1
A research proposal.
Multilevel governance and public accountability in Europe:
Which institutions, which practices, which deficit?
Prof. dr. M.A.P. Bovens ()
Prof. dr. P. ‘t Hart ()
Prof. dr. D. Curtin ()
Dr. M. van de Steeg ()
Utrecht School of Governance
Utrecht University
Bijlhouwerstraat 6
3511 ZC Utrecht
The Netherlands
Draft, made for CONNEX, Research group 2: Democracy and accountability in the EU
1. Introduction
This paper contains a research proposal on multi-level governance and public accountability in relation to EU decision-making. The proposal has been accepted, and will be funded by the Dutch Scientific Council, NWO. Soon, a post-doc and two Phd-researchers will start analyzing the extent to which European Council decisions, expert input and European agencies are matched by public accountability arrangements. The paper on public accountability presented at this CONNEX workshop by Mark Bovens provides the research project with a conceptual framework. At this stage, any comment would be helpful in further improving the research design. We are currently working on an operationalization of the public accountability concept.
2. Backgrounds and aims
The so-called displacement or diffusion of national, state-based `politics’ is one of the crucial contemporary `shifts in governance’ (Kohler-Koch, 2003). The diffusion and perhaps erosion of state-based politics proceeds along different paths (see Van Kersbergen and Van Waarden, 2001), but in this programme we focus on the trend towards regionalisation of public policymaking, in particular within the framework of the EU. In recent decades much of the agenda of European studies in various social science disciplines (IR, Political Science, Public Administration, Law) has been firmly focused on the trend toward European policymaking. Researchers have sought to document, explain and evaluate emerging practices of multilevel governance, both conceptually/theoretically (Marks et al, 1996; Sandholtz and Stone Sweet, 1998; Hooghe and Marks, 2001; Wiener and Dietz, 2004) and empirically (see Quermonne et al, 1996; Soetendorp and Hanf, 1998; Kassim, Peters and Wright, 2000a, b).
Among these scholars, there has long been concern that this development is not being matched by an equally forceful creation of appropriate accountability regimes (Schmitter, 2000). Accountability deficits are said to be a key cause of the low public visibility and legitimacy of the EU (Scharpf, 1999). The new `constitution’ addresses some of these problems, but it is doubtful whether it will also tackle the large and growing non-legislative component of European governance: i.e., technocratic and intergovernmental arenas such as comitology, the European Council/Council of Ministers, as well as the many non-majoritarian agencies. Gaping accountability deficits may continue to exist and even grow, compromising the legitimacy of the European polity (Curtin, 2004; Bergman and Damgaard, 2000). In contrast, others, such as Moravscik (2002, 2004), claim that compared to the way national democracies operate in practice, the EU does not suffer from a fundamental democratic deficit precisely because some of its critical institutions are relatively insulated from the vagaries of direct political interference.
Which of these competing claims is right, and what does this imply for institutional (re)design of existing and new accountability regimes? At present there is very little in the way of robust empirical findings that deal with these issues. This gap needs to be filled, and in a way that does justice to the full scope of the issue, and not just concentrates on the most visible and most intensely debated part of it. For that reason, we discern in this programme different distinct manifestations of `politics’ (defined as processes of deliberation and decision-making regarding the binding allocation of values for a community, Easton, 1965): the three core stages of the public policymaking process (design/preparation; choice; implementation), plus the increasingly salient category of crisis management, i.e. responding to unscheduled events and acute threats (cf. Howlett and Ramesh, 1995; Boin et al, 2005). These are not exhaustive (agenda-setting, electoral politics, and social mobilisation are not included), but nevertheless essential faces of politics in any governance process.
Policy choice and crisis management tend to be situated in frontstage, elite politics, in highly visible and classic arenas such as cabinets, parliaments and news conferences populated by elected representatives and politically appointed executives. Public policy design and implementation are more likely to involve civil servants, public organizations, experts and lobbyists operating in backstage bureaucratic arenas. In many national systems, specific accountability regimes have evolved over decades/centuries to accommodate these various modes of political choice and action. But how do these regimes adapt, now that at least part of the political action in each of these fields has been moving to the European level?
A full understanding of the extent and consequences of this dispersion of politics and governance requires insight into all of its manifestations. Scholarly attention has been overwhelmingly devoted to the classic politics of multilevel policy choice and law making. Much less effort has gone into studying the other manifestations of the europeification of politics. The politics of `crisis management’ has only just begun to draw attention as a substantive dimension of multilevel governance (cf Jorgensen, 1996; Kelstrup and Williams, 2000; Gronvall, 2001; Van Ham and Medvedev 2002; Olsson, Larsson and Ramberg, 2004). Likewise, the `backstage’ politics of policy preparation (committee processes and bureaucratic politics) in Europe remain underexplored relative to their importance in shaping the process and content of European public policymaking (see Peters, 1992; Van Schendelen, 1998; Richardson, 2000; Rhinard, 2002; Kassim, 2003; Page, 2003). The same goes for European-level implementation of EU policy (as opposed to `europeanisation’ of national policies, e.g. national transposition and implementation of EU law at the national level, see e.g. Featherstone and Radaelli (2003).
