progress reports to the Treasury and the Department of Trade and Industry on developments in account switching processes and access to branches for SME customers.

Phasing out PVC from public construction in Hesse In September 1992, the governing coalition of Greens and Social Democrats (SPD) in the German state of Hesse enacted a technical directive according to which future social housing projects should be free of polyvinyl chloride (PVC). The directive provided for an immediate ban of some PVC elements such as water and sewer pipes, and floor and wall coverings from public contract awards (Hessischer Landtag 1993a). A transition clause stipulated the phaseout of other products, including PVC-based door and window frames, from public calls for tenders beginning on 1 January 1995, thereby giving the construction and supply industries 18 months to adjust to the new requirements. By May 1993, the PVC-manufacturing and processing industry demanded that the regulation be dropped, arguing that it would lead to considerable disinvestments and redundancies (Die Tageszeitung, 23 July 1993). The government showed itself surprised by the industry’s concerns, as the directive was well-researched and backed up by a substantial body of expertise emphasizing the ease of PVC’s substitution with alternative materials and the non-negative employment effects of that substitution. Pledge costs From its inauguration, the coalition has developed a strong profile on environmental policy, with a particular emphasis on principles of precautionary and sustainable substance policy in public contract awards. As one might expect, the issue was particularly important for the Green coalition partner, and somewhat less so for the Social Democrats. Nonetheless, the phaseout of PVC from various applications constituted a core element of the government’s environmental policy strategy (Die Tageszeitung, 23 July 1993). Moreover, the minister in charge, Jörg Jordan, was himself a Social Democrat. The policy catered for a substantial and increasing demand among the electorate for the protection of the environment and the containing of man-made health and safety hazards. Throughout the 1980s, chlorine has been the one material most frequently involved in hazardous incidents. Sixty per cent of personal injuries related to these incidents are directly attributable to the release of chlorine (Darimont 1995: 285). PVC is also responsible for the release of dioxins and other toxic substances and smoke fumes in fire accidents. It increases the technical and financial burden on waste removal and a plethora of complex additives render many PVCs unsuitable for recycling. These hazards have gained high levels of public awareness so that PVC featured centrally among the substances causing concern and attracting negative publicity by the early 1990s. Overall, the coalition had incentives to deliver some of its promises relating to the environment. The phaseout of PVC from public construction was

Two real-world signalling games 85 considered an obvious and uncontentious starting point. The directive was intended by the government as a first step to implement a more general motion passed unanimously by Hesse’s parliament in May 1990 (during the predecessor government’s term) according to which all procurement and public calls for tender were to exclude the use of PVC in their terms. But while the issue was particularly important for the Green coalition partner it was less central to the Social Democrats’ ideological profile, who were the bigger and more powerful party in the coalition. This view is confirmed by policymakers from both sides of the debate at the time (personal communication with Karin Hagemann; interview with Hans-Jürgen Hielscher). Negative inducement effects PVC is the quantitatively single most important mass plastic. Its production and processing accounts for a large portion of chemical manufacturing in Germany, constituting a segment that has risen continuously since the 1960s (Umweltbundesamt 2001). In 1992, approximately 1.5 million tons of PVC were produced (Nolte and Joas 1995: 67). With almost half of the total use of PVC, the construction industry in turn constitutes the single largest end-user (Umweltbundesamt 2001). However, a comprehensive study by the Swiss consulting firm Prognos AG found that 95 per cent of all PVC uses in construction could be substituted using available alternatives (Prognos AG 1994: II–256–63). Prognos found further that, while some alternatives were more expensive, a large percentage of PVC uses could be substituted cost-effectively; that is, at no or very little net increase in costs, and with a potential increase in employment of 4 per cent. Windows and cables are part of the economically more costly segment, with substitution leading to cost increases of 36 per cent (ibid.). The higher costs for windows are due to the fact that wood is the main suitable alternative material, and cost calculations for window frame substitution include the expenses for more intensive maintenance efforts for wooden frames. Ackerman and Massey (2003) claim that cables can be manufactured cost-efficiently from a range of inputs. At any rate, all remaining construction inputs (floorings, pipes etc.) are suitable for cost-efficient substitution. Around 320 firms in Hesse were involved in producing, processing and installing the products that were supposed to be free of PVC in the future if the directive was implemented. These firms would have been affected in one way or another by the directive. They include five large and fifteen medium-sized enterprises, as well as around 300 small firms. As PVC processing formed only a minor part of the activities of the larger firms, these would have been affected only marginally (Jacob 1999: 38). The smaller firms, such as building contractors and material purveyors, faced compliance costs arising from the need to switch to alternative materials (Hanauer Anzeiger, 21 June 1993). According to the industry, the directive would have wiped out up to 10,000 jobs in firms that manufacture PVC-based building material or use it in construction (Wiesbadener Tagblatt, 15 June 1993). While in all likelihood the proposed regulation would

86 Two real-world signalling games not have had the severe negative impact claimed by the industry, the government showed itself disconcerted by those predictions. The threat of negative inducement effects was further exacerbated by an overall unfavourable economic situation: in May 1993, unemployment was at 167,000 – a 28 per cent increase on the same month of the previous year. Thus, from the perspective of the government at the time, potential negative inducement effects must have appeared considerable. Policy costs to business What costs would the policy entail for business if the government passed the proposal? Vinyl sidings and windows have been the fastest-growing uses of PVC (and the largest after pipes), experiencing double-digit annual growth rates throughout the 1980s (Ackerman and Massey 2003: 33). Industry representatives feared that a ban of these materials from public contracts in one state would set a precedent, opening the floodgates for further regulatory inroads into PVC use in other German states and, ultimately, at the federal level (personal communication with Günter Dunschen). The sector had already been under pressure since the late 1980s, when the total volume of chlorine going into plastics and other chlorinated products increased only marginally. A standstill, not to mention a fall, in the demand for chlorine had the potential of creating long-term cost implications for the entire chemical industry. The German-based (but globally operating) firms Bayer AG, BASF Group and the former Hoechst AG (subsequently broken up into Doorman and Aventis), for instance, rely heavily on the chlorinated polymer industry to absorb the millions of tons per annum of hydrochloric acid, a by-product of many of their core production processes. In the early 1990s, approximately 30 per cent of German primary chlorine production fed straight into PVC (Nolte and Joas 1995: 17–38). At the mercy of its end-user markets, the German chlorine industry saw its fortunes erode during the early 1990s and feared immense additional costs to the entire appendage of related industries in the wider chemical sector. Due to concurrent developments such as the recession of the early 1990s, the ban on chlorofluorocarbons, the declining demand for chlorinated solvents and the partial replacement of domestically produced PVCs by cheaper imports, annual primary chlorine production in Germany decreased by around 15 per cent from 3.4 million tons in 1989 to 2.9 million tons in 1993 (Darimont 1995: 285–6). Around the same time, the European chlorine industry’s rationalization processes resulted in plant closures in Scandinavia, Italy and the UK (Darimont 1995), which sensitized Germany’s chemical industry to potentially unfriendly political developments. Thus, much more was at stake for Germany’s chlorine industry than the costs attached to the continuation or discontinuation of the use of PVC components in Hesse’s public housing projects. While the actual directive meant little harm to either Hesse’s construction industry or Germany’s chlorine industry, the latter in particular perceived the Hesse PVC phaseout as a first battle within a much larger struggle about the future standing of PVC in particular, and

Two real-world signalling games 87 chlorinated polymers more generally. Thus, policy costs for business would have been considerable. Material lobbying costs The industry’s campaign against the directive in May and June 1993 resembled allout warfare, deploying the full repertoire of pressure politics available to resolute lobbyists. According to industry representatives involved at the time, it is impossible to quantify the costs of this campaign (personal communication with Günter Dunschen). However, a leading protagonist of the opposition to the policy in the Hessian parliament at the time recalls being impressed with the industry’s campaign and agrees that the associated costs must have been ‘relatively high’ (interview with Hans-Jürgen Hielscher). In order to illustrate the cost dimension implied in the struggle, it should suffice here to point out that mass demonstrations by chemical workers in June 1993 in Wiesbaden, the state’s capital and other Hessian locations were very costly endeavours for the firms. The protests were organized by a special lobbying organization of the PVC producers and processors, the Arbeitsgemeinschaft PVC und Umwelt and financed by the individual firms. They took place on weekdays and involved several thousand protesters turning out on paid leave on at least three occasions. According to the authors’ own estimates in consultation with the Regional Office for Hesse and Thuringia of the IG Bergbau, Chemie und Energie (Mining, Chemical and Energy Industrial Union), this implies deadweight labour time losses of between E500,000 and E1 million. It is at this point that the interactive nature of the signalling game is most relevant. Even a lobbyist willing to invest in a costly political campaign would have still faced only moderate costs compared with the long-term costs of a policy for which this particular directive could have been a mere precedent. Thus, even the considerable lobbying costs are still less than the expected policy costs. Because of potentially large negative inducement effects, the policymaker would be receptive to the industry’s arguments. A type-2 business lobbyist can in turn exploit this sensitivity to his advantage and expect to get away with a costly bluff. As the risk of detection would have been low, a pooling equilibrium can be expected in which the lobbyist always claims to be severely affected and the government always backs down. Outcome A vigorous lobbying campaign by the industry and high levels of publicity attached to the issue convinced the government to reconsider the policy. Meanwhile, the opposition parties scheduled a special hearing before the parliament’s Committee on Building and Urban Planning in early June 1993 (Wiesbadener Kurier, 14 June 1993). Responding to the government’s initial persistence, the industry increased the pressure, and soon Hesse witnessed one of the most vigorous political campaigns in the state’s post-war history. The firms and employers’ organizations convinced the chemical workers trade union and the works committees in the individual firms that slashing the directive was vital for the future

88 Two real-world signalling games employment of their members (Die Tageszeitung, 23 July 1993). Union executives then declared that the PVC phaseout would have disastrous effects on the labour market (Arbeitsgemeinschaft PVC und Umwelt e.V.). With the workers on board, the scene was set for the centrepiece of the campaign – a series of mass rallies in Wiesbaden, whose peak, on June 14, 1993, saw a turnout in excess of 4,000 workers (6,000 according to industry sources) from the chemical and plastics industry demonstrating outside Hesse’s parliament and in Wiesbaden’s main square (Frankfurter Allgemeine Zeitung, 15 June 1993). The protests took place at daytime on weekdays, and the workers had been given paid leave by their employers to attend. A letter-writing campaign cluttered the desks of Green and Social Democratic MPs. Full-page advertisements in the local and regional press hammered home the benefits of PVC production to the region and the alleged dire consequences of PVC’s partial ban as a building material (Die Tageszeitung, 11 November 1993). Furthermore, the campaigners managed to take the building material purveyors, building contractors and other micro businesses on board, even though for them it would arguably make little difference which particular materials the window frames and floor tiles they sell and install are made of, as long as they continue to sell and fit them (Hanauer Anzeiger, 21 June 1993). The opposition parties in the Hessian parliament demanded a roll-call vote on a bill to slash the directive, which was intended to put the Social Democrat MPs under the pressure of increased scrutiny from their working class voters (Hessischer Landtag 1993b). As a result, rifts ensued in late June, first among factions within the SPD parliamentary party, and later between the Greens and the Social Democrats, with the Greens defending the directive for some time, whereas the Social Democrats had by late June become willing to scrap it (Wiesbadener Tagblatt, 30 June 1993). Eventually, the Greens prized the continuation of the governing coalition higher than the policy, and on 22 July 1993 a compromise bill was debated and passed in parliament (Hessischer Landtag 1993c; Hessischer Landtag 1993d). Looking back two months after the events, the then head of government Hans Eichel (SPD) commented, ‘the preservation of jobs had to have priority’ (Hersfelder Zeitung, 16 September 1993).

Discussion These case studies are consistent with the predictions of the informational– structural model of business power. They refer to two equilibrium outcomes, one in which business tells the truth about the state of the world by declining to make a substantial lobbying effort, and another in which the high policy costs for business generate incentives for lying by means of a costly campaign. Both outcomes are difficult to explain in the context of the major approaches to the study of business political power outlined in the previous chapter. In the UK banking case, for instance, a policy network approach would have suggested a victory for the banks, which have traditionally enjoyed a privileged existence in the City–Treasury– Bank policy community. A pluralist analysis of this case might have also led to expecting the banks to prevail, as they have traditionally enjoyed more clout over

Two real-world signalling games 89 British politics than other industries. Likewise, theories of the structural power of business would have led us to expect the banks to carry the day due to their structurally privileged position, in particular with respect to the implementation of the policy. Thus, the factors that are emphasized by policy network, pluralist and structuralist explanations of business political power make the case of British small business banking a ‘hard case’ in the sense that the values of many contextual variables are distinctly unfavourable to the expectations generated by the informational account of political power. The application of the signalling model developed in Ch. 3 explicitly allows for the possibility that, despite its structurally privileged position, business sometimes has little choice but to back down and bear the costs of policy for the sake of preserving its good reputation as an advisor to political actors. This was indeed the case for the big four British high street banks. Turning to the case of PVC phaseout, pluralist models of interest group politics might well have warranted a prediction of business failure due to the high levels of publicity surrounding the policy struggle (cf. Gormley 1983). Structural power theories, with their emphasis on objective structural constraints would have had particular difficulties accounting for the outcome of this episode. On the one hand, the lack of objective compelling structural constraints would suggest a defeat for business. On the other hand, the classic theory of business’ structural power does not allow for business to lose out against non-business interests. A network or policy community analysis may also have produced inconclusive results, with predictions being contingent on whether the traditional policy community involving the chemical industry is emphasized (Grant, Paterson and Whitston 1988), or the discontinuity of that community due to the presence of a Green party coalition partner. Because of its explicit treatment of asymmetric information, the informational model outlined in Ch. 3 explains why the Hesse government decided to withdraw the policy even though it would not have led to the negative inducement effects predicted by business. The leading Green party parliamentarian involved in the episode at the time, Karin Hagemann, maintains that any claims about job losses were simply incorrect (personal communication with Karin Hagemann). It should be acknowledged, however, that proponents of the industry side continue to emphasize the economic damage that would have been caused by the policy (interview with Hans-Jürgen Hielscher). At any rate, the high expected costs that might have occurred if the industry’s predictions were truthful would have been damaging to the government, and this is what leads the informational–structural model to predict victory for business. Against a backdrop of rising unemployment and faced with a business community that behaves convincingly as if the situation was critical, the Hessian government was willing to heed the industry’s warnings. It could be argued that the classic theory of business’ structural power too would have predicted the outcome of the PVC case by reference to the major policymaker’s beliefs that the expected economic end electoral costs of the policy were high. In this sense, the informational theory of business political power does not so much contradict structural dependence theories as it complements them by providing the cognitive and strategic link between ‘objective’ economic–structural facts on the one hand, and actual political decisions and strategic behaviour on

90 Two real-world signalling games the other. But while the negative inducement effects might have been large, the likelihood of them materializing was not. This in turn highlights another difficulty of applying stylized game-theoretical models to the inevitable messiness of the real world. Given that the pooling equilibrium implies that policymakers cannot learn anything new from observing costly lobbying, the question arises how the Hessian government could have been brought to change its views on the policy. To accommodate this problem it could be argued that this case reflects a semiseparating equilibrium better than a pure pooling equilibrium – recall from the previous chapter that the two equilibrium types are in close proximity to one another with respect to the values of the relevant parameter. But because of the random element in the semi-separating equilibrium it is as difficult to predict as it is to identify in historical case studies, and post hoc rationalizations are not very satisfactory. Similarly, the case of British small business banking illustrates that evidence in support of the informational theory of business political power does not itself discredit notions of the structural dependence of the state on capital. While Gordon Brown’s policy clearly violates the interests of one fraction of capital – the oligopolistic small business retail banking sector – it can also be viewed as favouring what might be perceived as the interests of capital overall by strengthening other banks and SMEs, and by promoting competition, which benefits business and the economy as a whole. The informational–structural analysis of the two case studies in this chapter illustrates the extent to which strategic and communicative processes are required to serve as the link between structural parameters and political action, thereby explaining outcomes that are far from obvious from the viewpoint of other theories of business and politics. In allowing for sources of relevant information independent from the lobbyist and substituting a plethora of micro-businesses for a constituency of ‘ordinary people’, the working class or other non-business interests, the case of small business banking highlights the importance of the information gap between business and policymakers as an explanatory variable. In order to obtain a reliable picture of the potential adverse effects of regulation, the government went to great lengths to assess all aspects of the policy when it decided to have the matter thoroughly investigated by the Competition Commission. This process unearthed valuable information for the Treasury, on the basis of which it could then update its expectations about negative inducement effects independently from information supplied and interpreted by the banks. Thus, an important lesson from the analysis of this case is the extent to which the Competition Commission’s privileges of full insights into the banks’ books reduced the information asymmetry between the banks and the policymaker. This suggests that information asymmetry is a variable factor, rather than a constant, even within a given policy area. In highlighting the importance of business’ informationally privileged position, the banking case also sheds new light on the well-established importance of cohesion and unity among business to prevailing in political struggles. Central to network approaches to the study of business political power, business unity and structural equivalence are seen as prerequisites for the formulation of

Two real-world signalling games 91 common preferences and the unidirectionality of the use of political resources (Mizruchi 1992). From the perspective of the informational model of business political power, however, business unity and structural equivalence reduce the risk of sending conflicting signals, thereby diminishing the scope for policymakers to obtain independent information. The high-quality information and assessment provided by Cruickshank, the Competition Commission and various City analysts reduced the information asymmetry in favour of the policymaker. It is ironic that this determination of policymakers who ‘know better’, a characteristic that has become firmly associated with successive Thatcher governments (Richardson 2000: 1009–10), has found such a vivid incarnation in the Labour Party’s take on competition in small business banking. The information gap between business and policymaker also turned out to be decisive for the outcome of the case of PVC phaseout in Hesse. The explicit role of information asymmetry and the way strategic actors deal with this constraint given the values of the model parameters enables us to understand why business won even though the policy would probably not have had negative inducement effects. The net cost effect of the substitution of PVC with alternative materials would amount to a 2 per cent increase in the per-unit costs of social housing and similar projects. That would have amounted to average E1,150 per unit and constitutes a percentage rate far lower than the usual fluctuations on bids and costs in the construction industry (Bündnis 90/Die Grünen 1997). It should further be noted that the directive only referred to PVC used in social housing construction, which at the time constituted a segment of less than 15 per cent of the total construction market in Hesse. Thus, production costs for 15 per cent of the market could have been expected to rise by 2 per cent as a result of the directive. This increase, as the social-democratic MP Sieghard Pawlik pointed out at the time, would not have altered any of the incumbent government’s plans for social housing and other construction contracts (Hessischer Landtag 1993b). Nonetheless, faced with predictions of negative inducement effects, remaining uncertainty about the likelihood of negative inducement effects, observing a lobbyist that behaved convincingly as if the situation was critical, and therefore unable to know better, the Hessian government was unwilling to take a risk and instead gave in to the industry’s demands. The PVC case illustrates a further fundamental methodological problem involved in the empirical analysis of models of asymmetric information. As in this case business ‘won’ the signalling game, any assessment of the real costs of regulation inevitably involves counterfactual reasoning. This problem is here exacerbated by fact that in environmental politics, for every study predicting negative impact there is another one claiming the opposite. Moreover, one and the same study can often be read to support either of two opposing interpretations. According to a senior technical consultant, this was the case with the Prognos study cited above (interview with Rainer Nolte). Therefore, it is necessary to assess the predictive power of the informational–structural account of business power against more systematic data. The following two chapters report quantitative tests of micro and macro-level observable implications of the informational–structural

92 Two real-world signalling games model of business power. The micro-level tests in the following chapter deal with implications relating to the costs of lobbying, in particular reputation costs. The macro-level analysis in the chapter thereafter tests the predictive capacities of the informational structural variables in the model – the size of the information gap between business and politics, and the magnitude of negative inducement effects.

5

Reputation and informativeness in lobbying

According to the arguments and evidence presented so far, business lobbyists base their decision to lobby on the expected costs and benefits of lobbying, including considerations about the likelihood of success. In practice, this involves balancing the expected costs in terms of campaign expenses and possible damage to reputation against the expected benefits of averting an unfavourable policy or attaining a desired one. The case study on small business banking in the previous chapter suggests that policymakers will be more likely to override business’ messages if, ceteris paribus, they have relatively little to lose from incorrectly dismissing as idle babble what might really be an informative signal. This condition would be given, for example, in the case of individual legislators nearing term limits or retirement, as well as for those whose incumbent government is enjoying a postelection honeymoon period during which the extent to which they attract blame for undesirable socio-political developments is generally limited (MacKuen, Erikson and Stimson 1992). Policymakers also have reduced incentives to heed business’ advice if they have invested heavily in their reputation with regard to a particular issue, be that advancing the fate of the working class, fighting for a cleaner environment or whatever other issue happens to feature centrally in their election platform. Lobbyists, on the other hand, should be expected to place a bigger weight on their reputation in circumstances that have been characterized by close, well-nurtured relationships with policymakers. Good and trustful working relationships between firms and policymakers, between regulated and regulator, are the outcome of a long process of give and take, during which the provision of good information will have played an important role in the build-up of trust. But while it takes a long time to develop such a relationship, it can be lost in much shorter time (Coen and Grant 2006: 21). At the same time, the extent to which a business lobbyist can expect to get away with exaggerated claims of negative inducement effects is contingent on the political situation of the policymaker. If the policymaker enjoys high levels of public support and if her other policies have largely been perceived as successful, she will be in a position to act tough. As the model in Ch. 3 and the case study on PVC phaseout imply, making bold predictions about policy effects implies making occasionally untruthful assertions. Business actors will then be less inclined to make such bold assertions if the policymaker is likely to override them.

94 Reputation and informativeness in lobbying By contrast, if the policymaker’s political situation is precarious, business lobbyists can expect policymakers to be more responsive to their demands, in particular if these are expressed with some vigour. While these expectations are difficult to evaluate empirically, some microlevel implications of the informational–structural model of business power can be assessed quantitatively, using information that is already available to researchers. In the following sections, a number of hypotheses concerning the reputational costs of lobbying will be developed and tested, using individual-level data on the behaviour and characteristics of members of national parliaments in eleven European countries. This is followed by an analysis of the relationship between business contacts and influential impact of lobbyists on policymaking. The concluding section of the chapter discusses the empirical findings in the light of past and present research on business, politics and lobbying. According to the informational–structural model of lobbying, business lobbyists have incentives to make the negative repercussions of public policy at times sound more dramatic than they actually are. At the same time, business faces constraints on its capacity to promulgate negative inducement effects of policy untruthfully. These constraints arise from business’ vital, long-term interest in maintaining its reputation as a supplier of high-quality information (Berry 1997: 98–100). Policymakers, on their part, are suspicious of the possibility of misrepresentations by business. If, in a given policymaking and lobbying scenario, it turns out that the business lobbyist untruthfully warned the policymaker of negative inducement effects, business may end up damaging its relationship of trust with the policymaker. The policymaker will be disappointed and, according to the model, may react by inflicting an exogenous penalty for lying on business. This penalty can be interpreted as the behavioural consequences of a damaged reputation, such as the denial of access to the policymaker for future representations. Assuming that misinformation can only be detected if the policymaker actually passes the policy, by heeding business’ advice and withdrawing a policy the policymaker forgoes her chance to find out what would have really happened if she had pushed through with her plans. Therefore, when deciding whether or not to lobby, business will be careful to avoid situations in which untruthful predictions are likely to be exposed. By thus considering the potential reputational costs of lying in addition to the material costs of lobbying, the model outlined in Ch. 3 expands on previous signalling models in the literature. According to the nineteenth-century German caricaturist Wilhelm Busch (1975 [1865]), once a person’s reputation is damaged, he is free to say and do whatever he likes, because as far as his reputation is concerned things cannot get any worse from hereon. Similarly, if someone never had a particularly good standing with another person and if, regardless of his actions, prospects are poor for him to gain a reputable position in the eyes of the other person, he is fairly free to make immoderate claims without risking noteworthy damage to his reputation – for one cannot lose friends that one does not have. Applying this logic to the world of lobbying, the price to be paid for bold lobbying will vary with the quality and nature of the relationship between business and policymakers. Intuitively, where

Reputation and informativeness in lobbying 95 there is only a weak relationship of trust in the first place, business should feel less restrained from making bold claims. This implies, in turn, that business lobbyists can be more daring, and thus to some extent discount reputation costs, if and when they have a relatively poor relationship with policymakers. Conversely, lobbyists should feel obliged to restrict their messages to true claims more effectively when their relationship with policymakers has traditionally been characterized by mutual respect, trust and perhaps even a genuine interest in the views of the other party. Of course, even businesses and industries in rather precarious relationships with public officials must have some minimum level of credibility – otherwise their messages would be consistently ignored as uninformative babble. In turn, they will not find it worth their while to expend any money, time or effort on lobbying, resulting in an equilibrium of silence. Do different businesses vary with respect to their trustworthiness? One might think here of differences between, on the one hand, firms that feel some kind of obligation towards the public good, and on the other hand, ‘bad apples’ that have a tendency to cry wolf on every occasion to advance their interests. But the imperative to further the good of the firm in the light of unforgiving market forces and considerable risk and uncertainty will generally set rigid limits on the ability of individual owners and managers to display altruistic virtues with respect to any public policy that affects their firm’s operations and profits. Alternatively, it might be that some commercial activities are easier understood by policymakers than others, thus suggesting that the informational gap between regulator and regulated varies across different sectors. But few systematic characteristics appear plausible at the firm level. As there are at least two parties to every relationship, more meaningful differences in government–business relationships might surface by looking at the type of policymaker. It is widely accepted, and conforms to much common-sense knowledge, that policymakers vary systematically in the relationships they are able and willing to maintain with business (Grant 1993: 125–45). A typical source of such variation is of course party membership or partisan attachment. According to Wilson, ‘[in] most, if not all democratic nations, one political party is more closely associated with business than are others. … The Republican Party in the United States, the Liberal Democratic Party in Japan and the Conservative Party in Britain are all examples of political parties closely associated with business’ (Wilson 1990: 24). Just like electoral class cleavages have undergone a certain erosion in recent years, left–right patterns in the relationships between interest groups and political parties too have changed. As Jordan and Maloney (2001: 27) point out, while for much of the twentieth century the Labour Party had close ties with trade unions and also with left-of-centre organizations such as antinuclear groups, the Conservative Party was close to business as represented by the CBI. Since the mid 1990s, major changes have occurred in this pattern of party–group relations. The ‘New’ Labour Party under Tony Blair loosened its links with the trade union movement and increasingly embraced business, prompting in turn corresponding changes in the Conservative Party’s relations with interest groups. But despite these changes, established ties and relationships do not disappear quickly. Organizational ties in

96 Reputation and informativeness in lobbying particular tend to be of an enduring nature. It is still reasonable to assume then, that right-leaning and market-liberal politicians, parties and governments maintain closer relationships with the business community in their countries than their leftleaning colleagues. From the perspective of business, the reputational constraint is then stronger with respect to pro-business (i.e. right-leaning) policymakers, with whom business already enjoys a good relationship. This means that lobbyists interacting with politicians who are politically close will feel more compelled to telling the truth (Broscheid 2006: 99). Thus, we can expect left-leaning policymakers to be ‘lied to’ more frequently by business compared to their conservative counterparts. Similarly, members of business-friendly political parties should be expected to receive mainly sincere messages, while, say, environmentalist policymakers such as members of Green parties, would be more likely to receive misrepresentations about policy costs. Damage to business reputation as a provider of good information can be incurred in two ways, each of which renders the likelihood of lobbying contingent on the relationship between business and the policymaker. First, if the business– policymaker relationship is characterized by a high degree of co-operation and trust, much is at stake for business if untruthful lobbying goes wrong. The point has already been made that this should be the case if the policymaker is of a businessfriendly party. Regardless of party affiliations or ideological proximity government business relations will also be more trust-based in neo-corporatist systems of interest intermediation. Second, while business lobbyists can sometimes expect to get away with lying, they will rarely be certain as to whether a policymaker will heed the signal or call a bluff. However, they can expect the risk to be higher when the policymaker is of a left-leaning party background. Thus, on the one hand, business will remain silent more often when it enjoys a good standing with the policymaker; that is, when the policymaker is of the pro-business or non-left kind. On the other hand, the risk of detection itself is partly a function of the policymaker’s preferences. While left-leaning or environmentalist policymakers are not oblivious of business concerns, they may have stronger commitments and preferences that they balance against business’ lobbying. Thus, business lobbyists always have to take into account the extent to which existing government policy commitments may limit the success of their strategy (Coen and Grant 2006:22). The true amount of dishonest signals cannot of course be known or is at least extremely difficult to come by. But if untruthful signals are unequally distributed between left- and right-leaning recipients, this distribution would to some extent be reflected in the relative distribution of lobbying by the type of policymaker – whether truthful or not. Thus, while no data is available for measuring the relative frequencies of true versus untrue lobbying messages, individual-level measures of policymakers’ interaction with various types of lobbyists – business, trade unions and public interest groups – can be used to approximate the relative occurrence of untruthful lobbying. Recall that business is often itself not fully certain about the likelihood of negative inducement effects, but tends to deal with the remaining uncertainty in a conservative way. That is to say, businesses themselves will often

Reputation and informativeness in lobbying 97 not know with certainty whether their lobbying messages are strictly speaking true or untrue. However, they will be more careful with respect to spreading potentially untruthful signals when doing so might harm their reputation. Put differently, business lobbyists will scrutinize the truth content of their messages more carefully before sending a signal when the recipient is someone with who they have a good working relationship or when the risk of detection is high, which leads to an overall reduction of lobbying ‘noise’. Thus, while there will not be a perfect correlation between frequency of lobbying and frequency of lying (if there was, policymakers would be able to make precise inferences about everything that is known to business merely by observing its lobbying efforts!), the frequency of lying can be assumed to increase with the frequency of lobbying, albeit weighted by a factor unknown to us and to the policymaker. Therefore, frequency of lobbying serves as a useful approximation for ‘lying’.

