Global Stakeholder Democracy: Power and Representation Beyond Liberal States

GLOBAL STAKEHOLDER DEMOCRACY This page intentionally left blank Global Stakeholder Democracy Power and Representati...

Author: Terry Macdonald

45 downloads 1047 Views 976KB Size Report

This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!

Report copyright / DMCA form

DOWNLOAD PDF

GLOBAL STAKEHOLDER DEMOCRACY

This page intentionally left blank

Global Stakeholder Democracy Power and Representation Beyond Liberal States

T E RRY MACD ONA LD

1

3 Great Clarendon Street, Oxford ox2 6dp Oxford University Press is a department of the University of Oxford. It furthers the University’s objective of excellence in research, scholarship, and education by publishing worldwide in Oxford New York Auckland Cape Town Dar es Salaam Hong Kong Karachi Kuala Lumpur Madrid Melbourne Mexico City Nairobi New Delhi Shanghai Taipei Toronto With oﬃces in Argentina Austria Brazil Chile Czech Republic France Greece Guatemala Hungary Italy Japan Poland Portugal Singapore South Korea Switzerland Thailand Turkey Ukraine Vietnam Oxford is a registered trade mark of Oxford University Press in the UK and in certain other countries Published in the United States by Oxford University Press Inc., New York © Terry Macdonald 2008 The moral rights of the author has been asserted Database right Oxford University Press (maker) First published 2008 All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, without the prior permission in writing of Oxford University Press, or as expressly permitted by law, or under terms agreed with the appropriate reprographics rights organization. Enquiries concerning reproduction outside the scope of the above should be sent to the Rights Department, Oxford University Press, at the address above You must not circulate this book in any other binding or cover and you must impose the same condition on any acquirer British Library Cataloguing in Publication Data Data available Library of Congress Cataloging-in-Publication Data Macdonald, Terry. Global stakeholder democracy : power and representation beyond liberal states / Terry Macdonald.—1st published 2008. p. cm. Includes index. ISBN–13: 978–0–19–923500–1 1. Democracy. 2. Power (Social sciences) 3. Representative government and representation. 4. Political participation. I. Title. JC423.M153 2008 321.8—dc22 2008011601 Typeset by SPI Publisher Services, Pondicherry, India Printed in Great Britain on acid-free paper by Biddles Ltd., King’s Lynn, Norfolk ISBN 978–0–19–923500–1 1 3 5 7 9 10 8 6 4 2

For Tom and Zoe

This page intentionally left blank

Acknowledgements Many thanks go to Andrew Hurrell, David Miller, Henry Shue, Charles Beitz, Thomas Pogge, Kate Macdonald, and an anonymous reviewer for OUP, who have read drafts of this book at various stages of its development and provided incisive and challenging criticisms and comments. Thanks also go to Patti Lenard, Zoﬁa Stemplowska, Jonathan Symons, Daniele Archibugi, participants in the Nuﬃeld College Political Theory Workshop, and several former colleagues at Merton College, Oxford, and the Centre for Applied Philosophy and Public Ethics (CAPPE), at the ANU—all of whom have contributed stimulating conversations and valuable insights along the way. I am grateful also to Nuﬃeld College, Oxford, Merton College, Oxford, and CAPPE, ANU, for providing me with such supportive and motivating working environments during my research on this book. Finally, heartfelt thanks go to my family for their support, and in particular to Tom, for his generosity and patience as I have worked to complete this book. Terry Macdonald

This page intentionally left blank

Contents Introduction The issues The project The strategy The argument The limitations

1 1 4 7 12 16

PART I. DEMOCRATIC BOUNDARIES IN THE NEW GLOBAL POLITY 1. Democracy Beyond ‘Closed’ Societies 1.1. Introduction 1.2. The standard theoretical model of a ‘closed’ democratic society 1.3. Beyond ‘closed’ societies: A ‘realist’ democratic agenda 1.4. Public power and political participation in a ‘pluralist’ democratic order 1.5. Conclusions: Towards a ‘pluralist’ democratic global order

21 21 23 29 35 40

2. Public Power Beyond ‘Sovereign’ States 2.1. Introduction 2.2. Assessing arguments for state-based accounts of public power 2.3. Public power and the institutional practice of law-making 2.4. Public power and the use of force 2.5. Public power and centralized or ‘constitutionalized’ decision-making structures 2.6. Conclusions

43 43

3. The Public Power of NGOs in Global Politics 3.1. Introduction 3.2. Public power and inﬂuence in regulative social norm-building 3.3. Public power and the imposition of material constraints 3.4. Conclusions

62 62

4. From Nation-States to ‘Stakeholder’ Communities 4.1. Introduction 4.2. The idea of democratic ‘stakeholder’ communities 4.3. Delineating stakeholder communities in practice: stakeholders in the public power of NGOs 4.4. Conclusions

83 83 85

44 45 51 52 61

63 71 81

92 101

x

Contents PART II. DEMOCRATIC REPRESENTATION IN THE NEW GLOBAL POLITY

5. Global Social Choice Beyond Nation-State Representation 5.1. Introduction 5.2. A critique of social choice through ‘national’ representation 5.3. A theoretical critique of the justiﬁcatory Burkean framework 5.4. A critique of global social choice through state representation 5.5. A ‘liberal individualist’ justiﬁcation for state representation 5.6. Conclusions

105 105 107 117 120 122 138

6. Global Social Choice Through Multi-Stakeholder Representation 6.1. Introduction 6.2. The multi-stakeholder model of global constituencies 6.3. ‘Liberal pluralist’ interest representation as a justiﬁcation for the multi-stakeholder model 6.4. Strengths and weaknesses of the multi-stakeholder model of global representation 6.5. Conclusions: The need for a hybrid model of global social choice

139 139 140 146 150 160

7. Theorizing Global Representative Agency: Non-Electoral Authorization and Accountability 7.1. Introduction 7.2. Elections as mechanisms of stakeholder control 7.3. Authorization and accountability as mechanisms of democratic stakeholder control 7.4. The institutional mechanisms of authorization and accountability 7.5. Conclusions

170 177 192

8. Instituting Global Representative Agency: The Authorization and Accountability of NGOs 8.1. Introduction 8.2. The democratic authorization of NGOs in global politics 8.3. The democratic accountability of NGOs in global politics 8.4. The challenge of democratic equality 8.5. Conclusions

193 193 194 211 218 220

Conclusion

223

Index

231

163 163 165

Introduction THE ISSUES In November 1999 the streets of Seattle became the scene of huge public demonstrations against what protest organizers called ‘one of the most powerful, secretive, and anti-democratic bodies on Earth’. 1 The protestors’ target was the Ministerial Meeting of the World Trade Organization (WTO), which was to have launched the much-vaunted ‘Millennium Round’ of trade talks among the organization’s 135 member states. Instead, the talks collapsed under the shadow of the public demonstrations. On the ﬁrst day of the scheduled talks, an estimated 50,000 demonstrators ﬂooded the streets of Seattle. By evening, a civil emergency had been declared and hundreds arrested, but protesters continued to deliver their message to conference organizers: ‘This is what democracy looks like,’ they chanted, as they encircled the city’s prison to demand the release of the 500 protesters detained inside. 2 The demonstrations were organized by an international coalition of NonGovernmental Organizations (NGOs), representing a wide spectrum of divergent interests. While their objectives in relation to trade policy were often conﬂicting, many of the demonstrators were united by their demands for greater access to and inclusion within the WTO’s decision-making processes, framed in terms of calls for ‘a global citizen-based and citizen-driven democratic order’. 3 The Seattle protests brought the issue of global democracy for the ﬁrst time into the glare of prime-time global news. As such, they represented a watershed for the broad-based social movement seeking to salvage the democratic project in the context of the complex technological, economic, political, and social transformations commonly referred to as ‘globalization’. In recent years, networks of political activists have mounted various high-proﬁle campaigns seeking greater democratization of powerful international institutions such as the World Bank and the IMF, and increased ‘corporate accountability’ for the many powerful economic entities in the so-called ‘private’ sector. In part as a response to such political demands, many key global decision-makers—including national governments 1 ‘International Forum on Globalization’, Statement by Jerry Mander, President of the International Forum on Globalization, at a Press Conference Regarding the WTO, 4 November (1999 [cited February 2002]); available from http://www.ifg.org/media.html 2 John Vidal, ‘Real Battle for Seattle’, The Observer, 5 December 1999. 3 Vandana Shiva, The Historic Signiﬁcance of Seattle (1999 [cited February 2002]); available from http://www.wtowatch.org/library/admin/uploadedﬁles/Historic_Signiﬁcance_of_Seattle_The.html

2

Introduction

as well as leaders of international organizations, corporations, and NGOs—now recognize the need for democratic reform to ensure the stability, eﬀectiveness, and legitimacy of global decision-making. Although the Seattle protests thus served on one level to bolster NGO demands for the democratization of powerful international organizations such as the WTO, they had the additional eﬀect of highlighting NGOs’ own increasing power on the global stage, and raising parallel questions about the democratic legitimacy of NGOs themselves. The rise of NGOs as a force in global politics has been underway for many years; between 1960 and 1995, the number of international NGOs grew from roughly 1,300 to over 36,000, 4 and by 1999, the UNDP estimated the number of such NGOs to be around 44,000. 5 By 1996, the United Nations had adopted extensive formal provisions for NGO consultation and participation within UN institutions, 6 and these provisions served as part of a wider agenda for accommodating NGOs as recognized participants within a range of international institutions. 7 The inﬂuence of NGOs has also grown within the corporate sector, as many corporations have attempted to enhance their perceived public legitimacy by establishing ‘partnerships’ or ‘stakeholder dialogues’ with NGOs. These ‘partnerships’ and ‘dialogues’ entail various consultation mechanisms which grant NGOs some input into corporate decision-making in issue areas such as human rights and environmental policy. 8 In recognition of this growing role, NGO leaders were invited in 2000 to participate for the ﬁrst time in World Economic Forum discussions alongside political and business leaders. Moreover, in addition to these increasing roles within such high-level decision-making processes, many NGOs now exercise signiﬁcant forms of power through operational roles in various development and humanitarian contexts, in which they commonly engage in important public decision-making about issues such as service-provision and resource-distribution within communities. The political momentum behind NGOs’ political demands for participation in these various forms of public decision-making in global politics has been strengthened by high levels of public trust in NGOs, which in many parts of the world have begun to rival levels of public trust in both business and government. Indeed, 4 Boutros Boutros-Ghali, ‘An Agenda for Democratization: Democratization at the International Level’, in Barry Holden (ed.), Global Democracy: Key Debates (London: Routledge, 2000), 111. 5 United Nations Development Programme, Human Development Report 2000 (2000 [cited February 2005]); available from http://hdr.undp.org/reports/global/2000/en/pdf/hdr_2000_ch0. pdf p. 8. 6 See United Nations Secretary-General, Arrangements and Practices for the Interaction of NonGovernmental Organizations in All Activities of the United Nations System (United Nations, 1998 [cited February 2002]); available from http://www.globalpolicy.org/ngos/docs98/koﬁ998.htm 7 For a fuller discussion of these developments, see Boutros Boutros-Ghali and United Nations Dept. of Public Information, An Agenda for Democratization (New York: United Nations Dept. of Public Information, 1996); Daniele Archibugi, Sveva Balduini, and Marco Donati, ‘The United Nations as an Agency of Global Democracy’, in Holden (ed.), Global Democracy: Key Debates. 8 See, e.g., Kelly Currah, ‘How Corporations Absolve Their Sins’, Guardian, 28 August 2000; Andy Rowell, ‘Corporations “Get Engaged” to the Environmental Movement’, PR Watch Archives, 8/3 (2001).

Introduction

3

a recent survey of public opinion leaders in the USA, Europe, and Australia measured public trust in government, business, media, and NGOs, and found that NGOs are beginning to surpass all other institutions in terms of their public credibility. 9 An author of the study, Richard Edelman, told political and business leaders in a speech to the World Economic Forum in New York that NGOs are now the Fifth Estate in global governance—the true credible source on issues related to the environment and social justice. . . . Now companies must communicate with multiple stakeholders, especially NGOs, with speed, transparency, and an oﬀer of interactivity. 10

The growing inﬂuence of NGOs is underpinned by their perceived legitimacy, and this derives to some degree from their claims to serve as democratic representatives of global peoples. Consistent with this, it is commonly argued that as processes of ‘globalization’ transfer public political decision-making capacities from statebased to intergovernmental and corporate decision-making processes, NGOs are able to play an important democratic role as civic representatives in these new political domains. Many global leaders have responded to such arguments by embracing some of the claims to democratic legitimacy made by NGOs and their public supporters. Over the last decade or so, democratic arguments in favour of greater ‘civil society’ participation in global decision-making have been advanced at the highest level of international leadership. 11 Former UN Secretary-General Boutros BoutrosGhali has argued that NGOs are ‘a basic form of popular representation in the present-day world’, and that accordingly ‘[t]heir participation in international relations is, in a way, a guarantee of the political legitimacy of those international organizations.’ 12 Former United Nations (UN) Secretary-General Koﬁ Annan has also added his voice to the call for greater civil society participation, maintaining that for the UN and other international organizations ‘a partnership with civil society is not an option; it is a necessity.’ 13 Despite the rhetorical endorsement of these global leaders, there is still much scepticism about the democratic legitimacy of NGOs. Many critics of NGO participation challenge the legitimate status of NGO constituencies, labelling them 9 Edelman PR Worldwide, Edelman Trust Barometer 2005: The Sixth Global Opinion Leaders Study (January 2005 [cited March 2005]); available from http://edelman.com/image/insights/content/ Edelman_Trust_Barometer-2005_ﬁnal_ﬁnal.pdf 10 Edelman Survey: NGOs Trusted More Than Business or Governments (2002 [cited February 2005]); available from http://solidar.org/Document.asp?DocID=3598&tod=487 11 Prominent discussions include Commission on Global Governance, Our Global Neighbourhood: The Report of the Commission on Global Governance (Oxford: Oxford University Press, 1995); Commission on Global Governance, The Millennium Year and the Reform Process (1999 [cited February 2002]); available from www.cgg.ch/millenium.htm Boutros-Ghali and United Nations Dept. of Public Information, An Agenda for Democratization. 12 Quoted in Riva Krut, Globalization and Civil Society: NGO Inﬂuence in International DecisionMaking (Geneva: United Nations Research Institute for Social Development, 1997), 18. 13 United Nations, Partnership with Civil Society Necessity in Addressing Global Agenda, Says Secretary-General in Wellington, New Zealand Remarks (Press Release) (2000 [cited February 2002]); available from www.un.org/News/Press/docs/2000/20000229.sgsm7318.doc.html

4

Introduction

pejoratively as ‘interest groups’ with no special entitlement to representation. Critics of NGO participation also question how ‘representative’ NGOs actually are of the constituencies for whom they purport to speak. They assert that NGOs are unelected and unaccountable, arguing that unlike national governments, NGOs ‘need not answer to the broad public they claim to represent’. 14 Consequently, some commentators dismiss them as ‘self-appointed spokesmen for their cause’ 15 or in even more unsympathetic terms as ‘a self-serving coterie of elitists’. 16 Advocates of NGO participation therefore face the further challenge of establishing that NGOs are in some democratic sense ‘representative’ of the constituencies that can legitimately claim a voice in global decision-making, whichever these may be agreed to be. Since the Seattle protests, the general challenge of building global democracy and these more speciﬁc democratic challenges posed by NGO power have been closely intertwined in the global democratic consciousness. Reﬂecting further upon this connection is potentially of great theoretical utility, since the controversies surrounding the democratic legitimacy of NGOs, and wider debates about the democratization of powerful international organizations, transnational corporations (TNCs), and other global actors, are underpinned by parallel theoretical dilemmas and practical diﬃculties. The challenges associated with democratizing this wider range of global actors—like the speciﬁc challenge of democratizing NGOs—can best be met by moving away from traditional state-based models of democracy, and better coming to terms with the complex normative implications of institutional pluralism. Questions about the democratic legitimacy of NGOs are therefore of great signiﬁcance not only because of the considerable power exercised by NGOs themselves in the global domain. Even more importantly, global NGOs are of signiﬁcant democratic interest because they provide an ideal political lens through which to examine the larger theoretical and institutional questions of how we could conceivably build democratic institutions beyond liberal states, which have traditionally provided the central focus of democratic aspiration and critique.

THE PROJECT Despite the prominence of these democratic issues on recent global political agendas, they have to date received surprisingly little analytic attention from normative political theorists. Much of the existing literature on global political legitimacy is framed as ‘ideal’ theory, and accordingly focused on abstract questions of rights and duties, rather than on the ‘non-ideal’ questions of how political power should 14 Jenny Bates, ‘The Third Sector in Global Aﬀairs: Civil Society Deserves a Voice, but Not a Vote, in International Institutions’, New Democrat (2000). 15 Bruce Stokes, ‘New Players in the Trade Game’, National Journal, 18 December 1999. 16 Pranay Gupte, ‘Whose Cause Is It, Anyway?’, Newsweek, 6 December 1999.

Introduction

5

be constrained through representative institutions. Moreover, the widespread preoccupation of post-Rawlsian political theory with issues of distributive justice has directed the focus of much recent work on global political legitimacy, reinforcing the tendency of authors to neglect traditional democratic issues of power and representation in favour of issues related to the distribution of social and economic goods. Several works have proved exceptions to this general theoretical tendency, and have directly tackled the political questions of power and representation in global politics. However, these have often tended to proceed by extrapolating institutional schemes from the domestic to the global level, and in doing so have not dealt satisfactorily with the distinctive organizational pluralism of the global polity. Political theorists have been for so long accustomed to preoccupation with the state (and its associated ideas of ‘sovereign’ or ‘constitutionalized’ public power) that it has thus far proven diﬃcult to develop adequate theoretical responses to the powerful new roles of non-state actors on the global stage. The most prominent theoretical alternatives that have been oﬀered to the statism of ‘communitarian’ democrats have been various models of ‘cosmopolitan’ democracy. 17 These ‘cosmopolitan’ theories are extremely valuable in highlighting the normative and practical imperatives for adopting a global perspective on the democratic project. They are also valuable in exploring and articulating legitimate directions for future global institution-building. Although it may not be realistic to implement cosmopolitan democratic blueprints in full in the near future, some elements of them may be feasibly achievable. Moreover, ideal blueprints of this kind are helpful for orienting more long-term reform processes, even if they cannot be realized in the foreseeable future. As guides to contemporary projects of global institutional reform, however, their utility is limited in two key ways: ﬁrst, by their universalist ideals of a uniﬁed global polity, which ﬁt poorly with the fragmentation of existing global society; and second, by the constitutionalized structure of their institutional prescriptions, which I will argue in Chapter 1 has minimal prospects of implementation within a proximate time frame. My central project in this book is to develop a theory of global democracy that can transcend the existing impasse between ‘communitarian’ statism and ‘cosmopolitan’ universalism, and deal more satisfactorily with the organizational pluralism of contemporary global politics. The theoretical arguments I develop here share much more in common with cosmopolitan than with communitarian democratic theories, and can be understood as an attempt to adapt rather than to overturn the broader cosmopolitan democratic project. But as will become clear as my arguments unfold, my pluralist institutional framework of stakeholder 17 For a prominent ‘communitarian’ approach to the problem of global democracy, see David Miller, On Nationality, Oxford Political Theory (Oxford: Clarendon Press, 1995); for ‘cosmopolitan’ approaches, see David Held, Democracy and the Global Order: From the Modern State to Cosmopolitan Governance (Cambridge: Polity, 1995); Andrew Kuper, Democracy Beyond Borders: Justice and Representation in Global Institutions (Oxford: Oxford University Press, 2004); and Daniele Archibugi and David Held (eds.), Cosmopolitan Democracy: An Agenda for a New World Order (Cambridge: Polity, 1995).

6

Introduction

democracy diverges in normatively and practically signiﬁcant ways from some prominent cosmopolitan institutional models. Throughout the book, I develop the stakeholder model of global democracy by systematically reﬂecting upon the democratic dilemmas surrounding the power of NGOs in contemporary global politics. One obvious advantage of focusing the analysis on NGOs is that doing so can help to answer important practical normative questions about how democratic policymakers should assess and respond to the power exercised by NGOs—and by extension, other non-state actors—within contemporary global politics. In addressing these practical normative challenges, this book seeks to counter some of the sceptical challenges to the legitimacy of non-state actors, and to make sense of certain widespread intuitions that the political activities of non-state actors need not threaten democracy, but rather have the potential to satisfy certain rigorous democratic standards, if the right institutional reforms are undertaken. The other advantage of focusing analysis on NGOs in this way is that it can help us to develop more general democratic concepts and institutions that are suitable for our ‘pluralist’ global polity, by forcing us to rethink some standard state-centric assumptions about the fundamental elements of a democratic system. When we examine widespread intuitions about the way in which NGOs can attain democratic legitimacy in global public decision-making, it becomes evident that many of the theoretical conceptions invoked in these democratic arguments diverge signiﬁcantly from those inherent in traditional, state-centric, democratic models. First, the ‘public’ forms of power that NGOs exercise in global decision-making, and which common intuitions suggest should be subject to democratic standards of legitimacy, are quite diﬀerent in character from the ‘governmental’ forms of public power that are the focus of conventional state-centric democratic theories. The governmental forms of power that we traditionally see as the rightful focus for democratic scrutiny are generally underpinned by the use or threat of force, are invested with law-making capacities, and to a signiﬁcant degree are organizationally centralized, or at least constitutionalized. In contrast, many of the ‘public’ forms of NGO power that attract democratic criticism display few or none of these state-like characteristics. Second, the ‘constituencies’ that NGOs commonly claim to represent are quite a diﬀerent kind of entity from the territorially bounded populations that usually serve as electoral constituencies within democratic nation-states. Significantly, NGO constituencies—such as religious communities, indigenous peoples, women, environmental and human rights advocates, trade unionists, and business and professional communities—tend in general to be non-national and non-territorial in character, and moreover tend to overlap in terms of their individual memberships. Third, the kind of democratic ‘representation’ in which NGOs could plausibly claim to be engaged is quite diﬀerent in structure from the kind associated with representative institutions within democratic states. Whereas representation in state democracies is characterized primarily in terms of mechanisms associated with formal election procedures, NGOs can claim no such electoral mandate.

Introduction

7

Therefore, if it is to be plausibly argued that NGOs are ‘representative’ of their various constituencies, it is necessary to invoke a conception of representation that does not rely on the formal mechanisms of electoral authorization and accountability. In this light, it is evident that if we are to succeed in constructing a coherent framework for understanding the legitimate democratic role of NGOs in global decision-making, we must in the process confront fundamental theoretical questions about the character and scope of democratic public power, the basis for delineating democratic boundaries, and the institutional properties of democratic representation. We cannot merely seek to apply old theoretical conceptions of ‘governments’, ‘peoples’, and ‘representation’ to the new political phenomenon of global NGOs. Rather, we must re-examine the theoretical conceptions themselves, and consider how they must be reformulated if they are to accommodate the emerging norms and social realities of a globalizing world in which NGOs— as well as other non-state actors—play an increasingly signiﬁcant role. For this reason, an examination of the democratic debates surrounding NGOs can help to bring the broader theoretical questions of global democracy into focus, and chart a productive route to the development of a wider theoretical argument which could potentially be applied to a more extensive range of global actors. Through reﬂecting upon the democratic dilemmas surrounding the political power of global NGOs in this way, my arguments in this book challenge the widespread assumption that ‘sovereign’ power, nationally or territorially bounded societies, and ‘electoral’ processes are essential institutional foundations of a democratic system, serving, respectively, as the subjects, agents, and means of democratic control. In this book, I present a critique of these traditional democratic ideas, and proceed to rethink the democratic project from its conceptual foundations, posing the questions: What needs to be controlled? Who ought to control it? How could they do so? In the course of answering these questions, I develop a distinctive theoretical model of representative democracy based on the new concepts of ‘public power’, ‘stakeholder communities’, and ‘non-electoral representation’.

THE STRATEGY Before outlining in more depth the substance of these practical and theoretical arguments, it is helpful to say a little more about the ‘contextual’ method of theory-building I employ. By developing my theoretical arguments via examination of policy-debates surrounding the democratization of NGOs, I employ what can be described as a kind of contextual ‘reﬂective equilibrium’. The idea of reﬂective equilibrium has been employed by many political theorists since John Rawls to denote the outcome of a process of developing normative principles by making mutual adjustments between intuitions (or ‘considered judgements’, in Rawls’s terms) and some kind of ideal conceptions, or abstract theoretical

8

Introduction

formulations. The ‘correct’ principles, which are derived through this process of mutual adjustment, are those that most coherently reconcile our abstract ideal conceptions with our more speciﬁc intuitions, or judgements. The contextual method of seeking reﬂective equilibrium is distinctive insofar as it requires that the speciﬁc intuitions with which we attempt to reconcile our ideal theories should be derived not from hypothetical scenarios or thought experiments, of the kind favoured by many political theorists, but rather from concrete political problems, entangled as they are in all the complexity, ambiguity, and emotion of the real political world. 18 The primary advantage of this approach is that it best enables our political theories to fulﬁl their central function: to guide us about how to act politically, in constructing our political institutions and formulating our political policies. We do not need to know how we should act in hypothetical worlds, but rather how we should act in our own world, in the present time, confronted as we are with the concrete imperatives and choices imposed by political life. Principles derived directly from, and addressed directly to, this world will be more useful to us than principles derived and delivered behind veils of abstraction. John Rawls, a prominent advocate of more abstract forms of normative reasoning, claims that whereas the politician looks to the next election, and the statesman to the next generation, the philosopher should look to ‘the indeﬁnite future’. 19 The contextualist philosopher, more modestly, looks to the foreseeable future, since the practical problems that lie beyond this horizon are unknowable, while the problems within it are all too readily accessible. Principles derived contextually in this way will be more useful than those derived through more abstract reasoning processes for three key reasons. First, the meanings and prescriptive implications of our abstract theoretical formulations will be clearer, as they will have been interpreted, applied, and reformulated with respect to the very problems with which we are grappling, and on which we must make decisions and take action. Second, contextually derived principles will be more useful because they will embody normative insights that are implicit in institutions and practices but obscured through the process of theoretical abstraction. 20 Third, they allow us to incorporate more readily into our normative analysis the empirical assessments and constraints that are most evident in analysis of real rather than hypothetical examples, and that must always be recognized and accommodated within any practically viable and stable normative structures. 21 18 Iris Young calls this kind of historically and socially contextualized theorizing ‘critical theory’; see Iris Marion Young, Justice and the Politics of Diﬀerence (Princeton: Princeton University Press, 1990). 19 John Rawls, ‘The Idea of an Overlapping Consensus’, Oxford Journal of Legal Studies, 7 (1987), 24. 20 Joseph Carens makes similar points in his defence of a contextual method in political theory in Joseph H. Carens, Culture, Citizenship, and Community: A Contextual Exploration of Justice as Evenhandedness (Oxford: Oxford University Press, 2000). 21 Here I mean ‘stable’ in the colloquial political sense (incorporating elements of both normative agreement and strategic ‘modus vivendi’), rather than in the exclusively normative sense in which Rawls employs the term. See John Rawls, Political Liberalism, New edn. (New York: Columbia University Press, 1996).

Introduction

9

In relation to my project here, it can be seen that deriving democratic principles through a process of mutual adjustment between our intuitions about the legitimate political roles of NGOs, on the one hand, and our abstract theoretical conceptions of ‘public power’, ‘constituencies’, and ‘representation’, on the other, will generate principles that are of greater practical utility and normative validity than principles generated through more abstract reasoning processes. Principles generated in this fashion are inevitably more pertinent to the decisions we have to make about the problem of NGOs; they embody normative insights that are implicit in the actual practices that we employ to deal with NGOs in the real world; and they can accommodate and function within the empirical constraints inherent in the contemporary global order. I also need to explain here my choice of which (and whose) ideal theoretical conceptions should be invoked to engage our intuitions within a process of reﬂective equilibrium. As already noted, the abstract conceptions that I invoke and articulate in this book are ideals of democracy (and their constituent conceptions of public power, constituencies, and representation) that are embedded within real discourses about political legitimacy in global politics. These ideals are embedded to some extent in certain philosophical writings addressing the subject of global political legitimacy, but more importantly they are embedded also in the public and intra-organizational debates about legitimacy in which a wide range of global political practitioners are engaged. They have signiﬁcant currency within various international organizations (such as the UN and the World Bank), within many NGOs and in parts of the wider NGO community, among a broad range of grassroots citizen groups and public commentators, and even to some smaller degree among some state governments and transnational corporations. In the course of developing my democratic arguments in this book, I draw upon some of these practitioners’ contributions to the debate, although the sources I cite here only scratch the surface of this vast public discourse. I certainly do not want to make the claim that the ideal of Global Stakeholder Democracy is universally endorsed by participants within these public discourses on global political legitimacy. Rather, these discourses are composed of a jumbled mix of diﬀerent and incompatible norms and ideals, and characterized by a high degree of normative disagreement and disorientation. In this context, no single normative ideal could claim universal endorsement within global public discourse—neither Global Stakeholder Democracy nor any of its competitors. My claims about the connections between the Stakeholder model of global democracy and these public discourses about global political legitimacy are more modest: the model represents an attempt to draw out a range of concepts and norms with wide currency in these discourses, and tie them together within a coherent and potentially workable democratic framework. Once this model has been properly ﬂeshed out, and its various implications more fully revealed, it may well be that some of those who had unreﬂectively invoked some central concepts and norms on which the stakeholder model is based would reject it in favour of more conventional democratic models. But my hope is that demonstrating how these concepts and norms can form the basis of a coherent and workable democratic model

10

Introduction

will have the overall eﬀect of strengthening commitment to these ideas within global society. Either way, the project of articulating the central norms of Global Stakeholder Democracy should be a useful one, since it can help build a better understanding of what is at stake, normatively and practically, in the political contests among competing norms of global political legitimacy. We can see the democratic norms identiﬁed through the process I have described as ‘free-standing’ in the process of global public justiﬁcation, rather than derivative of or subordinate to some wider normative framework. Our global democratic values can be seen as ‘free-standing’ in two senses. First, they are ‘freestanding’ in terms of the source of their justiﬁcation; this means that democratic arguments can be publicly evaluated solely on the basis of their ‘ﬁt’ within the limits of public democratic discourse, in the manner that I have described. 22 Second, they are ‘free-standing’ in terms of the scope of their application; this means that global democratic arguments can be distinguished from other normative arguments with currency in certain global discourses (such as arguments about global ‘justice’, social ‘utility’, or some other more ‘comprehensive’ moral standard) in accordance with the particular range of human actions or institutions that each seeks to regulate within the global domain. Viewing global democratic principles as wholly ‘free-standing’ in this second sense is somewhat unconventional in the existing philosophical literature, because of a common tendency to subsume democratic arguments within broader frameworks of ‘global justice’. 23 Philosophers often observe that whereas ‘comprehensive’ or ‘ethical’ principles apply to all domains of individual action and social interaction, principles of ‘global justice’ apply only to some more limited domain, commonly characterized as a global ‘basic structure’, or something similar. 24 However, the analogous question as to the particular range of activities and relationships that is the rightful domain of democratic principles (as distinct from principles of ‘justice’) is rarely examined, perhaps because the question is assumed to be unproblematic for one of two reasons. First, the question of the scope of democratic principles is sometimes avoided by viewing democratic principles as secondary norms of ‘procedural justice’, derived from a broader theory of social justice. 25 By viewing principles of justice 22 Any further disagreements must then be resolved through political contestation rather than public justiﬁcation; by ‘political’ contestation here I refer to opposing assertions of power in pursuit of both conﬂicting strategic interests and conﬂicting moral commitments. 23 David Held, for example, suggests that global democratic theory should be grounded in a wider cosmopolitan theory of global justice in this way. See Held, Democracy and the Global Order: From the Modern State to Cosmopolitan Governance; David Held, ‘Law of States, Law of Peoples’, Legal Theory, 8 (2002); David Held and Mathias Koenig-Archibugi, Global Governance and Public Accountability (Oxford: Blackwell, 2005). 24 See Rawls, Political Liberalism; Thomas Pogge, Realizing Rawls (Ithaca: Cornell University Press, 1989); Charles R. Beitz, Political Theory and International Relations (Princeton: Princeton University Press, 1999). 25 Thomas Pogge characterizes democracy as ‘procedural justice’ in this way in Pogge, Realizing Rawls. John Rawls was the ﬁrst to make this move, by grounding his theory of justice in the values shared by citizens of a democratic society. See Rawls, Political Liberalism.

Introduction

11

as those regulative of the ‘basic structure’ of a paradigmatically closed democratic society, traditional democratic values of ‘political representation’, ‘political equality’, and so on are simply assimilated within the theory of social justice, such that they need not be accorded any independent scope or subject matter. 26 Second, to the extent that these more traditional democratic values are ascribed an independent subject, or domain, these values are commonly assumed to be selfevidently concerned with structures of ‘governmental’ decision-making within the (settled) constitutional constraints of a liberal state. 27 Both of these responses to the problem are unsatisfactory, however, when we are seeking to apply democratic principles to the domain of global politics. The problem with the former response—that of subsuming democratic principles within a theory of social justice—is that it limits democratic discourse to the conﬁnes of the ‘closed’ societies that have the kinds of intransient memberships and shared institutions that provide settings for wider systems of social justice. 28 Although some theorists of justice argue that global politics incorporates such a ‘global society’ along with such appropriate ‘institutional schemes’, 29 this position is sustainable only in limited contexts and respects, because of the unevenness of global solidarities and institutions. 30 Consequently, an insistence that traditional democratic principles cannot apply outside broader institutional systems of social justice would impose signiﬁcant theoretical limitations upon the scope of application of these principles in global politics. In practice, however, I would maintain that democratic values such as political representation and political equality are in fact publicly embraced and advanced by political actors in various kinds of relationships, independently of claims about the appropriate scope of 26 Charles Beitz refers to this democratically grounded theory of justice as ‘democratic theory in the broader sense’. See Charles R. Beitz, Political Equality: An Essay in Democratic Theory (Princeton: Princeton University Press, 1989), ix. Rawls takes this approach to the assimilation of theories of justice and democracy most clearly in his later work. See Rawls, Political Liberalism. 27 Charles Beitz refers to these more traditional issues of democratic theory as ‘democratic theory in the narrower sense’ (as distinct from the former ‘democratic theory in the broader sense’), and then makes precisely this assumption about the scope of these ‘narrower’ democratic values. See Beitz, Political Equality: An Essay in Democratic Theory, x–xi. Rawls appears to view the scope of these traditional democratic principles in this way in his early writing, insofar as he presents democratic mechanisms of public political control as subordinate to the constitutional principles of the liberal state, which, in turn, rest on a comprehensive doctrine of individualist liberalism. See John Rawls, A Theory of Justice, Rev. edn. (Oxford: Oxford University Press, 1999). 28 The precise role of pre-existing social institutions in generating obligations of justice is a disputed subject among theorists of global justice. Most agree that the existence of certain shared institutions is not a suﬃcient condition for principles of justice to apply, since there must also be a commitment to the more general cosmopolitan moral solidarism necessary to attribute moral signiﬁcance to institutions. Within such a broadly cosmopolitan moral framework, however, various substantive moral ideas that most think to be entailed in the ideal of justice (such as reciprocity and desert) can provide grounds for designating certain institutions as having a special role in generating substantive duties of justice towards others. See, e.g., Pogge, Realizing Rawls; Beitz, Political Theory and International Relations. 29 See Thomas W. Pogge, ‘An Egalitarian Law of Peoples’, Philosophy and Public Aﬀairs, 23/3 (1994); Allen Buchanan, ‘Rawls’s Law of Peoples: Rules for a Vanished Westphalian World’, Ethics, 110/4 (2000); Beitz, Political Theory and International Relations. 30 I discuss these issues further throughout Part I of this book.

12

Introduction

wider distributive principles of social justice. In the normative discourses of global politics, these distinctively democratic values thus possess a kind of ‘free-standing’ quality, which compels us to articulate and apply them as ‘free-standing’ at a theoretical level. The second common theoretical response to the question of the scope of democratic discourse—that of linking its scope to the structures of ‘governmental’ decision-making within a constitutional liberal state—is similarly untenable in the context of global politics. Although this solution to the problem may seem satisfactory to those democrats who are concerned only with the kinds of democratic dilemmas that arise within liberal states, it cannot suﬃce for democrats concerned with the problem of democracy in global politics, where many fewer state-like constitutional structures exist. As with the former common theoretical response, this latter view of the scope of democratic principles would imply the need for signiﬁcant restrictions on the application of democratic principles in global politics, given the limited scope of state-like constitutional structures and governmental agents in that domain. Thus, like the former view, this latter view is belied by the extent to which democratic principles are in practice invoked and applied in non-state contexts and to non-governmental actors in global politics. These objections to the standard theoretical viewpoints concerning the appropriate scope of democratic principles do not of course answer the question of how we should delineate this scope; ultimately this is not a methodological question but a normative one, and as such I address it in some depth in the course of developing my substantive argument in Part I of the book. These objections do, however, help to explain both how it is possible to construct a democratic argument that is ‘free-standing’ and independent of wider arguments about social justice or other moral obligations and also why we should think that it is desirable to approach democratic theory in this way, in the context of global politics.

THE ARGUMENT The argument presented in this book in response to these practical and theoretical challenges is developed in two parts, addressing the two key conceptual elements of a theory of global representative democracy. In Part I (Chapters 1, 2, 3, and 4), I tackle the task of delineating the legitimate boundaries of democratic societies in global politics, and identifying the legitimate subjects and agents of democratic control. I then turn in Part II (Chapters 5, 6, 7, and 8) to the problem of how to construct legitimate institutions for representative decision-making within such societies. In Chapter 1, I begin the task of developing a new account of legitimate global institutional boundaries by arguing that we need to abandon the assumption,

Introduction

13

entrenched within traditional liberal democratic theory, that democracy must take place within a ‘closed’ society. By ‘closed’, I mean to denote a society constituted, on the one hand, by a uniﬁed agent of public power advancing collective purposes as its subject of democratic control and, on the other hand, by a community of ‘citizens’ (sharing an exclusive and all-encompassing political status of civic membership) as its agent of democratic control. The paradigmatic case of the former is a state, and the paradigmatic case of the latter is a nation (or ‘people’, or similar social entity). Instead, we should embrace the prospect of a ‘pluralist’ liberal democratic order in global politics, composed of multiple agents of public power held to account by their multiple overlapping ‘stakeholder’ communities. What this requires, I argue, is that multiple state and non-state agents of public power in global politics should be identiﬁed as the legitimate subjects of democratic control in accordance with their autonomy-constraining impact on populations. Correspondingly, individual ‘stakeholders’ should be accorded participatory entitlements as the legitimate agents of democratic control, in accordance with the impact of this public power upon their lives. In Chapter 2, I develop and further defend my central proposition that what I call ‘public power’ is the legitimate subject of global democratic control. The material in this chapter is aimed at sceptical readers, who are not prepared to relinquish standard state-based accounts of public power on the basis of the arguments presented in Chapter 1. My additional arguments here attempt to counter some plausible theoretical objections to my autonomy-based criterion for identifying agents of ‘public power’ which are based on the claim—implicit in much liberal writing—that public power should be delineated in accordance with certain organizational features of statehood: centralized decision-making, the use of force, and law-making capacity. I argue that none of these organizational features embodies any uniquely ‘public’ political qualities, and that each of these organizational features qualiﬁes as ‘public’ only to the extent that it satisﬁes the autonomy-based criterion I have proposed. Having made these eﬀorts to bolster my autonomy-based criterion of public power, Chapter 3 then applies this criterion to the contemporary global domain, and illustrates its practical implications. Here, I identify some concrete forms of NGO power in contemporary global politics that can be designated as ‘public’ and rightly subjected to democratic contestation and control. In Chapter 4, I elaborate and illustrate my theoretical account—sketched initially in Chapter 1—of how the agents of democratic control should be identiﬁed in global politics. Here, I develop the argument that we should accord democratic agency (i.e. entitlement to exercise control over public power) to the members of ‘stakeholder’ communities, composed of all those individual ‘stakeholders’ whose autonomous capacities are constrained by the exercise of public power. My analysis here begins by elaborating the concept of a jurisdictional ‘stakeholder’ community—which is grounded in the global liberal democratic discourse surrounding attempts to democratize non-state actors such as NGOs—and explaining how the conception of democratic ‘community’ underpinning it diﬀers

14

Introduction

from the ideals associated with the ‘closed’ societal model of democracy. I then turn to the question of how this ‘stakeholder’ model can in practice be instituted. Here, I illustrate how this new conception of democratic communities could be applied in practice, by addressing some issues and dilemmas that commonly arise for practitioners who invoke the idea of ‘stakeholder’ communities in their attempts to democratize the activities of NGOs. In the next two chapters, I develop the ﬁrst part of my framework for representative decision-making in a democratic global polity, which comprises an account of legitimate social choice in a representative global democracy—that is, an account of the process through which stakeholder communities can identify the speciﬁc policies, or more generally the substance of the ‘interests’, that their representatives can legitimately advance through public policymaking. In Chapter 5, I challenge the widespread idea that legitimate social choice in global politics can be achieved solely through representation by nation-states—either as a means of aggregating individual preferences across territorial states or as a means of conducting deliberative decision-making among the cultural entities of ‘nations’. I begin my discussion in this chapter by explaining that any representative democratic theory of legitimate social choice must be built upon two kinds of argument: an explication of and justiﬁcation for the underlying normative conception of interest representation; and an explanation of how the proposed constituency boundaries can eﬀectively enact the ideal of interest representation in practice. I then develop a substantive critique of the conventional nation-state-based model of global representation, critically examining both some normative and practical justiﬁcations for this model that are based upon a ‘burkean’ model of interest representation and linked to the notion of a nation and some alternative justiﬁcations that are based upon a ‘liberal individualist’ model of interest representation and linked to the notion of a state. My analysis here highlights several signiﬁcant normative weaknesses of nation-state representation, and I conclude accordingly that nation-state representation cannot alone achieve legitimate social choice in global politics. In Chapter 6, I argue that achieving legitimate social choice in global politics instead requires the establishment of a decision-making process involving deliberation among multi-stakeholder representatives, which could redress the weaknesses of the nation-state model in crucial ways. I begin here by outlining the key institutional features of the ‘multi-stakeholder’ model, and explaining how it can employ the representation of ‘multi-stakeholder’ interests by NGOs either to supplement or to substitute for the representation of nation-state constituencies by governments. I then examine the theoretical basis of the multi-stakeholder model, and argue that it can be justiﬁed within the framework of a theoretical model of interest representation that is characterized here as ‘liberal pluralist’. I further examine the strengths and weaknesses of the multi-stakeholder model as a framework for global representation. At a theoretical level, I argue that it has stronger normative foundations than either the burkean or liberal individualist models which underpin the nation-state representative system. It does, however, have certain practical limitations, which mean that it would be unworkable on

Introduction

15

its own as the sole basis for delineating constituencies in public global decisionmaking institutions. Accordingly, I conclude that we should endorse a hybrid approach to delineating democratic constituencies in global politics—one that incorporates multi-stakeholder representation within more conventional structures of representation by nation-states. In the ﬁnal two chapters, I develop the second part of my framework for representative decision-making in a democratic global polity, which comprises an account of legitimate representative agency within a representative global democracy—that is, an account of how to identify the speciﬁc individuals or organizational agents that are entitled to wield public power as legitimate ‘representatives’ of stakeholder communities (in pursuit of the policies or general interests identiﬁed through social choice mechanisms). In Chapter 7, I argue that establishing legitimate representative agency in global politics does not require election mechanisms, of the kind that are employed to identify legitimate representatives within established nation-state democracies. Rather, legitimate representative agency can be established through alternative non-electoral mechanisms of authorization and accountability. My analysis in this chapter begins by elaborating a fundamental reason for viewing elections as a legitimate means of establishing representative agency: that elections have the capacity to deliver stakeholders some degree of political control over their representatives’ actions and hence some degree of political control over the public political decision-making in which they have a legitimate stake. It follows from this that, at least in principle, we could establish legitimate nonelectoral representative agency if it were possible to identify alternative institutional mechanisms that could deliver equivalent forms of political control to stakeholders. First, though, we would need to specify more clearly the character of the particular forms of political control that are instituted through election mechanisms. With this prospect in view, the chapter turns next to an analysis of these forms of stakeholder control, arguing that the speciﬁc forms of political control over representatives’ actions that elections deliver to stakeholders are those entailed in institutional mechanisms of authorization and accountability. I explain how appropriate mechanisms of authorization and accountability can deliver certain forms of political control to stakeholders, and defend the proposition that mechanisms of authorization and accountability are suﬃcient to constitute a conceptually complete model of democratic representation. Finally, having defended the claim that mechanisms of authorization and accountability (whether electoral or non-electoral) can establish legitimate representative agency, I elaborate the institutional mechanics of democratic authorization and accountability in greater theoretical detail. In Chapter 8, I then consider how non-electoral mechanisms of the kind elaborated in the previous chapter might function in practice in contemporary global politics. Here, I examine (in a tentative and preliminary fashion) some possible pathways for institutional reform via which NGOs could build stronger democratic credentials as representatives of various stakeholder groups. To this end, I draw on some examples of contemporary practice among NGOs to illustrate

16

Introduction

the general institutional form such mechanisms might take. I begin by considering how the two elements of authorization—delegation and empowerment— might potentially be enacted through non-electoral mechanisms so that NGOs can function as legitimate representatives of certain stakeholder constituencies. I then consider how the two elements of accountability—transparency and disempowerment—might potentially be enacted through non-electoral mechanisms. Finally, I discuss how it may be possible to meet the challenge of ensuring appropriate forms of democratic equality for all stakeholder participants, in the absence of the formal equality embodied in an election. I do not attempt to argue in this chapter that current practices are fully democratic, nor do I purport to be able to oﬀer fully developed institutional solutions. Rather, the purpose of the material presented here is to provide a focal point for exploration of some practical possibilities, and to point towards some potential future directions for innovation and reform.

THE LIMITATIONS Before getting started with this substantive argument, it is important to clarify a few issues regarding the parameters and limitations of the arguments that are to follow. First, I have said that although I explore and illustrate my theoretical arguments primarily by examining a single category of non-state actors— NGOs—I want to claim that the core concepts and norms of Global Stakeholder Democracy have broader applications, and can be invoked also in relation to other political actors such as TNCs and IOs. One implication of this is that we need not be overly concerned with precisely how the category of non-state actors called ‘NGOs’ is deﬁned for the present purposes. To focus my discussion, I draw mainly upon examples of NGOs that fall within fairly conventional deﬁnitions of the category—non-proﬁt groups engaged in humanitarian, development, and advocacy work, and so on. However, this does not reﬂect any underlying theoretical distinction that I am making between these and other organizations that might also be included as ‘NGOs’. Instead, I suggest that diﬀerent kinds of actors in global politics should attract diﬀerentiated democratic critiques based on the forms of power that they exercise over others in their global activities, rather than on some organizational typology that designates them as ‘NGOs’ or otherwise. Moreover, it may well be that the process of democratizing NGOs in accordance with the Stakeholder model would, in some cases, transform their organizational structures so signiﬁcantly that they would become quite diﬀerent institutional creatures from the contemporary NGOs whose standard organizational structures have become familiar over recent decades. To the extent that signiﬁcant organizational transformations would be required by the Stakeholder model— both within ‘NGOs’, and between NGOs and the broader communities in which they operate—my arguments here cannot be understood merely as an attempt

Introduction

17

to confer democratic legitimacy upon NGO structures and practices in their current forms. Rather, the process of democratizing a political entity must always transform the structure of the entity itself in some important respects; this is true in relation to the democratization of state entities, so we should not be surprised or puzzled to ﬁnd it true also in relation to non-state actors. The extent of the transformations to existing NGO structures required to achieve democratic legitimacy will likely vary greatly in diﬀerent cases and contexts. But irrespective of the extent of the organizational transformations required, focusing on NGOs in their current form as a starting point for theoretical analysis can help identify novel pathways to global democratization, which involve treating non-state actors quite diﬀerently from how traditional state-based democratic models have treated them. Despite not wanting the applications of my arguments to be limited to any speciﬁc category of organizational entities, it is true that the cluster of organizations I focus on for illustrative purposes does fall within the quite narrow category of non-state institutional entities deﬁned as ‘NGOs’ on standard organizational typologies. 31 An important limitation of this methodological strategy is that the material I present in this book may not be suﬃcient to illustrate clearly the broader applications of the theoretical framework I develop, to the satisfaction of a sceptical reader. I acknowledge that there are many signiﬁcant institutional diﬀerences between the kinds of NGOs I look at here and other kinds of nonstate actors wielding power in the contemporary global polity, like TNCs, IOs, and states wielding external power. Moreover, these institutional diﬀerences are likely to have some important implications for the forms of democratization that the Stakeholder model would prescribe in relation to each. These other actors diﬀer from NGOs in a number of important respects: the forms of power they wield and the institutions through which they wield it; the kinds of communities that would likely qualify as ‘stakeholders’ in their public power; and the forms of countervailing power that could be harnessed eﬀectively (via processes of authorization and accountability) to ensure their public decision-making is appropriately representative of stakeholder interests. These diﬀerences certainly generate a number of important theoretical and institutional challenges of their own, which I do not claim to tackle fully in this book. In the book’s Conclusion, I say something brief about these matters, but this discussion attempts only to ﬂag some key challenges that would arise in applying the Stakeholder model to other actors, and present a few tentative suggestions about how these challenges might be overcome. My hope is that this cursory discussion will be enough to convince readers that my claim about the broader applicability of the Stakeholder model is plausible, and deserving of further exploration and development as a project for future research, but I do not claim that it will be enough to defend this claim fully. This task is one that I must postpone to another occasion. 31 For a discussion of some such typologies, see Anna C. Vakil, ‘Confronting the Classiﬁcation Problem: Toward a Taxonomy of NGOs’, World Development, 25/12 (1997).

18

Introduction

Finally, it is worth saying something more about the sense in which readers can expect this book to engage in ‘realist’ analysis, as I have so far claimed it will do. In applying the ‘realist’ label to the theoretical arguments I develop, I do not intend to commit myself to delivering detailed blueprints for institutional reform, realizable at their existing level of speciﬁcation and ready for practical implementation. Rather, I mean only to commit myself to sketching a framework for global democracy that is realizable in broad outline—that is, a framework that sketches a realistic rather than a utopian path towards global democratization. What precisely the institutions of Global Stakeholder Democracy will ultimately look like once we advance further along this path is something I do not claim to be able to answer with a high degree of clarity or ﬁnality in what follows here. I do certainly undertake to elaborate the Stakeholder framework with speciﬁc institutional details to the extent that my own empirical expertise and institutional imagination permit. But it is also important to acknowledge that there is a limit to the kind of practical prescription that I can credibly engage in within a normative theoretical book of this kind. I think it would be unhelpful to try to prescribe detailed institutions without more thorough empirical investigation and analysis of various alternatives. And although empirical analysis of this kind would be crucially important to a practical project of building stakeholder democracy, it would require extensive original empirical research of a kind that is beyond the scope of this book’s ambitions. As a work of normative political theory, the arguments in this book aspire to contribute to practical institutional debates by conceptually framing—and helping to set a new agenda for—subsequent empirical investigation (and practical experimentation) in relation to how global democracy can best be achieved. I think that this kind of intellectual and practical agenda-setting is the kind of practical work that normative political theorists are best suited to perform. For this reason, if the book were to succeed in persuading empirical scholars and practitioners to take up the institutional challenges that it poses, then I would think it had made a useful practical contribution of the ‘realist’ character that I am committing myself to here.

Part I Democratic Boundaries in the New Global Polity

This page intentionally left blank

1 Democracy Beyond ‘Closed’ Societies 1.1. INTRODUCTION The ﬁrst challenge that we must confront in developing a framework for global democracy is the problem of political boundaries: how should we delineate the scope of democratic ‘societies’ within the global order? To be clear about the nature of the boundary problem I am concerned with here, I am not referring to the familiar question of moral boundaries—that is, the boundaries of ‘moral concern’, or of those to whom political institutions must be morally ‘justiﬁable’. 1 The political boundaries I am concerned with here, instead, are the concrete institutional boundaries that delimit the scope of democratic processes of social choice and political control. These institutional boundaries are constituted by answers to the following two questions. First, what is the scope of the political power that must be held to democratic account as the legitimate subject of global democratic control? In what follows, I refer to the legitimate subject of democratic control as ‘public power’. 2 Second, which populations comprise the decision-making ‘publics’ or ‘peoples’ entitled to wield this democratic control as the legitimate agents of global democratic control? Developing answers to these two questions must be the ﬁrst task of a theoretical project concerned with global democracy, since we cannot begin to devise legitimate institutions for global social choice and political control until we have identiﬁed the boundaries within which these institutional processes are to take 1 I take moral boundary questions of this kind to subsume most discussions of the scope of ‘justice’ (even when justice is conceptualized as ‘political’—or in other terms even when one is a ‘dualist’ rather than a ‘monist’ with respect to justice claims). This is because whether or not we think of our theories of justice as moral in any comprehensive or metaphysical sense, we generally agree at least that they specify the community of moral concern for the purposes of critical reﬂection on legitimate political institutions (which is the topic of my discussion here). A recent discussion of the scope of justice that invokes a ‘political’ conception is Thomas Nagel’s; see Thomas Nagel, ‘The Problem of Global Justice’, Philosophy and Public Aﬀairs, 33/2 (2005). The distinction between ‘monist’ and ‘dualist’ conceptions of justice is Liam Murphy’s; see Liam Murphy, ‘Institutions and the Demands of Justice’, Philosophy and Public Aﬀairs, 27/4 (1998). 2 The concept of ‘public power’ is rarely problematized or given clear articulation within democratic theory; rather, the term ‘public power’ is commonly employed loosely and interchangeably with a family of related but distinct concepts—including ‘authority’, ‘sovereignty’, ‘statehood’, and ‘governance’ or ‘government’. Accordingly, one of the functions of my analysis in Chapters 1, 2, and 3 will be to explicate the concept of ‘public power’ as distinct from these other related concepts, and to present speciﬁcally ‘public’ forms of power as the appropriate targets for democratization in global politics.

22

Democracy Beyond ‘Closed’ Societies

place. As such, the absence of satisfactory answers to these boundary questions is widely recognized, by empirical scholars and practitioners concerned with issues of institutional legitimacy, as a serious impediment to their analytic and reformist endeavours. 3 It is commonplace for democratic theorists to acknowledge institutional boundary questions of this kind as important, yet resist the notion that they could be resolved from within democratic theory itself. As Robert Dahl puts the point, ‘one cannot decide from within democratic theory what constitutes the proper unit for the democratic process’. 4 The reasons given for this usually invoke the notion that a democratic resolution of its own boundaries would be paradoxical, since democracy ‘depends on a decision-rule, usually some variant of majority rule, but the rule’s operation assumes that the question “majority of whom?” has already been settled’. 5 Certainly it is true that democratic theory cannot answer questions about institutional boundaries if we conceptualize democracy procedurally—that is, if we think it is deﬁned by the operation of some speciﬁed social choice procedure. This is not true, however, if we conceptualize democracy substantively—that is, as a system that gives institutional expression to the fundamental democratic values of individual autonomy and equality. Democratic theory viewed in this latter sense has the resources within it to say much more about the institutional boundary problem than is often thought. In this chapter, I present a new democratic framework for answering these institutional boundary questions, and in doing so I establish some clear theoretical foundations upon which my wider account of global democracy can be built. I argue that we need to abandon the assumption, entrenched within traditional liberal-democratic theory, that democracy must take place within a ‘closed’ society. By ‘closed’, I mean to denote a society constituted on the one hand by an organizationally uniﬁed (or ‘constitutionalized’) agency of public power as its subject of democratic control, and on the other hand by a community of ‘citizens’ (sharing an exclusive and all-encompassing political status of civic membership) as its agent of democratic control. The paradigmatic case of the former is a state, and the paradigmatic case of the latter is a nation (or ‘people’, or similar social entity). 3 See, for example, the discussions of this problem in Ruth W. Grant and Robert O. Keohane, ‘Accountability and Abuses of Power in World Politics’, American Political Science Review, 99/1 (2005); Robert O. Keohane, ‘Global Governance and Democratic Accountability’, in D. Held and M. KoenigArchibugi (eds.), Taming Globalization (2003); Benedict Kingsbury, Nico Krisch, and Richard B. Stewart, ‘The Emergence of Global Administrative Law’, Law and Contemporary Problems, 68/3–4 (2005); Nico Krisch, ‘The Pluralism of Global Administrative Law’, European Journal of International Law, 17/1 (2006). 4 Robert Dahl, ‘Federalism and the Democratic Process’, in J. Roland Pennock and John W. Chapman (eds.), Liberal Democracy (New York, London: New York University Press, 1983), 103. 5 Ian Shapiro and Casiano Hacker-Cordón, ‘Outer Edges and Inner Edges’, in Ian Shapiro and Casiano Hacker-Cordón (eds.), Democracy’s Edges (Cambridge: Cambridge University Press, 1999), 1.

Democracy Beyond ‘Closed’ Societies

23

Instead, we should embrace the prospect of a ‘pluralist’ liberal-democratic order in global politics, composed of multiple agents of public power held to account by their multiple overlapping ‘stakeholder’ communities. What this requires, I argue, is that multiple state and non-state agents of public power in global politics should be subject to direct democratic control in accordance with their impact on the capacities of populations to lead autonomous lives. Correspondingly, individual ‘stakeholders’ should be accorded participatory entitlements as the legitimate agents of democratic control with respect to these multiple powerful actors, in accordance with the impact of particular public political actors upon their capacities to lead autonomous lives. This argument does not constitute a complete theory of global democracy, since it does not provide an institutional model of the mechanisms through which collective political decisions can be legitimately reached, and power legitimately controlled, within a pluralist global order. These two sets of normative problems— those concerning the boundaries of democratic decision-making, and those concerning the internal mechanisms of this decision-making—are not wholly independent of one another, insofar as the same fundamental liberal-democratic ideals of individual autonomy and equality must underpin legitimate solutions to both. However, developing a detailed account of these political institutions is to a signiﬁcant extent a separate and subsequent theoretical problem, which I address in some depth in Part II of the book. In this chapter and the following three chapters, then, my central focus can cogently remain on the narrower—but more fundamental—problem of global political boundaries.

1.2. THE STANDARD THEORETICAL MODEL OF A ‘CLOSED’ DEMOCRATIC SOCIETY Although the institutional boundary problem has not been tackled extensively in the democratic literature, a number of democrats have addressed it via one of two polarized responses. Many democrats have endorsed a ‘communitarian’ solution to the boundary problem, which prescribes that institutional boundaries should correspond with nation-states (or ‘peoples’, or some similar concept). Several others have endorsed an alternative ‘cosmopolitan’ solution, which prescribes that institutional boundaries ought to be, at least in certain respects, global in scope. 6 6 For a prominent defence of ‘communitarian’ democratic boundaries, see David Miller, On Nationality: Oxford Political Theory (Oxford: Clarendon Press, 1995). For prominent ‘cosmopolitan’ accounts, see David Held, Democracy and the Global Order: From the Modern State to Cosmopolitan Governance (Cambridge: Polity, 1995); Daniele Archibugi and David Held (eds.), Cosmopolitan Democracy: An Agenda for a New World Order (Cambridge: Polity, 1995); Daniele Archibugi, David Held, and Martin Köhler, Re-Imagining Political Community: Studies in Cosmopolitan Democracy (Cambridge: Polity Press, 1998); Richard A. Falk and Andrew Strauss, ‘On the Creation of a Global People’s Assembly: Legitimacy and the Power of Popular Sovereignty’, Stanford Journal of International Law, 36/2 (2000);

24

Democracy Beyond ‘Closed’ Societies

Despite the obvious diﬀerences between these standard answers in terms of the scope of the institutional boundaries they prescribe, these existing answers do share certain crucial—though tacit and generally unexamined—assumptions about the structure that democratic boundaries must take. More speciﬁcally, I want to highlight (in order to subsequently challenge) the widespread theoretical assumption that democratic boundaries must delineate ‘closed’ societies of a particular kind. This standard conception of democratic boundaries views them as ‘closed’ with respect to both the structure of public power (the legitimate subject of democratic control) and the structure of the community of democratic decision-makers (the legitimate agent of democratic control). First, democrats commonly assume that public power within a democratic society must be ‘closed’ in the sense that it is structured as an organizationally uniﬁed framework of public power to which each member of the democratic society is equally subject. Paradigmatically, public power is concentrated in a ‘state’ agent, or ‘sovereign’ power, which exercises public power through law-making. As such, public power is thought to impact equally upon all members by virtue of the fact that each is equally subject to the same set of laws, through which public power is wielded. In this sense, we can say that it is ‘closed’ with respect to its organizational structure. This conception of the structure of public power is closely tied to a certain set of assumptions about where this public power comes from, in the ﬁrst place. The standard assumption is that public power comprises an institutional agency that has been actively constructed by the relevant ‘public’ (hypothetically, if not temporally, prior to any democratic process for controlling its exercise) to provide a collective political framework for the pursuit of a shared democratic ‘common good’ or ‘public interest’. This assumption is associated at a deep level with the ‘social contract’ tradition, as well as with ‘republican’ strands of democratic thought. This theoretical assumption about the origins of public power is reﬂected most clearly in conventional state-based models of democratization, which present the process of democratization as entailing two components. The ﬁrst of these components involves the creation of a state—comprising a centralized institutional apparatus within a given territory—with extensive regulatory powers. The statebuilding project is supposed to ensure that the power of the plural actors within ‘civil society’ is controlled and limited in such a way as to protect a designated democratic ‘common good’ or ‘public interest’. Through this ﬁrst phase of democratization, ‘public power’ (the subject of democratization) is concentrated in the state. As a product of this concentration of public power, the domain of ‘civil society’—composed of multiple individuals as well as associations such as NGOs and corporations—is ‘depoliticized’, through the state’s regulatory limitation of Thomas Pogge, ‘Cosmopolitanism and Sovereignty’, Ethics, 103/1 (1992); Allen Buchanan, Justice, Legitimacy, and Self-Determination: Moral Foundations for International Law (Oxford: Oxford University Press, 2004); Andrew Kuper, Democracy Beyond Borders: Justice and Representation in Global Institutions (Oxford: Oxford University Press, 2004). I review both communitarian and cosmopolitan responses to the boundary problem in Terry Macdonald, ‘Boundaries Beyond Borders: Delineating Democratic “Peoples” in a Globalizing World’, Democratization, 10/3 (2003).

Democracy Beyond ‘Closed’ Societies

25

the range of powers that these ‘private’ actors can wield over one another. In the second element of the democratization process, the public power of the state is brought under collective democratic control by all the individuals within ‘civil society’ who are subject to its regulatory restraints. In this way, the standard model of democratization suggests that a ‘closed’ framework of public power is constructed through a process of state-building, prior to the establishment of democratic procedures for political control. Although this standard account of the ‘closed’ structure of democratic public power is linked paradigmatically to the idealized structure of state power, it is reﬂected also in most cosmopolitan models of democratic boundaries. To avoid any ostensible misrepresentation of cosmopolitan arguments, I should emphasize that cosmopolitan democratic models do not require all public political decisions in global politics to be made by a single centralized decision-making agency, in the image of a conventional ‘sovereign’ decision-maker. (Correspondingly, they do not imply that all members of a cosmopolitan democratic society must be equally subject to every speciﬁc political decision.) On the contrary, they typically allow that many speciﬁc policy decisions will be made at more localized and disparate jurisdictional levels, by multiple agencies of public power operating within an overarching public political apparatus. Such a jurisdictionally divided framework of public power is structurally similar in key respects to ‘federalist’ frameworks of public power, of the kind we see in many nation-states (like the USA and Australia) and to some extent also in the emerging European Union. But despite the jurisdictional division of public power within cosmopolitan democratic models, global public power is still ‘closed’ in an important sense, insofar as all public power is entrenched within an encompassing constitutional structure, which allocates public political power across multiple decision-making jurisdictions in accordance with uniﬁed global democratic purposes, and which applies equally to all individuals. Such a global structure of public power need not be institutionalized through formal legal norms to count as ‘constitutionalized’ in the relevant sense, and indeed many cosmopolitans do not advocate such legal institutionalization of global public power. Let me clarify then what I mean instead when I talk about a constitutional structure of global public power, as something endorsed within conventional cosmopolitan democratic models. Establishing a shared framework of public power within a society is one of the central tasks performed by a democratic constitution. This is of course not the only task performed by a democratic constitution; at a more general level, the ‘constitution’ of a democratic society functions to specify the fundamental purposes and institutions of that society, and correspondingly the rights and duties of members. This includes speciﬁcation not only of the framework of public power but also of the scope of the community of decision-makers, and the structure of institutions for political decision-making. Here, though, I am not concerned with all functions of a democratic constitution in general, but only with the speciﬁc function of instituting a uniﬁed framework of public power. A democratic constitution functions structurally to constitute the agencies of public power within a society by specifying each of the following. To begin with,

26

Democracy Beyond ‘Closed’ Societies

it speciﬁes the constitutive purposes for which public power is to be wielded. This involves speciﬁcation of certain ‘public’ democratic purposes, and the prescriptive requirement that agencies of public power should be structured so as to best advance these purposes. In some contexts this implies the imperative to impose certain constraints or limits upon the purposes for which public power can be wielded, and the proscription of the exercise of public power in violation of these purposes. Limits imposed upon public power through the constitutional entrenchment of certain individual ‘rights’ would fall into this category. In other contexts, this implies the imperative actively to construct enabling agencies, with special capacities to advance public purposes. The constitutional construction of political agencies with special coordinating or administrative capacities would fall into this latter category. Second, a constitutional structure of public power entails speciﬁcation of the identities of the actors that can be empowered to wield such power. There are many and varied ways in which constitutions can restrict the range of actor identities that can be accorded access to the exercise of public power. To take one example, they can specify that particular public powers are to be reserved for members of particular families (as in democratic constitutional monarchies), or alternatively for members of particular ethnic or religious groups (as in ‘consociational’ democracies). To take another example, democratic constitutions can reserve access to a range of important public powers for citizens (as distinct from non-citizen residents), or for certain subcategories of citizens (such as those born rather than naturalized into citizenship, as in the case of the US Presidency). Third, a constitutional structure of public power entails speciﬁcation of the particular sources of empowerment that are to underpin and secure its exercise. Here, constitutions can entrench answers to questions of the following kind. Must public power be wielded through the instrument of law, and subject to the constraints of the ‘rule of law’, or can it be wielded extra-legally? Can force or the threat of force be employed in the exercise of public power, and if so what limits are there on the forms and circumstances of its permissible use? Are agents of public power to be accorded material resources to sustain their power, such as special property rights or sources of revenue (e.g. land titles and entitlements to impose taxation)? If so, under what conditions? Are public political agents to be granted special entitlements to access information about others’ activities (as are sometimes accorded through investigative police powers), or special entitlements to speech and dissemination of ideas (as are sometimes accorded through parliamentary privilege), as means of bolstering their public political capacities? Finally, a constitutional structure of public power entails speciﬁcation of whether and how public power is to be divided among multiple agencies. Should public power be centralized in a single (executive) agent? If not, should it be divided functionally (as in standard separation of powers among executive, legislative, and judicial branches)? And if so, what functional jurisdictions should be accorded separate public agencies? Should it alternatively (or additionally) be divided territorially, through devolution of public power to local jurisdictional

Democracy Beyond ‘Closed’ Societies

27

levels? If so, where should such devolved public agencies be located, and how should public powers be divided among them? Ultimately, diﬀerent answers to each of these four sets of questions will lead to the creation of diﬀerent structures of public power within a given democratic society. But the construction of any framework of public power based on coherent answers to these four sets of questions results in some kind of constitutional structure of public power in the sense I am concerned with here. With this understanding in mind, we can recognize how cosmopolitan democrats do generally endorse the necessity of a constitutional framework of public power of this kind, and in doing so incorporate a ‘closed’ structure of public power in the relevant sense. In line with this standard model, cosmopolitan David Held refers to the need for multilevel decision-making structured by a unifying framework of ‘cosmopolitan democratic law’, 7 which ‘sets the form and limits of public power—the framework in which public debate, deliberation and policy-making can be pursued and judged’. 8 On Held’s account, ‘public power’ is thus delimited through global constitutional protections devised to establish a ‘common structure of political action’ within which members of a democratic society can engage in collective decision-making. 9 In a related proposal, Thomas Pogge refers to the need for ‘sovereignty’ to ‘be widely dispersed in the vertical dimension’—across multiple territorial levels—in such a way that it best serves the functions required by the overarching collective purposes entailed in his cosmopolitan account of justice. 10 Similarly, Allan Buchanan places the importance of a ‘closed’ constitutional structure of public power at the centre of his cosmopolitan account of global political justice. 11 Once again, it is not clear that these cosmopolitans intend to endorse any legally formalized constitutional framework for global public power; it could well be that the notion of ‘cosmopolitan democratic law’ and similar concepts are only meant to denote informal norms and principles which underwrite global democratic practice. But whichever way we interpret their prescriptions, the same central point remains: the structure of global public power is presented as a target for active institutional design, as a key element of the cosmopolitan democratic project. In other words, global democratic legitimacy is seen to require—among other things—the construction of a ‘closed’ global structure of public power that is designed to limit, enable, identify, empower, and divide agencies of public power in optimal service of global democratic norms.

7

Held, Democracy and the Global Order: From the Modern State to Cosmopolitan Governance,

270. 8

9 Ibid. 154–5. Ibid. 200–5. Pogge, ‘Cosmopolitanism and Sovereignty’, 58. Another cosmopolitan democrat, Andrew Kuper, proposes a similar dispersal of public power through a global constitutional framework, but allows for functional as well as territorial dispersions of this public power. See Kuper, Democracy Beyond Borders: Justice and Representation in Global Institutions. 11 See Buchanan, Justice, Legitimacy, and Self-Determination: Moral Foundations for International Law. 10

28

Democracy Beyond ‘Closed’ Societies

In addition to this ‘closed’ model of public power, standard democratic theories conceptualize institutional boundaries as ‘closed’ also with respect to the community of public decision-makers, which serves as the collective agent of democratic control. Here, democrats commonly assume that the democratic decision-making community must be composed of a collective of democratic ‘citizens’ sharing an exclusive and encompassing status of civic membership. First, the community is conceptualized as exclusive in the sense that it has a discrete membership, whereby members share the status of citizen only with one other, and are not simultaneously members of multiple other political communities. 12 Second, the community is conceptualized as politically all-encompassing, in the sense that the political relationships taking place within its boundaries encompass all aspects of its members’ political lives. Crucially, this means that the political roles and responsibilities allocated to individuals as members of a ‘closed’ political community encompass the complete set of political roles and responsibilities allocated to each individual. As a result of this, there is no possibility of conﬂict between the political roles and responsibilities that an individual may possess via membership in multiple political communities. The concept of ‘citizenship’ is then deployed to denote membership in a ‘closed’ community with these two related characteristics. Paradigmatically, ‘closed’ decision-making communities of this kind are linked to the idea of a ‘nation’ or ‘people’, or similar social entity. But this conception of the democratic decision-making community as ‘closed’ is not conﬁned to communitarian democrats. Cosmopolitans, too, portray their global demos as ‘closed’ in this way, with common references by many to the notion of global ‘citizenship’, and the conceptions of exclusive and all-encompassing political membership associated with this term. 13 Just as the idea of a ‘closed’ or ‘constitutionalized’ structure of public power does not preclude the possibility of internal jurisdictional divisions (within a federalist constitutional framework or some such thing), the idea of a ‘closed’ democratic community of citizens does not preclude the possibility of multiple overlapping spheres of membership internal to the overarching democratic society. An individual can participate in decision-making processes, and possess a range of political roles and responsibilities, at multiple levels within the democratic society—for example at regional, national, state, and local levels. But this individual will still remain a citizen of a ‘closed’ democratic society so long as the various roles and responsibilities associated with each of these overlapping democratic memberships are mutually compatible and hierarchically ordered such that 12 The practice of dual citizenship represents an exception to this, but this practice remains controversial in many democracies precisely because it violates the assumption of exclusivity paradigmatic of citizenship in standard democratic models. 13 See, for example, James Bohman, ‘Cosmopolitan Republicanism: Citizenship, Freedom and Global Political Authority’, Monist, 84/1 (2001); Richard A. Falk, On Humane Governance: Toward a New Global Politics (Cambridge: Polity, 1995); Falk and Strauss, ‘On the Creation of a Global People’s Assembly: Legitimacy and the Power of Popular Sovereignty’; Richard A. Falk and Andrew Strauss, ‘Not a Parliament of Dreams’, Worldlink: The Magazine of the World Economic Forum (2002).

Democracy Beyond ‘Closed’ Societies

29

there are no conﬂicts among them. That is, the rights and duties of membership within various internal democratic communities are encompassed by, or subsumed within, the overarching framework of rights and duties constituting their status as ‘citizens’ within the democratic society as a whole. Within an idealized ‘constitutionalized’ cosmopolitan democratic society—just as within a federalist state structure—individuals’ democratic rights and duties are coherently ordered in this way, and it is for this reason that we can identify such a democratic society as ‘closed’, despite its many internal divisions. In sum, the standard conception of democratic boundaries as ‘closed’ is closely linked to the structure of democratic boundaries in a nation-state, and is perhaps a reasonable way of conceptualizing democratic boundaries in this context insofar as it reﬂects common (though certainly far from universal) facts about the structure of state power, and of national decision-making communities. 14 However, what is more noteworthy (and, I will argue, problematic) is that this conception has been abstracted and transposed to the global level by cosmopolitans also, as though ‘closed’ boundaries of this particular kind were essential for any legitimate democratic order.

