Functional Representation and Democracy in the EU The European Commission and Social NGOs Corinna Wolff
Corinna Wolff 2013 First published by the ECPR Press in 2013 The ECPR Press is the publishing imprint of the European Consortium for Political Research (ECPR), a scholarly association, which supports and encourages the training, research and cross-national co-operation of political scientists in institutions throughout Europe and beyond. ECPR Press University of Essex Wivenhoe Park Colchester CO4 3SQ UK All rights reserved. No part of this book may be reprinted or reproduced or utilised in any form or by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying and recording, or in any information storage or retrieval system, without permission in writing from the publishers. Typeset by Anvi Printed and bound by Lightning Source British Library Cataloguing in Publication Data A catalogue record for this book is available from the British Library ISBN: 978-1-907-301-65-0 www.ecpr.eu/ecprpress
Chapter Two What Does it Mean to Represent? Representation as such need not be political nor democratic. Following Hanna Pitkin s classical definition of representation as the making present in some sense of something which is nevertheless not present literally or in fact (Pitkin 1967: 8 9; emphases in original), 1 we can speak of political representation when a participant in the political process (widely understood) is participating by virtue of also giving political presence to actors other than herself. Such representation is democratic if it ensures input and output legitimacy as well as accountability in relation to the represented. These criteria are most familiar to us in the context of elected parliaments and governments. Yet, the political stage is also filled with unelected actors perceived or claiming to represent various functional constituencies. How are we to evaluate the assertion that the European Anti-Poverty Network (EAPN) represents socially excluded Europeans, given that their staff is not elected by their supposed constituents, nor is their organisation even known to most of them? Intuitively, many will still recognise them as representing the poor in some way. At the same time, it is obvious that the EAPN represents excluded people in a different way than a Member of Parliament represents either her constituency or the nation as a whole, and that EAPN s representing also differs from that of an equally unelected monarch. So how does it work? The question is a fundamental one, because the intended democratic value of involving such actors in governance processes rests first and foremost on the extent to which they legitimately represent [ ] some group or larger set of social interests (Saward 2005: 179). In a democracy, those taking part in governance processes do not in general participate on their own behalf, but speak for a larger group of citizens. Representation is thus the key to studying the democratic credentials of the involvement of functional organisations like EAPN in democratic policy making. Scrutinising representation is the prerequisite to dealing with questions of democracy, such as, for example, should and can functional actors be accountable, and if yes, in what way and to whom? These issues are discussed in the following chapters. In the present chapter I consider the question of how we know if someone represents a given constituency. I look at different theories of representation and the concepts and criteria they use, to assess if and in what kinds of ways unelected actors on the political stage in particular can legitimately be said to represent a certain functional constituency. 1. Plotke (1997: 27 30) criticises Pitkin for linking representation with absence and argues that it is plausible to say that in a given context, someone is present politically but not physically (Plotke 1997: 30). Although I am not entirely convinced by his reading of Pitkin in this case, I find this latter distinction clarifying and helpful.
24 Functional Representation and Democracy in the EU Chapter Three looks at the special case of the descriptive representation of disadvantaged groups, i.e. their representation by people who themselves belong to these groups. Descriptive representation has become an important issue for many enquiries in democratic representation, because the question of whether and in what sense a representative should resemble the represented, and what difference this makes for representation, goes to the core of what representation itself is about. What is more, descriptive representation is one mode of functional representation that is used in the social OMC. Chapter Four changes the angle of observation to address the question of in what ways functional representation can make governance processes more democratic. It works out the main issues and difficulties we encounter when analysing governance from the perspective of democratic legitimacy, and looks at the ways in which the different theories of representation relate to these issues. Actors who are representing functional constituencies are in most cases not elected by the represented. How, then, can we tell whether someone is legitimately representing a certain group of people? Legitimate in this context does not necessarily mean democratic legitimacy; it merely refers to whether an actor is accepted as a representative of a particular constituency. Different criteria are possible for determining this. In the following, I present three perspectives, each of which interprets in different ways what it means to represent. The first notion is based on a simple principal agent model. The second view builds on this, but widens it with deliberative concepts of democracy, and emphasises the role of ongoing communication between the representing and the represented. Thirdly, the theory of the representative claim challenges the fundamental assumptions of the former two perspectives and proposes a new view of what representation is about, while at the same time providing the concepts to integrate all three theories into one framework. Simple principal agent models 2 Simple principal agent models are a theory of delegation. 