COMING TO TERMS WITH ACCOUNTABILITY Combining different forums and functions in a multidimensional way

Transcription

1 Title: Subtitle: Authors: COMING TO TERMS WITH ACCOUNTABILITY Combining different forums and functions in a multidimensional way Tom Willems Wouter Van Dooren Abstract In this article we develop a theoretical argument that could lead to a more optimistic outlook on the present state of accountability. By moving beyond the unidimensional electoral and parliamentary means of holding governments to account, we attempt to gain a full understanding of the richness of the concept of accountability, without losing the clarity of its core meaning. If we combine the different forums and functions of accountability in a more open and flexible manner, the possibilities to hold power to account seem to be larger than ever. The main reason is that functions no longer depend on the well-functioning of a single forum and each forum serves multiple functions. In order to study accountability on a more systematic and empirical basis we urgently need an agreed upon conceptual framework. We aim to contribute to this much wanted coming to terms with accountability. Introduction Contemporary analyses of accountability, both academic and political, tend to be rather pessimistic. They generally contend that the central idea of ministerial responsibility has eroded and that the once solid democratic pyramidal chain of delegation is broken; and therefore leads to accountability gaps and even democratic deficits (e.g. Flinders, 2001; Goetz & Jenkins, 2001; Mulgan, 2003; Papadopoulos, 2007; Pierre, 2000; Rhodes, 1997). Questions like is it possible to safeguard accountability in horizontal public governance? are usually approached sceptically and answered negatively (Michels & Meijer, 2008). The pessimistic narrative risks idealizing the past, as if holding a government to account was ever that easy and straightforward; and it also risks leading to unrealistically high expectations and consequently to negative assessments of the present. In this article we offer a more inclusive and optimistic outlook on contemporary accountability. In order to do so, we build on its rather well-defined meaning: a process of being called to account to some authority for one s actions (Mulgan, 2000). Although this principal-agent approach is analytically useful, the problems start when confronting it with the practice of accountability. The many complex processes of calling to account do not always fit adequately into a principal agent scheme. By moving beyond its common electoral and parliamentary interpretation, we attempt to gain a better theoretical understanding of the richness of the concept of accountability, without losing the clarity of its core meaning. Pollitt & Hupe (2011) described accountability as a magical concept, based on its broad scope, great flexibility, normative attractiveness and being hard to oppose. It is one of those fashionable concepts with global appeal, which risks becoming analytically useless because it functions like an umbrella that covers various other, often highly conflicting and contested, concepts (Bovens, * Tom Willems, PhD Fellowship of the Research Foundation Flanders, Department of Political Science, University of Antwerp, Sint-Jacobsstraat 2, 2000 Antwerp, Belgium. tom.willems@ua.ac.be Wouter Van Dooren, Assistant Professor, Department of Political Science, University of Antwerp, Sint-Jacobsstraat 2, 2000 Antwerp, Belgium, wouter.vandooren@ua.ac.be

2 Schillemans, & 't Hart, 2008). Although magic concepts have some advantages, they also face serious limitations as scientific concepts (Pollitt & Hupe, 2011). In order to study accountability more systematically and empirically, more conceptual coherence into the diverging development of the literature seems warranted. In other words, to make accountability operational we first need to come to terms with it. That is exactly what this article aims to do and what it may add to the already vast amount of literature on this topic. The article is structured as follows. In the first section we distinguish between the two basic approaches to accountability. While starting from the notion of answerability, we gradually move beyond strict principal-agent thinking and open up to the complexity of accountability. The second section then identifies five accountability forums in which the multiple accountability processes are at play: the political, judicial, administrative, public and market forums. In the third section, we discuss three functions which accountability is supposed to have: the democratic, constitutional and performance function. The fourth section presents a conceptual framework combining different forums and functions. While unidimensional approaches usually attribute a particular function to a forum (for instance democratic function to political forum), we argue that forums may have many functions and functions may seek many forums. We believe that this model allows for precise analyses of accountability, without losing the richness of the concept or ignoring its complexity. We round up with some limitations of the framework and challenges for further research. Two Basic Approaches of Accountability Accountability is a key concept for understanding democratic governance. The question of how to hold governments to account has a long history. It has evolved from mere financial accounting into the much broader concept of good governance. In the 1940s, the Friedrich-Finer debate set the stage on the subject. Samuel Finer believed that accountability could only be guaranteed by maintaining hard external constraints, whereas Carl Friedrich argued that self-control is feasible based on broader array of soft internal norms and values. These two interpretations of accountability are echoed in much of the recent scholarly discussions (Acar, Guo, & Yang, 2008), and also resonate in the two basic approaches we distinguish below. Finer and beyond: Having to answer to a principal The first basic notion of accountability points to a condition of having to answer to an individual or body for one s actions (Flinders, 2001). In line with Finer, external constraints guide the agent s behaviour. It presumes a clear principal-agent relationship of bureaucracy that is accountable to its main principals: the citizens and their elected representatives. Mulgan (2003) defines accountability as a relationship of social interaction and exchange involving complementary rights on the part of the account-holder and obligations on the part of the accountor. He stresses a number of defining features. First, it is external which means that account is given to some other person or body outside the person or body being held accountable. Accountability is distinguished from responsibility which concerns internal norms and values. Second, it also involves social interaction and exchange. Third, it implies some rights of authority. The relationship is unequal, because the account-holder has some kind of moral authority over the accountor. Yet, this does not necessarily entail actual or formal power. Although accountability has been a key concept in the field for a long time, there is recently a tendency to expand its core meaning and a remarkable resurgence of popularity (Mulgan, 2000;