1.3. BEYOND ‘CLOSED’ SOCIETIES: A ‘REALIST’ DEMOCRATIC AGENDA Although most cosmopolitan models of democracy devised for application at the level of global politics tacitly suppose that democratic boundaries must be ‘closed’ beyond nation-states—just as they are within nation-states—I argue here that this unduly limits cosmopolitan thinking about the range of institutional possibilities for global democratization. In departure from the standard picture of democratic boundaries, I propose that a legitimate global democratic society should be constituted through boundaries of a very diﬀerent kind, which better reﬂect the ‘pluralist’ political character of the global order. Endorsing the ‘closed’ model of democratic boundaries at the global level is highly problematic, I argue, because of serious pragmatic diﬃculties with the requirement for public power to be wielded through a uniﬁed (constitutionalized) structure of the kind I have discussed. In our contemporary ‘globalizing’ world order, global public power is not currently allocated among regulatory agencies in accordance with any underlying democratic principles, or ‘constitutional’ structure. The existing system of ‘sovereign’ states and international law can perhaps be said to perform such regulatory functions to a limited degree. However, non-state actors such as Transnational Corporations (TNCs), Non-Governmental 14 It may be that we should not think of boundaries this way within many states either, since many do not ﬁt the ideal mould of the nation-states I have described. I do not resist this possibility, nor seek to pursue it; my interest is primarily in the global case so it is with respect to this case that I direct my critique of the ‘closed’ model of democratic boundaries.

30

Democracy Beyond ‘Closed’ Societies

Organizations (NGOs), and International Organizations (IOs), alongside a plurality of states, wield many of the forms of ‘public’ decision-making power (in ﬁelds of law-making, economic development, public service-provision, etc.) that democrats have traditionally sought to subject to public regulatory control. 15 The signiﬁcance of these plural non-state agencies of public power is that they tend to emerge in ad hoc, varied, and unstructured ways—that is, their public power is not the product of any project of grand institutional design. As such, the overall institutional structure of global public power lacks the kind of underlying principled order which a cosmopolitan democratic blueprint would seek to deliver. The present structure of global public power can thus be characterized as pluralist both because its organizational structure is fragmented and uneven in impact and because its structure is not designed to serve optimally any unifying democratic purposes (as a constitutional structure of public power would do). Cosmopolitans do not deny this contemporary global reality; their institutional proposals are presented merely as normative ideal standards, intended to guide the future development of the international system. As such, they prescribe that the existing global political pluralism should be countered through the creation of a new ‘closed’ structure of global regulatory agencies capable of limiting and controlling the exercise of these other actors’ power, in the same way that state agencies regulate the power of non-state actors within their own territorial boundaries. What this would entail, eﬀectively, would be a project of global constitutionbuilding—with the transfer of global decision-making power from these existing plural actors to some new constitutionally structured agencies. Following this path may provide a largely satisfactory solution at the level of purely ‘ideal’ theory. To the extent that such global constitution-building is feasible then, there is a strong case for maintaining that democrats should attempt to pursue it, and it is possible that over time there would be some modest successes in this endeavour. However, we should expect that the extent to which democrats are likely to succeed in such global constitution-building will be minimal, at least in the foreseeable future. Any attempt to transform substantially the structure of global public power would be a massive undertaking in social engineering, the enormity of which should not be underestimated, and the prospects of achieving extensive success in such a project would be dim. Shared democratic values at a global level cannot be straightforwardly converted into a new structural framework of global public power through some functionalist process of political transformation, driven by the intentional, instrumental, and far-sighted choices of reformist global political actors. This is because 15 See, for example, Susan Strange, ‘The Declining Authority of States’, in D. Held and A. G. McGrew (eds.), The Global Transformations Reader: An Introduction to the Globalization Debate (2000); Held, Democracy and the Global Order: From the Modern State to Cosmopolitan Governance; A. Claire Cutler, Virginia Hauﬂer, and Tony Porter, Private Authority and International Aﬀairs, Suny Series in Global Politics (Albany: State University of New York Press, 1999); A. Claire Cutler, Private Power and Global Authority: Transnational Merchant Law in the Global Political Economy (Cambridge: Cambridge University Press, 2003); Rodney Bruce Hall and Thomas J. Biersteker, The Emergence of Private Authority in Global Governance (Cambridge: Cambridge University Press, 2002).

Democracy Beyond ‘Closed’ Societies

31

the choices made by the political actors who drive historical processes of institutional development and transformation are not always (or even usually) intentional, instrumental, and far-sighted. For a start, political actors are commonly motivated by short-term goals of various kinds, instead of by the sort of longterm interests that are at stake in large-scale institutional development. Moreover, political actors commonly act on the basis of ‘logics of appropriateness’ rather than socially detached instrumentality, and this can inhibit the rational pursuit of long-term interests in institutional reform. Finally, many important political consequences of actors’ decisions are unintended rather than rationally calculated; since the extent to which political actions have unintended consequences increases with social complexity, this factor is likely to have signiﬁcant salience at the global level. 16 In other words, the fundamental distributions of power within the global order are a product not of any grand constitutional design, instituted through the coordinated pursuit of collective ‘public’ purposes, but rather of many uncoordinated factors (technological, sociological, economic, and political) which interact in unplanned ways to generate structural outcomes in terms of the framework of power to which populations are subject. That is, distributions of power arise to a large extent through the uncoordinated interplay of all the diﬀerent things that people do in the course of their lives—including the most personal, socially inconspicuous, and unplanned activities. Such social action by uncoordinated individuals and groups can lead, for example, to the development of new technologies, to new distributions of material resources, to new forms of communication and social mobilization, to new sources of political identity and solidarity, to new ideological movements and cultural value-systems, and so on. And in turn, these wider social transformations play key roles in generating and shaping opportunities for the exercise of power, thus contributing to the process of structuring the overall distribution of power in global politics. Because structural opportunities for the exercise of power are thus created—at least in signiﬁcant part—through such haphazard and complex sociological processes, the development of any new framework of global public power over time will inevitably be directed to a large degree by many highly contextual and path-dependent factors. It cannot be expected to conform to any functionalist logic arising simply from the identiﬁcation of shared cosmopolitan democratic values. It is for precisely these reasons that the existing structure of global public power has so far developed in the pluralist manner I have described, and the institutional dynamics that have generated the existing structure of global public power are likely to remain resistant to the future constitutionalization of global public power. 16 Paul Pierson, ‘The Limits of Design: Explaining Institutional Origins and Change’, Governance: An International Journal of Policy and Administration, 13/4 (2000); Paul Pierson, Politics in Time: History, Institutions, and Social Analysis (Princeton: Princeton University Press, 2004); John Gerald Ruggie, ‘Territoriality and Beyond: Problematizing Modernity in International Relations’, International Organization, 47/1 (1993).

32

Democracy Beyond ‘Closed’ Societies

As a result of this, the coordination problems associated with trying to transform a deeply pluralist structure of power into a ‘closed’ or ‘constitutionalized’ structure of power would seem insurmountable, at least within any but the longest of time frames. A democratic framework concerned with the practical political project of institutional reform must therefore confront and accommodate existing realities regarding the structure of global power, rather than expect that a project of democratization could begin by overturning and rearranging them in any fundamental way. I therefore propose that democrats should identify a more pragmatic path to global democratization, which can be accommodated within the existing institutional structures of global politics. Instead of seeing the creation of a ‘closed’ or ‘constitutionalized’ framework of public power as a necessary ﬁrst step in a project of global democratization, we should seek to democratize the framework of global power in its current pluralist structure. We should identify the multiple organizationally disparate agents that wield power within our existing pluralist global order of a kind that we think requires democratic regulation (states, NGOs, IOs, corporations, etc.) as direct subjects of democratic control (i.e. as agents of ‘public’ power), and then see the establishment of institutions to hold them to direct democratic account as a central task of global institutional reform. To understand the basis for this ‘pluralist’ account of global public power more fully, we must recognize that it is underpinned by a ‘realist’ approach to democratic theory, which accepts the structure of our existing political order as axiomatic in the democratic analysis, and takes it as a motivating precondition for the democratic project. This ‘realist’ approach is underpinned by a particular view of the purpose of democratic theory, which is to provide a theoretical framework for guiding the practical task of democratizing political life in the most eﬀective way possible under prevailing political circumstances. This pragmatic view of the democratic project has ﬁrm historical roots in the liberal-democratic tradition, which has from its outset been focused on the practical task of institutional reform and political transformation, and not only the task of philosophical critique. Theorists with nominally ‘ideal-theoretical’ methodological commitments do not always acknowledge the role of pragmatic considerations in normative argument. Moreover, functionalist and rational-choice empirical accounts of institution-building processes can further reinforce the assumption that conventional state-based democracy embodies some context-neutral normative ideal— by implying that democracy has historically developed within state structures because states provide the ideal institutional framework for performing democratic functions. Nonetheless, I suggest that the prominence accorded to a centralized framework of public power within mainstream normative democratic theories did not emerge solely (or even primarily) through the articulation and rational implementation of ‘ideal-theoretical’ constructions about how democratic public power ought to be structured within a social order. Rather, the historical development of democratic systems that emphasize the role of centralized regulatory agencies has been driven to a large degree by a contextual pragmatic

Democracy Beyond ‘Closed’ Societies

33

interest in imposing greater standards of legitimacy upon already existing uniﬁed public structures of sovereign power. Thomas Nagel has recently reiterated this point, in the context of a closely related set of debates about the prospects for global ‘justice’ more broadly conceived: [I]n thinking about the future, we should keep in mind that political power is rarely created as a result of demands for legitimacy, and that there is little reason to think that things will be diﬀerent in this [global] case. . . . First there is the concentration of power; then, gradually, there grows a demand for consideration of the interests of the governed, and for giving them a greater voice in the exercise of power. The demand may be reformist, or it may be revolutionary, or it may be a demand for reform made credible by the threat of revolution, but it is the existence of concentrated sovereign power that prompts the demand, and makes legitimacy an issue. 17

In the present context, the implication of this important insight is that the empirical facts about the present pluralist structure of power in the global order are of great relevance to the project of democratic theory—at least insofar as we seek democratic theory that can serve as a practical guide to institutional reform and institution-building in the real political world. 18 To prevent misunderstanding of this ‘realist’ approach, it is important to emphasize that it does not imply that overall structures of public power are of no democratic signiﬁcance, or that these structures should not in any sense be the target of democratic critique and reform. Rather, one of the long-term eﬀects of instituting democratic control over the multiple agents of public power that currently exist in global politics is likely to be an ultimate transformation— perhaps even an increasing ‘constitutionalization’—of the overall structure of public power. This is so because when agents of public power are brought under democratic control, the ‘public’ constituencies wielding this control are likely to demand changes to certain structural features of these agents, as an institutional means of increasing the likelihood that, over time, these public political agents will more fully and reliably act in the service of public interests. To take an example, the stakeholder constituency of a powerful corporation may demand that its current constitutive purpose (maximizing shareholder proﬁts) be structurally amended (through corporations laws, informal codes of conduct, etc.) to incorporate recognition of the importance of protecting certain stakeholder interests. Moreover, the development of formal accountability mechanisms as a means of holding the corporation to democratic account would ultimately result in structural changes to the decision-making processes and power-bases of the corporation in question, insofar as these changes would be necessary to accommodate the 17

Nagel, ‘The Problem of Global Justice’, 145. Although I concur with Nagel on this historical question about the importance of pre-existing power in prompting demands for democratic legitimacy and thus structuring existing democratic theories, I diﬀer from him in my answer to the normative question of which particular forms of preexisting power ought to be seen as prompting such demands. 18

34

Democracy Beyond ‘Closed’ Societies

consultative processes and sanctioning mechanisms that are inherent in accountability mechanisms. Over time, then, we would expect that the overall structure of public power in the global order would be signiﬁcantly transformed as a result of reforms to institute democratic control of existing agents of public power. It may be the case that these changes would eventually result in an overall structure of public power somewhat akin to the ideal blueprints articulated by cosmopolitan advocates of the ‘closed’ model of democratic public power. On the other hand, we may ﬁnd instead that the process of establishing democratic control of existing agents of public power leads to quite diﬀerent overall structures, of kinds that have not yet been imagined by philosophers but which nonetheless serve ideal democratic purposes through alternative (more practicable) structural means. For this reason, we should not expect the outcome of a ‘realist’ approach to yield any less ‘ideal’ democratic results over time. The key factor distinguishing the ‘realist’ approach from the standard approach, instead, is that the former articulates a clear reformist vision of how we can get from here to there—from a highly undemocratic global political order to a highly democratic one, through a legitimate process of democratization. 19 The articulation of ideal constitutional structures of public power would serve a more useful reformist purpose if there were some way to transpose these static ideal structures directly onto the political world straight from a philosophy classroom, bypassing the need for the dynamic process of political action and reform. Often cosmopolitans talk as though the question of transformative political agency is unproblematic— presumably because it is implicitly assumed that the ‘we’ engaged in democratic critique is allied with some benevolent superpower (either a single agent or some coordinated collective) willing and able to implement the just democratic blueprints devised by philosophers. However, even the most steadfast of optimists must surely be deeply sceptical about the prospect of locating such an omnipotent and benevolent hegemony in global politics, willing and able to subject its own power along with that of others to democratic control. And in the absence of such an agency, it seems clear that the top–down approach to democratization implied by the standard emphasis on grand constitutional design is untenable as a pragmatic theoretical framework for democratic reform. In contrast, the ‘realist’ approach I advocate here provides an approach to democratization that engages directly with the power of real global actors, by targeting its normative demands to existing political agencies. As such, it provides us with a more action-oriented and strategically viable theoretical framework for conceptualizing and pursuing the project of global democratic 19 The sense in which we can speak of this democratic control as conferring democratic legitimacy is thus somewhat diﬀerent from the sense in which ‘ideal’ theorists attribute legitimacy to the closed societal model of democracy. Whereas ‘ideal’ theoretical approaches characterize democratic legitimacy as something inherent only in the ideal democratic end-states resulting from a democratic reform process, the ‘realist’ approach I am advocating here characterizes democratic legitimacy as inherent in a (dynamic) legitimate reform process, within a society that lacks an ideal constitutional structure of public power but is undergoing democratization over time.

Democracy Beyond ‘Closed’ Societies

35

reform, capable of harnessing the real political dynamics of the contemporary global order in the pursuit of long-term democratic transformation. It should be clear from what I have said, also, that the ‘realist’ approach to delineating public power ought not to be seen as incompatible with a project of global democratic constitution-building. Rather, the two approaches to the project of global democratization can be complementary, and pursued simultaneously. To the extent that we can identify collective global purposes and establish the collective public political agencies to advance them, then we should certainly do so. All I am arguing here is that we should not see the practical limits of a project of global constitution-building as dictating the limits of the project of global democratization. Where global constitution-building fails—due to the enormous practical constraints I have pointed to here—democratization can still proceed, by targeting powerful agencies in their current pluralist structures.

1.4. PUBLIC POWER AND POLITICAL PARTICIPATION IN A ‘PLURALIST’ DEMOCRATIC ORDER If we are willing to make the pragmatic move I have so far defended, and accept that the structure of global public power will remain (for the foreseeable future at least) pluralist rather than uniﬁed, we are confronted with a normative theoretical challenge rather diﬀerent from the standard cosmopolitan project of ideal institutional design. Our new challenge is to provide normative criteria for specifying what forms of power count as ‘public’, and what communities of individuals should be empowered to participate in their democratic control. First, we need to articulate a normative criterion for identifying which of the many organizationally disparate forms of state and non-state power in global politics qualify as ‘public’ forms of power, such that they should be subject to democratic control. That is, we must provide an account of what makes some forms of power problematic and in need of special democratic legitimation, while other forms of power are unproblematic and permissible within a domain of ‘private’ conduct. Correspondingly, we must identify a criterion for delineating the jurisdictional boundaries of each agent of global public power, so that we know which populations should be empowered to hold each of these to democratic account. In response to the ﬁrst of these questions, I propose the central liberal value of individual autonomy as the appropriate standard to invoke in determining whether the impact of an act of power by one political agent upon another constitutes an act of public power. The liberal-democratic ideal has the value of individual autonomy at its core, and as such it is the protection of individual autonomy that should be the primary concern of liberal-democratic institutions. As such we can specify that power should be designated as ‘public’ and subject to democratic control whenever it impacts in some problematic way upon the capacity of a group of individuals to lead autonomous lives.

36

Democracy Beyond ‘Closed’ Societies

It is important to emphasize that this autonomy-based criterion for delineating the scope of public power makes a certain set of background assumptions about the underlying purposes of a democratic system. Here, I am articulating a criterion appropriate to a speciﬁcally liberal conception of democratic legitimacy, which sees the protection of autonomy (and conversely, the resistance of tyranny) as democracy’s fundamental purpose. Democratic institutions are of course not the only means of advancing this general liberal objective; it can also be pursued via a range of other (non-democratic) institutional devices—most notably, various constitutional constraints on power (such as bills of rights, separations of powers) and various mechanisms of ‘horizontal’ accountability, or ‘balance of power’ among powerful elites. These institutional avenues are not mutually exclusive, however, and I would maintain that democratic control is an especially eﬀective and reliable means of pursuing this liberal purpose. Democracy could of course be justiﬁed in other ways—for instance, as a system for protecting certain objectively speciﬁed human interests (either at an individual level or at an aggregative level). If we were to think of democracy as justiﬁed in some other such way, then the criterion for delineating the scope of public power would need to diﬀer accordingly. For instance, if we were to think that democracy’s purpose is to protect certain objective interests, then we would need to designate relevant threats to these interests, rather than threats to individual autonomy, as the basis for delineating public power. Accepting my autonomy-based criterion, then, requires ﬁrst accepting my general liberal orientation towards the democratic project. The scope of liberal-democratic principles is thus delineated in accordance with the normative function of liberal representative institutions: if the purpose of democratic institutions is to resist oppressive (autonomy-constraining) forms of power, it follows that it is these forms of power that should be subject to the normative constraints imposed through democratic institutions. Such democratic institutions function to confer legitimacy upon public power because they reinstate (to some degree) the autonomy of individuals, by according each individual a share in decision-making control. It must be recognized that the autonomy reinstated through representative institutions diﬀers from (and is less extensive than) that limited by public power insofar as it is expressed through collective rather than individual decision-processes. 20 Since a share in democratic control of public power does not fully compensate an individual for the autonomy constrained through the exercise of public power, democratic institutions can only confer legitimacy upon public power if it is assumed that the existence of some public power is necessary or unavoidable in the ﬁrst place. This could be assumed either on the normative grounds that some public power is desirable as a means for maintaining social order or on the pragmatic grounds that it is 20 It is for this reason that Robert Wolﬀ rightly rejects the capacity of democratic institutions completely to counteract the autonomy-constraining impact of public power. See Robert Paul Wolﬀ, In Defense of Anarchism (Berkeley, CA: University of California Press, 1998). In his analysis here Wolﬀ focuses on forms of public power he characterizes as ‘authority’.

Democracy Beyond ‘Closed’ Societies

37

a pervasive and unavoidable feature of social life. But if such an assumption is accepted, democracy confers liberal legitimacy insofar as it preserves individual autonomy to the greatest degree that is compatible with the existence of public power. If, then, we are to invoke some standard of impact on individual autonomy as the criterion for diﬀerentiating public from private acts of power, we must further specify the theoretical content of two aspects of this criterion: what elements of individual autonomy should be protected and promoted within a democratic polity? and what are the responsibilities of members of democratic polities to protect and promote the autonomy of others? These are big questions, and the task of articulating an operationalizable criterion for delineating public power would require ﬂeshing out complete answers to both. The answers to these questions cannot be wholly philosophical, however, so I cannot here undertake to provide full or deﬁnitive answers to either. Rather, an important implication of the democratic pluralism I am defending here is that we must accept there is no universally applicable set of understandings of autonomy, and (egalitarian) responsibility for autonomy which can be invoked to delineate democratic boundaries in all political contexts within the global domain. Instead, answers to these questions must to some degree be politically determined, within each of the speciﬁc contexts in which the need for democratic control arises. However, there does still need to be some conceptual core of autonomy and equality which is preserved and shared, common in all political contexts, in order for the political framework to count as ‘democratic’; otherwise, the notion of ‘democracy’ would lose any substantive meaning. Here, my purposes are limited to the task of sketching this conceptual core of the democratic public power criterion, and highlighting the residual normative questions on which ﬁnal political determinations of the scope of public power would then turn in each political context of democratic control. First, with respect to the value of individual autonomy, we can say in general terms that this is concerned with ensuring that individuals are empowered to formulate and eﬀectively pursue authentic personal goals. To be autonomous in the broad (etymological) sense of ‘self-rule’, an individual must be capable both of formulating authentic desires and of acting upon them in the real political world. 21 The conception of autonomy that can be employed to delineate public power within a liberal representative system, however, is narrower than this broad deﬁnition, insofar as it is a speciﬁcally democratic conception of autonomy—that is, one that can be protected and enacted through democratic political institutions. In some social contexts—of the kind discussed in certain debates within moral 21 In some recent literature on the concept of ‘autonomy’, the idea is deﬁned in relation only to the formulation of choices, and thus distinguished from the concepts of ‘freedom’ and ‘liberty’, which are deﬁned instead in relation to the capacity to act upon choices. See, for example, John Christman, ‘Introduction’, in John Philip Christman (ed.), The Inner Citadel: Essays on Individual Autonomy (Oxford: Oxford University Press, 1989); and Gerald Dworkin, The Theory and Practice of Autonomy (Cambridge: Cambridge University Press, 1988). I use the term here, however, in a broader democratic sense which incorporates both formulation of choices and action upon them.

38

Democracy Beyond ‘Closed’ Societies

philosophy—it is feasible to promote forms of autonomy that can be limited by aspects of the self, or aspects of the material world. In contrast, a speciﬁcally democratic ideal of autonomy must be conceptualized as a condition that can be limited only by the activity of other agents (and not by aspects of the self, or by aspects of the material world). Moreover, since public power is something that must always be exercised by agents, we need to think also about what is normatively entailed in the act of constraining another’s autonomy—that is, what will count as an act with impacts or eﬀects upon the autonomous capacities of another. To answer this question, we must identify the range of democratic responsibilities each agent possesses for protecting and promoting the autonomy of others. When one political agent fails to meet these responsibilities towards another, we can say that it has exercised public power over that other agent, since it is—through its (in)actions—responsible for some constraints upon the autonomy of that other. To avoid possible misunderstanding of my position here, it is important to emphasize that I do not intend, by linking public power to impacts of agents’ actions upon others’ autonomy, to attribute any intrinsic normative signiﬁcance to the common analytic distinction between ‘acts’ and ‘omissions’, or between ‘doing’ and ‘allowing’. Rather, I see this analytic distinction as a tool for invoking—in common-sense moral shorthand—fundamental and widely shared assumptions about the scope of agents’ responsibilities. As such, autonomy-constraining eﬀects for which we view a particular agent as clearly responsible we will characterize as emanating from the agent’s ‘actions’, and we will accordingly say that the agent has wielded ‘public power’. Conversely, autonomy-constraining eﬀects for which the agent is less evidently responsible we will say that the agent has merely ‘allowed’; in this latter scenario, we will not say that the agent has wielded ‘public power’. As such, it should be clear that I do not mean to link the criterion of autonomyconstraining impact to any substantive normative stipulations about the scope of diﬀerent actors’ responsibilities, based on any philosophical distinction between ‘negative’ and ‘positive’ duties, or some such thing. Rather, I think that it is helpful to employ the concept of ‘impact’ (of one agent upon another) as a basis for delineating public power only because this concept denotes (in common-sense moral shorthand) those eﬀects for which we broadly agree particular actors are responsible. In this way, the question of how responsibilities should be distributed remains largely open on my account of public power, and will be reﬂected in (rather than determined by) our conception of what it is for particular actors to ‘have eﬀects’ on others through their actions. All that is conceptually required with respect to the conception of responsibility underpinning a criterion of public impact is that it must be egalitarian in some central respect. That is, the account of responsibility for others’ autonomy must treat all individuals—at some level—with equal concern. This is necessary since the value of equality—alongside that of individual autonomy—is a foundational value within the democratic project, and a conception of social responsibility that

Democracy Beyond ‘Closed’ Societies

39

accorded greater priority to the autonomous capacities of some individuals than those of others would not meet the minimal standard required to qualify as ‘democratic’ in character. However, the substantive responsibilities that are implied by a democratic ‘equal concern’ for individuals’ autonomy can be interpreted quite diﬀerently in diﬀerent political contexts, and there can be a great deal of variation in the positions adopted on this issue without undermining the core democratic credentials of a political system. Beyond these basic conceptual elements that I have identiﬁed—with respect to the character of individual autonomy, and the range of responsibilities political actors possess for protecting the autonomy of others—the speciﬁcs of this public power criterion must be worked out in detail in each political context in which democratic control is to be established. As such, the fully operationalized public power criterion will vary in diﬀerent political contexts, to reﬂect the varying degrees of responsibility for others’ autonomy that can gain broad political consensus (and thus deliver legitimacy) in each. (I will say more about the implications of this democratic pluralism in Chapter 4.) If we invoke shared views about autonomy and (egalitarian) responsibility in this way as the basis for delineating global public power, which range of actors are likely to qualify, in many political contexts, as the legitimate subjects of democratic control? To begin with, application of such a criterion would certainly be likely to identify most forms of power wielded by states as ‘public’, since state power would generally be considered to have the requisite impact on others’ autonomy. We commonly view states as responsible for eﬀects upon individual autonomy that arise through a range of state activities such as law-making, law-enforcement, resource distribution, and so on. However, what is crucial to recognize here is that states are unlikely to be the only actors to satisfy this criterion of public power, since the criterion attributes states no intrinsic or exclusive public status. We can accordingly identify many examples of non-state power that would also appear likely candidates for designation as ‘public’, and subjection to democratic control. I do not elaborate these in this chapter; this task is postponed to Chapter 3, where I outline a range of NGO activities in contemporary global politics that can plausibly be viewed as wielding ‘public’ forms of power. Having sketched the basis for delineating the legitimate subjects of global democratic control, it remains to identify a corresponding criterion for delineating the decision-making communities that are to serve as the legitimate agents of global democratic control. In other words, how can we delineate the boundaries of the communities of decision-makers that should hold plural agents of public power to democratic account through appropriate institutions of social choice and political control? In accord with the above analysis of the scope of public power, I propose that democratic decision-making communities in global politics should be delineated based on an assessment of the impact of power upon the autonomous capacities of populations within the global order. Just as power qualiﬁes as ‘public’ whenever it impacts detrimentally upon the autonomous capacities of a group of individuals,

40

Democracy Beyond ‘Closed’ Societies

individuals should be entitled to participate in any political decision-making that impacts in problematic ways upon their autonomous capacities. 22 Accordingly, it is the communities of individuals whose autonomous entitlements are aﬀected by the exercise of power that should be identiﬁed as the legitimate agents of democratic control, with respect to the particular political actors that wield this public power. The institutional implications of this criterion for allocating participatory entitlements (delineating a democratic decision-making community) can perhaps be best understood with reference to a concept that has gained signiﬁcant currency in recent global democratic discourse—the idea of ‘stakeholder’ communities. In general, this idea embodies the proposition that political decision-making should be undertaken by the communities of ‘stakeholders’ who are aﬀected by given political decisions. A liberal-democratic account of shared responsibility for others’ autonomy can then be invoked to specify what kind of ‘eﬀects’ or ‘impact’ are required—both to generate an imperative for the democratic control of a political actor and correspondingly to delineate the population of individuals entitled to wield this control. It is by interpreting the concept of stakeholding through the lens of liberal-democratic values that we arrive at the criteria I have identiﬁed for delineating the legitimate subjects and agents of democratic control. (In Chapter 4, I elaborate this idea of stakeholder communities in more depth, and examine some practical institutional problems that arise in identifying the stakeholders of non-state actors such as NGOs.)

1.5. CONCLUSIONS: TOWARDS A ‘PLURALIST’ DEMOCRATIC GLOBAL ORDER The arguments I have presented in this chapter identify a strong imperative to abandon the long-established conception of democratic boundaries as necessarily ‘closed’, which has so far maintained a ﬁrm grip upon the theoretical and institutional agendas of liberal democrats in the global domain. Instead, I have defended an alternative ‘realist’ democratic framework, which seeks to allocate decisionmaking control to individuals in accordance with the impact of powerful agents’ actions upon their lives. Instead of seeing public power as something that must be constructed via a ‘closed’ constitutional structure, as the ﬁrst task of democratic institution-building, we should instead identify as ‘public’ any existing forms of power that impact problematically upon the autonomous capacities of others. 22 This general liberal intuition about the importance of autonomy in delineating the scope of participatory entitlements is shared by certain other liberal democrats, including John Rawls, A Theory of Justice, Rev. edn. (Oxford: Oxford University Press, 1999); Nagel, ‘The Problem of Global Justice’; Michael Blake, ‘Distributive Justice, State Coercion, and Autonomy’, Philosophy and Public Aﬀairs, 30 (2001); R. M. Dworkin, Sovereign Virtue (Harvard University Press, 2000); Held, Democracy and the Global Order: From the Modern State to Cosmopolitan Governance. However, the precise nature of their justiﬁcatory arguments diﬀers in some important ways from my own.

Democracy Beyond ‘Closed’ Societies

41

Moreover, instead of seeing the entitlement to political participation as something that should be allocated to all individuals equally by virtue of membership in a ‘closed’ democratic society, participatory entitlements should be accorded to individuals only in relation to those forms of power that impact in problematic ways upon their autonomy. This implies that global democracy can be achieved through a ‘pluralist’ democratic order, whereby multiple organizationally disparate state and non-state agents of ‘public power’ are subject to the democratic control of their respective communities of ‘stakeholders’. As I have said, these criteria for delineating democratic boundaries would require fuller elaboration before they could be operationalized in real-world institution-building. We would need to specify more fully not only which forms of individual autonomy are suﬃciently valuable to warrant protection through democratic institutions but also how (egalitarian) responsibilities for this protection should be distributed among the many state and non-state actors in global politics. At a deeper theoretical level, then, this framework for delineating democratic boundaries points us back to the same basic normative questions that underpin traditional approaches to the project of democratization, focused on building and controlling ‘closed’ or ‘constitutionalized’ frameworks of global public power. Traditional cosmopolitan approaches of this kind proceed by asking these same fundamental normative questions about the substance of democratic autonomy and social responsibility, and then invoke their answers as the basis for their democratic constitutional designs. The diﬀerence between my ‘pluralist’ model and these traditional accounts thus rests not on any divergent underlying normative questions (or divergent normative answers to these questions); rather, it rests on the level of abstraction and generality at which these questions are posed. First, there is a diﬀerence in the level of abstraction at which these questions are posed by each: by asking about the responsibilities of real existing actors to protect autonomy in the real existing political world (instead of asking, more conventionally, about the responsibilities of hypothetical actors in hypothetical democratic societies), the pluralist democratic model directs us to reﬂect upon which agents of public power and their decision-making communities should engage in real processes of democratization in contemporary global politics. Second, there is a diﬀerence in the level of generality at which these questions are posed by each: whereas the ‘closed’ model of global democracy requires a uniform set of answers to these questions to be applied in the delineation of democratic boundaries throughout the global polity, the ‘pluralist’ model permits diﬀerent answers to be applied in diﬀerent political contexts. Although each set of answers must conform to a core conceptual structure, the details can vary in each of the sets of political relationships where democratic control is to be established. The framework I have presented here for delineating global democratic boundaries raises many signiﬁcant questions, which require much more detailed consideration than I have been able to provide within the constraints of this chapter. In the following three chapters, I elaborate this framework further, and demonstrate how the concepts of public power and stakeholder communities can be applied

42

Democracy Beyond ‘Closed’ Societies

concretely to non-state forms of power in contemporary global politics. But even in the preliminary form in which I have presented it so far, we can see that the framework for a pluralist global democratic order that I have sketched out here can serve to articulate a new and promising direction for thinking about the democratic boundary problem. A key advantage of embracing ‘pluralist’ democratic boundaries is that doing so provides a means of transcending the impasse between ‘communitarian’ and ‘cosmopolitan’ democratic models that has dominated recent discussion of the global democratic problem. Political theorists have been for so long accustomed to preoccupation with the state (and its associated ideas of ‘sovereign’ power and ‘closed’ societies) that they have struggled to ﬁnd adequate theoretical responses to the powerful new roles of non-state actors on the global stage. On the one hand, communitarians can tend to serve as political stalwarts for an increasingly obsolete statist world order, while on the other, cosmopolitan models are somewhat hampered by their commitment to the unrealistic goal of global constitutionalism. The ‘pluralist’ democratic model that I have proposed here has the advantage of transcending the limitations of both ‘communitarian’ statism and ‘cosmopolitan’ constitutionalism. It achieves this by developing conceptual and institutional alternatives which deal directly and systematically with the myriad non-state actors—as well as the states and global institutions—that exercise signiﬁcant power in contemporary global politics. By so doing, this framework provides a theoretical basis for the progressive development of a new democratic reform agenda aimed at bringing greater democratic legitimacy to the exercise of global public power in all of its myriad institutional forms.

2 Public Power Beyond ‘Sovereign’ States 2.1. INTRODUCTION Claims about the democratic credentials of non-state actors such as NGOs— whether laudatory or derogatory—nearly always begin with some declaration about the scope of their power. Pronouncements about the power of NGOs in global politics are commonly framed in revolutionary language, characterizing the emerging power of NGOs as a signiﬁcant challenge to the dominant power of states within global politics. In one widely cited passage, Lester Salamon described the rise of NGOs as a ‘global associational revolution’, evoking connotations of a non-governmental insurrection against the power of states. 1 Echoing this suggestion, Jessica Mathews claimed that in the face of the growing power of NGOs, ‘the steady concentration of power in the hands [of states that] began with the Peace of Westphalia, is over, at least for a while.’ 2 United Nations SecretaryGeneral Koﬁ Annan has similarly compared the power of NGOs to that of the most powerful states, claiming that the global community of NGOs ‘has become the new Super-Power—the peoples determined to promote better standards of life in larger freedom’. 3 It makes good sense to preface a discussion of NGOs’ democratic legitimacy with a discussion of their power, because democratic norms (as I have argued) are rightly applied only to the political subject of ‘public power’. For this reason, any demands for the democratization of NGOs must begin by establishing that they exercise the kind of ‘public power’ that would warrant such democratic imperatives. In the following chapter, I demonstrate how myriad NGOs can indeed be seen to wield public power within the contemporary global political arena. But before doing so, I think it is helpful to say more to bolster my theoretical account of public power, by examining and refuting some plausible theoretical arguments in favour of the traditional view that public power is uniquely tied to the institutional properties of statehood. The arguments I present in this chapter are accordingly directed primarily to those sceptical readers who may at this point remain unconvinced of Chapter 1’s 1

Lester M. Salamon, ‘The Rise of the Nonproﬁt Sector’, Foreign Aﬀairs, 73/4 (1994). Jessica Mathews, ‘Power Shift’, Foreign Aﬀairs, Jan./Feb. (1997). 3 Koﬁ Annan, The Emerging Power of Civil Society, Text of Speech Delivered 14 July 1998 (United Nations, 1998 [cited February 2005]); available from http://www.un.org/News/Press/docs/1998/ 19980714.sgsm6638.html In this speech Annan refers to the community of NGOs as ‘global civil society’. 2

44

Public Power Beyond ‘Sovereign’ States

central claim about the independence of public power from institutions of ‘states’ or ‘sovereignty’. Readers who are already on board with this central theoretical proposition may ﬁnd little in this chapter to advance the book’s argument, and may wish to skim quickly through it, straight on to the following chapter. However, I take it that my attempt to sever the concept of public power from the concepts of ‘state’ and ‘sovereignty’ is suﬃciently subversive of widely held theoretical assumptions that some readers will not yet be convinced, on the basis of what I have said so far. There are several strands of argument, running through mainstream literatures in Political Theory and International Relations, which could be invoked to defend the traditional view that public power is necessarily linked in some way to institutional structures of statehood, or ‘sovereignty’; the material in this chapter represents my attempt to pre-empt and counter such arguments.

2.2. ASSESSING ARGUMENTS FOR STATE-BASED ACCOUNTS OF PUBLIC POWER The proposition that NGOs can exercise ‘public power’ may seem highly counterintuitive to some readers, who are accustomed to quite diﬀerent traditions of thinking about the status of non-state actors within democratic societies. It is customary in mainstream liberal traditions to take for granted separations among diﬀerent ‘spheres’ of society—most central of which is the separation of the ‘state’ (the locus of public power and authority) from ‘civil society’ (the locus of individual liberty and free association). As Michael Walzer argues, this division between state and civil society is a crucial partition within liberalism’s ‘world of walls’, which is a function of the liberal ‘art of separation’. 4 On this picture of the diﬀerentiated political roles within a liberal democracy, NGOs are identiﬁed as key players within the associational sphere of ‘civil society’, and thus precluded— eﬀectively by deﬁnition—from a role in the exercise of ‘public’ power. More broadly, the conventional association between ‘public’ power and the state is deeply entrenched in the liberal theoretical worldview and taken for granted in much liberal analysis. Robert Wolﬀ articulates the view of many liberals when he claims that ‘[p]olitics is the exercise of the power of the state, or the attempt to inﬂuence that exercise’, and consequently that political philosophy ‘must begin with the concept of the state’. 5 Even much liberal literature on ‘global governance’—a paradigm that explicitly ‘rejects the conventional statecentric conception of world politics and world order’ 6 —maintains a state-centric 4

Michael Walzer, ‘Liberalism and the Art of Separation’, Political Theory, 12/3 (1984), 315. Robert Wolﬀ, ‘The Conﬂict Between Authority and Autonomy’, in Joseph Raz (ed.), Authority, Readings in Social and Political Theory (New York: New York University Press, 1990), 20. 6 Held and McGrew, ‘Introduction’, in David Held and Anthony McGrew (eds.), Governing Globalization: Power, Authority and Global Governance (Cambridge: Polity Press, 2002), 9. 5

Public Power Beyond ‘Sovereign’ States

45

account of the distinction between ‘public’ and ‘private’ power. Writers in this literature commonly refer to the growing role of non-state actors in global governance as the ‘privatization’ of global ‘governance’ or global ‘authority’— terminology that maintains the association of ‘public’ with ‘state’, and ‘private’ with ‘non-state’. 7 The widespread association of the idea of ‘public power’ with the idea of the ‘state’ has the implication that public power is commonly assumed to entail a range of organizational characteristics linked to statehood. 8 There are three organizational characteristics of states that I suggest are especially strongly linked to the idea of public power in much theoretical writing. The ﬁrst of these is the reliance upon law-making and law-enforcement as the central institutional vehicle for the exercise of power; the second is a capacity to use force or violent coercion to sustain the exercise of power; and the third is a centralized or ‘constitutionalized’ structure of power. In what follows I raise and challenge some plausible arguments in support of the common belief that power must exhibit one or more of these characteristics to qualify as ‘public’ power. In doing so, I hope to demonstrate that none of these organizational characteristics of statehood is in fact essential for the exercise of public power, and to reinforce the case for adopting the account of public power proposed in Chapter 1.

2.3. PUBLIC POWER AND THE INSTITUTIONAL PRACTICE OF LAW-MAKING One important set of liberal arguments that could be advanced to defend the link between state power and ‘public’ power involves the claim that there is something special (and uniquely ‘public’) about the ﬁrst of the organizational characteristics of states I have identiﬁed: reliance upon law-making as the central institutional vehicle for the exercise of power. To clarify the way in which this organizational characteristic is distinct from the other two I will discuss (centralized power and the use of force), it is helpful to begin by emphasizing the diﬀerence between a focus on one particular theoretical account of law and law-making (one that 7 For an account of Held and McGrew’s use of this terminology, see Ibid. 10. See also articles by Mathias Koenig-Archibugi and Ngaire Woods in that volume. See also literature on what is sometimes referred to as ‘private authority’, which is the terminology used to refer to the authority (in the broad sense of legitimate power) of non-state actors in global politics. See A. Claire Cutler, Private Power and Global Authority: Transnational Merchant Law in the Global Political Economy (Cambridge: Cambridge University Press, 2003); A. Claire Cutler, Virginia Hauﬂer, and Tony Porter, Private Authority and International Aﬀairs, Suny Series in Global Politics (Albany: State University of New York Press, 1999); Rodney Bruce Hall and Thomas J. Biersteker, The Emergence of Private Authority in Global Governance (Cambridge: Cambridge University Press, 2002). 8 All such organizational accounts of public power are clearly consistent with ‘liberal’ conceptions of public power, whereby public power is something exercised by particular kinds of agents over others. On these organizational accounts, agents that wield public power are distinguished from agents that do not by one of (or some combination of) these characteristics of their organizational form.

46

Public Power Beyond ‘Sovereign’ States

reﬂects the other two criteria of statehood I have distinguished), and a focus on law-making per se. The close association between statehood and law-making has had the implication that some prominent theoretical models of law and law-making have characterized law as something that is enforceable (ultimately) through the use of violent sanctions and made by a uniﬁed ‘sovereign’ agent. 9 In this way, such theories connect the idea of law-making, at a deﬁnitional level, with the other two characteristics of statehood that I discuss below—the use of force and centralized power. In the context of global politics, however, most now recognize that law-making is actually a pluralist process in which multiple political agents participate, and moreover that normative rules need not be enforceable by violence to constitute law. 10 By characterizing law without reference to violent sanction mechanisms or a unitary sovereign, ‘law’ and ‘law-making’ can be theorized on their own terms, independently from theories of the state. The practice of law-making can be characterized without reference to either unitary power structures or the use of force, insofar as it involves a distinct institutional framework for the exercise of power, with four key characteristics. First, the practice of law-making is distinguished from certain other methods of exercising power by virtue of its concern with rule-making (as distinct from social control through broader norms, markets, or material social architecture). 11 Second, the rules of law are binding, in the sense that they make legitimate demands upon individuals for compliance—a characteristic that they share with moral rules, but not with others such as rules of protocol. 12 Third, even among such normative forms of rule-making, law-making is a distinctive institutional practice (and can thus be distinguished from morality) insofar as it involves distinctive forms of 9 ‘Command’ theories of law, such as those espoused by John Austin, claim that law must emanate from a single ‘sovereign power’ capable of issuing a command, rather than from some ‘uncertain aggregate of persons’. See John Austin, The Province of Jurisprudence Determined (London: J. Murray, 1832). Moreover, Austin and some later legal ‘positivists’ have argued that law requires enforcement through eﬀective sanctions, such as the use of force, to generate compliance. See also Hans Kelsen, General Theory of Law and State (New York: Russell & Russell, 1961). 10 First, the structure of law-making in global politics is radically decentralized; as Cassese puts it, ‘[t]he relations between States comprising the international community remain largely horizontal. No vertical structure has as yet crystallized, as is instead the rule within the domestic systems of States.’ See Antonio Cassese, International Law (Oxford: Oxford University Press, 2001), 5. Further, the criterion of force as the basis for international law, and for recognition as an authoritative actor within it, is no longer generally taken to require eﬀective control. As Anthony D’Amato has argued, international law can be enforceable through other mechanisms, which he characterizes as ‘reciprocal entitlement violation’. See Anthony D’Amato, ‘Is International Law Really “Law”?’, Northwestern University Law Review, 79/5–6 (1985). 11 See Joseph Nye and Robert Keohane, ‘Introduction’, in Joseph S. Nye and John D. Donahue (eds.), Governance in a Globalizing World (Cambridge, MA, and Washington, DC: Visions of Governance for the 21st Century and Brookings Institution Press, 2000), 12, for a discussion of the distinctions among various forms of social control, of which rule-making is only one. 12 See Anthony Clark Arend, ‘Toward an Understanding of International Legal Rules’, in Robert J. Beck, Anthony C. Arend, and Robert D. Vander Lugt (eds.), International Rules: Approaches from International Law and International Relations (Oxford: Oxford University Press, 1996), for a discussion of this point.

Public Power Beyond ‘Sovereign’ States

47

reasoning and normative argumentation. As Kratochwil puts it, law can be understood as ‘a choice-process characterized by the principled nature of the norm-use in arriving at a decision’, such that the distinctive mode of norm-use constitutes ‘legal reasoning as a particular “style” of practical reasoning’. 13 Finally, the practice of law-making must be characterized with respect to not only the speciﬁc contents of the reasoning styles that can be deployed but also the speciﬁc social contexts in which this deployment can eﬀectively occur. The distinct styles of argumentation through which law is constituted are not equally accessible to all social actors in all social contexts; rather, international legal discourse incorporates some stipulations with respect to how particular agents can gain recognition in the process of legal argumentation. Legal theorists of diﬀerent schools give diﬀering accounts of these: Hart characterizes these stipulations as ‘secondary rules’ of a legal system, 14 while McDougal and Lasswell point to the particular role of ‘specialized organs’ within social processes of law-making—specialized institutions that have been developed for the speciﬁc purpose of performing legal decision-making tasks. 15 In domestic legal systems, these institutions include legislative and judicial institutions; in global politics, the primary institutions are intergovernmental forums of various kinds, while other international organizations, courts, and agencies can play secondary and subsidiary roles. 16 With this understanding of the institutional practice of law-making in mind, we can recognize how law-making capacity is identiﬁed in much liberal literature as a deﬁnitive element of ‘public’ power. The suggestion is that power only qualiﬁes as public power if it is exercised through the creation and enforcement of laws. On this view, power exercised through other mechanisms—including ad hoc 13 Friedrich V. Kratochwil, Rules, Norms and Decisions: On the Conditions of Practical and Legal Reasoning in International Relations and Domestic Aﬀairs (Cambridge: Cambridge University Press, 1989), 18–19. 14 H. L. A. Hart, The Concept of Law, Clarendon Law Series (Oxford: Clarendon Press, 1961). 15 Myres S. McDougal and Harold D. Lasswell, ‘The Identiﬁcation and Appraisal of Diverse Systems of Public Order’, American Journal of International Law, 53 (1959), 10. See also Myres Smith McDougal, Studies in World Public Order (New Haven: Yale University Press, 1960); Myres Smith McDougal and Florentino P. Feliciano, Law and Minimum World Public Order: The Legal Regulation of International Coercion (New Haven: Yale University Press, 1961). 16 Intergovernmental institutions remain the key sites of international law-making because agreement among states (through treaties and custom) remains the key source of international law. Unilateral actions and declarations of states, certain judicial decisions and decisions by international organizations, and ‘general principles’ of international law can also play important roles. A wider range of institutions play a role in the formation of ‘soft law’—a ‘body of standards, commitments, joint statements, or declarations of policy or intention’, which do not impose legally binding obligations but nonetheless shape common guidelines and policies, and can gradually ‘harden’ into legally binding customary rules or treaty provisions. See Cassese, International Law, 117–61, for a discussion of the various institutions recognized within legal discourse as valid ‘sources’ of international law. See also Vaughan Lowe, ‘The Politics of Law-Making: Are the Method and Character of Norm Creation Changing?’, in Michael Byers (ed.), The Role of Law in International Politics: Essays in International Relations and International Law (Oxford: Oxford University Press, 2000), for a discussion of the role of ‘interstitial norms’—which ‘direct the manner in which competing or conﬂicting norms that do have their own normativity should interact in practice’—in international law, and the distinct institutional processes through which they are developed.

48

Public Power Beyond ‘Sovereign’ States

directives, non-legal rules, economic incentives, and social infrastructure—would not qualify as ‘public’. The prevalence of this view is reﬂected in the fact that the idea of public power is commonly used interchangeably with that of ‘authority’— a term that is predominantly deﬁned with reference to some idealized theoretical model of law-making. Deﬁnitions of ‘authority’, such as that provided by Joseph Raz, generally denote a form of power that is exercised through issuing rules that generate an obligation to obey—a form of power that is a key constitutive element of the institutional practice of law-making. 17 As Raz puts it, [t]he exercise of coercive or of any other form of power is no exercise of authority unless it includes an appeal to compliance by the person(s) subject to the authority. This is why authority is typically exercised by giving instructions of one kind or another. 18

In this book, I use the term ‘public power’ rather than ‘authority’ to avoid these legalistic connotations that the latter term has acquired. Nonetheless, it is helpful to highlight the semantic connection between these terms, since this clearly illustrates the deeper theoretical associations between law-making and public power that are commonly assumed. The proposition that law-making capacity is essential for the exercise of public power has some intuitive plausibility, insofar as it captures most instances of what we call ‘public’ power in the context of states. We do usually characterize all instances of law-making as ‘public’ power; moreover, within democratic states there are few instances of public power that do not involve law-making, since state power (generally taken as paradigmatically ‘public’) is exercised nearly exclusively through legal apparatus. This is true even of forms of state power that are exercised through control of public infrastructure (such as roads) via decisions about ﬁnancial expenditure, which do not directly involve issuing rules in the way we commonly think of these (such as criminal laws proscribing certain behaviours by individuals). It is true even of such examples as these, for two reasons: ﬁrst, public funding is commonly obtained through compulsory payment of taxes, as dictated by law; and second, government budgets and expenditure programmes require legislative authorization. However, establishing a contingent empirical correlation between law-making and public power in the context of state-based democracy is not enough to show that there is any necessary theoretical correlation. That is, identifying a correlation in the context of the state does not establish that there are good normative reasons for connecting the idea of public power to the capacity for law-making in global politics. We must therefore consider further: can we identify any convincing normative arguments for maintaining this connection in the global political domain? I can think of two plausible arguments that could be advanced for claiming that law-making capacity is essential for the exercise of ‘public’ power in any 17 A notable exception to this is the deﬁnition of authority provided by Robert Ladenson, which excludes reference to the obligation to obey. See Robert Ladenson, ‘In Defense of a Hobbesian Conception of Law’, Philosophy and Public Aﬀairs, 9/2 (1980). 18 Joseph Raz, ‘Authority and Justiﬁcation’, in Raz (ed.), Authority, 116.

Public Power Beyond ‘Sovereign’ States

49

democratic polity. First, it could be argued that law applies to all equally within a polity (at least formally), and in doing so it satisﬁes some egalitarian conception of ‘publicity’. While some non-legal forms of public control can operate in more uneven, partial, and ad hoc ways between and among various agents within a political community, laws formally apply equally to, and require equal compliance from, all members of some designated political community. For this reason, it seems to make at least semantic sense to refer to law-making power as ‘public’ in a way that more ad hoc forms of power, which impact unevenly upon members of some designated community, are not. Moreover, a normative argument could be made, in support of this semantic connection, that democratic principles (based in part upon the value of equality) can only be coherently applied to power structures that themselves apply equally to all. If the power structure that is the subject of democratic control does not apply equally to all, then it could plausibly be argued that the democratic control of these structures would fail to result in an equal democratic society (which some people may take to be the ultimate goal of democracy). Accordingly, it could be argued, only power that is ‘public’ in the sense of impacting equally upon all can be the rightful target of democratization. There are two points to make in response to this argument. First, even if we were to accept both the general normative claim that public power must be inclusive in the sense of impacting equally upon all within a society and the further proposition that all law-making satisﬁes this standard of equal impact, it would still not follow that law-making is the only kind of power that can have equal social impact in this way. Rather, it remains a theoretical possibility that there are some forms of non-state power that have as equal an impact upon the various individuals within a population as any legislation. Moreover, we can ﬁnd many empirical examples in practice of ‘public’ programmes and infrastructures—such as a public roadway system of the kind alluded to in a previous example—that can impact equally upon all, irrespective of whether the power to undertake them has been conferred through a legislative process or through some non-governmental organizational channels. My second response to this argument is simply to reject the normative claim that public power must be inclusive in the sense of impacting equally upon all within a society, since this claim is based on a mistaken view of the requirements of democratic equality. As I argued in Chapter 1 (and argue further in Chapter 4), a commitment to the democratic value of equality in a pluralist global order does not require public power to be constitutionally structured, in such a way that it impacts equally upon all individuals. Rather, a commitment to democratic equality requires only that public power is structured in accordance with a normative account of responsibility (for protecting others’ autonomy) which treats all individuals in a given democratic context with equal concern. This diﬀerence is signiﬁcant: whereas a law-based framework of public power treats all individuals equally at the institutional level of applying the same prescriptive rules to all, a pluralist framework demands equality only at the more abstract level of normative justiﬁcation. As such, a pluralist framework allows for the institutional

50

Public Power Beyond ‘Sovereign’ States

possibility that a structure of public power will create diﬀerent sets of constraints and opportunities for diﬀerent individuals that are subject to it (in circumstances where doing so does not violate the justiﬁcatory requirement of equal concern). If we accept my broader normative claim about the role played by the value of equality in structuring a framework of democratic public power, the equalitybased argument for linking the concept of public power to law-making capacity will be unconvincing. Another plausible argument for seeing law-making as an essential organizational characteristic of public power involves the proposition that law-making constrains the autonomy of those subject to it in some uniquely problematic way, and so is uniquely in need of democratic legitimation. Robert Wolﬀ seems to suggest something along these lines, by arguing that since law-making authority is ‘the right to rule’ and autonomy is ‘the refusal to be ruled’, authority and autonomy are irreconcilably in conﬂict. 19 The problem with this argument, however, is that even if we accept that law-making does constrain the autonomy of those subject to it in the special way that generates a demand for democratic legitimation, it does not follow that legal instructions must be the only forms of power that constrain autonomy in this special way. This is because social decisions and activities conducted through the vehicle of legislation, and those conducted outside of legislative institutions, can sometimes have wholly equivalent impacts on the autonomy of those subject to them. To take an example, we can consider two programmes of social expenditure that aﬀect individual autonomy (say, expenditure on the construction of a public roadway determining patterns of mobility in the community), one of which is performed by a state agent through the vehicle of legislation, and the other performed by a non-state actor outside the legislative process. We can assume, to avoid the introduction of complicating factors, that both programmes have the same funding source (say, an aid donation from some third party—to a state government in the ﬁrst instance and to an NGO in the second). If both of these expenditure programmes were to have equivalent impact on the autonomy of those aﬀected, an autonomy-based argument could provide no basis for designating the legislatively enacted programme as ‘public’ power (in need of democratic legitimation) and the non-state programme as ‘private’, and unproblematic from a democratic perspective. This suggests that although practices of law-making generally satisfy an autonomy-based criterion for delineating public power, non-legislative forms of power can also satisfy this criterion, and so the organizational characteristic of law-making is not in itself a criterion for public power. We can conclude here, then, that the organizational characteristic of law-making capacity has no special ‘public’ political status, and embodies no grounds for resisting the more general criterion of public power that I presented in Chapter 1. 19

Robert Wolﬀ, in Raz (ed.), Authority, 29.

Public Power Beyond ‘Sovereign’ States

51

2.4. PUBLIC POWER AND THE USE OF FORCE The next set of liberal arguments that could be advanced to defend the link between state power and ‘public’ power involves the claim that there is something special (and uniquely ‘public’) about the second of the organizational characteristics of states I have identiﬁed: the capacity for force or violent coercion. The state is commonly deﬁned, at least in part, according to the Weberian criterion of holding a monopoly of (legitimate) violence within a territory. 20 This is reﬂected in traditional accounts of ‘sovereign’ authority in international law. An historically important idea in international law has been that a political agent must exercise ‘eﬀective’ control over a territory in order to qualify for the status of ‘statehood’ or ‘sovereignty’, which has been achieved nearly invariably (though not necessarily exclusively) through the eﬀective use of force. 21 Correspondingly, it is sometimes thought that only power undergirded by capacities for force and violent coercion can qualify as ‘public’, insofar as it is ‘sovereign’ states of this kind that are identiﬁed as the appropriate targets for democratization. Even if the ‘monopoly’ element of this account of public power were to be abandoned, as I argue below in my defence of institutional pluralism that it should be (and if judgements about legitimacy were also set aside on the grounds that such judgements, at this stage of the analysis, would beg the democratic question), the use or threat of force could still remain as an essential organizational characteristic of ‘public’ power. As Elizabeth Anscombe articulates a version of this belief, ‘government is distinguished from authority in voluntary cooperative enterprises by “bearing the sword”, by its exercise, actual and threatened, of coercive force.’ 22 The suggestion here is that only power undergirded by the use of force or violent coercion requires democratic legitimation as ‘public’ power, while power undergirded by other, non-violent, social capacities should be regarded as ‘private’. At an intuitive level it seems plausible to link public power to the use of force in this way, since we do generally think that the use of violence to secure power requires democratic legitimation. But what liberal argument could be advanced in support of the designation of violent coercion as uniquely ‘public’? The most plausible such argument I can think of involves the claim that violence of any kind violates individuals’ autonomy in some special way, and so warrants special 20 See Max Weber, ‘Politics as Vocation’, in Max Weber, Peter Lassman, and Ronald Speirs (eds.), Political Writings (Cambridge: Cambridge University Press, 1994). 21 For a discussion of the traditional requirement of ‘eﬀectiveness’ and force in securing authoritative status under international law, see Cassese, International Law, 12–13. 22 Elizabeth Anscombe, ‘On the Source of the Authority of the State’, in Raz (ed.), Authority, 144. In this article, Anscombe is not seeking to give an account of public power, but rather one of ‘authority’, accordingly, she further argues that government must administer a system of justice to qualify as authoritative. The issue I am concerned with here, though, is not the legitimacy component of ‘authority’, but only the power component—and so the relevant point here is that the kind of ‘public’ power, for which she seeks this special form of legitimation, is that of force and violent coercion.

52

Public Power Beyond ‘Sovereign’ States

democratic legitimation for this reason. However, although the use or threat of violence is certainly a prime example of autonomy-constraining power, there seems no reason for thinking that it is the only kind of power that imposes problematic constraints upon individuals’ autonomy. On the contrary, most of us would readily accept that other (non-violent) forms of coercion, such as those involving the exploitation of an agent’s material deprivation, can restrict individuals’ autonomy in ways that are just as problematic. A starving person forced into the servitude of another to secure essential supplies of food, or a sick person forced into servitude to secure supplies of life-saving medicines, is hardly more autonomous than someone who is forced into servitude at the barrel of a gun. But if we are willing to accept the normative equivalence of these kinds of constraints upon individual autonomy, it follows that power wielded through these latter forms of coercion has just as great a need for democratic legitimation as does power wielded through the threat or use of violence. In other words, we could not conclude that the use of force is necessary for the exercise of public power; rather, that it is an instance of power that generally satisﬁes a broader autonomy-based criterion for identifying public power, of the kind outlined in Chapter 1. But the fact that the violent coercion exercised by states generally qualiﬁes as ‘public’ power does not preclude non-violent forms of power (wielded by both state and non-state actors) from similarly qualifying as ‘public’.

2.5. PUBLIC POWER AND CENTRALIZED OR ‘CONSTITUTIONALIZED’ DECISION-MAKING STRUCTURES The ﬁnal set of liberal arguments that could be advanced to defend the link between state power and ‘public’ power involves the claim that there is something special (and uniquely ‘public’) about the third of the organizational characteristics of states I have identiﬁed: centralized or ‘constitutionalized’ structures of political decision-making agency. The ideas of ‘centralized’ and ‘constitutionalized’ power are closely related insofar as both are concerned with the implications of the underlying structures that constitute political agency. The idea of ‘centralized’ power is more speciﬁc, however, since it entails not only that the power in question is constitutionally structured but also that these structures concentrate power within a single decision-making agency. Here I want to examine arguments for linking the idea of ‘public’ power both to this speciﬁc organizational form of constitutionalized power and to constitutionally structured power more broadly. This discussion is of course related to what I have already said about the role of constitutional structures of public power within the democratic project, but adds a new dimension to my analysis of the issue. In Chapter 1, I presented some arguments for rejecting the claim that a structure of power must be constitutionalized to qualify as ‘public’, but the arguments presented there were wholly pragmatic

Public Power Beyond ‘Sovereign’ States

53

in nature, and based on acceptance of a ‘realist’ philosophical orientation to the democratic project. Here, I want to consider some more ideal-theoretical arguments that could be advanced in support of insisting on the traditional connection between ‘public’ and ‘constitutionalized’ forms of power. First, let us consider what arguments might be advanced for the view that ‘public’ power must involve centralized power, whereby the power that is exercised over a society of individuals emanates ultimately from an institutionally uniﬁed agency of some kind. This characteristic of statehood is strongly linked to the traditional idea of ‘sovereignty’. 23 Such unitary power, which was once conceptualized as emanating from a single individual ‘sovereign’, is now more commonly linked to the agency of the ‘sovereign’ state. The agency of the state can be characterized as ‘unitary’ insofar as the authority structures linking the many individuals that comprise it are centralized through hierarchical relationships, which tie decisionmaking actors at the ‘centre’ to other social actors within this structure. What normative reasons could be presented for claiming that power emanating from a uniﬁed agency of this kind is uniquely ‘public’, and in special need of democratic legitimation? Very few arguments are explicitly advanced for this claim; when such normative signiﬁcance is attributed to political centralization, it is generally just assumed as axiomatic, through its association with the idea of the state and that of ‘sovereignty’. Nonetheless, there is a plausible argument that could be advanced for presenting institutional uniﬁcation of this kind as being crucial to the ‘public’ character of political power. It could be argued that a single dominant or monopolistic political agent at the centre of a system of power is necessary to produce responsible political agency, capable of exercising the sort of deliberate and coordinated actions that could eﬀectively be subject to democratic legitimation. Let me say more here about what I mean by the idea of responsible political agency. What I mean is that the power in question must be vested in organizational agents capable of producing outcomes based on intentional decisions (of the kind for which they can be held responsible), rather than distributed through diﬀuse social structures that are incapable of deliberate and responsible action. Without the ability to take responsibility for actions and their outcomes, a collective organizational entity cannot be expected to engage with and respond to the democratic demands of those subject to its power, and therefore cannot exercise the kind of ‘public’ power that we can reasonably seek to hold democratically accountable. 24 This is true at least if we accept that ‘ought implies can’, and that we 23

As I established earlier, ‘command’ theories of law claim that law must emanate from a single ‘sovereign power’ capable of issuing a command. For a broader discussion of the evolution and transformation of the unitary conception of sovereignty, see Daniel Philpott, Revolutions in Sovereignty: How Ideas Shaped Modern International Relations (Princeton: Princeton University Press, 2001). David Held has attempted to delink the conception of sovereignty from this unitary model, though this use of the idea of ‘sovereignty’ is somewhat unconventional in the literature. See David Held, ‘Law of States, Law of Peoples’, Legal Theory, 8 (2002). 24 I discuss the importance of responsible agency to democratic accountability in greater depth in Chapter 7.

54

Public Power Beyond ‘Sovereign’ States

cannot coherently apply democratic standards to those agents that are empirically incapable of meeting them. (It is, I would suggest, precisely this criterion that distinguishes an ‘NGO’ from a ‘social movement’, and explains why the former can reasonably be subject to demands for democratic legitimation within a liberal framework, while the latter cannot.) To exercise responsible agency of the kind that can feasibly be held democratically accountable, collective entities must have capacities to comprehend and engage in communication with other social actors, in order to formulate collective purposes guiding their organizational activity, and to take coordinated strategic action in pursuit of these purposes. A case could therefore be made for seeing the organizational characteristic of uniﬁed power as essential for the exercise of ‘public’ power, by advancing some empirical argument demonstrating that responsible agency, of the kind I have described, can only be achieved via a centralized structure of power. What empirical arguments of this kind could be found to advance such a proposition? First, it could be argued—in a manner akin to some ‘Hobbesian’ or ‘Realist’ writers in International Relations theory—that all political agents are motivated by the pursuit of self-interest, and so multipolar systems inevitably result in conﬂict rather than responsible political rule. By adopting such a Realist view, it could be argued that only unipolar systems (systems in which some unitary agent of power enjoys political supremacy) can achieve responsible political rule. In contrast, it could be argued that the pluralistic global domain is a system of anarchy whereby no agent has the kind of political supremacy that would enable it to exercise responsible power. Without a system of centralized power, the pluralistic domain of global politics would as such inevitably remain a Hobbesian state of perpetual war. Realists commonly argue that empires or ‘hegemony’ can moderate this state of conﬂict and achieve some responsible rule in global politics, since they restore some degree of political monopoly, but nonetheless it could be argued from this perspective that responsible rule can never be achieved in a genuinely multipolar—or ‘pluralist’—global order. 25 It certainly seems true that an organizational structure with some degree of centralization and hierarchy would provide an eﬀective way of facilitating the kind of communicative engagement and purposive activity that are necessary for responsible agency, since it would provide a collective entity with the requisite ability to coordinate decision-making and action. But the issue here is whether the centralization of power is the only way to achieve such responsible political 25

Although all realists deny the possibility of responsible political rule in a multipolar global order, some do claim that the more limited goal of ‘stability’ can be achieved in multipolar orders. See Kenneth Neal Waltz, Theory of International Politics (New York: Addison-Wesley, McGraw-Hill, 1979). However, the goal of responsible political rule, which is my focus here, is more extensive and demanding than the Waltz’s notion of ‘stability’. The idea of responsible rule that I am concerned with here does, however, have more in common with notions of ‘stability’ advanced by some ‘hegemonic stability’ theorists, such as Robert Gilpin. See Robert Gilpin, The Political Economy of International Relations (Princeton: Princeton University Press, 1987); Robert Gilpin, War and Change in World Politics, 1st Paperback edn. (Cambridge: Cambridge University Press, 1983).

Public Power Beyond ‘Sovereign’ States

55

agency. A range of arguments advanced in liberal institutionalist and constructivist International Relations literature on the concept of ‘global governance’ can be invoked to argue that the construction of a centralized political agency is not the only way to achieve such responsible agency. First, ‘liberal institutionalist’ arguments point to an alternative way in which responsible social decision-making and action can be achieved: through the establishment of various systems of rules and regimes that harness the harmonies of interest among the multiple actors in an interdependent global order. On this liberal account, the decentralization of power in a system only precludes purposive collective political action (and thus responsible collective political agency) if there are no institutions to coordinate the mutually beneﬁcial cooperative activities of social actors. 26 In the absence of centralized state-like power structures, it follows that responsible social decision-making and action in global politics can be achieved instead through instituting a range of coordinating rules and processes, commonly described as global ‘governance’. 27 The so-called ‘constructivist’ theorists of global governance reinforce these liberal arguments by documenting how coordinated social decision-making and action can also be achieved in the absence of centralized power through the operation of shared norms, values, and identities of various kinds. 28 In sum, liberal institutionalist and constructivist arguments suggest that systems of rules, institutional processes, and norms can serve as the functional equivalent of a centralized structure of public power, by regulating and constituting individual decisions in such a way as to facilitate responsible collective decision-making and action within the group. 29 26 See Robert O. Keohane, After Hegemony: Cooperation and Discord in the World Political Economy (Princeton, NJ: Princeton University Press, 1984); Andreas Hasenclever, Peter Mayer, and Volker Rittberger, Theories of International Regimes (Cambridge: Cambridge University Press, 1997); Stephen D. Krasner, International Regimes (Ithaca: Cornell University Press, 1983). 27 James Rosenau deﬁnes ‘governance’ as ‘systems of rule, as the purposive activities of any collectivity that sustain mechanisms designed to insure its safety, prosperity, coherence, stability, and continuance’. See James N. Rosenau, ‘Change, Complexity and Governance in a Globalizing Space’, in Jon Pierre (ed.), Debating Governance (Oxford: Oxford University Press, 2000), 171. Also see James N. Rosenau, ‘Governance in the 21st Century’, Global Governance, 1 (1995). For extensive discussion of the infrastructure of ‘global governance’, see Held and McGrew (eds.), Governing Globalization. 28 See, for example, John Gerald Ruggie, Constructing the World Polity: Essays on International Institutionalization (New York: Routledge, 1998); Friedrich V. Kratochwil and John Gerald Ruggie, ‘International Organization: A State of the Art on an Art of the State’, International Organization, 40/4 (1986); Alexander Wendt, ‘Anarchy Is What States Make of It: The Social Construction of Power Politics’, International Organization, 46/2 (1992); Martha Finnemore and Kathryn Sikkink, ‘International Norm Dynamics and Political Change’, International Organization, 52/4 (1998); Michael Barnett and Martha Finnemore, ‘The Politics, Power and Pathologies of International Organizations’, International Organization, 53/4 (1999); Thomas Risse, Stephen Ropp, and Kathryn Sikkink (eds.), The Power of Human Rights: International Norms and Domestic Change (Cambridge: Cambridge University Press, 1999). 29 On this account, the question would still remain of where these rules, processes, and norms themselves come from, and which actors exercise public power in determining them. I address this question in more depth in the following chapter with some illustrations of forms of norm-building power exercised by NGOs.

56

Public Power Beyond ‘Sovereign’ States

But even once we have set aside these ‘Realist’ arguments linking institutional centralization to responsible political agency in the sense I have described so far, it could be argued further that centralized power is necessary for ‘responsible’ agency in another important sense. If a political agent is to be eﬀectively subjected to democratic control it does not merely need to be responsible in the sense of having the capacity to formulate and act intentionally in pursuit of goals, it must also be responsible in the sense that it is possible for other social actors to hold it responsible. What this means is that it must be possible for other social actors to identify the key range of political actions the agent undertakes, and trace the key political impacts which result from these actions. Following from this, it could be argued that although the ‘global governance’ literature demonstrates that a pluralist order can generate responsible political agency in the primary sense, it does not show so clearly how responsibility for speciﬁc social outcomes produced through particular governance institutions can be attributed to speciﬁc agents within these collective institutions, such that they can be held responsible for these outcomes via institutions of democratic control. The problem here is that the multiplicity of actors participating in processes of global ‘governance’ may produce a vacuum of responsibility in this latter sense, since nobody is sure which agent or agents caused any given outcomes. 30 It is not usually diﬃcult to establish in general terms when political power is being wielded through governance structures. As a number of writers have argued, many— if not most—existing structures of ‘global governance’ are ‘distorted’ by power relationships, in the sense that they ‘[promote] the interests of the most powerful states and global social forces’, as a result of vast asymmetries of power among social actors. 31 In many cases it will be possible to identify the politically dominant actors within governance structures, and we would presumably say that these dominant actors should be attributed more responsibility than other participants for the outcomes produced by these structures. But it is one thing to recognize that certain actors are ‘dominant’ in general terms, and another—much more diﬃcult—task to trace the particular responsibilities of these powerful actors, for speciﬁc political outcomes in speciﬁc governance contexts. This problem is certainly signiﬁcant, and it may well be that there are many instances in which centralization is the best way of ensuring that responsibility can be allocated to speciﬁc agents for speciﬁc political outcomes, such that it is possible to hold decision-makers responsible for these outcomes. However, there are grounds for thinking that this will not be true all of the time, and that there 30

Criticisms of this kind are sometimes made about forms of ‘network governance’ in global politics. See, for example, Michael Zürn, ‘Global Governance and Legitimacy Problems’, Government and Opposition, 39/2 (2004). 31 David Held and Anthony McGrew, ‘Introduction’, in Held and McGrew (eds.), Governing Globalization, 13. See also Philip Cerny, ‘Globalization, Governance and Complexity’, in Aseem Prakash and Jeﬀrey A. Hart (eds.), Globalization and Governance (London: Routledge, 1999); Hurrell, ‘Global Inequality and International Institutions’, in Thomas Pogge (ed.), Global Justice (Oxford: Blackwell, 2001); Craig Murphy, ‘Global Governance: Poorly Done and Poorly Understood’, International Aﬀairs, 76/4 (2000); Cutler, Hauﬂer, and Porter, Private Authority and International Aﬀairs.