This same imbalance also exists in the field of accountability studies (cf. Harlow, 2002; Curtin, 2004), and that is where this programme seeks to make a contribution. The general aim of the programme is to investigate to what degree and how trends towards multilevel policymaking and implementation in Europe have been matched by correspondent changes and innovations in public accountability regimes and practices. In keeping with the distinctions made above, it undertakes research on accountability practices in four domains of politics. It entails systematic comparative empirical research on accountability regimes surrounding: European heads of government in both regular policymaking and crisis management (project 1); experts and/in committees engaged in policy preparation (project 2); European agencies engaged in policy implementation (project 3).
3. Conceptual and theoretical foundations
Before we present these three projects in greater detail, we briefly discuss the analytical agenda of the programme as a whole. Based on the applicants’ ongoing research into public accountability (Bovens, 1998, 2005a, b; ‘t Hart, 2005), EU constitutional architecture and accountability processes (Curtin, 2005), comparative political evaluation (Bovens, ‘t Hart and Peters, 2001), and the multilevel governance and public accountability literatures, the projects within the programme will be given a common conceptual foundations for empirical research.
a. Multi-level governance theory and research
The concept of ‘multilevel governance’ has arisen in opposition to conceptualisations of the EU that assume hierarchical relations between clearly delineated ‘levels’ of government (EU – national government – local government). By contrast, multilevel governance is characterised by: (a) the existence of decision-making centres at multiple levels of government; (b) that are not clearly hierarchically ordered (even though there may be formal relations between them); ad (c) whose decision-making processes are mutually intertwined (cf. Hooghe and Marks 2001; Kohler-Koch 2003; Kohler-Koch and Eising 1999). As a result, the EU harbours a complex system of intergovernmental relations on different ‘levels’ of government that has profound consequences for policy-making and implementation, as well as for accountability norms and practices.
b. Accountability theory and research
‘Public accountability’ is not just another political catch phrase; it also refers to legally and otherwise institutionalised practices of account giving (Bovens, 1998, 2005b). Accountability refers to a specific set of social relations that can be studied empirically. This raises taxonomical issues: when does a social relation qualify as ‘public accountability’?
Accountability can be defined as `a social relationship in which an actor feels an obligation to explain and to justify his or her conduct to some significant other’ (Day & Klein 1987:5; Romzek & Dubnick 1998:6; Lerner & Tetlock 1999:255; McCandless 2001:22; Pollit 2003:89). After having heard and questioned the actor on her or his conduct, the significant other comes to a judgement that might have consequences for the actor (Mulgan 2003, Bovens 2005c). This relatively simply defined relationship contains a number of variables. The actor, or accountor, can be either an individual or an agency. The significant other, i.e. the accountability forum or the accountee (Pollitt 2003), can be a specific person or agency, but can also be a more virtual entity (‘the public’).
It is important to distinguish the concept of public accountability from other concepts such as responsiveness and transparency that are often used in the same context (Mulgan 2003: 21; Harlow 2002: 185; Bovens 2005c). What distinguishes accountability from responsiveness, participation, or the legislative function of Parliament is that rendering accountability entails a justification of conduct. While participation takes place when a decision is being prepared, accountability comes in when decisions have been made. Transparency is also insufficient as an indicator of accountability (Schillemans & Bovens 2004: 28; Bovens 2005c). In order to be considered as an accountability arrangement, the information obtained via practices of transparency should be taken one step further. For an actor to be accountable, information is given to a forum, which then comes to a judgement that may have consequences for the actor in case it is negative.
Focusing mainly on more formal, institutionalised accountability relationships, this programme shall, where appropriate, include and compare the operation of not only political (specifically democratic, see Gustavsson, 1999), but also hierarchical, legal, and professional accountability types (Romzek, 2000). Hence in each project we shall specify the nature and locus of various accountability modes that together constitute an accountability regime by asking basic descriptive questions: who is rendering account, to which forum, by which means, and with which purposes? We can then situate them in a broader grid/typology of accountability regimes (see further Bovens, 2004).
Furthermore we open the black box of the accountability process. The relationship between the actor and the forum, the account giving, usually consists of at least three elements or stages. First of all, the actor must feel obliged to inform the forum about his conduct, by providing various sorts of data about the performance of tasks, about outcomes, or about procedures. Secondly, the information can prompt the forum to interrogate the actor and to question the adequacy of the information or the legitimacy of the conduct (debating phase). Thirdly, the forum usually passes judgement on the conduct of the actor. In case of a negative judgement the forum may impose some sort of sanctions on the accountor. These may be formal, such as fines, disciplinary measures, or dismissal, but they can also be implicit or informal (such as negative publicity). The three projects will ascertain to what extent the various accountability regimes entail each of these stages of an accountability process, and will study the relevant processes to examine how they unfold, and what pattern of relations between accountor and accountee exists.