3 Because the people cannot govern themselves, they delegate this task to agents. At the same time, however, the people remain the sovereign, i.e. the principal who is in control of her agents. 4 The agent needs to be authorised by the principal, but, mainly for 2. I speak of simple principal agent models because the application of these models is here restricted to their core elements that are relevant for the purpose of the study. Principal agent theories serve here as an analytical framework to understand how political representation in general comes about, rather than being meant as an accurate description or prescription of how political systems function in practice. For more complex accounts, see for example the contributions in Przeworski et al. 1999. 3. In this sense, the terms principal and agent were first used by Alexander Hamilton in The Federalist, synonymously with those of master and servant (Hamilton 1788: 379 80). For the historical context, see Grimm 2009: 38 9. 4. From a different angle, principal agent theory can also be seen as originating in the self-government vs. rule dichotomy, i.e. the question of whether representation is democratic at all. In this interpretation, self-government is equated with direct democracy, referring to the Athenian roots
What Does it Mean to Represent? 25 practical necessities (Dunn 1999: 342 3), she has a free mandate. Therefore, a central role is given to political accountability. Thus, we can say that someone legitimately represents a given constituency if that constituency has chosen and authorised her to act in their interest (Manin et al. 1999: 2 3, referring to Pitkin 1967; Przeworski 1999: 32), and if she is accountable to them. In realising these principles, the key role is usually given to elections. Since they are recurrent, they merge the act of authorisation and the exercise of accountability together, thus providing, simultaneously, an incentive for the agents to be responsive to the preferences of their constituents, and an instrument of control for the principals. At the same time, elections create an abstract equality among the people (one person, one vote) and thus also function to establish procedural fairness. In order for elections to fulfil their role, they need to be supplemented with other procedural norms, such as the freedom of information (Dunn 1999: 339 41), of opinion, and of expression (Manin 1997: 174 5, 237). Transparency in particular is an important prerequisite for principals to be able to make enlightened choices and effectively control their agents. As long as the two core criteria of authorisation and accountability are met, it is irrelevant who does the representing, because any representative will be induced by elections to act in the interest of the represented, and can be replaced if they perceive her to have failed to do so. Consequently, the main issue is not who the representatives are, but what they do, 5 and the former is not expected to have any decisive impact on the latter (Williams 1998: 138 9). This does not mean, though, that every participant in the political process has to be elected by the people. The legitimacy of functional representatives rests on two pillars: the internal legitimacy of their organisation and external legitimacy via the democratic chain of delegation. Internal legitimacy means that the functional representative legitimately represents her constituency, i.e. she is authorised by and accountable to them. Consequently, functional representatives can usually legitimately represent only membership-based organisations with formal structures. They can hardly be said to speak for those who are not organised, and can therefore mostly not legitimately represent an entire functional constituency, defined as all those affected. External legitimacy refers to whether it is legitimate to involve a functional representative in a political process, for which she has not been authorised by the principals of that process. In other words, under what conditions is she legitimately of democracy. Representation, on the other hand, is interpreted as separating the governed from the governing; it is thus not self-government, but mainly a substitute for direct democracy, which has become impossible in large-scale societies. Representative democracy, then, is not genuine self-government but rule, which is intrinsically humiliating and potentially harmful (Dunn 1999: 342). The classical question in this paradigm is what makes this rule legitimate? and the answer is: choosing the rulers, who through this become agents of the people. In other words, representative democracy is an oligarchy with elected rulers (Manin et al. 1999: 4 5; Sartori 1987: 111). 5. Cf. Pitkin 1997: 90; and cf. Phillips (1995: 4 5), who has termed this the politics of ideas, as opposed to the politics of presence (see Chapter Three).