Public Power Beyond ‘Sovereign’ States

57

are still many cases in which such responsible agency can be identiﬁed in the absence of centralized political structures. Most straightforwardly, there are many instances of multilateral interstate institutions in which the responsibilities of particular state participants can be quite easily identiﬁed. Sometimes this is because the responsibilities of each state are formally assigned and entrenched within the organizational apparatus, and sometimes it is because the power dynamics of interstate politicking are suﬃciently transparent to enable tracing of the informal decision-making processes leading to outcomes. It is also commonly possible to trace in similar ways the roles of particular non-state actors in governance processes—either because their roles and responsibilities are formally codiﬁed in some way or because their activities are suﬃciently stable and transparent to expose the scope of their inﬂuence upon (and thus responsibility for) outcomes. (I will discuss many concrete examples of such non-state power in the following chapter.) On the basis of these considerations, we can conclude that a centralized structure of power is not necessary to generate responsible political agency, in either of the senses I have discussed, since many forms of such responsible political agency can be identiﬁed within the existing pluralist global political order. However, these considerations also point us towards the recognition that many other political outcomes within a pluralist political order are not generated through responsible political agency of the kind that can be held to democratic account. So although we do not need to go as far as centralizing political decision-making powers, the pluralist global order in its present institutional conﬁguration does not generate as much responsible political agency as would be optimal for the establishment of eﬀective democratic control. If we look more closely at the institutional contexts in which examples of responsible political agency do arise in the contemporary pluralist order, we can identify an alternative set of institutional characteristics (distinct from centralization) which appear to be important in generating responsible agency of the kinds I have discussed. First, political roles in the exercise of power must have some fair degree of institutional stability if it is to be practical to identify them as ‘public’ and subject them to democratic control. Social actors can sometimes generate signiﬁcant and problematic political impacts through haphazard or ‘oneoﬀ ’ decisions or actions, but it will not be feasible or productive to subject these instances of power to democratic control if the agents in question do not engage in political actions of these kinds in a reasonably ongoing fashion, in a reasonably predictable way. One key reason for this is that institutions of democratic control—both ‘social choice’ institutions for collective stakeholder decision-making and ‘representative’ structures of authorization and accountability—take time to establish eﬀectively, and work best as means for controlling political decision-makers wielding power over a longer time frame. For example, a wealthy individual could make a decision to donate her entire fortune to some philanthropic project with signiﬁcant political impact, of the kind that we would usually seek to subject to democratic control—such as some major infrastructure project. But it would not be

58

Public Power Beyond ‘Sovereign’ States

appropriate to designate this individual as an agent of ‘public’ power despite this impact, given that the individual would possess no ongoing capacity for the exercise of public power. Once her fortune had been spent in this one-oﬀ act, her capacity for future actions with similar impact would be eliminated, so by the time eﬀective institutions of democratic control could be established in response to the exercise of power it would be too late for these institutions to wield control over any signiﬁcant political outcomes. Another reason it is important that those wielding power do so in a reasonably stable and institutionalized fashion is that stability of political roles will generally make it much easier for others to trace and identify the political impacts generated by particular agents’ actions, which is a necessary pre-requisite for holding political agents responsible. To put this in more general terms, institutional stability of political roles can contribute to a second institutional characteristic essential for responsible political agency: transparency of political roles within the institutional structure—that is, the availability and accessibility of information about the kinds of political actions undertaken by particular actors, and the kinds of political impacts these actions generate. Some degree of institutional stability is an important pre-requisite for establishing transparency of political roles, but such transparency can be promoted also through other means. One important means is the formal codiﬁcation of the political roles and responsibilities of particular actors. Sometimes codiﬁcation can take the form of formal legal instruments (such as those establishing the institutional capacities and roles of an International Organization), and sometimes it can take more informal forms (such as Charters, Codes of Conduct, Organizational Goals of the kind adopted by some NGOs and corporations). Broader forms of organizational transparency—documenting not only generalized roles but also the nature and impact of speciﬁc decisions and actions—can also play an important role in identifying appropriate targets for democratic control. Institutions that provide social recognition of these organizational roles and responsibilities (such as codes of conduct for organizations formulated and monitored by external social actors—like corporate codes maintained by NGOs and governments) can also generate ‘external recognition’ of political roles, and further enhance their transparency. 32 In sum, all institutional mechanisms focused on clarifying and making available information about the roles and responsibilities of particular political actors can play an important role in establishing the pre-conditions for instituting democratic control, and thus instituting forms of political agency that are ‘public’ in appropriate ways. The recognition that these various organizational characteristics are important to the kind of ‘public’ power amenable to eﬀective democratic control does not support the claim that public power must be centralized in structure, as I have said; however, it does lend some credence to the suggestion that a constitutionalized 32 The claim that ‘external recognition’ of political roles is important to the status of ‘public’ political agency mirrors the familiar idea that ‘external recognition’—alongside ‘eﬀective control’— is a key constitutive element of ‘sovereignty’—a prominent institutional form of public power.

Public Power Beyond ‘Sovereign’ States

59

structure of power, more broadly conceived, may be helpful in generating responsible political agency. A constitutionalized structure of power may be helpful in generating responsible political agency insofar as it can help to stabilize and codify political roles and responsibilities, and thus generate the kind of recognition and transparency of political roles and impacts that I have explained is important for responsible agency. The more stable and formally codiﬁed a structure of power, and the more entrenched the broader institutions of social recognition surrounding it, the easier it will be to attribute responsibility for the political impacts generated by the agencies constituted through it. Correspondingly, the easier it will be to identify these political agencies as ‘public’, and subject them to democratic control. Insofar as a project of political constitution-building seeks explicitly to achieve the task of increasing the stability, clarity, and recognition of roles and responsibilities within a political structure, it is undoubtedly true that building a constitutionalized structure of public power would be a highly eﬀective way of instituting responsible political agency. Nonetheless, it is not obvious that the required degrees of institutional stability, and of codiﬁcation and social recognition of political roles, would be suﬃcient to constitute a fully constitutionalized structure of public power. In other words, while constitutionalization might be a clear and eﬀective way of generating these institutional characteristics, it is not obvious that we need to take the development of these characteristics as far as to constitute ‘constitutionalization’, if our goal is only to achieve the degree of responsible agency necessary to facilitate eﬀective democratic control. The question of exactly what degrees of institutional stability and codiﬁcation and social recognition of political roles are needed for democratic control is one that I cannot answer with any precision here. Here I want only to make the broader theoretical observation that it is possible to achieve responsible political agency within a fundamentally pluralist political structure, so long as the pluralist structure demonstrates enough of the other institutional characteristics that I have identiﬁed. What are the implications of this, then, for the account of public power I presented in Chapter 1? My account there of what it is that makes power ‘public’ is focused exclusively on the impact of power (and not at all on its organizational structure), so the question arises: do the concessions I have made here to the importance of certain organizational characteristics point to the need to amend my account of public power, to incorporate this organizational dimension? I do not think that my account of public power does need to be amended in this way, given the analytic purpose my theoretical model is intended to serve. The account of public power I presented in Chapter 1 was focused on articulating the normative considerations that should be at play in the delineation of public forms of power. Nothing that I have said here about organizational structure undermines the basic point that there is a normative imperative to subject all forms of power to democratic control when they have the right kind of impact on the autonomous capacities of groups of individuals. Rather, the issues I have raised here are concerned with the practical limitations on eﬀorts to subject power to democratic control. What I have said here recognizes that although there

60

Public Power Beyond ‘Sovereign’ States

might be good democratic reasons to want to control all instances of autonomyconstraining power, it will not always be possible to do so; rather, it will likely be possible to do so only when the power in question is structured in appropriate ways. But I do not want to incorporate these practical organizational requirements into my deﬁnition of public power, because I think that to do so would undercut the critical utility of my theoretical model. That is, it would direct theoretical and reformist energies away from the task of developing innovative institutional means of subjecting the fullest possible range of political actors (with public impacts in the primary normative sense) to eﬀective democratic control. If we were to think of the institutional characteristics I have identiﬁed here as deﬁnitional components of ‘public’ power (i.e. as elements of the subject of democratic control) this would imply that such institutional structures must already exist as pre-requisites for a project of democratization. Given the critical function that I would like the Stakeholder model to serve, I think it is more helpful instead to present the development of institutional characteristics of this kind as objectives of a project of democratization. This is because if we were to demand that political actors need already to wield power via fully stable and transparent structures in order to qualify as eligible candidates for democratization, then many highly tyrannical political agencies would be immediately oﬀ the democratic hook. It is characteristic of many political tyrannies that they try to conceal the full extent of their political powers and impacts, and the full range of strategies they employ to secure their political control; clandestine modes of political control can sometimes be helpful tools for political tyrants, since they can make the target of political reform more elusive and thus less vulnerable to oppositional political forces. As a result, I think it is more helpful to see these institutional characteristics as elements of an institutional framework for achieving democratic control of public power, rather than as elements of the institutional subject of democratic control— that is, as elements of public power itself. In line with this, I present these institutional components of responsible political agency later in this book as elements of institutional mechanisms of authorization and accountability, devised to bring public power under democratic control, rather than as elements of public power. But I should emphasize again that the decision to view these institutional characteristics in this way—as elements of authorization and accountability mechanisms, rather than as elements of a structure of public power—reﬂects a strategic calculation about the most eﬀective strategy of advancing a project of global democratization, rather than any conceptual assessment of the democratic function of these institutional forms. It would be perfectly coherent, from a conceptual perspective, to view these institutional characteristics in either of the two ways I have indicated; there is no wholly conceptual answer to the question of where the line should be drawn between those institutional characteristics that constitute power as ‘public’ (i.e. as an appropriate target for democratization) and those institutional characteristics that invest power with democratic legitimacy (by subjecting it to eﬀective democratic control). Rather, exactly where we choose to draw this line depends on how ‘realist’ we are in our orientation to the democratic project, and where along a progressive ‘realist’ path of democratic development we

Public Power Beyond ‘Sovereign’ States

61

judge ourselves to be situated in any given political context. The further advanced we are along this path the more it makes sense to talk of these characteristics as elements of ‘public’ power, and the less advanced the democratic project the more it makes sense to talk of them as elements of democratic control. My decision to view these institutional characteristics in the way I do therefore reﬂects both the ‘Realist’ orientation to the democratic project I defended in Chapter 1 and the empirical judgement that global democratization is still at a very embryonic stage.

2.6. CONCLUSIONS In this chapter, I have countered some plausible arguments for maintaining that there is a necessary connection between public power and the organizational characteristics of states or ‘sovereignty’, and defended the view that public power should be delineated in accordance with the impact-based criterion outlined in Chapter 1. This theoretical approach to thinking about public power has both normative and empirical advantages over traditional statist approaches. Delineating public power in accordance with this criterion has clearer normative foundations than traditional organizational accounts, and it also has clearer application to the myriad patterns and structures of political interaction and domination in global politics. As Andrew Hurrell argues, the current global institutional order is a ‘hybrid’ one ‘in which solidarist and cosmopolitan models of governance coexist, usually rather unhappily, with many aspects of the old Westphalian order’. 33 Insofar as power is exercised globally, on a centralized state-like model of the kind advocated by some cosmopolitan democrats, or insofar as it is exercised within discrete territorial states, the account of public power that I have given here can subsume them. But my autonomy-based account can also be applied to those more patchy and partial forms of power—exercised by and among non-state actors—which traditional organizational accounts of public power exclude from their purview. 33 Andrew Hurrell, ‘Global Inequality and International Institutions’, in Pogge (ed.), Global Justice, 41.

3 The Public Power of NGOs in Global Politics 3.1. INTRODUCTION In the previous chapter I developed and defended my claim that there is no reason in principle for thinking that only states can wield public power. Rather, nonstate actors such as NGOs should be identiﬁed as agents of public power when they impact, in autonomy-constraining ways, on some population of individuals. Having made this eﬀort to bolster my theoretical arguments about public power against some likely objections, this chapter takes up the task of explaining how my account of public power might play out in practice. Here I draw on the impactbased theoretical criterion of public power presented in Chapter 1, to illustrate some of the ways in which NGOs can in fact be seen to exercise public power of this kind in contemporary global politics. In what follows I distinguish, and present some examples of, two key forms of ‘public power’ wielded by NGOs in contemporary global politics: inﬂuence in regulative social norm-building and the imposition of material constraints upon populations of individuals. The following analysis of NGOs’ public power is not intended to be precise or comprehensive; my aim is only to delineate a range of forms of public NGO power upon which we may wish to focus critical attention, as likely candidates for democratization. At a normative level, my account lacks precision insofar as the relevant normative standards (of valued individual autonomy and egalitarian responsibility for its protection) will necessarily be context-speciﬁc, and must be worked out politically in each context in which democratization is to be pursued. There is therefore no uniform or deﬁnitive interpretation of the normative criterion that I can invoke here to underpin a precise account of which forms of NGO power should count as ‘public’. Instead, my goal here is to draw on examples of NGO power that I think are underpinned by plausible interpretations of the relevant normative standards, around which I would expect there to be a fair deal of agreement in many political contexts. My account here of NGOs’ public power also lacks precision at an empirical level, insofar as it is beyond the scope of my analysis to conduct a rigorous assessment of the scope and impact of NGO power—although such empirical analysis would certainly deliver greater insight with respect to the range of speciﬁc empirical practices that would qualify as ‘public’. The more cursory analysis I present here does, nonetheless, serve to identify some prominent instances of

The Public Power of NGOs in Global Politics

63

NGO practices that are plausible candidates for identiﬁcation as ‘public’ forms of power in the contemporary global domain.

3.2. PUBLIC POWER AND INFLUENCE IN REGULATIVE SOCIAL NORM-BUILDING The ﬁrst important way in which individuals’ autonomous capacity to pursue their goals can be aﬀected problematically is through the operation of regulative social norms that (directly or indirectly) proscribe speciﬁc activities. This form of public power not only encompasses power wielded through the traditional institutional instruments of law-making but also incorporates power wielded through non-legal regulative norms. While constructivists in International Relations theory have documented many ways in which regulative norms can be deployed to wield some form of ‘power’ in international politics, not all power of this broad kind will qualify as ‘public’. Such power will only qualify as public when the imposition of regulative norms impacts upon the autonomous capacities of a population in some problematic way. Whether the imposition of regulative norms satisﬁes this criterion will generally depend on the strength of the norms in question, and the severity of the psychological costs and social sanctions associated with violating them. 1 I suggest that it is possible in global politics to identify some norms that are strong enough to place signiﬁcant and potentially problematic constraints on individuals’ capacities to pursue their goals. Of these, international legal norms and policy norms generated through the UN system would perhaps provide the clearest examples. 2 We can also ﬁnd examples of norms generated through more informal non-state institutions—such as corporate codes of conduct established by non-state actors and imposed through eﬀective advocacy—that can have a signiﬁcant constraining eﬀect on those actors subject to them. 3 I discuss some of these in more detail below. There are a few cases where the regulatory norms in question apply directly to individuals—as in the case of certain human rights norms pertaining to war crimes, for example. However, there are few examples of norms of this kind; more commonly, the subjects of international regulatory norms are collective agents 1 When I refer to ‘social sanctions’ here, I am excluding sanctions that involve the use of force or the deployment of material resources, since I deal with these later, as distinct forms of public power. 2 As I noted in the previous chapter, it is common to refer to ‘authoritative’ laws as wielding autonomy-constraining power over those individuals subject to them. 3 Again, it is important to stress that the constraints imposed through the weight of the norms themselves are not the only kind of constraints associated with the imposition of such strong norms. When these norms are enforced through force or economic sanctions, this imposes other constraints associated with the other two kinds of public power that I discuss elsewhere in this chapter. In this section, though, it is the constraints imposed through the weight of the norms alone—rather than the further constraints imposed through their violent or resource-based enforcement—that is the focus of my attention.

64

The Public Power of NGOs in Global Politics

such as states and corporations. In such instances, the autonomy of individuals can be aﬀected indirectly, through ﬂow-on eﬀects of a state or corporation’s compliance with the norms. For example, the autonomy of some individuals may be problematically aﬀected through a state’s compliance with certain formal environmental norms regulating ﬁshing practices, if these norms prevent them from engaging in traditional ﬁshing practices that they depend upon for their livelihood. Alternatively, the state’s compliance with liberal trade rules could have a similar impact upon such individuals, by subjecting them to forms of economic competition that make their traditional productive activities unproﬁtable. Similarly, the autonomy of some individuals may be problematically aﬀected by a corporation’s compliance with labour-related norms regulating issues such as child labour, if this compliance results in their loss of livelihood, either through sackings or through the retreat of the corporation from operations within their community. 4 In order to relate the constraining impact of global norms to the exercise of public power, we must take the further step of linking the operation of norms to the actions of particular political actors. What role in the production and maintenance of regulative norms must a political agent have, then, to be attributed the ‘public power’ that is wielded through the impact of these norms? I propose that any political agent can be said to exercise public power if it plays some prominent and inﬂuential role in the processes of production and maintenance of autonomyconstraining regulative norms. It is crucial to recognize here that the standard of inﬂuence must be a relative one, such that an agent exercises public power if its inﬂuence over the process is signiﬁcantly greater than that of other social actors that are subject to the norms. It is thus the asymmetry in the roles of diﬀerent social agents in generating regulative norms that enables us to say that one agent exercises power over others through this norm-building process. How, then, can NGOs and other political agents exercise such signiﬁcant inﬂuence within the processes through which strong global norms are produced and reproduced? Although this is not the place for a rigorous empirical analysis of these processes, it is possible here to point in general terms to certain processes of authoritative normative production and reproduction that are especially susceptible to signiﬁcant inﬂuence by powerful elements within society, and so are especially likely to serve as institutional vehicles for the exercise of public power. One way in which certain political agents can exercise signiﬁcant inﬂuence in these processes is by gaining privileged access to those institutional arenas that are designated as key sites for authoritative normative production. In many social and cultural contexts, various religious or educational institutions may provide examples of such key institutional sites for authoritative normative production. In global politics, I suggest that prominent instances of such sites include institutions of the UN system that are generative of ‘soft law’ of the kind outlined in Section 3.1, as well as some emerging non-governmental institutions that I discuss in more depth below. In situations where access to these institutions is distributed 4 See Institute of Development Studies, The Cost of Compliance: Global Standards for Small-Scale Firms and Workers (2003); available from http://www.ids.ac.uk/ids/bookshop/briefs/Pb18.pdf

The Public Power of NGOs in Global Politics

65

with suﬃcient asymmetry—that is, when it is concentrated very strongly in the hands of a few, while others subject to the norms are largely denied access—then we can say that those who wield signiﬁcant inﬂuence in these processes of norm production are exercising ‘public power’ over those who have minimal access to these productive processes. To what extent do NGOs exercise public power that is based on their privileged access to the authoritative institutional sites through which strong regulative norms in global politics are generated? In considering this question, it is crucial to distinguish clearly between such public norm-building power and private forms of power exercised by NGOs, which similarly operate by building norms in global politics. There is now an established body of literature highlighting the ways in which NGOs exercise power in global politics through various advocacy activities that succeed in inﬂuencing and transforming other actors’ ideas and values, and building social norms. 5 As Kathryn Sikkink argues, NGOs are sometimes very eﬀective in the exercise of ‘soft’ or ‘communicative’ power: This may take the form of information politics, symbolic politics, accountability politics, or leverage politics. It may be relatively dull reports, or lively street protests, or private meetings, but in all cases, the stress is on changing discourses and practices. By creating new issues and placing them on international and national agendas, providing crucial information to actors, and most importantly by creating and publicizing new norms and discourses, transnational advocacy groups help restructure world politics. 6

Although such advocacy activities are undoubtedly important forms of NGO ‘power’, they often constitute private rather than public forms of power, for two reasons. First, some of the norms generated through advocacy work may not be strong enough, in terms of the regulative constraint they impose upon others, for us to consider them problematic from the point of view of individual autonomy. Second, in many advocacy contexts NGO activities do not enjoy the kind of privileged access that I have identiﬁed for norm-building activity to be problematic with respect to other individuals’ autonomy, rather, many of these wider normbuilding processes are open to a broad range of social actors on a more inclusive basis. Accordingly, when NGOs alter others’ preferences and build norms by presenting ideas and arguments at grassroots levels, through technological and institutional vehicles that are widely available to other political actors, the communicative power of NGOs should be seen as private power—residing in the sphere of ‘civil society’—rather than as the kind of power that should be subject to public democratic contestation and control. If eﬀective NGO advocacy does not always satisfy the criterion for this kind of public power, what are the circumstances under which NGO activities do satisfy 5 See Margaret E. Keck and Kathryn Sikkink, Activists Beyond Borders: Advocacy Networks in International Politics (Ithaca, NY: Cornell University Press, 1998); Sanjeev Khagram, James V. Riker, and Kathryn Sikkink (eds.), Restructuring World Politics: Transnational Social Movements, Networks and Norms (Minneapolis: University of Minnesota Press, 2002). 6 Kathryn Sikkink, ‘Restructuring World Politics: The Limits and Asymmetries of Soft Power’, in Khagram, Riker, and Sikkink (eds.), Restructuring World Politics: Transnational Social Movements, Networks and Norms, 305–6.

66

The Public Power of NGOs in Global Politics

these conditions? I suggest here that they exercise this form of public power through privileged access to a range of International Organizations and other less formal institutional sites in which strong regulative norms—legal and non-legal— are produced in global politics. 7 One of the important ways in which NGOs exercise public power—and have potential to exercise it further, should certain proposed reforms be adopted—is by playing an inﬂuential role in the production of the international laws and key policy norms that are generated through the various International Organizations and interstate decision-making forums entailed in the UN system. At the centre of the formal relationship between the UN and NGOs is the consultative status accorded to NGOs within the United Nations Economic and Social Council (ECOSOC). 8 The participatory entitlements accorded to NGOs through this mechanism range from the minimal entitlements of access to agendas and observer status at a restricted range of meetings, to more substantial entitlements including the right to place items on the provisional agenda of the Council and of the Council’s subsidiary bodies, and the right to present statements to members of the Council or subsidiary bodies. 9 Most of the funds, agencies, and programmes within the wider UN system have also developed a wide range of mechanisms to work with NGOs, many of which are similar to the procedures and arrangements for ECOSOC consultative status. 10 Of particular note is the Joint United Nations Programme on HIV/AIDS (UNAIDS), which is the ﬁrst UN programme to include NGO representatives on its governing body as full participants, rather than simply as observers. 11 The International Labour Organization (ILO) also warrants special mention as the only tripartite institution within the UN system. Its constituents are state governments, employers’ organizations, and workers’ organizations, all three of which are represented on an equal footing within each of the ILO’s main bodies. Thus, the ILO fosters consultative relationships with NGOs representing employer and worker constituencies via their separate inclusion, as well as by inclusion of NGOs within national delegations. 12 NGOs can also gain accreditation to attend UN conferences and other occasional events, which facilitates considerable participation in informal sessions. 13 Within the WTO, the role of NGOs has developed separately and more restrictively, due to its recent evolution from a series of 7

It is important to note that I do not claim that NGOs play an inﬂuential role relative to all other political actors in the production of these key international norms. Such a claim would clearly be false, since states—as well as some other key actors—play far more signiﬁcant roles than NGOs in the production of many international norms. All I claim here is that NGOs enjoy special roles and forms of access to these norm-making processes, which result in their role being inﬂuential relative to many other political actors who are excluded from participation in the production of these norms. The process through which these key international norms are generated is thus a pluralist one involving a range of actors who all exercise public power over excluded parties, at levels proportionate to their power in the process. 8 For details of these arrangements, see United Nations Secretary-General, Arrangements and Practices for the Interaction of Non-Governmental Organizations in All Activities of the United Nations System (United Nations, 1998 [cited February 2002]); available from http://www.globalpolicy.org/ngos/ docs98/koﬁ998.htm p. 1–2. 9 Ibid. 10 Ibid. 5–10. 11 Ibid. 9. 12 Ibid. 13 Ibid. 16–17.

The Public Power of NGOs in Global Politics

67

multilateral trade negotiations to a fully ﬂedged institutional member of the UN system. 14 However, in 1996 the WTO General Council adopted new guidelines for NGO participation, and since then there have been signiﬁcant moves towards enhanced NGO access. 15 This incremental progress towards greater NGO access within the various institutions in the UN system has not been smooth, and there has been pressure from several national delegations within these institutions to wind back some reforms. 16 At the same time, many NGOs and supportive national delegations are working to institute a broader range of formal mechanisms for NGO participation. 17 The rhetoric accompanying these proposals for greater NGO participation within the UN system has spoken of ‘partnership’ and ‘full participation’. Koﬁ Annan and his deputy, Louise Frechette, regularly referred to the need for a ‘partnership’ between the UN and ‘civil society’, 18 and Boutros Boutros-Ghali claimed that the United Nations is no longer ‘considered to be a forum of sovereign states alone’, as ‘[n]on-governmental organizations are now considered full participants in international life.’ 19 Despite this rhetoric, proposals for NGO participation do not extend to granting NGOs equal status with nation-states within the public decision-making processes of global institutions. Signiﬁcantly, the monopoly of states on decision-making mechanisms such as votes and vetoes is not currently being challenged, and it is only in the deliberative aspects of decision-making that 14 For a comprehensive historical account of NGOs’ role in the WTO, see Steve Charnovitz, ‘Participation of Nongovernmental Organizations in the World Trade Organization’, University of Pennsylvania Journal of International Economic Law, 17/1 (1996). 15 For details of these changes, see World Trade Organization, The WTO and Civil Society: Comments by the Director-General to US NGOs (2002 [cited February 2002]); available from http:// www.wto.org/english/forums_e/ngo_e/ngospe_e.htm World Trade Organization, WTO and NGOs: Relations with Non-Governmental Organizations/Civil Society (2002 [cited February 2002]); available from http://www.wto.org/english/forums_e/ngo_e/intro_e.htm 16 See Global Policy Forum, NGOs and the United Nations: Comments for the Report of the Secretary General (1999 [cited February 2002]); available from http://www.globalpolicy.org/ngos/ docs99/gpfrep.htm p. 4–5; James A. Paul, NGO Access at the UN (1999 [cited February 2002]); available from http://www.globalpolicy.org/ngos/analysis/jap-accs.htm 17 For discussions of various proposed reforms to increase the level and quality of NGO access to participation in decision-making, see Steve Charnovitz and J. Wickham, ‘Nongovernmental Organizations and the Original International Trade Regime’, 29/5 (1995); Charnovitz, ‘Participation of Nongovernmental Organizations in the World Trade Organization’; World Trade Organization, The WTO and Civil Society: Comments by the Director-General to US NGOs (cited 2002b); Global Policy Forum, NGOs and the United Nations: Comments for the Report of the Secretary General (cited); Commission on Global Governance, The Millennium Year and the Reform Process (1999 [cited February 2002]); available from www.cgg.ch/millenium.htm Peter Willetts, ‘Consultative Status for NGOs at the United Nations’, in Peter Willetts (ed.), The Conscience of the World: The Inﬂuence of NonGovernmental Organisations in the UN System (London: Hurst, 1996); United Nations SecretaryGeneral, Views of Member States, Members of the Specialized Agencies, Observers, Intergovernmental and Non-Governmental Organizations from All Regions on the Report of the Secretary-General on Arrangements and Practices for the Interaction of Non-Governmental Organizations in All Activities of the United Nations System (United Nations, 1999 [cited February 2002]); available from http://www.globalpolicy.org/ngos/docs99/koﬁ99.htm 18 See various speeches at: http://www.un.org (cited March 2005). 19 Boutros Boutros-Ghali, ‘We the Peoples: Building Peace’, paper presented at the UN Department of Public Information Forty-Seventh Annual Conference of Non-Governmental Organizations, New York, 20 September 1994.

68

The Public Power of NGOs in Global Politics

greater NGO participation is proposed. In the language commonly employed to describe their emerging role, it is proposed that NGOs should be given ‘a voice, but not a vote’ in decision-making within the UN system. Another important way in which NGOs exercise public power through normbuilding is via non-state institutional processes outside the UN system. There have been many initiatives in recent years by NGOs, often acting in collaboration with governments and other organizations, that have been devised to establish and monitor various social and environmental standards for corporations. Collaborative initiatives among multiple non-state (and sometimes also governmental) actors include a range of very diﬀerent schemes such as AccountAbility1000, the Ethical Trading Initiative, the Fair Labor Association, the Global Alliance, the Global Compact, the Global Reporting Initiative, the International Forest Stewardship Council, and the Worker Rights Consortium. 20 Some prominent NGOs, such as Amnesty International and Social Accountability International, have also developed their own standards, which they seek to impose directly upon corporations. These include initiatives such as Amnesty International’s Human Rights Principles for Companies and Social Accountability International’s Social AccountAbility 8000 framework of labour standards. 21 These various initiatives—and the many others that have been developed in recent years—employ very diﬀerent strategies in developing norms and imposing them upon individuals in global politics. Some of these regulatory processes involve the formulation and imposition of quite formal norms in the form of codes of conduct, while other processes entail quite emotive public advocacy campaigns, yielding broader and more emotionally grounded normative standards. In a detailed analysis of the forms of public power that may be entailed in each of these norm-building processes, it would be important to recognize and take account of these important empirical diﬀerences among them. For the purposes of the cursory analysis I am undertaking here, however, the non-state character of these various initiatives provides suﬃcient commonality to warrant examining them together. The norms developed and imposed through these non-state norm-building mechanisms can acquire special authoritative status, and thus impose some 20 For a detailed discussion and evaluation of some aspects of these multistakeholder initiatives, see Peter Utting, Regulating Business Via Multistakeholder Initiatives: A Preliminary Assessment (2001 [cited March 2005]); available from http://www.un-ngls.org/documents/publications.en/develop.dossier/ dd.07%20(csr)/Section%20II.pdf See also http://www.codesofconduct.org/interest.htm Simon Zadek, ‘The Future of Nongovernmental Organizations in a World of Civil Corporations’, in David Lewis and Tina Wallace, New Roles and Relevance: Development NGOs and the Challenge of Change (Bloomﬁeld, CT: Kumarian Press, 2000); Melanie Beth Oliviero and Adele Simmons, ‘Who’s Minding the Store? Global Civil Society and Corporate Responsibility’, in Helmut K. Anheier, Marlies Glasius, and Mary Kaldor (eds.), Global Civil Society 2002 (Oxford: Oxford University Press, 2002); Dara O’Rourke, Outsourcing Regulation: Analyzing Non-Governmental Systems of Labor Standards and Monitoring (2002); available from http://web.mit.edu/dorourke/www/ PDF/OutsourcingReg.pdf 21 For more details, see http://web.amnesty.org/library/index/ENGACT700011998 and http://www. cepaa.org/SA8000/SA8000.htm (cited March 2005).

The Public Power of NGOs in Global Politics

69

regulative impact on global actors, in two ways. First, the kind of special authoritative status that is attached to formal international law and UN institutions can to some extent be replicated by those NGOs that have forged for themselves highly trusted ‘brands’ in the domain of global norm-building. NGOs like Amnesty International are among the organizations that are most trusted by the public to ‘do the right thing’ in political life, and the trust they are accorded can ascribe special weight to the norms that they develop and impose in global politics. 22 Second, as Vaughan Lowe argues in some depth, the norms promoted by NGOs can feed back into the formal international legal system by acquiring the status of ‘interstitial norms’ of international law. 23 Lowe argues that certain norms, such as ‘sustainable development’, lack normativity as elements of customary international law, but nonetheless exercise ‘an interstitial normativity by establishing the relationship between the neighbouring primary norms when they threaten to overlap or conﬂict with each other’. 24 Although these interstitial norms ‘have no independent normative charge of their own’, he argues that they can nonetheless have a signiﬁcant impact on the operation of primary norms by directing ‘the manner in which competing or conﬂicting norms that do have their own normativity should interact in practice’. 25 These norms diﬀer from traditional treaty- and custom-based international law in the sense that states are not the only agents who can contribute to their development. 26 Accordingly, NGOs that have forged a special authoritative status in global politics—such as Greenpeace in Lowe’s example of norms of sustainable development—can contribute to the development of legal norms through their wider participation in normative debate and regulation in global politics. So far, then, I have outlined some of the ways in which NGOs enjoy an inﬂuential role (relative to many actors in global politics) in the production of certain strongly regulative global norms. It may be that not all of the norms produced through these institutions and processes would constrain the autonomy of those aﬀected by them in problematic ways; however, many of these norms (and thus the powers exercised in creating them) do problematically impact on individuals’ autonomy. This is because decisions to apply particular norms often entail tradeoﬀs between diﬀerent public social goals, such as health care versus environmental protection, with consequent trade-oﬀs also among the resulting levels of autonomy of the various individuals aﬀected by these policies. NGO critics Roger Bate and Richard Tren argue that there are many ways in which the impacts of NGO norm-building activity can be detrimental to some 22

See Edelman PR Worldwide, Edelman Trust Barometer 2005: The Sixth Global Opinion Leaders Study (January 2005 [cited March 2005]); available from http://edelman.com/image/insights/content/ Edelman_Trust_Barometer-2005_ﬁnal_ﬁnal.pdf See also World Economic Forum, Results on The ‘Voice of the People’ Survey on Trust (2002 [cited March 2005]); available from http://www.weforum. org/site/homepublic.nsf/Content/Annual+Meeting+2003%5CResults+of+the+Survey+on+Trust 23 Vaughan Lowe, ‘The Politics of Lawmaking: Are the Method and Character of Norm Creation Changing?’, in Michael Byers (ed.), The Role of Law in International Politics: Essays in International Relations and International Law (Oxford: Oxford University Press, 2000). 24 Ibid. 216. 25 Ibid. 26 Ibid. 219.

70

The Public Power of NGOs in Global Politics

groups of individuals. In general, they argue that NGOs, through opposition to free trade, to low wages for poor workers and to various products and technologies, ‘pursue politics that are detrimental to the long-term growth of Africa and even . . . exacerbate poverty and increase death and disease’. 27 In relation to environmental campaigns against the use of DDT in combating malaria, they argue that NGO environmental campaigns against DDT use have resulted in more malaria deaths and poverty in aﬀected communities. In relation to campaigns to persuade drug companies to lower prices of life-saving medications, they argue that the NGO focus on drug companies diverted attention and resources away from more basic causes of poor drug access such as donor funding and health care infrastructure. Finally, in relation to NGOs’ opposition to the distribution of genetically modiﬁed (GM) food as aid in Africa, they argue that NGO campaigns contributed to African governments’ decisions to reject GM food aid, exacerbating the famines in their countries. 28 Similar arguments have been made in relation to the many other campaigns in which NGOs have been engaged. Previously, I mentioned environmental and labour rights campaigns, and noted the detrimental impacts these can have on the autonomy of some individuals. Especially compelling examples of such impacts can be drawn from the experience of past NGO campaigns that sought to impose norms prohibiting child labour within certain communities. In some situations, these campaigns were successful in removing children from the jobs in which they had previously been employed. However, some analysts have claimed that one eﬀect of this displacement was to drive some of these children into prostitution— an outcome that most would agree was detrimental to the autonomy of these individuals. 29 The question of whether each of the various norm-building activities engaged in by NGOs is beneﬁcial or detrimental overall to the aﬀected community is of course a subject of much dispute, since the harms and advantages conferred by such activities are rarely distributed equally among all aﬀected individuals. Whether the trade-oﬀs among individual interests entailed in the judgements underlying these activities are judged as overall right or wrong, however, the point remains that these trade-oﬀs involve problematic impacts upon the autonomy of some individuals in each instance. As such, these activities should be seen as ‘public’ forms of power that require democratic legitimation, whatever judgements we may make about their beneﬁts overall. 27 Roger Bate and Richard Tren, Do NGOs Improve Wealth and Health in Africa? (2003 [cited February 2005]); available from http://www.aei.org/docLib/20030624_bate.pdf 28 Ibid. 29 See United States Department of Labor, The Apparel Industry and Codes of Conduct: A Solution to the International Child Labor Problem? (1996 [cited March 2005]); available from http:// www.dol.gov/ILAB/media/reports/iclp/apparel/apparel.pdf G. Stoikov, ‘Presentation on Behalf of ILO to Hearing of the United States Department of Labor’ (Washington, DC: 12 April 1994). For a discussion of some such problems that can arise in the process of tackling the problem of child labour, see Jo Boyden, Birgitta Ling, and William E. Myers, What Works for Working Children (Florence: UNICEF International Child Development Centre, 1998).

The Public Power of NGOs in Global Politics

71

3.3. PUBLIC POWER AND THE IMPOSITION OF MATERIAL CONSTRAINTS Individuals’ democratic autonomy can be aﬀected by non-state political agents in global politics not only through the imposition of strong regulative norms but also through material obstruction that entails the use or threat of force and/or resource-deprivation. The forms of power associated with the coercive powers of state agents are subsumed within this category of public power, although the range of material constraints that qualify as ‘public’ on my account here is broader. There is a wide literature on the conditions under which the restriction of an individual’s capacity to pursue her goals through force or resource-deprivation constitutes a problematic constraint upon her autonomy (or ‘freedom’, or ‘liberty’). 30 I do not attempt here to resolve or contribute to these debates; my more limited task here is simply to outline the key types of constraints upon individuals that can usefully be distinguished as plausible candidates for identiﬁcation as ‘public’ forms of power. As I have indicated, in the ﬁrst instance we can distinguish between material constraints that are imposed through the use of force or violence of some kind and those that are imposed through deprivation of crucial social resources. Within these categories, we can further distinguish at a general level between those constraints that constrain an individual’s autonomy through direct mechanisms (acting directly upon individuals) and those constraints that constrain autonomy through indirect mechanisms (acting through eﬀects on the social environment in which the individual lives). In terms of direct imposition of material constraint, it is easy to see how force or violent coercion can be employed directly against an individual to constrain her freedom of action. In global politics, most uses of force or violent coercion against individuals would be considered problematic with respect to the autonomy of those individuals, and would qualify as ‘public’ in this way. It is also easy to see how a political agent can constrain the free action of others by controlling the distributions of crucial resources to a group of individuals, and depriving some individuals within this group of the resources they need. The precise kinds of resource-deprivations that would be considered problematic from the point of view of individual autonomy are of course contested. In general terms, though, we can say that resource-deprivations involve acts of public power in situations where certain individuals are denied access to crucial social resources (such as food and water supplies, public transportation infrastructure, health care facilities) by powerful actors who control access to these resources. The control of (and restriction of access to) some portion of 30 For some important discussions of these ideas, see Isaiah Berlin, ‘Two Concepts of Liberty’, in Isaiah Berlin (ed.), Four Essays on Liberty (Oxford: Oxford University Press, 1969); David Miller (ed.), Liberty, Oxford Readings in Politics and Government (Oxford: Oxford University Press, 1991); Charles Taylor, ‘What’s Wrong with Negative Liberty’, in Charles Taylor (ed.), Philosophy and the Human Sciences, Philosophical Papers, 2 (Cambridge: Cambridge University Press, 1985).

72

The Public Power of NGOs in Global Politics

crucial social resources does not constitute public power in circumstances where these resources are in ample supply, since such control would not then have the requisite autonomy-constraining impact on others. Similarly, control of (and restriction of access to) non-critical social resources (such as luxury consumer goods of some kind) does not constitute public power, since the resources are not crucial enough to be considered pre-requisites for individual autonomy. Precisely which resources are crucial to individuals’ autonomy, and precisely how readily available these resources must be to individuals, remain matters of dispute. Nonetheless, it is a criterion of this kind that we must seek, in identifying which forms of resource-control count as ‘public’ and which count as ‘private’ acts of power. The autonomy of individuals can be further constrained by the power of other political actors through the indirect imposition of material constraints, based on violence or on resource-deprivation. That is, the power of other agents can constrain the autonomy of individuals by playing a signiﬁcantly inﬂuential role in contributing to the creation of a social environment that is conducive to, or permissive of, violence or material deprivation. The wider social conditions aﬀecting the levels of violence and resource-deprivation to which individuals are subject are a result of many political agents acting in aggregate, and it would be a rare situation in which sole, or even primary, responsibility could be attributed to a single political actor. Nonetheless, at times it is possible to identify the roles of some political actors as particularly critical in creating social conditions that aﬀect individuals’ autonomy, and in such instances the power wielded by these actors over such conditions can be seen as ‘public’ forms of power. In most situations, states are key political agents responsible for violent social conditions, although sometimes agents such as armed rebel groups, international armies, and occasionally NGOs can play key roles in the exercise of this kind of public power. There is a wider range of political actors who have signiﬁcant inﬂuence over social conditions that determine the availability and distribution of resources; these can include governments and private corporations as well as some NGOs.

3.3.1. NGOs, Public Power, and the Use of Force and Violent Coercion First, we can examine the extent to which NGOs exercise public power in global politics through the direct use of force. There are of course some non-state actors in global politics that impose force directly upon individuals, such as mercenary organizations and guerrilla organizations. 31 I do not discuss these here, however, because where such non-state use of violence occurs it is far more common for those aﬀected to seek its outright eradication than it is for them to seek its democratization; here I conﬁne my attention to examples of NGOs’ public power 31 See Bernedette Muthien and Ian Taylor, ‘The Return of the Dogs of War? The Privatization of Security in Africa’, in Rodney Bruce Hall and Thomas J Biersteker (eds.), The Emergence of Private Authority in Global Governance (Cambridge: Cambridge University Press, 2002).

The Public Power of NGOs in Global Politics

73

that are likely to conform more closely with widespread democratic intuitions among relevant stakeholder populations. 32 The most important instances of violence-based NGO power arise, then, not when NGOs themselves wield force, but when their activities or decisions are supported or enforced through the use of force by others—generally states or statesponsored international armies and police-forces. Decisions made (in part) by NGOs—within interstate organizations for instance—can sometimes be enforced through the use or threat of violence by states. NGOs do not generally participate in the kinds of international decisions that involve the use of international force, since they have negligible inﬂuence in UN Security Council or other such securityrelated decision-making. There are, however, many international laws and policies that are implemented and enforced domestically within states by police wielding force. The support of state and international armies and police-forces is also essential for many humanitarian NGOs operating in crisis situations—especially crises in conﬂict zones. The capacity of NGOs to perform even basic humanitarian activities such as the distribution of food aid, within refugee camps and wider communities, can often depend on the use or threat of force by those state agents that provide security to the NGOs in these contexts. 33 In such situations, the use of force in the enactment of NGO decisions and more general facilitation of their activities makes the power of NGOs in these situations eﬀectively dependent—at least to a signiﬁcant degree—on the use of force. Even though the NGOs do not wield force themselves in these situations, force is nonetheless an aspect of their power because it is instrumental in implementing their decisions and facilitating their activities which impact upon individuals. Accordingly, the NGOs whose actions are materially facilitated and enforced through the violence of armies and police-forces should in some sense be accountable for the impacts upon individuals of the violent or coercive enactment of their decisions. Since in most cases the use of force or violent coercion upon individuals is considered to impose problematic constraints upon those individuals’ autonomy, the impacts upon individuals are of the kind that 32 It would be interesting to examine in more depth the reasons why such non-state power is generally excluded from the kinds of NGO power subject to claims of democratization. One key reason may be that the arguments for centralization of public power discussed in the previous chapter have much greater merit in relation to the use of force than they do in relation to other kinds of power. That is, it may be that democratic regulation of violence is diﬃcult to achieve without the centralization of power in agents such as states, while other forms of public power can be responsibly exercised by plural non-state actors operating within decentralized structures. On the other hand, it may be that the use of violence by non-state actors could rightly be considered ‘public power’ of the kind that should be subject to direct democratic control, but that common intuitions have not yet come to terms with this democratic possibility. 33 For some discussions of such situations, see Nicolas de Torrenté, ‘Humanitarianism Sacriﬁced: Integration’s False Promise’, Ethics and International Aﬀairs, 18/2 (2004); CARE USA, Care Decries Murders of Aid Workers in Afghanistan (2004 [cited March 2005]); available from http:// www.globalpolicy.org/ngos/aid/2004/0227murder.htm Joanna Macrae and Humanitarian Policy Group Overseas Development Institute (London England), The New Humanitarianisms: A Review of Trends in Global Humanitarian Action (London: HPG Overseas Development Institute, 2002).

74

The Public Power of NGOs in Global Politics

require democratic legitimation, and thus we should see activities of NGOs that are supported by agents of violence in this way as acts of public power. 34 Sometimes NGOs also exercise public power in global politics in circumstances whereby their activities have wider social impacts that lead to some individuals being subject to increased levels of violence. In particular, much has been written about the ways in which humanitarian NGOs operating in conﬂict situations can sometimes contribute (unwillingly, although predictably) to the perpetuation or even escalation of violence. Mary Anderson argues that humanitarian NGOs can contribute to violence in conﬂict situations in a variety of ways. First, the aid resources they bring into the community can be appropriated by warring parties and used to support the war eﬀort. Sometimes warring factions may ‘tax’ the NGO for the right to deliver aid; aid may be stolen and redirected to the war eﬀort; resources given to civilians may be passed on to those engaged in ﬁghting; NGO-built infrastructure such as road systems can be used by militaries to assist their operations; and NGO-trained locals may be conscripted into military service. 35 Peter Marsden has similarly argued that NGO activities in Afghanistan during the period of the Soviet occupation at times contributed to violence within that community by bolstering the war economy, and enabling warring parties to divert all their resources to conﬂict by relieving them of their responsibilities to provide for the civilian populations under their control. Moreover, he argues, they contributed directly to the war eﬀorts of the Mujahidin ﬁghters in this conﬂict by providing ﬁrst aid to wounded Mujahidin, and by making decisions to operate almost exclusively in Mujahidin-held areas of the country. 36 The role of NGOs in contributing to wider conﬂicts through their humanitarian work has also been widely highlighted in relation to NGO operations in refugee camps in Goma in 1994 and 1995. Around 200 NGOs played key roles in running refugee camps in Goma during that period, and these NGOs encountered widespread criticism because the camps were used as safe havens for Hutu ﬁghters, and as bases for some military operations against civilians. 37 34 In such situations, NGOs are not the sole agents of public power that are accountable for the use of force; the state and international forces directly wielding the force, and making key decisions about when and where the force should be deployed, are in most cases the primary agents of this public power and the primary agents to be held democratically accountable for it. When NGO decisions are enacted through the use of force, however, they join the direct agents of force as democratically accountable wielders of public power. 35 Mary Anderson, ‘Humanitarian NGOs in Conﬂict Intervention,’ in Chester A. Crocker, Fen Osler Hampson, and Pamela R. Aall, Managing Global Chaos: Sources of and Responses to International Conﬂict (Washington, DC: United States Institute of Peace Press, 1996), 343–54. See also Mary B. Anderson, Do No Harm: How Aid Can Support Peace—Or War (Boulder: Lynne Rienner Publishers, 1999). 36 Peter Marsden, ‘Afghanistan: State Disintegration and the Role of NGOs’, in Jon Bennett (ed.), NGOs and Governments: A Review of Current Practice for Southern and Eastern NGOs (Oxford: INTRAC, 1997). 37 Alexander Cooley and James Ron, ‘The NGO Scramble: Organizational Survival and Transnational Action’, International Security, 27/1 (2002), 28.

The Public Power of NGOs in Global Politics

75

Furthermore, Anderson argues that NGO activities can contribute to violent conﬂict by fuelling attitudes within communities that exacerbate the antagonisms between factions. NGO advocacy that publicizes human rights abuses can both provoke defensive responses from perpetrators and promote dehumanized images of perpetrators in the wider community. More materially, she argues that ‘NGOs must choose to employ some people (and not others), purchase goods from some (and not others), and target their aid toward some people (and not others); these decisions can fuel separate group identities, inequalities, and jealousies.’ 38 In such situations, we can say that NGOs indirectly impact upon the autonomy of individuals subject to this violence. Even though the impact is indirect rather than direct, however, the NGOs should still be held accountable for this impact, so long as it is suﬃciently predictable and preventable. It is important to emphasize that the role of NGOs in contributing to violence in these communities is far smaller than the role of direct perpetrators of violence, and often far smaller also than the role of state and international security-forces with more direct responsibilities for security situations. Nonetheless, the contributions of NGOs to security situations cannot be disregarded, since it can have real impact on the autonomy of some individuals who are subject to violence through NGOs’ involvement. Accordingly, the power that NGOs wield through their capacity to contribute to wider security situations within communities in which they operate should be regarded as ‘public’ and held to democratic account.

3.3.2. NGOs, Public Power, and the Control of Critical Resources Second, I will examine the public power that is exercised by NGOs when they control critical resources to which individuals have access—at both individual and social levels. One way in which NGOs can exercise resource-based public power is directly analogous to the way in which they can exercise public power through the use of force, as outlined above. Similarly, we can say here that NGOs exercise public power when international laws and key policy decisions that they have played a signiﬁcant role in producing are enforced through economic sanctions or other forms of economic pressure and coercion. For example, to the extent that NGOs play an inﬂuential role in the development of coercive programmes of ‘economic conditionality’, imposed by states or International Organizations such as the IMF, NGOs must be viewed as agents of this public power. Just as NGOs must be held accountable for those of their decisions that are enacted or enforced through violence by other parties, so they must be accountable for those of their decisions that are enacted or enforced through economic sanctions imposed by other parties. But the resource-based public power of NGOs is not limited to their reliance on the economic clout of other actors in the same way that their force-based 38 Mary Anderson, ‘Humanitarian NGOs in Conﬂict Intervention,’ in Crocker, Hampson, and Aall (eds.), Managing Global Chaos: Sources of and Responses to International Conﬂict, 348.

76

The Public Power of NGOs in Global Politics

public power is limited in this way. Rather, NGOs themselves wield substantial material resources in contemporary global politics, and deploy it with signiﬁcant material impacts on some target populations. According to some recent research, the NGO sector is now a huge economic force in global politics. In the thirtyﬁve countries surveyed in one study, the ‘civil society’ sector was found to have aggregate expenditures of US$1.3 trillion as of the late 1990s—a ﬁgure that represents 5.1 per cent of the GDP of these countries. Taken collectively, this makes the ‘civil society’ sector the world’s seventh largest economy, compared with the economies of states. 39 The overwhelming majority of the sector is engaged in service provision to communities, particularly in areas of education and social services. 40 The extent to which NGOs’ control and deployment of resources aﬀects individuals’ autonomy is dependent not on the absolute quantities of resources controlled, but the extent to which these resources are signiﬁcant to individuals, and suﬃciently monopolized by the NGOs. In many situations, the resources controlled by NGOs—such as private educational services in communities where public education is available, to take one example—are not critical in this way. For such reasons, it tends to be in developing countries rather than developed countries that NGOs’ resource-based power has the greatest impact on autonomy, because the NGO contributions are a higher proportion of the total, even though the absolute sums spent in these communities may often be smaller. It is therefore worth examining the role of NGOs in developing countries in particular. NGOs now play a major role in the provision of basic goods and services in developing communities, alongside states and international development agencies. From 1970 to 1985 the total development aid disbursed by NGOs increased tenfold, and by 1992 international NGOs were channelling over US$7.6 billion of aid to developing countries, out of a total of US$58 billion in overseas development aid disbursed that year. 41 By the end of the 1990s, between 15 and 20 per cent of total overseas development aid was being channelled through NGOs, 42 and the total value of this aid disbursed through NGOs was estimated at around US$10 billion annually. 43 This is signiﬁcantly greater than the amount of such aid disbursed through the UN system; in terms of net transfers, NGOs are collectively 39 Lester M. Salamon, S. Wojciech Sokolowski, and Regina List, Global Civil Society: An Overview (Baltimore, MD: Center for Civil Society Studies Institute for Policy Studies, The Johns Hopkins University, 2003), 13–15. These ﬁgures include not only the narrower group of organizations called ‘NGOs’ on some common organizational typologies but also a broader range of non-state institutions like schools and other educational organizations, which fall within common deﬁnitions of ‘civil society’ institutions. As I noted in the book’s Introduction, my focus here on forms of power rather than organizational structures means that these distinctions between ‘NGOs’ and other kinds of non-state institutions are not important to the arguments I am making here. 40 Ibid. 22. 41 John Greensmith, Trends in Fundraising and Giving by International NGOs (2001 [cited February 2005]); available from www.globalpolicy.org/ngos/fund/2001/12trends.htm 42 Ibid. 43 Leelananda Silva, Aid Cooperation, Discussion Paper for the UNDP/MDGD (1999 [cited March 2005]); available from http://magnet.undp.org/Docs/aidcoord/study1.htm

The Public Power of NGOs in Global Politics

77

the second largest source of development and relief assistance, second only to bilateral donors. 44 Given that NGOs also raise signiﬁcant funds of their own, and further contribute unquantiﬁed value through volunteer labour within the organizations, the real value of their contributions to the provision of crucial goods and services in developing countries is much higher. It is worth noting also that much of this aid money is channelled through ten or twenty big international NGOs, though the percentages channelled through southern NGOs are increasing. This is signiﬁcant because it means that there are several key NGOs that play an especially critical role in delivering resources to these communities. In evaluating the signiﬁcance of these contributions to the individuals within these communities—that is, in evaluating how critical these resource-provisions are to the autonomy of the individuals—it is necessary to examine both the extent to which there is dependence in developing communities upon resources disbursed by NGOs and the character of the goods and services being disbursed. A rigorous economic analysis of these issues is beyond the scope of my analysis here, though such an analysis would be very helpful in more clearly delineating the scope of NGOs’ public power. Nonetheless, we can point here to some ﬁgures that are at the least suggestive in identifying contexts in which NGOs can be said to exercise public power. First, identifying the extent to which communities depend on NGO-provided resources requires consideration of both the proportion of aid money disbursed through NGOs and the extent of reliance on aid money by developing communities to meet their resource needs. The signiﬁcance of development aid for many developing countries declined through the 1990s, as a result of a large increase in private capital ﬂows to developing countries. In 1990 these private capital ﬂows constituted only 34 per cent of total net resource ﬂows to developing countries, but by the end of the 1990s they accounted for 78 per cent. However, since 95 per cent of private capital ﬂows are received by only 26 developing countries and the balance of 5 per cent was spread among 140 countries, development aid still plays a critical role within many developing countries. 45 The role played by NGOs is particularly signiﬁcant in communities that are in conﬂict and post-conﬂict situations, and within certain remote, vulnerable, or marginalized groups within given countries, where NGOs are sometimes more eﬀective than state and international agencies in delivering critical resources. In Mozambique, to take one example, foreign aid in 1991 accounted for 78 per cent of the country’s GDP, making Mozambique the most aid-dependent country in the world. 46 Moreover, in Mozambique, a single NGO—World Vision International—recently became the largest donor and employer in the country, and possibly also the largest provider of welfare services. 47 Mozambique 44

Antonio Donini, Surﬁng on the Crest of the Wave until It Crashes: Intervention and the South (2000 [cited March 2005]); available from http://www.jha.ac/articles/a006.htm 45 Silva, Aid Cooperation, Discussion Paper for the UNDP/MDGD. 46 United Nations Department of Public Information, ‘Mozambique: Out of the Ruins of War’, Africa Recovery Brieﬁng Paper No. 8 (1993). 47 Donini, Surﬁng on the Crest of the Wave until It Crashes: Intervention and the South.

78

The Public Power of NGOs in Global Politics

provides an especially spectacular case of NGO domination of resource-provision, but NGOs play signiﬁcant roles in many other developing countries as well. In Kenya, for example—a state with a far more functional government sector than Mozambique—NGOs provide a high percentage of all health care services to the population. 48 The resources that NGOs play a key role in disbursing in developing communities are generally deployed in the provision of crucial goods and services, which most would agree are critical to securing individuals’ autonomy. Such services include not only food, medication, and water supplies but also health and education services, and infrastructure such as public roadways. In postconﬂict Afghanistan and Rwanda, for example, NGOs have played critical roles in delivering key health and educational programmes, which the fragile governments of these countries have been unable to provide. 49 In Afghanistan, NGOs have provided signiﬁcant support to the agricultural sector and played a key role in providing clean water in some regions. NGOs have also played important roles in areas such as waste disposal, repairs of basic infrastructure such as bridges and roads, and the clearing of mines. 50 It is clearly evident, then, that NGOs can play signiﬁcant roles within many communities, controlling many of the key resources upon which individuals depend to lead autonomous lives. However, while I have established that individuals’ autonomy can be aﬀected by NGO activities, it may not be immediately clear how these NGO roles can aﬀect individuals’ autonomy in a problematic way (thus qualifying as public power). Indeed, many people would argue that the impact of NGOs’ resource-provision upon individuals’ autonomy is overwhelmingly positive, and most would certainly agree that NGOs’ resource allocations in developing communities have many positive impacts. If we were to think that these impacts were positive for all aﬀected individuals, then it could be argued that this NGO power would not qualify as public. That is, if it were possible to identify certain communities in which the impact of NGOs’ resource allocations upon all individuals were incontestably positive, these NGO activities may not require democratic legitimation. In most real-world examples, however, the impact of NGOs’ resource-based power upon the capacity of some individuals to live autonomous lives is more contestable. Most straightforwardly, such NGO power can have a detrimental impact on individuals’ autonomy in cases where certain global norms, which are enforced through economic sanctions imposed by third parties, have detrimental 48 Alan Fowler, ‘Distant Obligations: Speculations on NGO Funding and the Global Market,’ Review of African Political Economy, 55 (1992); Richard Wamai, ‘NGO and Public Health Systems: Comparative Trends in Transforming Health Care Systems in Kenya and Finland,’ in International Society for Third Sector Research (ISTSR) Sixth International Conference (Ryerson University and York University, Toronto: 2004). 49 Bennett, NGOs and Governments: A Review of Current Practice for Southern and Eastern NGOs, 4. 50 Peter Marsden, ‘Afghanistan: State Disintegration and the Role of NGOs’, in Ibid. 20.

The Public Power of NGOs in Global Politics

79

impacts on individuals’ autonomy in the various ways discussed in Section 3.3.1. Beyond these situations of norm-enforcement, however, most detrimental impacts on individual autonomy of NGOs’ resource-based power arise as a result of the various ‘opportunity-costs’ of NGOs’ resource allocations in developing communities. First, it is important to consider the direct ‘opportunity-costs’ of NGOs’ activities, which arise from the diversion of resources, through NGOs, from other potential uses to which these resources could be directed. Many of the resources disbursed by NGOs in developing communities are acquired from private and governmental donors, who are seeking to assist the populations of developing communities and see NGOs as the most eﬀective vehicles for allocating these resources. If NGOs were not serving as intermediaries for these funds, then developing communities would not be deprived of all of these resources; rather, in many instances the resources now disbursed by NGOs would be delivered through diﬀerent channels—either directly by donors or, more probably, through local state institutions or International Organizations. In evaluating the impact of NGOs’ activities on the autonomy of recipient individuals, then, we should not be considering the impact of NGOs’ resource allocations relative to a situation where communities were entirely deprived of these resources. Rather, we should be considering the impact of NGOs’ resource allocations relative to situations in which the resources were allocated in a diﬀerent way within these communities— either by NGOs or by alternative organizations—through diﬀerent decisions with respect to priority programmes, and diﬀerent (more or less eﬀective and eﬃcient) management practices. Once we view the impact of NGOs’ activities in light of opportunity-costs as well as their actual outcomes, we see that the impacts for which they should be accountable are very similar to those of elected state governments with respect to resource allocation. In evaluating the impact of NGOs’ resource allocations on individuals’ autonomy, we must evaluate ﬁrst of all whether NGOs are disbursing the resources optimally overall—with the greatest eﬃciency and fairness, to achieve the best overall impact on the autonomy of all potential recipients. If they are not (or if this question is suﬃciently contested) then we should view their resource-based power as ‘public’ and in need of democratic control. Moreover, in evaluating NGOs’ impact on individual autonomy, we must consider the impact on each aﬀected individual, not just overall impacts; we must consider, that is, the impacts on particular individuals of the distributional decisions that NGOs make with respect to the allocations of resources. In allocating their available resources, NGOs commonly must decide how to prioritize particular programmes and populations—whether to focus on urban or rural populations, on health care or infrastructure projects, and so on. In making these distributional decisions, there will inevitably be not only winners but also losers—individuals who would have beneﬁted had diﬀerent priority decisions been made, but who failed to beneﬁt through the actual decisions. Given the impact of distributional decisions on these individuals’ autonomy, these decisions should also be regarded as

80

The Public Power of NGOs in Global Politics

‘public’ and subject to democratic contestation and control by aﬀected beneﬁciary communities. 51 Second, it is important to consider the indirect ‘opportunity-costs’ of NGOs’ activities, particularly those associated with their signiﬁcant role in delivering goods and services. Within some developing communities, there can be a major structural impact of this NGO role upon the longer-term capacity-building of local governance institutions (especially local state institutions). It is sometimes argued by critics of NGOs’ predominant role in resource-delivery to developing communities that the entrenched role of NGOs in key service-delivery roles can be counterproductive in the long term, since they erode—or at best stunt the development of—the capacity of local state institutions to perform these roles. This can lead in the longer term, some argue, to entrenched dependence on aid within these communities. As Kamal Malhotra argues, [o]nly governments and intergovernmental multilateral institutions are equipped to operate on the scale that is necessary if poverty-eradication, full employment and social integration are to be achieved in a sustainable manner. Assigning such roles to NGOs instead of to governments and intergovernmental organizations whose main duty this should be is, therefore, both inappropriate and counterproductive, especially in the long run. 52

To the extent that these criticisms may be valid, we can see that there is a potentially detrimental impact on many individuals’ long-term autonomy due to the lost opportunity for local capacity-building that may arise through NGOs’ operations. In relation to this potential problem, then, we can say that NGOs exercise public power through their delineation of the scope and character of their operations within communities, and the trade-oﬀs they make between the short-term resource-based beneﬁts they can deliver to populations and the longer-term needs of the communities to develop the institutional capacities required to meet their own needs. Many NGOs now recognize the importance of directing resources to capacity-building within their recipient communities, and include this among their key operational priorities, but judgements about how best to balance such capacity-building against short-term assistance are highly contested. According to Ricky Stuart of Oxfam, a ‘multi-horned dilemma confronting Northern international development NGOs is their role in direct implementation, versus their role in funding and “capacity-building” of local organizations, versus their role in partnering with Southern organizations to achieve global-level change.’ 53 The key point here is that the decisions NGOs make with respect to these trade-oﬀs can have signiﬁcant impacts on the autonomy of individuals—impacts that can be 51 In the following chapter, I discuss the question of how communities that have a democratic stake in such public power should be delineated. 52 Kamal Malhotra, ‘NGOs without Aid: Beyond the Global Soup Kitchen’, Third World Quarterly, 21/4 (2000), 665. 53 Rieky Stuart, ‘Ethics in Action: The Oxfam Response’, paper presented at the Ethics in Action: The Successes, Compromises, and Setbacks of Transnational Human Rights and Humanitarian NGOs, Workshop I, Carnegie Council on Ethics and International Aﬀairs, New York, 2002.

The Public Power of NGOs in Global Politics

81

detrimental for many in the short or long term if the wrong (for them) trade-oﬀs are made in allocating resources. Accordingly, these decisions by NGOs should be regarded as ‘public’ and held to democratic account by the aﬀected recipient communities. 54 Finally, in evaluating the extent of NGOs’ resource-based public power, we must recognize that the role of NGOs in delivering resources to needy communities, while signiﬁcant, is often secondary to the roles played by governmental and other international actors. Nonetheless, given the widespread signiﬁcance of their roles at a community level, many NGO activities in the areas of development and humanitarian work have signiﬁcant impacts upon some individuals’ autonomy, and should thus be considered ‘public’ and held to democratic account.

3.4. CONCLUSIONS In this chapter I have illustrated some plausible empirical grounds for my claim that non-state actors wield important forms of public power in contemporary global politics, and I have highlighted some key forms of ‘public’ NGO power. Both of the two key forms of public power that I have identiﬁed—signiﬁcant inﬂuence over processes of social norm-building, and the imposition of signiﬁcant material constraints—are strongly associated with the family of concepts to which ‘public power’ belongs (along with concepts of government, governance, authority, sovereignty, and statehood). My analysis here of public power diﬀers from analyses of these related concepts, however, in that it focuses on impact (on individual autonomy) as the criterion for identifying those instances of normbuilding, coercive, and resource-based power that are in need of democratic control. This analysis is important on one level because it identiﬁes some of the key forms of NGO activity that are the rightful targets of democratic critique, and that should serve as subjects for democratic control. But clearly delineating NGOs’ ‘public power’ is also important because it helps distinguish these public activities from other ‘private’ NGO activities, which should be seen as immune from demands for democratic accountability, or attributions of democratic deﬁcits. As I have argued, not all NGO activities satisfy the criterion of autonomy-constraining impact that makes them ‘public’ and in need of democratic control, and the common failure to clearly distinguish ‘public’ from ‘private’ activities can result in private NGO activities being subjected to unfounded democratic demands. The agenda of ‘NGO democratization’ and ‘NGO accountability’ has been enthusiastically embraced by several ‘right-wing’ think tanks and political opponents 54 NGOs operating in conﬂict situations face further dilemmas regarding how to balance material beneﬁts from resource allocations against potential detrimental eﬀects of the kinds discussed above. These wider trade-oﬀs, also, involve the kind of public power that require democratic control.

82

The Public Power of NGOs in Global Politics

of NGOs, as a means of weakening the public credibility and resource-bases of NGOs, which they perceive to be predominantly ‘left-wing’ in orientation. 55 Clear identiﬁcation of those organizations and activities that are legitimate targets of democratic demands can thus serve also as a basis for defending and strengthening those organizations and activities that do not qualify as public against any unwarranted political assaults. 55 Several articles critical of NGOs’ democratic legitimacy are available at http://www.ipa.org.au/ units/ngos.html (accessed January 2008) and at http://www.ngowatch.org (accessed January 2008). Criticisms of NGOs’ lack of democratic credentials have been used not only in public campaigns against these organizations’ reputations but also in lobbying for government policies restricting NGOs’ access to beneﬁts such as tax concessions associated with charity status in some countries. For discussion of these policies in the Australian context, see Ravi Tomar, Redeﬁning NGOs: The Emerging Debate, Current Issues Brief No. 5, 2003–4 (Information and Research Services, Parliamentary Library, Commonwealth of Australia, 2004 (cited); available from http://www.aph.gov.au/library/pubs/CIB/200304/04cib05.pdf

4 From Nation-States to ‘Stakeholder’ Communities 4.1. INTRODUCTION With its opening words, ‘We the peoples’, the preamble to the charter of the United Nations appeals evocatively to the democratic ideal of popular sovereignty. Gone is the statist language of ‘The High Contracting Parties’ that introduced the covenant of the organization’s predecessor—the League of Nations—and in its place is the language of global citizenship. It is not the sovereign states, but rather the ‘peoples’ of the world, whose authority the preamble invokes in establishing these global institutions. The choice of the plural ‘peoples ’ rather than the singular ‘people’ is also signiﬁcant. It implies that global citizens are invested with democratic status as members of communities, and not simply as individuals. These communities, which serve as democratic constituencies within the global polity, were originally understood to be nations (or at least those nations that enjoy the status of statehood). Accordingly, having invoked the authority of ‘the peoples’, the preamble delegates this authority to ‘our respective governments’ who, having ‘exhibited their full powers found to be in good and due form, have agreed to the present Charter of the United Nations’. Although the preamble’s proclamation that it was the peoples of the world who were creating the UN was primarily just a rhetorical ﬂourish in what was overwhelmingly a charter of sovereign states, it symbolized hopes that the organization could one day become something more. 1 And in recent years, global leaders have recalled these opening words in attempts to realize such progressive aspirations through democratization of the UN institutions and of wider global structures of public power. 2 The words have found particular resonance among advocates of increased participation in decision-making of non-state actors such as NGOs. UN Secretary-General Koﬁ Annan claims that the UN ‘derives its mandate to 1 This observation is also made in Commission on Global Governance, Our Global Neighbourhood: The Report of the Commission on Global Governance (Oxford: Oxford University Press, 1995). 2 Prominent discussions of democratization proposals by global leaders include Commission on Global Governance, The Millennium Year and the Reform Process (1999 [cited February 2002]); available from www.cgg.ch/millenium.htm Commission on Global Governance, Our Global Neighbourhood: The Report of the Commission on Global Governance; Boutros Boutros-Ghali and United Nations Dept. of Public Information, An Agenda for Democratization (New York: United Nations Dept. of Public Information, 1996).

84

From Nation-States to ‘Stakeholder’ Communities

work with civil society from the Charter itself and its opening words “We, the peoples” ’, 3 and accordingly argues that ‘a partnership with civil society is not an option; it is a necessity.’ 4 Former UN Secretary-General Boutros Boutros-Ghali similarly told a conference of NGOs that they are ‘a basic form of popular representation in the present-day world’, and consequently that ‘[t]heir participation in international relations is, in a way, a guarantee of the political legitimacy of those international organizations.’ 5 Only with NGO involvement, he declared, can the UN be ‘faithful to the urgent exhortation with which the preamble to the Charter begins: We the peoples of the United Nations’. 6 Invocation of the language of ‘the peoples’ in support of the participation of non-state actors such as NGOs in global decision-making has engendered some evolution in the meaning of the phrase. NGOs commonly stake their democratic claim to global public power on the grounds that they represent a multitude of intranational and transnational ‘stakeholder’ communities, such as religious groups, indigenous peoples, women, environmental and human rights advocates, trade unionists, business and professional communities, local communities, or the global poor. NGOs and their advocates—by claiming that ‘the peoples’ of the world include such non-national ‘stakeholder’ communities as well as national (or rather, nation-state-based) communities—have thus challenged the assumption that nation-states are the only ‘peoples’ that should count as democratic communities within the global polity. Although NGOs and their advocates are clear in claiming that certain ‘stakeholder’ communities have the kind of status that enables them to serve as legitimate agents of democratic control, the theoretical basis for this claim requires further elaboration. As I have previously explained, my proposition is that every individual whose autonomy is problematically constrained by the exercise of power is designated as a democratic ‘stakeholder’ in that power, and entitled to participate in its democratic control. My goal in this chapter is to build upon the arguments presented in Chapter 1 about the nature of a democratic ‘stakeholder’ community, and elaborate in more detail the implications of my claim that stakeholder communities should be delineated in accordance with the impact of public power upon populations of individuals. I begin my discussion in Section 4.1 by elaborating the concept of a stakeholder community, which is broadly derived from the global liberal-democratic discourse surrounding attempts to democratize non-state actors such as NGOs. Here, 3 United Nations Secretary-General, Arrangements and Practices for the Interaction of NonGovernmental Organizations in All Activities of the United Nations System (United Nations, 1998 [cited February 2002]); available from http://www.globalpolicy.org/ngos/docs98/koﬁ998.htm p. 2. 4 United Nations, Partnership with Civil Society Necessity in Addressing Global Agenda, Says Secretary-General in Wellington, New Zealand Remarks, Press Release (2000 [cited February 2002]); available from www.un.org/News/Press/docs/2000/20000229.sgsm7318.doc.html 5 Boutros Boutros-Ghali, ‘We the Peoples: Building Peace’, paper presented at the UN Department of Public Information Forty-Seventh Annual Conference of Non-Governmental Organizations, New York, 20 September 1994. 6 Ibid.

From Nation-States to ‘Stakeholder’ Communities

85

I explain how the concept of ‘community’ entailed in the stakeholder model diﬀers from those linked to the ‘closed’ model of democratic society that I discussed in Chapter 1. In Section 4.2, I then turn to the question of how this ‘stakeholder’ model of democratic community can be instituted, at a more practical level. To illustrate some of the considerations that must be invoked in such practical application of the stakeholder model, I identify and address some issues and dilemmas that commonly arise for practitioners who invoke the idea of ‘stakeholder’ communities in their attempts to democratize the activities of NGOs in global politics.

4.2. THE IDEA OF DEMOCRATIC ‘STAKEHOLDER’ COMMUNITIES The concept of ‘stakeholders’ in global public power is not a novel one; on the contrary, it is gaining widespread currency in contemporary discussions about the democratization of global political decision-making. Increasingly, public political agents including NGOs, corporations, and International Organizations consult a range of groups during their decision-making processes, on the grounds that these groups are ‘stakeholders’ in the decisions. According to a deﬁnition coined by the Environmental Council, ‘stakeholders’ are ‘people who have an interest in a particular decision, either as individuals or as representatives of a group’. 7 A democratic stakeholder community, then, is composed of all individuals whose interests are aﬀected (in a normatively relevant way) by the decisions made by a particular agent of public power. The idea of ‘stakeholders’ is sometimes used very broadly, to include individuals and groups with a very wide range of interests in a decision. However, it is arguable that in our highly interdependent world there are very few decisions that impact exclusively upon isolated groups, and that everybody is aﬀected at least to some minimal degree by all political decisions. Accordingly, there is a need to employ clearer normative criteria to specify the kinds (and intensities) of interests that qualify an individual as a democratic stakeholder in a decision. The arguments presented in Chapter 1 provide a rationale for embracing a more restrictive view of what should count as a normatively relevant interest or ‘stake’ in a decision. In accordance with the liberal representative conception of democracy presented there, we should narrow the conception of stakeholders down to those individuals who are subject to problematic impacts on their autonomous capacities. What this means is that the range of individuals who will qualify as democratic stakeholders is signiﬁcantly narrower than the range of individuals commonly identiﬁed as stakeholders in broader discussions of the concept by NGOs and other global political actors. To take a prominent example of the broader use of 7 Cited in Minu Hemmati, Multi-Stakeholder Processes: What Are They About? And How Should They Be Done? 4 (2001 [cited February 2002]); available from www.sustdev.org

86

From Nation-States to ‘Stakeholder’ Communities

the concept, an International Non-Governmental Organizations Accountability Charter adopted by Oxfam and several other major International NGOs deﬁnes NGOs’ stakeholders to include Peoples, including future generations, whose rights we seek to protect and advance; ecosystems, which cannot speak for or defend themselves; our members and supporters; our staﬀ and volunteers; organizations and individuals that contribute ﬁnance, goods or services; partner organizations, both governmental and non-governmental, with whom we work; regulatory bodies whose agreement is required for our establishment and operations; those whose policies, programmes or behaviour we wish to inﬂuence; the media; and the general public. 8

While this broader list of stakeholders may well provide a good account of the range of parties who have interests of some kind in NGOs’ decision-making, it is quite clear that not all the parties on a list of this kind will qualify as democratic stakeholders. It would be very hard to imagine circumstances under which members of NGOs’ regulatory agencies, for example, would be aﬀected in the autonomy-constraining sense required for designation as democratic stakeholders; in most cases also it would be hard to argue that ‘media’ institutions, ﬁnancial donors, and some others on this list would qualify as democratic stakeholders according to the criterion I have set out. But although this normative criterion does narrow down signiﬁcantly the range of plausible candidates for designation as democratic ‘stakeholders’ in relation to a given agent of public power, the criterion is still open to quite a wide range of diﬀerent interpretations. To begin with, there is a range of plausible interpretations of which forms of individual autonomy should be valued by liberals, and accorded protection through democratic institutions. There is an even wider range of plausible interpretations of what range of responsibilities particular political actors should be accorded for protecting the autonomy of others. I suggest that it is on this latter question that the most diﬃcult debates will turn, in identifying the democratic stakeholders of powerful particular actors. To see how the delineation of stakeholder communities will turn to a large degree on judgements about the distribution of responsibilities among diﬀerent political actors, we can consider for example the following question: how should we delineate the stakeholder community of a humanitarian NGO that controls certain resources for providing assistance in crisis situations? If this NGO decides to use these funds to provide assistance to the members of community A but not to the members of community B, who then are the stakeholders of this NGO? Are members of the beneﬁciary community, who are aﬀected by the direct actions of the NGO upon them, the only stakeholders? Or are the members of community B, who have been denied NGO assistance, also stakeholders by virtue of the fact that they have been adversely aﬀected (relative to what could have been if a diﬀerent decision had been made) by the NGOs’ policy decision to divert funds elsewhere? An answer to this question clearly depends on whether 8 See ‘International Non-Governmental Organizations Accountability Charter’; available from http://www.oxfam.org/en/ﬁles/INGO_accountability_charter_0606 (accessed January 2008).