Data: The 1996 European Members of Parliament Study The data used in this chapter are taken from the 1996 European Members of Parliament Study (Katz and Wessels 1999). They have been generated through a mail survey among members of the national parliaments of Belgium, France, Germany, Greece, Ireland, Italy, Luxemburg, the Netherlands, Portugal, Spain and Sweden. A total of 1,412 MPs returned the questionnaires. Split up by countries, response numbers range from 28 for Luxembourg to 317 for Germany, and return rates vary between 14.9 per cent in Italy and 90.3 per cent in Sweden. While these rather big differences in return rates are clearly not desirable, their impact will be minimal as the estimates of interest obtained in the analysis that follows concern relations between variables rather than univariate distributions. Moreover, weights are available by which the estimates will be standardized with respect to the subpopulations; that is, the national parliaments. The variable codings and survey items used are introduced as they become relevant in the analyses that follow. A detailed description of the data, survey questionnaire items used, and coding of variables is provided in Appendix B. In European polities, civil servants and ministers are generally more influential in shaping policymaking than parliamentarians. But MPs are by no means unimportant for business lobbyists. Even in the UK parliament, with its high degree of executive dominance, private members’ bills get heard, read and sometimes become enacted. In a survey by Rush (1990), 75 per cent of organizations, including firms, maintained regular or frequent contacts with one or more British MP. Even allowing for the fact that lobbyists harbour illusions about the importance of MPs and end up wasting their time and money, this figure should be much lower if MPs were irrelevant to the policy process. In most of the consensus systems of continental Europe represented in the dataset, the influence of parliamentarians, as measured for example by MPs’ impact on government bills, is usually higher than in the UK. Thus, most of the parliaments whose members have been surveyed by Wessels et al. are what Norton (1993: 50) calls ‘policy-influencing legislatures’. Crucially, the observation that parliamentarians are at least somewhat

98 Reputation and informativeness in lobbying influential is shared by organized interests. For example, Wilson (1987) found that French interest groups frequently lobby parliamentarians, while LaPalombara (1964) documents the same for Italy. MPs’ influence on policymaking – and thereby their attractiveness to lobbyists – can take a variety of routes. First, due to their position in the political system, varied memberships, osmotic legislative–executive boundaries and privileged access to information, parliamentarians are uniquely suited to link organized interests to other agents in the political-administrative system (Saalfeld 2000: 354). For example, 50 per cent of the interest groups in Rush’s aforementioned survey had asked MPs to arrange meetings with a minister. Thus, MP contacts are at least useful approximations for contacts with executive branch officials. Second, MPs have incentives to hold governments accountable in what Saalfeld describes as ‘a continuous election campaign conducted by all parties in the parliamentary arena’ (ibid.). And third, even opposition MPs can exert influence on government policy via parliamentary committees or extra-parliamentary bodies, such as neo-corporatist bodies, inter-ministerial committees of national and regional governments in federal states or other actors in the institutional environment (ibid.: 367). Thus, while parliamentarians may not be the most important political players, they occupy influential positions in the policy process. As a result, they are the target of much political activity by special interest groups such as business. Indeed, 34 per cent of MPs in the dataset used here report frequent contacts with business interests. Again, this figure should be much lower if business lobbying was restricted to ministers or MPs with special official duties, such as government whips. A first glance at the data suggests that, irrespective of the type of policymaker, business lobbies more frequently than public interest groups (e.g. environmental or consumer groups), but slightly less frequently than trade unions. The relative frequencies of MPs’ contacts with different societal interests vary between 24 per cent of MPs having frequent contact with public interest groups and 36 per cent reporting frequent contact with trade unions. Frequent business contacts are close to trade union contacts, with 34 per cent reporting these contacts. Frequent contact is defined here as an MP’s contact with the respective interest group ‘at least once a month’. It is possible that serious lobbying is only really tapped by more intensive interaction than by what can be achieved by monthly contacting. Therefore, distributions of interest group contacts ‘at least once a week’ were also inspected. Quite similarly to frequent contacting, these very frequent contacts are lowest for public interest groups (3 per cent) and highest for trade unions (8 per cent), with business once again in the upper chunk at 7 per cent. Thus, it is reasonable to assume that both monthly and weekly group contacts measure the same phenomenon. To check for the possibility that this assumption might not hold, separate analyses for frequent and very frequent lobby contacts were conducted. As these led to essentially identical results, only the result for frequent lobby contacts will be reported in the following sections. Assuming that policymakers’ ideological orientations or partisan attachments serve as reasonable proxies for the quality of the relationship between business

Reputation and informativeness in lobbying 99 and political decisionmakers, to what extent are patterns of business lobbying contingent on these orientations and attachments? To answer this question, relative frequencies of MPs that reported frequent contact with business representatives are examined using three alternative indicators of MPs’ ideological orientations and attachments: (1) whether they belong to a political party that is traditionally seen as an ally of business; (2) their self-placement on an ideological left–right scale; and (3) whether they view the environment as an important political issue. The relative frequency of lobbying should be higher for left-of-centre or environmentalist policymakers for two reasons. First, business is particularly keen to avoid sending untruthful messages to pro-business policymakers for reasons related to its reputation and good standing with a natural political ally. If, as has been argued above, total untruthful lobbying increases with the frequency of lobbying, business has incentives to restrain overall lobbying of business-friendly politicians. Second, because of their ideological outlook, pro-business policymakers are less likely in the first place to generate policy proposals that run counter to business interests or fail to correctly understand what observing business interest involves for specific policy areas. Pro-business policymakers thereby produce fewer occasions for which lobbying may be necessary in the first place. For these reasons, business is expected to lobby left-leaning and environmentalist parliamentarians more frequently than their more pro-business counterparts. Conversely, the informational model of business political power will receive support if MPs that are assumed to have close relations with business – that is, those of a right-of-centre political orientation or with little concern for the environment – are lobbied less frequently. However, the structural nature of the information asymmetry between business and policymakers implies that pro-business legislators will also be lobbied by business, if simply to become better informed about the details of the pro-business policies they seek to enact. Recall, moreover, that definitive proof of business’ exaggeration or misinterpretation concerning a policy’s negative consequences is only forthcoming if the policy is actually implemented. Because of politicians’ incentives to redeem policy pledges, left-leaning policymakers are more likely to push ahead with their policy irrespective of business’ representations. As a result, untrue claims are more likely to be exposed when made to ‘anti-business’ parties as opposed to ‘pro-business’ parties, who face lower pledge costs associated with pro-business policies and are therefore more likely to heed business advice. This, in turn, implies the exact opposite of the previous consideration: Faced with a greater likelihood of exposure, and assuming that misrepresentations increase with the overall frequency of lobbying, business will lobby left-leaning parties less frequently. In the absence of compelling theoretical reasons for favouring one of these two conflicting expectations over the other, the partisan and ideological identity of the policymaker is expected to make a difference of either direction for the frequency of business lobbying. Members of opposition parties might not be expected to be lobbied to the same extent or in the same way as members of government parties. Intuitively, we would expect government MPs to be lobbied more often than

100 Reputation and informativeness in lobbying opposition MPs, simply because the former are usually more important when it comes to policymaking. While, there are competing theoretical claims and conflicting evidence in the literature as to exactly when and why lobbyists should address incumbent policymakers or opposition MPs, most authors focus on the institutional and procedural set-up of legislative bodies (Krehbiel 1999; Grossman and Helpman 2001: 283–317). For the purposes of the present analysis it suffices to note that being a member of a government party will not be irrelevant to the frequency of lobbying. To control for this factor, separate estimates will be obtained for government and opposition MPs. Members are coded as ‘government MP’ if they belong to a party that participated in government at the time the surveys were conducted. For details of the coding schemes see Appendix B.

Lobbying, reputation and ideological proximity Table 5.1 displays relative frequencies for frequent contacts with various types of lobbyists – business, trade union and public interest – and grouped by government incumbency. The expectation that business-friendly MPs are lobbied less frequently by and large fails to be born out. Members of business-friendly parties are lobbied at least as frequently as other MPs, and often more frequently. This holds for government MPs as much as for opposition members. None of these differences are statistically significant.1 Looking at left–right ideological self-placement, fewer MPs who classify themselves as left-wing report to be frequently lobbied by business compared with their right-wing peers (Table 5.2). This finding holds regardless of whether they belong to the government or opposition parties. Overall, the differences are quite small and statistically insignificant. Similarly, as can be seen from Table 5.3, there are no significant differences in the lobbying patterns between environmentalist MPs and their non-environmentalist peers. Variation in the reputational variable and its relationship with lobbying behaviour can also arise from the institutional setting within which the relationships between business and parliamentarians are situated. The frequency and nature of business contacts for both government and opposition parties of all ideological types may be different depending on whether they take place in the context of a pluralist or a neo-corporatist system of interest representation and intermediation. In neo-corporatist systems, business is routinely consulted during the formulation Table 5.1 Business contact and party type

Government (140) Opposition (131)

Anti-business

Pro-business

F

30 33

44 40

3.12 0.94

Source: 1996 European Members of Parliament Study. Note: Cell entries are percentages of MPs who reported to have contact with business interest groups at least once a month. Weighted row counts in parentheses. Design-based F-test: *significant at 10%, **significant at 5%, ***significant at 1% (two-tailed).

Reputation and informativeness in lobbying 101 Table 5.2 Business contact and left–right self-placement

Government (140) Opposition (131)

Left

Right/centre

F

32 34

42 39

2.62 1.32

Source: 1996 European Members of Parliament Study. Note: Cell entries are percentages of MPs who reported to have contact with business interest groups at least once a month. Weighted row counts in parentheses. Design-based F-test: *significant at 10%, **significant at 5%, ***significant at 1% (two-tailed).

Table 5.3 Business contact and environmental platform

Government (140) Opposition (131)

Environmentalist

Non-environmentalist

F

37 37

37 35

0.00 0.09

Source: 1996 European Members of Parliament Study. Note: Cell entries are percentages of MPs who reported to have contact with business interest groups at least once a month. Weighted row counts in parentheses. Design-based F-test: *significant at 10%, **significant at 5%, ***significant at 1% (two-tailed).

and implementation of public policy. A core, if ideal-typical, feature of corporatist systems is the ‘permanent institutionalization of access’ (Schmitter 1974: 101). In the words of Traxler (1999: 56), corporatism ‘means that the state shares its public-order function with organized business and labour’. For public policymakers and business to enter into a partnership of this kind presupposes a certain level of mutual trust. What is more, this relationship is likely to further reinforce any existing trust between private and public actors. Therefore, the importance of maintaining a good reputation with policymakers should be higher where these integrated patterns of interest intermediation prevail. Whether MPs operate in a neo-corporatist or, alternatively, a more pluralist system of interest intermediation should therefore be irrelevant for the patterns of lobbying we observe. Numerous measures of corporatism or pluralism are now available (e.g. Lijphart and Crepaz 1991; Siaroff 1999) Of these, Siaroff’s scale of economic integration combines most aspects of previous attempts at measuring the pluralism– corporatism contrast. Averaging indices of the various dimensions over a five-point score, the scale solves most of the problems that have tainted previous measures of these concepts (Lijphart 1999: 176). Integration, in Siaroff’s understanding, speaks to the functional roles and behavioural patterns noted in the ideal types of neo-corporatism and pluralism outlined in Ch. 2. While Siaroff’s measure is encompassing in an institutional sense, it is explicitly not about other structural features, such as the level of unionization, or favourable contexts, such as the size or openness of a country, or the political role of social democracy (Siaroff 1999: 189). Because MPs’ ideological position and partisan attachment are measured as separate variables in the present analysis, Siaroff’s concept of integration is therefore particularly useful here.

102 Reputation and informativeness in lobbying To facilitate tabular analysis, neo-corporatist and pluralist dummy codes based on Siaroff’s measure for the degree of economic integration for the mid-1990s are assigned to MPs’ contexts. In the absence of a compelling substantive classification rule, the categories are assigned according to a country’s score relative
= 3.50; σX = 0.49). to the unweighted population mean of Siaroff’s scale (X Coding all those countries as neo-corporatist that are one-half standard deviations above the mean or higher yields a list of neo-corporatist countries that includes Germany, Luxemburg, the Netherlands and Sweden. Conversely, France, Greece, Ireland, Italy, Portugal and Spain scored lower than one-half standard deviations below the mean; they are hence coded as pluralist. Belgium is the sole country in the middle category of systems that are neither clearly neo-corporatist nor pluralist. The picture looks more interesting now that the type of interest group intermediation system is controlled for. As Table 5.4 reveals, government members (but not opposition members) in neo-corporatist systems are more likely to be in frequent contact with business if they are from pro-business parties. By contrast, there is no such relationship between MPs’ political identity and lobbying frequency in pluralist systems. Looking at left–right ideological self-placement, fewer government MPs who classify themselves as left-wing report to be frequently lobbied by business compared with their right-wing peers. This finding holds regardless of whether they belong to the government or opposition parties, but is statistically significant only in pluralist systems (Table 5.5). Finally, in pluralist systems, fewer environmentalist MPs report monthly business contacts compared with their non-environmentalist peers, while the opposite is true for neo-corporatist systems (Table 5.6). The percentage of MPs reporting frequent lobbying in pluralist systems is almost twice as big for non-environmentalists as it is for environmentalists. The exact opposite is the case in neo-corporatist systems. Here, over twice the proportion of environmentalist government MPs report at least monthly business lobbying (53 per cent), compared with other MPs (22 per cent). To sum up the findings of this section, the frequency of MPs’ contact with business lobbyists is significantly related to MPs’ ideological orientations, Table 5.4 Business contact, party type and system of interest intermediation Pluralism

Neo-corporatism

Anti-business Pro-business F

Anti-business

Government 41 (140) Opposition 36 (131)

Pro-business F

49

2.51 11

36

3.45*

45

1.42 30

25

0.06

Source: 1996 European Members of Parliament Study. Note: Cell entries are percentages of MPs who reported to have contact with business interest groups at least once a month. Weighted row counts in parentheses. Design-based F-test: *significant at 10%, **significant at 5%, ***significant at 1% (two-tailed).

Reputation and informativeness in lobbying 103 Table 5.5 Business contact, left–right self-placement and system of interest intermediation Pluralism

Government (139) Opposition (130)

Neo-corporatism

Left

Right/centre

F

Left

Right/centre

F

39

52

3.69*

17

30

2.03

36

44

2.75

30

27

0.08

Source: 1996 European Members of Parliament Study. Note: Cell entries are percentages of MPs who reported to have contact with business interest groups at least once a month. Weighted row counts in parentheses. Design-based F-test: *significant at 10%, **significant at 5%, ***significant at 1% (two-tailed).

Table 5.6 Business contact, environmental platform and system of interest intermediation Pluralism

Government (139) Opposition (130)

Neo-corporatism

Environ.

Non-environ.

F

Environ.

Non-environ.

F

25

46

3.53*

53

22

4.36*

36

42

0.45

34

27

0.15

Source: 1996 European Members of Parliament Study. Note: Cell entries are percentages of MPs who reported to have contact with business interest groups at least once a month. Weighted row counts in parentheses. Design-based F-test: *significant at 10%, **significant at 5%, ***significant at 1% (two-tailed).

the relationship of their party to business and their attitudes toward the environment. Reports of frequent contact are generally more likely to come from right-leaning, business-friendly and non-environmentalist MPs. But these results are highly contingent on the type of interest group intermediation system. Where differences in lobbying behaviour are related to partisan identity and ideological position, this is mostly the case in neo-corporatist systems. Environmentalist positions of MPs, however, have a significant negative effect on lobby contacts in pluralist systems, while environmentalist MPs in neo-corporatist system are significantly more likely to be contacted by business. These findings contradict the hypothesis according to which business-friendly, right-wing and non-environmentalist MPs should be targeted less frequently by business due to business’ incentives to maintain good relationship with traditional allies. The theoretical expectations with respect to MPs’ commitment to the environment are confirmed. This finding should alert students of interest group lobbying and environmental politics to the possibility that in neo-corporatist settings, parliamentarians with strong environmentalist platforms may be particularly susceptible to extreme levels of lobbying by business. Moreover, if, as these tests assume, frequency of contact is indicative of increased possibilities of misrepresentations,

104 Reputation and informativeness in lobbying environmentalist parliamentarians in neo-corporatist contexts are also more prone to being lied to. But overall, the contrary expectation that pro-business parties are lobbied more frequently than others because untruthful lobbying is less likely to be exposed fares better in these tests. Thus, the partisan identity and ideological position of MPs matters for the frequency with which they are contacted by business. The implication of the informational–structural model of business political power that the reputation of the lobbyist has a constraining effect on lobbying activity receives some support.

Business contacts and influential lobbying Policymakers listen to business lobbyists because the latter’s expertise and knowledge aids them in making informed policy decisions. According to the informational–structural account of business power, policymakers are interested in the information provided by business because, more so than other groups, business has privileged and first-hand insights into important parameters affecting the costs and effects of public policy. If this mechanism is to generate political power for business relative to non-business groups, the private information held and communicated by business actors must be deemed more valuable by policymakers than information provided by other groups – regardless how efficiently organized, vociferously active or well-connected the latter may be. This would provide business interests with disproportionate influence over public policy in addition to any possible advantages resulting from their structurally privileged position or their other power resources. Anecdotal evidence from political discourse suggests that the economic and social implications of the policy-relevant speech acts of business lobbyists carry more weight than those made by other organized interests. When, for example, Craig Barrett, the chief executive of Intel, dropped a remark to the Irish government about its tax policy, this was likely to have made a bigger impact on policy decisions than various grassroots campaigns against service charges, such as the ‘Anti Bin Tax Campaign’ staged by public interest groups aiming to scrap charges for household waste removal (Irish Times, 20 October 2003). The reasons for this are at the core of the structural power theories. Political actors do not shy away from predicting the disconcerting effects that any set of policies other than that condoned by business may have. Dwelling on the above example a little further, Intel, the world’s biggest semiconductor producer, informed the Irish government in 2004 that any change to Ireland’s cutthroat tax regime, which offers very attractive tax breaks and direct giveaways to foreign investors, would have negative inducement effects. In particular, Barrett warned that reduced generosity might induce Intel to reconsider investment plans concerning a E1.66 billion microchip fabrication plant in Leixlip, Ireland. This investment would bring with it the creation of up to 400 jobs within 18 months, not including the benefits of multiplier effects to the economy (Irish Times, 20 May 2004). Taking the size of Ireland’s national economy into account, it is not hard to see how an investment project of this dimension could make a big difference for the

Reputation and informativeness in lobbying 105 popularity of elected politicians. Warning that Ireland should be careful not to manoeuvre itself out of a competitive position, Barrett hinted at other countries and regions that failed to heed business’ political preferences and ended up paying a high price for such disobedience: ‘If you want an example of this type of nonbusiness friendly environment just look at California, which has priced itself out of competition for manufacturing facilities. It got so bad there that they effectively threw out their governor. So it is possible for this thing to happen’ (Irish Times, 15 June 2004). One can only imagine the shivers that the prospect of ‘being thrown out’ can send down the spine of any elected politician. Barrett advised that the Irish government should continue to offer ‘aggressive’ incentives to foreign firms, such as low rates of corporate tax, to attract investment and high-quality jobs. It is not difficult to see how such advice carries more weight than the threat from the bin charge protesters that the government will disgruntle its members unless the charges are dropped. While the bin charge protesters can create disturbances, the clout of business’ speech acts rests in their prediction of seriously disturbing prospects. Moreover, unlike ordinary pressure groups, capitalists do not actually have to threaten any protest action at all but can restrict their lobbying to forecasting the business decisions they would be likely to take in the event of unfavourable policy decisions. In the concrete case, these would have involved relocating the plant extension to alternative sites in Israel or the US. But at the same time, businesses like Intel are to a large extent the sovereign architects of their fate – it is their own future decision that their forecasts are about. If a company ‘predicts’ that a policy will force it to shut down a plant (or prevent the opening of a new one), it is ultimately this company alone that avails of the rights and powers to realize this prediction.2 If these implications of the informational–structural model of business political power hold, policymakers should regularly pay more attention to information received by business, while being quicker to ignore messages from non-business groups. A given frequency of interest group contact to policymakers should therefore have a higher impact on policy decisions when the contacting group is business and a lesser effect when the contacting group represents other interests, such as labour, consumers or environmentalists, ceteris paribus. Thus, irrespective of the fact that policymakers are suspicious of lobbyists’ temptations to misrepresent the costs of policy, contact by business lobbyists should lead to more influence over public policymaking than contact by other groups. This proposition can be tested using the European Members of Parliament study, which asks MPs about the extent to which they take into account the opinion of lobbyists when making policy decisions. In particular, we can compare the degree to which the influence parliamentarians are willing to grant lobbyists changes depending on whether they have frequent contact with business interests, trade unions and public interest groups, respectively. There are of course limits to the extent to which elected politicians can be relied upon to give accurate and sincere answers to such questions. It is plausible to expect that the influence conceded to lobbyists will generally be understated. But there are no reasons to suspect that such under-reporting would be in any systematic way related to the other variables of interest.

106 Reputation and informativeness in lobbying The focus is now only on members of government parties.3 In order to compare the influence of business contacts with other group contacts, changes in attributed lobby influence were obtained separately for the three types of interests groups. The second column in Table 5.7 shows the influence over their policy decisions that MPs ascribe to lobbyists when they are contacted less than once a month by business. The next column displays the gradations of influence if business contacts are reported at least once every month. The fourth column, finally, shows the difference in lobbyist influence. Columns five through seven repeat the same comparison for trade union contacts, and columns eight to ten show changing gradations of influence as contingent on public interest group contacts. The findings strongly support the proposition that business contacts are more influential than those with other groups, given the same frequency of contact. Asked how much, on a scale from 1 (‘very little’) to 7 (‘very much’) they take the opinion of lobbyists into account, 30 per cent of MPs contacted by business indicated taking the opinion of lobbyists into account when making political decisions. This figure reduces to 24 per cent for MPs’ contacts with trade unions, and 22 per cent for public interest groups. Moreover, MPs revealed strong increases in influence attributed to lobbyists when they had frequent contact with business. For example, the proportion of MPs ascribing ‘very much’ influence to lobbyists is virtually zero for those who have business contact less than once a month, but it increases to 4 per cent for those contacted by business at least once a month. We note an upward shift in the modal category of lobbyist influence for business contacts, but less so for other groups. Next, it is useful to draw comparisons between, on the one hand, the increase of ascribed lobbyist influence associated with frequent business contact and, on the other hand, the decrease in percentages of those who ascribe little influence to lobbyists associated with the same frequency of business contact. To facilitate this comparison, cell entries for ascribed lobbyist influence are divided up into two panels, separated by the scalar midpoint of 4, which denotes neither much nor little influence. Looking at MPs who ascribed lobbyist influence above the scalar midpoint (the italicized cell entries), the proportion of those ascribing influence to lobbyists rises from 16 per cent for those with less-than-monthly business contact to 29 for those contacted by business at least once a month. This represents a proportion increase associated with monthly or more frequent business contact of 13 per cent, as indicated by the figure in the bottom row. The difference that business contacts make to the weight given to lobbyists is statistically significant as shown by the design-based F-test.4 By contrast the increase in lobbyist influence associated with trade union contact is only 5 per cent, while that for public interest groups reduces to just over 1 per cent. We can say, therefore, that equal frequencies of contact to policymakers are associated with notably unequal degrees of influence of lobbyists over public policymaking, depending on whether the contacting lobbyist is business or some non-business group. Looking at the rows below the scalar midpoint of 4, we find the

16

Sum above midpoint Design-based F-test Wilcoxon Mann-Whitney test

29

4 6 19 19 16 19 17 13 3.67* 4.38***

4 3 6 −5 2 −2 −7 19

1 3 15 22 16 23 20 24

2 5 17 22 13 18 23 5 0.45 0.02

1 2 2 0 −3 −5 2

Diff.

21

1 4 16 22 17 20 20

No

22

3 6 13 23 8 22 26

Yes

1 1.39 0.44

1 3 −3 1 −10 2 7

Diff.

Public interest contacts (N = 657)

Source: 1996 European Members of Parliament Study. Note: Cell entries are percentages of governing party MPs’ responses to the question, ‘How much do you take the opinion of lobbyists into account when making political decisions?’ Robust standard errors in parentheses. Design-based F-test: * significant at 5%, ** significant at 0.1% (two-tailed).

0 3 13 24 14 21 24

Yes

No

Diff.

No

Yes

Union contacts (N = 657)

Business contacts (N = 657)

Very much 6 5 4 3 2 Very little

Influence of lobbyists

Table 5.7 Lobbyists’ influence and MPs’ contact with different interest groups

108 Reputation and informativeness in lobbying matching obverse pattern: frequent business contact is associated with a reduction by 8 per cent of the percentage of parliamentarians who ascribe little influence to lobbyists. By contrast, frequent contact with trade unions reduces the percentage of parliamentarians who ascribe little influence to lobbyists by only 5 per cent. For public interest groups, this reduction decreases further to a mere 2 per cent. For both union and public interest groups, frequent contact is not statistically related to lobbyist influence. Overall, these findings support the propositions derived from the informational model of business lobbying. Following the logic expressed in the previous section that left-leaning or environmentalist policymakers may be more prone to follow their ideological predisposition, mistrust business messages and override lobbyists’ preferences, the readiness of parliamentarians to take lobbyists’ opinions on board might also differ according to their ideological and partisan identity. To allow for that possibility, the analysis in Table 5.7 was replicated separately for left-leaning, right-leaning and pro-business MPs. The overall pattern from Table 5.7 is largely reproduced; the results are therefore not reported separately. In the real world of lobbying, policymakers are of course often targeted by numerous different lobbyists simultaneously (Heinz et al. 1993). Often, the different lobbyists will represent interests that oppose one another in the context of a particular policy struggle. For instance, policymakers considering an environmental protection measure will likely be subject to lobbying both by the environmental groups in favour of the policy, and by opposing interests, often from the world of industry, that engage in countervailing lobbying (Austen-Smith and Wright 1992; 1994). While the European Members of Parliament dataset does not contain information on individual policy issues, we can still note a general overlap of lobbying activity. Thus, 58 per cent of parliamentarians who have monthly or more contact with business groups also report a similar frequency of contact with trade unions. Of the same group of MPs, 41 per cent report frequent contact with public interest groups. Because of this co-variation of group contacts, it is possible that the isolated comparisons of the increases in lobby influence due to business versus other group contacts discussed above give a distorted picture of the extent to which business contact prompts policymakers to take lobbying seriously. A more reliable picture of the informational clout of business contact can be obtained by assessing its effect on influence granted to lobbyists against the background of other group contacts in a multivariate regression model. Like before, the dependent variable in this model is the degree of lobbyist influence on policymaking as indicated by MPs on a seven-point scale. Given the underlying dimension of social reality that is supposed to be expressed in scales of this kind, treating the ordinal scales as an interval measure is appropriate in the current context (Wang et al. 1999). Table 5.8 displays results from OLS regression of alternative models of lobbyist influence. The dependent variable in all six models is the extent to which MPs take the opinion of lobbyists into account when making political decisions, expressed on the seven-point scale displayed in Table 5.7. The first three models are bivariate equations estimating, one at a time, the effects of business, trade union and public interest group contacts on lobbyists’ influence over policymaking.

