Transparency Mechanisms: Building Publicness Into Public Services

JOURNAL OF LAW AND SOCIETY VOLUME 28, NUMBER 4, DECEMBER 2001 ISSN: 0263-323X, pp. 471–89 Transparency Mechanisms: Building Publicness into Public Services Lindsay Stirton* and Martin Lodge** Recent changes in patterns of public service provision, sometimes associated with the `regulatory state', have been said to have eroded citizenship and diminished accountability. This paper responds to these challenges by outlining a toolbox of four transparency mechanisms – information, choice, representation, and voice – as alternative devices that can be built into the architecture of public service regimes, to increase responsiveness and answerability. Using insights drawn from cybernetics and transaction cost analysis, this paper looks at the consequences of different choices of combinations of mechanisms in allocating authority in line with competing administrative doctrines of fiduciary trusteeship and consumer sovereignty. Attention is drawn to differences in `cost profiles' between different public services that can facilitate or inhibit consumer choice as a basis for understanding the suitability of different combinations of mechanisms to specific public services. A contingency model determining the suitability of particular mechanisms to particular services of different `cost-profiles' is presented. Given the variety of public services and among different public service architectures in the regulatory state, it is argued that this differentiated approach to transparency and accountability provides a more effective response to holding public services accountable than narrower traditional notions of political accountability. * Department of Law, London School of Economics and Political Science, Houghton Street, London WC2A 2AE, England ** Centre for Analysis of Risk and Regulation, London School of Economics and Political Science, Houghton Street, London WC2A 2AE, England Financial support for this research was provided by an Association of Commonwealth Universities/British Academy grant for international collaboration and by the Ford Foundation Endowment to the Social Sciences, University of the West Indies. Tony Prosser and Jurgen de Wispelaere both provided helpful comments and suggestions. For the remaining deficiencies and shortcomings each of the co-authors holds the other accountable. 471 ß Blackwell Publishers Ltd 2001, 108 Cowley Road, Oxford OX4 1JF, UK and 350 Main Street, Malden, MA 02148, USA INTRODUCTION The past three decades have seen a considerable change in the relationship between the state and the citizen. One of the key areas of visible change has been in the provision of public services. Public services are crucial to facilitating an individual’s economic activities and are thus central to the modern understanding of citizenship.1 These changes, which have been part of the so-called deregulation agenda and the introduction of New Public Management, have arguably led to a shift from the ‘positive’ to the ‘regulatory’ state.2 This shift towards the regulatory state is said to involve a shift from public to private ownership of public utilities and other social service infrastructure, an increasing emphasis on pro-competitive regulation by quasi-autonomous independent agencies, and a shift on the style of the organization of the ‘core’ public sector. This has also included the separation of policy-making and service delivery functions, the ‘reading-over’ of private management practices into public services and the emphasis on securing performance of service delivery through contracts. One consequence of these changes has been a notable increase in the diversity of arrangements and institutions through which public services are provided. It is often argued that these changes have often been made without careful regard to the accountability of public services under these decentralized arrangements. Martin Loughlin, for example, claims that these reforms to public services were part of ‘a strategy driven and fashioned almost entirely by a political-economic impetus and with virtually no legal or constitutional consciousness.’3 Most of the ‘constitutionalist’ critiques of the regulatory state have been in terms of the lack of accountability and transparency of 1 This contribution does not deal directly with the important question as to what counts as a public service. This issue has never been adequately dealt with within the framework of British constitutional system or its colonial derivatives where public service has for half a century been equated with public ownership, control, and financing of services. In a number of European jurisdictions, as well as in EU law there is a substantial body of law attempting to define the concept of public service, although once the notion is uncoupled from public ownership or natural monopoly, it does not offer much guidance. For an interesting argument concerning the incorporation of the European concept of service public into British law, as an unintended consequence of privatization, see T. Prosser, ‘Public Service Law: Privatization’s unexpected offering’ (2000) 63 Law and Contemporary Problems 63. 2 For the literature on the regulatory state see, for example, P. Day and R. Klein, ‘The Business of Welfare’ New Society, 19 June 1987; G. Majone, ‘From the Positive to the Regulatory State: Causes and consequences from changes in the modes of governance’ (1997) 17 J. of Public Policy 139; G. Majone, ‘The Regulatory State and its Legitimacy Problems’ (1999) 22 West European Politics 1; M. Loughlin and C. Scott, ‘The Regulatory State’ in Developments in British Politics, eds. P. Dunleavy et al. (1997) 5; C. Scott, ‘Accountability in the Regulatory State’ (2000) 27 J. of Law and Society 38. 3 Quoted in Scott, id., p. 40. 