3 Pollitt & Hupe, 2011). Koppell (2003) offers five dimensions of accountability: transparency, liability, control, responsibility and responsiveness. Accountability is, in his view, all those things combined. Those who build on the Finerian notion of accountability, point to the risk of overstretching the analytical concept. They argue that the principal agent definition leads to conceptual straightforwardness. The original core sense of accountability, signifying external ex post scrutiny, is sufficiently distinct from its extensions to warrant separate identification. Rubin (2006) explains the attempts to expand the notion of accountability by referring to a widespread anti-administrative, anti-bureaucratic movement. He underlines that accountability involves many of the features that are central to the modern administrative state and that many people find so unattractive about it: hierarchy, investigation, evaluation, rules, reporting, etc. He also stresses that accountability has a well-established meaning: the ability of one actor to demand an explanation or justification of another actor for its actions, and to reward or punish that second actor on the basis of its performance or its explanation. Clearly, his view needs an account-holder who is able to reward or punish, who is de facto a principal. Mulgan (2000) also distinguishes accountability from its related concepts. Transparency is a vital condition of accountability, but it does not capture the whole process. Accountability is linked to the internal responsibility of public actors. It is an important part of the institutional checks and balances system, but there are other types of control mechanisms besides ex post accountability. It is related to being responsive to wishes or needs of citizens by the government, but that does not mean that they are induced to do so by processes of calling to account. It also involves public explanation and discussion in a forum, but it is not the same as the dialogue between citizens in a deliberative democracy. He argues that these extensions, albeit not desirable, are understandable because they are so closely related to the practice of accountability. It seems that the influence of the hierarchical P-A approach is so inherently present that even its critics have difficulties escaping it. Philp (2009) for instance argues that the tendency to see accountability as a principal-agent (P-A) relationship should be resisted mainly because the bilateral relationship between P and A is too simplistic and it treats contingent conditions as necessary ones (like the required level of authority). Yet, his definition of accountability is interesting: A is accountable with respect to M when some individual, body or institution Y, can require A to inform and explain/justify his or her conduct with respect to M. A rigid P-A approach to accountability is indeed counterproductive and of limited use. Yet, he mentions the words Y can require A to. The account-holder is apparently in a position to make demands of accountability to the accountor. Again, this position is not necessarily based on formal or actual power. Although a formal P-A relationship is not always required in this view, some P-A ingredients are clearly there. Friedrich and beyond: Managing expectations of multiple principals In their reference article about the Challenger tragedy, Romzek and Dubnick (1987) claim that accountability can play a greater role than expressed by the idea of answerability. Holding someone answerable implies the presence of prior expectations for such actions or behaviour. They state that public administration accountability involves the means by which public agencies and their workers manage the diverse expectations generated within and outside the organization. Public administration has to deal with many different and often conflicting expectations, which lead to complex overlapping accountability relationships. Viewing it as a strategy for managing expectations, accountability is more than the actual fact of being held accountable. It is a continuous process of anticipation, identification, definition and responses to pressures which eventually lead to certain actions.