Reputation and informativeness in lobbying 109 Table 5.8 Determinants of lobbyist influence (1) Business contact

(2)

(3)

(4)

(5)

(6)a

0.48 (0.15)**

0.51 0.47 0.48 (0.16)** (0.16)* (0.12)** Trade union contact 0.14 0.06 0.06 0.27 (0.25) (0.30) (0.30) (0.25) Public interest contact −0.06 −0.20 −0.19 −0.23 (0.26) (0.29) (0.29) (0.26) Official in business org. 0.47 0.25 (0.26) (0.26) Capitalist −0.60 −0.55 (0.50) (0.49) Constant 2.91 3.03 3.11 2.93 2.91 2.03 (0.14)** (0.15)** (0.14)** (0.10)** (0.10)** (0.08)** Number of 657 657 657 657 657 657 observations R2 0.02 0.00 0.00 0.02 0.04 0.14 F 10.30 0.32 0.05 2.90 2.63 8.85 Prob. > F 0.01 0.58 0.82 0.10 0.14 0.01

Source: 1996 European Members of Parliament Study. Note: Standard errors in parentheses. *significant at 5%; **significant at 1% (two-tailed). a Ten country dummies included in equation but not shown. The baseline country is Belgium. The results confirm what has already emerged from the previous bivariate analysis in Table 5.7, namely that frequent business contacts make a difference to the extent to which parliamentarians give weight to lobbyists’ opinions in their policymaking activities, while contacts with trade unions or public interest groups are less relevant. In fact, only business contacts have any statistically significant effect on ascribed influence. Frequent business contact shifts the average influence attributed by MPs to lobbyists upwards by almost 20 per cent. While the average MP from a governing party assigns a lobbyist influence score of 2.9, this value increases to 3.4 when MPs are in monthly or more frequent contact with business. Against the backdrop of a seven-point scale, an increase in half a point is by no means a trivial effect. By contrast, the estimates for trade union and public interest group contact (models 2 and 3, respectively) are very small and fail to reach accepted levels of statistical significance. The coefficient for public interest group contacts even has the wrong sign. Moreover, the business model is the only of the three bivariate models that explains even a small proportion of the variance of lobbyist influence, as expressed in the R-square of 0.02. As legislators that are open to lobbyists from one area or interest group are also contacted more frequently by others, combining the bivariate models into a single multivariate equation invites the risk that the estimates are affected

110 Reputation and informativeness in lobbying by multicollinearity of the covariances of the independent variables. However, bivariate correlation between the group contact measures are only between 0.28 (business and public interest groups) and 0.37 (labour unions and public interest groups) and the overall pattern of the individual bivariate regression results extends robustly to the multivariate regression of model (4), suggesting no noteworthy multicollinearity. The coefficient for business contacts is again the only one that has a noteworthy and statistically significant effect on lobbyist influence. Consequently, adding trade union and public interest group contacts to the equation does not help to increase the model’s explanatory capacity compared to what is explained by business alone, as R-square remains at 0.02. It is possible that the policy communities and elite networks within which MPs operate also influence their policy decisions. At the most basic level, this would suggest that MPs base their decisions on the preferences of any interest groups of which they may be a member. If, as Miliband (1969) and Useem (1984) suggest, common upbringing and similarity in socialization foster shared beliefs among members of political and economic elites, any past affiliation of MPs with business groups should boost the extent to which they consider lobbyists’ opinion in their policy decisions. By the same token, the influence granted to lobbyists should vary depending on whether or not an MP is, or has been, an entrepreneur or ‘capitalist’ herself. Fourteen per cent of respondents in the survey are members of a business organization, while almost 6 per cent have been business people at some stage. In order to control for such network or elite effects, model (5) includes dummy variables indicating whether an MP is also a member of an organization representing the interest of business or whether they have been a capitalist prior to their first election to the parliament. Neither of these variables has a significant impact on the influence enjoyed by lobbyists, although having been an official in a business organization comes close to reaching statistical significance at the 5 per cent level and then adds about as much to reported lobby influence as do business contacts. Most importantly, business contacts retain their effect on lobbyist influence after the effects of policy network and elite variables have been taken in account. The final column reports estimates for the full model including the policy network controls, as well as ten dummy variables capturing the institutional and structural characteristics of the eleven countries. These include, above all, the degree of integration of the political–economic system of interest intermediation; that is, the pluralism–corporatism contrast, but also aspects of the electoral and party system that may impact on the sensitivity of policymakers to lobbying, as well as political culture or indeed any other (often historical) country-specific linkages between individual legislators, parties and interest group organizations that might affect the causal mechanisms under examination. The inclusion of the country dummies increases the R-squared to 0.14. Crucially, the coefficient for business contacts retains its magnitude and significance in the presence of the full set of country dummies. Thus, we can infer with a reasonable degree of confidence that the effect of business contact on the willingness of policymakers to take lobbyists opinion into account when making political decisions is robust and not contingent

Reputation and informativeness in lobbying 111 on particular institutional contexts. The insignificance of contacts with labour and public interest groups is also confirmed in the presence of controls for country effects.

Discussion This chapter has examined a number of individual-level implications of the informational–structural model of business political power. As suggested by the information account of influence, left-wing or environmentalist MPs, whose political attachments and platforms might be at odds with business interests, are lobbied less frequently than their pro-business or right-leaning counterparts. Because these parliamentarians have more incentives to override business preferences in their policy decisions, untrue claims are more likely to be exposed when made to ‘antibusiness’ parties as opposed to ‘pro-business’ parties. As a result, business will lobby left-leaning parties more cautiously – reflected here in less frequent lobbying. While the findings are only partly significant at conventional levels, this hypothesis has found some support in the data. The contrary expectation that left-wing members are lobbied more frequently than others, however, has not survived the tests. This implies that, when devising their lobbying strategies, business lobbyists already factor in the chances of the lobbying message achieving its desired outcome, effectively gearing their lobbying strategy towards both its effects on shortterm policy outcomes as well as on the lobbyist’s long-term reputational interests. The results from the first part of this chapter in particular call into question a major implication of the classic theory of business structural power. According to the logic of structural dependence theories, the incentives that constrain policymakers to maintain business confidence apply equally to all policymakers left, right and centre (Mitchell 1997: 67), in practice eliciting substantial convergence among governments of different partisanship (Dryzek 1996: 79). This, in turn, implies that business has no reason to lobby any one type of policymaker differently, or more frequently, or less so, than any other. Instead, they should all be lobbied the same. However, the evidence presented here clearly indicates that this is not the case. Business lobbyists do discriminate between MPs according to the latter’s partisan attachments, left–right ideological position or importance lent to environmental concerns. By contacting the more receptive right-leaning policymakers more frequently, business discriminates according to the likelihood of lobbying success and with a view to its reputation as a provider of good information. In the language of business structural power, if policymakers are likely to respond differentially to structural constraints depending on their ideological position and partisan identity, an incentive to efficient use of lobbying resources combined with the greater likelihood of ‘desired’ policy responses leads to more frequent lobbying of ‘pro-business’ parties. While the classic theory of business structural power implies relatively low levels of identical (and truthful) lobbying of left- and right-leaning parties alike, any risk that policymakers deviate from the equilibrium response of acceding to business preferences leads business to lobby more selectively. The risk of ‘disobeying’ policymakers is

112 Reputation and informativeness in lobbying explicitly contained in the informational–structural model of business lobbying. The empirical results reported here validate these implications of policymakers’ off-equilibrium behaviour and the responses from business. It also has become clear in the above analyses that the degree of neo-corporatist integration matters. However, it does not do so in a way that corresponds to past research on neo-corporatist patterns of policymaking. While the expectations from the theoretical model in Ch. 3 concerning the relative frequency of lobbying are partially corroborated by the evidence, the patterns differ depending on whether we look at pluralist or corporatist systems. Under the assumption that frequency of contact serves as a good indicator for frequency of misrepresentations, business lobbyists would appear to feel less constrained to lobby (and misrepresent) in pluralist systems, while their good name puts a higher constraint on their behaviour in neo-corporatist systems. This is supported by the fact that levels of frequent contact are somewhat higher across the board in pluralist contexts. In this sense, then, neo-corporatist structures can be seen as creating incentives for business to share its private information about policy costs more carefully, and by implication more truthfully, with policymakers. Turning to the analysis of the ascribed influence of business versus other group contacts, the evidence presented in the second part of this chapter suggests that frequent contacts with business representatives lead policymakers to be sensitive to lobbying. By contrast, equally frequent contact by trade unions or public interest groups does not lead to an increased receptiveness to information conveyed by lobbyists. Thus, the messages of business do indeed appear to carry more clout than those emitted by non-business interest groups, even if all groups lobbied at the same rate of frequency. The information rationale for lobbying and the relative informational–structural advantage enjoyed by business both come out of this as important factors in the explanation of business political power. However, the findings have to be viewed with caution. While the frequency of lobbying is skewed towards right-wing MPs, it is possible that the frequency of truthful lobbying is so skewed even more. In this sense, it could still be the case that while right-wing policymakers are lobbied more frequently, MPs from business-friendly political backgrounds are lied to less frequently. Therefore, the next chapter proceeds to the final step of the triangulated test design by evaluating the informational–structural model of business lobbying in the light of aggregate data on political outcomes at the macro level.

6

Structure, information and environmental regulation

The previous chapter examined some individual-level implications of the informational–structural account of business political power. But if business power is defined as the ability to get policymakers to produce policy in line with business preferences, the results of these behavioural patterns need to be examined. If the informational view of business power is to claim empirical relevance and verisimilitude, it therefore has to be tested at the macro level and with respect to political outcomes. At this level, the informational–structural model would imply that concessions by policymakers to business interests are more likely in contexts where the economic and informational constraints on policymakers’ capacities are considerable, and vice versa. The theory of structural dominance, with its claim that policymakers in democratic capitalist societies depend on business’ co-operation for a healthy and expanding economy, is an integral part of the informational– structural model. This implies a role for the economic evaluation of policymakers by citizens – the phenomenon of ‘economic voting’ – which is a core ingredient to both the theory of the structural dependence of the state on capital and the informational view of business political power. In the informational–structural model, the damage through economic voting that can be incurred by policymakers for picking a business-unfriendly policy features as the size of negative inducement effects. To the extent that structural constraints such as economic voting vary across political contexts, they should then function as correlates of policy outcomes. Positive finding in this regard would constitute evidence in support of both the theory of the structural dependence of the state on capital, and the informational account of business political power at the centre of the present study. The other main ingredient in the informational–structural model is information asymmetry. To the extent that this factor varies across political contexts and contributes to explaining policy outcomes, we would expect policymakers to cater more frequently to business interests the more they depend on business for data and expertise relevant for assessing the economic and electoral consequences of a policy. While the classic theory of the structural dependence of the state on capital helps to identify if and when objective structural constraints circumscribe the policymaking capacity of political decisionmakers, the informational model claims to reveal if and when policymakers believe that these constraints apply. Variations in

114 Structure, information and environmental regulation information asymmetry can be due to differences in the institutional capacities for efficient and independent policy-impact analysis. These can result from differential resources and research capacities of public bodies and civil administration, relative to the knowledge resources available to business. Variation can also originate in various institutions that entice business to provide policymakers with truthful and reliable information. As the analysis in the previous chapter suggests, incentives for business lobbyists to report truthfully on the assessment of policy effects will be higher in neocorporatist systems of interest intermediation, where close and institutionalized bargaining structures exist between governments and the major sectoral interest groups. Furthermore, the two elements – information asymmetry and structural constraints – can be expected to combine and interactively affect policymakers’ ability to devise optimal policy. Policymakers may find it expedient to heed business advice if a policy’s potential negative inducement effects would be large. They may also be more open to business advice if their own information base in a policy area is relatively weak. But they will be most receptive to lobbying if both conditions apply simultaneously; that is, if a lot is at stake in terms of policy costs and business informational advantage in the area is considerable. Therefore, any information effects should be amplified when the structurally induced costs of picking the wrong policy would be high. To assess the effects of informational and structural sources of business political power, this chapter analyses the determinants of environmental regulation in 23 countries. After examining what environmental regulation does for business, the following section identifies the factors that have been found in previous comparative research to affect the level and strength of environmental policymaking in different countries. Next, an empirical model of the causal relationships between informational and economic structural sources of business power on the one hand and the strength of environmental regulatory regimes on the other will be developed. After the necessary aspects of data sources and measurement have been discussed, the results are presented and interpreted.

Business preferences on environmental regulation Aiming to reduce negative externalities flowing from the actions of citizens and businesses, environmental regulation has cost implications for business (Golub 1998: 1). Increasing sensitivity to global economic competition and budgetary constraints makes governments wary of any form of regulation which might threaten economic growth, foreign investment, exports and employment creation. Regulations requiring firms to reduce emissions, increase recycling, pay more for energy or switch to more expensive fuels and inputs all raise the final price of their products, with the result that ‘green’ states may lose markets to ‘dirty’ states that lack similar environmental standards (Golub 1998: 4). An oft-heard criticism of environmental policies, therefore, is that environmental gains come at too high a cost, where costs are often represented in money, jobs and growth. Protection of the North American spotted owl was claimed to cost between

Structure, information and environmental regulation 115 20,000 and 140,000 jobs, mainly in the timber industry (Bezdek 1993). In a similar vein, the American Petroleum Institute blamed environmental restrictions for the loss of 400,000 jobs during the 1980s. The US Motor Vehicle Manufacturers Association argued that increasing fuel economy standards would cost 300,000 jobs, an assertion repeated by former US vice president Dan Quayle during the 1992 campaign (ibid.). Sugar growers in Florida said protection of the Everglades would cost 15,000 jobs locally (ibid.). US mining companies claimed that a planned reform of the Mining Law could cost 47,000 jobs in an industry that suffered a 51 per cent loss of its jobs between 1979 and 1992 (ibid.). Some economists assert that environmental regulation was a significant contributor to the US productivity slowdown of the 1970s (Christiansen and Haveman 1981; Gray 1987). Overall, however, economic impact estimates show a wide variance depending on the methods and models used and on the economic indicator selected. A common finding is that environmental policies increase inflation mildly and decrease productivity only very slightly. The same policies also tend to lead to slight net increases in employment (Tietenberg 1994). However, job losses clearly have occurred in specific industries, communities and regions as a result of environmental policies (Kraft 1996: 170). There may be fewer losses than initial estimates suggest, but, as Kraft points out, this is of little comfort for those who lose their jobs or to the businesses forced to close their doors (ibid.). But exaggerations seem to be the norm rather than the exception. At the macro level, environmental policies have only a very small impact on the economy, no matter what measure is used: inflation, productivity or jobs (Portney 1990). In the US, the Bureau of Labor Statistics found that during 1988, only 0.1 per cent of all layoffs were attributable to environment-related causes. As Bezdek (1993) reminds us, that means that 99.9 per cent of jobs lost in the US during that time were the result of other factors, such as free trade, adoption of new technologies and fundamental shifts in the economy. And the price to be paid for environmental policy in this view does not always have to be of an economic kind. Other negative inducement effects are regularly invoked in debates surrounding the efficiency of environmental policy. Wildavsky (1995), for example, claims that by ‘searching for safety’ and favouring governmental regulations that constrain economic investment and growth, environmentalists cause the population to be overall less safe. In the context of environmental policy, the frequently invoked trade-off between jobs and clean air is just a special case of the more general problem of negative inducement effects. The notion of a negative link between environmental protection and economic performance is of course neither clear nor undisputed. The ‘Porter hypothesis’ claims that environmental progress can be achieved without sacrificing competitiveness (Porter and van der Linde 1995). Jaffe et al. (1995) emphasize the productivity-enhancing effects of environmental regulation, which induces firms to install cleaner, more efficient technologies. Arrow et al. (1996) suggest that all major regulations should pass a cost–benefit test. Kraft (1996: 166) points out that some very stringent environmental regulations – such as the

116 Structure, information and environmental regulation removal of lead from gasoline and the phase-out of chlorofluorocarbons (CFCs) as a propellant in spray cans – could easily survive such a test, while others might not. In fact, the adoption of environmentally benign technologies and encouragement of energy efficiency, source reduction, waste recycling and similar efforts all have been claimed to improve employment because they tend to be more labour intensive than the polluting technologies and practices they replace (ibid.: 171). Even at the level of the individual firm, environmental regulation can create opportunities for business, as well as protectionist benefits through stricter environmental policies, leading business to support increased regulation. In practice, however, this is only rarely the case (Murphy 2004). The aforementioned historical examples illustrate that error-free evaluation can really only be made with the benefit of hindsight. High-profile policy shifts such as the phase-outs of CFCs and leaded car fuels drew severe criticism from the affected industries when they entered the legislative agendas in different countries. The arguments put forward made it clear that, in the eyes of industry, these policies would specifically not pass the test suggested by Arrow et al. Over the long run, environmentalists maintain, policies that produce environmentally sound and efficient technologies and processes are also the economically most efficient ones, at any rate if the full cost of environmental damages is part of the equation (Kraft 1996: 171). However, there are often short-term economic advantages to weak environmental laws (Levy 2005: 81–3). It is then not surprising to note that the bulk of business reactions to increased regulation highlight compliance costs and uncertainty. These are claimed to lead to cutbacks in research and development efforts, limit the innovative efforts of firms and undermine their general profitability. Firms also perceive environmental policy to place them at a competitive disadvantage. As a result, they tend to emphasize the costs of environmental policy and underestimate the benefits, which leads them to oppose tougher regulation more often than support it. This is not to say that business does not value the environment or has a preference for pollution. But rather than being subjected to new regulations and other government action business prefers to handle its environmental impact in a voluntary or perhaps self-regulatory manner. Firms and business associations resist regulation if they believe that it is possible to retain a status quo of no regulation or to obtain self-regulation and ‘market solutions’ instead of government regulation. Only if these alternatives are taken off the table, either by the government or by market developments, do we witness industry support for stricter environmental policy. The CBI, for instance, accepts that, to some degree or other, it has opposed every one of approximately 250 EU environmental directives passed over the past two decades (The Guardian, 17 August 2005). Thus, it is reasonable to assume that firms on average prefer less (or weaker) environmental regulation over more (or stricter) regulation. Equipped with this simplifying but plausible assumption about business preferences, it is possible to ascertain the relevance of the explanatory value of the structural and informational sources of business political power by

Structure, information and environmental regulation 117 adding them as predictor variables to a simple model of environmental regulatory strength.

Dowsing flooded fields for underground water What variables should be included in a standard model of environmental regulation? A considerable body of theoretical and empirical work has been produced on the economic and political determinants of environmental quality. While some studies are interested in explanations of biophysical environmental outcomes (e.g. Grossman and Kruger 1995), most treat environmental quality as a proxy for environmental protection (e.g. Scruggs 1999). For the present analysis, a more direct measure of environmental regulatory stringency is employed. This has the advantage of reducing the number of control variables that would otherwise be required if physical output measures were to serve as proxies for environmental regulatory efforts. Population density or energy efficiency data would for example be required to infer the strength of environmental regulation from emission data. These controls are not needed when regulatory stringency is measured directly. On the other hand, a number of standard variables that affect the outcome of environmental regulation – environmental quality – will also impact on the extent to which strict regulations are enacted in the first place. First, environmental policy outcomes display a robust relationship with societal wealth and economic development (Grossman and Kruger 1995). Economic development of the kind set off by the industrial revolution goes hand in hand with creating a range of environmental problems, generating demand for abatement and remedies of some form or another. In addition, the security of material existence and welfare generated by economic development leads to a gradual amendment of materialist value orientations with postmaterialist ideas among citizens, moving quality of life concerns higher up the agenda (Inglehart 1997). Second, past research has found consistent and robust evidence that a strong presence of green and left-libertarian parties in parliament is associated with stronger environmental protection (Neumayer 2003), suggesting that party politics matters for environmental policymaking. Third, consensual modes of policymaking have been positively associated with an enhanced ability of policymakers to legislate, monitor and enforce environmental regulations, while majoritarian institutions tend to have the opposite effect (Crepaz 1995; Scruggs 1999). And fourth, countries that are members of the EU will have a certain set of environmental protection measures in place that have been imposed on them by a supranational policymaker, often on the impetus of environmental leaders within the EU (McCormick 2001). Once cross-country variation in the strength of environmental regulation has been accounted for by these standard predictors, effects of the structural and informational sources of business political power can be estimated. The classic theory of business structural power holds that policymakers cater for business interests because their own political survival depends on the accumulation

118 Structure, information and environmental regulation process, with a healthy economy guaranteeing adequate state revenues and political support. In the political science literature, the link between the economy and policymakers’ survival is frequently conceptualized as economic voting. According to Mitchell, [e]conomic voting ensures that governments are acutely sensitive to variations in the major macroeconomic indicators. The significance to political leaders of the signals from the economy places business institutions and membership groups in a very strong position in the policy struggle. Mitchell 1997: 62 Similarly, in Smith’s account, incumbent politicians of all kinds … expect a strong economy to augment their odds for reelection and expect weak economic performance to mitigate their electoral prospects. … Given the abundance of evidence indicating that the health of the economy actually does affect American election outcomes, it should not be surprising that politicians perceive such a relationship. Smith 1999: 842–3 There is indeed evidence that they do, and policymakers are well advised to anticipate the effects of economic voting in their policy decisions (Kiewiet 2000). Unless it can be shown that damage to business profitability is marginal, this mechanism can be expected to compromise environmental policy goals (Dryzek 2000). The argument also implies that the effectiveness of business structural dominance varies depending on the strength of this mechanism. While structural power theories are usually cast in deterministic terms, economic voting varies from country to country. If economic voting reflects business structural dominance, governments should afford to place higher emphasis on environmental protection in countries in which the economic–electoral constraint is less significant. Accordingly, the strength of environmental regulation should decrease with the extent to which economic factors are predictors of electoral outcomes. It was argued in Ch. 2 that the economy affects the utility of policymakers not only through economically oriented voting behaviour but also through tax revenues and the contingent ability of policymakers to achieve their policy goals. Thus, besides electoral behaviour there is also a direct economic basis to the structural sources of business power. Structural characteristics of the national economy and their relation to the cost of environmental regulation should therefore play a role in determining the structural constraints on environmental policymaking capacity. For example, the extent to which policymakers embrace the environment as a topic and enact strong protective regulations has been claimed to be negatively contingent on the importance of high pollution industries (Crenson 1971). If the structural power thesis holds, the relative contribution of ‘dirty’ and energy-extensive industries to the national economy should affect the stringency of environmental policy negatively. Furthermore, arguments about capital flight

Structure, information and environmental regulation 119 to pollution havens and the ‘race-to-the-bottom’ metaphor suggest a relationship between environmental regulation and the importance of international trade for a country’s economy. The nature of the relationship between environmental politics and openness to trade is in fact more ambiguous. On the one hand, trade integration can increase environmental co-operation by promoting growth and technological development, thereby increasing public demands for environmental protection while reducing the pollution intensity of production (Vogel 1997). On the other hand, integration in global trade provides governments with incentives to keep the domestic cost burden on producers low and minimize the costs to business from strong environmental regulation (Péchoux and Pouyet 2002). Following this logic, dependence on international trade is expected to affect the strength of environmental regulation negatively. As much as the structural power thesis focuses on the consequences of policy, it ignores the possibility that information about consequences can be a problem. In Lindblom’s view (1977: 185), business is structurally privileged because it only needs to point out any undesired effects of a policy to persuade policymakers to rethink their ambitions. The problem with this view is that, even if all policymakers only want to enact legislation to benefit business and the economy while sidestepping polices they think will harm the economy, they are frequently uncertain about the likely consequences of various policies (Smith 1999: 861). As the theoretical argument in Ch. 3 has demonstrated, business actors have incentives and opportunities to exploit this situation to their advantage by exaggerating the costs of undesired policy. The problem for policymakers is that they often lack the information necessary to assess the veracity of business’ assessment of a policy’s negative inducement effects, while the relevant information is directly available to business. While some fundamental uncertainty regarding a policy’s true costs remains on both the policymaker’s and business’ side, business lobbyists have particularly high incentives to assess policy from a primarily conservative perspective. For example, in a retrospective evaluation of regulatory programmes, the US Office of Management and Budget (OMB) found that the benefits to health and the economy of environmental regulations significantly exceed the economic costs of complying with those regulations. For the decade from October 1992 to September 2002, the agency estimated the total annual benefits to lie between $120,753 billion and $193,163 billion, while the total costs of federal rules amounted to between $23,359 billion and $26,604 billion (Office of Management and Budget 2003: 7, Table 2). The finding is particularly remarkable as the OMB is part of the presidential executive office and has historically proven to be a sceptical watchdog accustomed to restraining the Environmental Protection Agency’s (EPA) regulatory enthusiasm. For the present concern, it is even more interesting that these findings run counter to deeply engrained beliefs about the economic costs of environmental policy. And indeed, several commentators on the OMB report maintain that the EPA typically underestimates the economic costs when proposing new regulations, for example by focusing on compliance cost to business rather than taking ‘full social costs’ into account (Office of Management and Budget 2003: 169). A former EPA administrator acknowledges that this is a widely held view which is, however,

120 Structure, information and environmental regulation wrong in most instances (Washington Post, 14 October 2003). He illustrates this point by reference to the Bush (Senior) administration’s Clean Air Act. Designed to bring about far-reaching changes to require cleaner cars, power plants, factories and fuels, the bill was debated in Congress in 1990. Various agencies and groups, including the EPA, the OMB, the Council of Economic Advisers (CEA) and the electric power industry, produced estimates of compliance costs. With the benefit of hindsight, it turns out that the industry ‘overestimated’ the costs of the policy by 650 per cent, while the joint projections of the EPA, OMB and CEA overestimated the true costs by ‘merely’ 300 to 400 per cent (Washington Post, 14 October 2003). Thus, evidence related to the economic impact of environmental policies is often ambiguous, and neither the regulator nor the regulated have all the data or the ability to interpret them. As a result of these information problems, business can sometimes achieve its political goals by virtue of its privileged information about the consequences of policy. As the case study on small business banking in the UK (Ch. 4) suggests, information asymmetry has to be treated as a variable factor, rather than a constant, even within a given policy area. Policymakers will be more susceptible to untruthful lobbying, the more they depend on business for the gathering and interpreting of pertinent data. At the aggregate level, this implies that policymakers should compromise their environmental policy goals more often to accommodate industry concerns if business’ information resources are large relative to those available to public officials. It follows that the stringency of environmental regulation should decrease with the degree of informational asymmetry between business and policymaker. Lastly, policymakers should be most receptive to business lobbying if structural parameters suggest potentially high policy costs and business’ informational advantage over policymakers is considerable. Therefore, any negative effects of information asymmetry on environmental regulation should be stronger if potential policy costs are high, and weaker if negative inducement effects would be milder. Quinn and Shapiro (1991: 868) warned that the demonstration of structurally originating political power is more difficult than finding direct evidence of more readily observable instruments of power like making campaign contributions or lobbying. To employ their metaphor, dowsing a field for underground water using a water-witching stick is not made easier by the fact that the political field is usually flooded with the overtly manifest exercise of business power. Therefore, the empirical model ought to control for business political action. A major political advantage for business has been identified in its ability to organize effectively. A relatively small number of group members combined with a concentration of benefits from collective action gives business much stronger incentives to organize for political action than larger, more diffuse groups such as consumers or taxpayers, over whom both costs and benefits are more widely dispersed (Olson 1965). In no policy area is this more evident than environmental protection, where the group of beneficiaries of strict regulation is large (all citizens), while the group bearing the costs of regulation is smaller and more concentrated (business). The organizational strength of business varies considerably across countries, to the extent that the degree of business organization makes a difference for political outcomes, the

Structure, information and environmental regulation 121 strength of environmental regulation should decrease with business’ organizational strength.