472 ß Blackwell Publishers Ltd 2001 public service arrangements.4 This issue has recently come to the fore in the international reform agenda as international technical bureaucracies such as the OECD and the World Bank have come to see transparency as a requirement of good governance in recipient countries.5 The activities of Transparency International, a Berlin-based international non-governmental organization, in promoting transparency as a key component of anticorruption reforms, has also served to place the issues of transparency and accountability at the centre of the debate of the reform of public services.6 This article attempts to redress that balance somewhat, by injecting a normative and constitutional perspective into the debate about the institutions for the provision of public services. We take a pragmatic look at the problem of holding public services accountable, recognizing the persistence of the political-economic realities that have given rise to the regulatory state.7 In particular, we argue that any approach to accountability in the new public sector must recognize the increased diversity of arrangements for public service provision. What this paper offers is an approach to structuring the complexity of governance arrangements in order to build in ‘publicness’ into their organization. Our approach is as follows. In the next section we address the relationship between the normative and analytical aspects of our argument. We then set out a ‘toolbox’ of four different transparency mechanisms – information, choice, representation, and voice – which can be incorporated into so-called ‘public service architectures’ in various ways. This ‘toolbox’ of basic transparency mechanisms allows us to explore important variations in the way in which public service architectures promote transparency and accountability. It is argued that through the use of these mechanisms rather than through state ownership or control per se that the public character of public services is maintained. After outlining the four fundamental mechanisms we review (and ultimately reject) the view that the rise of the regulatory state represents a fundamental shift in the nature of the interests of citizens which society recognizes as legitimate. Whether the regulatory state promotes these interests effectively depends on a complex set of choices. In particular, two distinct but connected choices are highlighted: the first concerns the way different mixes of transparency mechanisms are incorporated into these architectures in order to better promote the interests of citizens. The second concerns the emphasis given to alternative centres of control – public authorities or service users – by contrasting public service 4 See, for example, id.; C. Graham, Is There a Crisis of Regulatory Accountability? (1997); Majone, op. cit. (1999), n. 2. 5 World Bank, Governance: The World Bank’s Experience (1994); World Bank, World Development Report: The State in a Changing World (1997). 6 F. Galtung and J. Pope, ‘The Global Coalition Against Corruption: Evaluating Transparency International’ in The Self Restraining State: Power and Accountability in New Democracies, eds. A Schedler et al. (1999). 7 See, also, Majone, op. cit. (1999), n. 3. 473 ß Blackwell Publishers Ltd 2001 architectures. Insights from cybernetics8 and principal-agent analysis, it is argued, can clarify the nature of these choices, by highlighting the respective transaction costs exercising control within alternative architectures. In conclusion, we draw out some of the implications of this approach, indicating circumstances under which different mixes of transparency mechanisms most effectively promote the public interest in respect of different services. Thus, while we argue that regulatory regimes should be ‘transparent’, choices as to what can be regarded as desirably transparent public services are based on distinct assumptions and involve particular trade-offs. What emerges therefore is an advocacy of a good (transparent) public service which reflects particular circumstances. TRANSPARENCY AND ACCOUNTABILITY To what extent are institutional mechanisms of accountability and transparency ethically desirable or even ethically necessary in a liberal democracy? A useful starting point for an inquiry into the ethics of accountability is a contribution by Guillermo O’Donnell which provides the theoretical anchor for an important recent collection on accountability in new democracies.9 O’Donnell makes an important distinction between vertical and horizontal accountability. The main instruments of vertical accountability, on this account, are elections but instruments such as the articulation of social demands by the electorate and scrutiny of official acts by the media also fall into this category. The description ‘vertical’ applies because these are instruments through which individuals and groups affected by the actions and decisions of those in authority pass judgement on the exercise of authority. Horizontal accountability, on the other hand, refers to: . . . the existence of state agencies that are legally enabled and empowered, and factually willing and able, to take actions that span from routine oversight to criminal sanctions or impeachment in relation to actions or omissions by other agents or agencies of the state that may be qualified as unlawful.10 The important point raised by O’Donnell is this: while it may be taken as more or less self-evident that democratic principles justify vertical accountability, strong democracies also rely on mechanisms of horizontal accountability to make them effective. Beyond this basic (if important) 8 Defined as control theory as it is applied to complex systems. See entry on cybernetics at Britannica.com <http://www.britannica.com> (site visited 1 August 2001). 9 G. O’Donnell, ‘Horizontal Accountability in New Democracies’ in Schedler, op. cit., n. 6. 10 id., p. 38. We agree with Schmitter’s criticism that this formulation somewhat arbitrarily restricts the notion of horizontal accountability to state agencies of oversight and to illegality as the triggering mechanism (see P. Schmitter, ‘The Limits of Horizontal Accountability’ in Schedler, op. cit., n. 6). 474 ß Blackwell Publishers Ltd 2001 point, however, political theory approaches such as this provide no way of distinguishing between circumstances in which direct responsiveness to individual preferences – vertical accountability – is adequate, and those in which mechanisms for horizontal accountability are needed to enrich democracy: such detailed prescription requires an institutional turn. According to Hardin, ‘. . . we may say of an institution that its morality is reasonably defined by its purpose and likely effects . . . Institutional morality is similar to some variants of virtue theory: its content is functionally determined by the purpose it is to serve.’11 A normative analysis of issues of transparency in the new public sector requires a justification of the purposes that these are supposed to serve, as well as an analysis of how well those purposes are served by different instruments. Furthermore, the manner in which purposes are related to effects poses the most taxing questions. It is for this reason that normative