4 This approach puts a stronger emphasis on the internal dynamics, norms and values of accountors, and hence is more in line with the Friedrich position. Yet, unlike the ideal-typical Friedrich model, they usually do see a role for external control. Strong bureaucratic control is, however, one of many forms of control. Romzek and Dubnick (1987) did not denounce the importance of the basic notion of answerability. They claimed that limited, direct and mostly formalistic responses to demands generated by specific institutions or groups are just one of many forms of accountability. Their classification of four alternative systems of accountability is based on two critical factors: (1) whether the ability to define and control expectations is held by some specified entity inside or outside the agency and (2) the degree of control that entity is given over defining those agency s expectations. Instead of downplaying its significance, the authors still use control as crucial for understanding accountability. Acar et al. (Acar, Guo, & Yang, 2008) claim that this strategic approach for managing expectations is more realistic in present-day complex governance. Instead of relying merely on public administrations compliance with the rules and preferences determined by elected politicians and rewarding or punishing its performance, the strategic approach focuses on the way public administration manages the diverse expectations placed on it by many internal and external stakeholders. Many scholars have followed this loosened up line of reasoning. They stress that the traditional P-A interpretation of accountability, with its one-sided focus on formal control and sanctions, is becoming less suitable to grasp the real nature of public governance (Behn, 2001; Considine, 2002; Dowdle, 2006; Flinders, 2001; Hodge & Coghill, 2007; Mulgan, 2003; Scott, 2000). The general argument of the loosened up approach of accountability is that in complex and dynamic contexts, the reliance on vertical oversight and control will fail to ensure accountability because it is inflexible and formalistic. Furthermore, it tends to produce accountability systems that are risk-averse and retrospective in nature rather than pro- and interactive (Acar, Guo, & Yang, 2008). Michels & Meijer (2008) suggest there is a need for supplementary horizontal mechanisms of accountability in line with the structures of government becoming more horizontal as well. Schillemans (2008) also emphasizes the promising nature of new horizontal mechanisms. Grant & Keohane (2005) identify no less than seven accountability mechanisms in world politics; three of them (market, peer and reputation) are horizontal. They are called horizontal because accountholders are not true hierarchical superiors: like clients, stakeholders and peers. Nevertheless, the accountors are (or do feel) compelled to render account for their actions to these informal account-holders, which makes the process anything but noncommittal. Their impact, however, should not be exaggerated. Schillemans (2008) admits that they only work effectively in the shadow of hierarchy. Moreover, Michels & Meijer (2008) emphasize that they are not intended to replace hierarchical accountability, but are instead to function as extensions. A mixture of vertical and horizontal mechanisms can only work if certain requirements have been met, which is currently not always the case. Hodge & Coghill (2007) identify a pyramid of accountability practices, building further upon the work of Ayres & Braithwaite (1995) on regulatory practices. A full appreciation of accountability should therefore include the whole range of possible consequences (from strong to weak) at the disposal of the account-holder (Considine, 2002; Hodge & Coghill, 2007; Scott, 2000). The minimal threshold to speak of accountability, however, should always be kept in mind. There has to be a certain obligation to answer, no matter how small. Friedrich meets Finer?

5 The main advantage of defining accountability in a P-A relation with answerability as the main currency, is the conceptual clarity. Yet, when reconstructing how accountability works in practice, it evolves almost automatically into managing different expectations. It reflects a balancing act among multiple sources of control. Instead of a single principal-agent relationship, public officials and politicians are confronted with many principals with diverging and often conflicting expectations. Yet, they remain principle agent relations, albeit complex ones. Both approaches to accountability hence do not have to exclude each other. What follows is an attempt at a conceptual framework that reconciles the clarity of the P-A definition with the richness and complexity of real-world accountability. Forums of Accountability Bovens (2007) definition of accountability as a relationship between an actor and a forum, in which the actor has an obligation to explain and to justify his or her conduct, the forum can pose questions and pass judgment, and the actor may face consequences is a good starting point. Accountability has three indispensable components: (1) the actor should be obliged to inform the forum about his conduct, (2) there should be an opportunity for the forum to debate with the actor about his conduct as well as an opportunity for the actor to explain and justify his conduct and (3) both parties should know that the forum is able not only to pass judgment but also to present the actor with certain consequences. The last part of his definition ( may face consequences ) is crucial, but necessarily ambiguous. What is the status of these consequences? In theory, the lower limit of accountability is reached when an accountor is compelled, or feels compelled, to disclose information to the public, which has very limited possibilities to ask questions or pass judgement. Purely informal or voluntary transparency does not amount to accountability. Bovens labels the account-holder and accountor respectively actor and forum. In our view, the conceptual potential of the term forum is not fully exploited. Bovens claims that a forum can be an individual, an organization, institution or even an almost virtual entity like the public opinion. In this way, the idea of a forum looses much of its clarity. For an actor does not give an account to a forum, but gives an account to another actor in a forum. The term forum could evoke a strong and comprehensible image because it refers to the forum Romanum of ancient Rome; the marketplace which was the centre of public life. This gathering place was of great social significance, due to the hosting of diverse public activities, including political discussions, judicial and other businesses. The image of the forum Romanum captures the notion that accountability is in essence a process of discursive interaction in public. Building on this metaphorical meaning of a marketplace, we will use the term forum in a more general way; as a virtual meeting place where ideas, views or arguments of a set of actors on a particular issue can be exchanged. The term forum as used here is also related to the notions of regulatory space and audit space which generally express the same idea (Hancher & Moran, 1989; Kells & Hodge, 2010). Braithwaite (2008) calls them somewhat more abstractly nodes in networked governance. If we paraphrase Bovens definition, we can describe accountability as follows: a relationship between an accountor and an account-holder in a forum, in which the accountor has an obligation to explain and to justify his or her conduct, the account-holder can pose questions and pass judgment, and the accountor may face consequences. Black (2008) stresses that accountability relationships are relational and dialectical in nature. The interdependencies, which are particularly