From Nation-States to ‘Stakeholder’ Communities

87

a case can be made that the NGO is responsible for the autonomy of members of community B. It is not possible to identify any deﬁnitive or universally valid answers to such questions about the range of responsibilities individuals (and collective political agents) within the global order bear for the autonomy of others. As I explained in Chapter 1, the range of such responsibilities accepted by liberal democrats will vary in diﬀerent democratic contexts (i.e. in relationships between diﬀerent sets of political actors), so must be worked out separately in each, via appropriate forms of political debate and consensus-building. There is of course a great deal that philosophical discussion can contribute to political debates of this kind, and this broad set of philosophical questions—about the range of democratic responsibilities particular political actors should accept in particular relationships—is one that certainly warrants much further theoretical attention. It is beyond the scope of this book, though, to contribute any new arguments to such philosophical debates. My theoretical goal here is limited to articulating the broad conceptual framework of Global Stakeholder Democracy, and highlighting the substance of the underlying normative questions which must be resolved politically, within each speciﬁc political context in which democratic control of public power is to be established. (Though political resolution of these normative questions raises a number of additional practical challenges, about which I say a little more later in the chapter.) Beyond saying that the stakeholder criterion will need to be interpreted and operationalized within speciﬁc political contexts—in relationships between speciﬁc political actors—it is possible to say something more here about how this contextualism would play out in relation to the stakeholder communities of particular public political agents, such as NGOs. On this issue, it is important to highlight how the organizational complexities of NGO decision-making and activity, and the correspondingly complex patterns of political impact NGOs can have on varying populations, make the delineation of stakeholder communities a far from straightforward process. First, the one NGO might have diﬀerent groups of stakeholders implicated in diﬀerent elements of its activity. For example, an NGO such as Oxfam with a diverse range of operational activities might have multiple distinct stakeholder communities relating to diﬀerent areas of its activity and decision-making: one group of stakeholders in its emergency-relief operations, a diﬀerent set of stakeholders in its advocacy work on fair trade issues, and so on. For this reason it would be mistaken to think that the political task would be to identify one uniﬁed stakeholder community for each NGO (or other organizational entity wielding public power); it is not the case that a single NGO will always have a single stakeholder constituency. Rather, the normative objective should be to ensure that the right group of stakeholders is consulted in relation to the right set of decisions that a given organizational entity might be undertaking. Second, the one NGO might need to deploy diﬀerent interpretations of the stakeholder criterion to identify stakeholders in relation to the same type of its activity, but in diﬀerent political contexts in which that activity is undertaken.

88

From Nation-States to ‘Stakeholder’ Communities

The reason for this is that the precise interpretation of the stakeholder criterion that will apply to an NGO’s activity must depend in part on how problematic the impacts of its activities are viewed as being by the aﬀected populations in question, and these views might vary signiﬁcantly among diﬀerent aﬀected groups. To illustrate this point, we can take the above example of Oxfam campaigning on labour standards applicable to international trading relations. If Oxfam were to reﬂect upon who should count as stakeholders in its decision-making about what minimum labour standards to endorse through its advocacy work, it may begin by noting that these decisions will aﬀect the economic interests of a wide range of populations. These would include current workers in export-industries subject to the standards; potential workers elsewhere who might gain employment in these industries if labour standards were to be reformed (perhaps due to manufacturers relocating in light of recalibrated eﬃciency calculations); owners and managers of the companies engaged in the labour practices in question, whose proﬁts may be aﬀected; and families and wider communities dependent on income generated by these workers and corporate actors. Then, to determine who among these aﬀected parties would qualify as democratic stakeholders, it would be necessary to consider which of these various economic interests that might be detrimentally aﬀected implicate the autonomous capacities of the individuals concerned. A plausible way of making this determination might be to link the judgement to some operationalized benchmark of (autonomy-constraining) economic deprivation—which could be conceptualized as an agency-based (as distinct from welfare-based) account of ‘poverty’, or in some other way. Those individuals for whom the decision determined whether or to what extent they would end up living below this minimum benchmark—such as many workers, prospective workers, and their families—would thus count as democratic stakeholders. Those individuals who were not at risk of falling beneath this benchmark and had only higher-order interests in the decision—such as those corporate shareholders who might stand to lose only a small portion of their (large) proﬁts if tougher labour standards were to be imposed—would not count as democratic stakeholders. Once we start to think about delineating stakeholder communities according to a benchmark of this kind, it becomes clear that the operationalized criteria for identifying problematic impacts on individuals’ autonomous capacities will often need to vary in diﬀerent political contexts. In relation to the present example, the income levels and other speciﬁc measures of economic deprivation that would need to be invoked to operationalize a benchmark of autonomy-constraining deprivation would need to vary in diﬀerent social and political contexts in which Oxfam may operate, because of the signiﬁcance of socially relative deprivations for social dimensions of individuals’ autonomous capacities. For instance, it would seem implausible to use exactly the same standards of economic deprivation to evaluate the democratic stake of Southern-based workers and that of Northernbased shareholders or employees, since the ways in which speciﬁc economic deprivations are likely to impact upon individuals’ autonomous capacities in these very

From Nation-States to ‘Stakeholder’ Communities

89

diﬀerent social contexts may vary considerably, given diﬀerent levels of economic aﬄuence and diﬀerent value accorded to speciﬁc economic goods in each context. In this way, we can see that a single NGO may reasonably invoke variable speciﬁc criteria in relation to diﬀerent populations aﬀected by its decisions, in making assessments about how to identify the democratic stakeholders in these decisions. Although (such context-speciﬁc interpretations of) this normative stakeholder criterion must provide the primary basis for delineating stakeholder communities, a number of additional practical considerations may also need to come into play. These considerations mirror the sorts of practical considerations that can arise also in delineating public power, which I discussed in Chapter 2. In Chapter 2, I explained how the organizational characteristics of institutional stability, and of codiﬁcation and social recognition of political roles—while not normative requirements for designating a political actor as an agent of public power—are nonetheless required to some degree if the project of subjecting their power to democratic control is to be achievable in practice. Correspondingly, we can say here that—as a practical matter—a population of individuals will need to be aﬀected by a particular political actor (such as an NGO) in a reasonably stable, ongoing, and transparent fashion if these individuals are to qualify as democratic stakeholders. This is so because otherwise it is not going to be feasible to establish eﬀective institutions of democratic control linking them to power wielders in appropriate ways. As I discussed in Chapter 2, the fact of being subject to a one-oﬀ political impact by another actor will not be suﬃcient to constitute any workable democratic stakeholder community, since democratic institutions of social choice and control (of the kind I discuss in Part II of this book) can only work eﬀectively within political relationships that have some reasonable degree of stability over time. Moreover, it will not be possible for a population to establish eﬀective democratic control over forms of power that are insuﬃciently institutionalized to identify clearly which political decision-makers are responsible (and can be held responsible) for the relevant political impacts. Since these institutional characteristics embody pragmatic rather than principled requirements, it will necessarily be a matter of contextual strategic judgement what kind and degree of institutional stability and transparency the political impacts in question must have, in order for the aﬀected population to qualify as democratic stakeholders in any workable sense. I have so far sketched the central normative and practical questions that must be answered in delineating the ‘stakeholder’ communities that should be empowered to wield democratic control over agents of public power. However, the sense in which these collectivities of individuals constitute democratic ‘communities’ requires some further elaboration. This is required because these stakeholder groups constitute collectives of a somewhat diﬀerent kind from those that most democrats—who are committed to conventional ‘closed’ conceptions of democratic society—associate with the idea of a democratic ‘public’ or ‘people’ (or alternatively, a ‘demos’). Stakeholder communities will necessarily share some constitutive characteristics in common with traditional ‘closed’ democratic societies, of

90

From Nation-States to ‘Stakeholder’ Communities

the kind I discussed in Chapter 1, but there are also some important diﬀerences, which it is important to spell out clearly. To begin with, a fundamental pre-requisite for the existence of a democratic stakeholder community will be the existence of certain basic shared political values across the population in question: speciﬁcally, there must be basic consensus around the core constitutive democratic values of individual autonomy, and shared (egalitarian) responsibility for protecting the autonomous capacities of others within the population group. Agreement within the population in question on these core democratic values would be a pre-requisite for them to constitute a democratic ‘stakeholder’ community, since such agreement would necessarily underpin the legitimacy of any democratic institutions that would be established among them. 9 To be clear about the implications of this claim, it is not the case that everyone who shares these core democratic values will as such be members of the one stakeholder community; rather, particular stakeholder communities will be constituted from within the wider community of democrats sharing commitments to these core norms of democratic political legitimacy. For the framework of Global Stakeholder Democracy to be fully global it would be necessary for everyone to share a commitment to these core values at least at some very thin or minimal level. But the thick interpretations given to these core values can vary widely in diﬀerent contexts within the overarching global democratic society. The need for an underlying political consensus on core democratic values is of course shared in common with standard theoretical accounts of the constitutive features of a democratic community—so at this level the concept of a stakeholder ‘community’ is quite conventional. Since most democratic theories have been developed with the nation-state in mind, this kind of underlying democratic value consensus is sometimes talked about as though it must be embedded within a much thicker set of shared values and solidarities, usually associated with quite thick shared conceptions of egalitarian justice. 10 However, a democratic stakeholder community does not need thick shared values of this kind, and can instead accommodate quite broad social and cultural diversity, so long as there is some minimal thin agreement on core democratic values. Diﬀerences in the thick interpretations given to the core democratic values of individual autonomy and egalitarian responsibility will have implications for how particular stakeholder communities are constituted in these various political contexts; stakeholder communities will be delineated diﬀerently in political contexts where broader political values and solidarities are shared than in more diverse and less solidaristic contexts. The more extensive the range of egalitarian responsibilities towards others accepted within the group, the more forms of 9 The assumptions I am making here about the connection between political legitimacy and such underlying value consensus are articulated in greater depth in the Introduction to this book. 10 The democratic theories of John Rawls and Michael Walzer provide clear examples of this portrayal of democratic values as embedded within a broader social consensus on egalitarian justice. See John Rawls, Political Liberalism, New edn. (New York: Columbia University Press, 1996); Michael Walzer, Spheres of Justice: A Defence of Pluralism and Equality (Oxford: Blackwell, 1985).

From Nation-States to ‘Stakeholder’ Communities

91

power will qualify as ‘public’, and the wider the population of individuals that will qualify as ‘stakeholders’. But even when the range of egalitarian responsibilities endorsed within a group are quite narrow (such as within certain transnational political contexts where agreement may be limited to commitment to protect a very minimal set of basic rights) a democratic stakeholder community can still exist. The diﬀerence there will be, though, that the range of forms of power qualifying as ‘public’ will be relatively narrow, and the range of individuals qualifying as stakeholders will be correspondingly slim. Another important characteristic of stakeholder communities is that membership within them is non-exclusive. That is, membership within any given stakeholder community does not exclude simultaneous membership in other stakeholder communities. An individual can be a member of multiple stakeholder communities concurrently, and the populations of diﬀerent stakeholder communities can thus be overlapping rather than discrete. For any given individual, membership of a stakeholder community is partial rather than all-subsuming, with respect to her overall political identity. As a practical matter, it is the case that under certain conditions—such as those of a hypothetical ‘Westphalian’ world in which all public power is concentrated in sovereign states—existing stakeholder communities could be ‘closed’, rather than partial and overlapping. This could be so since in a ‘Westphalian’ state-based political order, in which all public power were concentrated in states wielding power over national populations, closed national communities would be the only groups to qualify as ‘stakeholder’ communities in accordance with the autonomy-based criterion I outlined previously. However, insofar as my empirical claims in the previous chapters about the pluralist structure of global public power are well-founded, it follows that stakeholder communities will, in practice, be ‘open’ and overlapping. This non-exclusive quality of membership is not unique to stakeholder communities, insofar as the multiple overlapping communities internal to a multilevel constitutionalized order (such as in a federal state or cosmopolitan structure) will have the same ‘open’ structure viewed independently from one another. However, the multiple communities within an overarching constitutionalized political order are still structured as part of a ‘closed’ community when viewed collectively: the substantive rights and duties constitutive of membership in each of these multiple communities are wholly compatible with the rights and duties constitutive of membership in each of the others, since they are all articulated through a cohesive overarching constitutional structure. As a result, there is no possibility of conﬂict between the political roles and responsibilities that an individual may possess via membership in these multiple political communities, so any given individual will experience no conﬂicts between the various elements of their overall political identity. In contrast, the multiple overlapping stakeholder communities within a ‘pluralist’ framework of Global Stakeholder Democracy do not form part of an overarching ‘closed’ or ‘constitutionalized’ global order that structures all of individuals’ political roles and responsibilities in a coherent and uniﬁed fashion. As a result, it may sometimes be the case that the concrete roles and responsibilities accorded

92

From Nation-States to ‘Stakeholder’ Communities

to particular individuals via their various political memberships (within multiple stakeholder communities) may generate conﬂicting demands on their political resources and energies, which are not neatly mediated and reconciled through a unifying constitutional framework.

4.3. DELINEATING STAKEHOLDER COMMUNITIES IN PRACTICE: STAKEHOLDERS IN THE PUBLIC POWER OF NGO S So far, then, I have explained how the concept of a stakeholder community diﬀers from that of a ‘closed’ democratic society, and how a ‘pluralist’ global democratic order, composed of multiple overlapping stakeholder communities, might diﬀer from a ‘constitutionalized’ global democratic order. In what follows, I consider some questions about how stakeholder communities could be identiﬁed in practice, by examining a number of problematic issues that can arise for practitioners seeking to identify—and accord participatory entitlements to—the members of NGOs’ stakeholder communities.

4.3.1. Institutional Mechanisms for Identifying Members of Stakeholder Communities The ﬁrst challenging issue that arises in thinking about how to accord participatory entitlements to democratic ‘stakeholder’ communities concerns the institutional mechanisms that could be employed in practice to identify those individuals or groups that qualify as stakeholders in relation to a given NGO (or other agent of public power). As I have said, the determination of which individuals will qualify as ‘stakeholders’ in the decision-making processes of a given ‘public’ political agent will necessarily rest upon potentially contested normative conceptions of individual autonomy and (egalitarian) responsibility, as well as on potentially contested empirical assessments of where such relationships of public power are to be found in the real world. For this reason, there can be no simple and uncontroversial philosophical formula for identifying the stakeholders of an NGO (or other political actor) in speciﬁc real-world contexts. Instead, the identiﬁcation of an NGO’s stakeholders must ultimately rest upon a process of political contestation, through which some degree of consensus can be reached among interested parties regarding both the appropriate ‘thick’ interpretations of the stakeholder criterion in that context and the relevant empirical issues (about the nature and scope of political impacts being imposed in that context). On this basis, all parties can then achieve some mutual recognition regarding the forms of public power that the NGO in question is exercising, and the identities of the stakeholders over whom it is exercised. The importance of such mutual recognition within the process of stakeholder identiﬁcation is consistent with some of

From Nation-States to ‘Stakeholder’ Communities

93

the claims articulated in policy-oriented discussions of stakeholder consultation; for instance, the standards developed by the organization AccountAbility1000 stipulate that it is important to ‘[a]llow stakeholders to assist in the identiﬁcation of other stakeholders’. 11 Such processes for identifying stakeholders could in some circumstances be enacted through direct dialogue between an NGO and the community of prospective stakeholders, particularly in circumstances where this community is relatively limited in scope. For example, if a ‘development’ NGO were seeking to identify stakeholders in a health care project within a local territorial community, prospective stakeholders could quite feasibly be limited to those individuals residing in the local area. A dialogue among locals—and between locals and the NGO— could then yield some agreement on which members of this community would be signiﬁcantly aﬀected by the policy (thus qualifying them as stakeholders) and which members would not. In many other circumstances, however, the community of prospective stakeholders would be much wider and a direct dialogue between these individuals and the NGO wielding public power could be more diﬃcult to achieve. An environmental advocacy NGO exercising public power through involvement in norm-building within international institutions, for example, could prospectively impact upon a huge global population. Here, it would seem entirely impracticable to include such a vast population in a discussion of which among them are aﬀected by the NGO so as to qualify as its stakeholders. In such circumstances, it may be necessary in some way to centralize the discursive process through which criteria of autonomy and responsibility are determined and empirical assessments of the scope of an NGO’s public political activities conducted. That is, instead of leaving each NGO and its prospective stakeholders to settle the question of communal membership among themselves, some kind of jurisdictional tribunal could be established—at national, regional, global, or functional sectoral levels—for settling disputes over the identity of a given NGO’s stakeholders. 12 By centralizing the process in this way—and thus centralizing the resources available for including wide populations in the deliberative processes required to determine criteria for membership—these logistical problems arising in relation to NGOs with wide prospective impact could to some degree be overcome. However, such a solution would have the drawback that the wider the population included in the dialogue, the more diﬃcult it may be to reach agreement on the relevant criteria, and the ‘thinner’ the conceptions of autonomy and egalitarian responsibility underpinning these criteria would probably need to be. 11 AccountAbility, Aa1000 Overview (1999 [cited March 2005]); available from http://www. accountability.org.uk/uploadstore/cms/docs/AA1000%20Overview.pdf 12 David Held advocates something similar, which he calls a ‘public issue Boundary Court’, in David Held (ed.), Democracy and the Global Order: From the Modern State to Cosmopolitan Governance (Cambridge: Polity, 1995), 279.

94

From Nation-States to ‘Stakeholder’ Communities

4.3.2. ‘Internal’ and ‘External’ Stakeholders A second problematic issue that arises in thinking about how to accord participatory entitlements to democratic ‘stakeholder’ communities concerns the pertinence of a distinction between what are called ‘internal’ and ‘external’ stakeholders, which is commonly utilized in practical discussions of political accountability in global politics. The purported distinction between these two categories of stakeholder is characterized in the Global Accountability Report in the following way: Internal Stakeholders include an organization’s staﬀ, its shareholders, its member countries, national organizations and, in the case of international NGOs, supporters. These stakeholders are part of the organization and operate (in part or whole) within the organization. External Stakeholders are individuals or groups who are aﬀected by an organization’s decisions and activities but who are not formally part of the organization. 13

This distinction between internal and external stakeholders can be of practical value once we move to the next stage in a representative system, having established the boundaries of the overall stakeholder community. That is, it can be useful to employ in the process of establishing institutional linkages between stakeholders and agents of public power, to ensure representation of the former by the latter. Discussion of the appropriate character of such institutional linkages—which establish the ‘representative agency’ of the agents of public power—is the focus of Chapters 7 and 8 of this book. As I argue there, the representative agency of public power must be achieved by institutional mechanisms of ‘authorization’ and ‘accountability’ that link agents of public power to their stakeholders. Even without much elaboration of these institutional mechanisms it is easy to see that the mechanisms of authorization and accountability that are likely to be eﬀective for stakeholders who are already members of the same organizational entity as those who are exercising public power over them may well be diﬀerent from those that would be eﬀective for stakeholders without such organizational ties to their agents of public power. However, despite this potential pragmatic utility, it is crucial to recognize that no normative weight should be attached to this distinction. At a normative level, all stakeholders are entitled to equal representation (in the sense that their interests should be weighted in proportion to the extent of their stake in the outcome of decisions), and this entitlement is in no way aﬀected by whether or not they 13

Hetty Kovach, Caroline Neligan, and Simon Burall, Global Accountability Report 1: Power without Accountability (One World Trust, 2003 [cited March 2005]); available from http://www. oneworldtrust.org/documents/GAP2003.pdf p. 3. Robert Keohane makes a similar distinction in his discussion of global accountability, though he does not use the term ‘stakeholders’. Keohane claims that we can speak of ‘internal’ accountability when ‘the principal and the agent are institutionally linked to one another’, whereas ‘external’ accountability involves ‘accountability to people outside the acting entity, whose lives are aﬀected by it’. See Robert Keohane, ‘Global Governance and Democratic Accountability’, in David Held and Mathias Koenig-Archibugi (eds.), Taming Globalization: Frontiers of Governance (Cambridge: Polity Press, 2003), 141.

From Nation-States to ‘Stakeholder’ Communities

95

have pre-existing organizational ties with the agent that is exercising public power over them. If the distinction between internal and external stakeholders is to be employed to underpin the establishment of diﬀerent representative mechanisms of authorization and accountability for diﬀerent stakeholders, it follows then that the mechanisms must be devised to deliver all stakeholders—internal and external—normatively equivalent representation. More concretely, we can say that an NGO making a decision that has public eﬀects upon both members (internal stakeholders) and non-members (external stakeholders) should neither privilege nor disadvantage its own members in decision-making. This is so even though it may be pragmatic for the NGO to recognize that diﬀerent institutional structures and practices will be useful in reaching these diﬀerent stakeholder groups and establishing the kinds of linkages necessary to deliver eﬀective representation to all stakeholders.

4.3.3. ‘Primary’ and ‘Secondary’ Stakeholders It is also worth considering the pertinence of another distinction that is commonly made in practical discussions of NGO accountability, between what are called ‘primary’ and ‘secondary’ stakeholders. This distinction is not a precise one, and there is no clear deﬁnitional consensus. It is generally employed quite broadly to distinguish between individuals with a ‘critical’ or ‘fundamental’ interest in the decisions made by an NGO wielding public power, and those individuals with more trivial or marginal interests in the NGO’s decisions. The former are designated as ‘primary’ stakeholders and the latter as ‘secondary’ stakeholders. More concretely, primary stakeholders are sometimes taken to be the ‘beneﬁciaries’ of NGOs’ activities (those who are the primary targets of NGOs’ political activities), while secondary stakeholders are taken to be other involved parties such as NGO donors and employees. 14 At an abstract level, the notion that valid distinctions can be made between stakeholders with critical and trivial interests in a decision is entirely sound. In the ﬁrst instance, it is sound because such a distinction is required to distinguish between democratic stakeholders—those with an interest in a decision that satisﬁes the autonomy-based stakeholder criterion—and other kinds of stakeholders who have more trivial interests and so have no democratic claim to representation, although they may have other (non-democratic) kinds of interest-based claims upon an NGO, founded on ﬁnancial, professional, or other such interests. If the 14 See Hugo Slim, ‘By What Authority? The Legitimacy and Accountability of Non-Governmental Organisations’, Journal of Humanitarian Assistance (2002 [cited March 2004]); available from http:// www.jha.ac/articles/a082.htm P. Townsley, Social Issues in Fisheries, FAO Fisheries Technical Paper No. 375 (FAO, 1998 [cited March 2005]); available from http://www.fao.org/DOCREP/003/ W8623E/w8623e05.htm Patrizio Warren, Survey-at-a-Distance of Assessment of Stakeholder Participation in FAO Field Programme (Sustainable Dimensions, Sustainable Development Department, FAO, 2001 [cited March 2005]); available from http://www.fao.org/sd/2001/PE0401a_en.htm for some discussions of this distinction.

96

From Nation-States to ‘Stakeholder’ Communities

distinction between primary and secondary stakeholders were taken to correspond with this distinction between those with autonomy-constraining (public) and non-autonomy-constraining (private) interests, then the distinction between primary and secondary stakeholders would be centrally relevant to the project of delineating stakeholder communities. Primary stakeholders would be those with membership in democratic communities and entitlement to democratic representation, and secondary stakeholders would be those whose interests in a decision were not suﬃcient to entitle them to democratic representation. In many cases, this usage would seem to ﬁt with more concrete applications of the distinction, whereby ‘beneﬁciaries’ are designated as primary and donors and employees are designated as secondary stakeholders. Although actors such as NGOs’ donors and employees clearly have some kind of interest in the NGOs’ activities, it would often be hard to make the case that their autonomous capacities are at stake. In certain instances, of especially impoverished donors or professionally vulnerable employees, for example, such an argument could possibly be made, but in general the ﬁnancial interests of donors and the professional interests of employees would not be suﬃcient to qualify them as democratic stakeholders of the NGO. In this way, we can see that the distinction between primary and secondary stakeholders (interpreted in this way) has some normative substance. However, since the whole category of ‘secondary’ stakeholders would, in this case, fall outside the scope of democratic institutions, the distinction adds nothing to the theoretical framework that I have already explicated, and is not of additional theoretical utility here. The distinction between primary and secondary stakeholders could, however, have some practical application within a democratic system if it were interpreted instead as distinguishing two categories of democratic stakeholders: those whose autonomy may be constrained signiﬁcantly and those whose autonomy may be constrained only marginally by a particular NGO’s decisions. This would certainly be plausible at a theoretical level, since it is often the case that some individuals have deeper autonomy-based claims than others. If such a distinction were to have any practical utility within a democratic system, however, we would need to be very clear about how such distinctions would translate into diﬀering status or roles for these diﬀerent stakeholders within decision-making processes. If such distinctions among democratic stakeholders were to play a role in representative decision-making structures, their role would be conﬁned to the operation of whatever mechanisms an NGO established to ensure that conﬂicts among various stakeholder interests were resolved democratically. In situations where members of a group of stakeholders desire the same outcomes from a decision, diﬀerences in the relative strengths of their interest should make no diﬀerence to the representation they are entitled to receive in decision-making processes. If the outcome desired by each stakeholder is the same, then the same decision by the NGO will eﬀectively represent all stakeholders, whether primary or secondary. It is only when the outcomes desired by primary stakeholders and the outcomes desired by secondary stakeholders conﬂict that the relative depth of

From Nation-States to ‘Stakeholder’ Communities

97

their interest becomes relevant. In such circumstances, each NGO must establish some (aggregative or deliberative) decision-procedure that accords appropriate weighting to the diﬀerential interests of their primary and secondary stakeholders, as I will next discuss.

4.3.4. Representing Conﬂicting Interests Within Stakeholder Communities So far in my discussion of stakeholder communities I have talked about them as unitary entities which—once appropriately delineated—can be represented as a whole by the relevant agents of public power (such as NGOs). As I have just noted, however, in many situations the interests of the various stakeholders within a stakeholder community will conﬂict, and it is thus far from a straightforward matter for a single agent of public power (such as an NGO) to represent adequately the myriad interests of its stakeholders. Ultimately, of course, an NGO will have to make a ﬁnal decision, which is supposed to be representative of the stakeholder community in its entirety; the question, then, is how could a single NGO legitimately represent a range of conﬂicting interests simultaneously? In answering this question, it is important to recognize that the way in which any given NGO agent of public power will be required to deal with this problem of conﬂicting interests can vary, depending on the nature of the decision-making context in which the NGO wields its public power. To explain how this is so, it is crucial to distinguish clearly between two very diﬀerent, although related, kinds of stakeholder community or ‘people’, which can demand representation by an NGO in diﬀerent decision-making contexts. The ﬁrst kind of community can be characterized as a jurisdiction, or jurisdictional community. The second kind of community can be characterized as a constituency. At a broad deﬁnitional level, the diﬀerence between a ‘jurisdiction’ and a ‘constituency’ is straightforward: whereas a democratic ‘jurisdiction’ comprises a group of individuals subjected to the same agent of public power, a democratic ‘constituency’ is a group of individuals who share the same representative agent within some public decision-making process. Sometimes the boundaries of these two kinds of community are coterminous, such that the represented group referred to as the ‘constituency’ is the entire population of a jurisdiction. When the agent doing the representing is the central public political agent in a representative democratic polity, then we will refer to the entire population within the polity as the ‘constituency’ in question. 15 More commonly, however, the represented groups referred to as ‘constituencies’ are subgroups within the overall democratic polity, each of which is accorded a separate representative agent in the legislature or other decision-making structures through which public power is exercised. 15 It is in this sense of the term, also, that David Held can refer to the populations aﬀected by the power of non-state agencies such as corporations and International Organizations—the populations falling within the ‘jurisdictions’ of these agents—as ‘transnational constituencies’. See Held, Democracy and the Global Order: From the Modern State to Cosmopolitan Governance.

98

From Nation-States to ‘Stakeholder’ Communities

One way of understanding the signiﬁcance of this distinction is to recall the conceptual division that underpins the structure of the two parts of this book. On the one hand, jurisdictional communities—along with the (corresponding) scope of public power—delineate the scope and boundaries of the overall democratic decision-making ‘system’ in a given decision-making context. It is questions of this kind that have been my focus in the ﬁrst part of the book. In contrast, the delineation of multiple constituencies occurs within these jurisdictional boundaries; questions about the boundaries of multiple constituencies (along with questions about the structure of representative mechanisms) are concerned with the decision-making procedures that should be employed internally to the overall democratic ‘system’ in a given context. It is questions of this latter kind that are my focus in the second part of the book. Another way of understanding the signiﬁcance of this distinction is to recognize that these two diﬀerent kinds of democratic community are associated with the two diﬀerent kinds of public political decision-making structure. The ﬁrst kind of decision-making structure can be characterized as unilateral, insofar as a given agent (state or non-state) makes decisions about how to wield its public power independently (rather than in collaboration with other public political agents within a decision-making process that distributes power among them according to some democratic decision-procedure). As I discuss in more depth in Chapters 5 and 6, such democratic decision-procedures—or ‘decision-rules’—are generally either aggregative (whereby the interests of each participating agent are summed to generate a decision) or deliberative (whereby each actor advances arguments, then all deliberate and reach a decision through agreement). In order for us to identify a particular form of NGO power as unilateral, it is not necessary for that NGO to be the only agent exercising public power over a given population. Rather, what is crucial here is the nature of the decision-making process through which the NGO determines how to exercise its public power. It is true that all public power exercised by an NGO in contexts where it is the only agent exercising public power over the given population would qualify as unilateral, since its decision-making in such contexts would necessarily be independent. However, NGOs also engage in independent decision-making in contexts where many other public political actors are also operating, but where the decisionmaking processes of these various actors are not coordinated in any democratic way. Examples of the former kind (where an NGO is the only agent exercising public power over the given population) are relatively limited, but may include instances in which NGOs exercise public power over isolated populations that other public political agents have failed to reach, such as in certain development or humanitarian contexts. Examples of the latter kind (where an NGO is one of many independently operating public political agents) are more commonplace, and would include the many instances of NGO public power in norm-building outside the UN system and outside other such formalized collective decisionmaking processes, some of which I discussed in the previous chapter. Examples would further include instances in which NGOs exercise public power through

From Nation-States to ‘Stakeholder’ Communities

99

material constraints in development or humanitarian contexts in which other public political actors play a role, but with insuﬃcient mutual organization for the decision-making to be coordinated in a democratic sense; I also discussed examples of this kind in the previous chapter. In the case of such a unilateral decision-making process, the agent with sole decision-making power has full responsibility for the decisions. Accordingly, this unilateral public political agent must represent all individuals and social interests that are subject to the public power it wields. In a process of this kind, then, the form of democratic community that is entitled to representation by the agent is a jurisdictional community. The second kind of decision-making structure within which an NGO can exercise public power can be characterized as multilateral, insofar as each of the public political agents exercising public power over a given population makes decisions in collaboration with others (rather than independently), within a decisionmaking system that distributes power among them according to some (aggregative or deliberative) democratic decision-procedure. NGOs engage in multilateral decision-making when they make decisions in collaboration with other actors, through processes that are coordinated suﬃciently to incorporate some kind of democratic decision-procedure (or ‘decision-rule’) that allocates particular responsibilities and entitlements to each participant and thus ensures fair representation for all designated constituents within the overall representative system. Within international politics there are relatively few examples of decisionmaking processes that allocate power among actors in accordance with any sophisticated democratic decision-rules, but there are many examples of such processes that employ various voting systems and deliberative structures with some democratic credentials. Some of the clearest examples include the norm-building public power exercised by NGOs within the structured decision-making processes of the UN system, and some similarly structured decision-making processes outside the interstate system—especially some of the collaborative regulatory initiatives among NGOs, states, and corporations that I discussed in the previous chapter. Conceivably, we could also include here those forms of NGO public power that are based on the imposition of material constraints in humanitarian and development contexts of the kind discussed in the previous chapter, provided there was coordination among the various operational actors that was both substantial and democratically structured. 16 In the case of such a multilateral decision-making process, the collaborative participation of multiple actors makes it possible to distribute responsibility for the representation of the various individuals and social interests within the jurisdiction among the multiple decision-making agents. Whereas in a unilateral decision-making process the agent of public power must take democratic responsibility for (and ‘represent’) all individuals and interests within the 16 For discussion of the operational cooperation between the UN and NGOs, see Secretary-General, Arrangements and Practices for the Interaction of Non-Governmental Organizations in All Activities of the United Nations System, 10–17.

100

From Nation-States to ‘Stakeholder’ Communities

jurisdictional community, this need not be so within a multilateral decisionmaking process. Although a multilateral process could, in theory, require each decision-maker to represent all individuals and interests within the jurisdictional community, most multilateral decision-making processes in fact distribute responsibility among participants, such that each decision-making agent is allocated sole or primary responsibility for representing a diﬀerent group of individuals and social interests. 17 These groups—each with its own representative agent advancing its interests in the decision-making process—are then referred to as multiple constituencies. Within such a multilateral process, a designated decision-procedure or decisionrule is responsible for ensuring that the interests of each constituency receive fair representation overall in determining the ﬁnal decisions. Having thus understood the basis of the distinction between these two diﬀerent kinds of democratic community, and recognized the distinct decision-making processes in which each kind of community possesses entitlements to representation, we can now see that the political conditions under which NGOs can claim to represent each kind of community (or conversely, the conditions under which each has a legitimate claim to representation by NGOs) are quite diﬀerent. First, jurisdictional communities possess entitlements to representation by NGOs under conditions where NGOs engage in unilateral (independent) public political decision-making with impact on a given population. In such cases, NGOs must take responsibility for representation of all of the social interests subject to their public power. Second, multiple constituencies possess entitlements to representation by NGOs under conditions where the NGOs engage in multilateral (collaborative) public political decision-making with impact on a given population. In such cases, the NGOs can take responsibility for the representation of only some rather than all of the social interests subject to their public power; that is, they can represent only certain interest-based constituencies, rather than all those subject to the power exercised by the multilateral process in which they participate. In this way, we can see that the nature of the decision-making context in which an NGO wields its public power determines the range of interests that it may be obliged to represent in any given instance. When an NGO exercises public power in certain types of multilateral public decision-making contexts, it may be required only to represent certain sub-jurisdictional constituencies in which 17 Depending on the decision-rule being employed, a representative can have responsibility for its own constituency exclusively (with no responsibility for representing the wider jurisdictional community); or it can have primary responsibility for representing its constituency, and secondary responsibility for representing the wider jurisdictional community. In the case of an aggregative decision-procedure, representatives of constituencies need not take into account the interests of the overall jurisdiction, since the overall interest is supposed to be satisﬁed through the aggregation of the partisan interests of all individuals by means of a vote. In a deliberative procedure with a decisionrule of consensus, representatives of constituencies present arguments primarily in pursuit of their own constituency’s interests, but in the ﬁnal decisive phase of the decision-making process they must also take into account the interests of the overall jurisdiction. Thus, in deliberative systems with consensus decision-rules, agents represent their constituencies primarily and the wider jurisdictional community secondarily.

From Nation-States to ‘Stakeholder’ Communities

101

outcome-interests are shared by all. In such circumstances, the problem of how to reconcile or adjudicate among the conﬂicting interests within its stakeholder community does not arise for the NGO, because the interests of its constituents are uniform rather than conﬂicting. It is only in unilateral public decision-making contexts, where an NGO is responsible for representing all jurisdictional stakeholders, that it is necessary to ﬁnd means of identifying some uniﬁed interest within the stakeholder community to be represented by the NGO. The problems surrounding this need are certainly not unique to global democracy; it is, rather, a central democratic problem widely discussed in the contemporary literature on democratic ‘social choice’ and in some other literature on the problem of identifying the democratic ‘common will’. Review or resolution of such debates is well beyond the scope of the present discussion; here, it is suﬃcient to point to the need for NGOs that have responsibilities for representing entire stakeholder jurisdictions to establish some type of democratic decision-procedure, taking fair account of the various conﬂicting interests within them. In some instances, it may be possible and appropriate for an NGO to establish a voting system of some kind, whereby all stakeholders can vote on crucial NGO decisions. In other instances, voting may be inappropriate or infeasible, and it may be preferable to establish some kind of deliberative decision-making process that accords representation to each conﬂicting interest. Ideally, this would entail a conversion of the unilateral decision-making process (conducted by the NGO independently) into a multilateral decision-making process (of the kind I discuss in Chapters 5 and 6), in which multiple actors, representing these various conﬂicting interests, could participate as decision-makers. However, the ‘realist’ orientation of my analysis here necessitates recognition that ideal solutions of this kind (which eﬀectively entail adjusting the boundaries of public power, insofar as they entail substantive transfer of power from a single NGO to multiple other decision-makers) will not always be possible. In such circumstances, an NGO may need instead to employ some kind of adjudicative process, whereby some impartial standards are employed to adjudicate among the conﬂicting stakeholder interests. Such a solution would not be ideal from a liberal representative point of view, but so long as the adjudicative standards were based as closely as possible on the shared values of the wider jurisdictional community in question, rather than derived from the private moral commitments or strategic interests of the NGO, it could nonetheless confer some degree of representative legitimacy.

4.4. CONCLUSIONS In this chapter, I have elaborated some theoretical and practical implications of my earlier claim that the legitimate agents of democratic control in global politics are what I have called ‘stakeholder’ communities. These stakeholder communities

102

From Nation-States to ‘Stakeholder’ Communities

should be delineated in accordance with the scope of liberal-democratic public power, such that all individuals whose autonomy is problematically constrained by the exercise of public power by an NGO (or other actor) are designated as members of its stakeholder community. My arguments here do not imply that more solidaristic communities—with thicker shared values, loyalties, and interests of the kind commonly associated with traditional ‘closed’ societal models of democracy—do not or should not provide the basis for democratic communities within the global domain. Undoubtedly, it is true that solidaristic communities of this kind exist at national and extranational levels—and to some smaller extent also at the global level—and that such communities have good reasons for seeking to establish (and subsequently democratize) collective agencies of public power through which to pursue their collective purposes. But this is not in any way incompatible with the stakeholder model of democracy that I have elaborated here. The stakeholder conception of democratic community does not require thick cultural solidarities to exist as the basis for a democratic community, but nor are such solidarities within democratic communities incompatible with the stakeholder conception. The stakeholder model requires minimal value consensus on the core democratic values of individual autonomy and (egalitarian) responsibility, but beyond this all that matters for the delineation of a democratic community is the manner in which individuals are aﬀected by the exercise of public power.

Part II Democratic Representation in the New Global Polity

This page intentionally left blank

5 Global Social Choice Beyond Nation-State Representation 5.1. INTRODUCTION In my discussion so far, I have developed an account of how we can identify both the agents of ‘public power’ that should be subject to democratic control in global politics and the ‘stakeholder communities’ that are entitled to wield this control through representative decision-making institutions. In the remaining chapters of the book, I consider how we could construct such legitimate representative institutions at the global level. This account of representative global decisionmaking must be composed of two parts. First, it requires an account of legitimate social choice in a representative global democracy—that is, an account of the process through which stakeholder communities can identify the speciﬁc policies, or more generally the substance of the ‘interests’, that their representatives can legitimately advance through public policymaking. Second, it requires an account of legitimate representative agency—that is, an account of how to identify the agents that are entitled to wield public power as legitimate ‘representatives’ of stakeholder communities (in pursuit of the policies or general interests identiﬁed through social choice mechanisms). In this chapter and the next I focus on the ﬁrst of these elements of a representative system: mechanisms for making legitimate social choice in global politics. Any account of democratic social choice must tell us how the multiple (possibly conﬂicting) interests within a democratic community can be fairly combined to reach a collective public decision. Our answer to this question will depend in part on our views about the nature of the social interests that should be accorded representation. Our answer will further depend on some account of the legitimate method for combining conﬂicting individual interests into a single social choice. At a theoretical level, this must involve some normative criterion for determining how the various interests, each possessing an entitlement to representation, should be weighted and incorporated within the ﬁnal decision. Since any democratic notion of legitimacy must incorporate an egalitarianism of some kind or another, this criterion must specify a conception of equality that can justify the decisionprocedure that is to be adopted. At an institutional level, we must then determine whether democratic equality requires that individual interests should be combined through aggregation (leading to some form of majority decision), or

106

Global Social Choice Beyond Nation-State Representation

alternatively reconciled through deliberation (leading to some form of consensus decision). An account of how multiple individual interests should be combined in a collective decision is suﬃcient to provide a complete account of ‘direct’ or ‘participatory’ social choice, but a framework for social choice in a representative democratic system involves an additional layer of complexity. In a representative democratic system, public decision-making is not undertaken directly by all members of the population, but rather by a smaller group of individuals who have been designated as the ‘representatives’ of the whole society. The process of translating individual interests into a collective decision is mediated by assigning representatives to various ‘constituencies’ within the broader society. Whereas an institutional framework for social choice in a direct democracy involves only a single procedure—a mechanism for combining individual interests into a single social decision—an institutional framework for social choice in a representative democracy involves two sets of procedures. First, it involves some process for dividing the overall society into constituencies, each of which is to be accorded a representative agent. Second, it involves a process for combining the interests advanced by these representatives into a single social decision. A defence of any speciﬁc model of representative social choice must therefore involve not only normative arguments for the underlying account of social choice at the individual level of analysis but also further instrumental arguments about how the proposed division of the population into multiple constituencies can eﬀectively enact the underlying ideal of social choice in practice. The most conventional account of how representative social choice should be achieved in global politics involves a framework of nation-state representation. On this account, each individual is to be designated membership of a single constituency (her territorial ‘nation’ or ‘state’), and represented solely by the representative agents of this constituency (generally her government, or her government’s diplomatic delegates) within global decision-making processes. Justiﬁcations for this conventional account are rarely advanced and defended explicitly; nonetheless, certain justiﬁcations are sometimes hinted at, or partially elaborated, and from these we can extrapolate some plausible rationales for this conventional view. Among these justiﬁcations, we can usefully distinguish between those that claim global social choice should be achieved through representation of individuals by ‘nations’ (which are principally cultural entities) and those that claim global social choice should be achieved through representation of individuals by ‘states’ (which are principally territorial entities). Each of these ‘nationalist’ and ‘statist’ models is a theoretical ‘ideal-type’, and many advocates of nation-state-based global representation may endorse some combination of the two. In this chapter, however, I treat these two ideal-typical models as distinct, in order to highlight two quite diﬀerent rationales that can be developed in support of the conventional ‘nation-state’ model of global representation. In what follows, I elaborate these two potential rationales for nation-state representation, by critically examining both the general (individual-level) accounts of social choice that underlie each model and the instrumental means by which their

Global Social Choice Beyond Nation-State Representation

107

division of global jurisdictions into national and state constituencies is intended, in practice, to realize the normative ambitions of these underlying ideals of interest representation. I further develop a critique of each of these potential rationales, thus preparing the theoretical ground for my defence in the following chapter of an alternative—or at least supplementary—model of ‘multi-stakeholder’ representation. In Section 5.1, I examine some normative and practical justiﬁcations for this model that are based upon the notion of a nation, and in Section 5.2, I examine some normative and practical justiﬁcations for this model that are based upon the notion of a state. Overall, my critique here of national and state representation should be read as an explication of its limits, rather than as a wholesale rejection of its legitimate role in global politics. One further point of clariﬁcation is worth noting here before I proceed with my substantive arguments. It is important for the reader to bear in mind that although I refer to all of the processes of social choice I subsequently discuss as ‘global’ decision-making processes, not all of them are global (in the sense of universal) in jurisdictional scope. Although some decision-making processes (such as those within the UN system) could be viewed as ‘global’ in scope, others are conducted at regional or local levels, or within functional rather than territorial jurisdictions, and so have a narrower jurisdictional scope. It is therefore only for purposes of terminological simplicity that I refer to decision-making processes at all these various jurisdictional levels within the global polity by the same general label of ‘global’ decision-making processes. The examples of decision-making processes that I draw upon for illustrative purposes in developing my argument do generally have a global jurisdictional scope, but it should nonetheless be kept in mind that the same arguments about legitimate social choice apply, to greater or lesser degrees, at various jurisdictional levels within the global polity.

5.2. A CRITIQUE OF SOCIAL CHOICE THROUGH ‘NATIONAL’ REPRESENTATION

5.2.1. The Nationalist Model of Global Representation Let us begin, then, by examining the theoretical conception of social choice that underpins the ﬁrst potential route to justifying nation-state representation, which is based on the idea that ‘nations’ are the legitimate constituencies for representative social choice. The concept of ‘nationhood’ is widely discussed and contested, but most conceptions tend to cluster around a common set of characteristics. These characteristics are helpfully delineated by David Miller in his extensive discussion of the role of nationality in democratic politics. 1 On Miller’s account, nations are cultural communities with a distinctive character: they are constituted by shared beliefs and mutual commitment, extended in history, active in character, 1

David Miller, On Nationality, Oxford Political Theory (Oxford: Clarendon Press, 1995).

108

Global Social Choice Beyond Nation-State Representation

connected to a particular territory, and marked oﬀ from other communities by a distinct public culture. 2 The national model for delineating global constituencies prescribes that whenever public power is exercised at jurisdictional levels beyond the national community itself—whether at global, regional, or transnational levels—nations should be accorded representation within these institutions as distinct democratic constituencies. Where national communities correspond with state boundaries, this would generally involve representation of national constituencies by the governments of their nation-states; in the case of stateless national communities of various kinds, other (non-state) representative agents would need to be identiﬁed. There is some degree of ambiguity in the nationalist model of global constituencies with respect to how representative agency should be conferred—that is, how an agent (such as a government agency) can acquire the status of being ‘representative’ of a national constituency. On this model, it is clear that the interests with entitlement to representation are ‘national interests’, and that representatives acquire their status by virtue of their capacity to discern and act on behalf of these. What is ambiguous, however, is whether or not the ‘national interest’ is something greater than—or at least separable from—the aggregate interests of its individual members. Some advocates of national representation (i.e. those with liberal individualist normative commitments) may want to maintain that national interests are derived directly from the interests of their individual members, and that the normative basis for national representation resides in this fact. Such liberal nationalists may want to insist that a legitimate national representative accordingly must possess some kind of institutionally determined mandate (such as an electoral mandate) derived through the participation of each individual within the nation. Not all advocates of national constituencies would agree, however, that national interests are reducible in this way to the interests of individual members; some would instead advance the proposition that national interests possess some ‘objective’ status, independent of any aggregation of individual interests. If the substance of the national interest is distinguished from the interests of individual members in this way, then it is not so clear that legitimate national representatives must possess any institutionally determined mandate derived through the participation of each individual within the nation. Rather, all that is clear is that representatives must, through membership of the nation, either embody or be competent to discern the substance of this organic national interest. What the necessary qualiﬁcations may be for discerning and representing national interests remains largely an open question. It is perhaps for this reason that advocates of national constituencies hold conﬂicting and sometimes ambiguous positions on the question of the representative status of unelected national governments. Moreover, I have said that a complete model of multi-constituency representation must incorporate not only an account of the interests entitled to representation and the legitimate basis for representative agency but also an account 2

Miller, On Nationality, 17–27.

Global Social Choice Beyond Nation-State Representation

109

of the decision-procedures through which representatives of these multiple constituencies should reach collective decisions on behalf of the overall jurisdictional community. Advocates of national representation in multilateral global decisionmaking generally agree that all national constituencies should be accorded equal representation in the process of reaching collective decisions. Sometimes this support for equal status for national representatives is combined with support for an aggregative decision-rule, such that decisions are made on the basis of ‘one nation one vote’. However, it is diﬃcult to see how such a decision-rule could be justiﬁed without invoking an organicist ontology of nationhood of some kind, since nations, rather than individuals, would on this model be the bearers of equal status and representative entitlements. 3 Since such a view would be prima facie incompatible with the ontological and normative individualism central to democratic thought, it is diﬃcult to see how a democratic rationale for such a model could be constructed. Instead, the particular justiﬁcation for national constituencies that I discuss here—which I think is the most plausible and persuasive that can be constructed—requires that the procedure by which national representatives reach collective decisions be deliberative rather than aggregative (for reasons outlined below). Accordingly, I suggest that the national model of constituencies should be taken to include a decision-procedure of collective deliberation leading to some kind of broad consensus among national representatives. Although many individuals and groups are politically committed to a representative system of this kind, in which a central role is accorded to national constituencies within representative global institutions, it is no easy task to formulate a theoretical framework to support the democratic justiﬁcation of this representative structure. Few rigorous attempts have been made to develop a theoretical justiﬁcation for national constituencies in global institutions, which is hardly surprising since the theoretical diﬃculties of reconciling the dual normative goals of nationalism and global democracy are well established. 4 In what follows, then, I try to present as plausible as possible an account of how such a democratic justiﬁcation could proceed.

5.2.2. A ‘Burkean’ Justiﬁcation for National Representation 5.2.2.1. The ‘Burkean’ Model of Interest Representation I propose that the most plausible way we can make democratic sense of this ‘nationalist’ account of representative social choice in global politics is by invoking 3 Peter Spiro similarly notes the organic ontology of nationhood underlying rationales for the principle of ‘one nation one vote’. See Peter J. Spiro, ‘NGOs and Participatory Legitimacy in the UN System’, paper presented at the ACUNS Annual Conference, New York, 1995, 3–4. 4 See David Miller, ‘Bounded Citizenship’, in Kimberly Hutchings and Roland Dannreuther (eds.), Cosmopolitan Citizenship (Basingstoke: Macmillan, 1999), for a discussion of the incompatibility of democratic nationalism and global democracy.

110

Global Social Choice Beyond Nation-State Representation

what Hanna Pitkin describes as a ‘burkean’ conception of interest representation. 5 In her classic study of the concept of representation, Pitkin delineates an ‘ideal-typical’ ‘burkean’ conception of interest representation and contrasts it with an idealized ‘liberal’ conception, which I discuss in Section 5.2.2.2 of this chapter. 6 On this ‘burkean’ account of interest representation, interests are characterized as unattached, objective, stable, and reconcilable through rational deliberation. On this view, the interests that should be the subject of political representation are ‘unattached’ to any particular group of individuals; they are abstractions such as ‘justice’ or ‘world peace’, which can be represented independently of the wishes or preferences of particular individuals or groups. Although the speciﬁc interests that warrant representation (on a ‘burkean’ account) are associated with particular groups whose livelihood they characterize, Pitkin argues that Burke presented them as having an objective reality independent of any individuals they might aﬀect. Hence, for Burke a group ‘partakes of ’ or ‘participates in’ an interest, it does not ‘have’ the interest. 7 As this conception of interest is abstracted from the wishes or opinion of any particular individuals or groups, interest is viewed as having an objective reality which is, as Pitkin puts it, ‘very much as we today see scientiﬁc fact: it is completely independent of wishes or opinion, of whether we like it or not; it is just so’. 8 One consequence of this objectivity of interest is that interests are viewed as stable and lasting, unlike opinion, which ‘tends to be hasty, passionate, prejudiced, [and] subject to violent but short-lived ﬂuctuations’. 9 The objectivity of interest on the burkean model has further consequences for the question of representative agency—that is, of how an agent can acquire the status of legitimate ‘representative’ of particular interests. Since burkean interests are objective, it follows that they can easily be identiﬁed by the reasoned thinking of honest and intelligent representatives. On the burkean model, then, we need not think of representative agency as deriving from any formal institutional connections between individuals and representatives. Rather, 5 The extent to which this ‘burkean’ model provides a true reading of Burke’s political theory, and the extent to which it imposes upon his theory an artiﬁcial coherence, is a disputed question. See James Conniﬀ, The Useful Cobbler: Edmund Burke and the Politics of Progress (Albany: State University of New York Press, 1994) for a critical discussion of Pitkin’s reading of Burke. For the present purposes, however, the basis of this model in Burke’s political writing is of little signiﬁcance; what matters here is the substance and validity of the model itself, as explicated by Pitkin. 6 Hanna Fenichel Pitkin, The Concept of Representation, 1st Paperback edn. (Berkeley: University of California Press, 1972). Although these two conceptions of interest representation are very diﬀerent, they are not so much clear alternatives as ‘ideal types’, situated at the polar ‘limits’ of a spectrum of conceptions of interest representation. Real systems of political representation usually draw elements from each and are situated somewhere between them on the spectrum. However, distinguishing these ideal types helps to illuminate the range of conceptual resources upon which democrats can draw in establishing systems of representation, and the institutional implications of choices among these conceptual possibilities. 7 Ibid. 174. 8 Ibid. 180. 9 Ibid. 181.

Global Social Choice Beyond Nation-State Representation

111

agents legitimately represent an interest when they accurately discern its objective substance. 10 This implies, then, an acceptance of a ‘trustee’ rather than a ‘mandate’ conception of representative agency. 11 Another important feature of such burkean interests is that they are not only objective but also additive, in the sense that all the various interests within a political community can be reconciled to comprise the overall interest of the community. While Burke was not always consistent or clear on this question, Pitkin argues that fundamentally he conceived of each ‘interest’ ‘as something objective and discoverable that can be ﬁtted into an orderly schema, the interest of the whole’. 12 This conception of interest is closely linked to the ideal of deliberative decisionmaking, which is also entailed in the burkean representative model. A legitimate decision-procedure, according to this model, is one in which the interest of the whole community is discerned by means of a reasoned dialogue in which the various partial interests and perspectives within the community are represented and taken into account. Hence, on the burkean account of interest representation, ‘[v]oting . . . is of no importance; what is required is that all the facts and arguments be accurately and wisely set forth.’ 13 It is not necessary therefore for all interests to have an equal number of representatives within any decision-making body, because if an interest has ‘even a single competent member in Parliament, it will be looked after, because it is not . . . [the representative’s] vote but his arguments that matter’. 14 The conception of individual equality that could be taken to underpin this deliberative decision-procedure warrants some closer consideration, since it is not a prominent feature of the burkean model as it stands. Some aspects of Burke’s broader political theory are in fact strongly hostile to certain forms of democracy and political equality, so it is not entirely clear that Burke himself would endorse an egalitarian reading of his representative ideal. 15 In general, Burke rejected democracy conceived as rule by popular will in pursuit of subjective interests, and further rejected the simple participatory conception of political equality associated with this democratic idea. Instead, he embraced rule by reason in pursuit of objective interests, by those individuals with the virtue and experience to discern 10

As Pitkin notes, Burke recognizes that representation of these ‘unattached’ interests must have some kind of ‘basis in the actual’, insofar as there must be ‘a communion of interest and sympathy in feelings and desires between those who act in the name of any description of people and the people in whose name they act, though the trustees are not actually chosen by them’. See Edmund Burke, The Works and Correspondence of the Right Honourable Edmund Burke (London: F. & J. Rivington, 1852), vol. IV. What is crucial, though, is that this does not necessitate any formal institutional connections between individuals and representatives. 11 See Pitkin, The Concept of Representation, ch. 7, for a discussion of the diﬀerence between ‘trustee’ (or ‘independence’) and ‘mandate’ views of representative agency. 12 Ibid. 187. 13 Ibid. 188. 14 Ibid. 15 See Edmund Burke and L. G. Mitchell, Reﬂections on the Revolution in France (Oxford: Oxford University Press, 1999) for some of his arguments against the ideal of political equality.

112

Global Social Choice Beyond Nation-State Representation

these. 16 Nonetheless, we can identify a certain limited form of egalitarianism that is at least consistent with the burkean representative ideal as I have outlined it; whether it is one that Burke himself would have endorsed is a question that need not detain us here. The burkean representative model, then, is compatible with two forms of equality, both of which are associated with the deliberative decision-process entailed in the model, although distinct from other forms of equality associated with more participatory deliberative procedures. In models of deliberative democracy in which all individuals participate in the deliberative decision-making process, citizens enjoy equality in the form of equal opportunity to participate in and aﬀect the outcomes of decisions. 17 In the burkean model, however, deliberation is undertaken not by all individuals but only by their representatives, so it cannot be a participatory form of equality that underpins the burkean model. Instead, we can identify two other forms of equality that are possessed by individuals within a burkean representative system. First, the burkean model is underpinned by equality of individuals insofar as it is assumed that the objectively discernable ‘common good’ of the community is something in which all citizens share equally. Thus, groups of individuals without their own representative agents nonetheless ‘have an equal representation, because you have men equally interested in the prosperity of the whole, who are involved in the general interest and the general sympathy’. 18 In this sense, individuals are equal insofar as they all partake equally in the common good that is advanced by representation. Second, the burkean model is underpinned by equality of individuals insofar as the various diﬀerent social interests within the community, in which diﬀerent groups of individuals partake, are all given equal deliberative consideration in the process of reaching a ﬁnal decision. In such a process, individuals are treated as equal in the sense that the arguments for advancing the various interests in which they partake are accorded equal weight in deliberations alongside arguments for advancing other social interests, and decisions are made on the purportedly impartial basis of reasoned argument. Equality in these two senses could be linked to either of two ‘deeper’ conceptions of individual equality distinguished by Charles Beitz in his work on political equality. 19 First, they could be linked to what Beitz calls a ‘best outcome’ 16 See Michael Freeman, Edmund Burke and the Critique of Political Radicalism (Oxford: Basil Blackwell, 1980), ch. 6, for a discussion of the democratic and anti-democratic elements in Burke’s thought. 17 See Jack Knight and James Johnson, ‘What Sort of Political Equality Does Deliberative Democracy Require?’ in James Bohman and William Rehg (eds.), Deliberative Democracy: Essays on Reason and Politics (Cambridge, MA: London: MIT, 1997), for a discussion of the value of equality within deliberative democratic models. 18 Edmund Burke and Ross John Swartz Hoﬀman, Burke’s Politics: Selected Writings and Speeches on Reform, Revolution and War, 1st edn. (New York: A. A. Knopf, 1949), quote from ‘Speech on Reform of Representation in the House of Commons, June, 1784’. 19 Charles R. Beitz, Political Equality: An Essay in Democratic Theory (Princeton: Princeton University Press, 1989).

Global Social Choice Beyond Nation-State Representation

113

view of equality, according to which the pursuit of a socially optimal outcome treats the objective interests of all individuals equally. Second, they could be linked to what Beitz calls a ‘complex proceduralist’ view of equality, according to which individuals are treated as equal persons, in the sense that decisions are made through procedures that each individual has equally good reasons to accept. 20

5.2.2.2. Burkean Interest Representation and National Representation To what extent, then, can this burkean conception of interest representation provide a theoretical basis for the nationalist model of global social choice? How it can do so is not immediately obvious, since it is true that in one sense the burkean model of constituencies requires representatives to represent the ‘common interest’ of the whole polity, rather than particular sectional interests within it. Given that this is so, the question arises of how the burkean account can provide a basis for the representation of separate national constituencies within the global polity. A key factor here is the burkean idea of ‘unattached’ interests, as outlined above, whereby the ‘common interest’ of a society does not exist as a socially undiﬀerentiated abstraction; rather, it is constituted through a cohesive aggregation of the various partial ‘unattached’ interests within it. Accordingly, so long as the ultimate goal of the institutional system of representation is the pursuit of the ‘common good’, the burkean account can at least accommodate some independent representation for the various ‘unattached’ sectional interests in the society. Given the strong role of the deliberative democratic ideal within the burkean account of interest representation, it could be argued—even more strongly— that the burkean ideal would require independent representation for each of the constituent social elements of a society’s ‘common interest’, in order to facilitate the kind of representative social dialogue through which this interest can be identiﬁed. 21 Consequently, although the ultimate subject of representation on the burkean model is the common interest of the democratic society, the immediate (direct) subjects of representation are—or can be—the various ‘unattached’ sectional interests within the community. 22 20 The burkean model would not satisfy the speciﬁc account of complex proceduralism defended by Beitz. However, an account of complex proceduralism that speciﬁed diﬀerent ‘regulative interests of citizenship’ from those speciﬁed by Beitz, in identifying procedures that all individuals could reasonably be expected to accept, could be compatible with the burkean model. 21 For a version of this view, see deliberative arguments for the representation of ‘social groups’ as distinct constituencies within democratic polities, developed by Iris Marion Young, Inclusion and Democracy (Oxford: Oxford University Press, 2000). See also Anne Phillips, The Politics of Presence, Paperback edn. 1998 (Oxford: Clarendon Press, 1995). 22 Since these ‘unattached’ sectional interests are abstracted from the individuals who ‘partake’ in them, these interests do not constitute ‘constituencies’ as such (if constituencies are deﬁned as groups of individuals). However, they can be viewed as constituencies if these are deﬁned, more broadly, as the legitimate subjects of representation within a democratic system.

114

Global Social Choice Beyond Nation-State Representation

A burkean framework could provide a basis for national representation, therefore, if ‘national interests’ were to be designated as the principal ‘unattached’ interests that together comprised the common good of global society at large. Such an approach to theoretically justifying national constituencies in global politics might seem somewhat surprising, since national interests are traditionally understood by International Relations theorists as paradigmatically conﬂicting, rather than as commensurable parts of a uniﬁed whole. Indeed, few theorists have attempted to develop this—or any other—theoretical defence of the role of national constituencies within representative global institutions, despite the fact that so many individuals and communities are politically committed to the central role of national constituencies within current structures of global representation. Nonetheless, such a burkean approach has some degree of plausibility, as it could be seen as a democratic equivalent of a more common kind of normative argument for placing nations at the centre of global public decision-making institutions. More speciﬁcally, a burkean structure of national constituencies appears to be broadly consistent with the kind of ‘developmental communitarianism’ discussed by Peter Sutch and others. 23 On this ‘developmental communitarian’ view, legitimate principles of international law and policy must be grounded in the consent of the various cultural ‘peoples’ or ‘nations’ within global society, and must be derived through some kind of dialogue among them. Although the theoretical details of the arguments invoked by various ‘developmental communitarian’ writers diﬀer signiﬁcantly from each other and from the burkean conception of global democracy, one central insight is shared by all: the normative goals of global society, whether conceived as principles of ‘justice’ or as a democratic ‘common good’, must be constituted through some kind of discursive accommodation of the primary ‘values’ or ‘interests’ of each national community. In this way, then, we can see that the burkean conception of interest representation can provide a framework for justifying the role of national constituencies within global institutions for representative social choice that is, at the very least, theoretically coherent.

5.2.3. A Democratic Critique of National Representation A critical examination of this justiﬁcatory framework for national representation must consider two questions. First, if we assume that the burkean theoretical model used to justify national representation is sound at a theoretical level, is national representation a practicable means of enacting the normative ideals of this model? Second, does the burkean model in fact provide a sound theoretical framework for conceptualizing the qualities of a democratic representative system? 23 See Peter Sutch, Ethics, Justice, and International Relations: Constructing an International Community (London: Routledge, 2001) for a discussion of this ‘developmental communitarianism’.

Global Social Choice Beyond Nation-State Representation

115

5.2.3.1. A Critique of the Practicability of the Nationalist Model There are some good empirical grounds for thinking that representation of nations within deliberative global decision-making processes can enact some key ideals of the burkean representative model. If we were to accept the proposition that a global ‘common good’ is composed of objectively identiﬁable ‘unattached’ social interests within the global community, national interests would seem prominent candidates for recognition as the relevant social interests. Nations are still very signiﬁcant political entities in contemporary global politics, and it would be diﬃcult to formulate an account of the key social interests within the global community that did not accord a prominent place to national interests. Nonetheless, there are two key ways in which the nationalist model fails to satisfy the normative requirements of the burkean model in practice. The ﬁrst is that, despite the importance of national interests, it is implausible to claim that national interests could be objectively identiﬁed as the sole ‘unattached’ social interests with signiﬁcance for individuals. The designation of nationality as the only, or even as the principal, social interest in which each individual ‘partakes’ is becoming increasingly obsolete and implausible in our globalizing world. In contemporary global society, individuals’ core social experiences and identities, and the interests that arise from these experiences and identities, are no longer generated exclusively or even primarily within national boundaries. Rather, increasingly transnational social, economic, cultural, and political relationships are generating transnational groups with shared experiences, identities, and interests. Insofar as the social interests of central importance to individuals are associated with shared material circumstances, we can see how transnational social interests have emerged as a result of various technological, economic, demographic, and other social dimensions of ‘globalization’. 24 Insofar as the social interests of central importance to individuals are associated with shared collective identities, we can see how transnational social interests have emerged as a result of significant transformations in patterns of global identiﬁcation. Ronnie Lipschutz, for example, argues that the current global system, based on people’s identiﬁcation with the nation-state as the primary social grouping, is being challenged and eroded through the emergence of ‘new forms of collective identity’, including feminist, gay and lesbian, and religious fundamentalist identities, as well as identities arising from issue-based movements such as human rights and environmentalism. 25 William Connolly presents a very similar picture of the character of emerging transnational identities, claiming that ‘cross-national, non-state democratic movements’ such as gay and lesbian, nuclear disarmament, and human rights movements ‘challenge a [nation] state’s monopoly over the allegiances, 24 See David Held, Democracy and the Global Order: From the Modern State to Cosmopolitan Governance (Cambridge: Polity, 1995) for a discussion of these transformations and their implications for the emergence of transnational interest groups—or ‘communities of fate’, to use Held’s term. 25 Ronnie D. Lipschutz, ‘Reconstructing World-Politics—the Emergence of Global Civil Society’, Millennium, 21/3 (1992), 415.

116

Global Social Choice Beyond Nation-State Representation

identiﬁcations and energies of its members’. 26 Accordingly, a system that accorded representation solely to national interests to the exclusion of these other signiﬁcant non-national social interests would fall short in practice of the burkean ideal of inclusive and equal representation for all key social interests. The second way in which national representation fails in practice to satisfy the normative requirements of the burkean model is associated with the requirement that national representatives reach decisions through consensus, on the basis of some objective ‘common good’. Although there may be instances where a common good can be identiﬁed by national representatives and decision reached through deliberative consensus, on many issues national interests are conﬂicting rather than compatible. On such issues (and there are many), the normative ideal that decisions should be reached through deliberative reconciliation of national interests is unattainable in practice. This diﬃculty in reconciling various national interests through deliberation could be overcome, of course, if contentious decisions were resolved through an electoral process in which each national representative were accorded a vote. Many existing processes of international decision-making within international organizations employ such a strategy for reaching decisions among national representatives, and this does provide a way out of any practical problems of deliberative deadlock. The deliberative element of the burkean model cannot simply be abandoned in this way, however, since the process of decision-making through deliberative consensus is crucial to the legitimacy of the model. In the ﬁrst instance, deliberative consensus is essential for the legitimacy of the burkean model because such consensus is supposed to establish the objective validity of the ‘common good’ advanced through the collective decision. If no consensus can be reached, and a decision is instead taken by a vote, then it is not clear that the decision would reﬂect any common good as distinct from the factional interests of the majority of nations. Second, decision-making through deliberative consensus is essential for the legitimacy of the model because this ensures that the various social interests in which diﬀerent individuals partake are considered and evaluated equally in the decision-making process. If decisions were taken instead through a vote among national representatives, no equal consideration would be ensured for the various interests in which diﬀerent individuals partake. Overall, then, we can see that the normative ideal of burkean representation cannot satisfactorily be achieved in practice through a system of national representation. The burkean model depends for its legitimacy on the process of making decisions through consensus on the common good, so deliberation cannot be replaced with a voting-based decision-procedure. Retaining this demand for decision through deliberative consensus, however, appears highly unworkable given the scope and depth of the conﬂicts among national interests that are still entrenched within the wider global community. 26

William E. Connolly, ‘Democracy and Territoriality’, Millennium, 20/3 (1991), 479.

Global Social Choice Beyond Nation-State Representation

117

5.3. A THEORETICAL CRITIQUE OF THE JUSTIFICATORY BURKEAN FRAMEWORK Not only does national representation fail to satisfy the normative ideals of the burkean model of interest representation at a practical level; the burkean model is itself ﬂawed at a theoretical level, insofar as its justiﬁcatory framework does not entail a suﬃciently democratic standard of legitimation. This is not to say that the burkean representative model has no advantages from the point of view of democratic legitimacy. Rather, as I will argue in the following chapter, the process whereby representatives deliberatively consider the merits of various social interests according to objective criteria is democratically beneﬁcial because it can entrench a form of equality that recognizes and takes into account in decisionmaking the diﬀerential intensity of the various interests implicated in a decision. Despite this advantage, however, the burkean model is unsatisfactory overall, because its complete reliance upon the notion that individuals’ interests can be objectively discerned fails to incorporate an adequate role for individual autonomy within the system of representative democratic decision-making. Since the burkean account is not explicitly committed to an objective conception of individual interests, it is helpful to explain brieﬂy how such a conception is nonetheless implicit. I have already explained how the burkean account invokes an objective conception of social interests: on the burkean account, both the ‘common interest’ of the community and the various ‘unattached’ sectional interests that constitute this common interest are conceived as objective. As social interests, these objective burkean interests are not ascribed to individuals as such; however, they are conceived as interests in which individuals ‘partake’. So although this model does not seek to identify which individuals partake in which social interests—and in this sense does not ascribe any particular objective interest to any given individual—it does ascribe objective interests to individuals in the more limited sense of identifying the ‘objective’ constituent social elements of all individuals’ interests. Thus, a burkean defence of national constituencies would necessarily invoke an objective conception of individual interests by designating a given national interest as the primary social interest constituting each individual’s interest, without consulting individuals themselves as to whether they did in fact see their ‘national interest’ as central in this way. As I will discuss in more depth in Section 5.4 and the following chapter, a representative system must rely to some degree on the possibility that representatives can discern such objective interests, since without this possibility they would be unable to exercise any initiative or independent judgement in the decisionmaking process. Instead, they would be bound entirely by explicit instructions from constituents. This would preclude the democratically beneﬁcial possibility of representatives engaging in deliberative forms of decision-making, since there would seem to be little point in a deliberative process in which representatives were prevented from amending their judgements in light of the reasons presented in the course of this process. If we are to accept the role of deliberation in representative

118

Global Social Choice Beyond Nation-State Representation

decision-making, then we must also allow representatives some scope to form independent judgements, informed by such deliberations, about the interests of their constituents. The notion of objective interests—and the associated notion of representation as ‘trusteeship’—must therefore play some role in any satisfactory theoretical model of democratic representation. Despite the need to accord some role to the idea of objective interests, a representative system that relies wholly upon this means of ascribing interests to individuals is problematic, because it denies individuals suﬃcient autonomous agency in the identiﬁcation of their own interests. As I argued in Part 1 of the book, the normative keystone of any liberal conception of democracy is the value of individual autonomy, and this requires that individuals are empowered to formulate and to eﬀectively pursue their authentic personal goals without problematic obstruction by other political agents. As I argued in Chapter 1, representative democratic institutions confer legitimacy upon public power because they operate to ensure that individuals’ autonomous capacities to formulate and pursue goals, which are undermined by the exercise of public power, are reinstated through the redistribution of these capacities to aﬀected individuals within a collective decision-making structure. If we see the legitimating role of representative institutions in this way, it becomes straightforward to recognize how the democratic value of autonomy necessitates each individual being accorded some capacity to identify, subjectively, the interests that are to be advanced by their representatives in the exercise of public power. This is so because, if the capacity to identify reﬂectively and to pursue actively one’s interests is seen as constitutive of individual autonomy, then it follows that any political system that transferred all such capacities to identify and pursue interests from individuals to their ‘representatives’ would be incompatible with the value of individual autonomy in political life. Not all proponents of liberal representative democracy concur with this argument that a legitimate democratic system must incorporate a signiﬁcant institutional role for individuals in the subjective identiﬁcation of the interests that are to be advanced by their public political representatives. Andrew Kuper, for instance, develops a theory of representative global democracy that rejects representation of subjective interests in favour of what he calls ‘substantive’ representation, which entails the representation of objectively discerned interests. 27 As a liberal democrat, Kuper is committed to the protection of individual autonomy. However, he points out that the subjective view of interests has certain normative disadvantages for individual autonomy, since ‘[p]references are adaptive even to the grimmest realities: the most deprived people often acquiesce in their lot, accepting that small mercies are all they deserve, and viewing those who have a direct hand in their deprivation as benevolent father-ﬁgures.’ 28

27 Andrew Kuper, Democracy Beyond Borders: Justice and Representation in Global Institutions (Oxford: Oxford University Press, 2004). 28 Ibid. 81.

Global Social Choice Beyond Nation-State Representation

119

As I noted brieﬂy in Chapter 1, however, liberal-democratic principles—which are underpinned by recognition of the value of individual autonomy—dictate that such problems of irrational and inauthentic preference formation cannot be addressed within the framework of representative institutions. These problems must be addressed through consensus-based institutions that assist deprived people to revise their self-abasing preferences, rather than through conﬂict-based (‘relational’) forms of public power that politically override their preferences with some contested—even if purportedly ‘objective’—account of interests. In other words, objective conceptions of interest can certainly be useful in identifying through consensus-based mechanisms how deprived people can best be assisted to form more autonomous preferences. However, whenever power is exercised in pursuit of contested interests, autonomous individuals must be politically empowered to take or leave such well-intentioned ‘assistance’ in identifying their own best interests. In response to such concerns, those who advocate the representation of objective interests could argue that, even if we think individuals should be empowered to determine subjectively their own ‘ends’ or goals, representatives are often best equipped to determine objectively the best political ‘means’ or strategy for pursuing these ends. 29 It is certainly true that there are strong reasons for individuals to choose to delegate some strategic elements of decision-making to representatives. As Kuper argues, in defence of ‘objective’ identiﬁcation by representatives of the strategic elements of interests, the ‘innate numerical abilities’ of many individuals, for example, may be far too limited for them to grasp and evaluate the strategic merit of certain complex ﬁnancial legislation. And even for individuals who possess the requisite innate abilities, gaining an understanding of certain policy proposals may ‘require so much time and eﬀort that there are strong reasons to have someone else acquire the necessary skills and make the necessary judgements’. 30 However, there is a crucial diﬀerence between claiming that individuals would be prudent to delegate certain strategic decisions to expert representatives, and suggesting (as Kuper appears to do) that expert representatives are entitled to override citizens’ subjective preferences because of their ‘objective’ strategic competence. A democratic commitment to a subjective view of interests allows us to recognize that prudent individuals should take advice from and delegate certain strategic policy decisions to experts, while insisting that they are entitled to reclaim control of these strategic decisions whenever they choose to do so. Ultimately, democratic individuals must be free to refuse the advice of ‘experts’, however imprudent we may think them; if a political conﬂict should arise between the desires of individuals and the judgements of experts, then it is the will of these constituents that must prevail. 29 Various arguments can be advanced for viewing the strategic dimension of interests as objective: it can be argued that individuals lack adequate time to formulate eﬀective strategy, or that they lack adequate information or expertise, or simply that they lack some innate ability. 30 Kuper, Democracy Beyond Borders: Justice and Representation in Global Institutions, 84.