analysis needs to take seriously arguments about the (un)governability of the public sector,12 as well as give thought to the instruments through which public purposes are hoped to be achieved. Clearly, the concept of transparency is closely related to notions of accountability, though the two are not identical. It is useful to compare the two ideas as a starting point for a discussion of the moral requirement of transparency. Andreas Schedler sees accountability as a two-dimensional concept, embracing what he calls ‘answerability’ or the monitoring and justification of the exercise of political power as well as ‘enforcement’ or the threat of sanctions for abuses of the exercise of power.13 It is tempting to see transparency as the goal of answerability in this sense.14 Although this captures some of the distinction, in our view transparency is a slightly wider concept than this, therefore requiring some additional refinement. Simply knowing what public service providers have decided and why they think they made a good decision is not a particularly valuable form of transparency, especially if citizens cannot exercise some input into public decisionmaking. We prefer to see the focal meaning of transparency as the existence of a two-way street in which public services may be described as transparent when they are responsive to service users as well as answerable to them. Indeed, answerability, which to Schedler requires agents to explain their decisions as well as to inform us about them, may imply some degree of 11 R. Hardin, ‘Institutional Morality’ in The Theory of Institutional Design, ed. R. Goodin (1996) 126–7. 12 See R. Mayntz, ‘Governing Failures and the Problem of Governability: Some comments on a theoretical paradigm’ in Modern Governance: New governmentsociety interactions, ed. J. Kooiman (1993). This has also been a crucial part of the ‘New Right’ critique of the concept of social justice. See F.A. Hayek, Law, Legislation and Liberty: a new statement of the liberal principles of justice and political economy vol. II (1982). 13 A. Schedler, ‘Conceptualising Accountability’ in Schedler, op. cit., n. 6, pp. 14–17. 14 As indeed Schedler does: ‘Accountability as answerability aims at creating transparency’ (id., p. 20). 475 ß Blackwell Publishers Ltd 2001 transparency in this sense if there are real limits on the range of activities that are capable of explanation. The idea of responsiveness is captured in the cyberneticians’ classification of essential capabilities of a control system. Such a system requires instruments for effecting change in the state of the world (generally referred to as ‘effectors’) as well as ‘detectors’ for providing data about the state of the world.15 Transparency in its fullest sense thus requires that citizens be able to exert an influence on (to ‘control’) the way that public services are provided, based on their views or preferences about how they are provided, as well as knowing about the decisions that are made. Transparency can be understood to serve two separate but related functions. The first is to ensure that public service providers respect both the positive and the negative rights of individuals. This instrumental justification for transparency of public services comes close to Bentham’s principle for good governance: ‘The more strictly we are watched, the better we behave.’16 Control over service providers is necessary to ensure the effective provision of services and to prevent abuse of authority by public officials. The second purpose relates more directly to democratic theory, which values participation by individuals in the decisions which affect them. Young expresses this point thus: Participative approaches to democratic theory hold that democracy is a hollow set of institutions if they only allow citizens to vote on representatives to far away political institutions and protect those citizens from government abuse.17 Transparency, on this view has moral value because it enhances individual autonomy by involving citizens directly in the process of making decisions which affect their lives and interests. Further, transparency enhances individual autonomy to the extent that transparent institutions are predictable, allowing individuals to order their own private choices knowing the way that these are affected by public decisions. The remainder of this article is an attempt to address this issue of how the design of public services can be made to promote transparency. Specifically we aim to argue that welfare rights and individual autonomy can be effectively promoted within the regulatory state, given appropriately designed public service architecture. Thus, we will first look at different mechanisms which are available to enhance transparency, then apply competing doctrines to this transparency ‘toolbox’ to highlight the substantial differences in terms of standard-setting authority and effecting and detecting instruments. In accepting Hardin’s argument that the morality of an institution must be understood in terms of its purpose and effects, it has 15 See C. Hood, The Tools of Government (1983) 3–4. 16 Quoted in C. Hood et al., Regulation Inside Government: Waste-watchers, quality police and sleaze-busters (1999) 3 and discussed at 5–6. 17 I.M. Young, ‘Political Theory: an overview’ in A New Handbook of Political Science, ed. R. Goodin and H.-D. Klingemann (1996). 476 ß Blackwell Publishers Ltd 2001 to be realized that the effects may be quite different in different circumstances. We therefore advocate a contingency approach to transparency according to which circumstances are identified in which alternative mechanisms will better promote transparency. Bringing a contingency approach to bear on problems of the normative analysis of institutions can be seen as an extension of the approach of the legal philosopher Lon Fuller in asking what are the forms and limits of different forms of social ordering.18 A FIRST GLANCE AT THE TOOLBOX: FOUR BASIC TRANSPARENCY MECHANISMS Before it is possible to make any claims as to how transparent a public service is, it is necessary to establish particular criteria for evaluation. This section illustrates four broad transparency mechanisms: voice, representation, choice, and information. This categorization may be seen as an expansion of Hirschman’s dichotomy of exit and voice.19 ‘Voice’ promotes the exercise of individual contributions to and redress from the provision of public services, while ‘choice’ includes all types of possibilities through which users can exercise ‘exit’, mostly by selecting other providers, but also in terms of choosing different types of political leadership. ‘Information’ facilitates the quality of user knowledge, enhancing in particular the exercise of voice and choice. ‘Representation’,20 in contrast, aims to provide an institutionalized interest for the service user in regulatory and wider policy decision-making in order to compensate for the potential over-representation of competing interests.21 Table 1 offers a brief survey of the ‘toolbox’ and its ‘transparency mechanisms’.22 18 L. Fuller, ‘The Forms and Limits of Adjudication’ (1978) 92 Harvard Law Rev. 353. 