6 vital in what she calls polycentric regulatory regimes, qualify the hierarchical approach significantly. Moreover, she points out that more attention should be devoted to the communicative structures which construct and contest accountability processes. FIGURE 1 attempts to capture the interactive process of accountability visually. 1.Information ACCOUNTOR 2. Debate ACCOUNT-HOLDER 3. Judgment Formal Informal FORUM In order to understand how it really works, we need to focus on the dynamics of the many different processes of accountability. It can be extremely complex. Governments are being called to account by diverse account-holders in different forums for different aspects of their conduct. The accountability process will be arranged differently each time with different actors that demand different kinds of information. In each process, different criteria to decide what defines accountable behaviour will apply and different kinds of consequences will be at the account holder s disposal. Governments are continuously being called to account within different forums at the same time. Many scholars have attempted to reflect the complex diversity of accountability by formulating typologies (Pollitt & Hupe, 2011). The overview in TABLE 1 shows a selection of the abundance of typologies. Author Different Forums/Mechanisms Dimensions of Accountability in Literature Romzek & Dubnick (1987) Bureaucratic Legal Professional Political Sinclair (1995) Political Managerial Public Professional Personal Stone (1995) Parliamentary Managerialism Judicial Review Constituency Market Deleon (1998) Bureaucratic Political Professional Anarchic Haque (2000) Externalformal (legislative, executive, ) External-informal (media, interest groups, ) Internal-formal (officials rules, codes of conduct, ) Internalinformal (professional ethics, peers, ) Klingner, Nalbandian, Politics Administration Markets Romzek (2002) Flinders (2003) Parliamentary Judicial Managerial Mulgan (2003) Political Judicial Other Grant & Keohane (2005) Hierarchical Supervisory Fiscal Legal Market Peer Reputation Dowlde (2006) Elections Bureaucracy Judiciary Transparency Market Mashaw (2006) Public Market Social Governance Scott (2006) Public law Markets & competition Networks & communities Design Hodge & Coghill (2007) Goodin (2008) Hard law & justice mechanisms State: hierarchy Formal sanctions from markets & regulators Market: competition Informal influences : clients, media, environment Non-profit : Cooperative networks Social, personal, ethical behaviours Bovens (2008) Political Legal Administrative Professional Social Koliba, Mills, Zia (2011) Democratic (politicians, citizens, legal) Market (shareholder/owner, consumer) Administrative (bureaucratic, professional, collaborative)

7 The common denominator of the typologies suggests an overview of forums in which the most relevant processes of accountability are being held: the political, judicial, administrative, public and market forum (see TABLE 2). While the political, judicial and administrative forums and accountholders are well-known in the literature, we would like to draw the attention to the other two forums. First, governments are also being held to account through critical debate and scrutiny in the public forum. In contrast to Bovens (2007), who uses public accountability in a general way to cover all kinds of accountability processes, we delineate it more narrowly: accountability of persons or institutions vested with authority toward criticism, questions, and commentary voiced in public by citizens or organized civil society (Steffek, 2010). In other words, public accountability is accountability towards a wider public. The public forum is conceptually distinct from the state; it is a place for the production and circulation of opinions that can be critical of the state. The existence of such a non-state or non-governmental, yet intrinsically public forum is thought essential to democratic governance. Second, markets have developed into a distinct forum in which government s actions and decisions are being held to account. This market-based accountability applies to owners and shareholders who can call the company s managers to account for its performance, and to consumers who can refuse to purchase or exit to another provider. Due to various NPM-reforms this peculiar accountability of the private sector gets also more prominent in the world of governments. It is important to note that the typology of forums serves analytical purposes. In practice interactions and overlaps between them should occur regularly. Moreover, a certain forum can have different outlets at local, regional, national and supranational level. Yet, by identifying those forums, we hope to be able to better describe and analyse these complexities. For the same purpose, a further addition to our conceptual vocabulary deals with functions of accountability. Type Forum To whom? How? Impact? Political Forum Citizens as voters Elections Campaign, election day, anticipation of elections (proactive effects) Strong Members of Parliament Principle of Ministerial Responsibility Strong Plenary assembly Questioning Ministers Parliamentary committees Scrutinizing legislation & budget (and approving it) Special committees (e.g. policy disasters) Investigating public policy problems Deciding on support/resignation of Ministers Direct democratic tools Dialogue with citizens Weak Referenda, citizen surveys/juries, etc. Potential decisive vote by citizens Political Parties Dialogue with party executive committee & party members Weak Judicial Forum Administrative tribunals Judicial courts Hear and decide appeals against government decisions or actions Formal power to disclose information and overrule decisions Strong Strong Administrative Forum Internal Audit Court of Audit Auditors General/Audit office Inspections Visitations Advisory councils Ombudsmen Commissioners Professional peer review Whistleblowers External Audit Regulatory & monitory bodies Private auditing firms In general: No formal power to sanction Informal power to influence, publicize, benchmark & criticize (naming & shaming, reputation) If any formal power, very specific & limited (like inspections, self-regulating professions, ) Weak Public Forum Market Forum Mass Media Organized civil society (formal) Ad hoc action groups or individuals (informal) Shareholders/owners Information channel & opinion and framing No formal power, but great influence Monitoring, critical dialogue, petitions, protest campaigns, strikes, etc. Naming & shaming, reputation Ibidem, growing in importance due to rise of the internet, increased digitalization, Freedom of information legislation, sunshine acts, etc. Profit, performance indicators Strong Weak Weak Strong Consumers Adapted and elaborated from (Mulgan, 2003) Competition, consumer panels & surveys, reputation, performance indicators Weak