120

Global Social Choice Beyond Nation-State Representation

We can conclude, then, that the democratic value of autonomy requires that individuals must have some ability to determine for themselves what interests their representatives should advance. Although representatives do sometimes need to make decisions based on their own assessments of their constituents’ interests, this cannot be to the total exclusion of a role for individuals to reﬂect upon, identify, and actively pursue their own interests. Rather, the value of autonomy requires that some kind of further mechanisms are enshrined to provide individuals with institutional opportunities to participate at some level in the process of identifying relevant social interests, in order to ensure that constituents concur, at least in broad substance, with the assessments made by their representatives. 31 In practice, this means that representatives cannot be viewed wholly as ‘trustees’, but rather must be constrained to advance some range of subjectively identiﬁed interests, embodied in a democratic mandate. Since the burkean model fails to accord any such recognition, we must conclude that it provides a theoretically unsatisfactory framework for conceptualizing legitimate democratic representation in global politics.

5.4. A CRITIQUE OF GLOBAL SOCIAL CHOICE THROUGH STATE REPRESENTATION

5.4.1. The Statist Model of Global Representation The national framework of representation that I have examined so far is, as I have said, only one way in which the conventional model of global social choice based on nation-state representation can be justiﬁed. The other potential route to justifying this conventional model is by conceptualizing these constituencies as states instead of as nations. Rather than viewing legitimate constituencies within the global polity as cultural entities of a particular (national) kind, this second view of interest representation identiﬁes them as discrete territorially circumscribed groups of individuals, or ‘states’. Having presented some arguments for rejecting the national model, I now examine this second, state-based, route to justifying the nation-state-based model of representative global social choice. Superﬁcially, this statist model has much in common with the national model. The statist model prescribes that whenever the right kind of multilateral public power is exercised at jurisdictional levels beyond the state community itself— whether at global, regional, or transnational levels—territorial states should be accorded representation as distinct democratic constituencies within these institutions. Since national communities are often taken to correspond with state boundaries, the boundaries of constituencies on both national and statist models would 31 I elaborate, in Chapters 7 and 8, an account of what kind of mechanisms these should be in global politics.

Global Social Choice Beyond Nation-State Representation

121

largely correspond: both models would generally require that constituencies be represented by nation-state governments or their diplomatic delegates. There are, however, important diﬀerences between the nationalist and statist models of global social choice. First, whereas the nationalist model is ambiguous about the connection between the ‘national interest’ and the interests of the individuals comprising the national population, the statist model is clear about the individualistic basis of the interests that are represented through state constituencies. Related to this, whereas the national model is ambiguous with respect to how representative agency should be conferred—that is, how an agent can legitimately acquire the status of ‘representative’ of a national constituency— the statist model is clear that legitimate representative agency must rest upon a mandate of some kind from the individuals who comprise the population of the state. An associated diﬀerence between the nationalist and statist models pertains to their underlying conceptions of equality. On the one hand, the nationalist model is underpinned by the premise that equal representation within decision-making processes should be accorded to (unattached) national interests in the global community. In contrast, the statist model is underpinned by the claim that such equal representation should be accorded instead to all individuals within the global community. The division of the global population into territorial state collectives, each with its own representative agent, is an institutional means for ensuring this outcome of equal representation for individuals. As Peter Spiro describes the rationale, this system prevents double-counting at the same time that it at least in theory gives every person some voice. Leaving aside dual nationals and stateless persons, each individual may thus be presumed to have his single representative at the international level. 32

A further consequence of this diﬀerence in the interests that are viewed as the key subjects of representation is that the statist model can potentially provide a rationale for aggregative processes of decision-making among representatives. Whereas the national model prescribes that legitimate decisions can only be reached through a deliberative process in which all national interests are taken equally into account, the statist model has the further potential to justify the use in the decision-making process of an aggregative mechanism that ensures equality for all individuals by according an equal vote to each individual’s state representative. As I explain below, such an aggregative system could only be legitimate under certain empirical conditions, such as states being of approximately equal sizes, or being accorded votes proportionate to their populations. Nonetheless, this decision-procedure is one that can, potentially or theoretically at least, be legitimated within an appropriately structured system of territorial statist constituencies.

32

Spiro, ‘NGOs and Participatory Legitimacy in the UN System’, 3.

122

Global Social Choice Beyond Nation-State Representation 5.5. A ‘LIBERAL INDIVIDUALIST’ JUSTIFICATION FOR STATE REPRESENTATION

5.5.1. The ‘Liberal Individualist’ Model of Interest Representation I propose that the most plausible way in which we can make democratic sense of this ‘statist’ account of legitimate constituency boundaries is by invoking what Hanna Pitkin describes as a ‘liberal’ conception of interest representation. In order to distinguish this model, as articulated by Pitkin, from the ‘liberal pluralist’ model of interest representation that I propose in the following chapter, I refer to Pitkin’s model here as a ‘liberal individualist’ model (rather than referring to it by the more generic label of ‘liberal’). 33 In contrast to the ‘burkean’ account—which characterizes interests as unattached, objective, stable, and reconcilable through rational deliberation—Pitkin’s liberal individualist account characterizes interests as attached to persons, subjective, unstable, and irreconcilably conﬂicting. On the liberal individualist model, representation is conceptualized in terms of the representation of individuals (or groups of individuals), and interests are conceived as ‘attached’ to these individuals and groups. These attached interests are also ‘subjective’ in the sense that the subjective will of the individuals concerned deﬁnes the interest to be represented. Pitkin identiﬁes such a subjective conception of interest in the liberal thought of James Madison, arguing that in Madison’s thought, the term ‘interest’ becomes almost interchangeable with Burke’s ‘opinion’ and ‘will’, or even ‘feeling’ and ‘sentiment’. . . . A man’s interest is what he thinks it is, just as his opinion is what he thinks. 34

Because interests are subjective and attached to individuals, they are prone to being unstable and shifting, since the opinion and will that determine them can be ﬁckle and changeable. The subjectivity of interests, and the attachment of interests to particular individuals and groups, is further associated with the liberal individualist idea that interests are pluralistic, conﬂicting, and unable to be reconciled— either with each other or with the overall interest of the community—through rational deliberation. Accordingly, the purpose of interest representation, on the liberal individualist conception, is not to take all interests into account in a rational process of deliberative decision-making, but rather to balance, restrain, and accommodate the various interests within a community, and hence to obviate their potentially divisive consequences. 35 33 Since Pitkin did not herself recognize any distinction between ‘liberal individualist’ and ‘liberal pluralist’ conceptions of representation, it is not entirely clear whether she would have thought the liberal pluralist forms I discuss later to be subsumed within her general ‘liberal’ framework, or more properly characterized as a distinct third form of representation. Nonetheless, the general ‘liberal’ model as Pitkin has in fact outlined it is individualist rather than pluralist in character (as I make this distinction here), and it is therefore appropriate to characterize her liberal model as taking a narrower, ‘individualist’, form. See Pitkin, The Concept of Representation. 34 Ibid. 192. 35 Ibid. 196.

Global Social Choice Beyond Nation-State Representation

123

The subjectivity of interest has implications both for the question of what is entailed in representative agency and for the question of the appropriate decisionrule to be employed in the representative system. First, the subjectivity of interest means that we cannot rely on the intelligence, competence, and public virtue of representatives to do the job of discerning the interests that should be pursued in public decision-making. Rather, individuals must be entrusted with the task of identifying interests for themselves. A clear implication of this is that legitimate representative agency can only be conferred through institutional mechanisms that connect individuals to their representatives in a way that enables individuals to communicate to representatives the interests they want to be pursued, and that binds representatives to the pursuit of these interests. Accordingly, the liberal individualist model entails what Pitkin calls a ‘delegate’ as distinct from a ‘trustee’ model of representative agency. The view of interests as unstable and conﬂicting also has an important implication for the issue of what procedure should be employed to produce a collective decision that represents the relevant political interests within the society. Since these various individual interests cannot be reconciled through deliberation (as burkean unattached interests can), the liberal model implies a commitment to some kind of aggregative decision-procedure that is capable of producing a collective decision without a need for any substantive commensurability of the various conﬂicting interests. In formulating aggregative mechanisms for collective choice, liberal individualists further invoke distinctive conceptions of what is entailed in the democratic requirement of equality in the collective decision-procedure. Generally, this is taken to mean that each individual is accorded equal opportunity to inﬂuence the outcome of a decision. It is common to interpret this as the requirement that each individual be accorded a vote of equal weight, which is aggregated with those of other individuals in reaching a decision. As Robert Dahl speciﬁes this requirement, [t]he condition of political equality is satisﬁed if and only if control over governmental decisions is so shared that, whenever policy alternatives are perceived to exist, in the choice of the alternative to be enforced as government policy, the preference of each member is assigned an equal value. 36

This conception of political equality, which Charles Beitz calls the ‘simple’ view of political equality, can be underpinned by more than one deeper conception of individual equality. 37 Two of the deeper conceptions that are identiﬁed by Beitz in his work on political equality are potentially compatible with the broader liberal individualist account of interest representation. On one view, which Beitz calls the ‘popular will’ view, equality of individual voting power satisﬁes the deeper normative goal of giving equal priority to the subjective interest of each individual; here, equal voting power is taken to satisfy some presumptively legitimate ‘social 36

Robert Alan Dahl, A Preface to Democratic Theory (Chicago: University of Chicago Press, 1967),

37. 37

Beitz, Political Equality: An Essay in Democratic Theory.

124

Global Social Choice Beyond Nation-State Representation

choice function’. A key assumption underlying this ﬁrst conception of political equality is that each individual possesses a single uniﬁed interest, embodied in her ‘preference’, and this interest is ﬁxed prior to the decision-making process. Second, there is what Beitz calls the ‘procedural’ view, according to which the equality of individual voting power satisﬁes the deeper normative goal of giving individuals equal procedural opportunities to participate in decision-making; here, it is not each individual’s interest, but rather her status as a citizen, which is accorded equal priority. 38 For the present purposes it does not matter which of these two deeper conceptions of equality we take as underpinning the liberal individualist account. Even having narrowed the range of possible conceptions this far, we can draw a clear contrast with the conception of equality that underlies the burkean account. Whereas the burkean model accords equal recognition to individuals’ objective interests, through advancement of their equally shared common interest and through equal consideration of the various social interests in which they partake, the liberal individualist model accords an equal role to individuals within an aggregative choice procedure, underpinned by either a ‘social choice function’ or a ‘fair procedure’ that is, itself, presumptively legitimate. This conception of individual equality underlying the liberal individualist model of interest representation has an important implication for the way in which individuals need to be distributed among constituencies within the representative system. The liberal individualist model does not require the division of the jurisdictional population into multiple constituencies; rather, the process of collective choice through aggregation can (and perhaps ideally should) occur at a jurisdiction-wide level, such that the jurisdiction comprises a single electoral constituency. However, there are often logistical and other practical reasons for conducting electoral processes on smaller, and often more territorially localized, scales. These considerations can provide a pragmatic rationale for conducting aggregative processes of social choice in two stages: ﬁrst, in the election of representatives at a constituency level; and second, in the process of aggregative decision-making among representatives about more detailed policy decisions. If aggregation is to be conducted in such a two-stage process, however, the demand for equality of individuals at a jurisdictional level implies that the introduction of a second, constituency-level, aggregative process should not distort or alter the ﬁnal outcome of this process from that which would have resulted from a jurisdiction-wide aggregation of individual votes. In order to prevent it from 38

When I refer here to the ‘procedural’ view of equality, I mean to include both the ‘simple’ and ‘complex’ forms of proceduralism distinguished by Beitz; it is arguable that either could be invoked to justify a liberal individualist representative model. The other conception of political equality that Beitz outlines could not, however, provide liberal individualist rationales for simple voting equality. On this other ‘best outcome’ view, according to which equality of individual voting power satisﬁes the deeper normative goal of according equal priority to the objective interest of each individual, equal voting power is taken to satisfy some presumptively legitimate ‘social welfare function’. However, this view, through its reliance on some account of objective interest, is straightforwardly incompatible with the liberal individualist commitment to the representation of subjective individual interests.

Global Social Choice Beyond Nation-State Representation

125

distorting the outcome in this way, constituencies must in the ﬁrst instance comprise groups of individuals of at least roughly equal size. If constituencies comprised groups of diﬀerent sizes, then clearly the votes of individuals from smaller constituencies would count for more, in the ﬁnal jurisdiction-wide aggregation of representatives’ votes, than the votes of individuals from larger constituencies. (This would be true at least so long as each representative agent is to be accorded equal voting weight within the legislature or equivalent representative body. As I discuss further below, it may sometimes be possible to accord diﬀerentiated voting weight to representative agents with diﬀerent constituency sizes, though this of course would create additional institutional complexity at a practical level.) Moreover, the requirement of equality among individuals at a jurisdictional level has implications for the way in which conﬂicting individual interests must be distributed across and between constituencies. Within a majoritarian liberal system of representation, the demand for individual equality would require that if a clear majority of the population in the whole jurisdictional community favoured a particular policy or party platform, then the constituencies should be delineated in such a way that the majority of decision-making representatives in the jurisdictional community also favoured this option. 39 It is possible, however, for constituency boundaries to be drawn in such a way that the overall majorityfavoured option fails to win a majority of decision-making representatives. Such a distortion could simply be achieved by drawing boundaries around geographical clusters of interests in such a way that the overall favoured option has a very large majority in a minority of constituencies, and the less favoured option has a small majority in a majority of constituencies. If such constituency boundaries were to produce signiﬁcant and predictable distortions of the overall decisionmaking process, they would be condemned by liberal individualists as constituting a democratically illegitimate ‘gerrymander’. 40

5.5.2. Liberal Individualist Interest Representation and State Representation To what extent, then, can this liberal individualist conception of interest representation provide a basis for the statist, or territorial, model of global constituencies? The most important implication of the liberal individualist account is that representatives of particular constituencies are empowered to act as the delegates of particular sectional interests within the community. 41 As Madison 39 In a party-based majoritarian system, we would expect that if the majority of individuals favoured a certain party, then that party would win government. 40 Gerrymanders can similarly occur within proportional liberal systems, when constituencies are structured so that the proportions of preferences in the population at large do not reﬂect the proportions of preferences advanced by decision-making representatives. 41 It is not necessary to the integrity of the liberal individualist model of representation that these sectional interests, embodied within particular constituencies, conﬂict with one another; consensus is certainly theoretically possible and permissible. The point, though, is that they can conﬂict without

126

Global Social Choice Beyond Nation-State Representation

saw the matter, such ‘factionalized’ representation was far from ideal, but was superior to the alternative, which he saw as representative agents pursuing their own personal interests, instead of those of any group. 42 On the liberal individualist account, constituencies can thus be understood to be subgroups of individuals within an overall polity, each possessing somewhat independent (and sometimes conﬂicting) interests. 43 With such an understanding of the liberal individualist view of constituencies in mind, it is easy to see how this theoretical framework can help us to make some sense of the ‘territorial’ model of global constituencies. Just as the liberal individualist model prescribes, territorial state constituencies are discrete groups of individuals whose collective subjective interests are represented alongside the subjective, possibly conﬂicting, interests of other groups within the global representative system. The liberal individualist model does not require that the multiple constituencies, embodying various factional interests, should be territorially constituted. A prominent alternative to territorial constituencies is provided by proportional electoral models, which eﬀectively delineate constituencies (groups of individuals with shared representative agents) in accordance with their expressed voting preferences. 44 Nonetheless, territorial constituencies can sometimes possess certain logistical advantages over a proportional system, since a proportional system for delineating constituencies would necessitate conducting elections at the jurisdictional level. In global politics, where jurisdictions (of global, regional, and transnational public powers) span vast distances and accommodate great social diversity (in terms of language, culture, educational attainment, etc.), conducting elections on such a scale could prove hugely diﬃcult. Accordingly, conducting an initial aggregative procedure (entailing the election of representatives) at more localized territorial levels might prove more feasible in practice. These practicalities could provide a rationale, then, for a decision to delineate constituencies within a liberal individualist representative system along territorial rather than proportional lines. 45 undermining the legitimacy of the representative system; in this respect, the liberal individualist conception of interest representation is at variance with the burkean model. 42 See Alexander Hamilton et al., The Federalist Papers, A Mentor Book (New York: New American Library, 1961). See also Anthony Harold Birch, Representation (London: Macmillan, 1972), ch. 5, for a helpful discussion of Madisonian ideas about interest representation. 43 Precisely how the polity should be divided up into these subgroups is an open question within the liberal model—one that I will consider in more depth in Section 5.5.3. 44 For some discussions of proportional representation, see John Stuart Mill, Considerations on Representative Government, People’s edn. (London: Longmans, Green and Co., 1865); Douglas J. Amy, Real Choices/New Voices: How Proportional Representation Elections Could Revitalize American Democracy, 2nd edn. (New York: Columbia University Press, 2002); Guy Lardeyret, ‘The Problem with PR’, Journal of Democracy, 2/3 (1991); Lani Guinier, The Tyranny of the Majority: Fundamental Fairness in Representative Democracy (New York: Free Press, 1994). 45 In democratic debates focused on state-bound polities, other rationales are commonly advanced for preferring territorial to proportional constituencies, prominent among which is that territorial constituencies (purportedly) tend to promote greater governmental stability. The evidence for this empirical proposition, however, is inconclusive. See Arend Lijphart and Don Aitkin, Electoral Systems and Party Systems: A Study of Twenty-Seven Democracies, 1945–1990 (Oxford: Oxford University Press, 1994).

Global Social Choice Beyond Nation-State Representation

127

If we could further make an empirical case that states have, or could acquire, the empirical characteristics necessary to enable them to satisfy the requirements of equity while functioning as distinct constituencies (i.e. if states could be constituted in equal sizes, or designated votes in proportion to population, and a distribution of conﬂicting interests across them not generative of gerrymanders), then the liberal individualist model could thus provide a theoretical framework for justifying the statist model of global constituencies.

5.5.3. A Democratic Critique of State Representation A critical examination of this justiﬁcatory framework for territorial statist representation must consider the same two questions that I asked in examining the justiﬁcations for national representation. First, is the division of the global polity into state-based constituencies a practicable means of enacting the normative ideals of this model? And second, how sound is the liberal individualist model at a theoretical level, as a framework for achieving a democratically legitimate framework for representative social choice?

5.5.3.1. A Critique of the Practicality of the Statist Model There are some good empirical grounds for thinking that representation of states within aggregative global decision-making processes can enact some key ideals of the liberal individualist representative model. The fact that states generally encompass populations that are institutionally integrated to some degree (in much of the world, to a signiﬁcant degree) means that they are generally quite practical candidates to serve as local democratic constituencies. Whereas instituting election mechanisms on a global scale would, as I have said, pose huge logistical diﬃculties, it is arguable that instituting them within states would be a more feasible task. This is not, of course, true for all states, and the fact that so many states in the contemporary world are undemocratic, and thus not representing all individuals within them in accordance with the liberal individualist ideal, presents a clear problem for this approach to justifying state constituencies within a framework of global social choice. Nonetheless, it is arguable that it would be diﬃcult to identify in the contemporary world any alternative political groupings that would provide more logistically viable locations for the operation of electoral politics. Despite these advantages of territorial statist constituencies, there are two key ways in which the statist model fails to satisfy the normative requirements of the liberal individualist model in practice. More speciﬁcally, there are two ways in which this territorial model fails to provide a framework for an equitable accommodation of the various conﬂicting interests in global society, of the kind required by the liberal individualist representative model. First, the system of representation by territorial states in practice deviates from the liberal individualist requirement that constituencies should be of roughly equal size, to ensure that each individual enjoys equal representation in the ﬁnal

128

Global Social Choice Beyond Nation-State Representation

decision-making process among representatives. It deviates in this way within the framework of existing states, because an equal vote is accorded to each state irrespective of the fact that the populations of various states are vastly diﬀerent. This problem is not insurmountable, however, since it could potentially be overcome if current decision-making structures were altered to accord appropriate weighting to states with more or fewer citizens. A range of proposals for according representation to states proportionate to their populations has been raised over the years in debates about reform of the UN and other public global institutions, 46 and such changes could, potentially, bring the territorial model of representation into closer conformity with liberal ideals. Second, the system of representation by territorial states fails to satisfy the liberal individualist ideal of equality in practice, because of the way in which conﬂicting interests are distributed among states in the contemporary world. I have already explained how, on the liberal individualist model, equality requires that conﬂicting interests in global society should be distributed in such a way as to avoid the problem of a gerrymander. To clarify the demands of this requirement, we can identify two possible ways in which conﬂicting interests could be distributed across territorial constituencies within the global polity, in accordance with the requirement of liberal individualist equality. First, the distribution of interests within each territorial constituency could reﬂect the distribution of interests within the overall global polity, such that each territorial constituency is a kind of ‘microcosm’ of the global polity. Second, the various conﬂicting interests within the overall global polity could be proportionally concentrated in the various territorial constituencies in such a way that, although each constituency has a distribution of interests that conﬂicted with that in the overall global population, the body of decision-making representatives has a distribution of interests equivalent to that within the global population at large. Bearing in mind these two theoretical patterns of interest distribution within global society (either of which would enable territorially delineated state constituencies to satisfy the liberal requirement of equity), we can examine why achievement of such interest distributions would be very diﬃcult in practice given the empirical constraints of global politics. The ﬁrst distribution of interests that could satisfy the equity requirement— a relatively homogeneous distribution of interests across all states such that each state constituency is a microcosm of global society—is clearly in stark contrast with the actual distribution of interests in real global society. In practice, many signiﬁcant conﬂicting interests within global society are concentrated in territorial areas (commonly states), such that the populations of diﬀerent states often have widely divergent and strongly conﬂicting interests. 46 See, for example, Association of World Citizens, Working Together to Build a World Community: Options for a People’s Assembly (2005 [cited March 2005]); available from http://www.worldcitizens. org/unoptions.html.2005 Angela Wood, Structural Adjustment for the IMF: Options for Reforming the IMF’s Governance Structure (Reform Watch, 2001); available from http://www.reformwatch.net/ ﬁtxers/120.pdf

Global Social Choice Beyond Nation-State Representation

129

To some extent, this territorial concentration of conﬂicting interests is inevitable in global society because of the potentially signiﬁcant impact of the natural environment and resources, both of which are strongly linked to territorial location, on a population’s interests. For example, territorial location can have a signiﬁcant and permanent impact on the nature of a population’s interests in global environmental policy: certain low-lying countries are disproportionately aﬀected by global warming issues; other territorial regions by water shortages; still others by pollution problems, and so on. To take another example, territorial location can have a signiﬁcant and permanent impact on a vast range of resource-related policy issues, such as those relating to the regulation of natural resource exploitation and trade, and those relating to distributive entitlements to the world’s natural resources. The signiﬁcant territorial concentration of social interests in global society does not result only from territorially diﬀerentiated natural environments; it also arises from territorially localized cultural values and political systems. Although cultures can be transformed over time, territorially localized cultural diversity is currently a fact of life in international society, and will remain so for a very long time to come, if not permanently. Cultural values of course have a huge impact on interests, conceived in the subjective liberal sense, since these values signiﬁcantly inﬂuence the goals that individuals and groups will identify as desirable. Accordingly, localized cultural diversity inevitably produces substantial localization of interests. Territorially based political institutions—in particular, the internal political institutions of territorial states—further reinforce the territorial localization of interest in global society. The character of local laws and policies adopted within each territorial unit can not only reinforce and perpetuate territorially bounded cultural values but also signiﬁcantly aﬀect the material prosperity and well-being of these populations, and hence their interests in relation to welfare-oriented global policies. Given that these material, cultural, and political factors perpetuate the territorial concentration of many important interests within global society, it is clear that this ﬁrst theoretically acceptable distribution of global interests—that is, an even distribution of conﬂicting interests across all territorial constituencies—is unattainable in practice. Although such an approach to delineating legitimate territorial constituencies could possibly work within relatively homogeneous polities such as certain ‘nation-states’, it cannot deliver liberal-democratic legitimacy to a polity as diverse as global society. Let us turn, then, to the other theoretically possible distribution of global interests that would be compatible with the liberal equality requirement—that is, the concentration of particular interests within particular territorial constituencies in such a way that the body of global representatives has a distribution of interests equivalent to that within the overall global population. This approach has the advantage of being able to accommodate some territorial concentration of interests within global society, which the previous approach was unable to do. However, in order for the equity requirement to be met here, it is not enough for there to be some kind of territorial concentration of interests. Interests must be

130

Global Social Choice Beyond Nation-State Representation

concentrated in the right proportions in the right places if the representative body is to end up proportionally representative of the various conﬂicting interests in global society. And in practice, conﬂicting interests within global society are not territorially concentrated in the right proportions and places to be compatible with an equitable structure of territorial constituencies. One clear example of an interest group whose territorial concentration in particular territorial states is insuﬃcient to provide proportional representation in global society is that of indigenous peoples. Indigenous populations in particular territorial locations share many of the same interests with indigenous populations elsewhere, with respect to issues such as protection of cultural rights, entitlements to land and other natural resources, special education and health care needs, and so on. We can therefore see indigenous peoples as a democratic ‘faction’ or ‘interest group’ of a kind that the liberal model seeks to accommodate equitably with other conﬂicting interests in global society. The UN estimates that there are at least 300 million indigenous people in global society. 47 Since this means that the population of indigenous people is close to 5 per cent of the total global population, this second approach to institutionalizing the liberal model would require indigenous interests to control 5 per cent of states sending representatives to global forums. In reality, however, indigenous populations are not concentrated in 5 per cent of the world’s territory; rather, they are distributed across more than seventy countries on ﬁve continents. 48 Consequently, although indigenous populations are in the majority in certain states (such as Bolivia and Guatemala), they are in most cases a minority population within a state. 49 Because the overwhelming majority of these indigenous people are in permanent minorities within their territorial states, a territorial structure of global constituencies permanently marginalizes their interests from global representative decision-making. This example illustrates clearly, then, how the social diversity and complexity in the global polity precludes this second potential means of satisfying the liberal requirement for equity. It is clear from this that both of the patterns of interest distribution within global society that could theoretically satisfy the liberal individualist requirement of equality are practically unattainable in our diverse and complex global society. Given the failure of the territorial model with respect to this fundamental normative purpose of a liberal individualist system of interest representation, we must conclude that the territorial model cannot in practice be justiﬁed within the terms of this liberal individualist representative model. 47 See United Nations, First Meeting of Permanent Forum High Point of UN Decade (Indigenous People Backgrounder, 2002 [cited February 2005]); available from http://www.un.org/rights/ indigenous/backgrounder1.htm 48 Ibid. 49 Even when they are in the majority, or when they comprise the largest population group, they tend to be socially and economically disadvantaged, and marginalized from mainstream political institutions, so their numbers do not translate into a dominant political position. This is not, of course, due to a problem with the constituency boundaries, but to internal political factors; however, the reality of this internal democratic exclusion serves to compound and reinforce exclusions that arise from the location of constituency boundaries.

Global Social Choice Beyond Nation-State Representation

131

5.5.3.2. Theoretical Limitations of the Justiﬁcatory ‘Liberal Individualist’ Framework Given the problems with enacting the liberal individualist representative ideal in practice through statist territorial constituencies, should we then conclude that a better approach would be to attempt to enact the representative ideal through some other institutional means? Should we, for instance, attempt to establish some kind of proportional system of representation at a global level, in order to secure equal representation for all individual interests irrespective of their territorial concentration? If it could be successfully implemented in practice, a proportional system would certainly address the problems of gerrymanders that I have identiﬁed as endemic to a statist system of representation. But even if this could feasibly be instituted at a global level (and I have already suggested that this is unlikely), I would further argue that this would still fail to constitute a satisfactory global representative system, because the liberal individualist model is itself ﬂawed as a theoretical framework for conceptualizing legitimate democratic representation. It must be acknowledged that the liberal individualist representative model does have certain theoretical advantages as a framework for representative social choice. In particular, the subjectivist account of interests upon which it is based is more compatible with the value of democratic autonomy than is the objectivist account of interests underpinning the burkean model, for the reasons I previously explained in my critique of the burkean model. There are, however, two key problems with the liberal individualist model, which undermine its theoretical adequacy in other ways. First, the simple conception of equality that underpins this model is normatively problematic because it fails to take account of the diﬀering intensity of individuals’ preferences. Whichever ideal of equality is taken to underlie the liberal individualist representative model, there are reasons to think that the diﬀering intensity of individuals’ preferences should be taken into account in reaching legitimate collective decisions through aggregative procedures. If one is committed to the egalitarian notion that the will of each individual (reﬂected in her subjective preference) should be taken equally into account in some social choice function, it seems to follow that the strong will of an individual who has an intense preference on a particular issue should be accorded more weight than the weak will of a relatively indiﬀerent individual. This is so because it would be hard to ﬁnd arguments for the proposition that the notion of an individual’s ‘will’ should be interpreted to focus exclusively on the qualitative aspects of the will, to the complete exclusion of its quantitative aspects. Similarly, if one is committed to the egalitarian notion that collective choice procedures should reﬂect recognition of the equal status of all individuals in the sense that they should be ‘fair’ to all individuals, there are compelling grounds for maintaining that stronger preferences should be accorded more weight. It is beyond the scope of my project in this book to develop a thorough philosophical defence of this claim I am making about the substance of egalitarian fairness, so for the present

132

Global Social Choice Beyond Nation-State Representation

purposes I will simply declare that I am invoking here a strong intuition that it would be unfair, in general, to consistently allow the weak preferences of an apathetic majority to ride roughshod over the intense preferences of a minority. Despite this normative imperative to take intensity of preference into account, the liberal individualist model does not provide any institutional mechanisms for doing so. Since this model is committed to the proposition that only subjectively identiﬁed preferences can be taken into account in collective decision-procedures, its implementation is dependent upon the impracticable requirement to measure interpersonal diﬀerences in sensations. Accordingly, it is not possible in practice to satisfy this normative ideal through any practically conceivable aggregative decision-procedures. Thus, as Robert Dahl argues, however strongly we may believe sensate intensity to be a fact, we cannot directly observe and measure interpersonal diﬀerences in sensations. Hence we cannot hope to establish any political rules to deal with problems of sensate intensity, ethically desirable as such rules possibly might be. 50

If we recognize the normative imperative to take account of preference intensities in collective choice procedures, we must therefore recognize that this presents an important challenge to the liberal individualist model as a representative ideal. A liberal individualist could respond to this argument about the problem of preference intensity by claiming that the equal weight that is accorded to individual preferences on the liberal individualist model is not meant to reﬂect equal concern for any aspect of individuals’ subjective sensate experience, as my argument so far has supposed. Rather, it could be argued, this equal weight is meant to reﬂect recognition of a certain objective fact about individuals’ equal status as members of a democratic polity: the fact that all individuals are equally subject to their common public power (conceived as their shared state, or its laws which apply equally to all), and as such have an equal stake in its decisions. 51 This argument may have some validity in those democratic nation-states in which individuals are aﬀected (more or less) equally by their common agent of public power—their state—and therefore share an equal stake in its decisionmaking procedures. In global politics, however, the empirical supposition that individuals are aﬀected equally by all those agents of public power in which they share a democratic stake is clearly unsustainable. It is widely recognized in empirical analyses of ‘global governance’ that the myriad forms of public power in global politics—including some non-governmental forms of the kind I outlined in Chapter 3—often impact quite unevenly upon the various individuals within their stakeholder jurisdictions. As David Held and Anthony McGrew put this point, the institutional architecture of global public power ‘has a variable geometry 50 Dahl, A Preface to Democratic Theory, 118. More generally, see Dahl, A Preface to Democratic Theory, ch. 4, for a discussion of the problem of diﬀerentiated intensity of individual preferences. 51 Only a liberal individualist committed to an underlying ‘proceduralist’ conception of equality could make this claim; a ‘common will’ conception of equality would not be compatible with such a line of argument, because of the recourse in this argument to some objective assessment of individuals’ interests as ‘citizens’, or members of a democratic community.

Global Social Choice Beyond Nation-State Representation

133

in so far as the relative political signiﬁcance and regulatory capacities of these infrastructures vary considerably around the globe and from issue to issue’. 52 A conception of equality in social choice, based on an underlying procedural notion of fairness, would accordingly need to recognize and incorporate this fact. At a general theoretical level, we can therefore recognize that the normative limitation of the liberal individualist model that arises from its failure to recognize diﬀering preference intensities is especially acute under circumstances in which individuals are aﬀected to diﬀering degrees by the public power in question. Given this general limitation of the liberal individualist representative ideal, it appears especially unsuitable for application within the contemporary global domain; this is so since individuals are commonly aﬀected to more greatly diﬀering degrees by certain agents of public power in global politics (including many NGOs) than is the case within many state-based democratic polities. The second key normative limitation of the liberal individualist model derives from its failure to make theoretical provision for the establishment of legitimate deliberative structures underlying the aggregative decision-procedures. Before I explain what I mean by the claim that the deliberative structures underlying aggregative decision-procedures must be legitimate, I must ﬁrst explain why we should think that aggregative decision-procedures need to be underpinned by deliberative structures of any kind. If liberals are committed to the view that representative institutions are mechanisms for legitimately regulating conﬂicts of interest, rather than for deliberatively identifying collective interests, it may seem plausible to think it follows that aggregative decision-procedures of the kind endorsed by liberal individualists need not (and perhaps cannot) be underpinned by any kind of deliberative structures. A liberal representative system cannot wholly disregard the need for collective deliberative agreement, since such agreement will be necessary at least to perform the minimal function of generating consensus about the constitutive features of the liberal representative choice procedures. The work of many ‘social choice’ theorists and deliberative democratic theorists in recent years can help us to be more precise about the functions that deliberative structures must perform, in order to constitute workable aggregative choice procedures of the kind endorsed by liberal individualists. The ‘impossibility’ results of social choice theorists such as Condorcet and Arrow famously established certain limits on the conditions under which meaningful collective choices can be reached through aggregative decision-procedures. 53 William Riker further argued that the decisions reached through aggregative procedures could be manipulated through the choice of the 52 David Held and Anthony McGrew, ‘Introduction’, in David Held and Anthony McGrew (eds.), Governing Globalization: Power, Authority and Global Governance (Cambridge: Polity Press, 2002), 9. 53 See Kenneth Joseph Arrow, Social Choice and Individual Values, 2nd edn., Monograph (Yale University. Cowles Foundation for Research in Economics), 12 (New Haven: Yale University Press, 1970); Jean-Antoine-Nicolas de Caritat Condorcet, Iain McLean, and Fiona Hewitt, Condorcet: Foundations of Social Choice and Political Theory (Aldershot: Elgar, 1994). When I refer here to the notion of collective choices being ‘meaningful’, I am referring to the requirements (speciﬁed more formally by Arrow) that they be both consistent and minimally responsive to individual preferences.

134

Global Social Choice Beyond Nation-State Representation

particular decision-procedure employed, and consequently that these decisions are either arbitrary or dictatorial. 54 On the basis of this work, it is possible to demonstrate that the possibility of collective aggregative decisions that are meaningful, non-arbitrary, and non-dictatorial depends upon the presence of one or many of several possible forms of underlying agreement, at either ‘substantive’ or ‘meta’ levels. 55 First, it is possible to achieve meaningful social choice under conditions whereby there is suﬃcient substantive agreement among individuals on the preferred decision outcome. Obviously, substantive consensus on the preferred outcome (unanimity) would constitute suﬃcient agreement here, but some degrees of substantive convergence short of unanimity may also enable meaningful collective choices through aggregation. 56 Second, it is possible to achieve meaningful social choice under conditions where there is meta-agreement among individuals on a shared ‘structuring dimension’, such that the preferences of the group of individuals satisﬁes what is called ‘single-peakedness’. 57 This means that individuals must agree on a particular ‘ordering of these options from “left”-most to “right”-most such that each individual has a most preferred position on that “left”-“right” dimension and prefers options less and less as these options get more and more distant from his or her preferred option’. 58 Third, it is possible to achieve non-arbitrary and non-dictatorial collective choice under conditions where there is meta-agreement among individuals on the appropriate aggregative procedure to employ in reaching a decision. Such agreement on the appropriate aggregative procedure must include agreement on two ‘meta’ issues. First, there must be agreement on the relevant agenda items— that is, the content of the various alternatives among which individuals must choose in the aggregative procedure. 59 Second, there must be agreement on the appropriate social choice function to be employed to aggregate individual interests into a collective outcome. 60 54 William H. Riker, Liberalism against Populism: A Confrontation between the Theory of Democracy and the Theory of Social Choice (Prospect Heights, Illinois: Waveland Press, 1988). 55 See Christian List, ‘Two Concepts of Agreement’, The Good Society, 11/1 (2001), for a discussion of ‘substantive’ and ‘meta’ forms of agreement and their role in social choice procedures. 56 The degree of convergence short of unanimity that can enable meaningful collective choices in this way is a disputed question. See William Gehrlein, ‘Social Homogeneity and Condorcet Winners: A Weak Connection’, paper presented at the Annual meeting of the Public Choice Society, Charleston, South Carolina, 2000, for a sceptical view of the utility of more minimal forms of substantive agreement in generating meaningful social choice. 57 See Duncan Black, ‘On the Rationale of Group Decision-Making’, Journal of Political Economy, 56 (1948). 58 See List, ‘Two Concepts of Agreement’. 59 Agreement on ‘agenda items’ is related to the agreement on ‘structuring dimensions’, discussed above. 60 Any of a range of aggregative social choice functions could be employed here. For a discussion of some prominent alternatives such as the borda count and the condorcet vote, see Iain McLean and Arnold B. Urkin (eds.), Classics of Social Choice (Ann Arbor: University of Michigan Press, 1995); Iain McLean, Public Choice: An Introduction (Oxford: Basil Blackwell, 1987).

Global Social Choice Beyond Nation-State Representation

135

Several deliberative democrats have recently developed some persuasive arguments for the view that pre-aggregative forms of agreement, of the kinds I have just outlined as pre-requisites for meaningful, non-arbitrary, and non-dictatorial aggregative decisions, can be generated through democratic deliberation. 61 The case for this view rests upon various empirical propositions that have not yet been adequately veriﬁed through empirical investigation. Nonetheless, the arguments for this view, and the empirical results attained so far, are highly suggestive, and give us good grounds for maintaining that representative institutions employing aggregative decision-procedures must incorporate some underlying deliberative procedures, in order to achieve the necessary pre-aggregative forms of agreement. We can distinguish three key arguments for the proposition that deliberative procedures can help to achieve the crucial forms of pre-aggregative agreement that I have outlined here. First, deliberation can help to achieve substantive agreement among individuals through several mechanisms. These mechanisms include the elimination of preferences based on false information or logic, through exposure in the deliberative process; the elimination of preferences that are morally repugnant to community members, through the shaming processes inherent in public deliberation; and the transformation of preferences through the need to defend preferences in the ‘impartial’ terms, based on ideas of public ‘reasonableness’ that are necessary for reasoned deliberation. 62 Second, deliberation can help to achieve ‘meta’ agreement among individuals on a shared ‘structuring dimension’, such that the preferences across the group can be characterized as ‘single-peaked’. Insofar as unstructured preferences are due to unreﬂectiveness and irrationality on the part of individuals, the deliberation process can induce greater preference structuration within the group by generating more reﬂective and rational individual preferences. 63 Moreover, insofar as unstructured preferences are due to issue complexity, or conﬂict over multiple issue-dimensions underlying a single set of choices, the deliberation process can induce greater preference structuration within the group by separating out the various underlying issue-dimensions on which there is social conﬂict, so that separate aggregative decisions can be taken on each. 64 61 In particular, see: D. Miller, ‘Deliberative Democracy and Social Choice’, Political Studies, 40 (1992); John S. Dryzek and Christian List, ‘Social Choice Theory and Deliberative Democracy: A Reconciliation’, British Journal of Political Science, 33/1 (2003); Iain McLean et al., ‘Can Deliberation Induce Greater Preference Structuration? Evidence from Deliberative Opinion Polls’, paper presented at the American Political Science Association meeting, Washington, DC, 2000; Cynthia Farrar et al., ‘Experimenting with Deliberative Democracy: Eﬀects on Policy Preferences and Social Choice’, paper presented at the Swiss Chair’s Conference on Deliberation, The European University Institute, Florence, Italy, 2004; J. S. Dryzek and C. List, ‘Social Choice Theory and Deliberative Democracy: A Response to Aldred’, 34 (2004). 62 See, for example, Miller, ‘Deliberative Democracy and Social Choice’; Robert E. Goodin, Reﬂective Democracy (Oxford: Oxford University Press, 2003). 63 See, for example, Dryzek and List, ‘Social Choice Theory and Deliberative Democracy: A Reconciliation’. 64 See Ibid.; Miller, ‘Deliberative Democracy and Social Choice’.

136

Global Social Choice Beyond Nation-State Representation

Third, deliberation can help to achieve agreement among individuals on the question of the appropriate aggregative procedure to employ in reaching a decision. Here, it can facilitate agreement both on the question of the relevant agenda items among which individuals must choose and on the question of the appropriate social choice function to be employed to aggregate individual preferences. The mechanisms through which deliberation can help to generate these forms of agreement are similar to those through which it can facilitate agreement on substantive preferences, as outlined above. 65 Once we recognize that any representative institutions involving aggregative decision-procedures must incorporate deliberative structures that can generate some or all of these various forms of agreement, we face the imperative to ensure that these deliberative structures are democratically legitimate. This is necessarily so, since reaching meta-agreement through deliberation could only be considered an improvement on other arbitrary or dictatorial means of settling these preaggregative meta-issues if the deliberative process were capable of conferring democratic legitimacy, and thus were itself non-arbitrary or dictatorial. For a deliberative process to confer democratic legitimacy in this way, it must reﬂect a democratic consensus within the jurisdictional community in which the aggregative procedure is to be employed. Accordingly, a central element of the democratic legitimacy of deliberative processes is the egalitarian requirement that the interests of all individuals within the jurisdictional community be accorded equal weight in the deliberative process. Deliberative democrats disagree about how participatory the process must be (whether each individual must participate, or whether a representative sample of the population can deliberate legitimately on behalf of the group), but all concur that the various competing interests in the jurisdictional population must each receive deliberative representation. As James Fishkin puts the point, ‘[f]or adequate deliberation a diversity of views comparable to the diversity of views on the issue characteristic of the population as a whole, must be represented in the deliberation’. 66 If deliberative processes underlying aggregative mechanisms are to produce the forms of agreement that can render the outcomes of aggregative procedures meaningful, non-arbitrary, and non-dictatorial, it is therefore essential that all competing interests within the jurisdictional community in which the aggregative procedure is to be employed must be included in the deliberative process. So far, then, I have established that legitimate aggregative decision-procedures require that there be underlying deliberative processes to generate the prerequisite forms of agreement, and moreover that these deliberative processes must be structured in a way that is inclusive of all relevant social interests, if they are to be capable of conferring democratic legitimacy. It remains ﬁnally to explain why we should see these issues as problematic for the liberal individualist model. 65

Miller, ‘Deliberative Democracy and Social Choice’. James S. Fishkin, ‘Deliberative Polling and Public Consultation’, paper presented at the Workshop on Deliberative Democracy in Theory and Practice, European Consortium for Political Research, Turin, Italy, 2002. 66

Global Social Choice Beyond Nation-State Representation

137

Most straightforwardly, the need for these legitimate underlying deliberative processes is problematic for the liberal individualist model because it fails to recognize and incorporate provisions for such deliberative processes within its theoretical framework for democratic representation. In defence of the model, it could be argued that formal institutions of democratic representation and the theoretical frameworks that underpin them need not incorporate provisions for such deliberative processes, since it can be reasonably assumed, as an empirical proposition, that the necessary deliberative processes will function eﬀectively at a more informal level, within the domain of what is commonly called ‘civil society’. If we were conﬁdent that this empirical proposition were true in all societies in which the liberal individualist representative ideal was to be applied—that is, if we were conﬁdent that inclusive deliberative institutions within a sphere of ‘civil society’ produced the requisite forms of agreement for aggregative decisionprocedures in all prospective democratic societies—then this defence of the liberal individualist ideal would have some force. However, since the liberal individualist ideal does not make any formal provisions for such deliberative inclusion, its normative credentials rest heavily on the validity of this empirical proposition within the particular jurisdictional communities in which it is to be employed. It would therefore be a matter of empirical contingency whether or not the liberal individualist model could be legitimately enacted in any given instance, depending on ways in which the deliberative process was structured outside the formal representative process, within the sphere of ‘civil society’. I argue in the following chapter that this empirical proposition is not in fact true in global politics, since the institutions of ‘global civil society’ are insuﬃciently inclusive to achieve legitimate pre-aggregative agreement within a global jurisdictional community, and that the liberal individualist ideal is therefore inappropriate in global politics. 67 Even before we have explained these practical arguments fully, however, we can draw the preliminary conclusion, at a purely theoretical level, that the liberal individualist ideal is ﬂawed because its institutional framework does not include the requisite deliberative democratic procedures to produce the necessary forms of pre-aggregative consensus, without which no legitimate democratic outcome can be generated. By failing to make theoretical provision for these crucial democratic functions, and instead entrusting them to the contingencies of informal ‘civil society’ institutions, the liberal individualist ideal is too radically ‘thin’ to constitute a viable representative ideal. Too much in this model depends upon unsubstantiated (and, as I will later argue, practically unwarranted) empirical assumptions about the necessary features and functions of background ‘civil society’ institutions within any prospectively democratic political community. We need, then, a fuller theoretical account of democratic representation that incorporates some minimal deliberative processes—at the very least those minimal deliberative processes that 67 In global politics, I later argue that we should instead employ a ‘multi-stakeholder’ representative model founded upon a ‘liberal pluralist’ representative ideal, which formally incorporates certain deliberative processes in ways that the liberal individualist model fails to do.

138

Global Social Choice Beyond Nation-State Representation

are necessary to provide the conditions for meaningful, non-arbitrary, and nondictatorial aggregative decision-making.

5.6. CONCLUSIONS In this chapter, I have argued that the conventional nation-state model of global representation is fundamentally ﬂawed from a democratic point of view, no matter whether its constituencies are conceptualized as territorial states in accordance with a ‘liberal individualist’ representative ideal, or whether they are conceptualized as cultural nations in accordance with a ‘burkean’ representative ideal. In view of this fundamental ﬂaw, there is a clear imperative for the development of an alternative or supplementary framework for global interest representation. Despite the normative and practical limitations of the nation-state model of global representation that I have discussed, I have acknowledged that representation of nation-states within public global political institutions does have some democratic credentials. On the one hand, representation of national interests within a ‘burkean’ framework incorporates some deliberative institutions that have important democratic advantages, and moreover it ensures the representation of some of the key social interests in which individuals partake. On the other hand, representation of territorial state constituencies within a ‘liberal individualist’ framework provides clear channels for individual input into the identiﬁcation of interests, and provides a form of democratic equality for individuals, albeit a limited one. My arguments here about the problems of the nation-state model should not, therefore, be taken to imply that the representation of nation-states within public global political institutions has no value and should be jettisoned entirely. The question of whether the nation-state model should be abandoned in this way depends on the viability of alternative representative models, and as I explain in the following chapter, the alternative ‘stakeholder’ model that I outline there is subject to certain limitations of its own. We cannot identify any prescriptive implications of the critique I have presented thus far; ﬁrst, it is necessary to elaborate and critically evaluate my proposed alternative model, to determine whether it could potentially substitute for, or at least supplement, the nation-state model of representation in global politics. It is this task, then, to which I turn in the following chapter.

6 Global Social Choice Through Multi-Stakeholder Representation 6.1. INTRODUCTION In the previous chapter, I critically examined two plausible democratic rationales for the widely accepted nation-state-based model of representative global social choice. The ﬁrst of these rationales was based on a ‘burkean’ defence of national representation, and the other on a ‘liberal individualist’ defence of state representation. Although each of these attempts to justify the nation-state model had some practical and theoretical appeal, I demonstrated that, ultimately, neither approach succeeded in articulating a practically viable or normatively defensible account of representative global social choice. In this chapter, I outline an alternative practical and theoretical framework for global representation—a ‘multi-stakeholder’ model based on a ‘liberal pluralist’ representative ideal—which I argue can redress the weaknesses of the nation-state model in crucial ways. I begin this analysis by outlining in Section 6.1 the key institutional features of the ‘multi-stakeholder’ model, and explaining how it can employ the representation of ‘multi-stakeholder’ interests by NGOs either to supplement or to substitute for the representation of nation-state constituencies by governments. In Section 6.2, I examine the theoretical basis of the multi-stakeholder model, and argue that it can be justiﬁed within the framework of a theoretical model of interest representation that I characterize here as ‘liberal pluralist’. In Section 6.3, I critically examine the strengths and weaknesses of the multi-stakeholder model as a framework for global representation. At a theoretical level, I argue that it has stronger normative foundations than either the burkean or liberal individualist models which underpin the nation-state representative system. It does, however, have certain practical limitations, which mean that it would be unworkable on its own as the sole basis for delineating constituencies in public global decisionmaking institutions. Accordingly, I conclude by advocating a hybrid approach to delineating democratic constituencies in global politics—one that incorporates multi-stakeholder representation within more conventional structures of representation by nation-states.

140

Global Social Choice Through Multi-Stakeholder Representation 6.2. THE MULTI-STAKEHOLDER MODEL OF GLOBAL CONSTITUENCIES

6.2.1. The Idea of Multiple Stakeholder Constituencies To avoid confusion about the nature of multiple stakeholder constituencies, it is important to reiterate the distinction between this concept, and that of a ‘jurisdictional’ stakeholder community, which I characterized in Chapter 4 as the legitimate agent of democratic control in relation to a particular form of public power. I explained there that a jurisdictional stakeholder community is composed of all those individuals who are problematically aﬀected by a given public political agent. As such, members of these jurisdictional decision-making communities share ‘an interest’ in a particular decision in the sense that the decision constrains their autonomous capacities. However, sharing an interest of this kind—that is, sharing the condition of being aﬀected in some autonomy-constraining way—does not necessarily imply a shared interest in the additional sense that any particular decision-outcome will be to the advantage of all individuals. For example, two individuals competing for the same scant critical resources dispensed by an agent of public power are both stakeholders in the decision, since the autonomy of both is aﬀected by its outcome. In such a situation, however, the two individuals ‘have an interest’ in quite diﬀerent decision-outcomes in the sense that diﬀerent decision-outcomes will advantage each, because each individual can only receive the resources at the expense of the other. Since members of decision-making communities can thus possess quite diﬀerent interests from each other with respect to the decisionoutcomes that will advantage them, we can say that they share an ‘outcome-neutral’ interest in a decision. In contrast, multiple stakeholder constituencies arise within jurisdictions in which public power is shared, according to some democratic decision rule, among multiple agents, each of which can assume responsibility for representing a particular subset of interests within the jurisdictional decision-making community. Thus, members of each of these multiple constituencies share not only an ‘outcome-neutral’ interest in a particular decision in the sense that their autonomous capacities are aﬀected by it; they further share ‘outcome-speciﬁc’ interests in these decisions, in the sense that a particular decision-outcome is to the individual advantage of each member of the constituency. In relation to any exercise of public power by a representative agent within a democratic social choice procedure, we can thus identify two distinct kinds of ‘stakeholder’ groups—both of which are commonly discussed (and sometimes, wrongly, conﬂated) within public and academic debates about the democratization of global public power—corresponding with two diﬀerent senses in which we can say that an individual ‘has an interest’ in a decision. First, we can identify a group of jurisdictional stakeholders whose autonomy is constrained by the public decision, whatever its outcome. We can then further divide this group of jurisdictional stakeholders into various overlapping groups of individuals comprising

Global Social Choice Through Multi-Stakeholder Representation

141

multiple stakeholder constituencies—those who share some advantage from a particular outcome of the decision. In Chapter 4, it was the former type of stakeholder group that was the focus of my discussion, whereas in this chapter my focus is on this latter type of stakeholder group. The claim that multiple stakeholder constituencies of this kind should be accorded representation within certain public decision-making structures has been prominent in debates about corporate ethics and corporate ‘governance’ for some time. 1 It is only during the last decade, however, that it has become established within formal international decision-making institutions. Principles of stakeholder participation were extensively explicated in a UN document for the ﬁrst time in ‘Agenda 21’—an agenda for sustainable development agreed to at the 1992 Earth Summit in Rio de Janeiro. 2 These principles have now been institutionalized by the establishment of ongoing ‘multi-stakeholder dialogues’ within the UN Commission on Sustainable Development. 3 The trend towards stakeholder representation has expanded into other areas of the UN system and is embodied in initiatives such as Koﬁ Annan’s ‘Global Compact’, which was devised to establish an ongoing dialogue between the UN and representatives of global business and other stakeholder groups. 4 In contrast to the conventional ideas of global constituencies as ‘nations’ or ‘states’, advocates of multi-stakeholder representation claim that individuals could be better represented if multiple stakeholder constituencies were to be identiﬁed and accorded their own representative agents within global institutions. On the multi-stakeholder model, representation is accorded not only to nation-state constituencies but also to multiple stakeholder groups such as women, farmers, and indigenous peoples, each composed of individuals aﬀected by the same decision, but aﬀected in diﬀerent ways such that their preferred decision-outcomes are diﬀerent. Some advocates of the stakeholder model of global representation may want to claim that stakeholder interests have a similar character to ‘national interests’, in the sense that they can be viewed as objectively discernible ‘unattached’ social interests. I do not think that this view would be theoretically defensible, however, because a democratic system cannot be founded upon the representation of wholly ‘objective’ interests, since democratically autonomous individuals must be empowered to play a central role in the identiﬁcation of their own interests 1 See Will Hutton, Stakeholding and Its Critics (London: IEA Health and Welfare Unit, 1997), for a discussion of some of the arguments surrounding stakeholder consultation in corporate decisionmaking. 2 United Nations Division for Sustainable Development, Agenda 21 (1999 [cited February 2002]); available from http://www.un.org/esa/sustdev/agenda21.htm The agenda outlines proposals for the inclusion of representatives of several ‘major groups’ seen to be aﬀected by decision-making processes related to environmental issues; the ‘major groups’ accorded this status are women, children and youth, indigenous peoples, NGOs, local authorities, workers and their trade unions, business and industry, the scientiﬁc and technological community, and farmers. 3 Further information about these dialogues can be found at www.unedforum.org and www.earthsummit2002.org/msp (accessed January 2008). 4 See www.globalcompact.org (accessed January 2008) for details of this initiative.

142

Global Social Choice Through Multi-Stakeholder Representation

(as I argued in my critique of national representation). Neither, though, can we see stakeholder interests as factionalized individual interests, of the kind that can be represented on a territorial statist model of representation. Unlike territorial constituencies, which accord each individual a unitary interest to be represented by a single territorial representative agent, multiple stakeholder constituencies can— and often do—have overlapping memberships. Multiple stakeholder interests do not, then, comprise the factional interests of particular groups of individuals; rather, individuals can possess many stakeholder interests, each of which can be represented by a diﬀerent representative agent. Moreover, just as some people may want to think of stakeholder interests as objective unattached interests, like national interests, some people may think that stakeholder representation involves a trustee model of representative agency. That is, it may be claimed that stakeholder representatives can objectively discern and speak on behalf of stakeholder interests without an institutionally ascribed mandate from any group of individuals. However, since I have said that a democratic framework of stakeholder representation must view stakeholder interests as primarily subjective, it follows that the conception of representative agency entailed in the stakeholder model must be—at least in some crucial respects—a ‘mandate’ one. That is, representatives of stakeholder constituencies must be institutionally delegated in some way, to ensure that they will advance the particular outcomeinterests that are speciﬁed by their stakeholder constituents. In practice, representative agency on the multi-stakeholder model is commonly ascribed to NGOs. Whereas governments of nation-states are generally identiﬁed as the representative agents of territorial or national constituencies, NGOs are generally identiﬁed as the representative agents of multiple stakeholder constituencies. In Chapters 7 and 8, I discuss the question of whether NGOs are in practice ‘representative’ of stakeholder groups, in terms of satisfying appropriate normative criteria for representative agency. At this point, it suﬃces to establish that if NGOs are to be accorded the status of stakeholder representatives, they must attain this status through some kind of institutional mechanisms that require them to advance the subjectively identiﬁed interests of their constituents, in accordance with a ‘mandate’ model of representative agency. Conventionally, elections are taken to be the appropriate institutional means for conferring representative agency of this kind, although as I argue in Chapters 7 and 8, various other, non-electoral, mechanisms are generally more appropriate means of establishing representative agency within a multi-stakeholder framework. The decision-procedure commonly associated with the stakeholder model of representation is a deliberative one, such that the representatives of multiple stakeholder constituencies are required to deliberate among themselves and reach some consensus on the ﬁnal decision. This decision-procedure is underpinned by a particular conception of equality, which shares elements in common with each of the conceptions of individual equality that underpin the national and state models. In the ﬁrst instance, individuals are accorded equal opportunities to identify the multiple stakeholder ‘outcome-interests’ that are to be represented within a deliberative decision-making process. This equal opportunity for individuals to

Global Social Choice Through Multi-Stakeholder Representation

143

identify for themselves those interests that will receive representation is in most respects akin to the form of equality that is accorded to individuals within a statist representative model. The key diﬀerence is that, on the multi-stakeholder model, individuals are accorded opportunities to identify multiple interests, instead of to identify a single unitary interest communicated by means of a preference in a ballot. Moreover, on the statist model, this equal opportunity is the sole form of equality accorded to individuals (since individuals’ preferences are then aggregated, via representative agents, to reach a decision), whereas on the stakeholder model, individuals must be accorded a further form of equality in the decision-making phase. This is because the multiple stakeholder interests identiﬁed by individuals are not aggregated to reach a decision, but rather are advanced by the representatives of these stakeholder constituencies in the course of a deliberative process. In the process of their decision-making deliberations, representatives must consider not only the claims that they have a mandate to make on behalf of their own stakeholder interests but also the claims made by other representatives acting on behalf of competing stakeholder interests within the overall jurisdictional community. In this phase of the representative process, then, individuals must be conferred a form of equality akin to that conferred within the nationalist model, whereby the various stakeholder interests of individuals are accorded equal consideration in the deliberative process.

6.2.2. Institutional Mechanisms for Delineating Stakeholder Constituencies Having thus outlined the characteristics of multiple stakeholder constituencies at a general theoretical level, we must further confront the question of what institutional mechanisms could be employed to delineate these constituencies in practice. NGOs exercising public power in multilateral decision-making settings commonly recognize the need to make some claims about who it is that they represent, but they are often very vague about the precise nature of these constituencies. This vagueness is illustrated by the results from one 1995 survey of international NGOs, which asked respondents to specify the groups for whom their organization claimed to speak. The NGO respondents provided a wide range of answers, some of which identiﬁed constituencies with speciﬁc memberships, such as ‘Equatorial Committee on Human Rights’, ‘Anglican Church’, and ‘indigenous aborigine’. Others, however, identiﬁed extremely vague constituencies with ill-deﬁned memberships, such as ‘marginalized people’, ‘the oppressed’, ‘poor women’, and ‘the workers’. 5 Some NGOs even characterized the subjects of their representation as interests that are wholly ‘unattached’ to individuals, such as 5 Benchmark Environmental Consulting, Democratic Global Civil Governance: Report of the 1995 Benchmark Survey of NGOs (Oslo: Ministry of Foreign Aﬀairs, 1996), 29–32.

144

Global Social Choice Through Multi-Stakeholder Representation

‘nature’, ‘peace’, and ‘migratory birds’. 6 It is certainly possible to recast these interests as ‘partially attached’, by characterizing their constituencies in terms of supporters or advocates of these unattached interests—that is, those individuals who have adopted these ‘unattached’ interests as their own. Even once the subjects of representation are reconceived in this way, however, the precise group of individuals that would qualify as stakeholders on this basis still remains very indeterminate. How, then, can an NGO establish more precisely the identities of the stakeholders with whom it must establish representative mechanisms of authorization and accountability? Consistent with my discussion of ‘interests’ in Chapter 5, the substance of multiple stakeholder interests cannot be ascertained through the application of any ‘objective’ criteria; rather, the substance of individuals’ interests must be determined subjectively by the individuals themselves. 7 It follows from this that the members of an NGO’s stakeholder constituency must be identiﬁed through some political process in which individuals declare the substance of their interests, and stake their claims to representation as members of a particular stakeholder constituency. It might initially appear that there would be no problem with identifying the memberships of such subjectively constituted constituencies in practice, since an NGO simply represents all those individuals who consider themselves to share the outcome-interests it advances in a decision-making process, and conversely, everyone who considers its political agenda to be in conﬂict with their own are excluded, ipso facto, from its constituency. If this were so, there would be no apparent reason why an NGO in a multi-stakeholder process would need to be authorized by, or accountable to, any particular group of individuals; the NGO could simply contend that all those who objected to its agenda evidently possessed diﬀering interests, and hence failed to qualify as its constituents. If this were so, however, a multi-stakeholder decision-making process would have to be able to accommodate an unlimited range of stakeholder groups, to ensure that all divergent views could receive adequate consideration; if the degree of disagreement among individuals were great enough, this would imply the requirement that a separate representative agent be accorded to each individual within the jurisdiction. However, if this view of multi-stakeholder decision-making were accepted, the distinction between a system of multi-stakeholder representation and a (voluntaristic) system of direct participation would be entirely dissolved. In multistakeholder decision-making on a very small scale, such as within a small local community, it might be feasible to implement such a (potentially) participatory model. However, a key reason why we would generally endorse a representative multi-stakeholder model (rather than a direct participatory model) of deliberative 6

Benchmark Environmental Consulting, Democratic Global Civil Governance, 29–32. Representatives must be bound by the mandate of individual constituents in representing the substance of constituents’ interests, but they may make objective assessments of interest-intensities in the decision-phase of deliberations. 7

Global Social Choice Through Multi-Stakeholder Representation

145

decision-making is that direct participation in decision-making is infeasible as a democratic ideal within very large jurisdictions. In many circumstances there are limitations on the number of stakeholder representatives that it is feasible to accommodate within a deliberative process, which means that stakeholder constituencies cannot be constituted on the basis of detailed speciﬁcations of policy outcome-interests shared by individuals. Rather, stakeholder constituencies must be based on broader ‘clusters’ of interests; such clusters emerge when groups of individuals agree on broad values, despite potential disagreements on more detailed policy questions. For example, an environmental NGO may be included within a multi-stakeholder decision-making process on the grounds that it represents the ‘environmental’ interest, broadly conceived. In such circumstances, this NGO would have an obligation to represent all those who identify ‘environmentalism’ as one of their key political interests. The NGO would then need to establish representative institutions of authorization and accountability to connect it with this constituency, in order to ensure that all self-identiﬁed ‘environmentalists’ can exercise some democratic control over the decisions that are made on their behalf by the environmental NGO that is empowered within the multi-stakeholder process. What is required in practice, then, is some kind of political process through which jurisdictional stakeholders can reach a consensus on a set of general categories of outcome-interests requiring representation in their jurisdiction, such that groups of individuals with relatively minor disagreements can nonetheless be represented as members of a single constituency. A range of diﬀerent mechanisms could be imagined to achieve this. Conceivably, some process of public nomination at the jurisdictional level could identify the key categories of stakeholder interests for which members of the jurisdictional community sought representation in a deliberative multi-stakeholder decision-making process. Perhaps it could be stipulated here that any given interest (such as ‘environmental protection’, ‘indigenous rights promotion’, ‘the economic welfare of farmers’) must have some given number of individual supporters to qualify for representation, and must demonstrate this support through some kind of public petitioning process. 8 Within these established interest-categories, any individuals and groups identifying with these interests would then be required to mobilize politically—either through something like a stakeholder register maintained at a jurisdictional level or through more informal lobbying processes—and to present themselves as members of the given outcome-interest constituency. Such a system would require that all stakeholders enjoyed some form of equal access to these processes of political mobilization, and achieving this could be problematic in some contexts. However, as I discuss in the ﬁnal section of this chapter and further in Chapter 8, it is not clear that problems of equal access would generally be greater for a multistakeholder model than they would be for an electoral model of representation operating with territorial constituencies. 8 Petitioning mechanisms of this kind, such as citizen-initiated referenda, are sometimes used in more established democratic institutions within territorial states.

146

Global Social Choice Through Multi-Stakeholder Representation 6.3. ‘LIBERAL PLURALIST’ INTEREST REPRESENTATION AS A JUSTIFICATION FOR THE MULTI-STAKEHOLDER MODEL

Whereas the current nation-state–based structure of global constituencies can be theorized within the frameworks of traditional liberal individualist and burkean ideals of interest representation, as discussed above, the representation of overlapping transnational ‘stakeholder’ constituencies requires a revised theoretical framework for conceptualizing democratic interest representation. To this end, I present here a theoretical ideal of interest representation that draws on key elements of the two traditional models, harnessing and harmonizing the theoretical strengths of each. I call this conception of interest representation a ‘liberal pluralist’ ideal, since its normative core is liberal, but the interests that are accorded representation within its framework are ‘pluralist’ rather than individualist, in a sense that I explain below. In this way, the liberal pluralist model that I outline here is distinct from the liberal individualist model that I discussed in the previous chapter as a basis for territorial state representation. This ‘liberal pluralist’ model of interest representation builds a bridge between Pitkin’s ‘ideal-typical’ burkean and liberal individualist conceptions, by invoking a distinction between partially and comprehensively attached interests. This liberal pluralist model accepts that interests are in important ways attached to, and subjectively determined by, individuals and groups of individuals; in this way, my model is related closely to the liberal individualist conception of interests. However, the liberal pluralist model challenges the way in which Pitkin’s liberal individualist conception of interests portrays them as ‘attached’ to particular, discrete groups of individuals. Moreover, the liberal pluralist model recognizes that although interests must be in the ﬁrst instance subjectively determined, the ﬁnal decision-making phase of representation requires decision-makers to make some objective evaluations of the relative intensities of the multiple subjective interests under consideration. On Pitkin’s liberal individualist model, interests are attached comprehensively to groups of individuals as well as to single individuals, in the sense that an interest taken in the abstract (the ‘interest of New Zealanders’, e.g.) comprehensively accounts for the interest of each of the individuals comprising the group. Accordingly, what is in ‘New Zealand’s interest’ is equivalent to what is in the interest of each individual New Zealander. This tendency to see an abstract group interest and the comprehensive interests of the group’s members as equivalent is evident in Madison’s association of the concept of conﬂicting ‘interests’ in society with the idea of ‘factions’—concrete groups of individuals in concrete conﬂict with each other within a community and its political institutions. 9 The liberal pluralist model rejects this view of interests as comprehensively attached to groups of individuals, on the grounds that an individual can 9 See Hanna Fenichel Pitkin, The Concept of Representation, 1st Paperback edn. (Berkeley: University of California Press, 1972), 191–203, for a discussion of Madison’s account of the relationship between factions and interests.

Global Social Choice Through Multi-Stakeholder Representation

147

concurrently be a member of many diﬀerent groups with attached interests. An individual can be a New Zealander, a woman, a consumer, an environmentalist, a Maori, and many other things besides; accordingly, her comprehensive individual interest will be determined only partially by the interest of any one of these groups of which she is a member. Rather than conceiving of each individual as possessing a single ‘interest’ encompassing her many needs, values, and opinions, we should instead recognize that each individual possesses multiple partial ‘interests ’, each relating to a diﬀerent aspect of her social life and identity. Under some circumstances—and in relation to certain decisions—an individual’s partial interests can come into conﬂict with each other, forcing her either to choose between them or to ﬁnd some kind of compromise or accommodation. For example, she may have an interest as a consumer in obtaining cheap imported running shoes, but a conﬂicting interest as a human rights advocate in preventing exploitative labour practices in manufacturing industries. If such an individual were asked to identify her preferences with respect to appropriate global norms on the regulation of labour practices, she would be forced to weigh these conﬂicting interests against each other and decide which should prevail, or how a compromise could best be reached between them. At other times, or in relation to other decisions, the various dimensions of an individual’s interest do not conﬂict, and the individual can pursue each one more or less independently of the others. If, for example, this same individual were to give up running, and thus lose her interest in purchasing cheap imported shoes, she would subsequently experience no internal conﬂict of interests in relation to global labour regulation, and could pursue her interest in human rights standards without internal conﬂict or compromise. In this way, Pitkin’s dichotomy between attached and unattached interests is eﬀectively broken down, since group interests are seen as attached only partially to any individual group member. Plural stakeholder interests are attached in the sense that it is possible to identify concrete individuals for whom the interest is partially constitutive of their comprehensive individual interests. But they are unattached in the sense that no speciﬁc individual can be identiﬁed whose comprehensive interest is equivalent to the abstract stakeholder group interest. These plural stakeholder interests also appear very much like ‘burkean’ interests insofar as they are partial interests that an individual takes into account during the internal ‘deliberation’ through which she formulates her subjective assessment of her overall (comprehensive) interest. Like burkean interests, stakeholder interests are parts of a uniﬁed whole, and reconcilable through deliberation. Although this multidimensionality of individual interest, and the correspondingly overlapping boundaries of interest groups, are not recognized in traditional normative accounts of liberal democracy, their recognition is central to much empirical democratic literature—in particular, that in the ﬁeld of liberaldemocratic ‘pluralism’. Of the many writers in this ﬁeld, David Truman has perhaps elaborated and defended this conception of interests and interest groups most clearly:

148

Global Social Choice Through Multi-Stakeholder Representation

The view of a group as an aggregation of individuals abstracts from the observable fact that in any society, and especially a complex one, no single group aﬃliation accounts for all of the attitudes or interests of any individual except a fanatic or a compulsive neurotic. No tolerably normal person is totally absorbed in any group in which he participates. The diversity of an individual’s activities and his attendant interests involve him in a variety of actual and potential groups. 10

The liberal pluralist model, then, provides a theoretical framework for incorporating this established empirical account of individual interests within a viable normative model of legitimate interest representation. This distinctive conception of the way in which interests are ‘attached’ to groups of individuals is not the only way in which the liberal pluralist model diﬀers from the liberal individualist model. It further diﬀers by accommodating some recognition of the need for objective assessments of interest in the ﬁnal phase of decision-making, although it maintains the liberal recognition that mechanisms for identifying stakeholder interests must treat them as subjective in the ﬁrst instance. To explain this diﬀerence, it is important to highlight the complex character of the decision-procedure entailed in the liberal pluralist model. The decisionprocedures entailed in both the burkean and liberal individualist representative models are relatively simple. On the burkean model, national ‘trustees’ are empowered both to identify the substance of national interests and to engage in deliberative decision-making processes through which these various interests are reconciled in a ﬁnal decision. On the liberal individualist model, territorial ‘delegates’ are issued with mandates with respect to the substance of their territorial constituents’ subjectively identiﬁed interests, and further committed to voting in accordance with these mandates in a ﬁnal aggregative decision-procedure. The liberal pluralist model combines elements of these two decision-procedures within a more complex decision-making model. In the ﬁrst instance, stakeholder representatives are issued with mandates identifying the substance of their stakeholder constituents’ subjective interests (through elections or legitimate nonelectoral mechanisms, of the kind I discuss in Chapters 7 and 8). Instead of these stakeholder representatives voting in accordance with their mandates in a ﬁnal aggregative decision-procedure, as on the liberal individualist model, stakeholder representatives make their ﬁnal decisions through a deliberative process in which all stakeholder interests, including those that they have been issued a mandate to represent, must be evaluated and taken into account. Thus, stakeholder representatives in this decisive deliberative process have two sets of responsibilities. First, they must advocate on behalf of the particular stakeholder constituencies that they have been accorded mandates to represent; at this stage of the decision-making process, the interests represented are the subjective interests of their particular stakeholder constituents. Second, they must make assessments of all the competing stakeholder interests under consideration, in terms 10 David Bicknell Truman, The Governmental Process: Political Interests and Public Opinion (Westport, CT: Greenwood, 1981), 508.