19 See A.O. Hirschman, Exit, Voice and Loyalty (1970). 20 This is similar to William Bishop’s discussion of ‘complaint’, that is, appeal to some controller who can exercise effective sanction, as distinguished from voice in Hirschman’s sense, which involves a direct ‘say’ in decisions. See W. Bishop, ‘Agency Cost and Administrative Law’ in A New Palgrave Dictionary of Economics and the Law, ed. P. Newman (1997). 21 See I. Ayers and J. Braithwaite, Responsive Regulation: Transcending the deregulation debate (1992). 22 In engineering science a mechanism is a physical system that can deform without the application of force. The term ‘mechanism’ is sometimes used in the social sciences to refer to describe complex patterns of causal relationships incorporating multiple causal chains (see collection in P. Hedstrom and R. Swedberg (eds.), Social Mechanisms: an analytical approach to social theory (1998)). There is some similarity with the approach taken here given the multiplicity of effects that particular transparency mechanisms can have in different situations. We use the term ‘mechanism’ in a slightly more restricted sense to explore alternative ways in which the abstract notion of transparency can be transposed into regulatory regimes and the effects that this has on the architecture of public services. 477 ß Blackwell Publishers Ltd 2001 Table 1: Classification of Transparency Mechanisms Input-oriented Output-oriented Individually exercised Collective provision Voice: Representation: Prevents lack of user consideration/responsiveness. Prevents dominance of producer/large user groups. Enables user participation and redress. Enables inclusion of ‘public interest group’ objectives. Via political and/or providers’ redress mechanisms, consultation and decision-making. Via consumer councils, consumer groups, regulatory objectives and agencies. Example: Complaint forms available for train services. Regulator’s road shows in the provinces. Example: ‘Energywatch’ as institutionalized consumer representative in the UK. Choice: Information: Prevents paternalism, exposure to monopoly. Prevents ‘lemon choices’ and other abuse of asymmetric information. Enables consumer selection of nature, quality and quantity of service/good. Enables sophisticated consumer choice. Via market competition, competition for the market, elections, contracts. Via benchmarking, naming and shaming, information centralization, competitive comparative advertising. Example: Sudden appearance of differently shaped and coloured phone sets as part of early telecom liberalization. Example: Chinese rule that forced physicians to display the number of patients who died in their care.23 These mechanisms differ in the extent to which they are exercised on an individual basis or are provided collectively. Information and representation are predominantly collective mechanisms, operating to make public services transparent to groups of service-users en masse, while voice and choice are mechanisms that can be individually exercised. Input-oriented mechanism point to an emphasis placed on process, while output-oriented mechanisms affect mainly the choice of product. In their basic form, these four mechanisms can be harnessed to a range of different substantive goals and could be utilized with different intentions and effects. Later sections discuss 23 C. Hood, Administrative Analysis: an introduction to rules, enforcement and organisations (1983) 178–9. 478 ß Blackwell Publishers Ltd 2001 such variety by introducing two competing doctrines of public service provision. We do not aim to proffer a fully exhaustive and mutually exclusive set of transparency mechanisms, but rather utilize this basic ‘toolbox’ to further the debate about the architecture of public services. Transparency mechanisms offer a crucial way in which the rules of the game are employed to enhance the ‘publicness’ of the regime and therefore have wider implications for the relationship between state and citizen in the delivery of public services. In understanding how the deployment of different transparency mechanisms can affect the values that are served by public service institutions, we need to understand the nature of the different transparency mechanisms not in isolation, but within the overall service regime. The following sections discuss how this ‘toolbox’ offers meaningful statements with regard to the building of publicness into the architecture of public services. First, we discuss whether the choice of transparency mechanisms reflects statements about citizenship and forms of social ordering. Second, we question how the application of two distinct doctrines for the design of public services impacts on the emphasis given to the four transparency mechanisms, their application as policy instruments, and their means to allocate authority. Thirdly, we consider distinct costs and benefits which are involved in the choice of allocating authority. ARCHITECTURES OF PUBLIC SERVICE As noted in the introduction, one key tenet in the literature on contemporary changes to the provision of public services has highlighted the impact of such shifts on the nature of citizenship. On this account, the administrative reforms of the last twenty years have a deep normative significance, representing a shift from a predominantly ‘political’ conception of citizenship towards an understanding of citizenship as economic agency. In the former, the users of public services are identified as the benefactors of certain positive obligations of the state. The economic agency conception of citizenship has been criticized as being somewhat narrower in scope than the ‘political’ conception, a denial that citizenship rights encompasses the social rights discussed earlier.24 Such ‘rights’ as are admitted by the economic agency account are essentially understood as analogous to the contractual rights of consumers in private law, as the quid pro quo obligation of the service provider to those who pay for public services. The view that a shifting emphasis among transparency mechanisms represents a restriction of the entitlement of citizens makes sense in terms of our approach outlined earlier. If the architecture of the new public sector has 24 See A. Barron and C. Scott, ‘The Citizens’ Charter Programme’ (1992) 55 Modern Law Rev. 526; J. Stewart, ‘The Rebuilding of Public Accountability’ in Accountability to the Public, ed. J. Stewart et al. (1992). 