8 Functions of Accountability Accountability as practice means managing many processes of information, discussion and judgment in different forums. Not only do the actors use many different tools, they also have many different expectations. In other words, accountability tools may have various functions as TABLE 3 illustrates. Author Aucion & Heintzman (2000) Scott (2000) Different Expectations Dimensions of Accountability in the Literature Control for abuse & misuse of public authority Economic values Provide assurance for use of public resources and adherence to public law & values Social or procedural values Learning for continuous improvement Continuity or security values Behn (2001) Finances Fairness Performance Klingner, Politics Administration Markets Nalbandian, Romzek (2002) Lodge Voice Choice Representation Information (2004) Bovens (2008) Acar, Guo & Yang (2008) Steets, Blattner (2010) Koliba, Mills, Zia (2011) Democratic: Control by citizens Mapping & manifesting expectations Compliance (with rules &regulation) Constitutional: Prevention of corruption and abuse of power Mobilizing & motivating (exante) Financial Learning: Enhance efficiency and effectiveness Monitoring & measuring progress and performance Mission of organization Democratic Market Administrative Modifying Democratic Participation Avoid use of authority Mobilizing & motivating (ex-post) Outcome Performance evaluation Market elements Impartiality Independence Accuracy & Quality Professionalism Accountability has three distinct functions. First, public authorities are held accountable for a variety of well-established rules and procedures to prevent unfairness or abuse of power constitutional function. Second, citizens (or by means of elected representatives) want to have the final say because the ultimate authority and ownership of the state rests with the citizens democratic function. It is not enough that public authorities act fairly and legally, citizens should be able to control and elect the public authorities in a meaningful way. These two dimensions are concerned with how government functions. Third, we also care what government actually accomplishes. We want to hold public authorities accountable for their results performance function. Peters & Pierre (2010) describe the first two functions as legitimacy by procedure and the third one as legitimacy by performance, elaborating the distinction between input and output legitimacy famously made by Fritz Scharpf (1999). They ask themselves if legitimacy by performance can compensate for a lack legitimacy by procedure? Their answer is negative; which stresses the genuine and independent character of both dimensions. The previous decades the issue of performance in the public sector became more prominent (Hood, 1991; Osborne & Gaebler, 1992; Savas, 2000; Van Dooren, Bouckaert, & Halligan, 2010). The increasing concern with performance expressed itself also in an explosion of auditing, monitoring and evaluating mechanisms which focus on economy, effectiveness, efficiency and value for money (Power, 1999). It extended beyond financial probity and due process into a wider

9 range of performance indicators. May (2007) claims that these new forms of regulation entail a shift from prescribing actions to regulating for results. Thus, the widely proclaimed rise of the regulatory state has led to a proliferation of various monitoring and regulating bodies designed to safeguard some public standards in the absence of direct ministerial control (Hood, James, & Peters, 2004; Scott, 2000). The consequences for the processes of accountability are significant. Accountability in the private sector applies mainly to shareholders and owners who can call the management to account for the company s profit and performance, but also to consumers holding companies to account through their choice or refusal to purchase (Mulgan, 2003). Competition is the basis for monitoring and rewarding compliant behaviour through loyalty and for punishing deviant behaviour through exit (Scott, 2006). In recent times three simultaneous trends managerialism, contractualisation, and privatization (all falling under the broad label of New Public Management ) had a profound impact on public sector accountability (Hodge, 2009) in two ways. First, market accountability emerged as an increasingly important and distinct forum and second, the general orientation towards performance, outputs and results has crept into the other forums as well. Hodge (2009) states that there is huge array of accountability bodies, including output specifications, legal contracts and private commercial incentives to manage commercial risks for performance. He concludes by saying that not only the guardians have changed under regulatory governance, there is also a change in what society expect of our guardians and governments. Dynamics of forums and functions A continuous rebalancing act Forums and functions interact in a dynamic way. Functions may be directly opposed, which means that satisfying one will necessarily affect another (Klingner, Nalbandian, & Romzek, 2002; Koliba, Mills, & Zia, 2011; Koppell, 2003). Multiple legitimacy claims and engaging in multiple accountability relationships can mean that attempts to make an organization accountable end in the accountability equivalent of the regulatory trilemma : they are ignored, co-opted, or destroy that which it is they seek to make accountable (Black, 2008). Lodge (2004) explains how three doctrines in public management (fiduciary trusteeship, consumer sovereignty, empowered citizenship) lead to different understandings of accountability, and correspondingly to different kinds of expectations. Fimreite & Laegreid (2009) describe accountability as a multi-dimensional concept: public governance through partnerships and networks needs to be accountable upwards to political sovereigns, horizontally to other agencies and downwards to citizens and clients. Although the welfare administration reform in Norway (NAV) they studied claims to strengthen accountability on all these dimensions, they conclude that in reality it is a major challenge to balance all these different accountability demands and to achieve them simultaneously. In particular the performance function of accountability seems to set off a delicate (re)balancing act for which most authors tend to have a rather negative outlook. Behn (2001) talks about an accountability dilemma ; a trade-off between accountability and efficiency. Jos and Tompkins (2004) speak of an accountability paradox, because these new mechanisms often threaten the very qualities that support responsible behaviour and judgment. May (2007) mentions potential accountability shortfalls which can undermine regulatory potential. He therefore stresses the importance of finding the right fit between regulatory circumstances and the design of regulatory regimes. Some authors state that democratic and constitutional accountability is de facto being sacrificed in the name of public sector efficiency (Flinders, 2001; Haque, 2000; Mulgan, 2003). The