Chapter Four Functional Representation and Democracy One of the main difficulties in dealing with the democratic aspects of governance processes is that settings tend to be complicated. They involve a variety of different actors on different levels, and the functional representatives among these are often unelected and represent fluid constituencies of mostly not formally organised citizens. In Chapter Two I have worked out the tools representation theories provide us with to deal with such complex contexts. Theories of representation are especially equipped to deal systematically with the relations between the different actors involved in particular, functional representatives, the members of their organisations, their wider functional constituencies (the affected), and the citizenry of the polity. They offer conceptual tools to tackle the question of who should be accountable to whom, and they lay down the constitutive role of the represented as the basic principle of democratic representation. Concerning the ways to realise this principle, they emphasise different aspects. Principal agent models stress the possibility of effective sanctioning ( throwing the rascals out ) as well as the procedural equality of the represented. Deliberative representation adds substantial fairness and the quality of deliberation between representatives and representees. The representative claim shifts the focus to how representation is built in rhetorical ways. It opens the concept of representation up to also accommodate new, evolving forms and understandings, while confirming the recognition by the constituency as the key benchmark for democratic representation. In other words, each perspective is based on slightly different concepts of what makes political representation democratic. The purpose of this study is not to test any of these theories, but to use them as analytic tools in studying the concepts underlying the Commission s argumentation. In this chapter, I use the tools of these representation theories to elaborate on in what ways governance processes involving functional representatives can be democratic, and what the critical issues are in this regard. I do this by analysing functional representation in policy making through the three dimensions of my working model of democracy, i.e. input legitimacy, output legitimacy, and political accountability (cf. Chapter One). Democratic legitimacy in governance is looked at from two perspectives: firstly, the democratic legitimacy of the whole political process, insofar as this depends on the functional representatives involved (i.e. legitimacy gained through functional representation); and secondly, the internal legitimacy of these representatives vis-à-vis their constituencies (i.e. legitimacy of functional representatives), which is of course an important part of the former aspect. Together with the previous two chapters, this provides the fundaments to analyse practices of functional representation as well as the concepts used to shape them.
56 Functional Representation and Democracy in the EU Output legitimacy Output legitimacy is usually seen as the key strength of governance, in comparison to traditional government. 1 It has also been a primary perspective of the governance literature. The basic assumption in analysing functional representation from this viewpoint is that the choice of participants in policy making has an impact on the outcomes. Against the background of the debates on overloaded government and the ungovernability of society (Peters and Pierre 2006: 31 7), the roles of various actors have been analysed with respect to the potency of different arrangements to deliver public goods or steer society. Output legitimacy in this context means better quality of political decisions as well as reaching them more efficiently. Output legitimacy is seen as a characteristic of the political process and the choice of participants, rather than being associated with the internal organisation of the involved actors. There are two main arguments for the enhanced output legitimacy of involving functional representatives. Firstly, functional actors control resources that are needed in policy-making processes and that governmental actors lack, such as information, technical expertise, and the capacity to ease implementation and improve compliance by communicating decisions to their constituencies. It would therefore be wrong to start from the premise that policy makers were struggling to resist functional organisations that are vigorously knocking on their door (Maloney et al. 1994: 19 22). On the contrary, executive administrations need functional representatives for their own success. The second main factor to which output legitimacy is attributed is based more on the reasoning of deliberative democracy and on the main mode of policy making in governance, i.e. negotiation. As the study of Sonja Wälti et al. has shown, negotiation in networks has a (potential) ability to mediate and overcome particularistic interests and enhance solidarity and community-building, and, on the other hand, to provide forums of exchange and mutual learning (Wälti et al. 2004: 107). Output legitimacy can thus also include fairer outcomes. Recently, the optimism concerning the potentials of governance in terms of output has been criticised for its underlying rationalistic problem-solving orientation, which tends to render other democratic values instrumental to the production of rational outcomes (Buchstein and Jörke 2007: 186 9; Greven 2007). It has been stressed that output legitimacy is not just performing well (Schmitter 2002: 57). In particular, it not only implies delivering results that citizens endorse, but also means that the represented support or at least accept the goals that are pursued (Christiansen et al. 2003: 13; Kohler-Koch and Rittberger 2006: 41; Bellamy 2010; cf. Wälti et al. 2004: 104). Hence, instead of simply equating 1. It is worth noting that it is somewhat questionable to compare the output legitimacy of these two regime types. Given the differences in historical context, including changes in the expectations towards public authorities and in the ways of conceptualising democratic legitimacy, it becomes problematical what criteria to apply. I seek to avoid such comparisons here and restrict my analysis to evaluations from a present-day perspective.