479 ß Blackwell Publishers Ltd 2001 diminished capacity to promote the rights and interests of citizens, then it follows that these rights have also been affectively diminished. However, the view that a ‘political’ conception of citizenship exists which competes with the ‘economic agency’ conception, with the latter being in the ascendancy, can be challenged both empirically and conceptually. Empirical evidence suggests that far from abandoning its responsibilities and obligations for the provision of public service, the state still maintains substantial obligations and tasks. The shift from ‘positive’ to ‘regulatory’ state has, rather, involved a shift in instruments and means to establish similar outcomes. In many countries, both in the developed and the developing world, market-based mechanisms for securing responsiveness to citizen preferences have supplemented, rather than supplanted traditional mechanisms of political accountability.25 Furthermore, regardless of the means of delivering public services, these do not deflect or deny the continuing obligations and functions of a public authority (the state) vis-à-vis its citizens. Consequently, we reject the underlying normative significance between political and economic conceptions of citizenship. Rather, we see voting and contracting as alternative means of participating in decisions concerning the provision of public services. Our task is to show how these methods of participation relate to, and help to define, different public service architectures. Hood contrasts doctrines of institutional design which favour ‘maximum feasible consumer sovereignty’ with the mainstream administrative argument which advocates a ‘fiduciary trusteeship’ approach to the design of public services.26 The main difference between them is the question of who – public authority or individual – should be at the centre of control concerning the provision of public services. This difference can be seen as one of the appropriate extent of delegations by citizens of the authority to 25 For an empirical comparison of the uses of different transparency mechanisms in public utilities regimes among European countries see M. Lodge ‘Regulatory Transparency: Towards a single citizen-consumer model?’ in Challenges to Democracy, ed. K. Dowding et al. (2001). A comparison between developing and developed countries is offered in M. Lodge and L. Stirton ‘Regulating in the Interest of the Citizen: Towards a Single Model of Regulatory Transparency?’ (2001) 50 Social and Economic Studies 75. Also, W. Sauter, ‘Universal Service Obligations and the Emergence of Citizens’ Rights in European Telecommunications Liberalisation’ in Public Services and Citizenship in European Law, eds. M. Freedland and S. Sciarra (1998) provides an interesting case study of how market mechanisms have led to a fuller realization of citizenship rights in the field of telecommunications. It may be that in some cases doctrines of privatization and decentralization have led to a reduction in voice and representation while not significantly expanding choice and information – for example, where vertically integrated state monopolies are transferred to the private sector without adequate restructuring or regulatory arrangements. The point here is that a reduction in accountability is not a necessary consequence of the use of market mechanisms in public service provision. It may however occur where insufficient attention is given to the design of public service architectures given the nature of the particular service. 26 Hood, op. cit., n. 23, pp. 169–94. 480 ß Blackwell Publishers Ltd 2001 take decisions about how public services should be provided. One, more traditional, approach advocates a high degree of delegation of decisionmaking power to political authority. This contrasts with currently fashionable doctrines emphasizing limited delegation and empowering service users to make their own decisions concerning the nature of the public services they enjoy. Hood states the conventional argument for the first approach as follows: Public services, it may be held, are best provided by producers acting as fiduciary trustees, not as market traders dealing with sovereign consumers. Such producers must have a wide discretion to judge the best interests of their clients. Clients may not be able to judge their interests for themselves because they are hopelessly divided, lack the relevant expertise, or cannot be trusted to take the long view or to weigh risks properly.27 This view of public services argues that key decisions should be taken by ‘experts’, rather than left up to the preferences of citizens. It advocates a form of organization that places public authority at the centre of control of public services. Of course, decisions ought to be made in such a way as to reflect the rights and interests of citizens, and citizens may be able to ‘punish’ decision-makers, in an election in the case of elected officials, for example, if they fall short of this standard. However, this view is still essentially a ‘high-delegation’ approach in which important decisions are taken out of the hands of individual service users. The fiduciary trusteeship doctrine can be contrasted with an approach to institutional design that emphasizes consumer sovereignty. On this account, citizens are the best judge of their own needs, and should therefore be able to take important decisions regarding the public services they consume for themselves. On this alternative account it is the service provider, rather than the service user who is not to be trusted. Self-interested providers (whether business or government) may attempt to exploit the client, if given the opportunity. Furthermore, the sovereign consumer account stresses the importance of appreciating the level of sophistication and capacities of individuals who are seen as being highly capable of making informed choices. This is a distributed model of control in which the overall level and mix of public services provided is the outcome of the many micro-level consumption decisions made by citizens as service-users. These two doctrines for the design of public services lead to substantially different architectures of public service provision, in particular in terms of their respective demands on protecting the rights and interests of citizens. These demands also impact on the different ways in which the basic tools for achieving transparency in institutional design should be applied. Although each of the four tools is applicable to both models of public service, the emphasis in each case will be different. The argument is summarized in Table 2. 