Each project will answer the following three questions:
1. What are for the object of study the formal structures and actual practices of (public) accountability? A map of accountability arrangements will be developed for the European Council, European agencies and experts in committees.
2. How can the accountability arrangement be assessed from a democratic, constitutional and learning perspective? These perspectives are developed further Mark Bovens’ paper for this workshop. From a democratic perspective, the question is whether the accountability arrangements offer enough incentives to the accountor to commit himself to the agendas of its democratically elected accountees. From a constitutional perspective, an accountability arrangement should create the checks and balances needed for a balance of power. From learning perspective, the assessment is based on the degree to which an accountability arrangement stimulates administrative bodies and officials to achieve a higher awareness of the environment, increases self-reflection and induces the ability to change.
3. Does the accountability arrangement respond to the demands put by the practice of multi-level governance? Does multi-level governance create specific problems for the accountability arrangement identified?
c. Reconstructing accountability practices in multilevel governance
Multilevel governance patterns lead to distinct accountability problems because the participation of multiple actors at different levels of government leads to a dispersion of influence and responsibility. Although the literature on democracy and accountability in the EU still focuses on the European level per se (Harlow 2002; Lord 1998; Moravcsik 2002), some are studying issues of accountability while taking into account the specific character of systems of multi-level governance (Papadopoulos 2003; Slaughter 2004).
The fact that the EU is a system of multi-level governance makes the design of accountability arrangements more difficult. One way of analysing the effect of multi-level governance on accountability is to distinguish between a vertical and a horizontal relationship between the accountor and the accountee. A vertical or top-down model of accountability mirrors a chain of delegation of authority in a series of principal-agent linkages: from voters to elected politicians to political executives to bureaucratic agencies, both at the national or transnational level. The key accountability problem is how principals can make sure that the agents act in line with the principal’s preferences, given that these agents may be motivated to maximize their freedom from scrutiny and oversight and that they often possess ample means – such as monopolies on information about performance – to do so (Strom, 2000). European multi-level governance exacerbates this problem. The EU adds on many occasions another level between the principal par excellence, namely the voter, and the numerous agents ultimately active on their behalf in European politics. The distance between principal and agent increases. Since agents act often collectively with colleagues from other Member States, it becomes even more difficult to assign clear responsibilities to an agent. In multi-level governance, agents are likely to have several principals, and each principal might be entitled to monitor only a part of agents’ activities on the European level. From the principal-agent perspective, multi-level governance may lead to fragmentation and distance, which could complicate accountability mechanisms.
This programme asks how given national/supranational accountees (principals) control the behaviour of national and supranational actors (agents) who operate in European governance arenas on their behalf. How do these accountees cope with the growing importance (formal competence and political prominence) and the institutional changes (composition and decision rules) of various transnational arenas in which these actors operate? And how do they accommodate the emergence of European-level institutions, often deliberately designed to stay aloof from majoritarian politics?
Some of the problems caused by multi-level governance in a top-down model of accountability may be overcome by arrangements based on a horizontal relationship between accountor and accountee. For example, from the constitutional perspective accountability can be considered as a bundle of rules and principles regarding checks and balances on public power that may be institutionalised in various ways (Curtin, 2005). The chief design issue here is to make sure that public power is always matched by public accountability, which can but need not take the top-down, principal-agent form central to the first model. A horizontal model is more open to non-hierarchical (such as professional, legal etc.) accountability forms, and non-delegatory political accountability forms (such as those involving direct, clientele-based accountees). The model’s key analytical question for this programme is which configurations of existing and new accountability forms and practices have been developed (and are still emerging) to act as checks upon the evolving politics of multilevel governance in (and beyond) the European Union system.
4. Relevance
As explained above, this programme is situated at the cutting edge of research in comparative, multilevel and European governance in political science, public administration and legal studies, the three disciplines represented by the programme’s three applicants. It addresses an issue – democratic accountability of and in the European Union - that is not only hotly debated among scholars but in public and political discourse as well.
5. Project 1: Controlling European leaders in policymaking and crisis management
In the first research project we examine how national government leaders are held accountable by political accountability forums prior to and following peak `occasions for decision’ (see, e.g., Paige, 1968) with regard to more or less regular, scheduled processes of European institution-building and policymaking, as compared to ad-hoc, unscheduled crisis management episodes. We also try to ascertain the significance (impact) attributed to these accountability processes for these leaders’ policy stances and political positions.
To our knowledge, no prior research of this kind exists. Still its significance is clear. If we accept that what European leaders collectively agree on and decide in European Council meetings and other, less formal gatherings, is highly consequential for the development of European public policy and for the institutional development of the European polity as a whole, and if we take into consideration that these European leaders do not have to answer directly to the European parliament or any significant European level public opinion for their postures and actions on the European stage, then it becomes very important to ascertain to what extent and effect national accountability regimes monitor, constrain and shape their behaviour, and lend legitimacy to it.
Our questions are:
a. To what extent and how is heads of government’ performance on the European stage is subject to democratic scrutiny, both prior to and after major summit meetings?