Global Social Choice Through Multi-Stakeholder Representation

149

of some standards of public ‘rationality’ and ‘reasonableness’ shared by all members of the jurisdictional community, in order to make a deliberative judgement of the most appropriate ﬁnal decision. 11 At this stage of the decision-making process, the interests they need to represent are those of all the stakeholders within the overall jurisdictional community. To a signiﬁcant degree, the deliberative process of reconciling these various stakeholder interests within the jurisdictional community can be undertaken without the need for invoking any objective assessments of these interests, since representatives should take as authoritative the claims made by each constituency’s own representative about the substance of the outcomes that are in the interest of each. However, objective assessments of the various stakeholder interests may need to be undertaken in one sense: representatives may need to make objective assessments of the relative intensities of conﬂicting stakeholder interests, in evaluating the strength of competing claims for consideration by various stakeholder constituencies. Based on awareness of both the substance and the intensity of stakeholders’ multiple interests, representatives can then apply appropriate shared epistemic and normative criteria (factual correctness, logical soundness, fairness, etc.) to the arguments advanced on behalf of various stakeholder constituencies, and thus reach consensus among themselves on a ﬁnal decision. In key respects, then, the decision-making structure on the liberal pluralist model entails a ‘mandate’ model of representative agency, like the liberal individualist model, since all judgements about the substance of stakeholder interests must be made by individual stakeholders themselves, and representatives are compelled to act as delegates of these subjective interests. The model also entails some degree of trusteeship, like the burkean model, since representatives are empowered to make objective assessments of the relative intensities of interests. However, this form of representative agency is only secondary and subordinate to the mandate aspects, since such objective assessments can only be applied within the constraints imposed by each representative agent’s mandate from her constituency. Just as the decision-procedure underpinning the liberal pluralist model combines elements of the burkean and liberal individualist models, so too does the conception of equality that underlies it. First, the kind of equality that individuals are accorded in the process of issuing mandates to their stakeholder representatives is akin, in key ways, to the equality accorded to individuals on the liberal individualist model. In both cases, all individuals are accorded equal opportunities to participate in the process of specifying the substance of the interests that their representatives must advance. The only diﬀerence here between liberal individualist and liberal pluralist models is that on the latter model individuals can specify multiple stakeholder interests, whereas on the former they can specify only a single unitary preference. 11 As I discussed brieﬂy in Chapters 5 and 6, such deliberative judgements can invoke both shared notions of ‘rationality’ and shared notions of ‘reasonableness’ based on a range of shared norms and principles, such as shared substantive goals and shared normative ideals of ‘justice’ or ‘impartiality’.

150

Global Social Choice Through Multi-Stakeholder Representation

Second, the kind of equality that individuals are accorded in the decisive deliberative process undertaken by stakeholder representatives is akin to that accorded to individuals on the burkean model. That is, both the substance and the intensity of their various stakeholder interests are taken equally into account by representatives in the deliberative process of decision-making. Overall, the only one of Beitz’s four deeper conceptions of equality that would be compatible with this liberal pluralist ideal would be a version of Beitz’s ‘complex proceduralist’ equality, according to which individuals are treated as equal persons, in the sense that decisions are made through procedures that each individual has equally good reasons to accept. 12

6.4. STRENGTHS AND WEAKNESSES OF THE MULTI-STAKEHOLDER MODEL OF GLOBAL REPRESENTATION In order to evaluate the democratic credentials of the stakeholder model of global representation as an alternative to representation by nation-states, we must critically examine both its theoretical and its practical features. I begin here by examining the normative credentials of the ‘liberal pluralist’ theoretical ideal underpinning stakeholder representation, and then consider the practical question of the extent to which this ideal can be enacted eﬀectively through multi-stakeholder representation in public global political institutions.

6.4.1. An Evaluation of the Liberal Pluralist Representative Ideal Underlying the Multi-Stakeholder Model The liberal pluralist representative ideal has several theoretical advantages over both the burkean and liberal individualist representative ideals. Its advantages relative to the burkean ideal are easy to recognize, since they are related to those of the liberal individualist ideal. Like the latter ideal, the liberal pluralist ideal has normative advantages over the burkean ideal insofar as it provides clear mechanisms for individuals to identify their own interests, rather than relying solely on the ability of representatives to discern their interests ‘objectively’. As I explained in the previous chapter, the provision of such scope for individuals to identify their own ‘subjective’ interests in the representation process is essential to ensure compliance with the central democratic value of individual autonomy. Since the liberal pluralist model (unlike the liberal individualist 12 A case could be made that the liberal pluralist model would be compatible with all three of the ‘regulative interests of citizenship’ speciﬁed by Beitz as critical in identifying the procedures that all citizens would have equally good reasons to accept. See Charles R. Beitz, Political Equality: An Essay in Democratic Theory (Princeton: Princeton University Press, 1989). It is, however, unnecessary for the present purposes to develop such an argument; the precise regulative interests underpinning the conception of equality associated with the liberal pluralist model is one that I leave unspeciﬁed here.

Global Social Choice Through Multi-Stakeholder Representation

151

model) permits some limited scope for representatives also to make objective assessments of interests, the contrast between the liberal pluralist and burkean models is less stark in this respect than that between the liberal individualist and burkean models. Nonetheless, the liberal pluralist ideal does ensure a signiﬁcant role for individual input in the identiﬁcation of interests, and is accordingly more compatible than the burkean model with the value of democratic autonomy. The liberal pluralist model has some important advantages not only over the burkean ideal but also over the liberal individualist ideal. First, whereas the liberal individualist model is unable to provide a framework for taking account of diﬀering intensities of individual interest, the liberal pluralist model can provide some scope for this important democratic function, by empowering stakeholder representatives to make objective assessments of the intensity (although not the substance) of individuals’ interests in the deliberative decision-making process. In my critique of the liberal individualist ideal in the previous chapter, I explained how the demand for individual equality generates a normative imperative for collective choice procedures to take account of varying intensities of individual interests. I further explained that this imperative is especially strong in global politics, since the standard state-centric assumption that individuals are aﬀected to an equally intense degree by their public political institutions—and therefore are likely to share an approximately equal intensity of interest in the decisions of the institutions—is clearly unwarranted in relation to many public political agents in global politics. Many of the diverse forms of public power wielded in our ‘pluralist’ global polity—by non-state as well as state actors— can have quite uneven impacts across their respective jurisdictional populations. When individuals are aﬀected to diﬀering degrees by an agent of public power, we have especially strong normative reasons for requiring representative mechanisms to take into account their varying intensities of interest in collective decisionprocedures. The capacity of the liberal pluralist model to take account of the intensity of interests, therefore, accords particularly signiﬁcant normative advantages to this model over the liberal individualist model, in relation to many agents of public power in contemporary global politics. It is important to acknowledge here that the scope provided on this model for representatives to make objective assessments of the intensities of individuals’ interests in the deliberative process comes at a certain normative cost, relative to the purely subjectivist process of interest representation on the liberal individualist model. As I have explained, the value of democratic autonomy requires that we should in general ensure that individuals have the greatest possible control over the identiﬁcation of their interests in the representative process. From this viewpoint, it appears to be an advantage of the liberal individualist model that it does not make any recourse to assessments of objective interest, since a less than public-spirited representative agent could conceivably cite dubious (but purportedly ‘objective’) assessments of interest intensity as grounds for dismissing or diminishing the force of any interest-claims that conﬂicted with her own interests. If representative agents were not permitted the licence to make such ‘objective’

152

Global Social Choice Through Multi-Stakeholder Representation

assessments, the possibility of such manipulation of the deliberative process would be minimized. As I have said, however, the liberal individualist model only provides channels for the subjective identiﬁcation of the substance of interests, and provides no channels of any kind for identifying interest intensity. Moreover, the liberal pluralist model only recognizes objectively discerned interests in relation to evaluations of interest intensity; like the liberal individualist model, it maintains that the substance of interests must be subjectively identiﬁed. Accordingly, these models are equivalent in ensuring that the substance of interests is identiﬁed by aﬀected individuals themselves, and the introduction of objective interests into the liberal pluralist model is allowed only insofar as it is necessary to facilitate recognition of varying interest intensity. In relation to the additional dimension of interest intensity, with respect to which the two models are distinguished, there appears to be a trade-oﬀ between the equality-related advantages of taking account of interest intensity and the autonomy-related problems associated with the objective identiﬁcation of this interest intensity. If we were to think that autonomy-based considerations should always take precedence over equality-based considerations, there may be some grounds for wanting to disregard the problem of interest intensity, as is done on the liberal individualist model. However, here we return to the problem that many global agents of public power have very uneven impacts across jurisdictional populations, such that there is great variance in the intensity of stakeholder interests that can claim entitlements to representation in their public decision-making processes. Accordingly, the equality-based problems with ignoring interest intensity are more severe and intractable in relation to these global agents of public power than they would be in many democratic nation-states. In global politics, then, we have especially strong equality-based reasons to think that the normative advantages of taking account of interest intensity outweigh the normative limitations associated with the objectivist mechanism for doing so. Another important normative advantage of the liberal pluralist model over the liberal individualist model is that it is able to make adequate theoretical provision for legitimate deliberative procedures. As I established in the previous chapter, a viable social choice procedure must incorporate some deliberative element, in order to ensure that the outcomes of any subsequent aggregative procedure will at least be meaningful, non-arbitrary, and non-dictatorial. Moreover, these deliberative procedures must themselves be legitimate, in that they must reﬂect a democratic consensus that has been reached through processes that are adequately inclusive of all aﬀected interests within the relevant jurisdictional community. As I have previously argued, the liberal individualist model is unable to make adequate theoretical provision for such legitimate deliberative processes, since it does not incorporate them within formal representative decision-making structures, but rather leaves them to the contingencies of political activity within the sphere of ‘civil society’. In contrast, the liberal pluralist model incorporates, within the formal representative decision-making processes of the relevant jurisdictional community, formal provisions for deliberation among all ‘partially attached’

Global Social Choice Through Multi-Stakeholder Representation

153

social interests. In so doing, the model ensures that the necessary deliberative elements of decision-making are suﬃciently inclusive of competing social interests to confer democratic legitimacy upon the decision-outcomes. Of course it is true that the liberal pluralist model accords a greater role to deliberative decision-making processes than that which I have argued is strictly necessary to ensure that subsequent aggregative processes are meaningful and legitimate. The liberal pluralist model does not incorporate any aggregative decision-making processes in addition to its deliberative procedures, thus entrusting all decision-making functions to these deliberative procedures. As I will later argue, this total reliance on deliberative procedures generates problems of a different, practical, character; at a purely theoretical level, however, we can recognize that this particular normative problem that is inherent in the liberal individualist model is not shared by the liberal pluralist ideal. As I suggested in the previous chapter, a liberal individualist could respond to this argument for preferring the liberal pluralist model by claiming that formal institutions of democratic representation need not incorporate formal provisions for such deliberative processes, since it can be reasonably assumed (as an empirical proposition) that the necessary deliberative processes will function eﬀectively at a more informal level, within the domain of what is commonly called ‘civil society’. I have claimed that this is unsatisfactory as a theoretical response, since it is unsatisfactory to leave such critical democratic functions to the contingencies of empirical circumstance; in this sense, the liberal individualist model is too radically ‘thin’ as a theoretical ideal of representation. Moreover, the force of my argument here for rejecting the liberal individualist model can be strengthened further by directly challenging the empirical assumption on which the legitimacy of the liberal individualist model in global politics must rest: that institutions of ‘global civil society’ can generate a suﬃciently inclusive deliberative consensus on the relevant pre-aggregative issues. To challenge this assumption, we must ﬁrst recall the nature of the preaggregative issues with which we are concerned here, on which agreement must be reached through deliberative processes. As I discussed in the last chapter, there are three kinds of issues on which deliberative agreement can provide a basis for meaningful and legitimate aggregative decision-making: the substance of preferred decision-outcomes; the underlying ‘structure’ of any disagreements about preferred decision-outcomes; and the appropriate agenda items and decisionrules to be employed in any aggregative procedures for resolving conﬂicts about preferred decision-outcomes. The question, then, is whether we have grounds to expect that inclusive deliberative agreement on these questions can be reached in global politics through deliberative structures within ‘global civil society’, without the need for any formalized inclusive deliberative processes within global representative institutions. Much has been written in recent years about the capacity of ‘global civil society’ to enhance democracy by creating participatory structures through which individuals can contribute to deliberative decision-making on issues such as these. ‘Global civil society’—conceptualized as a ‘supranational sphere of social and

154

Global Social Choice Through Multi-Stakeholder Representation

political participation’ 13 composed of various non-state organizations—is sometimes attributed signiﬁcant capacities to facilitate democratic forms of global public deliberation. It is certainly true that such institutions of ‘global civil society’ have been developing rapidly in recent years, and may well continue on this developmental path in future years. However, at the present time we cannot sustain very ambitious expectations of the capacity of ‘global civil society’ to facilitate extensive participatory deliberation. The structures of ‘global civil society’ are at this stage still very weak relative to those of many domestic civil societies, in the sense that the density of participatory organizations at the global level is much thinner than at domestic levels. 14 Moreover, to the extent that global civil society does facilitate public deliberation, it would seem fanciful at this stage to imagine that this ‘global civil society’ is inclusive of all social interests to anything close to the degree that would be required for it to confer any signiﬁcant democratic legitimacy. Rather, institutions of ‘global civil society’ have developed very unevenly across diﬀerent regions of the world, and many global ‘civil society organizations’ are far from participatory in structure. 15 In practice, despite the contributions of the various agents and institutional structures that constitute ‘global civil society’, it seems clear that many of the most eﬀective and inclusive forms of deliberation in global politics (beyond those deliberative processes incorporated within formal representative institutions) will not take place within a ‘global civil society’, but rather within various domestic civil societies, internal to states. The question that arises from this recognition, then, is whether deliberative structures in ‘civil society’ within nation-states could—in the absence of adequate deliberation within ‘global civil society’—perform the deliberative functions necessary to confer suﬃcient democratic legitimacy on global aggregative decisions. It may seem plausible, at ﬁrst glance, to think that this would be so. The suggestion here would be that individuals could deliberate with members of their nation-state 13 Helmut Anheier, Marlies Glasius, and Mary Kaldor, ‘Introducing Global Civil Society’, in Helmut K. Anheier, Marlies Glasius, and Mary Kaldor (eds.), Global Civil Society 2001 (Oxford: Oxford University Press, 2001), 3. 14 For analysis of the relative densities of global and domestic civil societies, see Lester M. Salamon, Global Civil Society: Dimensions of the Nonproﬁt Sector (Baltimore, MD: Johns Hopkins Center for Civil Society Studies, 1999); Lester M. Salamon and Helmut K. Anheier, The Emerging Nonproﬁt Sector: An Overview (Manchester: Manchester University Press, 1996); Lester M. Salamon, S. Wojciech Sokolowski, and Regina List, Global Civil Society: An Overview (Baltimore, MD: Center for Civil Society Studies, Institute for Policy Studies, The Johns Hopkins University, 2003). 15 For analysis of the diﬀering densities of civil society organizations in diﬀerent regions within the global polity, see Helmut Anheier and Hagai Katz, ‘Mapping Global Civil Society’, in Mary Kaldor, Helmut K. Anheier, and Marlies Glasius (eds.), Global Civil Society 2003 (Oxford: Oxford University Press, 2003); Helmut Anheier, Marlies Glasius, and Mary Kaldor, ‘Introducing Global Civil Society’, in Anheier, Glasius, and Kaldor (eds.), Global Civil Society 2001; Helmut Anheier, ‘Measuring Global Civil Society’, in Anheier, Glasius, and Kaldor (eds.), Global Civil Society 2001. For analysis of the various organizational forms of global civil society—many of which are far from participatory—see Helmut Anheier and Nuno Themudo, ‘Organizational Forms of Global Civil Society: Implications of Going Global’, in Helmut K. Anheier, Marlies Glasius, and Mary Kaldor (eds.), Global Civil Society 2002 (Oxford: Oxford University Press, 2002).

Global Social Choice Through Multi-Stakeholder Representation

155

communities about desired outcomes, the grounds for disagreement, and appropriate aggregative decision-procedures, and that their nation-state representatives could convey the outcomes of these deliberations in a subsequent global deliberative process among nation-state representatives. On this basis, the requisite forms of deliberative agreement on substantive issues, agendas, decision-procedures, and so on might be achieved at a global level. However, such a process would only be plausible under empirical conditions whereby the underlying ideas of rationality and reasonableness, which produce deliberative agreement, are substantively the same within local deliberative communities as within the overall jurisdictional community. As I established in the previous chapter, the deliberative process can lead to the requisite forms of agreement by drawing in various ways on certain shared ideas, including ideas of rationality and reasonableness. Shared ideas of rationality are central to the process whereby preferences based on false information or logic are deliberatively eliminated, and also to the process of reaching ‘meta-agreement’ on the structure of underlying disagreements. Shared ideas of reasonableness are central to the process of deliberatively eliminating preferences that are morally repugnant to community members, and also in the transformation of preferences through the need to defend preferences in impartial terms that all could reasonably accept. Further, it is crucial that the ideas of rationality and reasonableness that are invoked in the process of reaching deliberative agreement should reﬂect an underlying consensus within the jurisdictional community in which the aggregative procedure is to be employed (rather than a consensus in some other jurisdictional community or communities). This is because it is the appeal to ideas that are equally acceptable to all individuals bound by the aggregative procedures, in their public capacities as members of that jurisdictional community, which enables deliberative procedures to confer democratic legitimacy upon the decisions they produce. The question of whether the necessary legitimate agreement can be reached in global politics through deliberation conducted at more local (principally nationstate-bound) levels thus depends on whether the ideas of rationality and reasonableness on which deliberative agreement is based would be the same at each level of decision-making. The assumption that ideas of rationality (of the kind invoked in deliberative processes) are likely to be similar between local and global deliberative communities does not appear to present a serious problem. Although there may be some variations in the sorts of things that would be seen to constitute sound empirical ‘evidence’, or the sorts of claims that would be considered rationally ‘coherent’, such variations would probably not be substantial enough to undermine the democratic legitimacy of the deliberative outcomes. However, the assumption that the ideas of ‘reasonableness’ invoked in deliberative processes would be largely equivalent across local and global levels is much more problematic. The sorts of underlying substantive norms to which all people could ‘reasonably’ be expected to assent, and to which people could therefore ‘reasonably’ appeal in public argument, will sometimes vary signiﬁcantly at local and global levels. To illustrate this point, we can review the example of interest-claims

156

Global Social Choice Through Multi-Stakeholder Representation

made by indigenous peoples for various forms of self-determination, and consider the diﬀerent conceptions of ‘reasonableness’ that would probably be invoked in deliberation upon these claims, depending on whether the deliberations were conducted at a global level or within nation-states. Within nation-states (or at least within many liberal-democratic nation-states), public norms on the basis of which ‘reasonable’ claims can be advanced and recognized tend to be strongly individualistic; norms of self-determination for groups do not commonly play a strong role in internal political deliberations within democratic nation-states. 16 At the global level, however, norms of group recognition and collective self-determination are much stronger relative to more individualistic norms. For this reason, the range of self-determination claims for groups that would be considered ‘reasonable’ in deliberations within a nationstate would be more limited than the range of self-determination claims that would be considered ‘reasonable’ in global deliberations. If public deliberation upon issues of indigenous self-determination is conducted within nation-states, the participants in this deliberation will be not only inhabitants of a shared territorial space within a wider global jurisdictional community but also members of the jurisdictional community of their shared nationstate. Despite the fact that there are good grounds for believing that individuals’ public identiﬁcation and solidarity with a ‘global community’ are growing, it still remains true in most cases that individuals’ public identiﬁcation with their nation-state communities will be stronger than with the global. Therefore, when members of nation-states engage in public deliberation together on issues such as indigenous self-determination claims, it is highly likely that the shared norms and conceptions of ‘reasonableness’ that will be invoked by these individuals are those that they share as members of their territorial state, rather than those arising from their membership of a wider global jurisdictional community. Thus, when indigenous claims are raised in local deliberations, we may expect that these local norms are likely to be invoked, and self-determination claims deemed to be of lower priority in agenda setting than other, more individualistic, kinds of interest-claims. In contrast, we would expect that if the same interestclaims were made in a global forum, quite diﬀerent norms may be invoked in evaluating their merits. In the latter case, at a global level, we might expect an outcome according much higher priority to indigenous claims for certain forms of self-determination, because of the greater strength of self-determination norms in global politics. That is, we would expect that the diﬀerence in the substance of the shared norms underlying deliberative processes at local versus global levels would generate diﬀerent ‘reasonable’ decisions through deliberation at these diﬀerent levels. Since democratic principle requires that deliberative decisions are reached through invoking norms that are accepted as reasonable within the jurisdictional 16 Norms of self-determination for groups do have stronger status within nation-states in which certain minorities have succeeded in their demands for political recognition, but many indigenous communities have not achieved such recognition within their nation-states.

Global Social Choice Through Multi-Stakeholder Representation

157

community upon which the decisions will be binding, it would in this example be the norms of self-determination that should legitimately be invoked in evaluating indigenous claims, since these are the norms that operate within the global (rather than local) jurisdictional community. In this example, then, we can see that if global jurisdictional decisions on the entitlements of indigenous communities are to attain democratic legitimacy, deliberation upon these issues must take place at a global level, rather than only within local territorial constituencies. We could think of many additional examples to illustrate this point. In general terms, in societies where human rights are not prominent norms within the local political culture, groups that make interest-claims based on certain conceptions of human rights would be deemed less ‘reasonable’, and accorded lower priority, than would be the case if deliberation on these interest-claims were to take place at the global level, where there is greater recognition of certain fundamental human rights within the public political culture. Conversely, in societies where strongly individualistic norms are predominant within the local political culture, interestclaims based on such individualistic norms would be deemed more ‘reasonable’ and accorded greater priority in local deliberations than at the global level, where less strongly individualistic norms have greater public acceptance. This systemic problem with liberal individualist representation in global politics could explain the fact that governments are generally less trusted by their citizens in certain issue areas such as human rights, environmental protection, and social justice than they are in others such as the promotion of economic prosperity and physical security. 17 Relatedly, it could explain why individuals tend to pursue alternative avenues of political representation (such as supporting NGOs) far more in these ‘normative’ issue areas than they do in others. 18 It could be that, underlying this variation between diﬀerent issue areas in levels of trust accorded to governments as representatives, is the fact that public norms that regulate deliberation on reasonable priorities in these issue areas vary more widely between local and global levels than do public norms regulating deliberation on other issues. As a consequence, it may well be that the priority accorded to these issues by nation-state governments in global decision-making reﬂects the priority accorded to these issues in domestic politics, rather than the priority that their constituents believe they ought to be accorded at the global level. If this were so, it would seriously undermine the proposition that deliberation conducted at local levels provides a democratically satisfactory substitute for deliberative processes conducted at the level of the global jurisdictional community. The more general conclusion we can draw from this discussion is that when the public norms regulating deliberation at local constituency levels diﬀer from those operating at the wider jurisdictional level, the outcomes of deliberations at these diﬀerent levels may be diﬀerent. In such circumstances, it becomes crucial 17 Edelman has found that governments are trusted far less than NGOs in these key issue areas. See Edelman PR Worldwide, The Fourth Edelman Survey on Trust and Credibility (2003 [cited March 2005]); available from www.edelman.com 18 The overwhelming majority of NGOs operating in global politics fall into these or related issue areas.

158

Global Social Choice Through Multi-Stakeholder Representation

for the deliberative process to be conducted at the jurisdictional level at which the decisions are to be enacted; when the jurisdiction of the relevant public power is global, then the deliberation should take place at a global level. Given the structural weaknesses of existing institutions of ‘global civil society’, the only way of ensuring that such global deliberation will take place is to incorporate certain deliberative processes within the formal structures of the representative system. In this respect, then, the advantages of the liberal pluralist representative model relative to the liberal individualist model are clear. By ensuring that all relevant social interests within global society are represented through formal deliberative processes within representative institutions, the liberal pluralist model provides a mechanism for ensuring that adequate deliberation upon global jurisdictional decisions is conducted at the global level. In this way, the liberal pluralist model ensures—to an extent that the liberal individualist model is unable to do—that the priority accorded to the various competing interests with a stake in global decisions reﬂects the public norms within the global jurisdictional community, rather than those operating at local territorial levels.

6.4.2. An Evaluation of the Practicality of the Multi-Stakeholder Model The multi-stakeholder model of global representation has certain practical advantages relative to the nation-state model, in terms of the extent to which its underlying representative ideal can in practice be implemented through multi-stakeholder representation. In particular, the fact that the multi-stakeholder model represents individuals by issue-area rather than by territorial location or nationality ensures that it is better equipped than the nation-state model to accommodate the empirical reality of territorially dispersed interests within global society. Whereas the latter model assumes that individuals share their primary interests with those in their territorial nation-state, the multi-stakeholder model recognizes and accommodates the reality that individuals partake in both national and transnational social interests of various kinds. The implications of this are signiﬁcant, since the nation-state model provides inadequate representation of certain interests: non-national social interests, if the model is conceived as ‘nationalist’; and those individual interests that are underrepresented through territorial gerrymanders, if it is conceived as ‘statist’. These inadequacies seriously undermine the democratic legitimacy of the nation-state model. In contrast, the multi-stakeholder model can ensure an adequate degree of democratic inclusion by ensuring representation for all interests, whatever their territorial basis. Despite this important advantage, the multi-stakeholder model also has some practical limitations, particularly in its capacity to enact, in practice, the underlying liberal pluralist representative ideal. First, the multi-stakeholder model shares with the burkean model the problems associated with complete reliance upon deliberative processes as the mechanism for reaching collective decisions. Although there may be some instances in which a ‘common good’ could be

Global Social Choice Through Multi-Stakeholder Representation

159

identiﬁed by all stakeholder representatives and decisions thus reached through deliberative consensus, there are likely to be many issues on which competing stakeholder interests are directly conﬂicting and irreconcilable. On such issues, the complete reliance of the multi-stakeholder model upon deliberative decisionprocedures will be highly problematic, since failure to reach deliberative consensus would then result in stalemate and indecision. This problem is intractable within the parameters of the model, because the deliberative element of the decision-procedure cannot be simply abandoned in situations of stalemate, without compromising the model’s democratic credentials. This is so because the stakeholder model relies on the evaluative processes of deliberation to achieve equality for individuals in the ﬁnal decision. Since stakeholder interests, like national interests, do not correspond with the interests of any particular individuals, or discrete groups of individuals, according votes to stakeholder representatives would fail to accord equal consideration to individual interests. A second practical problem confronting the multi-stakeholder model arises from the fact that the model requires, as an empirical precondition, suﬃcient material equality among individuals within the jurisdictional stakeholder community to ensure that they all enjoy equal opportunity to mobilize politically and to advance their various outcome-interest-claims within global deliberations. This is a necessary precondition for the multi-stakeholder model because if stakeholder interests are to be identiﬁed subjectively, all parties must have equal opportunities for translating their subjectively assessed interests into political claims for representation. This precondition creates problems for the multi-stakeholder model, since in fact stakeholders often possess widely divergent levels of resources (ﬁnancial, cultural, political, etc.) that could assist them in the process of staking claims for representation. Accordingly, establishing the necessary level of equal access to stakeholder representation in the global domain could at times be very diﬃcult. It is possible that such inequality in stakeholders’ material resources would also pose practical diﬃculties for the statist model (underpinned by the liberal individualist representative ideal), since it is not at all clear that it would be straightforward to establish signiﬁcantly more equal access to institutional mechanisms for specifying the interests of a state constituency. To the extent that the statist model relies upon elections as a mechanism for identifying constituency interests and thus establishing representative agency, the model’s capacity to attain legitimacy under conditions of material inequality depends upon the likelihood that such inequality will undermine the possibility of establishing inclusive election mechanisms within territorial constituencies. It seems plausible that material inequalities would make it more diﬃcult to achieve inclusive election mechanisms than would be the case without such inequalities, but the extent to which this would undermine the inclusiveness of an electoral system at a global level is an open empirical question. Similarly, the capacity of the multi-stakeholder model to attain legitimacy under conditions of material inequality in global politics depends upon the extent

160

Global Social Choice Through Multi-Stakeholder Representation

to which such material inequality undermines the institution and operation of inclusive mechanisms for identifying interests and thus establishing representative agency. In Chapters 7 and 8, I outline certain non-electoral mechanisms for authorizing NGOs and holding them accountable as a means of establishing representative agency within the framework of the multi-stakeholder model. As I discuss there, these non-electoral mechanisms can be vulnerable to problems of unequal access, but there are no clear grounds for thinking that these problems of inequality will be greater for non-electoral than electoral mechanisms for establishing representative agency. The question of whether the multi-stakeholder model is more vulnerable than is the statist model to the problem of insuﬃcient inclusiveness under conditions of material inequality can ultimately be settled only by more rigorous empirical analysis of the diﬀerences between the particular institutions for establishing representative agency that are entailed in each model. However, it must be acknowledged that whichever way this question is resolved, it presents a signiﬁcant problem for the multi-stakeholder model when it is evaluated on its own terms, rather than relative to viable alternatives.

6.5. CONCLUSIONS: THE NEED FOR A HYBRID MODEL OF GLOBAL SOCIAL CHOICE In this chapter, I have demonstrated how the representation of multiple stakeholder constituencies within processes of global social choice can enhance the democratic legitimacy of these processes by redressing some of the normative limitations that are inherent within the conventional nation-state model of global representation (on both its nationalist and statist interpretations). In contrast to the nationalist interpretation of the nation-state model, the multi-stakeholder model empowers individuals to identify for themselves the substance of the interests that should be promoted in decision-making, and in so doing it ensures that individuals are accorded a suﬃcient degree of democratic autonomy. Moreover, its provisions for representing a range of social interests, beyond the narrow category of ‘national’ interests, ensure that the deliberative decision-making process takes appropriate account of all the various interests within the jurisdictional community. In contrast to the statist interpretation of the nation-state model, the multistakeholder model permits representatives to make objective assessments of the relative intensities of the various competing interests implicated in a decision, and to take these into account in decision-making; by this means, the model produces decisions that more adequately reﬂect an appropriate democratic regard for the equality of individuals. The multi-stakeholder model also redresses the failure of the statist model to incorporate adequate provisions for deliberative decisionmaking at a global level, and in doing so is able to ensure that decisions reﬂect the shared values within the global jurisdictional community, rather than values that

Global Social Choice Through Multi-Stakeholder Representation

161

are more appropriately applied internally to nation-state jurisdictional communities. Finally, the multi-stakeholder model avoids the problems of gerrymanders that arise as a result of the territoriality of statist constituencies, and in doing so provides a more practicable framework than the statist model for ensuring that all individual interests are accorded equal representation. These advantages of the multi-stakeholder model are, however, accompanied by certain practical limitations. First, the model’s reliance on deliberative procedures for reaching decisions creates practical diﬃculties for it, because this precludes aggregative resolution of deliberative deadlocks in circumstances where stakeholder representatives can reach no substantive agreement about how to resolve their conﬂicts of interest. Given the likelihood that such intractable conﬂicts of interest will occur quite frequently in global decision-making, we must recognize this as an important practical limitation of the multi-stakeholder representative model. Moreover, the assumption of this model that the members of stakeholder constituencies can take active responsibility for identifying their own interests requires a substantial degree of material equality among the members of competing stakeholder groups, since this is necessary to ensure that all have equal opportunities to secure representation. The absence of such material equality within the global community thus creates practical diﬃculties for the model. Whether these diﬃculties are more serious for the multi-stakeholder model than for the statist model depends on the inclusiveness of the mechanisms for establishing representative agency that are employed by each. I therefore return to this issue in my discussion of representative agency in the following chapters of the book. Overall, the clear normative advantages of multi-stakeholder representation that I have outlined here provide a strong democratic imperative for empowering multi-stakeholder representatives within public global decision-making processes, whenever it is feasible to do so. Representation of individuals by their nation-state governments is unable to accord individuals suﬃcient democratic autonomy and equality in the decision-making process, whereas multi-stakeholder representation is able—where it can feasibly be instituted—to ensure both autonomy and equality for individuals in public decision-making. However, since the multi-stakeholder model does not incorporate any aggregative procedures for reaching decisions in the face of deliberative deadlocks, in most circumstances it is unlikely to provide a ‘stand-alone’ framework for public global decision-making. Rather, multi-stakeholder representation will usually need to be implemented in conjunction with a representative framework that is underpinned by the liberal individualist ideal, to provide a basis for legitimate aggregative procedures. At the present time, probably the most practicable way of approximating a liberal individualist representative framework is through employing territorial state constituencies; despite the practical problems with the statist model, it is not clear how the liberal individualist ideal could more practicably be enacted through other means. It is unclear whether reversion to a statist representative framework would further ameliorate any of the practical diﬃculties associated with ensuring equal opportunity for individuals to seek representation for their

162

Global Social Choice Through Multi-Stakeholder Representation

multiple stakeholder interests. However, if an empirical case could be made for this proposition, this would provide an additional rationale for supplementing multistakeholder representation with representation of individuals through a system of state-based representation. In circumstances where the practical problems associated with enacting a multistakeholder representative framework cannot adequately be overcome, we should therefore employ a hybrid representative model, incorporating elements of both the statist and multi-stakeholder models. On such a hybrid model, as much decision-making as possible should be conducted through deliberative rather than aggregative procedures, and the deliberative process should be structured in accordance with the multi-stakeholder model. Representatives of multiple stakeholder communities—which may sometimes include national and territorial stakeholders as well as other stakeholder groups—should accordingly be empowered to reach as much deliberative agreement as possible in the ﬁrst phase of decisionmaking. Issues on which deliberative consensus among stakeholder representatives cannot be reached should then move to a second phase of decision-making, in which state representatives could employ appropriate aggregative procedures to resolve the problematic issues. A hybrid representative model of this kind would thus accord multi-stakeholder representatives ‘a voice but not a vote’, in line with the democratic slogan invoked by many advocates for the inclusion of NGOs in public global decision-making.

7 Theorizing Global Representative Agency: Non-Electoral Authorization and Accountability 7.1. INTRODUCTION Once we have identiﬁed the stakeholder groups that are entitled to representation in global public decision-making, through the processes discussed in earlier chapters, we are confronted with a further question: Which agents should be recognized as the legitimate representatives of these groups? Unelected agents such as NGOs sometimes claim that their power within global decision-making processes is legitimized by the fact that they are ‘representative’ of particular stakeholder constituencies. For example, the Executive Director of Friends of the Earth, Tony Juniper, has argued that NGOs can gain legitimacy in global decision-making by ‘representing’ certain interests that have ‘public support’, and thus ‘bringing voices to the table of policymaking that otherwise wouldn’t be there’. 1 The question I address in this chapter and the next is whether, and under what circumstances, there are any sound normative foundations for claims of this kind—that unelected agents such as NGOs can possess legitimate representative agency. In my previous discussion of ideals of interest representation, I established that a commitment to the value of democratic autonomy requires that legitimate representative agency be conceived in accordance with a ‘mandate’ model, such that a legitimate representative is one that represents substantive interests that have been subjectively identiﬁed by constituents. However, I have not yet speciﬁed the institutional arrangements that in practice must be entailed in such a ‘mandate’ model of representative agency. Nor have I considered the more speciﬁc question of whether, or how, unelected actors such as NGOs could satisfy the institutional demands of the mandate model, and thus attain the status of legitimate representatives of various stakeholder constituencies. It is this issue of the institutional requirements of the mandate model of representative agency, and its implications for the representativeness of NGOs and other such unelected actors, that is the focus of my analysis here. 1 Review of a Lecture by Tony Juniper, Executive Director of Friends of the Earth, 5 February (2004 [cited February 2005]); available from http://www.eci.ox.ac.uk/pdfdownload/cpp/ lecturereviewjuniper.pdf

164

Theorizing Global Representative Agency

The mandate model of representative agency is generally linked to electoral institutions, such that elections are commonly thought to be the only appropriate mechanisms for identifying legitimate representative agents of particular constituencies. Of course, the precise structure of the electoral system that is necessary to accord representative agency in a given political context must depend upon the nature of the interests that have a recognized entitlement to representation in that context. Within a ‘liberal individualist’ representative framework, the mandate model of representative agency could be enacted through election mechanisms by according each individual within a jurisdiction a single vote, to elect a single representative agent. Within a ‘liberal pluralist’ representative framework—of the kind advocated in Chapter 6—the mandate model of representative agency could be enacted through election mechanisms by according multiple votes to each individual, so that each can elect a range of representative agents to advance the various outcome-interests in which she partakes. If we were to accept, however, that such election mechanisms are the only institutional means for establishing legitimate representative agency, this would pose a serious problem for the argument that non-state actors such as NGOs can serve as representatives of stakeholder constituencies, because such non-state actors are rarely elected by their constituents. Moreover, instituting elections for non-state actors such as NGOs would in many cases be infeasible, for a range of normative, logistical, and organizational reasons. One reason why elections could not generally provide a feasible means of conferring representative legitimacy upon non-state actors is that the intensity of the autonomy-constraining interests that members of their stakeholder communities possess is usually not uniform across individuals. Accordingly, the aggregative decision-rules entailed in elections would be inappropriate mechanisms for collective decision-making, for the reasons discussed in Chapters 5 and 6. Another reason is that many of the elaborate logistical demands of free and fair elections (such as protections against electoral fraud) would be impracticable to establish within the territorially and politically dispersed constituencies of many NGOs, and other non-state actors. Finally, an enormously complex, costly, and confusing network of electoral processes would be required to establish separate electoral processes for each of the vast number of organizationally disparate non-state actors that may impact upon individuals in contemporary global politics, and this would impose an untenable burden on both stakeholders and all but the largest and best-resourced of non-state actors. For these reasons, the electoral view of representative agency is to a large degree incompatible with claims by unelected non-state actors such as NGOs to democratic representativeness. Many critics of NGOs’ public power make this point explicitly, arguing that NGOs have a ‘lack of democratic legitimacy compared to governments based on fair and free elections’. 2 The only way that NGOs could achieve a representative legitimacy equivalent to elected governments, they argue, 2 Peter Wahl, NGO-Transnationals, Macgreenpeace and the Network Guerrilla: On Some Recent Trends in International Civil Society (Shortened Translation of Article in Peripherie 71, 1998 [cited February 2002]); available from http://weedbonn.org/ngos/ngotrends.htm p. 2.

Theorizing Global Representative Agency

165

would be through the election of NGO leaders by their memberships (or wider stakeholder communities). 3 Should we conclude, as these NGO critics do, that unelected NGOs cannot be legitimate representative agents? Or are there other institutional mechanisms that can be employed to establish legitimate representative agency in the absence of electoral processes? In this chapter and the next, I argue that non-state actors such as NGOs can, in fact, serve as legitimate representatives of stakeholder constituencies, since representative functions can be performed not only by elections but also by nonelectoral mechanisms of authorization and accountability. In this chapter, I make this argument at a theoretical level, and in the following chapter, I illustrate my theoretical arguments by examining how NGOs can serve as legitimate representative agents of stakeholder constituencies through such non-electoral mechanisms. I begin my theoretical defence of non-electoral representation in Section 7.1 of this chapter by elaborating a fundamental reason for viewing elections as a legitimate means of establishing representative agency: that elections have the capacity to deliver stakeholders some degree of political control over their representatives’ actions and hence some degree of political control over the public political decision-making in which they have a legitimate stake. It follows from this that, at least in principle, we could establish legitimate non-electoral representative agency if it were possible to identify alternative institutional mechanisms that could deliver equivalent forms of political control to stakeholders. First, though, we would need to specify more clearly the character of the particular forms of political control that are instituted through election mechanisms. With this prospect in view, I turn in Section 7.2 to an analysis of these forms of stakeholder control. I argue that the speciﬁc forms of political control over representatives’ actions that elections deliver to stakeholders are those entailed in institutional mechanisms of authorization and accountability. I explain how appropriate mechanisms of authorization and accountability can deliver certain forms of political control to stakeholders, and I defend the proposition that mechanisms of authorization and accountability are suﬃcient to constitute a conceptually complete model of democratic representation. Finally, in Section 7.3—having defended the claim that mechanisms of authorization and accountability (whether electoral or non-electoral) can establish legitimate representative agency—I elaborate the institutional mechanics of democratic authorization and accountability in greater theoretical detail.

7.2. ELECTIONS AS MECHANISMS OF STAKEHOLDER CONTROL Why it is that so many democrats see elections as essential for the attainment of legitimate representative agency? Elections have emerged in recent democratic 3 See, for example Karsten Nowrot, ‘Legal Consequences of Globalization: The Status of NonGovernmental Organizations under International Law’, Global Legal Studies Journal, 6 (1999), 601.

166

Theorizing Global Representative Agency

history as the pre-eminent institutional model for liberal representative democracy, and they are central to the institutional design of all contemporary democratic states. 4 For many liberal democratic theorists, elections are now viewed as partially deﬁnitive of democracy. In keeping with this dominant electoral doctrine, regular elections are included by Robert Dahl in his inﬂuential account of the several deﬁning features of a liberal democratic system. 5 This pre-eminent status of elections is not due to their being the only institutional option on the table; other proposed mechanisms for democratically selecting representative agents have been advocated since the birth of representative democracy. 6 Of these, three are especially notable. First, an idea that has received much attention among democratic theorists is that representatives should be selected to ‘mirror’ the characteristics of those being represented—in terms of gender, ethnicity, and other such characteristics judged to be socially relevant. 7 This idea has been advocated most notably in some recent democratic debates focused on the need for special representation of disadvantaged and underrepresented social groups within democratic assemblies. 8 The applicability of this idea of ‘mirror’ representation is not conﬁned to debates about representing marginalized minorities within nation-states; Iris Young further applies this model of representation to global politics, arguing that global representation should be based on representation of the various ‘peoples’ of the world, each of which embodies its own distinctive identity and ‘perspective’. 9 In practice, special representation for certain social groups within a ‘mirror’ framework can be combined with election mechanisms in various ways—such as by according quotas of elected representatives to designated social groups. But since the selection of these ‘social 4 For a survey of contemporary democratic systems, see Arend Lijphart and Don Aitkin, Electoral Systems and Party Systems: A Study of Twenty-Seven Democracies, 1945–1990 (Oxford: Oxford University Press, 1994). 5 See Robert Alan Dahl, Modern Political Analysis, 5th edn. (London: Prentice-Hall International, 1991); Robert Alan Dahl, On Democracy (New Haven: Yale University Press, 1998). 6 It should be noted that not all proposals for non-electoral representation are focused on representative agency. Some proposals—such as Andrew Kuper’s proposals for ‘system-centric’ reforms to global public decision-making institutions—emphasize the role of institutional structures, rather than the actions of authoritative agents, in generating representative democracy. See Andrew Kuper, Democracy Beyond Borders: Justice and Representation in Global Institutions (Oxford: Oxford University Press, 2004). However, while such system-level structures are certainly an important element of any democratic system, they cannot replace the need for legitimate public political agency; the question of representative agency is an independent question, and is crucial to all representative systems, whatever wider institutional structures are developed to supplement these agent-centric representative mechanisms. 7 This ‘mirror’ representation is sometimes also referred to as ‘descriptive’ representation. See Hanna Fenichel Pitkin, The Concept of Representation, 1st Paperback edn. (Berkeley: University of California Press, 1972). 8 For discussions of such ‘mirror’ representation, see Will Kymlicka, Multicultural Citizenship: A Liberal Theory of Minority Rights, Oxford Political Theory (Oxford: Clarendon Press, 1995), ch. 7; Anne Phillips, The Politics of Presence, Paperback edition 1998 (Oxford: Clarendon Press, 1995); Iris Marion Young, Inclusion and Democracy (Oxford: Oxford University Press, 2000), Iris Marion Young, Justice and the Politics of Diﬀerence (Princeton: Princeton University Press, 1990). 9 Young, Inclusion and Democracy, ch. 7.

Theorizing Global Representative Agency

167

groups’ for special representation would nonetheless remain a distinct element of the process of selecting legitimate representatives, occurring prior to the electoral process, such ‘mirror’ representation is still recognizable as a distinct mechanism for selecting representative agents. Another, potentially related, suggestion is that representatives should be chosen at random from the general population—that they should be selected ‘by lot’. This model was employed to some degree in ancient Greek and Roman systems, and in early Renaissance Florence. 10 It has also been considered by other, more contemporary theorists—notably, John Burnheim, who has proposed a revised model of representation by lot. 11 Burnheim has argued that representative governance should be entrusted to a ‘statistical sample’ of the population, in much the same way that jurors are selected for judicial proceedings. Burnheim proposes that these statistical samples should be drawn from each of the social groups with an interest in the decision in question, rather than simply from the overall population; to this degree, his proposal incorporates elements of ‘mirror representation’ in conjunction with random selection. Nonetheless, the mechanism of random selection plays a signiﬁcant role in his theory of representative democracy. Finally, it has been suggested by some democratic writers that representatives should be selected (at least in part) on the basis of some special expertise, competence, or character, which would equip them to identify accurately, and to pursue reliably, the interests of their constituents. John Stuart Mill and James Madison— although both also are advocates of certain electoral systems—each displayed some degree of sympathy for this elitist view of legitimate representative agency. Mill advocated a system of proportional representation rather than a simple majoritarianism, in order to enhance the likelihood that good intellectuals would be chosen as representatives, 12 while Madison maintained that ‘[t]he aim of every political constitution is, or ought to be, ﬁrst to obtain for rulers men who possess most wisdom to discern, and most virtue to pursue, the common good of the society’. 13 These sentiments are echoed in more contemporary debates in global politics by those who endorse the representative legitimacy of certain ‘experts’ or ‘technocrats’ on the grounds that they are more competent than the uneducated global majority to identify and advance the interests of global stakeholders. 10 For detailed discussion of the role of this mechanism in these systems, see M. I. Finley, Democracy Ancient and Modern, 2nd edn. (London: Hogarth, 1985) and E. S. Staveley, Greek and Roman Voting and Elections, Aspects of Greek and Roman Life (London: Thames & Hudson, 1972). 11 John Burnheim, Is Democracy Possible?: The Alternative to Electoral Politics (Berkeley: University of California Press, 1989) and D. Mueller, R. Tollison, and T. Willet, ‘Representative Democracy Via Random Selection’, Public Choice, 12 (1972). 12 John Stuart Mill, Considerations on Representative Government, People’s edn. (London: Longmans Green and Co., 1865). See Nadia Urbinati, Mill on Democracy: From the Athenian Polis to Representative Government (Chicago: University of Chicago Press, 2002), for a helpful discussion of Mill’s elitism. 13 James Madison, in Alexander Hamilton et al., The Federalist Papers, A Mentor Book (New York: New American Library, 1961), quote from ‘The Federalist 57’. See also Bernard Manin, The Principles of Representative Government (Cambridge: Cambridge University Press, 1997), for discussion of the conﬂicting elitist and populist ideas in Madison and other early American democrats.

168

Theorizing Global Representative Agency

Given that there are such institutional alternatives to elections as mechanisms for establishing representative agency, what reasons do we have, as democrats, for viewing elections as normatively superior? Of course, there is no single overarching reason for preferring elections, as with all political decisions, a range of considerations come into play, pragmatic as well as principled. Nonetheless, I suggest that there is one reason that is especially central and forceful in convincing us to prefer elections to alternatives such as those identiﬁed above as a means of establishing legitimate representative agency. In identifying this reason, it is crucial to keep in mind that this chapter focuses only on the role of elections in establishing legitimate representative agency, and not on their role in aggregating individual interests as a means of social choice. It is, of course, true that one reason many democrats value elections is that they are thought to serve the utilitarian or egalitarian purpose of according equal weight to unitary individual preferences in an aggregative mechanism of social choice. However, this function of elections is not the topic here, since I addressed the issue of social choice in the previous two chapters, and in doing so defended a somewhat diﬀerent (though related) democratic strategy for making legitimate collective choices within and among constituencies. Once we have narrowed the focus for discussion in this way, then, we can recognize that elections are widely endorsed as a mechanism for delivering legitimate representative agency because of their capacity to give stakeholders a certain degree of political control over the actions of representatives who are invested with political authority. This rationale for electoral representation depends neither upon a claim that elected representatives are more competent or trustworthy individuals than others with respect to their capacity to advance stakeholders’ interests, nor upon a claim that elected representatives’ personal interests will better reﬂect the interests of particular demographic groupings or a random sample of the population at large. Rather, the rationale is based upon the value of giving stakeholders some active political role in deﬁning their own interests, in evaluating how successfully political authorities are advancing them, and in dictating by whom and within what constraints authoritative decisions aﬀecting these interests may be made. 14 To use somewhat more familiar (although potentially more misleading) terminology, we can say that the electoral model for establishing representative agency incorporates some degree of stakeholder participation in political decisionmaking—albeit participation of a very limited kind, which would not satisfy advocates of certain ‘direct’ or ‘participatory’ democratic models that have a more 14 I am claiming here only that this is the principal reason for valuing elections as a means of establishing representative agency; there are also other, well-established, reasons for valuing elections as a means to less ambitious normative ideals of democracy. In particular, elections are a key element of the Schumpeterian ‘competitive elitist’ model of democracy, and are valued in this model as a means for instituting ongoing competition among elites for political authority. Here, though, it is a certain system of ‘competition’, rather than an ideal of ‘representativeness’, that provides the normative basis for legitimate political power. See Joseph Alois Schumpeter, Capitalism, Socialism, and Democracy, 3rd edn. (London: Unwin University Books, 1952).

Theorizing Global Representative Agency

169

‘republican’ character. The limited scope of stakeholder participation within this representative model distinguishes it from deeper ‘participatory’ models; however, the inclusion of some degree of stakeholder participation—albeit it being quite limited—distinguishes the electoral model from the other representative models that I have identiﬁed here. If representatives were to be selected to reﬂect certain demographic groupings, or at random, or on the basis of special expertise or character, then those stakeholders not selected for oﬃce would have no opportunity to voice their concerns or have any input into the decision-making process. Further, those stakeholders selected for oﬃce would be free to act without the constraints imposed by speciﬁc electoral mandates, and stakeholders who judged their decisions as unsatisfactory would have no institutional means of forcing changes or removing them from oﬃce. Elections, in contrast, empower stakeholders both to specify policies that they want their representatives to pursue, and to challenge or remove representatives who fail to perform to their satisfaction. The value conferred upon a system of democratic representation by this kind of stakeholder control over representatives’ actions is derived in part from the straightforward pragmatic need to restrain potential self-interested behaviour of representatives. If stakeholders are not empowered politically to constrain the actions of representatives, then we would expect representatives more frequently and ﬂagrantly to employ their power in pursuit of their own interests rather than those of the public at large. In contrast, representatives who understand that their power is contingent upon the judgements of the stakeholder ‘public’ are likely to try harder to advance this public’s interest. A further signiﬁcant advantage of such stakeholder control derives from the epistemological imperative to ensure that representatives properly understand what the public interest is. Selection of representatives through systems of ‘mirror’ representation, by lot, or on the basis of special expertise or character, relies in each instance on the assumption that it is possible objectively to determine stakeholders’ interests. Since none of these mechanisms provides channels for stakeholders themselves to dictate which interests representatives must advance, the identiﬁcation of stakeholders’ interests is left in each case to representatives themselves. The assumption, implicit in these mechanisms, that representatives will be capable of identifying stakeholders’ interests without direct instruction by stakeholders could be defended in one of two ways, each of which invokes some notion of objective interests. Most simply, it could be argued that stakeholders’ interests are objectively knowable, and that the problem for representative institutions is to ensure that representatives are competent and appropriately motivated to identify and pursue them. A range of supplementary empirical arguments could then illustrate how either ‘mirror’ or ‘random’ selection best produce such competent and publicspirited representatives. Alternatively, a more nuanced explanation could be oﬀered, acknowledging to some degree the subjectivity of stakeholders’ interests, but, nonetheless, incorporating some ‘objective’ knowledge of these interests into the structure of the

170

Theorizing Global Representative Agency

representative institutions. It could be argued that individuals’ subjective interests are to some degree determined (or at least ‘conditioned’) 15 by their particular demographic characteristics, and therefore that the collective public interest can only be represented by a group of agents who are members of a range of demographic groups. On this rationale, the advancement of subjective individual interests is the normative purpose underlying the representation of these various demographic groups. However, the institutional method still requires that relevant social groups must be identiﬁed prior to the decision-making process in which subjective interests are revealed. The ‘objective’ identiﬁcation of relevant social groups prior to the decision-making process must therefore draw on some ‘objective’ knowledge about which social factors are most signiﬁcant in determining or ‘conditioning’ stakeholders’ interests. In contrast to mechanisms of ‘mirror’ and ‘random’ selection of representatives, election mechanisms can institute a system of interest representation that does not rely heavily upon ‘objective’ knowledge of stakeholders’ interests in these ways. Elections empower stakeholders to decide for themselves what their ‘interests’ in fact are, and to reject rulers who fail to advance these interests in ways that stakeholders perceive as satisfactory. A defence of electoral representation thus relies heavily upon a subjectivist liberal conception of interest representation— one version of which I defended in the previous chapter. If the arguments I have advanced for endorsing that version of interest representation are accepted, then this must be seen to provide good reasons for preferring electoral mechanisms for establishing representative agency to the several institutional alternatives outlined above. 16

7.3. AUTHORIZATION AND ACCOUNTABILITY AS MECHANISMS OF DEMOCRATIC STAKEHOLDER CONTROL I have illustrated so far that elections can provide stakeholders with a degree of political control over their public political representatives, and consequently that they are quite properly valued by democrats as a mechanism for delivering legitimate representative agency. It is crucial to recognize, however, that this does not logically imply that elections are the only eﬀective mechanism for delivering such political control to stakeholders; nothing in my analysis so far suggests that there may not be alternative mechanisms for providing such stakeholder control. So far, then, I have not defended elections per se, but rather described the valuable normative function of elections. In Section 7.3.1, I seek to characterize 15 Iris Young employs the idea of ‘conditioning’ rather than that of ‘determining’ to characterize the way in which interests depend on social group membership. See Young, Inclusion and Democracy. 16 As I discussed in depth in Chapters 1–6, the conception of interest representation underpinning elections can best be characterized as ‘liberal individualist’, while the conception of interest representation I defend can best be characterized as ‘liberal pluralist’. Although these diﬀer somewhat in their conception of individual interests, and correspondingly diﬀer in their primary decision-rules, they share a commitment to a subjectivist conception of interests.

Theorizing Global Representative Agency

171

this normative function in more general theoretical terms, thus providing theoretical foundations for a non-electoral representative system that has an equivalent normative function.

7.3.1. The Functions of Authorization and Accountability The theoretical terms that I use here to characterize the mechanisms through which elections deliver political control to stakeholders are authorization and accountability. Mechanisms of authorization and accountability are institutionally distinct but mutually complementary means of regulating the power relationships between rulers and ruled. More speciﬁcally, they are mechanisms for distributing power between stakeholders and public political agents in such a way as to ensure that the power exercised by public political agents remains subordinate, in some signiﬁcant respects, to the power of stakeholders. As Hanna Pitkin notes in her important discussion of political representation, these two kinds of mechanism occur at diﬀerent stages of the representative process. Authorization—the giving of authority to act—involves institutional arrangements that precede and initiate representation, while accountability—the holding to account of representatives for their actions—involves institutional arrangements that follow and terminate representation. 17 Both function to provide some stakeholder control over representatives’ actions by ensuring that representatives’ power is in some way causally contingent upon the conferral of legitimacy by stakeholders. However, the fact that they occur at diﬀerent stages of the representative process (one precedes public political acts while the other terminates them) does imply that they ensure the contingency of power in diﬀerent ways. On the one hand, mechanisms of authorization address agents who have potential or prospective power, and function to enable them to exercise certain forms of public power. The power of representatives is contingent, through authorization processes, upon the decision by stakeholders to delegate their power to these agents. On the other hand, mechanisms of accountability address agents who already have political power, and function to constrain their exercise of this power. The power of representatives is contingent, through accountability mechanisms, upon decisions by stakeholders about whether to apply or withhold sanctions upon unsatisfactory representatives. 18 17 Pitkin elaborates these understandings of representation in Pitkin, The Concept of Representation, chs. 2 and 3. 18 It is worth noting that sometimes authorization is incorporated into the idea of accountability— see, for example, Hetty Kovach, Caroline Neligan, and Simon Burall, Global Accountability Report 1: Power without Accountability (One World Trust, 2003 [cited March 2005]); available from http://www.oneworldtrust.org/documents/GAP2003.pdf This report rejects what it calls the ‘traditional’ conception of accountability as an ‘end-stage process’ in favour of what it describes as a ‘more open and participative conception of accountability’, which incorporates within the idea of accountability some key aspects of what I am calling ‘authorization’. However, it is analytically useful to maintain the distinction between these concepts, since the institutional mechanisms can in practice be separated from each other, as I discuss later in this chapter.

172

Theorizing Global Representative Agency

It is easy to recognize how these mechanisms function as the key institutional elements of electoral systems. Through the act of voting for a particular candidate, stakeholders ‘authorize’ certain agents to pursue a particular range of actions as their legitimate representative agents. Similarly, the subsequent opportunity to vote these agents out of oﬃce if their performance is unsatisfactory ensures that these agents are held ‘accountable’ for their public political actions. The mechanisms through which elections deliver political control to stakeholders, then, are these mechanisms of ‘authorization’ and ‘accountability’. As such, it is these general mechanisms, rather than elections per se, that we should see as central to legitimate representative agency. Elections are just one institutional instance of these more general mechanisms of authorization and accountability, and it is in this way that they are able to deliver stakeholders the control over representatives’ actions that is so highly valued by democrats.

7.3.2. Authorization, Accountability, and ‘Substantive’ Representation My claim that we should understand representative agency as constituted by mechanisms of authorization and accountability could be confronted by a plausible objection from those familiar with Hanna Pitkin’s classic analysis of the concept of representation, to whom my account here of democratic representation as constituted by mechanisms of authorization and accountability may seem theoretically incomplete. According to Pitkin, mechanisms of authorization and accountability alone are not enough to constitute democratic representation, since they are merely ‘formalistic’ mechanisms, which entail no ‘substantive’ account of what representatives must actually do. Since institutional mechanisms of authorization and accountability take place (respectively) before and after the representatives’ performance of public political actions, Pitkin argues that they exercise no constraints on the actions of representatives during the tenure of their oﬃce. Rather, she suggests that they allow representatives entirely unconstrained public power—they empower representatives to exercise political right without any speciﬁc substantive responsibilities to stakeholders. Pikin outlines her view of the theoretical limitations of authorization and accountability in the following way: Where one sees representation as initiated in a certain way, the other sees it as terminated in a certain way. Neither can tell us anything about what goes on during representation, how a representative ought to act or what he is expected to do, how to tell whether he has represented well or badly. Such questions do not even make sense in terms of formalistic deﬁnitions like the authorization and accountability views. 19

However, Pitkin’s argument here is misleading in so far as she characterizes authorization and accountability only as abstract ‘concepts’ of representation, rather than also as concrete institutional models for achieving representation in practice. 19

Pitkin, The Concept of Representation, 58.

Theorizing Global Representative Agency

173

Authorization and accountability can best be understood as elements of a politically grounded democratic theory, which entails both abstract normative purposes (with respect to interest representation) and concrete institutional prescriptions outlining the political arrangements best capable of achieving these purposes in the real political world. Authorization and accountability are thus mechanisms for ensuring a particular kind of substantive representation in practice: ﬁrst for empowering representatives with some kind of conditional public power, with respect to the interests that they are entitled to advance; and subsequently for ensuring that these representatives exercise their public power in a manner acceptable to stakeholders. 20 Once we recognize authorization and accountability as elements of a concrete institutional—rather than an abstractly conceptual—deﬁnition of democratic representation, it becomes easier to recognize that these mechanisms are, in fact, able to incorporate and instantiate certain substantive ideals of representation. I have already illustrated (in Section 7.1 of this chapter) how election mechanisms (the dominant method of establishing authorization and accountability) depend for their democratic justiﬁcation upon a liberal subjectivist ideal of substantive representation. Here, I argue further that, in fact, this substantive liberal ideal is institutionally enacted by all mechanisms of authorization and accountability—of which elections are simply one institutional example. In Section 7.3.1, I demonstrated how election mechanisms speciﬁcally enact this substantive ideal of representation by establishing concrete mechanisms for stakeholder control over the conduct of representatives during their public political tenure. Here, I want to show how this feature deﬁnes mechanisms of authorization and accountability more generally. As empirical mechanisms, authorization and accountability are not designed merely to precede and terminate substantive representation in a ‘formalistic’ sense, as Pitkin argues. Rather, they are designed to inﬂuence its substantive conduct throughout. 21 First, through mechanisms of ‘authorization’, stakeholders can exercise ongoing control over the actions of representatives during the tenure of their public political oﬃce by attaching various conditions to the public power that is delegated to them. By ‘conditions’, I am referring here not only to negative ‘limits’ on the range of forms of power that the representatives can exercise (i.e. constitutional constraints on executive or legislative power), although such 20 At one point Pitkin seems to acknowledge this, noting that ‘[o]ne might say that the accountability view is a practical or empirical hypothesis, mistakenly disguised as a conceptual view.’ Ibid., 58. However, she does not appear to recognize the signiﬁcance of this insight and does not develop the point further. 21 In describing the functions of authorization and accountability in the following discussion, I am presenting an idealized normative account, rather than a descriptively comprehensive empirical account of the way elections function in practice. In practice, real electoral systems commonly deviate to some degree from representative ideals, as all real political systems deviate to diﬀering degrees from the normative ideals that underpin and regulate them. It is, nonetheless, helpful to articulate the mechanisms of these ideals, so that we can evaluate the extent to which diﬀerent agents and systems comply or deviate in the real world—and more speciﬁcally, so that we can later evaluate the extent to which NGOs comply with or deviate from these ideals.

174

Theorizing Global Representative Agency

constitutional limits can certainly constitute important instances of such conditional authorization when these limits are electorally imposed through constitutional referenda. In addition, I am referring to positive stipulations of the sorts of policies that representatives are authorized to implement. This function of ‘authorization’ is clearly one of the mechanisms through which elections give stakeholders some degree of control over representatives’ actions. Elections ideally serve not merely to give a rubber stamp of legitimacy to preestablished agents of public power; they also serve to confer power on certain agents chosen by stakeholders to exercise public power, and to specify some concrete conditions upon the power that may be exercised. In the context of electoral ‘authorization’, these sorts of positive speciﬁcations of particular policies which a representative is authorized to pursue are commonly referred to as a ‘mandate’. In practice, stakeholders generally accord representatives suﬃcient power to pursue a particular range of public policies—not unrestricted power to pursue whichever policies they may choose. Mandates are generally broad and, of course, allow representatives more or less scope for initiative and elaboration, but they, nonetheless, allow stakeholders to specify, through the act of authorization, what it is that representatives are authorized to do. Authorization thus gives stakeholders ongoing inﬂuence over representatives’ actions, since it empowers them to dictate (more or less speciﬁcally) what interests representative agents are empowered to advance. There are various terms within the vocabulary of political scientists to describe the way in which mechanisms of ‘authorization’ lead representatives to develop and pursue their policies on the basis of stakeholders’ signalled preferences. Many democratic theorists—including Robert Dahl—use the term ‘responsive’ to denote the adoption of policies by a government on the basis of stakeholders’ signalled preferences. In electoral systems, electoral mandates based on candidates’ pre-election policy ‘platforms’ generally provide the primary preference signal from stakeholders, although other kinds of signals can also play important roles within electoral democracies. 22 Representatives that act in accordance with electoral signals are also sometimes referred to as ‘mandate-responsive’, or ‘receptive’. 23 These terms all describe the character of a representative system in which mechanisms of authorization are functioning eﬀectively, since they denote the particular democratic property of systems in which representatives’ actions are conditioned by certain preference signals of stakeholders that precede their authoritative acts. 22

As Manin, Przeworski and Stokes observe, these signals can include public opinion polls and various forms of direct political action such as demonstrations and letter campaigns, as well as votes cast in formal elections. See Adam Przeworski, Susan Carol Stokes, and Bernard Manin, Democracy, Accountability, and Representation (Cambridge: Cambridge University Press, 1999), 9. These other kinds of signals allow individuals to specify the policy content of representatives’ authorized power in between elections through other mechanisms of authorization. I will discuss the signiﬁcance of these non-electoral mechanisms later in this chapter. 23 Manin, Przeworski and Stokes call responsiveness to electoral signals ‘mandate-responsiveness’, as distinct from simple ‘responsiveness’ to non-electoral signals. See Ibid.

Theorizing Global Representative Agency

175

It is important to note here that historically the idea of ‘authorization’ has not always entailed this kind of conditionality imposed by stakeholders upon the public political decisions and actions of their rulers. On the Hobbesian model of ‘authorization’, which is the focus of Pitkin’s discussion of the concept, mechanisms of authorization simply transfer the power of stakeholders to their rulers, without any speciﬁcation of what rulers may do with their power. As Pitkin rightly argues, the kind of unconditional power exercised by rulers ‘authorized’ on such a Hobbesian model would not qualify as democratic despite some formal ‘authorization’ process having taken place. 24 When I refer to ‘authorization’ as a component of democratic representation, then, I am not referring to the nonconditional ‘Hobbesian’ mechanisms discussed by Pitkin. Rather, I am referring to mechanisms of conditional authorization of the kind that generate ‘responsive’ representatives—such as those entailed in the granting by stakeholders of an electoral mandate. This kind of democratic authorization is thus better conceptualized as a process of delegating power than as a process of transferring power, since the idea of delegation implies the conferral of power for a particular purpose. Mechanisms of accountability can also enable stakeholders to exercise control over the substantive actions of representatives during the tenure of their authoritative oﬃce, but they do so in a somewhat diﬀerent way. Whereas mechanisms of authorization empower stakeholders to inﬂuence the conduct of representatives by ensuring that they are aware of the conditions attached to their authority, mechanisms of accountability inﬂuence the conduct of representatives by ensuring that representatives can anticipate subsequent sanctions, which they know will be imposed if they fail to act in a manner acceptable to stakeholders. The example of electoral accountability clearly illustrates the way in which accountability mechanisms can give stakeholders some control over representatives’ authoritative actions. If we assume that representatives (or at least the political parties, movements or alliances to which they belong) are generally motivated to retain their power over time, it follows that representatives’ anticipation of the sanctions that are implicit in electoral defeat can empower individuals to exercise a degree of control over representatives’ action. A representative who wishes to be re-elected will be strongly motivated to make eﬀorts during the tenure of her oﬃce to satisfy stakeholders’ substantive expectations. 25 It is necessary here also to specify the nature of the sanction that democratic accountability imposes upon unsatisfactory representatives. In theoretical discussions of ‘accountability’ it is common to deﬁne this with reference to the general idea of a ‘sanction’ or ‘punishment’. Robert Behn makes this claim in clear terms, arguing that ‘[t]hose whom we want to hold accountable have a clear understanding of what accountability means: Accountability means punishment.’ 26 However, 24

Pitkin, The Concept of Representation. There are also many non-electoral mechanisms of accountability that can exercise control over the substantive actions of representatives, just as there are non-electoral mechanisms of authorization that can exercise this control. I discuss these mechanisms and their signiﬁcance later in this chapter. 26 Robert D. Behn, Rethinking Democratic Accountability (Washington, DC: Brookings Institution Press, 2001), 3. 25

176

Theorizing Global Representative Agency

for our purposes here it is crucial to recognize that speciﬁcally democratic accountability requires the more speciﬁc sanction of revoking an agent’s capacity to exercise public power. 27 Democratic mechanisms of accountability must have the eﬀect of somehow annulling the power exercised by the representative agent. Such accountability is democratic because it confers upon stakeholders a countervailing power, with which they can constrain the power exercised by their authoritative rulers. It is certainly the case that there are forms of ‘accountability’ that do not involve this particular sanction; we may think, for example, that a criminal has been held ‘accountable’ in some sense through being punished, even if this punishment does nothing to prevent recidivism. However, speciﬁcally democratic accountability is distinguished from other kinds by its linkage to public power, and the requirement that the relevant sanction must involve some ultimate ability to take public power away from those who are exercising it illegitimately. We can see how this quality of democratic accountability is reﬂected in mechanisms of electoral accountability. Elections provide democratic accountability not merely by punishing unsatisfactory incumbent governments, but more speciﬁcally by removing them from oﬃce, and thus removing their capacity to exercise public power in future. We may talk about governments being ‘punished’ by electorates when they are voted out of oﬃce, but we do not think that any punishments other than the loss of oﬃce could suﬃce as a matter of course. We would be reluctant to call a system ‘democratic’ if election losses consistently resulted only in punishments such as pay-cuts, loss of prestige, public protests, or (if the performance of representatives suﬃciently outraged public opinion) assassination attempts upon the incumbent rulers. This is the case even although these other sanctions may be just as eﬀective a means of punishing rulers—some of them probably more so, if reliably imposed. We may think that such sanctions would have held these representatives accountable in some retributive sense, but we would not think that they had been held accountable democratically if representatives suﬀering electoral defeat within a political system consistently retained their public political oﬃce. 28 We can see, then, that it is theoretically sound to view mechanisms of authorization and accountability as comprising a normatively substantive model of representation, and not merely a ‘formalistic’ account. Authorization enables stakeholders to exercise ongoing control over representatives’ actions by establishing 27 Robert Keohane and Joseph Nye also stress the centrality of power to the idea of accountability. See Robert O. Keohane and Joseph S. Nye, Democracy, Accountability and Global Governance (2001 [cited March 2004]); available from http://www.ksg.harvard.edu/prg/nye/ggajune.pdf See also Robert Keohane, ‘Global Governance and Democratic Accountability’, in David Held and Mathias KoenigArchibugi (eds.), Taming Globalization: Frontiers of Governance (Cambridge: Polity Press, 2003); Robert Keohane, ‘Governance in a Partially Globalized World’, in David Held and Anthony McGrew (eds.), Governing Globalization: Power, Authority and Global Governance (Cambridge: Polity Press, 2002). 28 There would, of course, be other normative objections to sanctions such as assassination attempts, but we can recognize that the democratic objection would rank among them.

Theorizing Global Representative Agency

177

conditional public power, while accountability enables such ongoing stakeholder control by establishing revocable public power.

7.3.3. Conclusions I have argued so far that we should recognize authorization and accountability— rather than elections per se—as the legitimate basis for representative agency, since elections are just one institutional instance of these more general institutional mechanisms. The way in which elections serve the valuable normative function of giving stakeholders some control over representatives’ actions is by institutionally enacting a system of authorization and accountability. As discussed above, acceptance of mechanisms of authorization and accountability as the basis for legitimate representative agency requires that we accept the general mandate model of representative agency that is entailed in liberal frameworks of interest representation. It also requires that we accept certain empirical propositions about the likelihood that these mechanisms will motivate representative agents to pursue the public interest, rather than their own, during the tenure of their oﬃce. However, since these normative and empirical views underlie the common democratic commitment to electoral representation, most democrats should ﬁnd these presuppositions to be relatively uncontroversial. Once we have recognized that legitimate representative agency can, in principle, be established through non-electoral as well as electoral mechanisms of authorization and accountability, the task of identifying such non-electoral mechanisms can be undertaken in two stages. First, it is necessary to examine more closely the various functional elements of authorization and accountability mechanisms, and on this basis to formulate a generalized theoretical account of these institutional mechanisms. Second, this generalized theoretical account of the functions of authorization and accountability can be applied in particular contexts—such as to particular NGOs and their stakeholder constituencies—to identify appropriate mechanisms of authorization and accountability, which would be functionally equivalent to elections as means for establishing legitimate representative agency. In Section 7.4 of this chapter, I undertake only the ﬁrst of these two forms of analysis, and present a generalized theoretical account of the functions of authorization and accountability. My discussion of how this framework could be applied to confer representative legitimacy upon NGOs in global politics is postponed to the following chapter.

7.4. THE INSTITUTIONAL MECHANISMS OF AUTHORIZATION AND ACCOUNTABILITY My following outline of the prospective structure of non-electoral representative mechanisms involves some theoretical elaboration, at quite a general and

178

Theorizing Global Representative Agency

abstract level, of the various institutional elements of democratic authorization and accountability. Some degree of abstraction is unavoidable in any such generalized account of authorization and accountability, because particular mechanisms must always be highly context-contingent. Particular mechanisms of authorization and accountability must vary both with the speciﬁc forms of power being exercised and with the speciﬁc forms of stakeholder constituency that arise in each given situation. As Hugo Slim argues, with respect to NGO accountability, the right accountability mechanism for a human rights organization working on labour rights in a European country with high levels of literacy and well developed unions is not likely to be appropriate for one working on labour rights with working children in South Asia where literacy rates are low, child labour remains non-unionised and where obvious activism may carry extreme personal risks. In other words, accountability procedures cannot be realistically expected to be uniform across a wide range of NGO activity. 29

Accordingly, I do not propose any detailed institutional blueprints here; instead, I outline the key institutional elements of authorization and accountability mechanisms at a more generalized level. Many elements of the mechanisms I outline here will be familiar to readers well-versed in literatures on formal representation and political ‘accountability’, and such readers should be able to digest the contents of my account with a fairly quick skim. However, I elaborate the functional components of authorization and accountability mechanisms in more detail than is common in these literatures. Moreover, I emphasize more strongly the particular characteristics of these mechanisms that are necessary to deliver democratic forms of legitimacy; this is signiﬁcant since these mechanisms (especially those of ‘accountability’) are sometimes given non-democratic justiﬁcations, and may diﬀer in certain institutional respects when designed to serve non-democratic normative objectives. In relation to this last point, it is worth saying a little more about how my analysis here is related to broader discussions—in public political discourse as well as in academic literature—about ‘accountability’ in global public decision-making in general, and accountability of NGOs in particular. 30 A substantial literature 29 Hugo Slim, By What Authority? The Legitimacy and Accountability of Non-Governmental Organisations (Journal of Humanitarian Assistance, 2002 [cited March 2004]); available from http://www. jha.ac/articles/a082.htm 30 See Michael Edwards and David Hulme, Non-Governmental Organisations’ Performance and Accountability: Beyond the Magic Bullet (London: Earthscan, 1995); Keohane and Nye, Democracy, Accountability and Global Governance (2008); Kovach, Neligan, and Burall, Global Accountability Report 1: Power without Accountability (2003); and Peter Raynard, Mapping Accountability in Humanitarian Assistance (ALNAP, 2000); available from http://www.alnap.org/pubs/html/ mappingacc.html#b19. Slim, By What Authority? The Legitimacy and Accountability of NonGovernmental Organisations (2002); N. Woods and A. Narlikar, ‘Governance and the Limits of Accountability: The WTO, the IMF, and the World Bank’, International Social Science Journal, 53/4 (2001); Ngaire Woods, ‘Making the IMF and the World Bank More Accountable’, International Aﬀairs, 77/1 (2001); and David Held and Mathias Koenig-Archibugi, Global Governance and Public Accountability (Oxford: Blackwell, 2005).