27 id., p. 181. 481 ß Blackwell Publishers Ltd 2001 Table 2: Contrasting Doctrines of Public Service Provision Fiduciary Trusteeship Model Consumer Sovereignty Model Voice Participation in public service. Enforcement of contractual obligations. Redress. Choice Political competition. Market competition. Representation Ministers, agencies. Consumer interest groups. Information White Papers, Annual Reports, requirements to give reasons. Benchmarking, comparative information. Within the fiduciary trusteeship model of public service provision, transparent institutions are those that force public service providers to justify their decisions, actions or behaviour (primarily on an ex post basis). Therefore, the primary emphasis across transparency mechanisms is given to voice and representation. Less important, and at best facilitating the exercise of voice and representation, are choice and information. Among the mechanisms of primary importance, voice is exercised through individuals attempting to influence the decisions of public officials through active participation in public services, for example, by presenting evidence to public inquires or other information-gathering bodies, for example, participation on service-user councils. This is an essential means by which public decision-makers determine the view of their ‘principals’, the service users. Of similar significance, representation highlights the central oversight role played by ministers, bureaucracies, and politically accountable agencies in deciding appropriate standards of service as well as in monitoring and enforcing their compliance. In the fiduciary trusteeship model, choice is a relatively underemphasized mechanism, mainly consisting of the opportunity to vote for elected officials. It is however debatable whether voting offers an effective mechanism given the negligible impact of an individual’s vote and also the ‘bundled’ nature of party platforms. For these reasons critiques based on the dilution of citizenship in the regulatory state, whether because of the privatization of public ownership or the delegation to quasi-independent technocratic agencies, may be overstated, relying on an overly-optimistic expectation of what can be achieved by traditional channels of political accountability. Information as a mechanism is of similar secondary importance in enhancing the effectiveness of other mechanisms. The publication of consultative documents can increase the effectiveness of ex ante input by individuals and consumer and other representative groups. In addition, the legal duty of decision-makers to give reasons may further constrain the discretion of decision-making by representative institutions. 482 ß Blackwell Publishers Ltd 2001 Transparency mechanisms designed along lines advocated by the consumer sovereignty model emphasize the importance of choice and information. Transparency is mainly associated with the range of issues over which individuals can make their own choices concerning the nature, extent, and quality of consumed public services as well as the quality of information on the basis of which these choice-decisions are made. This emphasis on information includes both the ‘benchmarking’ of different products as well as the explicit listing of alternative consumer packages. This may be seen as a narrower conception of transparency, since it demands ‘merely’ that the actions of public officials and service providers are visible, provided that a range of alternatives are available to the individual. However, from the service user’s perspective, this model of transparency allows for an ex ante as well as ex post say in the way in which public services are provided. Since individual service users are the central decision-makers within the consumer sovereignty model, a priori representation plays a less important role in ensuring transparent public services. Representation offers a mechanism to establish ex ante influence on the design of the regulatory regime (that is, an advocacy role) while also offering, for example, by consumer interest groups, collective action solutions to the monitoring of services. Finally, voice mechanisms add to the strength of the choice mechanisms by guaranteeing the handling of complaints and offering redress where particular aspects of public services fall below mandated service standards. Both representation and voice could therefore be said to be supportive of the other mechanisms. Thus, without claiming to represent all possible combination of transparency mechanisms, this section has shown that the choice among different forms of social ordering leads to clear implications for the selection and emphasis given to different types of transparency mechanism. However, it does not suggest that a shift between models leads to a deterministic reduction of ‘transparency’; rather, the shift relates to a dominantly qualitative change in which transparency mechanisms are applied and weighted. CHOOSING TRANSPARENCY MECHANISMS To see how the application of the four basic tools can bias the allocation of de facto authority within public service regimes, we need to understand the way in which different transparency mechanisms operate to establish control. There are three indispensable elements to any control system (whether in nature, in machines or, as in our case, public services): a director, a detector, and an effector.28 The director element represents a means for establishing 28 See A. Dunsire, ‘A Cybernetic View of Guidance, Control and Evaluation in the Public Sector’ in The Public Sector: Challenges for Co-Ordination and Learning, ed. F.-X. Kaufman (1991); Hood, op. cit., n. 15. 