10 common outlook on the state of accountability is rather negative. This is illustrated by the much diagnosed accountability gap or even democratic deficit of governance (e.g. Flinders, 2001; Goetz & Jenkins, 2001; Mulgan, 2003; Papadopoulos, 2007; Pierre, 2000; Rhodes, 1997). Yet, some NPM-advocates claim that the focus on performance does not necessarily imply a degrading of democratic and constitutional expectations. On the contrary, by upgrading performance a healthy balance can be found and democratic governance could even be improved (Bishop, Kay, & Mayer, 1994; Osborne & Gaebler, 1992; Savas, 2000). Responsiveness towards users and clients, combined with greater managerial accountability for results, acts as a promising substitute for diminished direct political control. Steets (2010) claims that the balance between the three functions is restored instead of disturbed. Performance in the public sector matters and may stand on equal footing with the other functions. Assessing accountability must be done according to different accountability standards, not all of them democratic ones. It is important to make the distinction between the different forums and different functions they have. Elaborating on this argument, we believe that a more optimistic outlook on the present state of accountability is possible and even warranted. A misleading assumption in the negative narrative about accountability is the unidimensional relationship between the forum and the function of accountability. The political forum is generally supposed to satisfy the democratic expectations, the judicial forum the procedurally ones. The public and administrative forums have only a secondary, supplementary role to play. Every forum thus serves a well-described central function. The recent rise of performance seems to disturb this composure, because it does not fit well in this scheme. Performance used to be a side issue; it was mainly a preoccupation of private companies, not of bureaucracies. Now a new market forum emerges in a public context and expectations regarding performance intrude the other forums. Balancing forums and functions before the performance era We argued that accountability for performance disturbed the unidimensional relationship of forum and function. Yet, some evidence seems to suggest that even in the past, the relationship between forum and function was more troubled. The traditional principle of ministerial responsibility for instance presupposes that the democratic function is essentially fulfilled by the political forum. It is the pyramidal chain of delegation that works as chief democratic safeguard (Ström, 2000). Scholars have, however, identified important flaws of the principle of ministerial responsibility (Hodge, 2009). Barberis (1998) describes it as a ruling fiction whereby ministers shelter behind civil servants and vice versa. Nevertheless, albeit a fiction, ministerial responsibility is still ruling. The principle of ministerial responsibility has such a strong symbolic and theoretical power that it remains the frame of reference for many. We have been, as Hodge (2009) says, very slow to move away from the central idea of ministerial responsibility. The principle is not only imperfect in practice because the main sanction (ministerial resignation) is applied rarely, arbitrarily and unpredictably. It faces more and serious problems. Although elections are a vital way to hold governments to account and influence the behaviour of politicians in different ways, they have some serious limitations as processes of accountability (Manin, Przeworski, & Stokes, 1999; Papadopoulos, 2003). First, elections are not necessarily retrospective. They can also be forward looking; voters can reward politicians with the best policy programme for the future. Second, elections as accountability require politicians who face re-election and genuine political competition. Otherwise the need to explain and discuss their performance is rendered