Theorizing Global Representative Agency

179

has developed in recent years on this subject of ‘accountability’, but much of it is concerned with accountability of non-democratic kinds. Although some writers emphasize the connections between democratic norms and accountability mechanisms, many express scepticism about the possibility of achieving democratic accountability in global politics and instead focus on other, non-democratic, forms of accountability. 31 My discussion here certainly relates to, and is subsumed within, this wider topic of accountability, but it is important to remember that here I am concerned only with its speciﬁcally democratic forms. ‘Accountability’, in general terms, is a broad concept incorporating myriad forms of ‘answerability’ or ‘responsibility’, which vary in accordance with a range of factors: the identity of the principals to whom an agent is accountable (such as donors, staﬀ members, states, public ombudsmen, and stakeholders); the substance of the regulative standards these principals impose (such as standards concerned with cost-eﬀectiveness, operational procedures, and substantive social impacts); and the character of the institutional mechanisms through which the accountability is achieved (ﬁnancial auditing, organizational restructuring, informal communication, etc.). In accordance with the important diﬀerences among forms of accountability, there can be a signiﬁcant variation between the normative purposes that are served by each, and that can be invoked to justify each. For example, ﬁnancial accountability to donors may be justiﬁed by reference to some normative commitment to eﬃciency, or to the bindingness of contracts, while social accountability to a designated ombudsman (with a given social agenda) may be justiﬁed by reference to quite diﬀerent normative ideals of social justice or utility. Here, I am concerned with the accountability of public political agents to stakeholders (and their authorization by stakeholders), in so far as this serves the democratic purpose outlined above: that of delivering control of the public political decisions to the stakeholders who are subject to them. Accordingly, the institutional mechanisms of authorization and accountability that I outline here are not intended to provide an exhaustive account of all forms of ‘accountability’ that may be desirable in global politics. Rather, they are intended only to explicate the narrower subset of these mechanisms that can serve as institutional vehicles for democratic representation. 32

31 See Edwards and Hulme, Non-Governmental Organisations’ Performance and Accountability: Beyond the Magic Bullet; Keohane and Nye, Democracy, Accountability and Global Governance (2008); Ruth W. Grant and Robert O. Keohane, ‘Accountability and Abuses of Power in World Politics’, American Political Science Review, 99/1 (2005); Robert O. Keohane, ‘The Concept of Accountability in World Politics and the Use of Force’, Michigan Journal of International Law, 24/4 (2003), for sceptical views of the prospects for democratic accountability in global politics in general, and the democratic accountability of NGOs in particular. 32 Whereas the normative property of political agents achieved through democratic mechanisms of authorization and accountability can be referred to as ‘representativeness’, the normative property of political agents achieved through non-democratic mechanisms of authorization and accountability can be characterized with the broader term ‘responsiveness’.

180

Theorizing Global Representative Agency

7.4.1. Mechanisms of Democratic Authorization The process of ‘authorization’ is best understood here as comprising two distinct elements. First, it involves mechanisms of delegation, for specifying the range of public political tasks that representatives are entitled to perform. Second, it involves mechanisms of empowerment, for according representatives the capacities to perform these tasks eﬀectively.

7.4.1.1. The Function of Public Delegation In the ﬁrst instance, authorization should be thought of as a process of delegation, through which groups of stakeholders transfer their collective decision-making entitlements to particular representative agents. Implicit in the idea of delegation is a speciﬁcation of the task for which responsibility is being transferred, so the nature of this task must be clearly negotiated between principals (stakeholders) and agents (representatives) as part of the delegatory process. Delegation mechanisms are thus a way of establishing a kind of ‘public contract’ between stakeholders and their political representatives, which speciﬁes the conditions on which stakeholders agree to delegate power and on which representatives accept responsibility to exercise it. In other words, this contracting process speciﬁes the substance of the democratic mandate that is conferred upon representatives by stakeholders and acknowledged by representatives as democratically binding. Within the framework of established democratic states, we generally think of representative agents’ (i.e. MPs’ or governments’) mandates in terms of quite speciﬁc policy-platforms, involving a range of key values and decisions that the representative is committed to pursuing within the conﬁnes of the institutional structures of the state. Within states, prior agreement on the constitutional structure of public decision-making is enshrined in key institutions of the state itself. Since the broad roles and responsibilities of representatives are already established through the structure of the state, it is only the more speciﬁc questions of policy that we usually think of as appropriate subjects for democratic mandates. However, broader ‘constitutional’ questions regarding representatives’ public roles and responsibilities can also be appropriate subjects for democratic mandates; certainly if we take a democratic view of the legitimate origins of constitutional structures (as I do here), then this is how we should view the matter. 33 In global politics, however, there are few established state-like constitutional structures delineating the appropriate general roles and responsibilities of agents such as NGOs within the global political order. For this reason, we must include many of these broader ‘constitutional’ responsibilities, along with more speciﬁc responsibilities for speciﬁc policy decisions, as appropriate subjects for the democratic mandates of these public political agents in global politics. 33 See Jon Elster and Rune Slagstad, Constitutionalism and Democracy (Cambridge: Cambridge University Press in collaboration with Maison des Sciences de l’Homme Paris, 1988), and Russell Hardin, Liberalism, Constitutionalism, and Democracy (Oxford: Oxford University Press, 1999), for discussions of various accounts of the legitimate origins of constitutional principles.

Theorizing Global Representative Agency

181

In most electoral systems, the process of delegation involves prospective representatives developing platforms and proposals for what they would do in power, in response to the views of citizens as perceived through various processes of citizen consultation and negotiation. 34 Some elements of this delegation process are initiated by the prospective representatives, who may solicit citizen input through channels such as political party structures, focus-groups, and other forms of public opinion polling. Other elements of this process are initiated by groups of citizens; various forms of lobbying and political activism by ‘interest groups’ and private individuals are examples of this. These processes do not generally involve legal formalization of speciﬁc electoral pledges or promises, but democratic convention dictates that there should be broad correlation between what is promised and what is done, such that citizens generally consider signiﬁcant deviation from such promises to be democratically illegitimate. Nonetheless, in electoral democracies the negotiation process between citizens and prospective representative agents is formalized to some degree (and some elements of it are accorded legal status) in two ways. First, and most importantly, the citizens’ end of the contracting process—the process through which they transfer their decision-making entitlements and thus ‘authorize’ representatives— is formalized through the casting of votes at an election. The key purpose of this formalization is to increase the likelihood that processes of delegation will succeed in being open to and inclusive of all citizens. 35 Second, the representatives’ end of the contracting process—the process through which representatives accept responsibility for their democratic role—is formalized through some kind of ‘swearing-in’ process, in which elected representatives make certain pledges to exercise their public power in the service of the democratic public. The purpose of this formalization is more symbolic and aﬃrmatory than substantive, in the sense that it seeks publicly and unambiguously to aﬃrm representatives’ commitments to public service that typically have already been made informally in the lead-up to the election. Although delegation is in a sense prior to representative decision-making, this does not mean that it must occur at a single point in time and confer upon representatives an eternal and irrevocable authorization. Rather, just as the process of representation is ongoing, so too can be processes of delegation. That is, delegatory ‘contracts’ can be revised as circumstances change, and as citizens’ demands upon their representatives change with them. Indeed, such delegatory ‘contracts’ can be revised by means of ongoing dialogue between citizens and public political agents throughout the tenure of political oﬃce.

34 At least, the latter—non-constitutional—kind of mandate is established in this way; as I have said, though, we can view the former kind of mandate as established in this way also, if we take a democratic view of the legitimate origins of constitutional structures. 35 As I have said, elections seek to ensure inclusiveness by collectivizing individual stakeholder signals through an aggregative mechanism for ‘social choice’. However, as argued in Chapters 5 and 6, deliberative mechanisms of inclusion are more appropriate for a ‘pluralist’ system of stakeholder representation of the kind that I advocate.

182

Theorizing Global Representative Agency

How, then, could delegatory processes, normatively equivalent to those achieved through elections, be established in global politics to create such mandates, and thus link agents of public power such as NGOs with the stakeholders they are obligated to represent? It may occasionally be possible to establish electoral mechanisms through which a stakeholder community could confer a mandate upon a non-state public political actor such as an NGO, but under many other circumstances this would be infeasible, for the general reasons outlined earlier. However, this need not pose a problem for the representative legitimacy of the actor in question, provided that the normative functions performed by the electoral process—that is, inclusion of stakeholders in the negotiation and imposition of a mandate, and public aﬃrmation of the responsibilities of representatives— can be achieved in other ways. These democratic purposes can in fact be achieved through non-electoral participatory processes that provide all stakeholders with an equal opportunity to contribute to the speciﬁcation of their representative agents’ mandates. For practical purposes, it is helpful to see delegation as comprised of two institutional elements, abstracted from electoral processes of delegation, which represent the two sides of the ‘contracting’ process. First, non-electoral delegation processes must incorporate mechanisms for facilitating stakeholder signals, representing the stakeholders’ side of the contracting process. Through mechanisms of ‘stakeholder signalling’, stakeholders could specify the tasks for which they are delegating entitlements to exercise public power, and the conditions that they are imposing upon the exercise of this power. 36 This would require participatory mechanisms through which stakeholders could communicate their preferences to their public political representatives in ways that articulate clear instructions. Just as participatory electoral mechanisms within states provide channels through which citizens can express preferences and deliver instructions to governmental representatives, non-electoral forms of delegation in global politics must similarly incorporate participatory processes as the mechanisms through which stakeholders can deliver their instructions to agents of public power. It is crucial to recognize here that any legitimate set of such mechanisms for stakeholder participation must incorporate some procedures for ensuring inclusiveness of all stakeholders in the formulation of a mandate. We cannot merely talk in general terms about the need for consultation; there must always be consultation that provides equal representation to all conﬂicting interests within the jurisdictional community. 37 Whether public political actors exercise their public power unilaterally or multilaterally, they will need to establish some such processes within their constituencies—either electoral processes or, more commonly, some kind of multi-stakeholder deliberative processes—to ensure that all conﬂicting 36 In Chapter 6, I explained how the multi-stakeholder model of interest representation can provide a coherent account of ‘public’ or collective choice of a kind that could be utilized here. 37 In Chapter 6, I discussed the particular sense in which individual stakeholders can be considered as ‘equals’ within a liberal-pluralist, multi-stakeholder, representative system.

Theorizing Global Representative Agency

183

interests are considered during the establishment of a mandate for their public power. In addition to mechanisms for ‘stakeholder signalling’, which correspond functionally to the electoral processes in which citizens cast votes at elections, the delegatory ‘contract’ established between public political agents and stakeholders has a second dimension involving the representatives’ side of the contracting process. Just as representatives in electoral systems must give some clear indications of the representative responsibilities that they are willing to accept (either formally through a ‘swearing-in’ process, or informally in the lead-up to elections), stakeholder representatives must also engage in some functionally equivalent processes, which we can refer to as public receptivity to stakeholder signals. Through mechanisms of public receptivity, representative agents must publicly acknowledge their receipt of and response to stakeholder signals. That is, they must make clear which of the various conditions that stakeholders seek to impose upon their exercise of public power they would be willing to accept. As Hugo Slim argues in relation to NGOs, each public political agent ‘must show clearly how the agency is responding to what it has learnt and what its stakeholders are telling it.’ 38 Stakeholders can then assess and respond further to the commitments of these prospective representatives, and determine whether or not to authorize them on this basis. Through such a process of negotiation, stakeholders can identify representative agents who will agree to exercise public power on acceptable conditions, and subsequently authorize these agents to act on their behalf. The democratic imperative for incorporating some such public receptivity by public political agents within the delegatory process arises because it is necessary for representatives to be clear about the nature of the tasks and decisions that they are willing and able to undertake, in order to provide a contractual basis for the ascription of responsibility to them for speciﬁc tasks and decisions. If speciﬁc mandates are to be considered binding upon an agent, there must ﬁrst be some contractual expression of willingness on the part of this agent to assume the responsibilities entailed in the execution of the mandate. This is not to say that a public political agent such as an NGO that has not accepted a speciﬁc mandate has no responsibilities to its stakeholders, and is thus free to take or leave any conditions that stakeholders seek to impose in accordance with its own independent wishes. Rather, as argued in Part I, an agent wielding public power over a group of stakeholders possesses general responsibilities to them by virtue of the autonomy-constraining impact of its activities upon these individuals. However, the responsibilities of such an agent towards its stakeholders that stem simply from its exercise of public power over them remain in an important sense unspeciﬁed in the absence of a delegatory contract. Without a contractually established mandate that speciﬁes the substantive content of the agent’s responsibilities, withdrawal of its public political impact, rather than compliance with the substance of the mandate, remains a legitimate route for it to take. If, 38 Slim, By What Authority? The Legitimacy and Accountability of Non-Governmental Organisations (2002).

184

Theorizing Global Representative Agency

for example, a population of stakeholders were to make some demand upon an NGO that it was unwilling or unable to fulﬁl, the NGO would be violating none of its general responsibilities (based on the rationales discussed in Part I of this book) if it were to refuse the request and cease its activities within the community entirely. 39 The speciﬁc substantive content of a public political agent’s responsibilities to take certain decisions on behalf of a stakeholder group, rather than simply to withdraw from engagement with it, must be derived in this case from the delegatory contract. The delegatory contract thus establishes the responsibilities that an agent must satisfy with respect to its stakeholders if it is to exercise public power over them with democratic legitimacy. An agent wielding public power over some group of individuals, then, must always choose between these two legitimate alternatives: either to accept and comply with a democratic mandate imposed by stakeholders; or to desist from exercising any forms of public power over the group in question. Through such a process of negotiation—involving both stakeholder signals and public receptivity—stakeholders can select representative agents who agree to exercise public power on conditions acceptable to these stakeholders, and authorize these agents to act accordingly on their behalf.

7.4.1.2. The Function of Public Empowerment Mechanisms Delegation mechanisms alone are not suﬃcient to constitute authorization; the other key component of authorization is public empowerment. Empowerment mechanisms are those through which the delegatory ‘contract’ between stakeholders and representatives is materially facilitated by an actual transfer of power to the delegated representatives, which enables these agents to engage in public political decision-making. As I argued in Chapter 3, there are two central kinds of capacities that can empower political agents to exercise public power. The ﬁrst of these involves capacities to exercise signiﬁcant inﬂuence over processes of regulative social norm-building, and the second involves capacities to impose material constraints—whether violence- or resource-based—upon individuals. Empowerment, then, can be achieved by transferring to delegated agents of public power (such as NGOs) whichever such capacities are required for them to perform the public political tasks for which they have been accorded responsibility by stakeholders. It is crucial to recognize here that such empowerment processes must always be contingent upon prior processes of democratic delegation if the authorization mechanism of which they are a part is to be democratic. It is not the existence of empowerment mechanisms per se, but rather their contingency upon processes 39 This is true, at least, so long as the accepted range of responsibilities underpinning the public power criterion in the context in question (in the manner discussed in Part I) incorporates some kind of distinction between ‘positive’ and ‘negative’ responsibilities, linking public power to the latter kind. Although, in principle, a theory of political responsibility may make no such distinction, I would suggest that as an empirical matter most accounts of responsibility with wide political currency do make some distinction of this kind.

Theorizing Global Representative Agency

185

of delegation from stakeholders, that ensures their democratic function. This is because the process of authorization would fail to deliver stakeholders suﬃcient political control over these agents for the process to attain representative democratic credentials, if public political agents were not empowered on the basis of their commitment to a mandate imposed by stakeholders. In most democratic states, the empowerment of representatives is contingent upon electoral processes of democratic delegation. Power must be delegated through electoral victory before governments are accorded control of public resources (communicative and material), before stakeholders can be obliged to obey their laws, and before state enforcement agencies can enforce these laws. The institutional apparatus of the state plays a central role here in coordinating the democratic empowerment process. Once an election has been won, the empowerment of the newly delegated representatives by stakeholders is easily achieved by handing over control of the relevant public political instruments (legislative, executive, and bureaucratic), which constitute the institutional apparatus of the state. Of course, a range of political actors must cooperate in enacting this empowerment process. Perhaps most importantly, participants in the various executive institutions of government (militaries, police forces, bureaucracies, etc.) must be committed to recognizing and enforcing the decisions of these representatives in order to facilitate their exercise of public power. To some extent, also, members of the wider jurisdictional community contribute to the empowerment of these elected representatives, since their willingness to accept and cooperate with the decisions of these elected leaders can be crucial in entrenching their public political capacities. The extent to which individual stakeholders (as distinct from these other agents of ‘the state’) contribute to the process of empowerment is not of normative signiﬁcance, however; it is not essential that stakeholders actually participate in the empowerment, so long as they have participated in the delegation process upon which the empowerment is contingent.

7.4.2. Mechanisms of Democratic Accountability The key purpose of accountability mechanisms, like authorization mechanisms, is to give stakeholders some control over the activities of public political agents such as NGOs. The main diﬀerence between them is that in the case of accountability mechanisms, the institutional channels for stakeholder input operate subsequent to, rather than prior to, the actual exercise of public power. We can best understand the process of ‘accountability’ by seeing it as constituted by two distinct elements, corresponding in some ways with the two elements of authorization outlined above. First, accountability involves mechanisms of transparency—that is, mechanisms that enable stakeholders to identify what representative agents are doing, and have previously done, with their public power. Second, accountability involves mechanisms of public disempowerment—that is, mechanisms for disabling public political agents, through an eﬀective sanction

186

Theorizing Global Representative Agency

imposed upon them at the instigation of stakeholders when they consider it appropriate, to minimize the agent’s capacity to continue exercising public power.

7.4.2.1. Public Political Transparency The central purpose of transparency mechanisms is to facilitate rigorous stakeholder evaluation of the performance of public political agents. The ﬁrst institutional characteristic that is necessary to facilitate this is transparent public role delineation—that is, clear allocation of public political power to speciﬁed agents. It is necessary for stakeholders to have some knowledge of what powers are wielded by which agents, and what each actor’s role is within an overall public political decision-making process, so that these stakeholders have some basis on which to allocate responsibility for those decisions. Responsibility must be clearly allocated here so that stakeholders can hold the appropriate agents accountable for particular decisions. As Peter Raynard argues, ‘[a]ccountability can fall at the ﬁrst hurdle if there is a lack of clarity in roles and responsibilities, or the attribution of them.’ 40 Thus, the function of mechanisms for transparent role delineation within accountability processes is somewhat related to the function of receptivity mechanisms within authorization processes. They are related in so far as both function to demarcate responsibilities for decisions that attach to particular agents of public power. They diﬀer, however, both in terms of the kind of responsibility that they demarcate, and in terms of the sense in which they can be said to ‘demarcate’ these responsibilities. As I explained in my earlier discussion of receptivity mechanisms, the contractual commitments that are entailed in the delegatory process serve to demarcate responsibilities for pursuing particular political purposes in the exercise of public power. Here, we can say that the delegatory process (in particular, the commitment to a mandate by an agent of public power) demarcates responsibilities for fulﬁlling this mandate in the sense that the contract provides the normative source, or justiﬁcation, of the responsibility. In contrast, the function of mechanisms of transparent role delineation is to demarcate responsibilities of public political agents of the other, general, kind that I distinguished in my earlier discussion. That is, the function of mechanisms of transparent role delineation is to demarcate the responsibilities that an agent of public power possesses by virtue of, and in accordance with, the scope of the public power it wields over a population of stakeholders. Since the normative source of this responsibility arises from the autonomy-constraining impact of this power, we cannot say that mechanisms of transparent role delineation demarcate the responsibility in the sense of providing its normative source, or justiﬁcation. Rather, these mechanisms demarcate the responsibility only in the sense that they publicly expose and clarify the forms of public power, and hence the scope of the general responsibilities, that particular political agents wield in a given context. 40

Raynard, Mapping Accountability in Humanitarian Assistance (2000).

Theorizing Global Representative Agency

187

In an ideally functioning democratic system, in which mechanisms of authorization and accountability operated harmoniously and complementarily, the theoretical distinction between these two forms of responsibility demarcation would not be an important one, since the two sets of responsibilities would coincide in practice. In such a system, no agent would exercise public power without having received prior authorization from stakeholders, and so the forms of public power allocated through the conferral of a mandate would transparently specify the full range of forms of public power that must be held accountable. In a wellfunctioning democratic state, for example, transparent role delineation of public oﬃcials is achieved both through mandates embedded in democratically sanctioned constitutions and other forms of political convention, and through more speciﬁc electorally established mandates, since these together demarcate the full range of forms of public power that are exercised within the state-bound society. 41 In global politics, however, mechanisms of authorization and accountability are not always so harmoniously aligned and coordinated, because the forms of nonelectoral delegation that I advocated in Section 7.4.2 (at both ‘constitutional’ and non-‘constitutional’ levels) are not always eﬀectively established. In the absence of clear delineation of public political roles through eﬀective mechanisms for public delegation, stakeholder control must instead be achieved through freestanding accountability mechanisms, which begin with some pre-democratically established forms of public power, and institute stakeholder control mechanisms that operate subsequent to the exercise of this power. That is, transparent public role delineation, as an element of an accountability process, must be achieved through mechanisms that can operate independently of authorization processes in some way. 42 The second institutional characteristic that is necessary to facilitate stakeholder evaluation of the performance of representatives is transparent public political action. 43 This is accepted as quite straightforward and standard in discussions of accountability, and is most commonly associated with the need to disclose both the outcomes of decision-making processes (the substance of the decisions that have been taken) and the means employed to enact them (sources of ﬁnancing, budget details, and so on). These aspects of transparent public political action are certainly important, but there are two additional aspects that are less commonly emphasized, and which, therefore, warrant special emphasis here. One of these commonly neglected aspects of transparent public political action is the requirement that, wherever possible, public political agents should give details not only of the outcomes of their decision-making processes, but also of the reasons for their decisions. This can often be achieved through measures such 41 The most serious problems with respect to this form of accountability tend to arise within democratic states in relation to bureaucratic power. See Behn, Rethinking Democratic Accountability, for a discussion of this problem in bureaucratic contexts. 42 I discuss the appropriate democratic relationship between mechanisms of authorization and accountability in more depth below. 43 Using more common but conceptually less precise terminology, we could call this ‘transparent governance’.

188

Theorizing Global Representative Agency

as provision of public access to the minutes of key decision-making meetings, details of evidence presented to these meetings, internal performance evaluation reports, and so on. 44 Explication of reasons for decisions enhances transparency by ensuring that the decision-making process (as distinct from its ﬁnal outcomes) is open to stakeholder scrutiny. Such transparency of decision-making processes is crucial to the facilitation of rigorous stakeholder evaluation of the performance of public political agents, since it can help stakeholders to ascertain the particular facts and arguments that are most relevant to a particular decision. Providing stakeholders with access to the information and arguments that have motivated a particular decision, so that they can themselves assess these reasons, makes it easier for them to engage critically in the deliberation over decision-making, and thus to evaluate eﬀectively the rectitude of decisions that have been taken. The other aspect of transparent public political action that warrants emphasis here is the requirement that access to information about the process and outcomes of decision-making should be available to stakeholders at an accessible cost. Accessing information about decision-making—that is, taking political advantage of transparency provisions—will inevitably come at some cost to stakeholders, if only the minimal time required to ﬁnd, absorb, and process the information. More commonly, accessing information about public political action will require further resources, including not only time but also some of the following: mobility (particularly if travel is required to reach sites where information is available, such as conferences, organizational headquarters, etc.); education and expertise (ranging from basic literacy through to more technical forms of knowledge); technology (especially communication technologies including computer and internet access); and money (to fund these other means of access, and to pay any direct fees associated with obtaining information). It is therefore imperative that public political agents try to minimize the costs to stakeholders of accessing the available information, to ensure that that transparency is achieved in practice and not just in principle. This poses particular diﬃculties in global politics because the degree of material deprivation experienced by billions of stakeholders worldwide makes it diﬃcult for them to access even those forms of information that come at relatively low costs, such as those requiring only basic literacy and time to read readily available materials.

7.4.2.2. Public Disempowerment The public disempowerment element of accountability mechanisms serves the same normative function as the public empowerment element of authorization mechanisms. The diﬀerence, of course, is that these public disempowerment mechanisms operate subsequent to rather than prior to the exercise of public 44 Although such transparency in decision-making processes has been sometimes neglected, certain writers have highlighted its importance. See, for example, Woods, ‘Making the IMF and the World Bank More Accountable’; Angela Wood, Structural Adjustment for the IMF: Options for Reforming the IMF’s Governance Structure (Reform Watch, 2001); available from http://www.reformwatch. net/ﬁtxers/120.pdf

Theorizing Global Representative Agency

189

power; they serve to disable, rather than to enable, the public power of political agents in accordance with the expressed wishes of stakeholders. Whereas public empowerment entails the conferral upon delegated representatives of certain capacities which enable them to perform their delegated tasks, public disempowerment entails the imposition of sanctions which eﬀectively annul these capacities and thus limit these agents’ ability to continue performing their public political activities. Just as in the case of public empowerment, it is not necessary for the power of public political agents to be annulled by stakeholders through a centralized public political apparatus, connected to a centralized (electoral) signalling process. In the absence of such a centralized public political apparatus at the disposal of stakeholders, the possibility of democratic accountability in global politics instead depends on two institutional features, which can operate in a decentralized fashion. First, there must be some kind of mechanisms through which stakeholders can de-legitimize public political agents through coordinated stakeholder signals. In an electoral system, an election loss provides the relevant signals, but alternative signals could, in principle, be employed here; these signalling devices must serve an equivalent normative function to those of delegation, but operate in the reverse fashion (revoking rather than conferring mandates). Second, there must be measures in place to ensure that those organizationally disparate actors who have the capacity to impose the sorts of sanctions that can eﬀectively annul the public power of representatives will wield this capacity contingent upon de-legitimizing signals by stakeholders. Here, as in the case of public empowerment mechanisms, it is important to recognize that it is not essential that all individual stakeholders actually participate in the imposition of sanctions entailed in public disempowerment. Just as the public disempowerment of politicians who have lost elections in democratic states can be achieved primarily by the withdrawal of support by various agencies of the state (bureaucracies, police forces, and so on) rather than by any direct action by the general citizenry, it is acceptable for actors other than stakeholders to impose disempowering sanctions upon agents of public power in global politics, so long as such sanctions are contingent upon delegation by stakeholders. All that matters for democratic purposes is that stakeholders participate in the delegation process upon which the process of public disempowerment is contingent. Moreover, just as the forms of empowerment appropriate to particular representatives will depend on the capacities required to perform their delegated tasks, the forms of sanction required as part of accountability mechanisms will similarly depend on the nature of the power exercised by the public political agent in question. In global politics, there are multiple public political actors exercising multiple forms of public power, which means that the kinds of sanctioning measures that will be required to annul the exercise of public power are likely to diﬀer considerably between cases. For instance, the sanctions required to annul the power of an NGO that exercises trust-based discursive power within a deliberative decision-making forum would be very diﬀerent from those required to annul the public power wielded by an NGO through the deployment of ﬁnancial resources.

190

Theorizing Global Representative Agency

We must not fall into the trap, then, of attempting to devise a universalized, or ‘one-size-ﬁts-all’ framework for democratic accountability, since the sanctioning measures appropriate in one instance may be excessive in a second instance and patently inadequate in a third. Devising concrete mechanisms of democratic accountability for global public power must always be a highly context-sensitive and context-speciﬁc endeavour.

7.4.3. The Relationship Between Authorization and Accountability Mechanisms So far I have outlined mechanisms of authorization and accountability independently of each other, but in this ﬁnal section it is worth giving further consideration to how, ideally, they would relate to one another within an institutionally integrated democratic order. Sometimes, authorization and accountability can each operate eﬀectively without the other, conferring democratic legitimacy upon public political agents. Typically, however, each is weak without the other and is limited in the depth of democratic legitimation that it can confer. Accordingly, the two forms of mechanism are best understood as mutually complementary. In electoral institutions within democratic states, authorization and accountability generally function together, in a complementary way. Regular elections provide the institutional framework for stakeholder delegation and sanction, while wider systems of law and political convention enact the outcomes of these processes, so as to empower and disempower representative agents in accordance with the expressed stakeholder will. In global politics, however, the absence of stable and centralized frameworks of public political power makes this more diﬃcult to achieve. First, it is worth highlighting the ways in which authorization mechanisms can be weak in the absence of accountability mechanisms, due to the weak forms of control that they allow stakeholders to exercise over representatives’ actions. Most straightforwardly, authorization mechanisms confer only weak forms of control in circumstances where it is diﬃcult for stakeholders to specify detailed conditions and tasks as part of the delegation process—either because the circumstances to which representatives must respond are unforeseeable, or because stakeholders lack the requisite knowledge and expertise to formulate speciﬁc mandates. In such situations, stakeholders may have to issue quite vague and expansive mandates to their delegated representatives, and in the absence of accountability mechanisms to hold these public political agents in check this may leave stakeholders with quite minimal forms of control. Also, and perhaps more importantly, the forms of control given to stakeholders by authorization mechanisms can be weak in situations where public political agents have independent or pre-existing power-bases, and hence do not rely on the ‘enabling’ function of empowerment mechanisms. This may be problematic, for example, in relation to those public political agents that are ﬁnancially selfsuﬃcient—such as the minority of NGOs that rely only on the resources provided

Theorizing Global Representative Agency

191

by a decision-making core membership—and that accordingly can wield public power independently of donor funding (at least in so far as their public power depends upon ﬁnancial resources, rather than upon further resources such as trust). 45 In such instances, the control exercised by stakeholders over their representatives must rely on their representatives willingly honouring the ‘contract’ established through the delegation process. Sometimes—when established public political agents have internalized democratic norms strongly enough to override their own interests—the imperative to honour this ‘contract’ will be suﬃcient to deliver the requisite control to stakeholders. However, when democratic norms are not suﬃciently strong to override the self-interested motivations of such agents, accountability mechanisms are required in addition to achieve this stakeholder control. Conversely, accountability mechanisms in the absence of authorization mechanisms can be weak in delivering political control to stakeholders in quite a diﬀerent way. Accountability alone is weak because it is unable to deliver positive forms of political control to stakeholders—that is, it is unable to deliver to them the capacity to direct the process of decision-making in a forward-looking, constructive, and creative way. Without the capacity to issue any kind of instructions to representatives prior to their authoritative action, stakeholders are denied the control entailed in political initiative. Rather, accountability mechanisms operating alone relegate stakeholders to the role of political reactionaries whose power is limited to the obstruction and frustration of the political initiatives and activities of others. The important democratic powers to create political options and set agendas, to formulate proposals and prospective policies, and more broadly to imagine and articulate new visions of political possibility, are all omitted from the purview of democratic accountability. Although accountability alone can eﬀectively deliver stakeholders’ ‘negative’ control over their representatives in a way that authorization cannot always do, its incapacity to deliver this ‘positive’ control is a signiﬁcant democratic weakness. Further, accountability mechanisms can be diﬃcult to establish under certain circumstances—particularly in relation to forms of public power that are newly established or that have an otherwise fragile and vulnerable basis. For instance, demanding mechanisms of accountability can be diﬃcult to achieve in organizations such as small grass-roots NGOs, which may lack resources or capacities for meeting these demands. In such cases, authorization alone can confer some democratic legitimacy, if the organization can reasonably claim to be acting within the constraints of a previously established democratic mandate from its stakeholders, performing delegated tasks in the service of stakeholders’ interests. In global politics, we should strive to establish both forms of mechanism in relation to each agent of public power. However, where this is not possible, either one alone can—if functioning eﬀectively—still generate some degree of democratic 45 In the next chapter, I discuss in greater depth the resources upon which NGOs can depend for their exercise of public power, and the implications of this for establishing mechanisms of public empowerment and disempowerment.

192

Theorizing Global Representative Agency

representativeness, albeit of a somewhat weaker form. In situations where the global community is seeking to constrain and limit the activities of already powerful actors, such as powerful states, corporations and NGOs, accountability mechanisms can play a primary role, and confer some representative legitimacy even in the absence of authorization. In situations when a community of stakeholders is seeking to establish power and enable public political agents to perform certain tasks and satisfy certain mandates, authorization can be of primary importance, conferring some representative legitimacy even in the absence of accountability. To achieve both together would require both stabilizing and strengthening the institutional framework of public power in global politics, so that there would be clear institutionalized agencies with which stakeholders could engage and establish eﬀective ongoing democratic linkages of both kinds. For now, though, we may sometimes have to accept the weaker forms of democratic legitimation that come with each alone.

7.5. CONCLUSIONS In this chapter, I have outlined a general model of democratic representation that has the potential to be applied within the global polity without a need for formal election mechanisms, and without a need for the establishment of centralized state-like structures of global public power. Whereas it is commonly taken as axiomatic that democratic representation requires elections—and that unelected agents, including NGOs, necessarily lack democratic legitimacy—I have argued here that this view of legitimate representative agency is too narrow. Elections are not intrinsically valuable; rather, I have argued that democrats rightly value them because of the particular forms of political control over the actions of public political agents that they can give stakeholders. Moreover, I have argued that we should recognize authorization and accountability—rather than elections per se—as the legitimate basis for representative agency, since elections are just one means of institutionalizing these more general functions. I have also sketched and provided rationales for the key institutional components that mechanisms of authorization and accountability must possess, if they are to satisfy standards of democratic legitimacy. If, then, we are seeking non-electoral mechanisms for establishing legitimate representative agency, we must devise non-electoral mechanisms of democratic authorization and accountability, in order to deliver stakeholders some control over the actions of their public political agents. In this chapter, I have outlined the general form and structure that such non-electoral mechanisms must take, by breaking down the functions of authorization and accountability into their constituent functional parts. In the following chapter, I examine some of the more concrete forms that such institutions could potentially take in practice, if they are to establish non-state actors such as NGOs as legitimate representative agents of stakeholder constituencies within contemporary global politics.

8 Instituting Global Representative Agency: The Authorization and Accountability of NGOs 8.1. INTRODUCTION In the previous chapter, I argued that democratically legitimate representative agency can be established not only through elections, but also through nonelectoral mechanisms of authorization and accountability. I also set out the broad institutional framework through which this could be achieved. In this chapter, I examine in greater depth how such non-electoral mechanisms might be able to function, in practice, in contemporary global politics, as a means through which powerful global actors could establish legitimate representative agency in relation to their respective stakeholder constituencies. As I have previously noted, non-electoral mechanisms of authorization and accountability could potentially be employed to confer democratic legitimacy upon a range of agents of public power in global politics, including IOs, TNCs, and unilateral state actors, as well as NGOs. Here, though, I draw only upon institutional practices surrounding NGOs, to provide my focal-point for exploring and illustrating some prospective avenues through which non-electoral representation might be instituted, in practice. I would like to be able to illustrate the workability of my model of stakeholder representation by presenting some examples of democratic authorization and accountability functioning fully in contemporary practice to confer legitimacy upon NGOs. At the present time, however, most practices of NGO authorization and accountability fall short of the democratic standards I set out in the previous chapter; Global Stakeholder Democracy is not yet up-andrunning in any real-world settings (at least, not any that I could ﬁnd in research for this book). However, there are some examples of NGO practice that have progressed important steps along an institutional path that could potentially lead towards the establishment of the kind of stakeholder representation I have in mind. Although I am unable to present illustrations of stakeholder representation in action in fully democratic forms, I instead present some speculative discussion of how such democratic representation could plausibly work in relation to NGOs, where possible combined with examples of existing NGO practices that embody some aspects of the non-electoral representative ideal. In

194

The Authorization and Accountability of NGOs

doing so, I hope to illustrate the kinds of directions for institutional reform that could bring us closer to realizing democratic stakeholder representation, in practice. In Section 8.1, I consider how non-electoral mechanisms of delegation and empowerment (which together constitute democratic authorization) could be achieved, in practice, and in doing so I present some examples of existing institutional practices that embody some elements of democratic authorization. In Section 8.2, I consider how non-electoral mechanisms of transparency and disempowerment (which together constitute democratic accountability) could be achieved, in practice, and in doing so present some examples of existing institutional practices that embody some elements of democratic accountability. In Section 8.3, I discuss how it may be possible to meet the challenge of ensuring equality for all stakeholder participants in these representative processes, in the absence of the formal equality embodied in an election. As I go through my examples of contemporary NGO practice in this chapter, I highlight the elements of existing mechanisms of NGO authorization and accountability that satisfy requirements of democratic representativeness, as well as those elements that fall short of democratic standards. Where possible, I raise suggestions about plausible directions for reform to bring the least democratic elements of these practices closer in line with the normative requirements of Global Stakeholder Democracy. In doing so, I do not go so far as to prescribe detailed blueprints for institutional reform. The ideas I present are only partial and preliminary, and are intended more to point towards productive directions for further empirical investigation and institutional experimentation, than to set out ﬁnal solutions. My ambitions here are limited to illustrating the broad reformist path that such institutional developments could take, and stimulating further discussion of reformist possibilities by practitioners and empirical scholars—who are better positioned than I to formulate detailed proposals for enacting the ideal of Stakeholder Democracy in particular political contexts.

8.2. THE DEMOCRATIC AUTHORIZATION OF NGO S IN GLOBAL POLITICS The ﬁrst set of mechanisms that an NGO can employ to accord its stakeholders some control over its public political decision-making are those of authorization, which can empower stakeholders to negotiate and establish a democratic mandate for the NGO. As I argued in the previous chapter, this involves both mechanisms of public political delegation by stakeholders and mechanisms of public political empowerment.

The Authorization and Accountability of NGOs

195

8.2.1. The Public Political Delegation of NGOs What kind of delegatory processes could be established in global politics to connect NGOs’ wielding public power with the stakeholders they are obligated to represent? First, we must be clear about the character of the mandates that must be imposed upon NGOs through the delegatory process. As I discussed earlier, the fact that background agreement on the constitutional structure of public decision-making is enshrined in key institutions of the state means that, within established democratic states, we generally think of representative agents’ (i.e. MPs’ or governments’) mandates in terms of quite speciﬁc policy-platforms. In global politics, however, there is a general absence of established state-like constitutional structures delineating the appropriate general roles and responsibilities of NGOs within the global political order. Consequently, we must include these broader ‘constitutional’ responsibilities, along with more speciﬁc responsibilities for speciﬁc policy decisions, as appropriate subjects for NGOs’ democratic mandates. Mandates of the former ‘constitutional’ kind can be accorded to NGOs—at least to some degree—through general ‘codes of conduct’, which can be applied to NGOs by stakeholders within a range of functional sectors, geographical areas, and so on. In certain sectors, various examples of this kind have been developed for NGOs in recent years. Within the humanitarian sector, most notably, such codes and protocols include the Sphere project, ALNAP, and the Humanitarian Accountability Project. 1 These standards have often been developed by states, UN agencies, and NGOs themselves with minimal stakeholder involvement, so may not provide examples of democratically established mandates. Moreover, these codes of conduct do not specify the public responsibilities of NGOs as comprehensively as would be democratically desirable, such that NGOs’ public responsibilities remain uncodiﬁed to a signiﬁcant extent. The existence of multiple sets of standards and codes, with overlapping jurisdictions and competing authority, further blurs the content of public expectations about the responsibilities that NGOs must discharge through their exercise of public power. However, these kinds of codes do embody some of the sorts of general standards that might be entailed in such mandates if they were to be developed in appropriate dialogue with stakeholders, and as such can give us some indication of the sort of reformist path that might be followed to develop more fully democratic mandates of this ‘constitutional’ kind. Mandates accorded to NGOs can also sometimes be of the other, more speciﬁc and policy-relevant, kind. Within multilateral decision-making structures such as UN institutions, for example, the broad ‘constitutional’ structure that delineates the general roles and responsibilities of participating NGOs are already established. The kind of mandate that such NGOs’ stakeholders may impose, then, 1 For more detailed information on these initiatives, see http://www.sphereproject.org http://www.alnap.org/ and http://www.hapinternational.org/en (all accessed January 2008).

196

The Authorization and Accountability of NGOs

is necessarily limited to the more speciﬁc questions of policy being determined within the framework of these institutions. As I established in the previous chapter, the ﬁrst kind of institutional mechanisms that are required to enable stakeholders to impose mandates of these kinds are stakeholder signalling mechanisms, through which stakeholders can specify the tasks for which they are delegating entitlements to exercise public power, and the conditions that they are imposing upon the exercise of this power. There is a limit to how far we can generalize on the question of what forms of participation will be appropriate for an NGO to establish, in practice, to constitute such stakeholder signalling mechanisms, since these will vary in diﬀerent contexts, depending on the informational and communicative resources, social infrastructures, educational attainments, and so on of the prospective stakeholder populations. Clearly, it is important for an NGO to take into account these factors in formulating participatory mechanisms in each particular context. Moreover, it is crucial that stakeholders themselves are involved in determining what forms of consultation may be suitable to them. As argued in a report by AccountAbility1000, it is essential to [i]nvolve stakeholders in deﬁning the terms of the engagement [with agents of public power]. The terms will include, but are not limited to the issues covered, the methods and techniques of engagement used, the questions asked, the means of analyzing responses to questions and the stakeholder feedback process. 2

Although generalized blueprints for such methods of engagement are not appropriate, it is possible, at least, to illustrate some possible directions for democratic reform by examining forms of stakeholder participation (albeit not fully democratic) that have been employed by NGOs. Existing forms of stakeholder participation do not generally involve participation of stakeholder groups that have been delineated in accordance with the principles advocated in this book; more commonly they involve wider groups of so-called ‘stakeholders’ that include individuals with broader (non-public) interests in the NGO (such as some staﬀ and donors). Nonetheless, these forms of ‘stakeholder’ involvement embody some aspects of the participatory mechanisms that could constitute processes of democratic delegation, and as such can point us down some promising paths towards institutional reform. One way in which many development NGOs have incorporated processes for consulting stakeholders within public decision-making structures is through the use of participatory methods commonly referred to with terms such as Participatory Rural Appraisal (PRA) and Participatory Learning and Action (PLA) processes. These participatory methods have been devised as a means of garnering stakeholder views on the appropriate policies for NGOs to adopt in their exercise of their public power within the stakeholder communities, and ultimately also for empowering stakeholders by giving them some degree of control over 2 AccountAbility, Aa1000 Overview (1999 [cited March 2005]); available from http://www. accountability.org.uk/uploadstore/cms/docs/AA1000%20Overview.pdf

The Authorization and Accountability of NGOs

197

the outcomes of decisions. 3 As characterized by the World Bank, these methods involve ‘a growing family of participatory approaches and methods that emphasize local knowledge and enable local people to make their own appraisal, analysis, and plans.’ 4 These decision-making methods are employed by many NGOs in urban as well as rural development contexts, and can incorporate a range of procedures and tools. They generally involve a series of open meetings with stakeholders, in which context-appropriate procedures are employed to facilitate dialogue among stakeholders and NGOs, and development of joint policy plans. In practice, these procedures have commonly included methods such as semistructured interviewing, focus group discussions, preference-ranking, and visual and spatial mapping of stakeholders’ social and historical experiences, as well as of stakeholders’ conceptions of, and preferences regarding, policy alternatives. 5 A range of similar mechanisms have been advocated and utilized at various times by many NGOs and commentators. AccountAbility1000—a research and advocacy institute—has outlined a range of mechanisms that it claims should be employed to engage stakeholders in decision-making. In its Global Accountability Report, the institute proposes guidelines for participatory processes involving ‘representative’ samples from stakeholder communities, including: one-to-one interviews (both face-to-face and distance); group interviews; focus groups; workshops and seminars; public meetings; and questionnaires carried out via face-toface discussion, letter, telephone, Internet or other techniques. 6 Some NGOs have made some moves towards actually incorporating a few such participatory practices in their decision-making procedures. Oxfam (Great Britain), for example, recently introduced reforms to ensure that the NGO’s trustees—members of its governing body, or ‘council’—take decisions that are ‘inﬂuenced by the views of its stakeholders’. 7 The methods employed to achieve this outcome are twofold. First, Oxfam established a stakeholder ‘assembly’, comprised of around 200 individuals deemed to be representative of the wider stakeholder community, to deliberate upon key policy issues. Second, it undertakes regular ‘Stakeholder Surveys’, which provide another channel through which stakeholder needs and opinions can be identiﬁed. 8 3 For discussions of such processes, see Institute of Development Studies, The Power of Participation: PRA and Policy (IDS Policy Brieﬁng, Issue 7, 1996 [cited February 2005]); available from http://www.ids.ac.uk/ids/bookshop/briefs/brief7.html Institute of Development Studies, The New Dynamics of Aid: Power, Procedures and Relationships (IDS Policy Brieﬁng, Issue 15, 2001 [cited February 2005]); available from http://www.ids.ac.uk/ids/bookshop/briefs/brief15.html World Bank, Participation and Civic Engagement in Poverty Reduction Strategy (2005 [cited February 2005]); available from http://www.worldbank.org/participation/PRSP/alp.htm World Bank, World Bank Participation Sourcebook (2005 [cited February 2005]); available from http://www.worldbank.org/wbi/sourcebook/sba104.htm 4 World Bank, World Bank Participation Sourcebook (2005). 5 Institute of Development Studies, The Power of Participation: PRA and Policy (1996). 6 AccountAbility, Aa1000 Overview (2003). 7 Oxfam, Organisation of Oxfam Great Britain (2005 [cited March 2005]); available from http://www.oxfam.org.uk/about_us/organisation.htm 8 Ibid.

198

The Authorization and Accountability of NGOs

In many political contexts, establishing processes that are suﬃciently inclusive of all stakeholders is likely to pose signiﬁcant institutional challenges, as a result of both resource disparities and other social power relations within stakeholder communities. As Alan Fowler argues, ‘creating a systematic way of engaging multiple stakeholders is unlikely to be free of conﬂict and risk, especially where diﬀerent levels of authority and power exist.’ 9 Such problems can to some degree be addressed through investment, on the part of NGOs and other agents of public power, in ‘capacity-building’ within stakeholder communities. As Fowler argues, ‘the process of stakeholder engagement should help to build the capacity of the parties involved, particularly the poor and marginalized whose ability to put forward their interests is most likely to be constrained.’ 10 Some commentators have further argued that overcoming such problems of inclusion can be assisted through the spread of communication technologies, claiming that ‘as access to technology grows, communities feel they can speak for themselves through video and the internet as well as more traditional arenas such as marches and demonstrations’. 11 Despite optimistic assessments such as these, there are, nonetheless, certain to be some political contexts in which NGOs exercise public power where it would be unrealistic to hope for signiﬁcantly inclusive forms of stakeholder consultation— notably humanitarian activity in crisis situations. Mukesh Kapila, head of the Conﬂict and Humanitarian Aﬀairs Department in the Department of International Development in London, rejects attempts to employ participatory processes in crisis situations, on the grounds that this would be incompatible with the more immediate goal of saving lives. He argues (somewhat rhetorically), that . . . a person stuck at the bottom ﬂoor of a building that has collapsed wants to get out of that building quickly. The last thing that person wants is a participatory, evaluative approach to the rescue. . . . We are insulting our beneﬁciaries by oﬀering them something that is, ﬁrstly, not deliverable and secondly, probably not actually relevant in the immediate circumstances. . . . There are less acute circumstances where there could be greater participation. I believe in all that. However, I think we must be very, very clear about what can be delivered and what cannot be delivered and not waste that time trying to deliver something, which could turn out to be a pie in the sky. 12

In crisis situations in which inclusive stakeholder participation in decisionmaking is an unrealistic goal, Kapila is right to acknowledge that the prospects for political delegation, and hence representative legitimation, are slim. It is important to recognize, though, that the democratic imperative for stakeholder 9 Alan Fowler, ‘Assessing NGO Performance: Diﬃculties, Dilemmas and a Way Ahead’, in Michael Edwards and David Hulme (eds.), Non-Governmental Organisations Performance and Accountability: Beyond the Magic Bullet (London: Earthscan, 1995), 151. 10 Ibid. 11 Michael Edwards, David Hulme, and Tina Wallace, ‘NGOs in a Global Future: Marrying Local Delivery to Worldwide Leverage’ (paper presented at the NGOs in a Global Future Conference, Birmingham, 1999). 12 Mukesh Kapila, ‘Incentives for Improved Accountability’, Refugee Survey Quarterly, 21/3 (2002), 242.

The Authorization and Accountability of NGOs

199

consultation should always be retained as an ideal and, as crisis situations stabilize, attempts should be undertaken to establish delegatory processes as quickly and eﬀectively as possible. The examples of stakeholder participation I have presented here do not provide illustrations of fully democratic stakeholder signalling mechanisms. Existing participatory mechanisms of these kinds are not adequately inclusive of stakeholder interests; moreover, they are not adequately deliberative in structure, such that the stakeholder preferences expressed through these forms of consultation do not generate appropriately democratic collective decisions within the relevant stakeholder communities. However, participatory mechanisms of these kinds represent some steps along the kind of reformist path that could lead to democratic forms of stakeholder delegation. In thinking about the best ways to develop these kinds of consultative mechanisms in more democratic directions, there is much room for creativity, on the part of both NGO practitioners and members of stakeholder communities. There is no reason to think that these mechanisms should not include electoral processes, in the limited circumstances where their establishment may be both normatively appropriate and practically feasible. 13 However, many of the necessary institutional innovations will need to involve the construction of appropriate forms of deliberative democratic decision-making within stakeholder communities. Some recent scholarly work and practical experimentation relating to methods for institutionalizing deliberative democracy in general—through mechanisms such as ‘deliberative polling’, ‘citizen juries’, and so on—could also provide some very helpful theoretical and institutional tools for developing appropriate deliberative mechanisms within stakeholder communities. 14 The kinds of deliberative mechanisms explored in this recent work have mostly been devised for application within national and local democratic communities rather than global or transnational stakeholder communities, and would no doubt need to be adapted in various ways to the particular circumstances in varying political contexts; nonetheless, I think scholars of deliberative democracy would have much to contribute to a project of Global Stakeholder Democracy through ideas and initiatives of these kinds. As I established in the previous chapter, the second set of mechanisms that are required to establish eﬀective stakeholder delegation are mechanisms of public receptivity, through which prospective NGO representatives clearly indicate which, of the various conditions that stakeholders seek to impose upon their 13 I discussed the considerations that would constrain their appropriateness and feasibility in the previous chapter. 14 See, for example, Archon Fung and Erik Olin Wright (eds.), Deepening Democracy: Institutional Innovations in Empowered Participatory Governance (London: Verso, 2003); John Gastil and Peter Levine (eds.), The Deliberative Democracy Handbook: Strategies for Eﬀective Civic Engagement in the Twenty-First Century (San Francisco: Jossey-Bass, 2005); James S. Fishkin and Robert C. Luskin, ‘Experimenting with a Democratic Ideal: Deliberative Polling and Public Opinion’, Acta Politica, 40/3 (2005); James S. Fishkin, The Voice of the People: Public Opinion and Democracy (New Haven, CT: Yale University Press, 1995); James S Fishkin, Democracy and Deliberation: New Directions for Democratic Reform (New Haven, CT: Yale University Press, 1991); and Robert E. Goodin, ‘Sequencing Deliberative Moments’, Acta Politica, 40/2 (2005).

200

The Authorization and Accountability of NGOs

subsequent public power, they would be willing to accept. In my earlier discussion of public receptivity, I identiﬁed the source of the democratic imperative for public receptivity as the need to establish a contractual basis for the ascription of responsibility to them for speciﬁc tasks and decisions. Once we turn to look at the practical enactment of delegatory mechanisms, we can recognize that a further imperative for incorporating some such public receptivity by NGOs within the delegatory process arises for more pragmatic reasons. Receptivity is important not only to ensure that NGOs clearly accept or reject the substantive mandates issued unilaterally by stakeholders, but also to provide NGOs with some scope for negotiation with stakeholders, to achieve mandates that both parties are willing and able to endorse. Although, in principle, a stakeholder community would be entitled to demand that an NGO either comply with its unilaterally formulated demands or else desist from its public political activity, in practice stakeholder groups may value some of the functions performed by the NGO and be reluctant to require its withdrawal. Consequently, stakeholders will commonly be willing to engage in a negotiation process with an NGO, through which a mandate can be established that is acceptable to the NGO as well as to its stakeholders. In this way, NGOs can, through their substantive responses to stakeholder demands, challenge and revise those aspects of stakeholder demands which they are unable or unwilling to satisfy. There are pragmatic advantages to this negotiation process, since the exercise of public power by NGOs is commonly subject to constraints that stakeholders may not, otherwise, understand or appreciate. For instance, in most political contexts NGOs are constrained by their limited ﬁnancial resources for serving stakeholder communities; as a result, it is essential for NGOs clearly to explicate, in a negotiation process with stakeholders over the content of their mandate, the material limits upon what they can be expected to accomplish. As Norman Uphoﬀ observes, NGOs have the burden of mobilizing and managing funds which will keep their operation solvent. If beneﬁciaries could decide how much they would receive and when, such organizations would likely be bankrupt within a year. 15

NGOs also commonly confront other (non-ﬁnancial) constraints, arising from the need to navigate a range of social and political factors that are beyond the control of the NGO. Thomas Linde, from the humanitarian NGO Médecins Sans Frontières, argues that ‘humanitarian work often consists, to a considerable extent, in negotiating with the “beneﬁciaries” on what the . . . [organization] can and cannot do for them, in spite of their insistent demands.’ 16 And even in the absence of external constraints on NGOs’ actions, challenges to stakeholders’ demands will sometimes arise simply from diﬀerences of judgement between beneﬁciaries 15 Norman Uphoﬀ, ‘Why NGOs are not a Third Sector: A Sectoral Analysis with Some Thoughts on Accountability, Sustainability and Evaluation’, in Edwards and Hulme (eds.), Non-Governmental Organisations Performance and Accountability: Beyond the Magic Bullet, 20–1. 16 Thomas Linde, ‘Accountability and Independence in Humanitarian Action’, Refugee Survey Quarterly, 21/3 (2002), 326.

The Authorization and Accountability of NGOs

201

and NGOs. As argued in a report by the International Council on Human Rights Policy, ‘[m]any NGOs have received advice from local activists or people experiencing abuse that seems foolhardy and unlikely to be eﬀective. There may be quite legitimate diﬀerences of judgement.’ 17 In situations such as these, then, mechanisms of receptivity are essential to ensure that NGOs and stakeholders are able to reach agreement on democratic mandates that are both consistent with the wishes of the stakeholders, and acceptable to the NGO agents who will be responsible for carrying them out.

8.2.2. The Public Empowerment of NGOs In the previous chapter, I established that delegation mechanisms alone are not suﬃcient to constitute authorization; the other key component of authorization is public empowerment. What is crucial in legitimate public empowerment mechanisms is a link between delegation processes and the transfer to representatives of the resources and capacities necessary for the exercise of public power; stakeholders themselves need not be the agents to transfer these resources and capacities, but it is crucial that whichever agents make these transfers do so in accordance with the mandates established by stakeholders. As Alnoor Ebrahim argues, participatory processes of the kind discussed above can contribute to the empowerment of stakeholders, but these processes alone do not deliver stakeholders suﬃcient control over key NGO decisions. He acknowledges that ‘tools such as participatory appraisal and asset mapping . . . can, at least in part, reverse or moderate conventional relations of authority and power.’ 18 However, he further emphasizes that ‘[a]ctual sharing of power . . . would require . . . [going] even further—not only by requiring dialogue and open access to all project-related information,’ but also by according stakeholders some control over the ultimate decisive phases of the policy-making process. 19 A link between mechanisms of delegation and those of public empowerment is, for this reason, essential in order to ensure that consultation with stakeholders is not merely cosmetic, but rather serves as a democratic instrument of stakeholder control. As I established in the previous chapter, public empowerment in democratic states is generally performed by agents of the state—such as armies, police-forces, and bureaucrats—which empower particular agents on the basis of their success in election processes. In global politics, however, there is no such centralized apparatus for making and enforcing public rules, or for mobilizing public resources, as means for empowering agents that have been accorded mandates 17 International Council on Human Rights Policy, Deserving Trust: Issues of Accountability for Human Rights NGOs (2003 [cited February 2005]); available from http://www.ichrp.org/ac/ excerpts/136.doc 18 Alnoor Ebrahim, ‘Accountability in Practice: Mechanisms for NGOs’, World Development, 31/5 (2003), 819. 19 Ibid.

202

The Authorization and Accountability of NGOs

by their stakeholders. As I discussed in Chapter 1, there are few, and generally weak, centralized mechanisms for enforcing public political rules in global politics. Moreover, there are few centralized ‘public’ resources to be allocated to delegated representatives, since there is no global taxation system and are few functionally equivalent means of raising public global funds, and there are also few centralized mechanisms for regulating the use of force and delivering physical security. Instead, public political actors in global politics—including NGOs—are empowered to mobilize resources and to secure compliance with certain rules by a range of organizationally disparate actors (individual states, International Organizations, donor agencies, corporations, etc.) rather than by any centralized ‘state-like’ institutional apparatus. In order to examine how NGOs could achieve democratic public empowerment in global politics—that is, in order to examine how they could be publicly empowered in a way that is contingent upon delegation by stakeholders—we must therefore consider the prospects for two kinds of processes that together would serve an equivalent normative function to that of governmental inauguration (public empowerment) within a state. First, it is necessary to identify the range of mechanisms through which public power can be conferred upon NGOs; and second it is necessary to consider how these empowerment mechanisms could be adequately connected to prior processes of delegation by stakeholders.

8.2.2.1. Instituting Non-Electoral Public Empowerment in Global Politics 1: Mechanisms for NGO Empowerment In Chapter 3, I outlined some key forms of public power that NGOs can exercise in global politics, based both on their capacities to wield inﬂuence in norm-building processes and on their capacities to impact materially on populations through the mobilization and deployment of resources or through complicity in the use of force. Recognition of these forms of public power is helpful in the identiﬁcation of empowerment mechanisms, in so far as these forms of public power constitute the outcomes of public empowerment processes (since public empowerment, by deﬁnition, equips agents to exercise public power). Here, however, I am interested in the diﬀerent, although related, question of the processes by which NGOs came to be conferred with these various forms of public power. That is, I am concerned here with the ways in which other political agents in global politics can act to enable (or refrain from enabling) NGOs’ exercise of these forms of public power in the ﬁrst place. The point I want to highlight here is that NGOs are not entirely self-reliant in sustaining their capacities for public power, but rather depend for these capacities on active support from other political agents. Although NGOs can be wielders of public power in global politics, their public political capacities can be vulnerable to the decisions of other agents about whether or not to provide the requisite support. I will outline these various vulnerabilities and then go on to suggest that, if they are connected in the right ways to processes of stakeholder delegation, these vulnerabilities have the potential to provide leverage through which

The Authorization and Accountability of NGOs

203

stakeholders can exercise some control over the decision-making of these NGOs. To demonstrate these vulnerabilities, it is helpful to examine each of the forms of public power I outlined in Chapter 3, and highlight the ways in which these capacities commonly depend upon the support of other political actors in global politics. As in Chapter 3, this discussion must be based on the recognition that a thorough analysis of the sources and mechanisms of NGOs’ public power would require a rigorous empirical investigation, of the kind that is beyond the scope of this theoretical project. My intention here is thus the more limited one of outlining some plausible possibilities, with a view to highlighting some potential mechanisms for NGO empowerment. The ﬁrst form of public power that NGOs commonly wield in global politics is that exercised through dominance within processes for the production of key global norms—including both formal laws norms generated within the UN system, and more informal norms generated through other regulatory mechanisms. To some degree, the capacity of NGOs to exercise this form of public power is dependent upon various material resources (funds for accessing communications technologies, transportation, production of advocacy materials, payment and training of employees, etc.). In this respect, this NGO capacity can be considered alongside other forms of public power that are dependent on deploying material resources, which I discuss below. However, although some material resources are usually necessary to enable NGOs to exercise this form of public power, they are generally not, in themselves, suﬃcient to ensure NGOs’ eﬀectiveness in inﬂuencing norm-building processes. In most situations, NGOs seeking to exercise public power through normbuilding will further depend in critical ways upon the resource of political trust, which is invested in them by other actors in the norm-building process. Alternatively, rather than using the language of trust, we could conceptualize the resource that NGOs require here as credibility, or good reputation—both of which are properties conferred upon an NGO through the trusting behaviours of others. Trust can be deﬁned in a variety of ways, but for the present purposes it can be understood simply as a disposition of conﬁdence in some aspect of an agent’s character or ability, or in the likelihood that the agent will perform some task. 20 In order for an NGO to exercise inﬂuence in norm-building processes, it must attain quite speciﬁc forms of trust from other political agents—trust in its good character and in its competence in the task of public regulation. Such trust is necessary for an NGO to exercise inﬂuence in norm-building, since it is this that leads other relevant actors—states, corporations, International Organizations, and wider jurisdictional populations—to accept the regulatory demands of the NGO as binding in any sense. As Joseph Nye argues, the kind of ‘soft power’ that is 20 Jessica Miller deﬁnes trust as ‘a feeling of security in the benevolence and competence of the trustee in a particular sphere of interaction’. See Jessica Miller, ‘Trust: The Moral Importance of an Emotional Attitude’, Practical Philosophy, 3/3 (2000). Diego Gambetta deﬁnes trust as ‘a particular level of the subjective probability with which an agent will perform a particular action. . . . ’. See Diego Gambetta, ‘Can We Trust Trust?’, in Diego Gambetta (ed.), Trust: Making and Breaking Cooperative Relations (Oxford: Basil Blackwell, 1988).

204

The Authorization and Accountability of NGOs

entailed in the norm-building process in global politics depends strongly on these resources of trust, credibility, and reputation, since these provide them with the crucial beneﬁts of public attention and normative recognition. In contemporary global politics, he argues, ‘credibility is the crucial resource and an important source of soft power’. 21 Accordingly, the increased inﬂuence of NGOs in global regulatory processes in recent years is commonly linked by analysts and commentators to the rise in levels of trust invested in these NGOs by other social actors. 22 According to research conducted by PR ﬁrm Edelman among opinion leaders in Europe, the US, Canada, China, Brazil and Japan, trust in NGOs ‘to do what is right’ has risen over the past ﬁve years, and NGOs are now the most trusted institutions (compared with government, business, and media) in this group of countries. 23 Another recent poll conducted by research organizations GlobeScan and the Program on International Policy Attitudes similarly found NGOs to be the most trusted organizations; worldwide, NGOs were trusted by 65 per cent of those polled. 24 Recognition of the role of such trust in enabling NGOs to exercise public power through norm-building is critical to the understanding of the relevant public empowerment mechanisms, since it is reliance on this trust which ensures that NGOs cannot exercise this form of public power until they have gained the trust of other political actors. It is, of course, clear that NGOs must depend upon others for this critical resource in norm-building, since trust is, by deﬁnition, something that cannot be acquired or wielded unilaterally. Rather, it is a kind of social resource that depends, for its deployment, upon the willingness of others to act in a particular (trusting) way towards the bearer of the trust. In recognition of the vulnerability associated with dependence upon this trust, many of the UK-based NGOs in a recent survey cited potential damage to their own reputation, or to the reputation of other NGOs in the sector, as a factor imposing ‘some constraint’ on their activities. 25 In this way, then, we can see that the conferral of trust is a 21 Joseph S. Nye, Public Diplomacy in the 21st Century (The Globalist, 2004 [cited March 2005]); available from http://www.theglobalist.com/DBWeb/StoryId.aspx?StoryId=3885 See also Joseph S. Nye, Soft Power: The Means to Success in World Politics (New York: Public Aﬀairs, 2004). 22 Although this point is commonly made by political practitioners and analysts, there has not been a great deal of academic analysis of this subject. For some analysis of the key academic discussions pertaining to this subject—and the limitations of this analysis—see Helmut Anheier and Jeremy Kendall, Trust and Voluntary Organisations: Three Theoretical Approaches (Civil Society Working Paper 5, 2000 [cited February 2005]); available from http://www.lse.ac.uk/collections/CCS/pdf/CSWP5webversion.pdf 23 The relative levels of trust in these various types of institutions varied somewhat by country, but NGOs were the most trusted institutions overall. See Edelman PR Worldwide, Edelman Trust Barometer 2005: The Sixth Global Opinion Leaders Study (January 2005 [cited March 2005]); available from http://edelman.com/image/insights/content/Edelman_Trust_Barometer-2005_ﬁnal_ﬁnal.pdf 24 Poll Finds High Level of Trust in UN and NGOs; Continued Discontent with US (2004 [cited February 2005]); available from http://www.globalsolutions.org/programs/glob_engage/news/ globescan_june04.html 25 See Ashridge Centre for Business and Society and National Council for Voluntary Organisations (Great Britain), Survey of Large UK-Based Non-Governmental Organisations and Voluntary

The Authorization and Accountability of NGOs

205

key mechanism through which NGOs can be empowered to exercise public power through global norm-building. The second form of public power that NGOs can wield in global politics is that exercised through the imposition of material constraints upon individuals— either through contribution to the use of force or violence, or through contribution to the deprivation of crucial social resources. Just as public power exercised through norm-building can depend in part on the provision of material resources, so too public power exercised through the imposition of material constraints can sometimes depend in part on the conferral of trust. For example, the trust of a local population may be important in establishing sufﬁciently cooperative relationships within the community for a development or humanitarian NGO to engage eﬀectively in operations there. As Serge Duss of World Vision argues, ‘[f]or non-governmental organizations like World Vision to work eﬀectively in post-conﬂict situations, we must establish a close and trusting relationship with the communities we serve.’ 26 Notwithstanding this, probably the most important resources that NGOs require in order to exercise public power of this second kind are themselves material in character, and NGOs generally depend heavily on other political actors (states, International Organizations, private donors or corporations, etc.) to provide them with these resources. Perhaps most importantly, NGOs are usually highly dependent on donor funding for nearly all of their activities—especially those that involve mobilizing and deploying resources (development assistance, humanitarian aid, and so on) within other communities. In Chapter 3, I presented some ﬁgures on the sums of money expended by NGOs in the exercise of public power; here, it is important to highlight the sources of these funds, upon which NGOs’ exercise of public power is dependent. Although there may be some examples of small NGOs that are largely self-funded (in the sense that the decision-makers within the organization also provide the bulk of the funds), the overwhelming majority of NGOs are heavily dependent on the provision of funds from other actors. The breakdown of funding sources for NGOs varies somewhat across functional sectors, and across geographical locations, but most NGOs depend upon some combination of three funding sources: fees (from sources such as memberships, subscriptions, service-charges, etc.), donations from governments of states, and donations from private philanthropists. 27 Since many forms of public power exercised by NGOs are dependent in this way on the provision Organisations (2001 [cited March 2005]); available from http://www.hillandknowlton.co.uk/ publications/ncvorep.doc 26 See Committee on Government Reform Subcommittee on National Security, Emerging Threats, and International Relations, United States House of Representatives, Statement of Serge Duss, Director Public Policy and Advocacy, World Vision, 18 July 2003, for a discussion of the importance of trust in fostering the conditions for eﬀective humanitarian operations. 27 Lester M. Salamon, S. Wojciech Sokolowski, and Regina List, Global Civil Society: An Overview (Baltimore, MD: Center for Civil Society Studies Institute for Policy Studies, The Johns Hopkins University, 2003), 28–33.

206

The Authorization and Accountability of NGOs

of resources by other political actors, we can straightforwardly identify this external funding as a key mechanism through which NGOs are accorded public empowerment. In addition to dependence upon external funding sources, NGOs can also sometimes depend upon the provision by others of the conditions necessary to maintain their physical security, as a prerequisite for conducting their operational activities in development and humanitarian contexts. In circumstances where one actor (usually a state government, but sometimes other actors such as international administrations or non-state militias) exercises dominant control over the use of force within in a territory, the capacity of an NGO to exercise public power through imposing material constraints will require at a minimum the permission of this actor to carry out its work. In many circumstances, NGOs are denied this permission, or granted it only conditionally. 28 In other circumstances—in particular where a single agent (such as a state government, international administration, or non-state militia) does not exercise adequate control over the use of force within the territory—permission will not be enough to empower NGOs to operate, and they will further require more direct physical protection and support, to secure them against violence or resistance by third parties. Under some circumstances, NGOs can hire their own security personnel, and occasionally do this. 29 Such practices are limited, however, and to a large extent NGOs are dependent on others—military forces or externally provided security personnel—to provide conditions that are suﬃciently secure to enable them to work. NGOs are generally reluctant to cooperate too closely with such external military and security personnel, since this can be seen to undermine their independence and neutrality; nonetheless, NGOs do sometimes need to resort to such protection in conﬂict situations. 30 The provision of requisite physical security, then, must be recognized as a further means through which NGOs can be empowered by other political agents, to confer upon them their capacity to exercise public power over populations.

28 See Refugees International, Humanitarian Access to Darfur Limited, Despite Sudan’s Claims (2004); available from http://www.globalpolicy.org/ngos/aid/2004/0213darfur.htm for a discussion of constraints imposed by the government in Sudan, for example. 29 See Charles Rogers, The Changing Shape of Security for NGO Field Workers (Together Magazine, no. 57, 1998); available from http://domino-201.worldvision.org/worldvision/pr.nsf/ stable/NGOsecurity and Joanna Macrae and Humanitarian Policy Group Overseas Development Institute (London England), The New Humanitarianisms: A Review of Trends in Global Humanitarian Action (London: HPG Overseas Development Institute, 2002), 9. 30 For some discussions of such circumstances, see Nicolas de Torrenté, ‘Humanitarianism Sacriﬁced: Integration’s False Promise’, Ethics and International Aﬀairs, 18/2 (2004), and CARE USA, Care Decries Murders of Aid Workers in Afghanistan (2004 [cited March 2005]); available from http://www.globalpolicy.org/ngos/aid/2004/0227murder.htm Macrae and Overseas Development Institute (London England), The New Humanitarianisms: A Review of Trends in Global Humanitarian Action.

The Authorization and Accountability of NGOs

207

8.2.2.2. Instituting Non-Electoral Public Empowerment in Global Politics 2: Connecting Empowerment to Stakeholder Delegation Having now established some possible ways in which NGOs can be empowered by other agents according them the capacities necessary for the exercise of public power, the remaining question is the following: How could it be said that NGOs empowered in these ways are empowered democratically? That is, how can NGOs be empowered in ways that are contingent upon processes of delegation by stakeholders? As I explained earlier, the empowerment component of democratic authorization mechanisms does not only entail the conferral of the requisite capacities for public power; it further requires that this empowerment process is connected in adequate ways to stakeholder signalling mechanisms, such that empowerment is contingent upon prior commitment to a democratic mandate. As I also explained earlier, the empowerment of representatives within democratic states is generally contingent upon electoral processes of democratic delegation, through which control of a centralized (state) apparatus of public power is handed over to representatives only after they have secured an electoral mandate. The connection between empowerment and delegation is not so straightforward in relation to NGOs’ exercise of public power, however, since NGOs can depend, in the ways I have just outlined, upon a range of organizationally disconnected actors (states, donor agencies, International Organizations, private individuals etc.) rather than upon a centralized process for public empowerment. Moreover, whereas public empowerment mechanisms in a democratic state can be straightforwardly linked to a centralized (electoral) process for stakeholder signalling, the decentralized process of NGO empowerment would require a correspondingly decentralized set of connections to stakeholder signals, such that each of the various agents contributing to an NGOs’ public empowerment would need separately to secure assurances that the NGO would subsequently act in accordance with a mandate from stakeholders. Under what circumstances could we claim, or expect, that each of these various empowering agents would be likely to impose stakeholder-based mandates upon NGOs in this way? It seems plausible to suggest that these various agents contributing to NGOs’ public empowerment will generally impose conditions of some kind upon their empowerment of NGOs. There are commonly quite stringent conditions attached to donor funding, for example, as well as to the other forms of empowerment that I have discussed. 31 What seems more diﬃcult, however, is to ensure that these various agents will impose conditions established by stakeholders in the form of a mandate, rather than impose conditions that serve their own private interests. Since NGOs’ political capacities are conferred upon them by a range of organizationally disconnected actors such as states, corporations, and private donors, there would be some reasons to expect that these myriad actors are also 31 See Edwards and Hulme, Non-Governmental Organisations Performance and Accountability: Beyond the Magic Bullet for discussion of the ‘multiple accountabilities’ NGOs may possess towards donors, employees, and the myriad other actors who empower them in various ways.