Functional Representation and Democracy 57 output legitimacy with efficiency, attention has to be given to how the goals, which are then realised as outputs, have been defined, and more generally, who has had a chance to influence the process leading to particular outputs (cf. Føllesdal and Hix 2006: 544 6; Bellamy 2010). Such a stronger definition links output legitimacy tightly to input legitimacy: While the democratic idea of the political process depends on the contingent articulation of the preferences of equal citizens or their representatives which inherently and incrementally leads to the construction of political problems and a possible agenda, the new approach of participatory governance rests on the very traditional premises of technocratic politics. In this framework, problems are given or defined by governance agencies and framed before the problem-solving process has even started. (Greven 2007: 242) Furthermore, there are also trade-offs between output legitimacy on the one hand and input legitimacy and accountability on the other hand. This is especially the case if output is conceptualised mainly in terms of efficiency, because broad access to group negotiation may conflict with the effectiveness of these negotiations (Immergut 1995: 204; cf. Dahl 1994). B. Guy Peters and Jon Pierre therefore speak of a Faustian bargain in the sense that core values of democratic government could be traded for accommodation, consensus and the purported increased efficiency in governance (Peters and Pierre 2004: 85; cf. Peters and Pierre 2006: 37). Hence, when analysing the output equality of a process, attention has to be paid to the definition of the concept and its balance with the other two dimensions of democracy, including particularly how participants are chosen and how the process is carried through (e.g. publicity, openness to alternatives). 2 Closely related to this, the above-mentioned accommodation and consensus is another alleged virtue of governance that has come under criticism. As such, there is no reason to assume that the affected parties views should be compatible. Rather, being affected presumably entails having an interest in the matter, which may well be opposed to the interests of other actors involved. If the main goal is reaching a consensus efficiently, there is a danger that actors outside the main players mainstream thinking are excluded from governance networks (Wälti et al. 2004: 101 7; Wälti and Kübler 2003; Papadopoulos 2007: 481). In other words, accommodation is not an inherent system characteristic, but something that has to 2. The relation of output legitimacy with other democratic values has been a prominent topic of discussion in the debate around the democratic legitimacy of the EU. The father of the concepts of input and output legitimacy, Fritz W. Scharpf, argues that the legitimacy of the Union can be founded primarily on output legitimacy as long as it deals only with uncontroversial issues that do not require zero-sum redistribution (Scharpf 1999: 22 3; for a related argument, see Majone 1998). More recently, the notion that a lack of input legitimacy (and accountability see Føllesdal 2011) can be compensated for by output legitimacy has come increasingly under fire. In parallel, the argumentation for a balance of input and output legitimacy has been turning to conceptualising the issue more from a perspective of representation. See, for example, Føllesdal and Hix 2006; Bellamy and Castiglione 2011.
58 Functional Representation and Democracy in the EU be worked for, 3 and that can be arrived at in different ways, for example through inclusive deliberation aimed at learning and mutual understanding, or through restricting access. Such different strategies affect the outcomes of the process. In particular, there is a risk that the interests of those excluded from policy making are not taken into account, or, worse, that negative externalities of decisions are primarily imposed on those groups who are not involved. 4 This leads us to the concept of outcome equality, i.e. the idea that not only should everybody have the possibility of equal input, but also the outputs should treat everybody with equal concern (cf. Phillips 2004). Put differently, when looking at the output legitimacy of a process, one should also ask whom the outputs benefit. All this suggests that the output legitimacy of functional representation is conditional on a number of factors, such as pluralistic enough governance networks (Wälti et al. 2004: 105, 107; Benz and Papadopoulos 2006: 10) with cooperative participants holding the required resources. These conditions can be compatible with each other and with other norms of democratic legitimacy, but they do not need to be. Trade-offs can also affect the involved associations internal organisation. For example, if participants are mainly chosen on the grounds of efficiency as perceived by executive officers, this tends to favour highly professionalised organisations with only thin supporter participation and control (cf. Maloney 2008; Saurugger 2006; Sudbery 2003). By adapting themselves to such expectations, functional associations can increase their influence, yet at the cost of their input legitimacy and accountability. At the same time, less professional and more participative organisations are disadvantaged. In view of this criticism of overly output-oriented perspectives, one should, however, be careful to avoid the other extreme and neglect the significance of output legitimacy for democracy. Besides the obvious fact that listening to citizens and being accountable to them does not amount to much if no approved outputs are produced, output legitimacy is also essential for the legitimacy of political rule and the system as a whole. As well as that, high levels of welfare for broad parts of the population reinforce and sustain attitudes necessary for democracy (Greven 1993: 404 7). In sum, when looking at arguments building on the output legitimacy of functional representation, it is important to examine first of all how output legitimacy is conceptualised in relation to the represented (for example, efficiently achieving conditions deemed beneficial for them or pursuing such policies as the represented wish). Secondly, we have to ask in what relation this stands to the 3. In view of the predominantly optimistic governance literature, in particular with respect to the expected harmoniousness of processes, Pierre (2000b: 245) suggests that conflict management and resolution in governance, as well as governance failure, should be studied more. 4. The contributions in Government and Opposition 46(1) therefore argue that functional representation should be tightly controlled by a system of territorial representation, in order to make sure that the public interest and unrepresented interests are taken into account (Bellamy and Castiglione 2011; Føllesdal 2011; Héritier and Lehmkuhl 2011; Weale 2011). See also Bellamy 2010; Papadopoulos 2010: 1040.