11 meaningless. Third, ideal type citizens vote after a careful process of information and deliberation. This is a very demanding and unrealistic assumption. Fourth, in democracies where executive power is shared in broad-based coalitions and elections are proportional, guaranteeing accountability through elections is rather doubtful. In addition, the formal powers of parliament to hold government to account can be seriously eroded in practice for several reasons. Party discipline can tie down both hands of members of parliament, especially in countries with broad coalition governments and influential political parties. Moreover, incompetence or unwillingness to devote time and energy in studying complex public policy in general can also affect it. Therefore, relying solely on competitive party elections and representative parliaments to get democratic accountability, relying solely on the political forum to get democratic accountability is limited and ill-fated. Braithwaite (2008) rejects the unidimensional electoral conception of nationally sovereign democracy for application to a world of networked governance. Multiple nodes of popular contestation at strategic intersections of networks should be set up. Democratic governance is no longer simply a way of handling the power of elected governments by electoral, parliamentary and constitutional means. Other forums need to step in and play a genuine role in establishing and guarding democratic governance: the public forum with its power to influence public opinion and reputation, non-elected courts which uphold rights of citizens, a public administration acting on public interests and with a clear public ethos, markets leading to forms of corporate social responsibility (CSR), or more generally processes of publicization in which private actors increasingly commit themselves to traditionally public goods and values (Freeman, 2006). They can function synergistically and mutually reinforce each other. The multidimensional relationship, applied to the democratic forum The increasing complexity of accountability is not only due to the inclusion of accountability for performance, but also because the forums get more crowded by actors and tools and the interchange between the different forums intensifies. We will illustrate this theoretical argument by elaborating on the democratic function of accountability, because here the dynamic interplay between different forums is most explicit, unexpected and interesting. A similar analysis could however be applied to the other functions of accountability. Contemporary accountability results from multidimensional relationships where every forum has multiple functions and every function has multiple forums at its disposal. This intermingling of accountability forums and functions challenges the idealistic notion of accountability in the past. Koppell (2005) labels the many, often conflicting, accountability expectations the multiple accountabilities disorder (MAD). It is a prominent example of an important current in the literature, where past accountability is depicted as being legible and clear and contemporary accountability as chaotic and obscure. Yet, this negative reading of accountability is not unison. Some scholars have recognized the ambiguous and complex nature of accountability practice, without drawing the evidently negative conclusion of disorder, deficit, disease, or worse (Aucoin & Heintzman, 2000; Behn, 2001; Considine, 2002; Fimreite & Laegreid, 2009; Hodge & Coghill, 2007). Others have suggested ways to facilitate these complicated accountability processes and to turn its multiplicity into an advantage. Several related concepts have been issued; extended accountability based upon interdependent and redundant networks of accountability (Scott, 2000); aggregate accountability to illustrate the necessary variety of accountability mechanisms (Freeman, 2000); 360 accountability in which each would be accountable to all the others (Behn, 2001); hybrid forms of accountability which bridge the vertical-horizontal divide (Goetz & Jenkins,

12 2001); spontaneous hybrid accountability regimes (Scott, 2006); a network approach to accountability to include promising new tools like judicial and ombudsmen networking (Harlow & Rawlings, 2007). Braithwaite (2008) suggests that a combination of nodal governance of networks from below and metagovernance of networks by institutions of representative democracy such as courts can provide superior accountability and superior transparency than either approach alone. The superiority comes from covering the weaknesses of hierarchical accountability with the strengths of horizontal accountability and vice versa. (emphasis added) Accountability is then accomplished by widening circles of opportunities to vote for people who represent our interests, to contest power and domination, and to deliberate in our own voice at many nodes of governance. The importance of electoral and parliamentary democracy is not denied, but if that is all there is then we are likely to end up with a thin set of protective measures, according to Braithwaite. All nodes of governance need enough autonomy and capacity to check others so that a multiplicity of separated powers can network to avoid domination. We need many strong kinds of circles we can join, or put differently, we need vibrant autonomous forums which can complement and strengthen each other. Keane (2009) calls this the advent of the monitory democracy, where public scrutiny and control of all kinds of decision makers is primordial. It is defined by the rapid growth of many different kinds of post-parliamentary mechanisms. In contrast to what is often suggested, they should not be treated as small supplements but as genuine scrutiny options. TABLE 4 illustrates the challenged relationships between forums and functions of accountability. Type of Forum Traditional Function New Functions Political Forum Democratic (major) 1. Democratic 2. Constitutional 3. Performance Judicial Forum Constitutional (major) 1. Constitutional 2. Democratic 3. Performance Administrative Forum Constitutional (minor) 1. Constitutional 2. Performance 3. Democratic Public Forum Democratic (minor) 1. Democratic 2. Constitutional 3. Performance Market Forum Performance (major) 1. Performance 2. Democratic/Constitutional