483 ß Blackwell Publishers Ltd 2001 Table 3: Control Architectures Centre of Control (establishes goals) Primary Detector (checks course) Primary Effector (makes necessary changes) Public Authority Voice Representation Individual Authority Information Choice the purposes of the system. Detectors represent an array of instruments through which data about a situation are passed to the centre of control. Effectors are instruments through which instructions from the control are fed into a situation. The allocation of authority among different centres of control depends on the effectiveness of the detectors and effectors through which competing centres of control – individual service users or public authority – can exercise direction over the system. Table 3 sets out the way in which different applications of the four transparency mechanisms operate as instruments of control when applied to the two models of public service provision. What is broadly defined as ‘public authority’ represents any organizational form which has public recognition (by law or otherwise) to represent service users or to enforce authority to enhance the transparency of the provided service to the user. This form of control to some extent represents a delegation of authority from the individual user. According to this ideal-type typology, we have highlighted the dominant mode according to the centre of control and detector/effector dimensions. This is not to suggest that these will be the sole mechanisms nor does it mean, in practice, that these are effective or dominant or that they will be exercised in a ‘public interested’ way reflecting organized and unorganized interests. However, we assume that political and quasi-independent authorities are to some extent responsive to individual concerns29 and are likely to search and interpret information following the detection of user preferences through expressed ‘voice’ options. More generally, decisions about how to incorporate transparency into public services are therefore choices about architectures of control: who do we want to exercise decision-making authority over how public services are provided. In all cases there exists a trade-off between the advantages and disadvantages between decision making at different levels of authority. In different circumstances the nature of this trade-off will bias the decision in favour of a fiduciary trusteeship model or a consumer sovereignty approach. This section discusses the trade-off between allocating authority at the individual or public level and addresses what variables define when certain tasks or functions should be delegated from the individual to a public authority. It is argued that the calculus of delegation differs according to the 29 See, also, S. Peltzman, ‘Towards a More General Theory of Regulation’ (1976) 19 J. of Law and Economics 211. 484 ß Blackwell Publishers Ltd 2001 ‘cost profile’ of any public service. In some cases, consumers will have little difficulty in choosing, becoming informed and expressing their preferences. In other domains, the ‘cost profile’ to the individual will be much higher. Such cases are usually associated with limited competition and low overall salience, thus making the acquisition of information for exercising choice for the individual much higher. Such a ‘cost profile’ offers support for a potential delegation towards public authority; however, at the same time, it is countered by the costs of delegating authority. Arguably, delegation from individual to public authority – and hence from consumer sovereignty towards fiduciary trusteeship – should take place when the costs of the exercise of individual authority outweigh its benefits. Delegation is therefore primarily justified in terms of costs of individual action that prohibits the effective exercise of one’s consumer sovereignty, in particular, information and control. At the same time, any delegation to a public authority bears considerable challenges not only in terms of prescribing choices, but also in terms of exercising those functions according to the preferences of the ‘delegating’ consumers.30 As already noted, the main costs of individual action are those of information and control costs. The former range from costs of evaluation and comparison, the application of choice and an awareness of the potential impact and consequences of a consumer’s choice. Control costs in contrast relate to the likely inability of an individual to control provision of public services. To (mis)use Bentham’s rule, cited earlier, ‘the closer we watch, the more costly it gets’. Such costs range from the pure activity of monitoring the provision of public services to the acquisition of expertise to be able to evaluate these activities to those of gaining institutional access. While many attempts have been made to include individual participants into the control process, for example, in hearings, power of action emerges from the degree of institutionalized access and the associated resources rather than from individual intervention. At the same time, the question of ‘why delegate’ needs to consider the associated potential pitfalls of delegation and whether these outweigh the costs of relying on the individual. Two main sources can be identified, namely the costs of ‘corruption’ (defined here very widely in terms of any activity which undermines the ‘public interest’ goals of the regulatory regime) and paternalism. The potential costs of ‘corruption’ are mainly those of ‘capture’ of the public authority by the industry, of ‘loose cannon’ or other action by the public authority which utilizes its discretion, either to pursue different priorities or to focus on high visibility rather than ‘undesirable’ functions and the costs of political intervention which is used to advantage certain electoral constituencies over others. 30 For the literature on legislative delegation see, for example, D. Epstein and S. O’Halloran, Delegating Powers (1999); M. Horn, The Political Economy of Public Administration (1995). 