Measuring structural constraints and information asymmetry Data on these factors have been compiled for 23 rich industrialized OECD countries and Israel. For the most part, these countries have enjoyed uninterrupted liberal democratic regimes since the mid-1970s, when environmental issues started to make first inroads into national political agendas following the publication of the Club of Rome’s study, The Limits To Growth, in 1972.1 Unless indicated otherwise, the sample year is 2000. Commonly used measures of environmental policy output include expenditure figures for environmental programmes, pollution abatement and control expenditures and physical pollution levels. But the strength of regulatory regimes will not be directly captured by these measures unless an expanded set of statistical controls is applied. Even then, the problems of inferring regime strength from physical outcomes are immense. A more direct measure of environmental regulatory stringency is available in the form of the World Economic Forum’s (WEF) ‘Stringency of Environmental Regulations’ index. Based on a survey of business leaders for the 2001 Global Competitiveness Report, the index measures environmental regime stringency as perceived by political and business elites in the countries under study (World Economic Forum 2002). Respondents were asked to rate the overall stringency of the environmental regime in their country on a seven-point scale. The survey was carried out by the WEF in collaboration with the Center for International Development and the Institute for Strategy and Competitiveness at Harvard University. Answers from an average of 61 respondents per country were averaged to obtain country scores, which can then be treated as quasi-interval measures. For the 23 countries in the present analysis, the resulting scale of environmental regulatory stringency ranges from 3.74 in Greece to 6.73 in the Netherlands. The main caveat to using elite perceptions as an indicator of regulatory strength is that, from a business perspective, the regulatory burden may be uniformly exaggerated, particularly in response to a survey conducted by the WEF. The WEF is the global peak organization of the world’s economic and political elites in the post-Keynesian era. Holding annual meetings in the elite seclusion of Davos, Switzerland, the WEF plays a leading role in promoting neoliberal policy programmes, economic globalization and the worldwide liberalization of finance and trade. Indeed, the organization claims credit for the launching of the Uruguay round of the General Agreement on Tariffs and Trade (GATT), which culminated in the World Trade Organization. According to Huntington (1999), ‘Davos people control virtually all international institutions, many of the world’s governments and the bulk of the world’s economic and military capabilities’. Given the organization’s unambiguous ideological orientation, data generated on its behalf through surveying business elites may not be an unbiased measurement of environmental regulatory strength. In a sense, this lack of objectivity is desirable, as we are interested in measuring the extent to which regulatory mandates come into

122 Structure, information and environmental regulation existence that may be perceived as too strict by business. Respondents may also have a tendency to overestimate their countries’ environmental regime strength because concern for environmental issues has such positive connotations that people might want to identify their countries with it. Either source of bias might lead to a compression of the differences between countries. Therefore, some validation with hard data may be useful. The measure is strongly correlated with actual physical pollution levels. Across the 23 countries, the regime index correlates negatively with urban total suspended particulate concentration (r = −0.71, p = 0.001) and urban sulphur dioxide concentration (r = −0.56, p = 0.007). This suggests that elite perceptions of environmental regime strength are in some systematic relation to country differences in actual environmental policy. Following the conceptualizations in the literature (e.g. Mitchell 1997), the structural sources of business dominance should be reflected in the strength of the economic determinants of electoral behaviour. Economic voting is measured using voter judgments on trends in general economic conditions as predictors of vote choice for or against incumbent candidates. Past findings are inconclusive as to whether voters take their clues from the wider economy and behave, in the terminology of the field, ‘sociotropically’, or follow a pattern of pocket-book voting (‘egotropic’ voting), as was originally reported by Kinder and Kiewiet (1979). While rational voters should be prospective in their assessment of election candidates and how the latter would perform in office once elected, most findings indicate that measures of retrospective economic evaluation by citizens work better (LewisBeck and Paldam 2000: 114–5). Given citizens’ considerable difficulties associated with the ex ante evaluation of politicians’ competence, it is reasonable to assume that voters derive clues from observing the past performance of incumbents. Therefore, citizens’ retrospective evaluations of the economy are used as the independent variable in a model for the assessment of the size of the economic vote. Using data on 21 elections in 16 countries from the Comparative Study of Electoral Systems (CSES, Module I) dataset, logit estimates of the effect on vote choice of voters’ perceptions of how the economy changed over the previous year are obtained.2 The dependent variable is the dichotomized incumbent-versus-opposition vote choice. Controlling for standard predictors such as party and candidate approval scales and socio-demographic factors, the logit coefficient represents the extent to which variation in citizens’ perceptions of economic performance can explain the probability of observing a vote for the incumbent party or parties. The resulting distribution of the economic vote across the 21 elections for which data are available is shown in Figure 6.1. The pattern is for the most part in line with what both common sense and past cross-sectional research on economic voting would suggest: Great Britain exhibits strong patterns of economic voting, Germany and the US are somewhat in the middle, while Belgium and Denmark display no measurable signs of economic voting. Overall, cross-country variation is not very large. With the notable exception of Great Britain, differences in country estimates tend to be smaller than the 95 per cent confidence intervals around country estimates.3 The other sources of structural constraints on environmental regulations are the relative contribution of ‘dirty’ and energy-extensive industries to the national

Structure, information and environmental regulation 123 1

0

−1

−2 Upper/lower 95% confidence band Logit coefficient

G

BR SC 97 O ES 97 P C 00 AN PR 9 T 7 N 02 ZL A 9 W US 6 G 9 ER 6 IS 98 L N 99 LD U 98 SA ES 96 P SW 96 SW Z 0 0 E N 98 O R JP 97 N Q 96 BE BC LW 97 BE 9 9 EG LF 9 9 ER D 98 N K 98

−3

Election

Figure 6.1 Country scores of economic voting

economy and a country’s dependence on international trade. Of these, the first is measured by the manufacturing share of GDP, taken from the United Nations Conference on Trade and Development Handbook of Statistics. Trade dependence is defined as exports plus imports divided by GDP and taken from the Penn World Tables. Both variables should affect the stringency of environmental policy negatively. Information asymmetry cannot be observed directly (Kennan and Wilson 1993: 96). Even if this variable is not thought to be strictly unobservable, measurement is very difficult and various proxies are employed in the literature. In analyses of investment markets, for example, information asymmetry between a firm’s management and its shareholders is hoped to be captured by the degree of dispersion in analysts’ forecasts (e.g. Healy and Palepu 2001: 416–27). The idea is that as the amount of information asymmetry increases there is likely to be less consensus among financial analysts about the future performance of a firm. Similar approximations are difficult to identify for asymmetric information between regulator and regulated. Indeed, information asymmetry has never been measured before in political analysis. In the absence of established solutions, three alternative proxies will be used. First, the relative informational clout of business vis-à-vis the state will be captured by the percentage of gross domestic expenditure on research and development (R&D) financed by industry. This measure is based on the source of funding for R&D rather than who performs it; it therefore also measures the extent to which, for example, university studies on environmental problems and their abatement costs are financed by business. Apart from a

124 Structure, information and environmental regulation small third category of ‘other national sources’ accounting for an OECD average of 4.5 per cent, the industry share of national R&D expenditure is the inverse of government R&D expenditure, thus reflecting the informational balance between government and private business. Environmental regime strength is expected to decrease with business’ share of R&D. Next, the institutionalized relationships between policymakers and business characteristics of neocorporatist systems of interest intermediation can induce business to lobby more truthfully in a variety of ways. First, neocorporatism can reduce the information gap by encouraging the sharing of information. In neocorporatist systems, sharing information with policymakers is less risky for business. According to Hall and Soskice (2001: 47–8), the relative empowerment of firms vis-à-vis governments in co-ordinated market economies also encourages firms to share information with policymakers, as firms in these systems can be more confident that governments will not abuse the newly acquired information. Conversely, business lobbyists face fewer incentives to provide accurate information in pluralist systems and have more reasons to be wary of sharing data with regulators and other government officials. Second, by involving business in institutionalized and cooptative relationships with government agencies, neocorporatism not only reduces policymakers’ monitoring costs but also increases the reputational costs attached to untruthful lobbying (Broscheid 2006: 95). This reduces the leverage of information asymmetry by providing business with disincentives from using privileged information strategically. Overall, neocorporatism facilitates informative communication between business and policymakers. Therefore, both the magnitude of information asymmetry as well as its effect on the policy process will decrease with the degree of political–economic integration as expressed in Siaroff’s measure for neocorporatism. The measure is expected to have a positive effect on environmental regime strength. Third, the presence of counteractive lobbyists can enhance the informativeness of lobbying (Austen-Smith and Wright 1992; 1994; Lohmann 1993). While countervailing groups may not match the political clout of producer interests, the latter’s influence can be checked to the extent that business’ information on compliance costs is complemented by information from alternative sources. Analogous to the monitoring role of interest groups in principal–agent relations between elected officials and bureaucrats, environmental movement organizations can serve as de facto auditors whose activities shift verification costs from governments to societal actors (cf. Banks and Weingast 1992). In this way, counteractive lobbyists increase the risk of detection for liars and induce more truthful behaviour by both groups of lobbyists. The presence of lobbyists with environmental agendas should therefore countervail the effect of information asymmetry between business and government. The strength of the environmental movement is measured as the percentage of environmental group members among the population, with data taken from the 1999–2001 World Values Survey. Environmental group strength is expected to increase the strength of the environmental regime. Despite the prominence of organizational strength in the literature on special interest politics, no commonly accepted cross-country measure for this source of business power exists and researchers have had to make do with less-than-optimal

Structure, information and environmental regulation 125 solutions. One such measure has been suggested by Bischoff (2003), who uses the density of business organizations, defined as the number of trade associations per million inhabitants. The problem with this measure is that numerous business associations exist in name only; they have no staff or money and do not engage in any political activity. Moreover, it is by no means clear that a large number of organizations is indicative of business strength. On the contrary, a proliferation of small business associations might reflect a lack of business unity and organizational strength. As the relationship between formal density and actual organizational strength varies across countries, this measure is problematic. However, business power can be measured indirectly by recording countervailing power. A major determinant of business strength is the relative strength of organized labour. The cohesion of business representation has been characterized as a reaction to labour cohesion (Mitchell 1997: 101–4). Historically, there is a strong structural correspondence in the differentiation and power of labour and business interest group systems in the advanced industrialized countries (Crouch 1993). The organizational strength of business can therefore be tapped by measuring the organizational strength of labour unions, for which reliable cross-national data are widely available. This is measured as trade union density, defined as the percentage of the non-agricultural workforce with membership in a trade union.4 Union density is expected to have a negative effect on environmental regime strength. Finally, economic development is GDP per capita, measured in US dollar purchasing parity. Much of the literature on the environment and the economy assumes – and finds more or less robust evidence for – a non-linear relationship between economic development and pollution levels (the so-called ‘environmental Kuznets curve’). As the countries in the present sample are among the wealthiest in the world, the per capita GDP of even the poorest of them is above the turning points for the effect on economic development on environmental outcomes, which have been estimated to be as low as $12,000 for CO2 emissions (cf. Moomaw and Unruh 1997). Thus a linear specification of the link is applied here. Data sources, coding details and summary statistics of all variables are provided in Appendix C.5

Sources and limits of business power: International trade, neocorporatism and eco groups Given the small number of observations available for analysis, potentially influential outliers merit special attention. Reported membership in environmental organizations, for example, is extremely high in the Netherlands. Another outlier is Greece, which combines the weakest environmental regime as well as exceptionally low scores on the economic structural variables. In the bivariate analysis, Greece is therefore excluded from correlations of regulation and economic–structural factors, while the Netherlands are left out when the effect of environmental movement strength is assessed. But even then, support for the structural sources of business power is mixed at best, as can be seen in Figure 6.2. Economic voting is negatively correlated with environmental regime strength, suggesting that the economic vote mechanism makes policymakers cautious of policies that might upset business confidence in countries like Spain and Portugal,

126 Structure, information and environmental regulation 7

NL

DE DK

6.5

SE

CH

IS AU FR BE NO US NZ CA IT JP

6 5.5

UK IE

5

ES

4.5 4

b = −1.048; se = 0.739; R-sq = 0.10; N = 20

0

Stringency of environmental regulations

PT

IL

0.2

0.4

0.6

0.8

1

Strength of economic voting 7 6.5

NL NO

6

AU

DK IS

AT

NZ UK

IT

JP IE

5

ES PT IL

4.5 4

FI CH

BE CA

US

5.5

DE

SE

FR

b = −0.005; se = 0.033; R-sq = 0.00; N = 22

10

15 20 25 Manufacturing share of GDP

7

DE DK SE AT FI NO CH FR IS US AU NZ CA UK JP IT

6.5 6 5.5

30

NL BE

IE

5

ES PT

4.5 4

IL

b = 0.000; se = 0.004; R-sq = 0.00; N = 22

0

50

100 150 Trade dependence

200

Figure 6.2 Environmental regime strength and structural sources of influence

as well as Ireland and the UK. This relationship is not very strong, however, and statistically significant only at the 10 per cent level (one-tailed). Both the structural composition of capital and the importance international trade, by contrast, seem to be unrelated to environmental policymaking. The picture looks better for the informational sources of business power. Business’ share of national R&D is related to environmental policymaking, but the

Structure, information and environmental regulation 127 7 AT

6

IS CA AU IT

NZ

NL NO FR

DE FI CH SE BE US

DK

UK JP IE

5

ES PT

4 3

GR b = 0.027; se = 0.010; R-sq = 0.02; N = 21

Stringency of environmental regulations

30

40 50 60 Business share of R & D

7

NL DE US

6

CA

IS FR

UK

AU

NZ

CH

BE

IT

DK FI

AT SE NO

JP

IE

5

ES PT

4 3

70

IL

GR b = 0.467; se = 0.146; R-sq = 0.33; N = 23

2

2.5

3 3.5 Neocorporatism

DE NO

6

SE

CH

IS FR

US

AU CA IT

JP

4.5

DK

AT

FI

6.5

4

BE

NZ

UK

5.5 IE

5 4.5

ES PT b = 0.065; se = 0.029; R-sq = 0.22; N = 20

0

5 10 Environmental group strength

15

Figure 6.3 Environmental regime strength and informational sources of influence

effect is in the wrong direction (Figure 6.3). Rather than eroding policymakers’ capacity to enact tough environmental policy, a strong role of business in society’s knowledge production appears to give a boost to environmental protection. The other two measures of informational power, however, perform as expected. Both the degree of neocorporatist integration and the strength of the environmental movement are significantly and positively associated with environmental regime strength.

128 Structure, information and environmental regulation Environmental regulation is determined by a whole host of other variables that in turn are related to factors such as R&D expenditures or the strength of new social movements and that should therefore be controlled for. To analyse the effects of the structural and informational factors on business political power in the area of environmental regulation in a multivariate setting, a standard model of environmental regime strength is estimated first. The various measures of business power are then added one at a time. The results are reported in Table 6.1. Because environmental regime stringency has a significant negative skew, the errors will not be normally distributed. To deal with this violation of the OLS assumptions, estimation is by least-squares regression with Huber–White robust standard errors (White 1980). As the hypothesized effects are directional, one-tailed tests of significance are appropriate. In order to maximize the degrees of freedom available for model estimation under small-N conditions, all available cases are used for estimation, including Greece and the Netherlands. To nonetheless control for the impact of these outliers, jack-knifed estimation of the variance–covariance matrices is applied in an alternative procedure to the Huber–White estimator. This procedure involves estimating 23 new 22-country models (20 countries in the case of the economic voting model and 21 for models including the business share of R&D and strength of the eco movement) by omitting one country at a time, producing ‘jack-knifed’ estimates for each coefficient. The resulting coefficient for each independent variable is the mean of the 23 22-country coefficients, and its standard error is the standard deviation of the variable formed by the 23 22-country estimates. The jack-knifed estimates confirm the findings reported in Table 6.1 and are displayed in Appendix C. The parsimonious baseline model (1) performs reasonably well, accounting for two-thirds of the observed variation in environmental regulation. Economic development and left-libertarian party vote have the expected positive effects on environmental regime strength. While the coefficient estimates for EU membership and majoritarian constitutions point in the right direction, they have rather large standard errors. Environmental regulation becomes weaker the more majoritarian the institutional framework of a country is, but with the exception of models (7) and (9) the effect is not statistically significant. EU membership increases environmental regime strength, but the coefficient is significant only in models (5) and (6). Overall, the relative performance of these standard predictors is in line with the findings of previous studies (Scruggs 1999; Neumayer 2003). Business organizational capacity, as represented by union density, is unrelated to environmental regulation. Only some of the effects of the structural and informational indicators of business political power are robust to multivariate control. Economic voting is no longer significant, and its coefficient now has the wrong sign. As before, the GDP share of manufacturing industries is unrelated to environmental regulation. With basic controls in place, trade dependence now turns out to be significantly negatively related to environmental regime strength. While the effect is not very large, it is by no means insignificant. An increase in the international trade share of a country’s GDP by 100 percentage points (e.g. the difference between the trade dependence

(3)

(4)

(5)

(6)

(7)

(8)

(9)

0.133 0.131 0.129 0.131 0.134 0.109 0.118 0.119 0.111 (0.027)∗∗∗ (0.029)∗∗∗ (0.026)∗∗∗ (0.025)∗∗∗ (0.030)∗∗∗ (0.035)∗∗∗ (0.024)∗∗∗ (0.036)∗∗∗ (0.035)∗∗∗ −0.051 −0.064 −0.041 −0.060 −0.018 −0.046 −0.330 −0.014 −0.243 (0.072) (0.087) (0.110) (0.078) (0.068) (0.082) (0.127)∗∗∗ (0.075) (0.104)∗∗ 0.090 0.087 0.090 0.091 0.109 0.089 0.068 0.080 0.076 (0.028)∗∗∗ (0.032)∗∗∗ (0.031)∗∗∗ (0.028)∗∗∗ (0.025)∗∗∗ (0.028)∗∗∗ (0.022)∗∗∗ (0.030)∗∗∗ (0.020)∗∗∗ 0.288 0.281 0.231 0.264 0.406 0.163 0.232 0.153 0.225 (0.196) (0.211) (0.216) (0.178)∗ (0.165)∗∗ (0.207) (0.183) (0.223) (0.186) 0.002 (0.007) 0.201 (0.772) 0.008 (0.034) −0.005 −0.005 (0.002)∗∗ (0.002)∗∗ 0.006 (0.009) 0.428 0.376 (0.135)∗∗∗ (0.150)∗∗∗ 0.012 0.014 (0.007)∗∗ (0.006)∗∗ 2.055 2.056 2.080 1.972 2.295 2.406 1.251 2.436 1.822 (0.697)∗∗∗ (0.710)∗∗∗ (0.822)∗∗ (0.901)∗∗ (0.742)∗∗∗ (0.892)∗∗∗ (0.694)∗∗ (0.970)∗∗ (0.939)∗∗ 23 23 21 23 23 22 23 22 22 0.66 0.65 0.62 0.65 0.70 0.59 0.76 0.61 0.75 10.54 8.33 8.79 8.99 16.79 4.75 12.68 13.11 18.40 0.000 0.000 0.001 0.000 0.000 0.008 0.000 0.000 0.000

(2)

Source: See Appendix C Note: Robust standard errors in parentheses. ∗ significant at 10%; ∗∗ significant at 5%; ∗∗∗ significant at 1% (one-tailed).

N Adjusted R2 F Prob. > F

Environmental group strength Constant

Business share of R&D Neocorporatism

Manufacturing share of GDP Trade dependence

Economic voting

Union density

EU member

Left-libertarian vote

Majoritarianism

Per capita GDP

(1)

Table 6.1 Effects of business power resources on environmental regime strength

130 Structure, information and environmental regulation of Norway and that of Ireland) is associated with a weakening of the environmental regulatory regime by a half point on the seven-point scale. While the (wrong) bivariate association between business’ R&D share and environmental regulation disappears in the presence of statistical controls, neocorporatism and environmental movement strength continue to display the expected positive effects on regulatory strength. A unit increase on Siaroff’s scale of economic integration is associated with a half unit increase on the seven-point scale of regulatory stringency. However, the effect of the environmental movement is very small. To achieve a half-point improvement in regulatory stringency, a country would have to experience a boost in environmental group membership of over 40 percentage points, which would represent a gain in movement strength from no membership at all to that reported for the Netherlands. Most importantly, and remarkably given the small number of observations, the effects of trade dependence, neocorporatism and environmental group strength persist when the three measures are included simultaneously (model (9)). Thus, the constraining effects of economic variables on policymakers’ capacities to regulate environmentally harmful behaviour do not appear as rigid as is sometimes claimed. Neither the extent to which voters punish politicians for economic slumps nor the importance of the energy-intensive manufacturing sector is related to the strength of national environmental regulatory regimes. By contrast, two of the three indicators of informational constraints contribute significantly to explaining differences in countries’ environmental regimes. As neocorporatism facilitates the flow of information among the parties to tripartite institutions, information asymmetry is lower in the more integrated and co-ordinated market economies, enabling policymakers here to enact stricter environmental regulation without unfounded fears of damaging business confidence. At the same time, neocorporatism increases the potential costs attached to overstating negative inducement effects, providing business actors in integrated and co-ordinated market economies with incentives to lobby more truthfully. The idea that business can achieve favourable political outcomes by virtue of its informational privileges also receives support through the positive effects of the strength of environmental groups – that is, of countervailing lobbyists – on environmental regime strength. It was argued above that any effects of information asymmetry on environmental regulation should be stronger if potential policy costs are high. This implies an interactive effect of informational and economic–structural variables. To test this interaction, multiplicatory terms have been created of the variables that proved influential in the additive models in Table 6.1. Trade dependence has been dichotomized along the median; highly trade-dependent countries are those whose trade share of GDP is above 75 per cent. The expectation is that the effects of neocorporatism and environmental group strength, respectively, are stronger in those countries that are highly dependent on international trade. The results are displayed in Table 6.2. The interactive expectation bears out for environmental group strength. The coefficient for this variable increases slightly in size from 0.012 (0.014 in the combined additive model (9)) to 0.017 for the group of highly trade dependent countries only. By contrast, neocorporatism does not affect

Structure, information and environmental regulation 131 Table 6.2 Interactive effects of business power resources on environmental regime strength (1) Per capita GDP Majoritarianism Left-libertarian vote EU member Trade dependence Neocorporatism Environmental group strength Trade dependence × Environmental group strength Trade dependence × Neocorporatism Constant N Adjusted R2 F Prob. > F

(2)

0.109 0.119 (0.032)∗∗∗ (0.048)∗∗ −0.272 −0.054 (0.111)∗∗ (0.079) 0.079 0.096 (0.020)∗∗∗ (0.030)∗∗∗ 0.206 0.322 (0.191) (0.207)∗ −0.005 −0.007 (0.002)∗∗∗ (0.002)∗∗∗ 0.361 (0.136)∗∗∗ 0.015 (0.007)∗∗ 0.017 (0.004)∗∗∗ 0.061 (0.084) 2.009 2.696 (0.905)∗∗∗ (1.245)∗∗ 22 22 0.77 0.65 28.99 20.08 0.000 0.000

Source: See Appendix C Note: Robust standard errors in parentheses. ∗ significant at 10%; ∗∗∗ significant at 1% (one-tailed).

∗∗ significant

at 5%;

environmental regulation interactively with trade dependence. This might reflect a true absence of an interactive effect. But with the degrees of freedom available for estimation severely stretched, this non-finding might also be due to overfitting. These results are also confirmed by jack-knife analysis reported in Appendix C.

Discussion In conceptualizing the electoral–economic constraint as the magnitude of economic voting, this chapter has explicitly tapped a causal mechanism that is frequently assumed in the literature but has hitherto escaped empirical investigation. But just like the importance of the manufacturing sector for national wealth creation, this key indicator of the structural source of business political power fails to explain cross-country variation in the strength of environmental regulation after standard economic, political and institutional determinants of regulation have been taken into account. This suggests that business influence

132 Structure, information and environmental regulation over politics might not take on the form implied by prominent theories of business structural dominance. However, the negative effects of trade dependence on environmental regime strength suggest that integration in global trade provides governments with incentives to keep domestic regulatory costs low, lending some modest support to the notion of structural dependence. The role of incomplete information – tested here for the first time since that variable started to feature prominently in the theoretical literature on special-interest politics 15 years ago – fares better overall. The idea that business can achieve favourable political outcomes by virtue of its informational privileges receives support through the significant positive effect of environmental group strength, implying that business power derived from private information on policy effects decreases with the presence of counteractive lobbyists in the form of environmental groups. This variable becomes even more important in interaction with trade dependence, suggesting that business informational clout, as well as countervailing factors, increases when the potential negative inducement effects over which business claims privileged information are high. Moreover, the strong positive effect of neocorporatism suggests that the institutionalized interaction between policymakers and business in neocorporatist systems increases the reputation constraints for business, inducing firms to lobby more truthfully and thereby reducing the leverage of information asymmetry. These results are to be viewed with caution. The indicators of business power used here are proxies for variables that are extremely difficult to measure directly. Moreover, the number of observations available for estimation is small. With the bare minimum degrees of freedom available for model estimation – and even less in model (9) of Table 6.1 and the interactive models in Table 6.2, there is a risk of overfitting that compromises the generalizability of these results. But as yet no better alternatives to systematically investigate the effects of business power on environmental policy performance, or indeed any other area of regulation, have been put forward in the literature. The analysis presented in this chapter should thus be understood as a first step towards the empirical assessment of information asymmetry and structural power in government–business relations and the importance of these factors for actual policymaking. Furthermore, the finding that the informational factors of business power perform better than economic–structural ones in explaining environmental regulatory stringency does not mean that objective structural constraints are irrelevant. Regardless of any business influence over policymaking, society regularly faces real trade-offs between environmental regulation and other goals. Money spent on the environment is not available in the short term for other social purposes such as education, healthcare or infrastructural investment. Not even the promise of a blissful future of sustainable development can entirely eliminate such choices, Moreover, the trade-offs often become starker as environmental policy deals with marginal gains in environmental quality (Tietenberg 1994). Of course, many environmentalists would object on ethical grounds to evaluating environmental policy choices in terms of economic cost–benefit analyses. At any rate might one question whether any economic calculus measures and weighs adequately all relevant costs

Structure, information and environmental regulation 133 and benefits, since the benefits are typically harder to estimate and society may heavily discount long-term benefits. But business actors regularly complain that policymakers overregulate their activities in an overly zealous effort to reduce risks to public health and the environment, thereby imposing unnecessary burdens on individual firms and ultimately on the economy as a whole. Likewise, workers in affected industries and the communities in which they live often blame environmental policies for threatening their livelihood (Kraft 1996: 157). There are winners and losers in policy choices, and each side will take an active interest in evaluations that may maintain, alter or eliminate programmes affecting them (ibid.: 159). In these situations, policymakers engage in a cost–benefit calculus with respect to the support they enjoy from constituencies, the electorate and business. When business can make credible claims that its interests are aligned with those of the electorate, its preferences will weigh heavier in the utility functions of policymakers.

7

Conclusions Information, structure and the prospects for democracy under capitalism

Informational models of special interest politics have been around for some time now. In the economics literature, informational asymmetries are frequently identified as obstacles to efficient regulation and rent-seeking opportunities for special interests. In the political science literature, however, information has predominantly been treated as a tradable asset – an entry ticket for special interests seeking access to policymakers – while little attention has been paid to the content of that information. Only relatively few authors have also analysed the strategic use of information in political science analyses of lobbying, and very rarely have their models been subjected to empirical examination. There are also good reasons to suspect that business is qualitatively different from non-business interest groups when it comes to using knowledge and expertise strategically in politics. This book has therefore cast a closer look at the role of information and its asymmetric supply for capitalists’ ability to influence the policy process.