13 Conclusion The majority of the literature on the present state of accountability is rather sceptical. The democratic deficit or accountability gap thesis is dominant in diverse (sub) disciplines like public administration, international relations, development studies and EU studies. A good illustration of this common negative disposition is found in a recent volume of Government and Opposition on the relationship between new modes of EU governance and democracy (Bellamy, 2011). The typical line of reasoning is that new, more horizontal ways to hold governments to account are only welcomed as long as they are embedded within old modes of government that enjoy more traditional forms of democratic accountability. In other words, they must operate under the shadow of hierarchy offered by the standard democratic mechanisms of competitive party elections to representative parliaments to function properly. A post-parliamentary basis for accountability is only acceptable as a minor supplement to the formal institutions of a liberal representative democracy. This article presents a literature review that paints a more positive picture. Core concepts in our vocabulary are forums and functions of accountability. The forums are political, judicial, public, administrative and market forum. The functions are the democratic, constitutional and performance function. While unidimensional approaches usually attribute a particular function to a forum, we argue that forums may have many functions and functions may seek many forums. This multidimensional approach to accountability suggests a more nuanced assessment of accountability old and new. First, the traditional accountability mechanisms such as ministerial responsibility have many pitfalls which make the process of holding governments effectively to account much more difficult and complicated than often assumed. Secondly, the idealistic notion of accountability in the past leads to unrealistically high expectations and consequently to negative assessments of accountability nowadays. This is well illustrated by democratic accountability, which clearly does not depend only on the political forum. The other forums intervene and could even strengthen the overall democratic claim. In sum, the argument of this article suggests that the possibilities to hold power to account may be larger than ever. The main reason is that accountability functions no longer depend on the well-functioning of a single forum. Yet, two challenges remain regarding this optimistic narrative on accountability. First, more accountability is not always the same as improved accountability. Many public officials complain about existing accountability overloads (Bovens, Schillemans, & 't Hart, 2008). They refer to the increasingly long catalogue of rules and criteria prescribing their conduct. Not only do they complain about the frequency of the reviews, they also denounce the nature and course of the evaluation process. Adding new layers to the already gigantic piles of accountability mechanisms creates more red tape and less public gains, according to some public officials. In addition, citizens and action groups often complain that most information is too technical. Documents should be made more understandable for a larger public. If one wants to hold governments effectively to account, it is crucial to that information is both condense and comprehensible. Second, some avenues for further work need to be discussed. (A) Proceed in theory. Because many studies on democratic governance focus on participative and dialogical policymaking (e.g. Bevir, 2010; Fung, 2003, 2006; Skelcher & Torfing, 2010), the full potential of accountability as democratic value is insufficiently recognized. Increased bottom-up direct participation and deliberation will at best lead to minor adjustments at the margins because it concerns a rather limited micro-level (Goodin, 2008). Moreover, most pleas in the name of participation do not demand direct collective decision-making of citizens but institutionalized ways to contest and

14 deliberate on public policy making (Braithwaite, 2008). Secondly, the enduring dominance of public sector management with its focus on performance and results will only accentuate ex post accountability as valid democratic reference point. Thirdly, the strong call for a transparent government has opened the doors for more accountability, which will not be closed again in the foreseeable future. On the contrary, the information revolution will in all probability further reinforce the pressure for transparency. (B) Proceed in practice. We need to develop strategies to facilitate many different and complex processes of accountability. It is, not sufficient to describe accountability as something complex, with certain patterns of interdependency and redundancy. Now we have to deal with these processes, guide them and turn them into a mutually reinforcing dynamic. Now we need to establish which strategies will work under which circumstances. (C) Proceed in research. An important requirement to get valuable insights in these strategies is more systematic, comparative and cumulative empirical analysis on accountability. Yet, the foundation of such a scholarly ambition is an agreed upon conceptual framework. This article may be a modest contribution to this coming to terms with accountability. References Acar, M., Guo, C., & Yang, K. (2008). Accountability when hierarchical authority is absent: views from public private partnership practitioners. The American Review of Public Administration, 38(3), Aucoin, P., & Heintzman, R. (2000). The Dialectics of Accountability for Performance in Public Management Reform. International Review of Administrative Sciences, 66(1), Ayres, I., & Braithwaite, J. (1995). Responsive regulation: transcending the deregulation debate Oxford: Oxford University Press. Barberis, P. (1998). The New Public Management And A New Accountability. Public Administration, 76(3), Behn, R. (2001). Rethinking Democratic Accountability. Washington D.C.: The Brookings Institution. Bellamy, R. (2011). Symposium on Democracy and New Modes of Governance. Government and Opposition, 46(1), Bevir, M. (2010). Democratic Governance. Princeton, N.J.: Princeton University Press. Bishop, M., Kay, J., & Mayer, C. (Eds.). (1994). Privatization and Economic Performance. Oxford: Oxford University Press. Black, J. (2008). Constructing and contesting legitimacy and accountability in polycentric regulatory regimes. Regulation & Governance, 2(2), Bovens, M. (2007). Analysing and Assessing Accountability: A Conceptual Framework. European Journal of Law, 13(4), Bovens, M., Schillemans, T., & 't Hart, P. (2008). Does Public Accountability Work? An assessment tool. Public Administration 86(1), Braithwaite, J. (2008). Regulatory capitalism : how it works, ideas for making it work better. Cheltenham, UK: Edward Elgar. Considine, M. (2002). The End of the Line? Accountability and Governance in