485 ß Blackwell Publishers Ltd 2001 Paternalism means that public authorities are likely to make choices which are collectively binding and do not represent the interests of all affected users. In particular in an increasingly differentiated society, the ability of any public authority to reduce ‘paternalism’ costs to a minimum are likely to be limited. Discussions with regard to the choice of delegating authority require a consideration of both the costs to the individual who aims to exercise her consumer sovereignty. An assessment of different public service’s ‘cost profiles’ (of both the costs of individual action and the likely costs resulting from delegating authority) is therefore likely to lead to a differentiated calculus as to the allocation of individual from public authority. While the advocates of the fiduciary trusteeship doctrine argue that the consumer cannot be trusted and requires institutional support, the debate has often lacked a focus on the potential disadvantages and costs of delegation. Depending on the estimated or perceived extent of the costs of relying on the delegation to public authorities, these costs alone are arguably a sufficient justification for advocating a large extent of consumer sovereignty, and, hence, individual authority. Given the different ‘cost profile’ of public services, the old distinction between ‘state’ and ‘market’, that is to say, a public administration model which relies on command and control versus a model based on market mechanisms, has become increasingly irrelevant and requires a more differentiated analysis. The argument of this section is presented schematically in Figure 1. CONCLUSION With the emergence of the regulatory state traditional doctrines of accountability have proved inadequate to deal with the diversity and complexity of the new public sector. This paper has attempted to address some of the problems of making public services accountable by suggesting a more complex understanding of the ‘publicness’ of public services. By relying on the concept of the transparency mechanism, we have aimed to highlight a toolbox for laying down institutional rules of the game which promote horizontal as well as vertical accountability. The availability and accountability of public services is an essential part of a modern understanding of what it means to be a citizen. The existence of transparency mechanisms allocating authority to public authority as well as to individual service users serves what Schmitter calls ‘. . . the necessity for democracy to protect itself from its own potential for self-destruction.’31 Where individuals face prohibitively high monitoring costs, the allocation of control with respect to public services to agencies of horizontal 31 Schmitter, op. cit., n. 10, p. 59. 486 ß Blackwell Publishers Ltd 2001 487 ß Blackwell Publishers Ltd 2001 Figure 1: Contingency Model of the Effects of Transparency Mechanisms for Services of Different ‘Cost-Profiles’ accountability serves to buttress democratic principles rather than to erode them. That we reach a similar conclusion to O’Donnell on this point ought to give some confidence in that conclusion, given the differences in our respective starting points and approaches to the problem. A second argument of this paper has been the importance of doctrinal choices and cost calculus. Building transparency into an institution’s design along fiduciary trusteeship lines will result in a very different architecture for the delivery of public services than when transparency follows the consumer sovereignty model. The ‘fiduciary trusteeship’ argument regards the marketplace as extremely costly for the individual and thus requiring public authority, while the ‘consumer sovereignty’ model sees constraints on individual authority at least as costly as the costs of exercising individual choice. Allocating authority between alternative centres of control is a question of choosing the right mechanism, matching the appropriate doctrine to the problem situation. In cases where individuals do not face prohibitive decision-making costs, a consumer sovereignty architecture is likely to be the superior form of social ordering. The state fulfils its obligation of respect for citizens, by supporting the decisions of individuals and enabling them to act on the basis of those decisions. It is where individual decision making is prohibitively expensive that we are likely to come up against the conflict that Goodin highlights between respecting people’s dignity and respecting people’s choices.32 Here a fiduciary trusteeship architecture for public services will be appropriate. Extensive voice and representation can ensure that ‘paternalistic’ decision-making is responsive to citizen well-being and that corruption and drift are limited. As pointed out earlier, in many countries both in the developing and developed world, the design of public services incorporates all four mechanisms to different degrees. This supports the argument of this article that there is no ‘one-size-fits-all’ approach to making public services transparent. Using a wide selection of mechanisms may enhance the protection of citizen welfare and autonomy to the extent that hybrid architectures can improve upon the limitations of either ‘pure’ consumer sovereignty and fiduciary trusteeship. In fact, the point could be put somewhat more strongly. This argument could be based on W. Ross Ashby’s first law of cybernetics according to which any system for controlling a situation must be able to proliferate at least as much internal variety as the situation itself is capable of producing: more succinctly, ‘only variety can destroy variety’.33 Applied to the design of public services this might mean, among other things, that where public services themselves raise complex problems, a single locus of control may lack the channel capacity to effectively govern a public service. Arguably, it is primarily through the opposition of competing centres of 32 R. Goodin, Political Theory and Public Policy (1982) 79–81. 33 S. Beer, Decision and Control: The meaning of operations research and management cybernetics (1966) 279; see, generally, id., pp. 275–82. 488 ß Blackwell Publishers Ltd 2001 control that it becomes possible to achieve transparency of a highly complex public service regime.34 Given the various possibilities for incorporating transparency that the four basic mechanisms in our toolbox suggest, any argument which claims that public sector reform towards the regulatory state erodes the meaningful provision of citizenship needs to be critically examined. While we see questions about legitimacy, accountability and transparency as genuine, there is no good reason to suppose that citizenship is necessarily eroded by regulatory state developments. Indeed through matching the complexity of the new public sector with an innovative application of transparency mechanisms, it is even possible that the emergence of the regulatory state might lead to a fuller realization of citizenship than the traditional welfare state, respecting universal access as well as individual choice. The challenge is to make that possibility a reality. 34 See, also, A. Dunsire, ‘Modes of Governance’ in Kooiman, op. cit., n. 12; Dunsire, op. cit., n. 28. 489 ß Blackwell Publishers Ltd 2001