Information asymmetry in the repertoire of political resources To get to the ground of what might be special about business, explanations of business power in terms of corporate political action, political institutions, elite networks, ideology and structural dominance were explored. The review of pluralist, neocorporatist and elite approaches suggested that they yield important insights into political realities but ultimately fail to provide satisfactory explanations of the sources and limits of business political power. And while these approaches acknowledge the role of information in special interest politics, they reduce it to that of a tradable asset (pluralism), the black-boxed content of concertation and co-ordination (neocorporatism) or an entry ticket to policy networks. The review of the ideological dimension of business influence over public policy has led to a rejection of the notion that business can successfully and on a large scale manipulate citizens’ and policymakers’ preferences to bring them in line with its own interests. While this rejection is justified on conceptual and ontological grounds, the ideological dimension hints at an important factor in the analysis of business political power: the ability of business interests to manipulate other actors’ beliefs about facts, and thereby their ability to connect means to ends in a rational manner.

Conclusions 135 Chapter 2 also revisited claims that capitalists’ privilege over investment and allocation of society’s productive assets enables them to constrain the set of feasible options available to policymakers and, by implication, to citizens who mandate elected officials to devise and implement policy on their behalf. While these structural constraints are important parameters for the policymaking capacities of elected officials, their explanatory value in terms of power is limited, as both the ubiquity of business political action and observations of business’ political defeat reveal. Moreover, referring to systematic advantages arising from these constraints as power invites serious conceptual problems. It was suggested that these privileges and their effects on citizens and policymakers are better conceptualized as problems of structural dominance and subordination. Nevertheless, the structural theories of business dominance produced crucial insights into the need for policymakers to observe economic parameters and anticipate business reaction to planned policy. Carville’s 1992 epiphany that the economy matters for political success was neither new nor unintuitive. Harold Wilson stated a quarter of a century earlier that ‘[a]ll political history shows that the standing of the government and its ability to hold the confidence of the electorate at a general election depends on the success of its economic policy’ (Financial Times, 9 March 1968), and we can safely assume that his was not the first utterance of this link between politics and economics. But what has been missing from the structural power theories is an account of the political actions that connect strategic actors to the structural constraints under which they operate; political outcomes that are the results of – often strategic – political decisions. If structural variables are supposed to play a role in producing these outcomes, it has to be shown how they relate to political action. The informational–structural account of business power claims to provide that explanatory link. Treating the different approaches to the study of business and politics as complementary rather than competing, Ch. 3 set out to explain business political power as contingent on the interaction of informational and structural constraints on public policymaking. Equipped with privileged access to a range of policy-relevant information, business actors use this information as an important power resource in conjunction with the other resources reviewed in Ch. 2; for example, money and organization, membership in elite networks and, above all, a structurally privileged position in the policy process. Their informational privileges enable capitalists to obtain desired political outcomes through lobbying even if alternative policies would be preferred by fully informed voters and policymakers. While theories of the structural dependence of the state on capital stipulate that business gets what it wants because policies it does not want would also be detrimental to voters and policymakers, the informational–structural model of business power asserts that business actors succeed in politics if they can make policymakers believe this to be the case – even when in fact it is not. Because policymakers are aware of both their relative informational disadvantage and business incentives for bending the truth sometimes, the use of political resources in lobbying can quickly become a delicate undertaking, requiring business actors to devise careful strategies involving a mixture of boldness and restraint.

136 Conclusions While informational and structural constraints can be strategically utilized by business actors in their political actions, their use is itself subject to reputational constraints, such as those resulting from business participation in policy communities or neocorporatist institutions. This logic has been expressed in a signalling game that predicts business actors’ ability to effect political outcomes by reference to their capacity to manipulate the beliefs of policymakers regarding the consequences of policy. The central parameters determining this ability are a policy’s expected costs and benefits according to business and policymakers, the costs of lobbying and the levels of credibility and trust that characterize the relationship between business actors and policymakers in the relevant policy area. Because the choice of business strategy as well as its likelihood of success are thus related to the central parameters in the informational–structural model, behavioural implications can be deduced that are suitable for empirical investigation. In a progression of levels of analysis, from detailed case studies to aggregatelevel, structural data, empirical tests have been carried out to assess the extent to which the strategic use of private information matters for business power over political decisionmaking. Analytical narratives of two case studies employed detailed evidence from real-world lobbying scenarios that are consistent with two major equilibrium outcomes of the signalling game: one in which a business lobbyist tells the truth about the state of the world, and another in which high policy costs for business generate incentives for lying. The subsequent micro-quantitative analysis provided statistical tests of individual-level implications of the theory. Because lobbyists must build the success chances of their political strategies into their decision rules, business actors have been shown to lobby left-leaning parties with reduced frequency. As untrue claims are more likely to be exposed when made to ‘anti-business’ parties as opposed to ‘pro-business’ parties, the former are lobbied more carefully. This contradicts the claim by Austen-Smith and Wright (1994) that interest groups lobby primarily policymakers who are opposed to their goals. The analysis has also shown that frequent contacts with business representatives lead policymakers to be receptive to the information conveyed by lobbying – something that works significantly less well for trade unions and public interest groups. This confirms the implication of the informational–structural model that the messages of business carry more clout than those emitted by non-business interest groups, even if all groups lobby at the same level of frequency, because only business actors enjoy a prerogative over decisions concerning important matters such as investment, production and pricing. The core claim of the informational–structural model of business power – that business’ informational advantage matters for political outcomes – has been tested in Ch. 6. While the expectation that relaxed economic–structural constraints allow policymakers to enact stricter environmental policies has survived the test only when the constraint is conceptualized as trade dependence, the central empirical implication of the informational model has found more support. Two of three alternative measures of business’ informational clout showed the hypothesized effects on policy. The positive effects of neocorporatism on environmental regime strength suggests that the institutionalized interaction between policymakers and business

Conclusions 137 in neocorporatist systems reduces both the degree and the leverage of asymmetric information by facilitating information flows between government actors and private interests and by inducing business actors to lobby more truthfully. While the existence of a relationship between neocorporatism and environmental governance is as such not new (Scruggs 1999), the causal mechanisms connecting the two variables were far from clear. Interest representation in neocorporatist systems is conducted by the most powerful interest groups, who can settle their conflicts at the expense of groups and issues that are less well organized and resourced (Offe 1981: 128–9). This casts doubt on the notion of a direct favourable effect of neocorporatism on environmental policymaking, leading Neumayer (2003) to suggests that it might be a myth to believe that neocorporatism is good for the environment. The findings generated in Ch. 6 confirm the positive link between neocorporatism and environmental governance established by earlier research. It was argued that the causal link between neocorporatist institutions and environmental policy consists in these institutions’ ability to reduce both the magnitude and the efficacy of information asymmetries between producers and policymakers. This function of neocorporatism has been confirmed by the elite survey analysis in Ch. 5, which provided individual-level evidence in favour of the claim that neocorporatist institutions reduce the likelihood on (untruthful) lobbying. Lastly, the effect of environmental groups on the strength of the regulatory regime corroborate the claim that business power derived from private information on policy costs decreases with the strength of counteractive lobbyists. Again, a positive effect of environmental group strength on environmental governance is neither new nor unintuitive. But information provided by environmental groups affects government behaviour in two principal ways: by raising public awareness of environmental problems, and by providing information on abatement and compliance costs independent from industry. The first mechanism is the stuff of pluralist accounts of pressure politics, which highlight the importance of lobbying campaigns and mobilization efforts to promote an issue onto the political agenda. The second mechanism is an upshot of the informational–structural model of business power and concerns the fate of an issue once it has made its way onto the agenda. This aspect has so far received only scant attention in the literature on environmental policymaking. The findings presented in this book suggest that, overall, informational accounts of business power perform at least as well as organizational and structural explanations in predicting both business political behaviour and policy outcomes. Informational privilege is not power. But just like organizational strength and structural privilege, private information on policy effects is a useful resource for the exercise of political power by business actors, to the extent that if business actors can successfully utilize this resource in their political strategies, they have power. How does this result relate to the main theoretical approaches to the study of business and politics? While network and elite analyses assume that it is who you know that determines whether or not you get what you want in politics, the informational–structural account of business power claims that political success is a function of what you know. Critics of the network approach have suggested

138 Conclusions that network formation should be analysed as a dependent variable (Dowding 1995; König and Bräuninger 1998). The results of the preceding analysis suggests that the quantity and quality of an actor’s private information might go a long way to explaining why that actor is – or is not – part of a policy network. The informational–structural view also highlights another aspect of policy networks and communities that has hitherto been largely overlooked. Membership in policy networks or privileged positions in the institutions of co-ordinated market economies not only bestows political influence on business actors, it also subjects them to reputational constraints on their ability to influence policy outcomes. The informational–structural view of business power shares with pluralist theories of corporate political action the claim that the chances of political success increase with the amount of resources – money, skills and time – invested into a political campaign. But while pluralists view the link between business political action and outcomes as a quid pro quo relationship – the more contributions business makes, the more political benefits it can purchase – the informational– structural view treats costly political action primarily as a means to confer credibility on policy-related messages that might otherwise be dismissed by policymakers as self-serving babble. This implies a non-linear relationship between lobbying effort and political gain. The informational analysis of business power also opens up a new perspective on the role of group size and collective action problems in the study of interest groups. The informational view concurs with Olson’s (1965) expectation that business has advantages in politics by virtue of being a small group compared to labour or consumers. But while for Olson this is the result of business’ ability to overcome free-rider problems with relatively greater ease than the larger groups, the analysis of lobbying as strategic information transmission suggests that business is more likely to succeed because its members find it easier to co-ordinate their speech acts, thereby minimizing policymakers’ access to independent information or deviating second opinions. As ideal-typical categorizations, the distinctions between the various approaches to explaining business power made here are, of course, somewhat artificial. In the real world of business and politics, the different aspects of political reality that are isolated by the respective theories, including the informational perspective explored here, are often congruent. And business actors frequently utilize multiple avenues for exerting political power simultaneously, with the expectation that the strategies emphasized by the different theoretical approaches mutually reinforce each other. An example of such multi-strategy political action is a lobbying campaign directed at the British government by British American Tobacco (BAT). In a dual effort to dissuade the government from raising tobacco tax and to avert an inquiry into allegations that the firm was colluding with cigarette smugglers, the chief executive of the world’s second biggest tobacco firm, Martin Broughton, lobbied Tony Blair and the trade secretary, Stephen Byers, by means of classic backroom contacts (The Guardian, 27 October 2004). At the same time, the company made use of personnel resources from ‘revolving door’ strategies ensuring that two former senior officials at the Department of Trade and Industry are on BAT’s payroll. These hirings were successfully used to approach their

Conclusions 139 former departmental colleagues. Prior to this strategy, BAT’s requests for private meetings with government officials had in fact been declined. Even previous attempts at playing the ‘revolving door’ card involving the former Conservative chancellor Kenneth Clarke, who was deputy chairman of BAT from 1998 to 2005, had remained unsuccessful – highlighting the role of party affiliation in revolving door strategies as well as in lobbying more generally. Lastly, Broughton’s access to Blair and Byers was significantly aided by the fact that Broughton figured officially not as the BAT chief executive, but as a representative for the Multinational Chairmen’s Group, an exclusive club whose members also include the bosses of BP, Diageo, Shell, Unilever and Vodafone. This role elevated Broughton onto a shortlist of people invited to breakfast with the British prime minister, highlighting the importance of business’ ability to overcome collective action problems even across sectors. Once communications between the industry and government had been switched to a friendlier mode, Broughton was able to provide information on the subject matters of smuggling and tobacco taxation. To dissuade the government from its plan of raising taxes on cigarettes, Broughton predicted negative inducement effects, claiming that the raise would encourage increased smuggling. In Broughton’s words, ‘[t]he chosen tax policy contains within it the seeds of its own destruction’ (The Guardian, 27 October 2004), highlighting the importance of privileged insights into the mechanisms responsible for a policy’s expected costs and benefits. This episode illustrates how in the real world of lobbying any one aspect of business political resources will hardly be sufficient on its own to explain political outcomes. Above all else, however, it exemplifies how information, structure and political action combine to produce desirable outcomes for business. This logic of multiple causation means that business’ grip on public policy is not as rigid as structural accounts of business dominance imply. Where strategies have to be employed to make gains, there is always room for losses to be incurred. The last section therefore looks at the implications the informational–structural account of business power has for the state of democracy in capitalist countries.

Outlook: Can policymakers’ information problems be alleviated? The findings presented in this book strongly suggest that democracy in capitalist society is inherently vulnerable to a pro-business policy bias that allows business actors to prevail when their preferences diverge from those of other groups. The point is sometimes made that, in the age of institutional investors, ownership of large corporations is so widely dispersed among actors, including many ordinary citizens through saving schemes, pension funds and employee share options, that the distinction between owners and non-owners becomes increasingly problematic. The argument could be made, therefore, that if big business dominates politics by virtue of influencing policymakers and channelling investment to wherever its returns are maximized, then really what we are talking about is the structural dominance by a multitude of citizens who

140 Conclusions have stakes in these schemes. This argument is flawed for two reasons. First of all, it overstates the degree of dispersal of capital in Western nations. In most countries, the bulk of assets continue to be owned by a small elite, which, if anything, is accumulating more and at a faster pace than the population as a whole. Moreover, the argument of shareholder democracy underestimates the difficulties for small-time investors in communicating their interests to fund managers, not to mention to corporate directors and managers (Mitchell 1997). But even if assets were as widely distributed as some suggest and small-time shareholders had improved mechanisms at their disposal to hold fund managers and corporate executives accountable, we would still be dealing with an invasion of profit-maximizing principles of decisionmaking into an area where, according to widely held understandings of representative democracy, decisions should be made through effective democratic procedures; that is, non-market modes of decisionmaking. In a democracy, citizens’ entitlements to participate in these decisionmaking procedures derive not from asset ownership but from citizenship. To restore democratic legitimacy, therefore, the pro-business policy bias needs to be rectified. But because the roots of this bias lie in the combination of structural privilege and information asymmetry in favour of business interests, political reforms aimed at restricting party and campaign contributions, corruption, lobbying or the many other channels through which business actors exert influence over political processes and outcomes will inevitably be limited in their effectiveness. As long as the institution of private property in the means of production exists along the lines we are familiar with today, information gaps between political and economic leaders will continue to empower the latter. Business predictions about policy effects are predictions made by one set of actors about their own future behaviour. The leaders of capitalist firms, by definition, make their own business decisions; and what they intend to do is private information unless and until they decide to disclose their intentions to others. Others, including policymakers, can make informed and often rather accurate guesses about a firm’s intrinsic efficiency, its costs and its likely future actions and performance. But any guess about its management’s intentions is only ever just that – a guess. The policy-relevant information at the centre of this book is privately owned in the literal and legal sense of that term. A protected sphere of informational privacy is part of the capitalist property right over productive assets, which usually involves very little by way of transparency. Outsiders do not usually have legitimate claims to a firm’s internal data, for the simple reason that it is quite literally ‘not their business’. It is generally assumed that capitalism cannot function without this protected informational realm, and the legal protection of private information about production and production costs is a constituent institution of capitalism. Transgressions of informational property rights by other private actors are acts of commercial or industrial espionage and as such heavily punishable offences. Their violation by the state is tantamount to partial expropriation. The informational privilege can therefore not be taken away from business without fundamentally altering the nature of the capitalist institution of private property.

Conclusions 141 But as this book has shown, informational advantage does not guarantee business’ victory in politics. Structural determinists could of course still argue that even policies which are disliked by business interests are actually beneficial for them, given their ‘real interests’, or ‘in the longer term’, if they become enacted by a well-informed or resolute policymaker. This idea underlies Poulantzas’ theory of the relative autonomy of the state in capitalism, according to which the state ‘takes charge, as it were, of the bourgeoisie’s political interests and realizes the function of political hegemony which the bourgeoisie is unable to achieve’ (Poulantzas 1973: 284). Much empirical research on the relationship between environmental regulation and capitalist performance can indeed be interpreted to be supportive of this view. Studies of the effects of environmental regulation on productivity have suggested that while the policies in question are almost universally rejected by business, they frequently turn out to enhance productivity and profitability of the affected industries if they are enacted in spite of business opposition (Berman and Bui 2001). The reason why businesses shun the implementation of productivityenhancing technology is that the short-term costs of upgrading plant equipment appear as too high a threshold to overcome voluntarily. In the terminology of the model formulated in Ch. 3, it is because the ‘mild’ policy costs (β) are too high even if the ‘severe’ policy costs (α) are low, or even negative as stricter environmental policy turns out to be of overall benefit to business. It is this notion – and the empirical evidence appearing to support it – of capitalists’ seeming inability to recognize their own ‘real interests’ and take the appropriate action, and instead having to rely on the government to force productivity-enhancing technology upon them against their will, that provides Poulantzas’ argument with such powerful appeal. At the same time it sheds a gloomy light on the Schumpeterian notion of the creative and daring entrepreneur. But, as was argued above, that position is entirely unfalsifiable. Moreover, it is ultimately not relevant to the problem of business political power at the centre of the present analysis. For even if government action overriding business preferences is still in business’ best interest, the action can only become possible because business power, defined in terms of the intentional notion of bringing about desired outcomes, is sufficiently low for business’ preferences to be overridden. Democratic policymakers can sometimes overcome business’ political privilege because information asymmetry is variable rather than constant. The availability of independent information can go a long way to alleviate the informational disadvantage of policymakers. This was illustrated by the case of small business banking regulation in Ch. 4. Moreover, even if the degree of information asymmetry was constant, the informational–structural account of business power allows for strategic interaction between business and policymakers, which can lead to the enactment of policies that maximally correspond to the preferences of voters and policymakers regardless of possible opposition by business. And lastly, institutional remedies are available that alleviate the extent to which information asymmetry can be exploited by special interests. In this way the informational– structural account of business power is fundamentally distinct from the structural mechanisms underlying business influence in theories of structural privilege or

142 Conclusions state dependence. While explanations of the structural dependence of the state on capital are essentially deterministic, the informational model of business political power brings political action back into the study of business and politics while simultaneously acknowledging the powerful constraints on public policymaking that exist in capitalist societies. Variable degrees of information asymmetry and the possibility of business defeat in policy struggles suggest at least five principle solutions to policymakers’ information problems. Two are actor-centred, while the other three are institutional in nature. The first actor-centred solution centres on the role of counteractive lobbyists. This is firmly embedded in the pluralist analytical apparatus and suggests a simple recommendation for non-business interests: groups that do not want policymakers’ beliefs about policy costs to be dominated by business advice should organize and become politically active to provide countervailing information. As suggested by the logic in Austen-Smith and Wright’s (1992; 1994) analysis and corroborated in Ch. 6, counteractive lobbyists can make a difference to policy outcomes. But there are limits to the efficacy of counteractive lobbying. As this book has demonstrated, if one lobbyist is a business actor making predictions about their own future behaviour, the situation is not quite as straightforward as counteractive lobbying models suggest. If an energy corporation’s predictions about policy costs are couched in terms of market loss, unemployment and economic downturn while the countervailing lobbyist is an environmental group questioning business’ cost predictions while highlighting the plight of polar bears, policymakers are likely to pay more attention to the business side. A second actor-centred solution builds on the availability of independent sources of information such as that contributing to the policymaker’s ability to override the banks’ forecasts in the case study in Ch. 4. Policymakers can invest in obtaining information independently from business interests. Sloof (1997) explores situations theoretically in which policymakers can acquire information independently of the lobbyist if they are willing to pay a price for the necessary investigation and verification of lobbyist’s claims. While the case of small business banking in the UK has addressed this possibility, it has treated the Treasury’s ability to obtain independent information as exogenous without problematizing the costs of becoming informed. Information costs need not be high, however. Binder (1985), for example, suggests that policymakers can obtain relatively low-cost information clues through monitoring firms’ dividends, which reveal information about past performance, or through the market value of the firm’s stock, which reflects expectations about future profits. And while firms may have incentives to emphasize their problems when talking to policymakers, they have at least equally high incentives to emphasize how well they are doing when communicating with shareholders. This is why the informativeness of signals can increase with the number of audiences (cf. Farrell and Gibbons 1989). These mechanisms can be difficult to use in practice, however, as stock prices cannot reveal information about subsidiaries of diversified companies (Kennan and Wilson 1993: 96). To address the information problem in their policy area, therefore, environmentalists have long demanded institutional solutions aimed at

Conclusions 143 forcing corporations to provide better public access to information, particularly scientific data in the form of ‘freedom of information’ and ‘right to know’ legislation (Paehlke 1988). While the institution of private property includes the protection of information held by owners, it is ultimately governments that determine the terms and extent of the privileges conveyed by property rights. Thus, a first institutional solution would consist in legislation requiring firms to disclose more information about their intrinsic productivity or efficiency more accurately. A recent prominent example of a far-reaching redefinition of informational property rights is the Sarbanes–Oxley Act in the US. Following a number of high-profile collapses including that of Enron, the act was passed in 2002 by the US Congress to enhance corporate governance and thereby restore public confidence. It has introduced significant changes in both management’s reporting responsibilities and the scope and nature of the responsibilities of auditors. Among other things, the act demands more detailed and timely disclosure of financial information. Such redefinitions of property rights do not go unchallenged by business. Since the passage of the Sarbanes–Oxley Act, the US business community has expressed substantial concerns about compliance costs. A survey of 224 public firms in July 2004 about the direct costs of complying with the information disclosure requirements found that the average first-year cost estimate is almost $3 million for roughly 26,000 hours of internal work and 5,000 hours of external work, as well as additional audit fees of $823,200 (Zhang 2007). However, there are more diverse opinions about how the act affects firms’ bottom line, with some authors suggesting that it is value-increasing (see Linck, Netter and Yang 2006 for an overview), while Zhang (2007) suggests that the act is overall value-decreasing. Whatever the truth about the act’s cost implications, the debate highlights the fact that regulations aimed at reducing information asymmetry are likely to be viewed as costly by business. And just like other policies, such measures are also likely to be subjected to lobbying that takes advantage of existing information asymmetries. Ultimately, passage of the Sarbanes–Oxley Act was aided by the exceptional blow to its reputation that American business incurred when the full scale of the Enron, Tyco and WorldCom scandals unfolded. Mitchell (1997: 174) distinguishes four dimensions to business’ legitimacy problems: the legal dimension, the ethical dimension, the economic or efficiency dimension and the traditional loyalty dimension. At least the first three of these dimensions were affected when the bill passed the House and Senate with overwhelming majorities, suggesting an exceptional historical context for business regulation. While the above discussion has focused on the relationship between business actors and policymakers, it is not difficult to see how business informational privilege also implies a knowledge advantage vis-à-vis voters. If voters had better insights into firms’ intrinsic efficiency and could gain more independent estimates of policy costs, they would be less likely to turn against incumbent politicians whose policies antagonize business. And of course, once voters are better informed about policy costs, such information will be easier available to policymakers as well. This suggests a second institutional solution involving various devices for direct and deliberative citizen participation. Proponents of increased direct

144 Conclusions democratic procedures argue that referendums can lead to a narrowing of the information gap between special interests, voters and democratic political institutions and decisionmaking bodies. Frey and Bohnet (1994) find a positive political impact of direct democracy on the quality of policy outputs. They explain these benefits by the higher intensity of discussion in the run-up to a referendum: the debate among citizens and between politicians and the electorate is strengthened, which leads to an overall increase in the level of publicly available information. As alternatives are evaluated by citizens rather than by elites only, the process of communication in direct democratic systems limits free-riding and rent-seeking by special interests. The third institutional solution flows directly from the findings presented in Chs 5 and 6. Policymakers operating in a neocorporatist environment are better positioned to avoid being fooled by informationally advantaged business lobbyists. But while this might be good news for policymakers (and citizens) in countries with neocorporatist systems of interest intermediation, it is difficult to derive policy recommendations for everybody else. While legislation requiring firms to disclose information pertaining to intrinsic efficiency and performance might meet with resistance from the business world, the Sarbanes–Oxley Act demonstrates that such resistance can be overcome. And while new direct democratic procedures are barely implemented overnight, the use of referendums and initiatives has increased steadily during the last quarter of the twentieth century (Craig, Kreppel and Kane 2001). But macro-institutional structures such as neocorporatist systems are hardly at the disposal of the government of the day. Relatively little is known to date just why and how neocorporatist systems have come into being in some countries but not others, although recent work suggests a link to electoral institutions and party systems (Martin and Swank 2007). At any rate, a glance at the failed experiment of implementing corporatist institutions in Britain during the 1960s and 1970s (Grant 1993: 30–1) suggests that institutional change of the required quality and scale would be a very challenging and lengthy endeavour. Thus, while solutions to policymakers’ information problems in democratic capitalism are not easily achieved, there is scope for reform-minded politicians and non-business interest groups to create a political landscape and institutional environment that makes it harder for business actors to realize excess gains at the expense of other actors’ political goals by virtue of their informational advantage. While five possible solutions have been sketched out, it is up to future work to explore their feasibility in greater detail. A number of analytical issues also remain for future research. First, the informational–structural account of business power analysed here assumes that business actors already have the information readily available that policymakers are lacking to assess policy effects. But in reality, much interest representation is done by industry associations on behalf of firms, even in predominantly pluralist systems or liberal market economies. These organizations have to collect data from their members and process these before they can use them to influence the policy process. This is costly work, and it requires that collective action problems of firms be overcome. Not only do

Conclusions 145 firms have good reasons not to share their private information with policymakers; they often are also wary to share information with other firms or with industry organizations. Moreover, even individual firms face situations in which they first have to investigate a policy’s expected effects before deciding whether to signal these to policymakers. To accommodate this more realistic setting, the model outlined in Ch. 3 would have to be extended by assuming that a special interest group is imperfectly informed but can acquire information by investigating the state of the world for a price. Second, while lobbying costs are exogenous in the present analysis, information about a lobbyist’s use-value to the policymaker might be revealed if the lobbyist or the politician is able to determine these costs endogenously. This would open up scenarios in which policymakers can set prices for access that reflect how important it is to make the right decision. Ainsworth (1993) suggests such price setting as a device for policymakers to control lobbyists and reduce the occurrence of untruthful signals. The problem in the context of business political activities, as this book has argued, is of course that material access costs are often trivial for business lobbyists, which is precisely the reason why mimicking genuine lobbyists is so frequently rational for them. But it would be left for future analysis to examine whether endogenous lobbying costs could lead to an overall increase of informativeness of lobbying. Finally, throughout the book, business lobbyists have been portrayed as pursuing a conservative strategy; that is, one whose goal is to avert a disliked policy change and preserve the status quo. While it is highly plausible to treat the avoidance of a future ‘bad’ as equivalent to the attainment of a future ‘good’, as long as both can be expressed in the same currency, there may be situations in which this assumption cannot be upheld. This would be the case for businesses seeking new property rights in areas such as telecommunications, where non-monopoly provision has only recently become conceivable as a result of technological change. Furthermore, whether a lobbyist seeks to bring about policy change or avert it may have implications for policymakers’ pledge costs. Politicians promise to do certain things and they frequently enumerate these in election manifestos and similar statements. But they do not usually promise to resist future, unforeseen demands for policy change from business. There is also room for improving the empirical analysis of informational– structural accounts of business power. Inspiration can be drawn from studies on pledge fulfilment. Early quantitative pledge studies were conducted by Page (1978) for the US and by Rose (1980) for the UK. Rose, for example, identifies specific commitments to act (‘we will cut unemployment’), thus excluding vague and general statements like ‘an economy based on more jobs’, in order to see whether the party in government subsequently passed appropriate legislation or took specific action to fulfil them. While Rose found that parties tend to stay away from ‘big’ pledges, elsewhere scholars have found that explicit policy pledges were made – and subsequently fulfilled – on quite specific and important matters (Thomson 1999). Indicators of business informational advantage could be used to predict relative frequencies of pledge redemption and reneging.