208

The Authorization and Accountability of NGOs

likely to be setting political agendas, since it is commonly recognized that ‘he who pays the piper calls the tune’. At the present time, it is no doubt true that most of the agents who empower NGOs do in fact do so in a way that primarily serves their own agendas, rather than those agendas established by stakeholders in the form of a democratic mandate. There are, however, certain grounds for thinking that this need not always be so, and that appropriate reforms to the structure of NGOs’ empowerment processes could connect them to stakeholder delegation mechanisms in more satisfactory ways. It is possible to distinguish here two separate problems that are likely to confront any such attempts to ensure that the public empowerment of NGOs is adequately connected to mechanisms of stakeholder delegation. The ﬁrst problem that needs to be confronted here arises from the imperative to provide each of the actors contributing to an NGO’s public empowerment (donors, states, etc.) with suﬃcient motivation to confer these capacities in accordance with stakeholder mandates, rather than in accordance with their own private interests. Although it may initially seem reasonable to be sceptical about the prospect of this, such scepticism appears to have weaker foundations once we recognize that an analogous theoretical problem arises also in relation to the motivation of key actors within states to empower representatives on the basis of electoral mandates. In the context of the democratic state, we can similarly ask: what can motivate both general citizens and key executive agents—military leaders, police chiefs, bureaucrats, and other administrative and enforcement agencies—to respect and comply with electoral mandates, rather than advancing selﬁsh interests through various forms of political insurgency, such as (in extreme cases) military coups? Certainly, compliance by key actors with electoral mandates is far from a universal phenomenon within states; rather, this phenomenon characterizes the subset of more stable democratic states. It would therefore seem reasonable to expect that the same sociological factors that motivate such compliance within democratic (as distinct from non-democratic) state-bound jurisdictions could motivate compliance with democratic imperatives within wider jurisdictions at the global level. 32 A characterization of the conditions necessary to motivate such compliance would of course require a complex sociological analysis of prospective mechanisms for global ‘democratization’, of a kind that is beyond the scope of the present discussion. Some analysts of ‘democratization’ processes stress the need for institutions of ‘civil society’ to generate the communal values and solidarities that are required to motivate individuals to sustain collective democratic institutions. As Michael Walzer argues, ‘[t]he civility that makes democratic politics 32 When I refer to ‘sociological’ factors here, I mean to include those social practices constitutive of legal systems and other institutions of ‘the state’ that play important roles in motivating compliance with democratic norms. Although we can, in some contexts, think of ‘the state’ as a unitary agent, or political actor, in this context it is more helpful to see it as the label applied to a particular set of institutionalized social practices; viewing the state in this way casts its institutions as a valid subject of sociological analysis, and thus helps to highlight the analogies between practice of democratic compliance within state-bound societies, and practices of democratic compliance beyond state-bound societies, in global politics.

The Authorization and Accountability of NGOs

209

possible can only be learned in the associational networks [of “civil society”]’. 33 Analysts from a ‘rational-choice’ perspective emphasize instead that the interests of the various participants in relationships of public power (often conceived in this literature as primarily economic interests) must be served by a framework of democratic institutions; or at the very least, rational-choice analysts stress the need for the interests of politically dominant participants to be served by democratic institutions. 34 Finally, examining the matter from a ‘constructivist’ perspective, we could point to the independent causal role of commitments to democratic norms themselves in motivating political actors’ democratic compliance. 35 At a very simple level, though, this analogy with domestic democratization processes can lead us to recognize that if democratic norms were to be strongly enough entrenched in global society (arising through independent ideational processes and/or from whichever solidaristic ties and rationalistic imperatives we think are necessary preconditions for this), the problem of motivation would not be as serious as it might initially appear. In strongly democratic states, a signiﬁcant part of what motivates individuals and corporations to pay their taxes and thus relinquish control of some of their private resources is a commitment to democratic norms, which recognize the right of the democratic community to mobilize the resources necessary to empower its delegated representatives to act. Moreover, a commitment to democratic norms is a strong part of what motivates enforcement agencies—such as armies and police forces—to enforce the laws of the democratic government, rather than enforcing only laws that serve their own private interests (which could be achieved by measures ranging from selective law enforcement through to the staging of a full-scale military coup). On this basis, we should expect that there would be a reasonable likelihood of successfully persuading private actors to transfer some of their power to democratically delegated representatives, if these actors were suﬃciently committed to democratic norms at a global level. In accordance with this expectation, it is possible to ﬁnd some examples, in practice, of circumstances under which NGOs have been required, by the donors that empower them, to act in accordance with stakeholder mandates of various kinds. In one study conducted on an NGO development project in Thailand, for example, Craig Johnson found that ‘villagers in southern Thailand were able to inﬂuence the terms on which an internationally funded NGO prioritized and allocated aid within their community’, and that a key reason for this was that the NGOs’ donors applied certain ‘stipulations aimed at

33

Michael Walzer, ‘The Idea of Civil Society’, Dissent 38/2 (1991). See, for example, Ronald Rogowski, Rational Legitimacy: A Theory of Political Support (Princeton: Princeton University Press, 1974), and George Tsebelis, Nested Games: Rational Choice in Comparative Politics (Berkeley: University of California Press, 1990). 35 See, for example, Samuel P. Huntington, The Third Wave: Democratization in the Late Twentieth Century (Norman: University of Oklahoma Press, 1991), and Barrington Moore, Social Origins of Dictatorship and Democracy: Lord and Peasant in the Making of the Modern World (Boston: Beacon Press, 1993). 34

210

The Authorization and Accountability of NGOs

encouraging and demonstrating [stakeholder] participation’ as a condition for the NGOs’ funding. 36 At the present time examples of this kind are few and far between, and as such a key challenge confronting any eﬀort to institutionalize stakeholder representation would be that of promoting the deeper internalization of democratic norms by the myriad political actors with capacities to empower and disempower others in these kinds of ways. However, we cannot take seriously the possibility of global democratization via any mechanisms (those of Stakeholder Democracy or any of its alternatives) if we do not ﬁnd it plausible to think that democratic norms are capable of providing behavioural motivations to political actors within global politics—ultimately to a degree equivalent to that within domestic democratic domains. I do not want to claim that this sociological precondition for global democracy is satisﬁed at the present time, but I see no reasons for thinking that it should not become realizable as we progress further down a developmental path of global democratization. The second problem arising here is that the decentralized structure of processes for public empowerment of NGOs—that is, the fact that NGOs are generally empowered by myriad organizationally disconnected actors, rather than through any centralized state-bound processes of governmental inauguration—presents certain diﬃculties of coordination. The centralized character of governmental inauguration in a democratic state makes it possible to connect public empowerment processes with a unitary and centralized set of (electoral) mechanisms for delegation. In contrast, the various organizationally disconnected actors empowering NGOs would each need to employ independent mechanisms of stakeholder consultation to identify the mandate on the basis of which empowerment should occur. Each donor agency, for example, would need to identify the relevant stakeholders of each NGO it sought to fund, and seek a mandate from these constituents about the conditions that should be attached to the NGOs’ funding. Such a decentralized process could be costly and overly burdensome on the resources of both donors and stakeholders. This would be, especially, so in cases where an NGO depended upon multiple other actors for its empowerment (say, multiple donors, as well as the host government of a territory in which it operated), since it could be that the NGOs’ stakeholders would then be required to participate in multiple consultative processes, established by each of the various empowering actors. In order to overcome such diﬃculties, it may be necessary for practitioners to think creatively about ways in which some coordinated approach could be agreed upon by the various participants in processes of NGO empowerment, perhaps through the establishment of some kind of functionally based or geographically based sector-wide mechanisms for identifying and consulting stakeholders.

36 Craig Johnson, ‘Towards Accountability: Narrowing the Gap between NGO Priorities and Local Realities in Thailand’ (Overseas Development Institute, Working Paper No. 149, 2001).

The Authorization and Accountability of NGOs

211

8.3. THE DEMOCRATIC ACCOUNTABILITY OF NGO S IN GLOBAL POLITICS

8.3.1. The Transparency of NGOs’ Public Power As I previously established, the ﬁrst institutional characteristic that is necessary to facilitate stakeholder evaluation of the performance of public political agents is transparent public role delineation—that is, clear allocation of public political power to speciﬁed agents. What mechanisms, then, can be employed to establish transparent public role delineation in global politics? In relation to NGOs’ public power, these may potentially take the form of some kind of ‘code of conduct’—whether such codes are formulated by an NGO unilaterally, by a group of NGOs at a sectoral or geographical level, or by some independent body seeking to impose accountability upon practices of NGOs. Codes of conduct can help to demarcate transparently the roles and responsibilities of particular NGOs by codifying them within some formalized and publicly accessible institutional charter. Existing NGO codes of conduct are characterized in a range of ways—as ‘technical standards’, ‘professional quality models’, organizational ‘statutes’, or in other terminology—and these various kinds of codiﬁed standards can each help, to greater or lesser degrees, to deliver some transparency in relation to NGOs’ roles. The forms of codiﬁcation of NGO responsibilities that may be appropriate will vary in diﬀerent circumstances, depending on the nature of a given NGO’s work, and the character of the wider political structures in which it operates. To some degree, codiﬁcation can be achieved through transparency on the part of each individual NGO about the roles and decision-making structures entailed in its own activity. Moreover, it is possible to establish greater clarity of roles than currently exists within particular NGOs, as well as within wider network structures. In a draft report on the accountability of human rights NGOs, the International Council on Human Rights Policy stipulates that each individual NGO should be required to be explicit about what it intends to do (in its ‘mission’ or ‘statement of objectives’) and what values it expects to respect. It should also be clear and explicit about how it makes decisions and who is responsible for them: this means making information public about the NGO’s governance, explaining who makes decisions about what and who is responsible for implementing them. 37

Usually, though, clear demarcation of the roles and responsibilities of NGOs cannot be eﬀectively achieved by single NGOs acting independently, since such uncoordinated processes can generate both great gaps and discrepancies among the responsibilities accepted by various organizations operating in a particular 37 International Council on Human Rights Policy, Deserving Trust: Issues of Accountability for Human Rights NGOs (2003).

212

The Authorization and Accountability of NGOs

context, and an impenetrable plethora of informational sources about the distribution of these responsibilities. Instead, greater clarity can often be promoted through the development of broader codes and standards of various kinds applying to NGOs collectively. Some of the clearest forms of codiﬁcation and formalization of NGO roles and responsibilities that so far have been achieved, in practice, have been in the humanitarian sector. In recent years in this sector, a range of standards in various forms have been developed, including the Humanitarian Charter (or ‘Sphere Standards’), the InterAction protocol on coordination, and the Humanitarian Accountability Project (HAP). 38 There have also been some moves towards developing similar standards in the development sector, but these remain, so far, more embryonic. 39 The forms of codiﬁcation of responsibilities appropriate to functional domains such as human rights advocacy are likely to diﬀer somewhat because of the ﬂuidity of many campaign networks, and the diﬃculties this could present to increased formalization of particular actors’ roles within the norm-building process. Nonetheless, as Lisa Jordan and Peter van Tuijl argue, this ﬂuidity need not entirely preclude the development of more transparent processes for delineating NGO responsibilities, since [a]s campaigns develop, relationships tend to become formalized. They can even get to the point where they have statutes such as in the case of World Rainforest Movement (WRM) or the International NGO Forum on Indonesian Development (INFID). Action committees, a memorandum of understanding, the production of joint newsletters etc, represent varying levels of formalization of mutual relationships in NGO advocacy. 40

The impact of these various initiatives in clearly demarcating roles and responsibilities has been limited somewhat by the fact that many NGOs have not adopted and implemented them, and the extent of implementation has had little monitoring. 41 Moreover, as Peter Raynard argues, establishing clearly deﬁned responsibilities cannot be achieved by NGOs acting alone, but must be achieved through agreement and codiﬁcation at a wider ‘system’ levels, involving other key actors apart from NGOs, such as donors, UN agencies, and the ICRC. Raynard recognizes that demarcating responsibilities among these myriad actors ‘can be 38 For more details of these initiatives, see Humanitarian Accountability Project, Why Humanitarian Accountability? Contextual and Operational Factors (2001 [cited March 2005]); available from http://www.hapinternational.org/pdf_word/844-RevisedCGcleanII.pdf Hugo Slim, ‘By What Authority? The Legitimacy and Accountability of Non-Governmental Organisations’ (Journal of Humanitarian Assistance, 2002 [cited March 2004]); available from http://www.jha.ac/articles/a082.htm and Peter Raynard, Mapping Accountability in Humanitarian Assistance (ALNAP, 2000 [cited]); available from http://www.alnap.org/pubs/html/mappingacc.html#b19 See also http://www.sphereproject.org/ http://www.interaction.org/disaster/NGO_ﬁeld.html http://www.hapinternational.org/en (all accessed January 2008). 39 Slim, By What Authority? The Legitimacy and Accountability of Non-Governmental Organisations (2002). 40 Lisa Jordan and Peter van Tuijl, ‘Political Responsibility in Transnational NGO Advocacy’, World Development, 28/12 (2000), 2055–56. 41 Humanitarian Accountability Project, Why Humanitarian Accountability? Contextual and Operational Factors (2001).

The Authorization and Accountability of NGOs

213

done to a certain extent through Memoranda of Understanding or Joint Protocols,’ but rightly points out that if overly bureaucratic and dysfunctional institutional structures are to be avoided, it is necessary ‘to somehow make this clear at a system (multilateral) level.’ 42 As I previously explained, the second institutional characteristic that is necessary to facilitate stakeholder evaluation of the performance of representatives is transparent public political action. Transparent public political action, conceived in terms of the various imperatives I discussed in the previous chapter, can be achieved by NGOs through a variety of mechanisms. First, it is possible for single NGOs, acting unilaterally, to establish more transparent processes within their own organizational structures. NGOs acting unilaterally can achieve some degree of transparency in public political action through mechanisms such as the publication of annual reports, and the dissemination via other channels (like internet sites) of information such as budgetary data, details of decision-making processes (such as meeting agendas and minutes, etc.), details of current and planned operational activities, and critical analysis of the outcomes of past activities. Further transparency can be achieved by NGOs acting unilaterally through the establishment of independent institutional elements within the organizational structure of the NGO—such as an independent board of trustees that is responsible for critically monitoring and disseminating information on the dayto-day decision-making of the organization. 43 Many NGOs currently fall short of the standards entailed in transparency mechanisms such as these, but existing practices embody some elements of the transparency that would be required for democratic accountability, and can give us some sense of the kinds of institutional reforms that would be needed to strengthen NGOs’ democratic legitimacy in this regard. 44 Eﬀective transparency in public political action cannot usually be achieved through such unilateral actions alone, however, for three reasons. First, some form of external scrutiny or monitoring—either by agencies established collectively by groups of NGOs, or by independently established ‘ombudsmen’ of some kind—is generally required to ensure the comprehensiveness and accuracy of the information presented by NGOs about their activities. Second, some coordination or centralization of processes for making this information available to stakeholders is generally necessary to ensure that the access costs to stakeholders are not prohibitive (in terms of time, expertise, and material resources). Third, the small and 42

Raynard, Mapping Accountability in Humanitarian Assistance (2000). See Edwards and Hulme, Non-Governmental Organisations Performance and Accountability: Beyond the Magic Bullet, for discussions of such mechanisms. 44 For critical analysis of transparency processes in contemporary NGO practice, see Hetty Kovach, Caroline Neligan, and Simon Burall, Global Accountability Report 1: Power without Accountability (One World Trust, 2003 [cited March 2005]); available from http://www.oneworldtrust.org/documents/GAP2003.pdf According to one UN-commissioned report, some NGOs that achieve good levels of transparency include the US-based Coalition for Environmentally Responsible Economies (CERES) and Oxfam, WWF and Save the Children in the UK. See SustainAbility et al., ‘The 21st Century NGO: In the Market for Change’ (London: SustainAbility, 2003). 43

214

The Authorization and Accountability of NGOs

poorly resourced scale on which so many NGOs operate places serious constraints on the range of transparency mechanisms that they can sustainably operate. 45 Some attempts have been made in recent years to address these diﬃculties in ensuring transparency, through the establishment of external and sector-wide mechanisms for scrutinizing and monitoring NGO activities, and subsequently reporting the ﬁndings of this process to stakeholders. Perhaps the widest range of initiatives has been undertaken within the humanitarian NGO sector. 46 One important recent initiative has been the Humanitarian Accountability Project (HAP), formerly the Humanitarian Ombudsman Project, which was established to monitor the compliance of humanitarian NGOs with various codes of conduct, of the kind I discussed earlier in relation to transparent public role delineation. 47 Other recent initiatives with similar purposes include the Active Learning Network for Accountability and Performance (ALNAP), which undertakes a range of activities including reviews of all published evaluations of humanitarian NGO performance, 48 and the People in Aid project, which requires reporting and accounting on compliance by those NGOs that adhere to its code of conduct in humanitarian action. 49 While fewer such initiatives have been established in non-humanitarian ﬁelds of NGO activity, some forms of ‘social auditing’ have been employed by a number of NGOs, such as Oxfam GB, and these mechanisms are intended to perform similar functions of external scrutiny, evaluation and reporting through a somewhat diﬀerent set of institutional tools. 50 Transparency mechanisms such as those outlined above are still in an embryonic form, and in most cases would need signiﬁcant development to meet the standards of transparency required for the attainment of democratic credentials. Nonetheless, these examples can help to illustrate the sorts of institutional developments that we should work towards in order to deliver transparency of the kind required for eﬀective mechanisms of democratic accountability.

8.3.2. The Public Disempowerment of NGOs As I established in the previous chapter, the second institutional component of accountability is public disempowerment, which entails the imposition of certain sanctions that eﬀectively annul certain political resources or capacities that enable an actor to perform public political functions, and thus limit the agent’s ability to 45 For discussion of some problems arising from the small scale of many NGOs’ operations, see Mick Moore and Sheelagh Steward, ‘Corporate Governance for NGOs?’, Development in Practice, 8/3 (1998). 46 For discussion of some such initiatives, see Jeﬀ Crisp, ‘Humanitarian Values, Accountability and Evaluation’, Refugee Survey Quarterly, 21/3 (2002). 47 For details of this initiative, see http://www.hapinternational.org (accessed January 2008). 48 For details of this initiative, see www.alnap.org (accessed January 2008). 49 For details of this initiative, see http://www.peopleinaid.org (accessed January 2008). 50 For critical discussion of mechanisms of ‘social auditing’ in NGO practice, see Ebrahim, ‘Accountability in Practice: Mechanisms for NGOs’; Raynard, Mapping Accountability in Humanitarian Assistance (2000).

The Authorization and Accountability of NGOs

215

continue performing its public political activities. To reiterate a key point, I made in that earlier discussion, individual stakeholders need not all participate directly in the imposition of these sanctions; rather, they can be imposed by any agents, so long as they are imposed in accordance with stakeholder signals. Although devising concrete mechanisms of democratic accountability for global public power is always a highly context-sensitive and context-speciﬁc endeavour, as I have explained, it is possible to illustrate brieﬂy the kinds of mechanisms that could achieve public disempowerment in some contexts, by highlighting how key capacities underlying NGOs’ public power can be withdrawn by those political agents who have conferred them. As I discussed in my analysis of mechanisms for public empowerment above, the key resources upon which NGOs depend to exercise public power in global politics are those of trust, or reputation, as well as ﬁnancial resources and some degree of physical security. We can examine, then, the ways in which NGOs can be disempowered through the withdrawal of these resources by those other political agents who provide them, contingent upon stakeholder signals that deem the NGOs’ performance to be in some way unsatisfactory. First, it is important to recognize that the resources of trust and credibility, upon which NGOs depend for the exercise of certain forms of public power, are readily vulnerable to withdrawal if an NGO’s performance is deemed to be unsatisfactory. As Margaret Gibelman and Sheldon R. Gelman argue, ‘in the third sector, public trust is a key component. When this trust is compromised, the costs to status, reputation, and funding can be signiﬁcant.’ 51 The vulnerability of NGOs to loss of trust is further emphasized by Joel Fleishman, chairman of the Markle Foundation, who claims that ‘[t]he greatest threat to the not-for-proﬁt sector is the betrayal of public trust, and the disappointment of public conﬁdence.’ 52 Miklos Marschall from the NGO Transparency International similarly observes that although NGOs generally enjoy a good reputation and trust among the public . . . we all know how fragile and vulnerable that trust can be. Many years are needed for NGOs to build up a good reputation, and it takes one bad move to lose it. 53

The risk of loss of trust is especially acute for NGOs who have invited clear public expectations through the acceptance of a speciﬁc mandate through prior mechanisms of delegation. As Alan Fowler observes, in relation to Oxfam’s imperative to comply with the directives emanating from its recently established stakeholder ‘assembly’, ‘the step taken by Oxfam contains a clear moral imperative to take the 51 Margaret Gibelman and Sheldon Gelman, ‘Very Public Scandals: Nongovernmental Organizations in Trouble’, Voluntas: International Journal of Voluntary and Nonproﬁt Organizations, 12/1 (2001), 60. 52 Quoted in SustainAbility et al., ‘The 21st Century NGO: In the Market for Change’, 45. 53 Miklos Marschall, Legitimacy and Eﬀectiveness: Civil Society Organizations’ Role in Good Governance (2002 [cited October 2004]); available from http://www.globalpolicy.org/ngos/ credib/2003/0529legit.htm

216

The Authorization and Accountability of NGOs

assembly seriously or risk being publicly derided and written oﬀ as a dishonest organization.’ 54 The withdrawal of trust by key social actors can serve to disempower NGOs in a range of ways. 55 First, it can disempower NGOs directly, through undermining their credibility in the deliberative processes through which many global norms are generated, and thus undermining the power that they can wield in these processes. Moreover, the loss of trust can lead to withdrawal of other key resources upon which NGOs commonly depend—particularly donor funding. The link between trust and donor funding is highlighted by the ﬁndings of a survey conducted by the Independent Sector, indicating that donors who express high trust in NGOs give ∼50 per cent more annually than those who express low levels of trust. 56 Margaret Gibelman and Sheldon Gelman highlight an even wider range of ways in which an NGO can be disabled by the loss of trust in its honesty and competence: The aftermath goes beyond the bottom line in contributions and the compromise in regard to public trust. NGOs raise money on the basis of donor trust, a trust shaken to the core by the revelations of improprieties. There are long-range implications for any NGO that ﬁnds itself the subject of allegations of wrongdoing when trust has been abrogated. For the NGO, long-term implications include diﬃculties raising money, loss of board members who want to disassociate and save their individual good name, and calls for greater accountability by citizens and government. 57

Not all decisions to withdraw resources such as funding from NGOs depend upon the degree of trust that is invested in them. As I established earlier, the idea of ‘trust’ is generally conceptualized as a disposition of conﬁdence towards some agent with respect to their likely future actions. Although resources will sometimes be withdrawn from NGOs on the basis of such unfavourable assessments of prospective actions, they can also be withdrawn solely on the basis of unfavourable assessments of past actions—that is, for punitive rather than precautionary reasons. But whatever the grounds for the withdrawal of these resources—whether punitive or precautionary—such withdrawal can serve as an eﬀective instrument for disempowering NGOs whose performance has been in some way deemed to be unsatisfactory. In my earlier discussion of public empowerment mechanisms, 54 Alan Fowler, ‘Assessing NGO Performance: Diﬃculties, Dilemmas and a Way Ahead’, in Edwards and Hulme (eds.), Non-Governmental Organisations Performance and Accountability: Beyond the Magic Bullet, 150. 55 The particular social actors whose trust in NGO depends upon for its capacity to exercise public power will vary in diﬀerent contexts. For example, when an NGO seeks to exercise public power within a deliberative interstate decision-making forum, it will depend heavily on the trust of participating state decision-makers; in contrast, when it seeks to exercise public power through the implementation of some development project within a local community, it may depend less on the trust of state actors, and more on the trust of private donors, and the trust of those local groups and individuals whose cooperation is crucial for eﬀective implementation. 56 See Independent Sector, Keeping the Trust: Conﬁdence in Charitable Organizations in an Age of Scrutiny (2002 [cited March 2005]); available from http://www.independentsector.org/pdfs/trust.pdf 57 Gibelman and Gelman, ‘Very Public Scandals: Nongovernmental Organizations in Trouble’, 60.

The Authorization and Accountability of NGOs

217

I highlighted the extent to which NGOs are dependent on the provision of resources and support by others, and thus highly vulnerable to the withdrawal of this support. The vulnerability of NGOs to loss of the trust of those agents who provide the resources upon which they depend is highlighted by the ﬁndings of a recent survey of UK-based NGOs, which established that the primary factor NGOs cited as exercising constraints on their activities was cost—in terms of time and money. 58 If the resources necessary to meet these costs are withdrawn on the basis of unsatisfactory performance, then, an NGO can straightforwardly be disempowered, and thus held accountable. The ﬁnal issue arising in relation to the public disempowerment of NGOs is the question of how this disempowerment can be achieved democratically—that is, how we can ensure that processes of public disempowerment are contingent upon mechanisms of stakeholder signalling, rather than upon unilateral decisions by the political actors who furnish NGOs with support. The establishment of appropriate connections between processes of public empowerment and prior stakeholder signals confronts the same diﬃculties of motivation and coordination that arise in relation to mechanisms of public empowerment, and the same potential solutions. As I discussed in relation to public empowerment above, those political actors who have the capacities to empower and disempower NGOs (states, donors, etc.) must ﬁrst be suﬃciently democratically motivated to seek and respond to stakeholder views on whether to disempower NGOs, through commitment to democratic norms, and perhaps also through appropriate solidaristic connections and alignment of interests. Furthermore, in order to facilitate this at a practical level, these political actors then require some coordinated procedures for eliciting stakeholder opinion so that the various political agents contributing to the empowerment of a given NGO or group of NGOs do not need to establish separate—and thus cumulatively complex and burdensome—processes for stakeholder signalling. Just as many of the political agents who publicly empower NGOs in contemporary practice do not always do this on the basis of stakeholder signals, it is no doubt true that the public disempowerment of NGOs is also commonly performed in accordance with the interests of the empowering agents, rather than in accordance with the expressed interests of stakeholders. Michael Edwards, David Hulme, and Tina Wallace rightly point out that many NGO decisions and organizational reforms are, in practice, commonly driven by donors’, rather than stakeholders’, agendas. 59 As Alnoor Ebrahim argues, however, this situation could be improved by more systematic incorporation of mechanisms of stakeholder participation within the processes through which donors (and other empowering agents) evaluate NGOs’ performance and make decisions about future support. He argues here that since 58 Ashridge Centre for Business and Society and National Council for Voluntary Organisations (Great Britain), Survey of Large UK-Based Non-Governmental Organisations and Voluntary Organisations (2001). 59 Edwards, Hulme, and Wallace, ‘NGOs in a Global Future: Marrying Local Delivery to Worldwide Leverage’.

218

The Authorization and Accountability of NGOs

. . . communities are unable to hold NGOs or donors accountable by threatening to withdraw funding or by imposing conditionalities . . . systematic involvement of communities in evaluating NGOs and funders is a key mechanism that can serve to increase their leverage. 60

Just as in the case of public empowerment mechanisms, the establishment of eﬀective mechanisms for linking public disempowerment mechanisms to stakeholder signalling processes would be a diﬃcult project, and would require some creativity on the part of practitioners to devise an eﬀective approach to the identiﬁcation and consultation of stakeholders. Nonetheless, the small steps towards this outcome that are already underway in the sector provide some basis for optimism about their future prospects. At the very least, examination of the potential for these developments can highlight the theoretical possibility that such mechanisms could serve as institutional constituents of a democratic system of non-electoral accountability.

8.4. THE CHALLENGE OF DEMOCRATIC EQUALITY Finally, it is important to address a potential challenge that could be posed to the non-electoral model of representation I have proposed here, concerning the extent to which it is compatible with democratic ideals of equality. The claim that NGOs can serve as the legitimate representative agents of stakeholder constituencies in global decision-making by means of non-electoral mechanisms of authorization and accountability could appear vulnerable to challenge on the grounds that nonelectoral mechanisms are unable to achieve adequate equality for all stakeholders, in the sense of equal access to inﬂuence within decision-making processes. The concern here might be that inequalities in the resources (ﬁnancial, linguistic, educational, etc.) available to diﬀerent stakeholder groups within global jurisdictions can lead to the exclusion of poorly resourced groups from participation in stakeholder signalling processes, and the domination of decision-making by those with greater resources to mobilize politically in these participatory processes. Ngaire Woods, for instance, argues that including NGOs in international decision-making undermines equality in representation because it serves to ‘magnify Northern views—both outside of governments, and through governments— in the international organizations, adding yet another channel of inﬂuence to those peoples and governments who are already powerfully represented.’ 61 This claim is bolstered by the evidence of the 1995 Benchmark survey of NGOs, in which the overwhelming majority of respondents identiﬁed inequalities of access among diﬀerent NGOs as a signiﬁcant problem. Here, respondents identiﬁed

60

Ebrahim, ‘Accountability in Practice: Mechanisms for NGOs.’ Ngaire Woods, ‘Making the IMF and the World Bank More Accountable’, International Aﬀairs, 77/1 (2001), 99. 61

The Authorization and Accountability of NGOs

219

larger NGOs, English-language-run NGOs and Northern NGOs as especially dominant (76, 75, and 71% of the respondents respectively). 62 However, while it is no doubt true that the vastly unequal distribution of resources in global politics undermines the prospects for establishing genuinely equal and inclusive participatory processes of non-electoral authorization and accountability, there are few persuasive grounds for believing that problems of equal access would generally be more serious for a non-electoral than for an electoral model of representation. The proposition that electoral representation more eﬀectively ensures equal and inclusive participation might appear plausible at a superﬁcial level, since it could be argued that elections ensure equality for participants by according a vote of equal weight to each individual, irrespective of her resources. It could be argued here that non-electoral mechanisms incorporate no such formal mechanisms for equal individual input, and are thus more vulnerable to domination by well-resourced groups of stakeholders. This argument would be misguided, however, for two reasons. The ﬁrst and most signiﬁcant problem with this argument is that the aggregative decision-making process linked to such formal electoral equality fails to satisfy the egalitarian imperative to take account of varying intensities of interest within a collective decision-making process, which I established in Chapters 5 and 6. As I discussed in those earlier chapters, democratic equality requires that each individual’s interest must be accorded weight proportionate to the intensity of her interest in the outcome of the decision. The purely formal equality of an electoral system, which fails to take account of widely divergent intensities of the conﬂicting interests of diﬀerent individuals and groups, fails to satisfy this broader egalitarian imperative. Second, even in circumstances in which the various stakeholders in a decision share a relatively equal intensity of interest, such that formal electoral equality would be compatible with the broader egalitarian requirement for consideration of interests proportional to their intensity, it is still not clear that formal electoral equality would overcome problems of resource inequality any more satisfactorily than non-electoral methods of participation. The force of the argument that equal inﬂuence in decision-making can be achieved through formal electoral equality— that is, through equality in the weighting accorded to each individual’s vote— rests upon a problematic assumption: that the range of decisions reached through participatory voting are somehow critical or fundamental in the overall decisionmaking process. That is, the argument depends upon the assumption that the decisions stakeholders make through voting have a signiﬁcant impact on the overall outcomes of public political decision-making processes. Such an assumption could potentially be viable in small and politically homogeneous jurisdictions, in which the range of conﬂicting interests and competing policy alternatives is small, and thus it is feasible for stakeholders to vote on a full and inclusive range of the contending policy options. In large and heterogeneous 62 Benchmark Environmental Consulting, Democratic Global Civil Governance: Report of the 1995 Benchmark Survey of NGOs (Oslo: Ministry of Foreign Aﬀairs, 1996), 16–18.

220

The Authorization and Accountability of NGOs

jurisdictions, however, in which the range of conﬂicting interests and competing policy alternatives is large, it would be prohibitively costly for stakeholders to evaluate and vote upon all the competing policy alternatives within the conﬁnes of an electoral process. Under such circumstances, it is thus inevitable that many of the most critical political decisions in electoral systems are made in agendasetting phases of decision-making, whereby the options that are to put to a vote are determined. 63 For practical reasons, then, the range of options put to popular vote in large and pluralistic jurisdictions, of the kind that commonly arise in global politics, must be only a tiny subset of the possible options that could be chosen. In global politics, the signiﬁcance of agenda-setting process to the ﬁnal decisionoutcome in many democratic jurisdictions is therefore likely to be enormous. Once we recognize that democratic equality requires inclusion of all individuals not only in a formal voting process but also in the process of agenda-setting, the egalitarian advantage conferred upon electoral mechanisms by the formal equality of voting is greatly diminished. Since electoral mechanisms for achieving inclusion in processes of authorization and accountability concern themselves only with equality in decision-making among predetermined options, and not also with equality in the determination of these options themselves, elections will yield outcomes only as inclusive as the prior agenda-setting processes. But once we turn our critical attention to the agenda-setting processes preceding and undergirding the egalitarianism of elections, we must recognize that these processes are vulnerable to domination by the actors with the greatest resources for political mobilization, in the same ways and for the same reasons that the non-electoral mechanisms I have outlined here are so vulnerable. Without equal and inclusive processes for agenda-setting, then, the formal equality entailed in a vote confers little more than an egalitarian façade upon the underlying resource-distorted decision-making process.

8.5. CONCLUSIONS The general framework for non-electoral democratic representation that I have set out in this chapter and the last can serve two functions. First, it can provide a theoretical basis for some degree of optimism regarding the prospects for democratizing global public power, since this framework could more feasibly be applied to existing structures of global public power than could conventional electoral institutions. Electoral systems were devised to establish legitimate representative agency within a historically and institutionally very speciﬁc form of polity—the 63 The signiﬁcance of the power that is exercised in this phase of decision-making has been inﬂuentially highlighted by Peter Bachrach and Morton Baratz, ‘Two Faces of Power’, The American Political Science Review, 56/4 (1962). See also Steven Lukes, Power: A Radical View, 2nd edn. (Basingstoke: Palgrave Macmillan, 2005) for a discussion of this form of power.

The Authorization and Accountability of NGOs

221

territorial nation-state. Elections work well within such polities, because these polities embody both a ﬁxed and stable site of public power (the state) and a ﬁxed and stable constituency (the nation), which can be simply linked through electoral mechanisms. In global politics, however, the complexity and ﬂuidity of both public power and democratic constituencies make electoral institutions unsuitable in many circumstances. The non-electoral framework that I have presented here is much more institutionally adaptive, and hence better able to accommodate the complex and ﬂuid structures and dynamics of global politics. Although the representative model outlined here is more theoretically promising than electoral alternatives in terms of its potential for implementation in practice, many public political agents in global politics nonetheless fall short of even these more realistic democratic standards. Accordingly, the second function that can be served by this model is as a basis for political reforms to democratize global politics. Although I present this representative model as a feasible and attainable ideal for contemporary global politics, I have not tried to argue that stakeholder representation is currently employed in any mainstream NGO practices, in fully democratic forms. Moreover, I have not attempted to answer here any empirical questions regarding the extent to which NGOs currently succeed or fail, in practice, to conform with the standards entailed in this representative ideal. In most contexts, it is undoubtedly true that existing forms of stakeholder participation in NGO decision-making are not broad enough, inclusive enough, or binding enough to constitute adequate mechanisms of democratic authorization. Moreover, it is true that in most current contexts transparency mechanisms are too weak, and disempowerment processes neither closely enough connected to stakeholder preferences, nor consistently enough imposed, to constitute adequate mechanisms of democratic accountability. However, these deviations, in practice, from the theoretical ideal of representation that I have articulated here do not in themselves undermine the plausibility of the ideal itself, any more than deviations, in practice, from traditional electoral models of representation within the many non-democratic and weak-democratic states in global politics undermine the plausibility of traditional democratic ideals. The institutional framework that I have presented here is very general and abstract in form, and I have not attempted to develop a detailed institutional blueprint for practical application. Such development is not possible at a general theoretical level, since elaboration will necessarily be context-dependent in two respects. First, at a normative level, detailed elaborations of these mechanisms of authorization and accountability will need to vary in accordance with the particular normative conceptions of inclusiveness and equality that are endorsed by the stakeholders subject to public power in diﬀerent contexts. Second, at a pragmatic level, the particular mechanisms that will be able to provide eﬀective channels for authorization and accountability will diﬀer in each political context, in accordance with the varying resources (in terms of language, education, mobility, time, money, and so on) available to the NGOs and the stakeholder

222

The Authorization and Accountability of NGOs

groups involved in the particular relationships involved. To some degree, the task of institutional elaboration could be assisted by more rigorous empirical investigation, to develop a clearer picture of how these mechanisms could work in diﬀerent contexts, and to some degree the task of institutional elaboration will fall upon the NGOs, donors, states, and stakeholders interacting in each set of jurisdictional relationships. Ultimately, then, it is at the level of political practice, not political theory, that further details of a non-electoral model of representation must be proposed and evaluated.

Conclusion The theory of Global Stakeholder Democracy developed in this book embodies a framework that can be employed both to assist with the democratic evaluation of the myriad political actors wielding power within the contemporary global polity, and to guide the formulation of institutional strategies for their future democratization. The account of democratic societies developed in Part I demarcates the range of political activities that should be subject to democratic regulation, as well as the corresponding ‘peoples’ that are entitled to exercise democratic control over these forms of public power through collective decision-making institutions. The account of democratic representation developed in Part II delineates the multiple constituencies that are entitled to representation within representative processes of global social choice, and outlines some institutions through which particular political agents could establish representative credentials with respect to these constituencies. The arguments and illustrations I have presented here are not suﬃcient to constitute a fully developed blueprint for global democratic reform, and many questions remain which would need to be resolved through further empirical research and practical experimentation. More work must be done, in particular, to devise operationalizable mechanisms for delineating the boundaries of ‘public power’ and ‘stakeholder communities’ in diﬀerent political settings, and for achieving eﬀective institutions for non-electoral authorization and accountability. As I have emphasized throughout the book, the solutions to these practical challenges will necessarily vary in important respects in diﬀerent political contexts. As such, the practical implementation of the theoretical framework of Global Stakeholder Democracy would require more detailed institutional elaboration in each political context, in order to take account of context-speciﬁc normative conceptions and practical constraints. My hope is, however, that the arguments presented in this book are suﬃcient to articulate the key criteria and considerations that must be invoked in each such process of context-speciﬁc elaboration and implementation. As a work of normative political theory, this book aspires only to contribute to practical institutional debates by conceptually framing—and helping to set a new agenda for—subsequent empirical investigation (and practical experimentation) in relation to how global democracy can best be achieved. It is also important to reiterate here that although my illustrative analysis in this book has focused on the subset of non-state actors commonly called ‘NGOs’, these are not the only kind of actor exercising public power over populations beyond

224

Conclusion

the territorial boundaries of democratic states. The framework of Global Stakeholder Democracy can be further applied to the broader range of actors exercising public power in global politics. These other actors include, most notably, Transnational Corporations (TNCs), whose activities can impact not only on individuals’ economic welfare but also upon the wider environmental and social conditions of their lives; International Organizations (IOs), which can sometimes engage in important political decision-making independently from the governments of member states; and states themselves, which commonly exercise signiﬁcant forms of economic, social, and sometimes even military power over populations external to their own territorial jurisdictions. As I noted in the Introduction to this book, these other actors diﬀer from NGOs in a number of important respects: the forms of power they wield and the institutions through which they wield it; the kinds of communities that would likely qualify as ‘stakeholders’ in their public power; and the forms of countervailing power that could be harnessed eﬀectively (via processes of authorization and accountability) to ensure their public decision-making is appropriately representative of stakeholder interests. In this book, I have not presented an analysis of which forms of power exercised by these other actors may qualify as ‘public’, who their stakeholders may be, and what particular representative mechanisms may be suitable for connecting each to their stakeholder populations. Nonetheless, I suggest that the general theoretical framework I have developed here provides the normative criteria and analytic categories that we would need, in order to begin answering these various concrete questions about the broader democratization of global politics. I suspect that this claim about the broader applicability of the Stakeholder model will appear most plausible in relation to International Organizations, since they share many relevant institutional characteristics in common with NGOs; as such, it is not diﬃcult to imagine how the kinds of institutions I have discussed in this book could be employed to democratize their power. Similarly, my claim that the Stakeholder model could be applied to the exercise of power by states beyond their territorial boundaries poses few special conceptual diﬃculties— although it does, of course, raise serious practical diﬃculties given the entrenched propensity of states (even democratic states) to exercise external power in the pursuit of ‘national interests’, without regard for appropriate democratic restraint. As a sociological matter, commitment to democratic norms would need to be signiﬁcantly strengthened at the global level (and more strongly internalized at domestic levels as motivators for external state action) before stakeholder control of externally wielded state power could become a realizable democratic prospect. The category of actors that perhaps raises the most signiﬁcant conceptual diﬃculties for the application of the Stakeholder model is that of TNCs. These diﬃculties arise primarily because the kinds of public power wielded by TNCs diﬀer in some important respects from those wielded by NGOs, such that they are not vulnerable to the same methods of empowerment and disempowerment (as elements of democratic authorization and accountability) that I have suggested

Conclusion

225

could work for NGOs. More speciﬁcally, the public power wielded by TNCs depends less upon public trust than does that of NGOs, and it also relies less on ﬁnancial support from donor agencies. Because of such diﬀerences, it may initially appear diﬃcult to employ mechanisms of non-electoral authorisation and accountability as means for subjecting the public power of TNCs to democratic control. But upon closer examination, we can see that the diﬀerences in the types of public power exercised by TNCs do not preclude the application of principles of Stakeholder Democracy; rather, these diﬀerences just mean that the sorts of institutional mechanisms that would be needed to achieve eﬀective democratic authorization and accountability would be quite diﬀerent for TNCs than for NGOs in many cases. To begin with, whereas NGOs wield signiﬁcant forms of public power via norm-building institutions as a result of widespread public trust (as I discussed in Chapter 3), TNCs participate to a much lesser degree in developing international law and other authoritative norms. As a result, the fact that corporations cannot (like NGOs) be sanctioned by withdrawal of public trust does not in itself generate any democratic problems, since corporations do not exercise nearly as much trust-based public power in the ﬁrst place. Corporations wield forms of public power more similar to those of NGOs via their impacts upon the distribution of material resources within communities, as well as via the public impacts of certain ‘externalities’ of this economic activity. With respect to the ﬁrst of these forms of corporate public power, the fact that TNCs allocate material resources to a large degree through market institutions, while NGOs allocate them primarily through more bureaucratic institutions, has no bearing on the ‘public’ character of these impacts, according to the criterion of ‘publicness’ that I have identiﬁed. With respect to the second, externalities such as environmental impacts, and impacts on security situations within certain territorial populations, provide examples of the kind of thing I have in mind. If TNCs are to be held to democratic account by the stakeholders in these forms of public power, then there must be some way to restrain their capacities to engage in these economic activities if they prove to be insuﬃciently responsive to their stakeholders’ interests. This cannot be achieved simply by withholding donor funding, as in the case of NGOs. However, it would be wrong to conclude that TNCs are thus ﬁnancially ‘independent’ and invulnerable to political sanction. On the contrary, the economic power of TNCs is contingent in important ways both upon the decisions of investors and consumers, and upon various legal privileges which they are accorded by their host states and societies. First, corporations depend upon willing buyers for their goods and services, and investors in their business; coordinated consumer and shareholder sanctions provide an important mechanism through which corporate activity can be eﬀectively held to account. Moreover, corporations now enjoy an array of legal privileges, including not only property rights and tax advantages but also ‘limited liability’ for their owners—a privilege which helps them to access the large quantities of investment capital which they commonly rely upon for their economic activities. At the same time, they are protected from

226

Conclusion

public scrutiny by laws which enable them to withhold important information about public impacts on the basis that it is private, ‘proprietary’, information. The potential to change the legal regimes that facilitate corporate activity while protecting it from public scrutiny thus creates ample opportunity for holding corporations accountable for their economic activities, provided there is suﬃcient political will on the part of lawmakers to undertake appropriate reforms. Reforms to tort law (to enable stakeholders to hold corporations accountable in civil law for their public political impacts) could also provide mechanisms for corporate accountability to stakeholders. Much more would need to be said about the speciﬁc powers and vulnerabilities of TNCs—as well as other powerful global actors such as IOs and states—before I could claim to have fully established the applicability of the Stakeholder model to these other actors. If I were to attempt this task with the same level of detail with each of these other categories of actors as with NGOs, the book would need to be very much longer than it presently is. I think it would be too ambitious a task for a single book not only to develop and defend a theoretical framework for Global Stakeholder Democracy, but also to demonstrate how it could be applied to every category of powerful actor on the contemporary global stage. For this reason, I have chosen to focus, for illustrative purposes, on a single category of actor (NGOs). But I hope that the little I have said here will be suﬃcient to convince readers that the applicability of the Stakeholder model to TNCs and other actors— as well as NGOs—is at least suﬃciently plausible to warrant further exploratory work on the subject. I think that application of the framework to other categories of actors would constitute a very interesting and important project; however, it is not one that I have been able to pursue myself within the constraints of this book. My suggestion that the framework of Global Stakeholder Democracy can be applied to a wider range of actors beyond NGOs—including states, when their public power extends to populations beyond their territorial borders—raises the additional question of how it relates to more conventional democratic frameworks focused on ‘closed’ nation states. To what extent does my global democratic framework present a direct challenge to established state democracies (or, at least, to the established ideals of these democracies), and to what extent can the two democratic models coexist harmoniously within a multi-jurisdictional global order? First, at a theoretical level, this global democratic model does not preclude the operation of state democracies within their territorial boundaries, to the extent that they do actually function as the ‘closed’ political communities envisioned by conventional democratic theories. Nation-state democratic models can legitimately coexist with other (global and supra-territorial) democratic jurisdictions in global politics, so long as each maintains congruence between the boundaries of its public political agents and those of its solidaristic (self-determining) population. The traditional ideal of a ‘closed’ nation-state democracy will only be challenged, then, to the extent that it does not, in practice, function as the ‘closed’ political society presupposed by the traditional ideal. Other jurisdictional spheres

Conclusion

227

will only intrude within the democratic state to the extent that powerful agents within it (such as state agencies or powerful NGOs) exercise public power external to the state’s own jurisdictional boundaries, or alternatively to the extent that powerful agents external to its jurisdictional boundaries (such as International Organizations, powerful NGOs, or other state agents) exercise public power over some or all of its citizens. Certainly, it is true that such public political penetration of state boundaries is extensive within the contemporary global arena, and the signiﬁcance of democratic state institutions is accordingly diminished. It is important to recognize here, however, that the democratization of these non-state forms of public power is not the source of the erosion of democratic states; rather, the diminution in the democratic signiﬁcance of states is a response to the reality that the theoretical ﬁction of the ‘closed’ society and the unitary ‘sovereign’ state is an unsustainable ideal in our globalizing world. At a more practical level, it is true that certain tensions can arise between the global representative institutions that I have outlined here and state-based democratic institutions, in so far as their coexistence can impose competing and conﬂicting demands on the political energies and allegiances of individual participants. It may be, then, that eﬀorts to democratize non-state forms of public power could to some degree further erode the strength of democratic institutions within states, as the participatory resources of individuals are divided among multiple jurisdictional levels and as their participation at these multiple levels dilutes the strength of their political allegiances to the national political communities. This process is, however, an inevitable cost of the social complexity with which we all now live. Unless we imagine that it could be both feasible and desirable either to command a retreat to an isolationist world of closed territorial societies, or to impose a centralized global ‘state’, then a complex multi-jurisdictional global democracy of the kind I have advocated here is the only viable option. Finally, a brief comment is warranted concerning a particular form of challenge that could confront my argument that this representative democratic model is the appropriate normative framework for evaluating certain activities of nonstate actors such as NGO in global politics. Many practitioners and analysts reject democratic standards as the appropriate criteria to employ in evaluating the legitimacy of NGOs, arguing that we should evaluate their activities in terms of alternative criteria such as ‘functional eﬃcacy’, ‘justice’, ‘rights-protection’, ‘morality’, and so on. Some NGO practitioners, such as Miklos Marschall from NGO Transparency International, have argued along these lines that ‘we need civil society organizations not because they “represent the people”; we need them because through them we can get things done better.’ 1 It is certainly important to acknowledge that it is sometimes entirely appropriate to invoke non-democratic normative criteria of this kind in evaluations of NGO activity. However, what is crucial to my argument is the recognition 1 Miklos Marschall, Legitimacy and Eﬀectiveness: Civil Society Organizations’ Role in Good Governance (2002 [cited October 2004]); available from http://www.globalpolicy.org/ngos/credib/ 2003/0529legit.htm

228

Conclusion

that democratic standards, and the alternative normative standards invoked by commentators such as Marschall, are appropriately applied to diﬀerent aspects of NGOs’ activities. It would be a mistake to imagine that we could identify one master-principle that could be applied to regulate all activities in which an NGO—or other actor—can be engaged; rather, diﬀerent normative principles apply to actors depending on the particular roles that they are performing in their social relationships. As I noted in Chapter 3, NGOs can perform many diﬀerent roles in global politics, some of which are ‘public’ because their activities constrain the autonomy of others, but many of which are ‘private’ because their activities do not have such autonomy-constraining impact. Democratic standards, as I have argued throughout this book, are the appropriate standards to invoke only in relation to the public forms of NGO activity. Within the category of ‘private’ roles and activities, a range of diﬀerent kinds of normative principles may apply, depending on the other characteristics of their relationships with others. One important kind of non-public role that an NGO can perform is as a contractee, participating in either formal or informal contract relationships with actors who provide funds (or other forms of support) to the NGO on the basis of certain expectations about how these resources will be deployed. Such roles can be associated with a diﬀerent kind of normative obligation, grounded in the normative bindingness of contracts. Another kind of non-public role that an NGO can perform is as an employer, and this role can be associated with various nondemocratic obligations (of a ﬁnancial, professional, or broader welfare-oriented character) towards employees. At other times, to the extent that it is not bound by such democratic, contractual, or professional obligations, an NGO can act as an independent moral agent (as a group of individuals with uniﬁed moral beliefs and agendas). In these roles, an NGO is obliged to comply with whichever moral values, and pursue whichever moral outcomes, it believes to be correct; in private roles such as these, then, moral criteria of the kind articulated by Marschall and others may be entirely appropriate standards to apply in evaluating the NGO’s behaviour. In light of these multiple roles that NGOs can perform, one of the key challenges that any NGO (or other global actor) will confront is to avoid engaging in roles that generate conﬂicting obligations with respect to any given activity. There are various ways in which conﬂicts between the moral imperatives associated with diﬀerent roles could arise. To take one example, an NGO may be engaged in a contractual relationship with a donor, which demands that it pursue certain policies that conﬂict with the expressed wishes of the NGO’s stakeholders (i.e. the wishes of the community of individuals whose autonomy is constrained by the enactment of these policies). Here, the NGO would experience a conﬂict between the private obligations arising from its contractual relationships and the public democratic obligations associated with its exercise of public power. To take another example, an NGO’s stakeholders may demand some set of policies that conﬂict with the private moral commitments of the NGO’s decision-makers, such as policies that involve forms of gender or racial discrimination. Here, the NGO

Conclusion

229

would experience a conﬂict between its public democratic obligations, and its private moral commitments. When confronted with these kinds of conﬂicts among the various obligations associated with its diﬀerent roles, an NGO should in some way seek to modify or disengage from one of these roles, in order to defuse the conﬂict. For example, in the former example where an NGO has contractual obligations that conﬂict with its democratic obligations to stakeholders, it must alter the substance of its relationship with either the donor or the stakeholder community to circumvent this conﬂict. This could involve either altering the terms of the contract with the donor (or refusing the funds if their conditions cannot be altered), or ceasing engagement in the public activities to which stakeholders object. In relation to the latter example, in which an NGO has moral commitments that conﬂict with democratic obligations to stakeholders, it could either persuade its stakeholders to modify their demands in relation to the discriminatory policies, or else discontinue its public political activities in the disputed policy area. The possibility of these sorts of normative conﬂicts, and the approach to resolving these conﬂicts that I have prescribed, are not unique to the domain of global politics; conﬂicts of precisely these kinds also arise for individuals and other social actors interacting within democratic states. Public political agents (such as elected politicians) within democratic states can also experience conﬂicts between the democratic obligations associated with their public political roles, and the private obligations associated with other roles they may occupy in their wider lives. For instance, it would be possible for a politician’s democratic obligations to conﬂict with certain professional or contractual obligations she may have to the shareholders of a corporation for which she is a board member, or to an employer who continues to pay her some kind of salary while she holds her public political oﬃce. It is precisely to avoid conﬂicts of this kind that many democratic states impose restrictions upon the kinds of private professional and ﬁnancial dealings that public oﬃcials may have during the tenure of their public oﬃce. Similarly, it is entirely possible for a politician’s democratic obligations, arising from her electoral mandate, to conﬂict with certain private moral values to which she may be strongly committed. In such circumstances, we would expect a principled individual either to persuade her constituents to revise their instructions on the contested policy, or to resign her post and thus renounce the public power that she wields in the contested policy area. We can thus see that the multiple imperatives arising from the myriad roles of NGOs within global society are clearly analogous to the multiple imperatives faced by agents of public power in established democratic states, and they can be confronted and resolved through equivalent political strategies. Democratic principles and wider normative ideals to which groups and individuals may be committed can sometimes come into tension, in global politics just as in domestic state-bound societies. A free-standing democratic analysis, isolated from consideration of these wider normative imperatives, cannot provide any answers about how particular agents should reconcile their democratic and non-democratic

230

Conclusion

values in particular circumstances; it cannot determine, for example, whether an NGO should in a given instance choose to comply with certain demands by stakeholders that conﬂict with its moral principles, or whether it should instead choose to discontinue its public political activities within that community. Free-standing democratic analysis can, however, conﬁrm the necessity that an NGO make such a choice when confronted with these conﬂicting imperatives. As I have acknowledged throughout this book, the contemporary arena of global politics is an enormously diverse and complex political universe, relative to the more socially homogenous and institutionally integrated world of the territorial nation state. Because of this, the task of democratizing its myriad forms of public power poses a momentous challenge to modern democrats. The protesters in Seattle, chanting democratic slogans and demanding democratic reforms, helped to generate some political momentum for confronting this challenge. But to meet this challenge satisfactorily, we must be prepared to revise not only our global institutional practices but also the core categories and conceptions of our statebased democratic theories. But, although the democratic conceptions of ‘public power’, ‘stakeholder’ communities, and non-electoral ‘representation’ that I have defended in this book require signiﬁcant adjustments to traditional state-centric democratic ideals, they are not radical or revolutionary in the global domain. To the contrary, they are at the heart of contemporary global democratic discourse, and embedded within the embryonic institutional practices that are emerging to enhance the democratic accountability of NGOs and other agents of global public power. For this reason, then, a reﬂective understanding of our existing global democratic principles and our embryonic global democratic practices is suﬃcient to reveal how the political activities of NGOs and other non-state actors can, as former UN Secretary-General Boutros Boutros-Ghali has claimed, help us remain ‘faithful to the urgent exhortation with which the preamble to the [UN] Charter begins: We the peoples’. 2 2 Boutros Boutros-Ghali, ‘We the Peoples: Building Peace’, paper presented at the UN Department of Public Information Forty-Seventh Annual Conference of Non-Governmental Organizations, New York, 20 September 1994.

Index Aall, P. R. 74–5 accountability 1, 15, 165, 185–90, 211–18 and disempowerment 185 function of 171–2 mechanisms of 60, 170–9, 191 and punishment 175–6 and substantive representation 172–7 and transparency 185 AccountAbility1000 68, 93, 196, 197 Active Learning Network for Accountability and Performance (ALNAP) 195, 214 adjudicative process 101 advocacy activities 65, 68 Afghanistan 74, 78 aid 76–7, 80 Aitkin, D. 126, 166 Amnesty International 68, 69 Amy, D. J. 126 Anderson, M. B. 74 Anheier, H. K. 68, 154 Annan, Koﬁ 3, 43, 67, 141 Anscombe, E. 51 Archibugi, D. 2, 5, 23 Arend, A. C. 46 Arrow, K. J. 133 Austin, J. 46 authoritative normative production 64–5 authority 48, 50 see also public power authorization 15, 165 function of 171–2 Hobbesian model 175 mechanisms of 60, 170–9, 180–5, 190–1 and substantive representation 172–7 autonomy: and authority 50 constraints 71–81; indirect 72, 75; marginal and signiﬁcant 96; material 71–8, 205 deﬁnition 37–8 and democratic control 59–60 indirect impact 64 individual 35–8, 90–1, 117–18 Bachrach, P. 220 Balduini, S. 2 Baratz, M. 220 Barnett, M. 55 Bate, R. 69–70 Bates, J. 4

Beck, R. J. 46 Behn, R. D. 175, 187 Beitz, C. R. 10–11, 112–13, 123–4, 150 Bennett, J. 74, 78 Berlin, I. 71 Biersteker, T. J. 30, 45, 72 Birch, A. H. 126 Black, D. 134 Blake, M. 40 Bohman, J. 28, 112 boundaries 12–13, 21–2 decision-making 23, 28–9, 39–40 delineation 41 global 35 boundary problem: communitarian solution 23, 28, 42 cosmopolitan solution 23, 25, 27, 29, 30, 41, 42 Boutros-Ghali, Boutros 2, 3, 67, 83–4, 230 Boyden, J. 70 Buchanan, A. 11, 24, 27 Burall, S. 94, 171, 178, 213 Burke, Edmund 110–11 burkean model 109–14, 148 democratic legitimacy in 117–20 states in 120–1 Burnheim, J. 167 Byers, M. 47, 69 Carens, J. H. 8 Cassese, A. 46, 47, 51 centralized power 45, 46, 53–8 see also constitutionalized power Cerny, P. 56 Chapman, J. W. 22 Charnovitz, S. 67 child labour 70 Christman, J. P. 37 citizenship, dual 28 civil society 44, 153–4 closed society 22, 40 deﬁned 13 standard model of 23–9 codes of conduct 63, 195, 211 Commission on Global Governance 83 communitarian statism 5 competitive elitist model 168 Condorcet, J.-A.-N. de Caritat 133 conﬂict situations 73–5, 77–8 Conniﬀ, J. 110

232 Connolly, W. E. 115–16 consensus decision-rules 100 constituencies 97–8, 106 delineating 143–5 multiple 100, 140–3 NGOs 6 overlapping membership 142 territorial 126, 161 constitution building 35 constitutionalized power 45, 52–3, 58–9 see also centralized power constraints see autonomy contextual reﬂective equilibrium 7–8 contract 200 Cooley, A. 74 coordination 210, 217 cosmopolitan theories 5–6 crisis situations 198–9 critical resource control 75–81 Crocker, C. A. 74–5 Culter, A. C. 30, 56 cultural diversity 129 Currah, K. 2 Cutler, A. C. 45 Dahl, R. A. 22, 123, 132, 166, 174 D’Amato, A. 46 Dannreuther, R. 109 DDT 70 de Torrenté, N. 73, 206 decision-making: boundaries 23, 28–9, 39–40 deliberative 14, 111–12, 142–3, 158–9, 199 global public 178–9, 180 mechanisms of 23 representative 12–13 decision-making structure: multilateral 99–101 unilateral 98–9, 100–1 delegates 123, 148 delegation 16, 175 and empowerment 207–10 functions of 180–4 NGOs 195–210 see also mandates deliberation 14 interest aggregation 135–6 deliberative process 100, 101, 148 legitimate 133, 137, 152–3 in liberalist pluralist model 153 see also decision-making democracy: global 35 ideals 9 justiﬁcation of 36 nation-state 226–7

Index pluralist 35–40, 41, 42 representative 106 democratic control 12 and autonomy 59–60 and pluralism 57–8 democratic stakeholder communities 85–92 democratic theory, realist purpose 32 democratization: legitimate process of 34 and public power 49 state-based models 24–5 Department of International Development 198 developing countries 76 developmental communitarianism 114 disempowerment 16, 185, 188–90, 214–18 distributional decisions 79–80 Donahue, J. D. 46 Donati, M. 2 Donini, A. 77 donor funding 205–6 Dryzek, J. S. 135 Duss, S. 205 Dworkin, G. 37 Dworkin, R. M. 40 Ebrahim, A. 201, 217–18 economic deprivation, benchmark of 88–9 economic pressure 75 Edelman, R. 3, 157, 204 Edwards, M. 178, 179, 198, 200, 207, 213, 216, 217 elections 15, 116, 164, 201, 220–1 and control 165–70 and empowerment 174, 207 and equality 219–20 mechanisms 173 models 126 public contract 181, 185 Elster, J. 180 empowerment 16, 184–5, 201–10 and delegation 207–10 mechanisms 202–6 sources of 26 Environmental Council 85 equality 37, 105, 143 of access to inﬂuence 218–20 complex procedural 113 diﬀering concepts of 121, 149–50 and elections 219–20 a foundational value 38–9 individual 111–13 and intensity of preference 131–3 material 159–60, 161 and pluralism 49–50 political 123–4 procedural view of 124

Index Ethical Trading Initiative 68 externalities 225 Fair Labor Association 68 Falk, R. A. 23, 28 Farrar, C. 135 Feliciano, F. P. 47 Finley, M. J. 167 Finnemore, M. 55 Fishkin, J. S. 136, 199 Fleishman, J. 215 force 45, 46, 51–2, 71 see also violence Fowler, A. 78, 198, 215, 216 Frechette, Louise 67 free trade 70 Freeman, M. 112 Fung, A. 199 Gambetta, D. 203 Gastil, J. 199 Gelman, S. R. 215, 216 gerrymander 125, 128, 161 Gibelman, M. 215, 216 Gilpin, R. 54 Glasius, M. 68, 154 Global Accountability Report 197 Global Alliance 68 global civil society 153–4 global democracy 35 global governance 55–6 global justice 10 global politics, though local deliberation 154–8 Global Reporting Initiative 68 global representation: deliberative or aggregative 109 nationalist model 107–9 statist model 120–1 Global Stakeholder Democracy 9–10, 87, 90–1, 193–4, 199, 223–4, 226 globalization 1 GlobeScan 204 Goma 74 Goodin, R. E. 199 Grant, R. W. 22, 179 Greenpeace 69 Greensmith, J. 76 Guinier, L. 126 Gupte, P. 4 Hacker-Cordon, C. 22 Hall, R. B. 30, 45, 72 Hamilton, A. 126, 167 Hampson, F. O. 74–5 Hardin, R. 180 Hart, H. L. A. 47

233

Hart, J. A. 56 Hasenclever, A. 55 Hauﬂer, V. 30, 45, 56 Held, D. 5, 10, 22–3, 27, 30, 40, 44–5, 53, 55–6, 93–4, 97, 115, 132–3, 176, 178 Hemmati, M. 85 Hoﬀman, R. J. S. 112 Holden, B. 2 Hulme, D. 178–9, 198, 200, 207, 213, 216–17 human rights abuses 75 Humanitarian Accountability Project (HAP) 195, 212, 214 Humanitarian Charter 212 Huntington, S. P. 209 Hurrell, A. 56, 61 Hutchings, K. 109 Hutton, W. 141 hybrid representative model 162 ICRC 212 IMF 75 indigenous populations 130 institutional capacity building 80 institutional stability 57–8, 89 InterAction 212 interests: additive 111 aggregation: or deliberation 105–6, 135–6; two-stages 124–5 in a burkean model 115–16 clusters 145 common 113, 116, 117 competing 136, 160 conﬂicting 97–101, 105–6, 116 global distribution 128–30 intensity of preference 131–3, 151–2, 160, 164, 219 national 108, 114, 115–16 non-national 158 objective assessments of 148, 149 objective/abstract/unattached 110, 113–14, 115, 117 outcome-neutral 140 outcome-speciﬁc 140 partially and comprehensively attached 146–7 sectional 125–6 subjective 122–3 of transnational groups 115–16 International Council on Human Rights Policy 201, 211–13 International Forest Stewardship Council 68 International Labour Organizationn (ILO) 66 international legal discourse 47 International NGO Forum on Indonesian Development (INFID) 212

234 International Organizations (IOs) 224 International Relations, realist 54–5 interstate institutions 57 intuitions 7 Johnson, C. 210 Johnson, J. 112 Jordan, L. 212 Juniper, T. 163 jurisdictional boundaries 35 jurisdictional community 97–8, 99, 100 jurisdictional stakeholders 140–1 jurisdictional tribunal 93 Kaldor, M. 68, 154 Kapila, M. 198 Katz, H. 154 Keck, M. E. 65 Kenya 78 Keohane, R. O. 22, 46, 55, 94, 176, 178, 179 Khagram, S. 65 Kingsbury, B. 22 Knight, J. 112 Koenig-Archibugi, M. 22, 45, 94, 176, 178 Köhler, M. 23 Kovack, H. 94, 171, 178, 213 Krakochwil, F. V. 47 Krasner, S. D. 55 Kratochwil, F. V. 55 Krisch, N. 22 Krut, R. 3 Kuper, A. 5, 118–19, 166 Kymlicka, W. 166 labour standards 88 Ladenson, R. 48 Lardeyret, G. 126 Lasswell, H. D. 47 law-making 45–50 equal application 49–50 key characteristics 46–7 and public power 48–50 Levine, P. 199 Lewis, D. 68 liberal individualist model 122–38, 146–7, 148 theoretical limitations 131–8 liberal institutionalism 55 liberal pluralist model 14, 139, 146–50 evaluation of 150–8 liberal-democratic principles 36 Lijphart, A. 126, 166 Linde, T. 200 Ling, B. 70 Lipschutz, R. D. 115 List, C. 134, 135 List, R. 76, 205

Index Loew, V. 47 Lowe, V. 69 Lugt, R. D. V. 46 Lukes, S. 220 Luskin, R. C. 199 Macdonald, T. 24 McDougal, M. S. 47 McGrew, A. G. 30, 44–5, 55–6, 132–3, 176 McLean, I. 134, 135 Macrae, J. 73 Madison, J. 122, 125, 167 Malhotra, K. 80 mandate model 163–4, 177 mandates 120, 148, 149, 174, 180 in global politics 182–4 see also delegation Mander, J. 1 Manin, B. 167, 174 market institutions 225 Markle Foundation 215 Marschall, M. 215, 227–8 Marsden, P. 74, 78 Mathews, J. 43 Mayer, P. 55 Médecins Sans Frontières 200 Mill, J. S. 126, 167 Miller, D. 5, 23, 71, 107, 109, 135 Miller, J. 203 Moore, B. 209 Moore, M. 214 motivation 208–9, 217 Mozambique 77–8 Mueller, D. 167 multi-stakeholder model 14, 139–62 justiﬁcation for 146–50 practicality of 158–60 representative 144–5 Murphy, C. 56 Murphy, L. 21 Muthien, B. 72 Myers, W. E. 70 Nagel, T. 21, 33, 40 Narlikar, A. 178 nation, compared to state 106 nation-state democracy 226–7 nation-state representation 106–7 nationalist model: global representation 107–9 practicability of 115–16 nationhood, characteristics of 107–8 Neligan, C. 94, 171, 178, 213 NGO Transparency International 215, 227 Non-Governmental Organizations (NGOs): accountability 178–9, 211–18 aid distribution 76–7

Index authorization 194–210 choice of 16 coalition 1 in conﬂict situations 73–5, 77–8 conﬂicting obligations 228–9 constituencies 6 constraints on 200–1 and critical resource control 75–81 criticised 69–70 decentalized structure 210 delegation 195–210 democratization 62, 81 disempowerment 214–18 distributional decisions 79–80 detrimental impact 69–70, 78–81 donor funding 205–6 empowerment 201–10 in global politics 62–82 inappropriate democratic standards 227–8 increasing power 2 inﬂuence on international organizations 66 legitimacy 3–4, 192 norm-building process 203–5 physical security 206 public and non-public activity 228 public power 6, 43, 65–70 as representatives 163–5 as stakeholder communities 84 stakeholder participation 196–9 as stakeholder representatives 142, 143–4 transparency 211–14 vulnerability 202–3 non-state agencies 29–30 see also Non-Governmental Organizations (NGOs); Transnational Corporations (TNCs) norm-building process 203–5 norms, interstitial 69 Nowrot, K. 165 Nye, J. 46, 176, 178, 179, 203–4 Oliviero, M. B. 68 opportunity cost 79–80 O’Rourke, D. 68 Oxfam 86, 87, 88, 197, 214, 215 participation 13, 168–9 Participatory Learning and Action (PLA) 196 Participatory Rural Appraisal (PRA) 196 Pennock, R. J. 22 People in Aid 214 Phillips, A. 113, 166 Philpott, D. 53 Pierson, P. 31 Pitkin, H. F. 110–11, 122, 146, 166, 171, 172, 173, 175

235

pluralism 4, 5 and democratic control 57–8 and equality 49–50 pluralist democracy 35–40, 41, 42 Pogge, T. W. 10–11, 24, 27, 56 political agenda, transformative 30–2, 34 political context 87–9 political control 165, 168–9, 170 political dominance 56 political mobilization 145, 159 political roles: codiﬁcation 58 external recognition 58 transparency of 58 political tyrannies 60 poltical contestation 92 Porter, T. 30, 45, 56 poverty 70, 88 power: privatization of 45 see also centralized power; constitutionalized power; private power; public power Prakash, A. 56 private power, advocacy activities 65 procedural justice 10 Program on International Policy Attitudes 204 Przeworski, A. 174 public advocacy 68 public contract 180–4 public power 13, 21 and autonomy-constraining impact 64–5 conditional 173–4, 177 constitutional structure of 25–8 context of 39 and democratization 49 and force 51–2 government 6 identifying criteria 35–7 and law-making 48–50 legitimacy 36–7 liberal conceptions of 45 of NGOs 6, 43, 65–70 non-state agencies 30 pluralist 29–30, 32–40 revocable 175–6, 177 source of 24–5 in stakeholder communities 89 state-based accounts 44–5 transformation of structures of 33–4 public receptivity 199–201 punishment, and accountability 175–6 rationality 155 Rawls, J. 7–8, 10–11, 40, 90 Raynard, P. 178, 186, 212, 213

236 Raz, J. 48, 51 realist analysis 18, 29–35, 40 reasonableness 111, 155–7 receptivity 199–201 recognition, external 58 Rehg, W. 112 representation: burkian model 109–13 by expertise 167 by lot 166–7, 169–70 global 107–8 legitimate 105, 163–4 in the liberal individualist model 122–5 mirror 166–7, 169–70 multi-constituency 108–9 multi-stakeholder 107 nation-state 106–7 proportional 167 state 125–7 substantive 118–19, 172–7 representatives: legitimate 15 mandate-responsive 174–5 NGOs as 142, 143–4 responsibilities 148–9 responsible political agency 53–7 responsibility 38, 89, 148–9, 200 allocation 186, 215 demarcation 187 egalitarian 90–1 Riker, J. V. 65 Riker, W. H. 133, 134 Risse, T. 55 Rittberger, V. 55 Rogers, C. 206 Rogowski, R. 209 Ron, J. 74 Ropp, S. 55 Rosenau, J. N. 55 Rowell, A. 2 Ruggie, J. G. 31, 55 rule-making 46 Rwanda 78 Salamon, L. M. 43, 76, 154, 205 sanctions 189–90 Schumpeter, J. A. 168 security 206 self-determination, norms of 156–7 self-interest 54 Shapiro, O. 22 Shiva, V. 1 Sikkink, K. 55, 65 Silva, L. 76, 77 Simmons, A. 68 Slagstad, R. 180 Slim, H. 95, 178, 183, 212

Index Social Accountability International 68 social choice 14 democratic 101 direct or participatory 106 legitimate 105 meaningful 134 social justice 11–12 social norms: asymmetry in generating 64–5 regulative 63–70 social resource deprivation 71–2 social sanctions 63 Sokolowski, S. W. 76, 205 sovereignty 33, 53 Sphere project 195 Sphere Standards 212 Spiro, P. J. 109 stability 57–8 stakeholder: criteria 85–8 inclusiveness 182 signalling 182, 183, 196, 207, 217 stakeholder communities 13–14, 40–1 concept of 84 conﬂicting interests 92, 96, 97–101 context 87–9 distribution of responsibility 86–7 identifying members 92–3 internal and external members 94–5 non-exclusive membership 91 practical issues 89 primary and secondary members 95–7 shared political values 90–1 stakeholder model 5–6 application of 224–5 states: compared to nations 106 of diﬀering sizes 127–8 representation 125–38 statist model: aggregative process 121 global representation 120–1 practicality of 127–30 Staveley, E. S. 167 Steward, S. 214 Stewart, R. B. 22 Stoikov, G. 70 Stokes, B. 4 Stokes, S. C. 174 Strange, S. 30 Strauss, A. 23, 28 Stuart, R. 80 Sutch, P. 114 swearing-in process 181 Taylor, C. 71 Taylor, I. 72

Index Themudo, N. 154 Tollison, R. 167 Townsley, P. 95 Transnational Corporations (TNCs) 224–5 transparency 16, 89, 185, 186–8 cost of 188 mechanisms 214 NGOs’ 211–14 of political roles 58 transparent public political action 213–14 transparent public role delineation 186–8, 211 Tren, R. 69–70 Truman, D. B. 147–8 trust 2–3, 69, 157, 203–5, 215–17 trustees 118, 142, 148, 149 United Nations 2, 212 Agenda 21 141 Charter 230 Commission on Sustainable Development 141 Global Compact 68, 141 language of 83–4 legal and policy norms 63 Programme on HIV/AIDS (UNAIDS) 66 relationship with NGOs 66–8, 195 Security Council 73 soft law 64 Uphoﬀ, N. 200 Urbinati, N. 167 Urkin, A. B. 134 Vakil, A. C. 17 van Tuijl, P. 212

237

Vidal, J. 1 violence: democratic regulation of 73 non-state 72–3 see also force voting system 101 Wahl, P. 164 Wallace, T. 68, 198, 217 Waltz, K. N. 54 Walzer, M. 44, 90, 208–9 Wamai, R. 78 Warren, P. 95 Weber, M. 51 Wendt, A. 55 Westphalian political order 91 Wickham, J. 67 Willet, T. 167 Willets, P. 67 Wolﬀ, R. P. 36, 44, 50 Wood, A. 128, 188 Woods, N. 45, 178, 218 Worker Rights Consortium 68 World Bank 197 World Economic Forum 2 World Rainforest Movement (WRM) 212 World Trade Organization (WTO) 1, 66–7 World Vision 205 World Vision International 77 Wright, E. O. 199 Young, I. M. 8, 113, 166, 170 Zadek, S. 68 Zürn, M. 56