146 Conclusions Furthermore, simplified replications of the case studies in Ch. 4 could be implemented with numerical measurement and sufficient frequency to allow quantitative analysis. To that end, archival searches of news media could be carried out to compile a list of policy proposals that have cost implications for business. Perusing these sources, as well as consultation lists, parliamentary minutes and press releases of business associations, trade unions and public interest groups, could lead to the identification of the key political actors – policymakers, firms and business associations and public interest groups – that have an interest in the proposal. For each actor, data about their policy preferences, the action taken to affect the outcome of the policymaking process and the perceived costs and benefits attached to alternative political strategies could then be collected using survey questionnaires or interviews. This book has only kick-started research into the important informational dimension of business political power. More work lies ahead to put the findings generated here on firmer empirical and theoretical ground.

Appendix A Interviews and personal communications

Günter Dunschen Email correspondence, 14 and 18 June 2004 Technology and Environment Officer at the Wiesbaden Chamber of Trade

Karin Hagemann Email correspondence, 27 June 2004 Former Member of the Hessian parliament (1991–7) and Economic Policy Speaker for the Green Party parliamentary faction

Hans-Jürgen Hielscher Interviewed by telephone on 23 June 2004 Former Member of the Hessian parliament (1987–96), Parliamentary Secretary and Environmental Policy Speaker for the Free Democratic Party (FDP)

Ciaran Higgins Interviewed in Dublin on 4 October 2002 Executive at the Irish Software Association and Consumer Electronics Distributors Association

Martin Lord Interviewed by telephone on 27 April 2004 Head of Government Relations at HSBC and Senior Manager in the Strategic Development Unit

Dr Rainer Nolte Interviewed by telephone on 29 April 2004 Managing Director of ECOTEC Institut für chemischtechnische und ökonomische Forschung und Beratung, Munich

148 Interviews and personal communications

Nick Palmer, MP Email correspondence, 28 April 2004 Member of the UK House of Commons (1997–) and former Member of the Treasury Select Committee

Marc Schnerr Interviewed in Berlin on 2 September 2002 Press Officer at the German Hotel and Restaurant Association (DEHOGA – Deutscher Hotel- und Gaststättenverband) (2000–4)

Appendix B

Coding and summary statistics of variables used in Ch. 5

Main data source Survey data on members of parliament is taken from the European Members of Parliament Study 1996. This study was co-ordinated by Bernhard Wessels (Social Science Research Centre, Berlin) and Jacques Thomassen (University of Twente) and consists of surveys in eleven parliaments of member countries of the EU, carried out by country teams. The national studies were administered as mail surveys. Fieldwork was carried out between April 1996 and July 1997. Representativity has been tested for party, gender and age. The data are available from the Zentralarchiv für empirische Sozialforschung, Cologne, Germany (Study No. 3079). For more details see Katz and Wessels (1999).

Table B.1 Coding of variables used in Ch. 5 Variable name

Question wording

Coding

Lobbyist influence over political decisionmaking

‘How much do you take the opinion of [lobbyists] into account when making political decisions?’ ‘How often do you have contact with the following interest groups?’ (industry, trade and commerce or banking and insurance) ‘How often do you have contact with the following interest groups?’ (trade unions)

Seven-point scale, ranging from 1 (‘very much’) to 7 (‘very little’)

Frequent business contact

Frequent trade union contact

Frequent public interest group contact

‘How often do you have contact with the following interest groups?’ (Consumer associations, environmental organizations)

Coded 1 if respondent ticked ‘at least once a month’ or ‘at least once a week’; 0 otherwise Coded 1 if respondent ticked ‘at least once a month’ or ‘at least once a week’; 0 otherwise Coded 1 if respondent ticked ‘at least once a month’ or ‘at least once a week’; 0 otherwise (continued)

150 Coding and summary statistics of variables used in Ch. 5 Table B.1 cont’d Variable name

Question wording

Pro-business party

Parties were classified into pro-business and other parties based on the party families defined by the Zentrum für Europäische Umfrageanalysen und Studien (ZEUS) at the University of Mannheim. Parties were coded as 1 (‘pro-business’) if they belong to either Christian Democrats, Liberals or Conservatives, and ‘0’ if otherwise. ZEUS coding: 1 Social Democrats 2 Christian Democrats 3 Liberals 4 Radical Left 5 Conservatives 6 Greens 7 Regional 8 Nationalists Coded left (=1) if ‘In political matters some people self-placement between talk about “left” and “right”. 1 and 4, 0 otherwise; Where would you place yourself right (=1) if on the following scale?’ self-placement between Ten-point scale, from 1(=left) 7 and 10, 0 otherwise to 10 (=right) ‘Here is a list of political issues. Respondents were coded [Prompt list.] Which of these as environmentalists issues do you consider the most (=1) if they named important, which comes second, ‘protection of the and which comes third?’ environment’ as the first or second most important issue ‘Do you hold or have you ever Coded 1 if MP answered held office in [a business ‘yes, at the moment’ or organization]?’ ‘yes, in the past’; 0 otherwise Coded 1 if MP answered ‘Please indicate your occupation ‘yes’ to any of these immediately prior to your first categories; 0 otherwise election to the parliament, using the following list (business employer, up to 10 employees; business employer, between 11 and 50 employees; business employer, with more than 50 employees)

Left–right self placement

Environmentalist

Official in business organization

Capitalist

Coding

Coding and summary statistics of variables used in Ch. 5 151 Table B.1 cont’d Variable name

Question wording

Coding of additional vriables: Variable name Coding Government party Coded 1 if the MP’s parliamentary party was part of a government coalition or was the only governing party at the time the survey was conducted in the respective country

Neocorporatism

Pluralism

Coded 1 (= neocorporatism) if one-half SD above the mean or higher on Siaroff ’s integration scale; 0 otherwise (X¯ = 3.50; σX = 0.49) Coded 1 (= pluralism) if one-half SD below the mean or lower on Siaroff ’s integration scale; 0 otherwise (X¯ = 3.50; σX = 0.49)

Coding

Data source Roberto Ortiz de Zárate’s European Governments dataset (© Copyright ZPC, Roberto Ortiz de Zárate, 1996–2004). Available from: [Accessed 14 July, 2004] Siaroff 1999

Siaroff 1999

Appendix C Coding and summary statistics of variables used in Ch. 6

Table C.1 Coding of variables used in Ch. 6 Variable

Coding

Data sources

Stringency of environmental regulations

‘The stringency of overall environmental regulation in your country is (1 = lax compared to most other countries, 7 = among the world’s most stringent)’ Union membership as a percentage of the non-agricultural workforce (1995)

World Economic Forum 2002

Union density

Neocorporatism Economic voting

Manufacturing share of GDP

Integration scores for mid-1990s Logit coefficient of perception of economic performance (‘Would you say that over the past twelve months, the state of the economy in [country] has gotten better, stayed about the same or gotten worse?’) on vote choice for incumbent

ILO (1997) World Labor Report 1997–98: Industrial Relations, Democracy and Social Stability, Geneva: International Labor Office; Figures for Denmark, Iceland, Italy, Spain, Sweden and Switzerland are for 1994; Bulgaria, Canada, Germany and Ireland: 1993; Mexico: 1991 Siaroff 1999, Table 5 Comparative Study of Electoral Systems (CSES), module I

United Nations Conference on Trade and Development (2004) UNCTAD Handbook of Statistics, Geneva: United Nations

Coding and summary statistics of variables used in Ch. 6 153 Table C.1 cont’d Variable

Coding

Data sources

Trade dependence

Total trade (exports plus imports) divided by GDP Percentage of gross domestic expenditure on research and development financed by industry Percentage of people stating membership or activism in conservation, environment, ecology or animal rights groups

Penn World Table 6.1

Business share of R&D

Environmental group strength

Per capita GDP (in million USD)

Left-libertarian vote

EU member

Per capita GDP in million USD adjusted for purchasing power across countries Percentage of votes at elections to lower chamber for left-libertarian parties

Coded 1 for countries that were EU members by 2000; 0 otherwise

Eurostat (Key indicators on EU policy – Structural indicators – Innovation and Research)

1999 World Values Survey. Data for Australia, Japan, Korea, Norway, Switzerland and the US are from 1995–7 wave; data for Canada and Iceland are for 1990–1 Penn World Table 6.1

Swank, D. (2002) Political Strength of Political Parties by Ideological Group in Capitalist Democracies. 21-Nation Pooled Time-Series Data Set. Available from [Accessed 19 July 2005.] Data for Iceland is from Electionworld.org, Elections around the world. Available at http://www.electionworld.org, [Accessed October 2005].; data for Israel from Psephos Adam Carr’s Election Archive; available at http://psephos.adam-carr.net/ [accessed October 2005]

154 Coding and summary statistics of variables used in Ch. 6 Table C.2 Summary statistics of variables used in Ch.6

Environmental regime stringency Per capita GDP (in 1000 USD) Majoritarian Left-libertarian vote EU member Union density Economic voting Manufacturing share of GDP Trade dependence Business share of R&D Neocorporatism Environmental group strength

N

Mean

Std. Dev.

Min

Max

23 23 23 23 23 23 21 23 23 22 23 22

5.87 24.69 0.32 4.18 0.61 32.84 0.33 17.48 79.59 53.13 3.22 7.59

0.81 4.18 1.01 3.76 – 19.79 0.20 4.57 37.35 14.30 0.99 9.07

3.74 16.33 −1.39 0.00 0.00 6.10 0.00 10.00 20.10 27.05 1.88 0.00

6.73 33.89 1.87 11.00 1.00 77.20 0.93 30.00 175.56 72.46 4.63 44.00

Appendix D Jack-knifed estimates of regression models in Ch. 6

2.055 (1.045)∗∗ 23 0.66 7.40 0.001

0.133 (0.041)∗∗∗ −0.051 (0.112) 0.090 (0.046)∗∗ 0.288 (0.280)

2.056 (1.061)∗∗ 23 0.65 5.81 0.001

0.131 (0.043)∗∗∗ −0.064 (0.126) 0.087 (0.051)∗∗ 0.281 (0.303) 0.002 (0.008)

(2)

2.080 (1.222)∗ 21 0.62 5.33 0.003

0.201 (1.016)

0.129 (0.042)∗∗∗ −0.041 (0.151) 0.090 (0.049)∗∗ 0.231 (0.305)

(3)

1.972 (1.306)∗ 23 0.65 6.09 0.001

0.008 (0.047)

0.131 (0.042)∗∗∗ −0.060 (0.130) 0.091 (0.048)∗∗ 0.264 (0.296)

(4)

2.295 (1.082)∗∗ 23 0.70 11.84 0.000

−0.005 (0.003)∗∗

0.134 (0.045)∗∗∗ −0.018 (0.104) 0.109 (0.043)∗∗ 0.406 (0.242)∗

(5)

(7)

2.406 (1.220)∗∗ 22 0.59 3.56 0.017

0.006 (0.009)

1.251 (0.991) 23 0.76 8.99 0.000

0.428 (0.171)∗∗

0.109 0.118 (0.048)∗∗ (0.036)∗∗∗ −0.046 −0.330 (0.111) (0.162)∗∗ 0.089 0.068 (0.042)∗∗ (0.041)∗ 0.163 0.232 (0.287) (0.273)

(6)

Source: See Appendix C. Note: Robust standard errors in parentheses. *significant at 10%; **significant at 5%; ***significant at 1% (one-tailed).

N Adjusted R2 F Prob. > F

Constant

Environmental group strength

Neocorporatism

Business share of R&D

Trade dependence

Manufacturing share of GDP

Economic voting

Union density

EU member

Left-libertarian vote

Majoritarianism

Per capita GDP

(1)

Table D.1 Jack-knifed estimates of effects of business power resources on environmental regime strength (9)

0.012 (0.008)∗ 2.436 (1.273)∗∗ 22 0.61 9.84 0.000

0.376 (0.170)∗∗ 0.014 (0.013) 1.822 (1.181)∗ 22 0.75 7.08 0.000

−0.005 (0.003)∗∗

0.119 0.111 (0.048)∗∗ (0.050)∗∗ −0.014 −0.243 (0.097) (0.125)∗∗ 0.080 0.076 (0.040)∗∗ (0.030)∗∗∗ 0.153 0.225 (0.277) (0.254)

(8)

Jack-knifed estimates of regression models in Ch. 6 157 Table D.2 Jack-knifed estimates of interactive effects of business power resources on environmental regime strength

Per capita GDP Majoritarianism Left-libertarian vote EU member Trade dependence Neocorporatism

(1)

(2)

0.109 (0.045)∗∗ −0.272 (0.134)∗∗ 0.079 (0.032)∗∗ 0.206 (0.267) −0.005 (0.002)∗∗ 0.361 (0.162)∗∗

0.119 (0.068)∗∗ −0.054 (0.102) 0.096 (0.043)∗∗ 0.322 (0.286) −0.007 (0.003)∗∗

Environmental group strength Trade dependence × Environmental group strength Trade dependence × Neocorporatism Constant N Adjusted R2 F Prob. > F

0.015 (0.012) 0.017 (0.006)∗∗∗ 2.009 (1.144)∗∗ 22 0.77 16.48 0.000

0.061 (0.103) 2.696 (1.695)∗ 22 0.65 7.91 0.000

Source: See Appendix C. Note: Robust standard errors in parentheses. *significant at 10%; **significant at 5%; ***significant at 1% (one-tailed).

Notes

1 Introduction 1 Sometimes a distinction is made between preferences – what people want – and preferences about preferences – what people want to want, given what they really want – in order to explain people’s choice under various constraints, risks and uncertainties. The respective concepts have been termed first- and second-order preferences (Sen 1977) or preferences and metapreferences (Hirschman 1984). For the bulk of the present analysis, these distinctions will not be necessary.

2 Groups, institutions, networks, ideology or structural dependence: What drives business power? 1 Of course, as Barnes points out, all power is invisible as such, including the firstdimensional, oppositional form of power at work in pluralist accounts of politics. The difference is that in the first dimension there is ready-made positive evidence in the form of the explicit preferences and outcomes of the decision-making process (Barnes 1993: 200–5). 2 In a new chapter added to the second edition of his book on power, Lukes deals with some of these critical points. But while he agrees with the need for providing evidence for deliberate manipulation, he also rejects the equation of power with purposive action that is undertaken in the present study and insists on the validity of functional explanation through ‘feedback mechanisms’ (Lukes 2005: 136). 3 Note that this concept of negative inducement is somewhat analogous, but not similar, to Wilson’s negative inducement in the context of bargaining scenarios. For Wilson (1961: 292), a negative inducement is given ‘if action in accordance with A’s intention, although no more attractive absolutely than before the change was made [by A], is nevertheless more attractive [to B] relative to the other possibilities that now exist’. That is to say, A generates negative inducements for B to choose X if he makes Y even more unattractive for B than X, even though X may already be quite unattractive in absolute terms. 4 Re-estimating Garrett’s counterfactual analysis using superior simulation techniques, King, Tomz and Wittenberg (2000) falsify his claim that left-leaning governments always outspend rightwing governments. However, their results do support Garrett’s finding that globalization has intensified the relationship between partisan politics and government spending (King, Tomz and Wittenberg 2000: 355–7).

Notes 159

3 An informational–structural model of business power 1 While cheap talk acts have been shown to affect the beliefs, and hence the actions, of their recipients under certain circumstances, positive incentives for truthful revelation through cheap talk are usually contingent on a very limited degree of divergence in the players’ preferences (Crawford and Sobel 1982) or on the presence of a second recipient whose interests are different from the first (Farrell and Gibbons 1989). An example of the latter scenario could involve a firm trying to convince a policymaker of its perilous situation in order to avert a pending costly policy while simultaneously convincing its shareholders or a potential buyer of its value and thriving performance. In the context of a general analysis of business political power, these conditions cannot always be assumed. 2 Bargaining is different from lobbying in one important aspect. In bargaining games, the cost imposed by a party rejecting an offer and thereby retaining the status quo always affects both parties’ payoffs negatively (Kennan and Wilson 1993: 52). This dimension is usually absent in last-minute lobbying against a specific policy. 3 While the canonical form signalling game involves two players, in principle there is no restriction on the number of actors and extensions involving three or more players have been analysed; cf. Grossman and Helpman 2001: 143–70. 4 Bayes’ rule requires an actor to update her belief about a hypothesis A in the light of new evidence B in such a way that her posterior belief P(A|B) is calculated by multiplying her prior belief P(A) by the likelihood P(B|A) that B will occur if A is true. Formally, P(B |A) P(A) P(A |B) = . P(B) 5 Note that it is the lobbying effort itself that is costly here, but not the gathering of information, as is assumed in some studies. While these studies have modelled information acquisition as being costly for the lobbyist (e.g. Austen-Smith 1993), it is assumed here that being informed is costless for business. Firms of course have to pay for at least some of the policy-relevant information they want to acquire. However, the rationale for the assumption of zero information-acquisition costs for firms in the present analysis is that many of the relevant expenses will have already been made by the time the firm decides whether or not to lobby against a particular policy. 6 Descriptions of all equilibria of this game, as well as formal proofs of the two sample equilibria, are reported in Bernhagen and Bräuninger (2005). To facilitate computation, Bernhagen and Bräuninger use a slightly different equilibrium concept – perfect Bayesian equilibrium. In the game presented here, with two possible types of a business lobbyist and each actor acting only once, the sets of perfect Bayesian equilibria and sequential equilibria coincide (Fudenberg and Tirole 1991).

4 Two real-world signalling games 1 At the time, no excessive pricing was found in Scotland or Northern Ireland.

5 Reputation and informatives in lobbying 1 The reported F statistic uses the survey design correction proposed by Rao and Scott (1984). 2 In the event, Intel did go ahead with the plant expansion plans, despite the fact that a proposed grant offer of E170 million from the Irish government had to be withdrawn

160 Notes as it would have violated EU restrictions on state aid. An environmental challenge to the investment also appears to have been averted. The Irish heritage trust, An Taisce, and local planning officers and residents have lodged objections to the project on the grounds of the impact on flood plains, failure to provide alternative modes of transport, inadequate waste water treatment and lack of an overall traffic-management strategy. Weighing off environmental versus economic benefits, the environment minister, Dick Roche, has come out strongly against An Taisce’s appeal, saying, ‘[t]he Intel investment in this country is critically important and any organization wishing to raise objections would want to consider very carefully the implications of what it’s doing’ (Sunday Times, 14 August 2005). 3 Relative frequencies of contact for opposition MPs were also obtained and found to be on the whole no different from those for government MPs. 4 Because the influence scale is ordinal and not normally distributed, a non-parametric test might be more appropriate. The alternatively used and reported Wilcoxon Mann-Whitney rank sum test also leads us to reject the null hypothesis that the influence ascribed to lobbyists is the same for MPs with frequent business contacts and those without.

6 Structure, information and environmental regulation 1 The countries are Australia, Austria, Belgium, Canada, Denmark, Finland, France, Germany, Greece, Iceland, Ireland, Israel, Italy, Japan, the Netherlands, New Zealand, Norway, Portugal, Spain, Sweden, Switzerland, the UK and the US. 2 Respondents were asked, ‘Would you say that over the past twelve months, the state of the economy in [country] has gotten better, stayed about the same or gotten worse?’ 3 After obtaining the economic vote estimates, figures for the two successive Spanish elections (1996 and 2000) were averaged to obtain the country-unique economic voting score. For countries with important regional electoral cleavages, separate estimates were first obtained for each region and subsequently combined into a single country score of economic voting, weighted by the regions’ population size. This procedure has been applied to Belgium 1999 (Flanders versus Wallonia), Canada 1997 (Quebec versus the Anglophone regions), Germany 1998 (East versus West) and Great Britain 1997 (Scotland versus England and Wales). Because for four elections the logit coefficients have a positive sign, the entire set of estimates was linearly transformed by subtracting the highest positive estimate (for the 1998 Danish election) from each data point. Finally, absolute values were obtained for the transformed scale. Missing data for France, Greece, Ireland, Italy and Switzerland have been imputed by assigning to these countries the arithmetical midpoint between the two nearest values identified from a country ranking of economic voting reported in Duch and Stevenson (2006). 4 An alternative measure of labour strength is the percentage of private sector workers covered by collective bargaining agreements. The two measures are congruent in some states (e.g. Sweden and the US), but differ in others, notably France. The use of trade collective bargaining coverage instead of union density does not alter the results reported here. 5 A replication dataset is available from the author at: http://www.abdn.ac.uk/∼pol209/ data.htm

References

Ackerman, F. and Massey, R. (2003) ‘The economics of phasing out PVC’, unpublished research paper. Somerville, M.A.: Global Development and Environment Institute, Tufts University. Ainsworth, S. (1993) ‘Regulating lobbyists and interest group influence’, Journal of Politics 55: 41–56. Akard, P.J. (1992) ‘Corporate mobilization and political power: the transformation of US economic policy in the 1970s’, American Sociological Review 57: 597–615. Alvarez, R.M., Nagler, J. and Willette, J.R. (2000) ‘Measuring the relative impact of issues and the economy in democratic elections’, Electoral Studies 19: 237–53. Arbeitsgemeinschaft PVC und Umwelt e.V. [No date]. ‘10 Jahre AgPU’. Austen-Smith, D. (1992) ‘Strategic models of talk in political decision making’, International Political Science Review 13: 45–58. Arrow, K.J. et al. (1996) ‘Is there a role for benefit–cost analysis in environmental, health, and safety regulation?’, Science 272: 221–2. Austen-Smith, D. (1993) ‘Information and influence: lobbying for agendas and votes’, American Journal of Political Science 37: 799–833. Austen-Smith, D. and Wright, J.R. (1992) ‘Competitive lobbying for legislators’ votes’, Social Choice and Welfare 9: 229–57. Austen-Smith, D. and Wright, J.R. (1994) ‘Counteractive lobbying’, American Journal of Political Science 38: 25–44. Bachrach, P. and Baratz, M.S. (1970) Power and Poverty: Theory and Practice, New York: Oxford University Press. Bank of Ireland et al. (2003) Branch Access Feasibility Study. Banks, J.S. (1991) Signaling Games in Political Science, Chur, CH: Harwood Academic Publishers. Banks, J.S. and Weingast, B.R. (1992) ‘The political control of bureaucracies under asymmetric information’, American Journal of Political Science 36: 509–24. Barnes, B. (1993) ‘Power’, in R. Bellamy (ed.) Theories and Concepts of Politics: An Introduction, Manchester: Manchester University Press. Baumgartner, F.R. and Leech, B.L. (1998) Basic Interests: The Importance of Groups in Politics and in Political Science, Princeton: Princeton University Press. Baumgartner, F.R. and Leech, B.L. (2001) ‘Issue niches and policy bandwagons: patterns of interest group involvement in national politics’, Journal of Politics 63: 1191–213. Bearden, J. and Mintz, B. (1987) ‘The structure of class cohesion: the corporate network and its dual’, in M.S. Mizruchi and M. Schwarz (eds) Intercorporate Relations, Cambridge: Cambridge University Press.

162 References Beck, U. (1992 [1986]) Risk Society: Toward a New Modernity, London: Sage. Becker, G.S. (1983) ‘A theory of competition among pressure groups for political influence’, Quarterly Journal of Economics 98: 371–400. Bentley, A.F. (1993 [1908]) ‘The process of government’, in J.J. Richardson (ed.) Pressure Groups, Oxford: Oxford University Press. Berger, P.L. (1992) ‘The uncertain triumph of democratic capitalism’, Journal of Democracy 3: 90–9. Berman, E. and Bui, L.T.M. (2001) ‘Environmental regulation and productivity: evidence from oil refineries’, Review of Economics and Statistics 83: 498–510. Bernhagen, P. and Bräuninger, T. (2005) ‘Structural power and public policy: a signaling model of business lobbying in democratic capitalism’, Political Studies 53: 43–64. Berry, J.M. (1997) The Interest Group Society, 3rd edn, New York: Longman. Bezdek, R.H. (1993) ‘Environment and economy: what’s the bottom line?’, Environment 35: 1004–18. Binder, J.J. (1985) ‘Measuring the effects of regulation with stock price data’, Rand Journal of Economics 16: 167–83. Birnbaum, P.H. (1984) ‘The choice of strategic alternatives under increasing regulation in high technology companies’, Academy of Management Journal 27: 489–510. Bischoff, I. (2003) ‘Determinants of the increase in the number of interest groups in Western democracies: theoretical considerations and evidence from 21 OECD countries’, Public Choice 114: 197–218. Black, D. (1948) ‘On the rationale of group decision-making’, Journal of Political Economy 56: 23–34. Block, F. (1977) ‘The ruling class does not rule’, Socialist Revolution 33: 6–28. Bonardi, J.–P., Hillman, A. and Keim, G. (2005) ‘The attractiveness of political markets: implications for firm strategy’, Academy of Management Review 30: 397–413. Börzel, T.A. (1998) ‘Organising Babylon: on the different conceptions of policy networks’, Public Administration 76: 253–73. Bowles, S. and Gintis, H. (1986) Democracy and Capitalism: Property, Community, and the Contradictions of Modern Social Thought, London: Routledge & Kegan Paul. Bresnahan, T. and Reiss, P. (1991) ‘Econometric models of discrete games’, Journal of Econometrics 48: 57–81. Broscheid, A. (2006) ‘Public choice models of business lobbying: distributional and informational accounts’, in D. Coen and W. Grant (eds) Business and Government: Methods and Practice, Opladen: Barbara Budrich Publishers. Bündnis 90 / Die Grünen (1997) Musterantrag zur PVC-Substitution mit Erläuterungen, available at: [Accessed 23 August 2007]. Busch, W. (1975 [1865]) Max and Maurice: A Moral Tale in Seven Parts, London: AbelardSchuman. Cairncross, F. (1992) Costing the Earth, Boston: Harvard Business School Press. Caldeira, G.A., Hojnacki, M. and Wright, J.R. (2000) ‘The lobbying activities of organized interests in federal judicial nominations’, Journal of Politics 62: 51–69. Carpenter, D., Esterling, K. and Lazer, D. (1998) ‘The strength of weak ties in lobbying networks’, Journal of Theoretical Politics 10: 417–44. Carver, T. (2003) ‘Socialism–Marxism’, in R. Axtmann (ed.) Understanding Democratic Politics, London: Sage. Castles, F. (ed.) (1982) The Impact of Parties, London: Sage. Caul, M.L. and Gray, M.M. (2000) ‘From platform declarations to policy outcomes: changing party profiles and partisan influence over policy’, in R.J. Dalton and M.P. Wattenberg

References 163 (eds) Parties without Partisans: Political Change in Advanced Industrial Democracies, Oxford: Oxford University Press. Cerny, P.G. (1999) ‘Globalization and the erosion of democracy’, European Journal of Political Research 36: 1–26. Chari, R.S. and Cavatorta, F. (2002) ‘Economic actors’ political activity in “overlap issues”: privatization and EU State Aid control’, West European Politics 25: 119–42. Chin, M.L., Bond, J.R. and Geva, N. (2000) ‘A foot in the door: an experimental study of PAC and constituency effects on access’, Journal of Politics 62: 534–49. Christiansen, G.B. and Haveman, R.H. (1981) ‘The contribution of environmental regulations to the slowdown in productivity growth’, Journal of Environmental Economics and Management 8: 381–90. Clawson, D., Neustadtl, A. and Scott, D. (1992) Money Talks: Corporate PACs and Political Influence, New York: Basic Books. Clawson, D., Neustadtl, A. and Weller, M. (1998) Dollars and Votes: How Business Campaign Contributions Subvert Democracy, Philadelphia, PA: Temple University Press. Coen, D. and Grant, W. (2006) ‘Managing business and government relations’, in D. Coen and W. Grant (eds) Business and Government: Methods and Practice, Opladen: Barbara Budrich Publishers. Coleman, W.D. (1988) Business and Politics: A Study of Collective Action, Kingston, Ontario: McGill–Queen’s University Press. Competition Commission (2002) The Supply of Banking Services by Clearing Banks to Small and Medium-sized Enterprises, London: Stationary Office. Cookson, P.W. Jr. and Persell, C.H. (1984) Preparing for Power: America’s Elite Boarding Schools, New York: Basic Books. Craig, S.C., Kreppel, A. and Kane, J.G. (2001), ‘Public opinion and support for direct democracy: a grassroots perspective’, in M. Mendelsohn and A. Parkin (eds) Referendum Democracy: Citizens, Elites and Deliberation in Referendum Campaigns, Basingstoke: Palgrave Macmillan. Crawford, V.P. and Sobel, J. (1982) ‘Strategic information transmission’, Econometrica 50: 1431–51. Crenson, M.A. (1971) The Un-politics of Air Pollution: A Study of Non-Decisionmaking in American Cities, Baltimore, MD: Johns Hopkins University Press. Crepaz, M.M.L. (1995) ‘Explaining national variations of air pollution levels: political institutions and their impact in environmental policy-making’, Environmental Politics 4: 391–414. Crouch, C. (1993) Industrial Relations and European State Traditions. Oxford: Clarendon Press. Cruickshank, D. (2000) Competition in UK Banking: A Report to the Chancellor of the Exchequer, London: Stationary Office. Culhane, P. (1981) Public Lands Policies, Baltimore, MD: Johns Hopkins University Press. Dahl, R.A. (1957) ‘The concept of power’, Behavioral Science 2: 201–15. Dahl, R.A. (1958) ‘A critique of the ruling elite model’, American Political Science Review 52: 463–69. Dahl, R.A. (1961) Who Governs? Democracy and Power in an American City, New Haven: Yale University Press. Dahl, R.A. (1971) Polyarchy: Participation and Opposition, New Haven: Yale University Press. Dahl, R.A. (1985) A Preface to Economic Democracy, Berkeley: University of California Press.

164 References Dahl, R.A. (1989) Democracy and Its Critics, New Haven: Yale University Press. Darimont, T. (1995) ‘Konversion Chlorchemie’, Zeitschrift für Umweltchemie und Ökotoxologie 7: 280–6. Denzau, A.T. and Munger, M.C. (1986) ‘Legislators and interest groups: how unorganized interests get represented’, American Political Science Review 80: 89–106. Department of Trade and Industry (2002) ‘Competition in banking services for small and medium-sized enterprises’, Press release, 14 March. Domhoff, G.W. (1975) The Bohemian Grove and Other Retreats: A Study in Ruling-Class Cohesiveness, New York: Harper Torchbooks. Domhoff, G.W. (1998) Who Rules America? Power and Politics in the Year 2000, Mountain View, CA: Mayfield. Dowding, K. (1995) ‘Model or metaphor? A critical review of the policy network approach’, Political Studies 43: 136–58. Dowding, K. (1996) Power, Buckingham: Open University Press. Downs, A. (1957) An Economic Theory of Democracy, New York: Harper. Dryzek, J.S. (1996) Democracy in Capitalist Times: Ideals, Limits, and Struggles, Oxford: Oxford University Press. Dryzek, J.S. (2000) Deliberative Democracy and Beyond, Oxford: Oxford University Press. Duch, R.M. and Stevenson, R. (2006) ‘Assessing the magnitude of the economic vote over time and across nations’, Electoral Studies 25: 528–47. Elster, J. (1982) ‘Belief, bias and ideology’, in M. Hollis and S. Lukes (eds) Rationality and Relativism, Oxford: Basil Blackwell. Elster, J. (1983) Sour Grapes, Cambridge: Cambridge University Press. Elster, J. (1985) Making Sense of Marx, Cambridge: Cambridge University Press. Elster, J. (1986) ‘Introduction’, in J. Elster (ed.) Rational Choice, Oxford: Blackwell. Farrell, J. and Gibbons R. (1989) ‘Cheap talk with two audiences’, American Economic Review 79: 1214–23. Ferejohn, J.A. (1990) ‘Information and elections’, in J.A. Ferejohn and J.H. Kuklinski (eds) Information and Democratic Processes, Urbana, IL: University of Illinois Press. Fiorina, M. (1981) Retrospective Voting in American National Elections, New Haven: Yale University Press. Frey, B.S. and Bohnet, I. (1994) ‘Direct-democratic rules: the role of discussion’, Kyklos 47: 341–54. Friedman, M. (1962) Capitalism and Freedom, Chicago: University of Chicago Press. Friedrich, C.J. (1963) Man and His Government: An Empirical Theory of Politics, New York: McGraw–Hill. Fudenberg, D. and Tirole, J. (1991) ‘Perfect Bayesian equilibrium and sequential equilibrium’, Journal of Economic Theory 53: 236–60. Fukuyama, F. (1992) The End of History and the Last Man, London: Penguin. Garrett, G. (1998) Partisan Politics in the Global Economy, Cambridge: Cambridge University Press. George, A.L. and Bennett, A. (2005) Case Studies and Theory Development in the Social Sciences, Cambridge, MA: MIT Press. Getz, K.A. (1997) ‘Research in corporate political action: integration and assessment’, Business & Society 36: 32–72. Giddens, A. (1979) Central Problems in Social Theory: Action, Structure and Contradiction in Social Analysis, London: Macmillan. Glasgow University Media Group (1980) More Bad News: Volume 2 of Bad News, London: Routledge.

References 165 Godwin, R.K. and Seldon, B.J. (2002) ‘What corporations really want from government: the public provision of private goods’, in A.J. Cigler and B.A. Loomis (eds) Interest Group Politics, 6th edn, Washington DC: CQ Press. Golub, J. (1998) ‘Global competition and EU environmental policy: introduction and overview’, in J. Golub (ed.) Global Competition and EU Environmental Policy, London: Routledge. Goodman, J.B. (1991) ‘The politics of central bank independence’, Comparative Politics 23: 329–349. Gormley, W.T. (1983) The Politics of Public Utility Regulation, Pittsburgh, PA: University of Pittsburgh Press. Granato, J. (1996) ‘The effect of policy-maker reputation and credibility on public expectations: an application to macroeconomic policy changes’, Journal of Theoretical Politics 8: 449–70. Grant, W. (1993) Business and Politics in Britain, 2nd edn, Basingstoke: Macmillan. Grant, W. (2000) Pressure Groups and British Politics, Basingstoke: Macmillan. Grant, W., Paterson, W.E. and Whitston, C. (1988) Government and the Chemical Industry: A Comparative Study of Britain and West Germany, Oxford: Clarendon Press. Gray, W.B. (1987) ‘The cost of regulation: OSHA, EPA and the productivity slowdown’, American Economic Review 77: 998–1006. Green, P. (1985) Retrieving Democracy: In Search of Civic Equality, London: Methuen. Grossman, G.M. (1993) ‘Pollution and growth: What do we know?’, Discussion paper no. 848, London: Centre for Economic Policy Research. Grossman, G.M. and Helpman, G. (2001) Special Interest Politics, Cambridge, MA: The MIT Press. Grossman, G.M. and Krueger, A.B. (1995) ‘Economic growth and the environment’, Quarterly Journal of Economics 110: 353–77. Hall, P.A. and Soskice, D. (eds) (2001) Varieties of Capitalism: The Institutional Foundations of Comparative Advantage, Oxford: Oxford University Press. Haq, G. et al. (2001) ‘Determining the costs to industry of environmental regulation’, European Environment 11: 125–39. Harrison, K. (1999) ‘Racing to the top or the bottom? Industry resistance to eco-labelling of paper products in three jurisdictions’, Environmental Politics 8: 110–37. Hart, D.M. (2002) ‘Business is not an interest group (and, by the way, there’s no such thing as “business”): On the study of companies in American national politics’, Faculty Research Working Papers Series, RWP02–032, John F. Kennedy School of Government, Harvard University. Hayek, F.A. von (1944) The Road to Serfdom, Chicago: University of Chicago Press. Healy, P.M. and Palepu, K.G. (2001) ‘Information asymmetry, corporate disclosure and capital markets: a review of the empirical disclosure literature’, Journal of Accounting and Economics 31: 405–40. Heclo, H. (1978) ‘Issue networks and the executive establishment’, in A. King (ed.) The New American Political System, Washington, DC: American Enterprise Institute. Heinz, J.P., Laumann, E.O., Nelson, R. and Salisbury, R.H. (1993) The Hollow Core: Private Interests in National Policy-Making, Cambridge: Harvard University Press. Hessischer Landtag (1993a) Technische Wohnungsbaurichtlinie 1993 der Hessischen Landesregierung, Landtagsdrucksache 13/4403, Wiesbaden: Hessische Landesdruckerei. Hessischer Landtag (1993b) Protokoll der Plenarsitzung 13/068 [16. June 1993], Wiesbaden: Hessische Landesdruckerei.

166 References Hessischer Landtag (1993c) Dringlicher Antrag der Fraktionen der SPD und der GRÜNEN betreffend Verwendung umweltfreundlicher Materialien im Wohnungsbau, Landtagsdrucksache 13/4582, Wiesbaden: Hessische Landesdruckerei. Hessischer Landtag (1993d) Protokoll der Plenarsitzung 13/072 [22. July 1993], Wiesbaden: Hessische Landesdruckerei. Hicks, A. (1988) ‘National collective action and economic performance: a review article’, International Studies Quarterly 32: 131–53. Hillman, A.J., Keim, G.D. and Schuler, D. (2004) ‘Corporate political activity: a review and research agenda’, Journal of Management 30: 837–57. Hinich, M.J. and Munger, M.C. (1994) Ideology and the Theory of Political Choice, Ann Arbor: University of Michigan Press. Hirschman, A.O. (1984) ‘Against parsimony: three easy ways of complicating some categories of economic discourse’, American Economic Review 74: 89–96. HM Treasury (2000) Chancellor of the Exchequer’s Budget Speech – 21 March 2000, London: Stationary Office. Available at: [Accessed 23 August 2007]. HM Treasury (2002) ‘Eight main clearing banks agree to competition commission remedies’, Press release, 31 October, London: Stationary Office. HM Treasury (2004) Microeconomic Reform in Britain: Developing Opportunities for All, Basingstoke: Palgrave. Hobbes, T. (1996 [1651]) Leviathan, Cambridge: Cambridge University Press. House of Commons Treasury Committee (2002) Banking, Consumers and Small Businesses, Fifth Report of Session 201–02, Volume I, HC 818–1, London: Stationary Office. Hunter, F. (1953) Community Power Structure: A Study of Decision Makers, Chapel Hill: University of North Carolina Press. Huntington, S.P. (1999) ‘Cultures in the 21st century: conflicts and convergences’, Keynote Address at Colorado College’s 125th Anniversary Symposium, 4 February, 1999. Available at [Accessed 23 August 2007]. Hyland, J.L. (1995) Democratic Theory: The Philosophical Foundations, Manchester: Manchester University Press. Inglehart, R. (1997) Modernization and Postmodernization, Princeton: Princeton University Press. Jacob, K. (1999) Ökologische Modernisierung und Strukturwandel in der Chemischen Industrie: Der Fall Chlorchemie, Berlin: Forschungsstelle für Umweltpolitik am Otto– Suhr–Institut für Politikwissenschaft der Freien Universität Berlin. Jaffe, A.B., Peterson, S.R., Portney, P.R. and Stavins, R.N. (1995) ‘Environmental regulation and the competitiveness of US manufacturing: what does the evidence tell us?’, Journal of Economic Literature 33: 132–63. Jänicke, M. (1992) ‘Conditions for environmental policy success: an international comparison’, The Environmentalist 12: 47–58. Jessop, B. (1996) ‘Interpretive sociology and the dialectic of structure and agency’, Theory, Culture & Society 13: 119–28. John, P. (1998) Analyzing Public Policy, London/New York: Continuum. John, P. (2000) ‘Policy networks’, in K. Nash and A. Scott (eds) Blackwell Companion to Political Sociology, Oxford: Blackwell. Jordan, A.G. (2001) Shell, Greenpeace, and the Brent Spar, Basingstoke: Palgrave.

References 167 Jordan A.G. and Maloney, W. (2001) ‘Britain: change and continuity within the new realities of British politics’, in C. Thomas (ed.) Political Parties and Interest Groups: Shaping Democratic Governance, Boulder, CO: Lynne Rienner. Katz, R.S. and Wessels, B. (eds) (1999) The European Parliament, the National Parliaments, and European Integration, Oxford: Oxford University Press. Kennan, J. and Wilson, R. (1993) ‘Bargaining with private information’, Journal of Economic Literature 31: 45–104. Key, V.O. (1964) Politics, Parities and Pressure Groups, 5th edn, New York: Crowell. Key, V.O. (1966) The Responsible Electorate: Rationality in Presidential Voting, 1936–1960, Cambridge, MA: Belknap Press. Kiewiet, D.R. (2000) ‘Economic retrospective voting incentives for policymaking’, Electoral Studies 19: 427–44. Kinder, D. and Kiewiet, D.R. (1979) ‘Economic discontent and political behavior: the role of personal grievances and collective economic judgments in congressional voting’, American Journal of Political Science 23: 495–527. King, G., Tomz, M. and Wittenberg, J. (2000) ‘Making the most of statistical analyses: improving interpretation and presentation’, American Journal of Political Science 44: 347–61. Kitschelt, H. (1994) The Transformation of European Social Democracy, Cambridge: Cambridge University Press. Knoke, D. (1990) Political Networks: The Structural Perspective, Cambridge: Cambridge University Press. Knoke, D., Pappi, F.U., Broadbent, J. and Tsujinaka, Y. (1996) Comparing Policy Networks, Cambridge: Cambridge University Press. Kollman, K. (1998) Outside Lobbying: Public Opinion and Interest Group Strategies, Princeton, N.J.: Princeton University Press. König, T. and Bräuninger, T. (1998) ‘The formation of policy networks’, Journal of Theoretical Politics 10: 445–71. Kraft, M.E. (1996) Environmental Policy and Politics: Toward the Twenty-First Century, London: Harper Collins. Krehbiel, K. (1999) ‘Pivotal politics: a refinement of nonmarket analysis for voting institutions’, Business and Politics 1: 63–81. Kreps, D.M. and Wilson, R. (1982) ‘Sequential equilibria’, Econometrica 50: 863–849. Lacy, D. and Paolino, P. (1998) ‘Downsian voting and the separation of powers’, American Journal of Political Science 42: 1180–99. Laffont, J.–J. and J. Tirole (1986) ‘Using cost observation to regulate firms’, Journal of Political Economy 94: 614–41. LaPalombara, J. (1964) Interest Groups in Italian Politics, Princeton, N.J.: Princeton University Press. Lasswell, H.D. (1958 [1936]) Politics: Who Gets What, When, and How, New York: Meridian. Laumann, E.O. and Knoke, D. (1987) The Organizational State: A Perspective on National Energy and Health Domains, Madison: University of Wisconsin Press. Le Fanu, J. (2004) ‘The UK banking market: consolidation, competition and the Cruickshank Report’, Die Bank 3: 150–4. Lehmbruch, G. (1979) ‘Liberal corporatism and party government’, in P.C. Schmitter and G. Lehmbruch (eds) Trends Toward Corporatist Intermediation, London: Sage.

168 References Levy, D.L. (2005) ‘Business and the evolution of climate regimes: the dynamics of corporate strategies’, in D.L. Levy and P.J. Newell (eds) The Business of Global Environmental Governance, Cambridge, MA: MIT Press. Lewis-Beck, M.S. and Paldam, M. (2000) ‘Economic voting: an introduction’, Electoral Studies 19: 113–21. Lewis-Beck, M.S. and Stegmaier, M. (2000) ‘Economic determinants of electoral outcomes’, Annual Review of Political Science 3: 183–219. Lijphart, A. (1999) Patterns of Democracy: Government Forms and Performance in Thirty-Six Countries, New Haven: Yale University Press. Lijphart, A. and Crepaz, M.L. (1991) ‘Corporatism and consensus democracy in eighteen countries: conceptual and empirical linkages’, British Journal of Political Science 21: 235–46. Lindblom, C.E. (1977) Politics and Markets: The World’s Political–Economic Systems, New York: Basic Books. Linck, J.S., Netter, J.M. and Yang, T. (2006) ‘The effects and unintended consequences of the Sarbanes–Oxley Act, and its era, on the supply and demand for directors’, paper presented at the Annual Meeting of the American Finance Association, Boston MA, 6–8 January. Available at http://papers.ssrn.com/sol3/papers.cfm?abstract_id=687496> [Accessed 23 August 2007]. Lipset, S.M. (1959) ‘Some social requisites of democracy: economic development and political legitimacy’, American Political Science Review 53: 69–105. Lipset, S.M. and Schneider, W. (1987) The Confidence Gap: Business, Labor and Government in the Public Mind, Baltimore: Johns Hopkins University Press. Little, D. (1991) Varieties of Social Explanation: An Introduction to the Philosophy of Social Science, Boulder, CO: Westview Press. Lohmann, S. (1993) ‘A signaling model of informative and manipulative political action’, American Political Science Review 87: 319–33. Lohmann, S. (1998) ‘An information rationale for the power of special interests’, American Political Science Review 92: 809–27. Lowi, T.J. (1964) ‘American business, public policy, case-studies, and political theory’, World Politics 16: 677–715. Lukes, S. (2005) Power: A Radical View, 2nd edn, London: Macmillan. Lupia, A. (1992) ‘Busy voters, agenda control, and the power of information’, American Political Science Review 86: 390–403. Maass, A. (1951) Muddy Waters: The Army Engineers and the Nation’s Rivers, Cambridge: Harvard University Press. MacDonald, E. and Schoenberger, C.R. (2006) ‘The world’s most powerful women’, Forbes.com. Available HTTP: [Accessed 23 August 2007]. MacKuen, M.B., Erikson, R.S. and Stimson, J.A. (1992) ‘Peasants or bankers? The American electorate and the US economy’, American Political Science Review 86: 597–611. Macpherson, C.B. (1973) Democratic Theory: Essays in Retrieval, Oxford: Clarendon Press. Maloney, W.A., Jordan, G. and McLaughlin, A.M. (1994) ‘Interest groups and public policy: the insider/outsider model revisited’, Journal of Public Policy 14: 17–38. Marsh, D. and Rhodes, R. A. W. (eds) (1992) Policy Networks in British Government, Oxford: Clarendon Press. Martin, C.J. (1989) ‘Business influence and state power: the case of US corporate tax policy’, Politics and Society 17: 189–223.

References 169 Martin, C.J. and Swank, D. (2007) ‘The partisan origins of coordinated capitalism: employer organizations and party systems in the age of innocence’, paper presented at the Annual Meeting of the Midwest Political Science Association, Chicago, April 12–5. Marx, K. (1977 [1852]) ‘The Eighteenth Brumaire of Louis Bonaparte’, in D. McLellan (ed.) Karl Marx: Selected Writings, Oxford: Oxford University Press. Marx, K. and Engels, F. (1977 [1848]) ‘The communist manifesto’, in D. McLellan (ed.) Karl Marx: Selected Writings, Oxford: Oxford University Press. McClosky, H. and Zaller, J. (1984) The American Ethos: Public Attitudes Toward Capitalism and Democracy, Cambridge: Harvard University Press. McCormick, J. (2001) Environmental Policy in the European Union, Basingstoke: Palgrave. McFarland, A.S. (1991) ‘Interest groups and political time: cycles in America’, British Journal of Political Science 21: 257–84. Meier, K.J. (1988) The Political Economy of Regulation: The Case of Insurance, Albany: SUNY Press. Miliband, R. (1969) The State in Capitalist Society, New York: Basic Books. Miliband, R. (1977) Marxism and Politics, Oxford: Oxford University Press. Mills, C.W. (1956) The Power Elite, New York: Oxford University Press. Mitchell, N.J. (1997) The Conspicuous Corporation: Business, Public Policy, and Representative Democracy, Ann Arbor: University of Michigan Press. Mitchell, R.B. and Bernauer, T. (1998) ‘Empirical research on international environmental policy: designing qualitative case studies’, Journal of Environment and Development 7: 4–31. Mizruchi, M.S. (1982) The American Corporate Network, 1904–1974, Beverly Hills: Sage. Mizruchi, M.S. (1992) The Structure of Corporate Political Action: Interfirm Relations and Their Consequences, Cambridge, MA: Harvard University Press. Mizruchi, M.S. and König, T. (1986) ‘Economic sources of corporate political consensus: an examination of interindustry relations’, American Sociological Review 51: 482–91. Moe, T.M. (1997) ‘The positive theory of public bureaucracy’, in. D.C. Mueller (ed.) Perspectives on Public Choice, Cambridge: Cambridge University Press, 455–80. Moene, K.O. (1993) ‘Contested power’, in D. Copp, J. Hampton and J. Roemer (eds) The Idea of Democracy, Cambridge: Cambridge University Press. Mokken, R.J. and Stokman, F.N. (1976), ‘Power and influence as political phenomena’, in: B. Barry (ed.) Power and Political Theory: Some European Perspectives, London: John Wiley. Moomaw, W.R. and Unruh, G.C. (1997) ‘Are environmental Kuznets curves misleading us? The case of CO2 emissions’, Environment and Development Economics 2: 451–63. Moore, G. (1979) ‘The structure of a national elite network’, American Sociological Review 44: 673–92. Morriss, P. (1987) Power: A Philosophical Analysis, Manchester: Manchester University Press. Morton, R.B. (1999) Methods and Models: A Guide to the Empirical Analysis of Formal Models in Political Science, Cambridge: Cambridge University Press. Morton, R. and Cameron, C. (1992) ‘Elections and the theory of campaign contributions: a survey and critical analysis’, Economics and Politics 4: 79–108. Mullen, I. (2001) ‘Prosperity and morality: a partnership with government’, The New Statesman 14 (679): R8–R12. Murphy, D. (2004) The Structure of Regulatory Competition: Corporations and Public Policies in a Global Economy, Oxford: Oxford University Press. Nash, J. (1951) ‘Noncooperative games’, Annals of Mathematics 54: 286–95.

170 References Neumayer, E. (2003) ‘Are left-wing party strength and corporatism good for the environment? Evidence from panel analysis of air pollution in OECD countries’, Ecological Economics 45: 203–20. Nolte, R.F. and Joas, R. (1995) Handbuch Chlorchemie I – 1992– Produkte, Unternehmen, Stoffflüsse, Berlin: Umweltbundesamt. Norton, P. (1993) Does Parliament Matter?, London: Harvester Wheatsheaf. Offe, C. (1981). ‘The attribution of public status to interest groups’, in S.D. Berger (ed.) Organizing Interests in Western Europe: Pluralism, Corporatism, and the Transformation of Politics, Cambridge: Cambridge University Press. Office of Fair Trading (2003) ‘OFT sends government progress report on banks’, Press release, 2 June, 2003. Office of Management and Budget (2003) Informing Regulatory Decisions: 2003 Report to Congress on the Costs and Benefits of Federal Regulations and Unfunded Mandates on State, Local, and Tribal Entities, Available from [Accessed 23 August 2007]. Olson, M. (1965) The Logic of Collective Action: Public Goods and the Theory of Groups, Cambridge, MA: Harvard University Press. Olson, M. (1982) The Rise and Decline of Nations: Economic Growth, Stagflation, and Social Rigidities, New Haven: Yale University Press. Paehlke, R. (1988) ‘Democracy, bureaucracy and environmentalism’, Environmental Ethics 10: 291–308. Page, B.I. (1978) Choices and Echoes in Presidential Elections: Rational Man and Electoral Democracy, Chicago: University of Chicago Press. Page, B.I., Shapiro, R.Y. and Dempsey, G. (1987) ‘What moves public opinion?’, American Political Science Review 81: 23–43. Palmer, H.D. and Whitten, G.D. (2000) ‘Government competence, economic performance and endogenous election dates’, Electoral Studies 19: 413–26. Pateman, C. (1970) Participation and Democratic Theory, Cambridge: Cambridge University Press. Péchoux, I. and Pouyet, J. (2002) ‘Regulatory competition, transboundary pollution and international trade’, in L. Marsaliani (ed.), Environmental Economics and the International Economy, Dordrecht: Kluwer Academic Publishers. Peltzman, S. (1973) ‘An evaluation of consumer protection legislation: the 1962 drug amendments’, Journal of Political Economy 81: 1049–91. Peltzman, S. (1976) ‘Toward a more general theory of regulation’, Journal of Law and Economics 19: 211–40. Pertschuk, M. (1982) Revolt Against Regulation: The Rise and Pause of the Consumer Movement, Berkeley, CA: University of California Press. Pijnenburg, B. and Thomas, C.S. (2004) ‘Comparative interest group studies’, in C.S.Thomas (ed.) Research Guide to U.S. and International Interest Groups, Westport, CT: Westview Press. Polanyi, K. (2001 [1944]) The Great Transformation: The Political and Economic Origins of Our Time, Boston: Beacon Press. Polk, A. (2002) ‘How special interests shape policy: a survey’, Working Paper No. 0206. Zurich: Socioeconomic Institute, University of Zurich. Polk, A. and Schmutzler, A. (2005) ‘Lobbying against environmental regulation vs. lobbying for loopholes’, European Journal of Political Economy 21: 915–31. Polsby, N.W. (1960) ‘How to study community power: the pluralist alternative’, Journal of Politics 22: 474–84.

References 171 Polsby, N.W. (1980) Community Power and Political Theory: A Further Look at Problems of Evidence and Inference, 2nd edn, New Haven: Yale University Press. Popper, K.R. (1966) The Open Society and its Enemies, vol. 2, 5th edn, London: Routledge & Kegan Paul. Porter, M.E. and van der Linde, C. (1995) ‘Toward a new conception of the environmentcompetitiveness relationship’, Journal of Economic Perspectives 9: 97–118. Portney, P.R. (1990) ‘Policy watch: economics and the Clean Air Act’, Journal of Economic Perspectives 4: 173–81. Posen, A.S. (1995) ‘Declarations are not enough: financial sector sources of central bank independence’, in B. Breanne and J. Rotenberg (eds) NBER Macroeconomics Annual, Cambridge, MA: MIT Press. Potters, J. and van Winden, F. (1992) ‘Lobbying and asymmetric information’, Public Choice 74: 269–92. Poulantzas, N. (1973) Political Power and Social Classes, London: NLB and Sheed and Ward. Prognos AG (1994) Endbericht: Konversion der Chlorchemie, Wiesbaden: Hessisches Ministerium für Umwelt, Energie und Bundesangelegenheiten. Przeworski, A. (1991) Democracy and the Market: Political and Economic Reforms in Eastern Europe and Latin America, Cambridge: Cambridge University Press. Przeworski, A. (1999) ‘Minimalist conception of democracy: a defense’, in Ian Shapiro and C. Hacker-Cordón (eds) Democracy’s Value, Cambridge: Cambridge University Press. Przeworski, A. (2003) States and Markets: A Primer in Political Economy, Cambridge: Cambridge University Press. Przeworski, A. and Sprague, J. (1986) Paper Stones: A History of Electoral Socialism, Chicago: University of Chicago Press. Przeworski, A. and Teune, H. (1982) The Logic of Comparative Social Inquiry, 2nd edn, Malabar, Florida: Krieger. Przeworski, A. and Wallerstein, M. (1988) ‘Structural dependence of the state on capital’, American Political Science Review 82: 11–29. Quinn, D.P. and Shapiro, R.Y. (1991) ‘Business political power: the case of taxation’, American Political Science Review 85: 851–74. Rao, J.N.K. and Scott, A. J. (1984) ‘On Chi-squared tests for multiway contingency tables with cell proportions estimated from survey data’, Annals of Statistics 12: 46–60. Redford, E.S. (1969) Democracy in the Administrative State, New York: Oxford University Press. Rhodes, R.A.W. and Marsh, D. (1992) ‘New directions in the study of policy networks’, European Journal of Political Research 21: 181–205. Rhodes, R.A.W. and Marsh, D. (1994) ‘Policy networks: ‘defensive’ comments, modest claims, and plausible research strategies’, paper prepared for the Annual Conference of the Political Studies Association, Swansea, March 29–31. Richardson, J.J. (2000) ‘Government, interest groups, and policy change’, Political Studies 48: 1006–25. Richardson, J. and Jordan, G. (1979) Governing Under Pressure: The Policy Process in a Post-Parliamentary Democracy, Oxford: Martin Robertson. Ringquist, E.J. (1993) ‘Testing theories of state policy-making: the case of air quality regulation’, American Politics Quarterly 21: 320–42. Rodrik, D. (1997) Has Globalization Gone Too Far?, Washington, DC: Institute for International Economics.

172 References Roscoe, D.D. and Jenkins, S. (2005) ‘A meta-analysis of campaign contributions’ impact on roll call voting’, Social Science Quarterly 86: 52–68. Rose, R. (1980) Do Parties Make a Difference?, Chatham, NJ: Chatham House. Rose-Ackerman, S. (1999) Corruption and Government: Causes, Consequences, and Reform, New York, NY: Cambridge University Press. Rush, M. (ed.) (1990) Parliament and Pressure Politics, Oxford: Oxford University Press. Saalfeld, T. (2000) ‘Members of parliament and governments in Western Europe: agency relations and problems of oversight’, European Journal of Political Research 37: 353–76. Sabatier, P.A. and Jenkins-Smith, H.C. (1993) Policy Change and Learning: An Advocacy Coalition Approach, Boulder, CO: Westview. Sabato, L. (1984) PAC Power: Inside the World of Political Action Committees, New York: Norton. Safire, W. (1993) Safire’s New Political Dictionary, New York, NY: Random House. Salisbury, R.H. (1975) ‘Interest groups’, in F.I. Greenstein and N. Polsby (eds) Handbook of Political Science, Volume IV, Reading, MA: Addison–Wesley. Salisbury, R.H. (1984) ‘Interest representation: the dominance of institutions’, American Political Science Review 78: 64–76. Salisbury, R.H., Johnson, P., Heinz, J.P., Laumann, E.O. and Nelson, R.L. (1989) ‘Who you know versus what you know: the uses of government experience for Washington lobbyists’, American Journal of Political Science 33: 175–95. Sanchez, C.M. (1997) ‘Environmental regulation and firm-level innovation: the moderating effects of organizational and individual-level variables’, Business and Society 36: 140–68. Satz, D. and Ferejohn, J. (1994) ‘Rational choice and social theory’, Journal of Philosophy 91: 71–87. Sayre, W.S. and Kaufman, H. (1965) Governing New York City: Politics in the Metropolis, New York: Norton. Schattschneider, E.E. (1975 [1960]) The Semisovereign People: A Realist’s View of Democracy in America, Hinsdale, IL: Dryden P. Schlozman, K.L. and Tierney, J.T. (1986) Organized Interests and American Democracy, New York: Harper & Row. Schmitter, P.C. (1974) ‘Still the century of corporatism?’, Review of Politics 36: 85–131. Schmitter, P.C. (1981) ‘Interest intermediation and regime governability in contemporary Western Europe and North America’, in S.D. Berger (ed.) Organizing Interests in Western Europe: Pluralism, Corporatism, and the Transformation of Politics, Cambridge: Cambridge University Press. Schmitter, P.K. (1982) ‘Reflections on where the theory of neo-corporatism has gone and where the praxis of neo-corporatism may be going’, in P.K. Schmitter and G. Lehmbruch (eds) Patterns of Corporatist Policy-Making, London: Sage. Scholtz, J.T. and Feng-heng, W. (1986) ‘Regulatory enforcement in a federalist system’, American Political Science Review 80: 1249–70. Schumpeter, J.A. (1976 [1942]). Capitalism, Socialism, and Democracy, London: Allen and Unwin. Scruggs, L.A. (1999) ‘Institutions and environmental performance in seventeen Western democracies’, British Journal of Political Science 29: 1–31. Selten, R. (1965) ‘Spieltheoretische Behandlung eines Oligopolmodells mit Nachfrageträgheit’, Zeitschrift für die gesamte Staatswissenschaft 121: 301–24. Sen, A.K. (1977) ‘Rational Fools: A critique of the behavioral foundations of economic theory’, Philosophy and Public Affairs 6: 317–44.

References 173 Shepsle, K. (1972) ‘The strategy of ambiguity, uncertainty, and competition’, American Political Science Review 66: 555–68. Siaroff, A. (1999) ‘Corporatism in 24 industrial democracies: meaning and measurement’, European Journal of Political Research 36: 175–205. Skocpol, T. (1980) ‘Political response to capitalist crisis: neo-Marxist theories of the state and the case of the new deal’, Politics and Society 10: 155–201. Skocpol, T. (1985) ‘Bringing the state back in: strategies of analysis in current research’, in P. Evans, D. Rueschemeyer and T. Skocpol (eds) Bringing the State Back In, New York and Cambridge: Cambridge University Press. Sloof, R. (1997) ‘Lobbying when the decisionmaker can acquire independent information’, Public Choice 77: 199–207. Sloof, R. (1998) Game-theoretic Models of the Political Influence of Interest Groups, Boston: Kluwer Academic. Sloof, R. and van Winden, F. (2000) ‘Show them your teeth first! A game-theoretic analysis of lobbying and pressure’, Public Choice 104: 81–120. Smith, M.A. (1999) ‘Public opinion, elections, and representation within a market economy: Does the structural power of business undermine popular sovereignty?’, American Journal of Political Science 43: 842–63. Smith, M. A. (2000) American Business and Political Power: Public Opinion, Elections, and Democracy, Chicago: University of Chicago Press. Stigler, G.J. (1975) The Citizen and the State: Essays on Regulation, Chicago: University of Chicago Press. Su, T., Neustadtl, A. and Clawson, D. (1995) ‘Business and the conservative shift: corporate PAC contributions, 1976–1986’, Social Science Quarterly 76: 20–40. Sunstein, C. (1998) ‘Health–health trade-offs’, in J. Elster (ed.) Deliberative Democracy, Cambridge: Cambridge University Press. Svallfors, S. (1997) ‘Worlds of welfare and attitudes to redistribution: a comparison of eight Western nations’, European Sociological Review 13: 283–304. Swank, D. (1992) ‘Politics and the structural dependence of the state in democratic capitalist nations’, American Political Science Review 86: 38–54. Therborn, G. (1977) ‘The rule of capital and the rise of democracy’, New Left Review 103: 3–41. Thomson, R. (1999) The Party Mandate: Election Pledges and Government Pledges in the Netherlands, 1986–1998, Groningen: University of Groningen PhD thesis. Tietenberg, T. (1994) Economics and Environmental Policy, Aldershot: Edward Elgar. Traxler, F. (1999) ‘The State in Industrial Relations: A cross-national analysis of developments and socioeconomic effects’, European Journal of Political Research 36: 55–85. Truman, D.B. (1963 [1951]) The Governmental Process. Political Interests and Public Opinion, New York: Knopf. Ulph, A. (1997) ‘International trade and the environment: a survey of recent economic analysis’, in H. Folmer and T. Tietenberg (eds) International Yearbook of Environmental and Resource Economics 1997/98, Cheltenham: Edward Elgar. Umweltbundesamt (2001) Action Areas and Criteria for a Precautionary, Sustainable Substance Policy Using the Example of PVC, Berlin: Erich Schmidt Verlag. Useem, M. (1984) The Inner Circle: Large Corporations and the Rise of Business Political Activity in the US and UK, New York: Oxford University Press. van den Berg, A. (1988) The Immanent Utopia: From Marxism on the State to the State of Marxism, Princeton: Princeton University Press.

174 References van Winden, F. (1999) ‘On the Economic Theory of Interest Groups: towards a group frame of reference in political economics’, Public Choice 100: 1–29. Vogel, D. (1989) Fluctuating Fortunes: The Political Power of Business in America, New York: Basic Books. Vogel, D. (1996) Kindred Strangers: The Uneasy Relationship Between Politics and Business in America, Princeton, NJ: Princeton University Press. Vogel, D. (1997) Trading Up: Consumer and Environmental Regulation in a Global Economy, Cambridge: Harvard University Press. Wang, S.–T., Yu, M. –L., Wang, C.–J. and Huang, C.–C. (1999) ‘Bridging the gap between the pros and cons in treating ordinal scales as interval scales from an analysis point of view’, Nurse Researcher 48: 226–29. Weber, M. (1964) The Theory of Social and Economic Organization, New York: Free Press. White, H. (1980) ‘A heteroskedasticity-consistent covariance matrix estimator and a direct test for heteroskedasticity’, Econometrica 48: 817–38. Whitehead, L. (1992) ‘The alternatives to ‘liberal democracy’: a Latin American perspective’, Political Studies 40: 146–59. Wildavsky, A. (1995) But Is It True?: A Citizen’s Guide to Environmental Health and Safety Issues, Cambridge, MA: Harvard University Press. Williams, J.T. and Collins, B.K. (1997) ‘The political economy of corporate taxation’, American Journal of Political Science 41: 208–44. Wilson, F.L. (1987) Interest Group Politics in France, Cambridge: Cambridge University Press. Wilson, G.K. (1990) Business and Politics: A Comparative Introduction, 2nd edn, New York: Macmillan. Wilson, G.K. (2003) Business and Politics: A Comparative Introduction, 3rd edn, New York: Macmillan. Wilson, G.K. (2006) ‘Thirty years of business and politics’, in D. Coen and W. Grant (eds) Business and Government: Methods and Practice, Opladen: Barbara Budrich Publishers. Wilson, J.Q. (1961) ‘The strategy of protest: problems of negro civic action’, Journal of Conflict Resolution 5: 291–303. Wilson, J.Q. (1981) ‘Democracy and the corporation’, in R. Hessen (ed.) Does Big Business Rule America?, Washington DC: Ethics and Policy Center. Wood, S. (2001) ‘ Business, government, and patterns of labor market policy in Britain and the Federal Republic of Germany’, in P.A. Hall and D. Soskice (eds) Varieties of Capitalism: The Institutional Foundations of Comparative Advantage, Oxford: Oxford University Press. World Economic Forum (2002) The Global Competitiveness Report, 2001–2002, New York: Oxford University Press. Worsham, J. (1997) Other People’s Money: Policy Change, Congress, and Bank Regulation, Boulder, CO: Westview Press. Zhang, I.X. (2007) ‘Economic consequences of the Sarbanes–Oxley Act of 2002’, Working paper, Available at: http://ssrn.com/abstract=961964

Index

Page numbers in italics are figures. Page numbers followed by ‘t’ are tables. access: denial of 94; and interests groups 55; and networks 37; and privilege 50; privileged 25–7, 38, 60, 98, 135 adaptive preference formation 42–3 advertising 13 agency capture theory 23 Ainsworth, S. 58, 145 Austen-Smith, D. 58, 142 Bachrach, P. 27–8 banking signalling game 79–84, 88–9, 142 Baratz, M.S. 27, 27–8 bargaining 158 Barnes, B. 28, 158 Barrett, Craig 104–5 BAT (British American Tobacco) 138–9 Baumgartner, F.R. 51 Bayes’ rule 67, 71, 159 Bearden, J. 35 Beck, U. 78 Becker, G.S. 23 Belgium 5, 33, 97, 102, 109, 160 Bentley, A.F. 24–5, 45 Bernhagen, P. 70–5 Bezdek, R.H. 115 bias, mobilization of 28 Binder, J.J. 142 Bischoff, I. 124–5 Blair, Tony 138–9 Block, F. 47, 53 Bond, J.R. 26 Bonnet, I. 144 Braun, Ludwig-Georg 11 Bräuninger, T. 70–5 Brent Spar 50 British American Tobacco (BAT) 138–9 Broadbent, J. 38

Broughton, Martin 138–9 Brown, Gordon 79, 80, 90 Busch, Wilhelm 94 business actors (definition) 19 business confidence factor 10, 24, 54–5 business (definition) 19 Byers, Stephen 138–9 ‘calculated heroism’ 54 Caldeira, G.A. 55 Cameron, C. 28, 30 Canada, 50, 152t 160 capital flight 119 capitalism 2–3, 19, 47, 151t; and constraints on policymaking 24, 30, 135; and democracy 5–8, 10, 30, 35, 40; see also capitalist democracy capitalist democracy 2–3, 5, 13, 49, 153t; and information 8–9, 54, 144 ‘capitalist pluralism’ 33; see also pluralism; ‘capitalist statism’ 33 capturing see agency capture Carpenter, D. 37 Caul, M.L. 50, 76 causation 14–15 Center for International Development 121 Cheney, Richard 37 Chin, M.L. 26 Citigroup 26 class: commercial 6; and ideology 43; negotiation between 7, 35; and networks 39; and structural dependence of the state on capital 47 Clean Air Act 120 co-ordinated market economies 32–4 ‘command-and-control’ policies 78 commercial classes 6

176 Index communism 3, 4 confidence 12, 44, 111, 125, 130; and business confidence factor 10, 24, 54–5 confidence in business 12, 44, 55 consequences of policies 13, 49–50, 57 constraints on policymaking 24, 30, 135 consumer protection 23 contacts 100–11, 100–103t, 107t, 149–51t contributions see finance convergence of ideology 50 Cookson, P.W. Jr. 36 corporatism see neocorporatism corruption 24, 46 costs in lobbying 73–4, 73–4, 100–4 costs in lobbying/lobbyists, costs 73–4, 73–4, 100–4 counteractive lobbyists 124, 132, 137, 142 credibility: of lobbyists 11–12, 20, 60–2, 70–2, 95; of policymakers 52; see also lies Dahl, R.A. 5, 6, 18, 26–7 ‘decision method’ 27 democracy 5, 18–19; and capitalism 6, 8, 30, 35, 40, 49 democratic capitalism 8–9, 33, 39, 54, 144; and information 8–9, 144; and networks 22; and structural dominance 113 democratization 6–7 Denmark 33, 122, 152t, 160 Department of Trade and Industry 80, 84, 138 Dowding, K. 38, 50, 75 Dryzeck, J.S. 18 Dunschen, Günter 86, 87, 147 economic elites 3, 25 economic voting, 44–5, 118, 123t, 125, 154t, 155t 160 economy 43–5, 51, 114–17, 135 education 36 ‘egotropic’ voting 122 elite network theory 36 elite networks 34–9 elites 2, 3, 7, 9, 25; and education 36; and influence 36; integration 38; networks 34–9; social interaction between 24, 35 Elster, J. 42–3, 47 Engels, Friedrich 5, 22

Environmental Protection Agency (EPA) 119–20 environmental regulation 16–18, 23, 78, 152t, 154t, 155–7t; and the economy 114–17; and European Union (EU) 117; and informational asymmetries 120; and left–right ideology 117, 128, 129t, 131t; and negative inducement effects 113–15, 119–20; and neocorporatism 136–7; and preferences 114–17 equality 2, 5–6, 27 equilibrium 66–9, 73, 77, 88, 136; pooling 68, 73, 87, 90; separating, 72, 72t 75, 82–3, 90; sequential 66, 67, 72, 159 Esterling, K. 37 European Union(EU): and environmental regulation, 117, 160, 153–4t; members of parliament (MPs) 97–104 ‘false consciousness’ 41 finance 10–11, 23–5, 28, 46, 51, 62–3; regulation of financial services 15, 20 Finland 33, 160 ‘flight of capital’ 8, 17 Forbes 1, 3, 4 four dimensions to business’ legitimacy 143 France 8, 33, 36, 97, 102, 160 free-riding 29, 144 ‘freedom of information’ 143 Frey, B.S. 144 Friedman, M. 6 Friedrich, C.J. 44 Galbraith, John Kenneth 61 game-theoretic models 15; see also signalling games Garrett, G. 50–1, 158 GDP, 123, 125, 126t, 128, 129t, 130, 131t, 152–4t, 156t, 157t General Agreement on Tariffs and Trade (GATT) 121 General Motors 61–2 Germany 32–3, 97, 102, 160, 152t; economic voting 122; financial services regulation 15, 20; flight of capital 8 Geva, N. 26 globalization 7–8, 51, 121, 158 Godwin, R.K. 30 ‘government MPs’ 99–100 Gray, M.M. 50, 76 Greece 21, 97, 102, 125, 128, 160

Index 177 Hagemann, Karin 85, 89, 147 Hart, D.M. 62–3 Hayek, F.A. von 6 Heclo, H. 37 Heinz, J.P. 37–8 Hesse 84–8, 91–2 Hewitt, Patricia 80 Hicks, A. 33 Hielscher, Hans-Jürgen 85, 87, 89, 147 Higgins, Ciaran 56, 59, 60, 61, 147 Hillman, A.J. 3, 11 Hobbes, T. 5, 74–5 Hojnacki, M. 55 Hunter, F. 35 Huntington, S.P. 121 Hyland, J.L. 41 ideology 39–43, 99; and contacts 100–4; convergence of 50; and informational symmetries 56; and voters 52 ‘incentives-based’ policy 78 industrial relations 62; see also labour unions influence 5–6, 18, 55, 97–8, 160; and elites 36; left–right ideology, 107t 108–9; lobbying, 22, 24, 26, 104, 112, 149t; and neocorporatism 33 influence function 23 information 12, 30; and democratic capitalism 8–9, 144; informativeness 93–112; and interest groups 64; legal protection of 59; and lobbying 12, 55, 140; and neocorporatism 33, 124; and policymakers 11, 57, 59, 139–46; as private property 140, 143; and rational choice 12–13; and voters 30; see also credibility; informational asymmetries; informational structural model; lies informational asymmetries 52, 78, 99, 113–14, 137, 142; and environmental regulation 120, 123–4; and ideology 56; informational structural model 56–63; and policymakers 135 informational–structural model 94, 135–8; asymmetries 56–63, 113; and industrial relations 62; and signalling models 63–75 informativeness 93–112 Institute for Strategy and Competitiveness 121 institutional groups 30 intentional interference 8–9 intentionality 18

intentions 60, 140, 141 inter-organizational communication networks 37 interest group capture theory 23 interest groups 9, 11–12, 46, 58–9, 98, 137; and access 55; capture theory 23; inactivity of 29; and information 64; and neocorporatism 31–4; and pluralism 24–31; and policymakers, 25, 28–9, 38, 149–50t; and signalling games 64; see also lobbying interests 27, 29, 61; left–right ideology 26, 136 interference, intentional 8–9 international trade 23, 119, 122–3, 125–33 intrafirm democracy 5 Ireland 97, 102, 128, 160; economic voting 125; lobbying 104–5 Israel 121, 160 Italy 33, 86, 97, 98, 102, 160 Jaffe, A.B. 115 Japan 33, 38, 95, 160 Jenkins-Smith, H.C. 37 Jessop, B. 14 John, P. 38 Jones, Digby 17, 35, 48 Jordan, G. 37, 38–9 Keim, G.D. 3, 11 Kennan, J. 59, 62, 70 Key, V.O. 29 Kiewiet, D.R. 122 Kinder, D. 122 King, G. 158 Knoke, D. 37, 38 König, T. 34 Kraft, M.E. 78, 115 Labour party 52 labour unions 7, 26, 47–8, 62, 149–50t, 152t, 154t, 155t LaPalombara, J. 98 Lasswell, H.D. 3 Laumann, E.O. 37–8 ‘law of anticipated reactions’ 44 law enforcement 53 layoffs 115 Lazer, D. 37 Leech, B.L. 51 left–right ideology 50–1; and environmental regulation, 117, 128, 129t, 131t; influence, 107t 108–9;

178 Index interests 26, 136; proximity, 100–104, 100–103t; self-placement 95–6, 111–12, 99, 108–9, 150t, 156–7t legal protection of information 59 ‘liberal corporatism’ 33 liberal market economies 32 lies 61, 103–4, 112, 135; and information 58, 75, 124; and left-wing politicians 96; reputation 70, 71, 94–7; see also credibility Lindblom, C.E. 5, 19, 57, 74 Lipset, S.M. 6 lobbying games see signalling games lobbying/lobbyists 9–11, 52, 55, 61; and contacts, 100–11, 100–103t, 107t, 149–50t; counteractive 124, 132, 137, 142; credibility of 11–12, 20, 60–2, 70–2, 95; and influence 24, 26, 104, 112; 24, 26, 104, 112, 149t; and information 12, 55, 58, 140; ‘loophole’ 30; and pluralism 24; and policymakers 93; see also reputation; signalling games Lohmann, S. 58–9 ‘loophole lobbying’ 30 Lord, Martin 80, 82, 147 Lowi, T.J. 77 Lukes, S. 40–3, 46, 158 Lupia, A. 30 McFarland, A.S. 30, 31 McLaughlin, A.M. 38–9 Macpherson, C.B. 7 Maloney, W.A. 38–9 manipulation 40–3, 134, 158 market economies 32 market entry 23 Marsh, D. 38 Marx, Karl 5, 46–7 mass media 40 measuring structural constraints 121–5 median voter theorem 22 members of parliament (MPs), 97–104, 106–11, 107t, 109t Merkel, Angela 1, 3 Mills, C.W. 35 minimum wage legislation 23 Mintz, B. 35 Mitchell, N.J.: ‘calculated heroism’ 54–5; economic voting 118; four dimensions to business’ legitimacy 143; on influence 22; on information 57; on interests 61; on structural dependence 10

Mizruchi, M.S. 34–5, 39 mobilization of bias 28 Moene, K.O. 4 money see finance Morton, R.B. 28, 30 MPs (members of parliament), 100–4, 106–11, 107t, 109t Nash equilibrium 66–7 negative inducement effects 49–50, 54, 56–62, 71–5, 158; in banking signalling game 81; and environmental regulation 113–15, 119–20; and neocorporatism 130; in PVC case study 85–6, 90; and reputation 93–6, 104; and signalling games 77, 81, 83, 85–6, 87, 89–92 Nelson, R. 37–8 neo-pluralists 31 neocorporatism 31–4, 96, 134, 151–2t, 154t, 155t; and environmental regulation 124, 136–7; and interest groups 31–4; and negative inducement effects 130; theory 24 Netherlands 33, 50, 97, 102, 121, 125, 160 networks 34–9, 54, 110, 134–8 Neumayer, E. 137 New Haven, Connecticut 26–7, 27 Nolte, Rainer 91, 147 non-decisionmaking 27–8 non-intentional dominance 8–10 Norton, P. 97 Norway 33 Office of Management and Budget (OMB) 119–20 Olson, M. 7, 23, 29 opposition parties 99–100 Page, B.I. 52, 145 Palmer, Nick 83, 148 Pappi, F.U. 38 parties, political 50–1, 95, 99–100, 151t Pateman, C. 18 payoffs, 64–74, 65t, 66t, 67t, 72t, 159 Peltzman, S. 23, 49 Persell, C.H. 36 Peterson, S.R. 115 ‘play it safe’ 56 pledges 52, 71–4, 99; cost of 80–1, 84–5; fulfillment of 145 pluralism 24–31, 70, 134, 151t pocket-book voting 122 ‘policy advocacy coalition’ 37

Index 179 policy communities 36–9, 89 policy formation, theory of 23 policy network approach 38 ‘policy-influencing legislatures 97–8 policymakers 26; and ‘calculated heroism’ 54; and capitalism 9–10; and consequences of their policies 49–50, 57; credibility of 52; and economic voting 118; and information 11, 57, 59, 139–46; informational asymmetries 135; and interest groups 25, 28–9, 38; and lobbyists 93; and neocorporatism 31; and policy communities 37–8; and policy effects on the economy 45, 135; and signalling games 62, 64; see also MPs Political Actions Committees (PACs) 26 Polk, A. 11, 64 Polsby, N.W. 24 Popper, K.R. 13 ‘Porter hypothesis’ 115 Portney, P.R. 115, 116 ‘positive political theory’ see rational choice Poulantzas, N. 47 power (definition) 18; see also influence preferences 40–3, 114–17 pressure 24–5, 27, 28 price-fixing 23 private information 140, 143 ‘privileged position of business’ 24 privilege; and access 25, 38, 50, 60, 98, 135; structural 141–2 Przeworski, A. 19, 44–5, 47, 77 ‘public choice’ see rational choice public interest groups see interest groups ‘push and resistance’ 24 PVC case study 84–8, 91–2 ‘quadripartite corporatism’ 33 Quinn, D.P. 10–11, 50, 120 rational choice theories 12–15 re-election 26, 29, 30, 57–8, 75 redistribution of wealth 7, 22, 40, 44 regulation of financial services, finance 15, 20 reputation 136; costs in lobbying 100–4; and informativeness 93–112; and lobbying 69–75, 82–3; and negative inducement effects 93–6, 104 retrospective voting 30

‘revolving door politics’ 36–8, 138–9 Rhodes, R.A.W. 38 Rice, Condoleezza 1, 3 Richardson, J. 37 right–left ideology see left–right ideology Rinquist, E.J. 23 Rose, R. 145 Sabatier, P.A. 37 Salisbury, R.H. 30, 37–8 Sarbanes-Oxley Act 143–4 Schlozman, K.L. 25 Schmitter, P.K. 31 Schnerr, Mark 52, 60–1, 70, 148 Schuler, D. 3, 11 Schumpeter, J.A. 6, 18, 141 Schwarzenegger, Arnold 25, 26 second-dimensional power 27–8 Seldon, B.J. 30 Shapiro, R.Y. 10–11, 50, 120 Shepsle, K. 52 Siaroff, A. 32, 33 signalling games 62, 66–7t, 72t, 76–9, 135–66; on banking 79–84; models of lobbying 63–9; PVC case study 84–8; and reputation 69–75 Sloof, R. 64, 75 Smith, M.A. 118 ‘social choice’ see rational choice ‘social democratic corporatism’ 30, 33 social interaction between elites 24 socialism 2 Spain 97, 102, 125, 160 special interest groups see interest groups standards, loss of 7, 8, 17 state-centred theory 52–3 Stavins, R.N. 115 Stigler, G.J. 23 strategic-relational approach 14 ‘structural coercion’ 75 structural constraints 9–14, 46–52, 89, 113–14, 132, 135–6 measuring 121–5 ‘structural dependence of the state on capital’ 24, 44–51, 75, 141–2; and class 47; and information 57 structural dependence theories 10–11 structural dominance 10, 15–16; and democratic capitalism 113; and the economy 113, 118, 131; and informational asymmetries 134–5, 139; and informational–structural models 75; and pluralism 31

180 Index ‘structural equivalence’ 39 structural Marxism 46–7 structural privilege 141–2 ‘structurally privileged position of business’ see structural dependence theories Sunstein, C. 48–9 survey data on members of parliaments, 149–51t Sweden 32, 33, 97, 102, 160 Switzerland 22, 33, 121, 160 ‘systematic luck’ 75

United States of America (US): and economic voting 122; environmental regulation, 23, 114–15, 119, 126t, 127t 160; financial service regulation 15; fundraising 25, 28; and ideology 39–40; and liberal market economies 32; lobbying/lobbyists 70; and networks 36; New Haven, Connecticut 26–7; Office of Management and Budget 49; pledges 145; Sarbanes-Oxley Act 143 unity 34–6, 39, 51, 90–1

‘taken for granted’ interests 29 Teune, H. 77 theories; elite network 36; interest group capture 23; median voter 22; neocorporatism 24; policy formation 23; rational choice 12–15; state-centred 52–3 Therborn, G. 7 think tanks 40 third-dimensional power 40–3, 46 Tierney, J.T. 25 Tomz, M. 158 trades unions 7, 26, 47–8, 62, 149–50t, 154t, 155t Truman, D.B. 29 Tsujinaka, Y. 38 two-dimensional power 28

van Winden, F. 75 Vogel, D. 30, 31 voters 22, 44; and ideology 52; and information 30; and interest groups 30 voting 23; economic 44–5, 118, 122–3, 123t, 125, 154t, 155t 160; ‘egotropic’ 122; pocket-book 122; retrospective 30

unintended consequences 13 unions, 7, 26, 47–8, 62, 149–50t, 154t, 155t United Kingdom (UK): and ‘capitalist pluralism’ 33; and economic voting 125; and education 36; environmental regulation, 126–7t; financial service regulation 15; and liberal market economics 32; parliamentary study 97, 160; and ‘revolving door’ politics 37; small business banking 79–84, 88; think tanks 40

wage legislation 23 Wallerstein, M. 44, 47, 72 Washington DC 26 Weber, M. 53 welfare state 7 Wildavsky, A. 115 Wilson 8, 11, 30, 31 Wilson, Charles E. 61–2 Wilson, F.L. 98 Wilson, G.K. 95 Wilson, Harold 135 Wilson, J.Q. 158 Wilson, R. 59, 62, 70 Wittenberg, J. 158 women 1, 3, 4 World Economic Forum (WEF) 121 World Trade Organization 121 Wright, J.R. 55, 142 Zhang, I.X. 143