Agencification and the EU as a Regulatory State: a Framework for the Study of European Community Agencies
Helena Ekelund, University of Nottingham
Paper prepared for the UACES Student Forum 8th Annual Conference, Nottingham 19th-20th of April
Agencification and the EU as a Regulatory State: a Framework for the Study of
European Community Agencies
In recent years we have witnessed a number of changes in the institutional structure and power relations between the various institutions of the European Union (EU). One notable change is the rapid increase in the number of European Community Agencies, henceforth referred to as ‘agencies’. Currently, the EU has some twenty agencies, the majority of which have been established during the 1990s and 2000s. In my PhD project I aim to explain this development. At the conceptual level, this will involve dealing with the concepts and misconceptions of the debates on agencies. At the empirical level, it will involve mapping and comparing of the European Community Agencies. By studying the establishment of agencies we can gain important insights into how the EU actually works and, hence, my project will also contribute to the debates on the transformation of governance in Europe.
One strand of governance research is the study of the rise regulatory policy-making, which characterizes the EU as a ‘regulatory state’ (c.f. Kohler-Koch and Rittberger 2006: 35). In this paper, I will review literature on the regulatory state and agencies. I will elucidate the main components within the regulatory state concept and relate these to the EU and the study of Community Agencies. Particular attention will be given to the normative implications of agency establishment and analytical dimensions to consider in the study of agency creation.
The Regulatory State
The rise of the regulatory state refers to a process through which the regulatory function of the state gain prominence at the expense of redistribution and macroeconomic stabilization. In Europe we can also see how the traditional European model of public ownership have given way for the American model of private ownership subjected to rules created and upheld by regulatory agencies. Whilst acknowledging the fact that there are differences between the public administrations of various European countries, it makes sense to speak of an ideological cleavage between the United States and the Europe regarding the principles of public sector management. This cleavage can be attributed to differing ideological views on the functioning of the market. The traditional American view is that the market normally functions well, and that it is only justified to interfere with the market in clear cases of market failure (Majone 1996: 50). Subsequently, the management of public utilities has been left largely in private hands, and the threat of market failure has been addressed by subjecting the private owners to regulation, which is developed and upheld by agencies or commissions (Majone 1996: 15). In Europe, by contrast, there has been more suspicion against the market. The traditional response has been public ownership, i.e. that the state manages the provision of public utilities. According to Majone (1996: 11), it was generally assumed that public ownership gave governments increased ability to regulate their economies and to protect public interests. However, the opinion in Europe changed as public ownership did not seem to deliver the expected benefits, and, in the 1980s, there was a macro-economic paradigm shift from Keynesianism to more neo-liberal solutions (Műller and Wright 1994: 2). In practical terms, this led to privatisation of industries, but also “the creation of new instruments of regulation and the establishment of new regulatory authorities” (Levi-Faur 2005: 19, c.f. Gilardi 2002; Moran 2001; Thatcher 2002a). However, the regulatory state does not only aim to stop market failure by engaging in economic regulation. Social regulation, such as worker and consumer protection, has become increasingly important (c.f. Moran 2002: 394-395).
When studying the regulatory state it may prove useful to break up the concept into smaller parts, which can be operationalised more easily. Minogue (2002: 654) reiterates six key variables for the design of an effective regulatory state that Majone outlines in an article from 1999. They are:
(i) the extent to which decisions are delegated to an independent agent rather than taken by the political principal
(ii) the nature of the structure of governance itself, particularly in determining the agent’s degree of independence from the political process
(iii) the rules that specify the procedural framework eg reason giving requirements, consultative processes (sic)
(iv) the scope for political principals to overrule agency decisions
(v) the relative autonomy of financial resources
(vi) the extent of ex post monitoring, eg legislative oversight, judicial review, citizen’s complaints procedure (sic)
The EU as a Regulatory State
In my PhD thesis I intend to systematically analyse the idea of the EU as a regulatory state through the application the concept of the regulatory state to the study of Community Agencies. The idea of the EU as a regulatory state can be attributed to Majone, at the core of whose vision we find the idea of market regulation being a more important task than income redistribution and macroeconomic stabilisation. Majone (1996: 63) argues that it is important to distinguish between regulatory policies and non-regulatory policies, for example income redistribution. The key difference between them is that regulatory policy-making is less dependent on budgetary constraints, and states’ budgets are to a large extent dependent on states’ ability to tax and level of tax revenues. The EU budget is rigid and small in comparison to the budgets of member states, and the EU has no independent tax power (Majone 1997: 150). Thus, it is very difficult for the EU to engage in non-regulatory policies, and if the Commission wishes to increase its competencies, which Majone (1996; 1997) assumes that it does, it has to do so through increased regulatory activity. The EU can do so because most of the costs that arise from regulatory policy-making are borne not by the regulators but by the regulated, wherefore budgetary considerations only impose soft constraints on regulators (Majone 1996: 64). At EU level the distinction between regulatory and non-regulatory policies in terms of costs is further exacerbated by the fact that most costs for implementation of EC rules are directly or indirectly borne by the member states (Majone 1996: 64).
Majone’s ideas find their analytical expression in a supply and demand model of regulation, which borrows from public choice models of bureaucracy and neofunctionalism (c.f. Majone 1996: 64-68). On the demand side of the model we find multinational and export-oriented companies, public interest groups and member states. As predicted by neofunctionalists, the functional needs of the single market have required substantial transfer of policy-making power to the EU level (Majone 1996: 66). Companies tend to favour EU level regulations over national ones as they remove the costs resulting from demands to adhere to different national standards. Public interest groups also demand EU level regulation when they are unable to make their national government impose the regulation they wish for. This is typically the case of interest groups, for example environmentalist and consumer protection groups, from countries with low health and safety standards (Majone 1996: 67). However, according to Majone (1996: 67-68) the strongest demands for EU level regulation come from the member states, which may gain by having their national standards incorporated into EU law. For example, a country with high standards of health and safety protection in the work place would see the comparative advantage of countries with low levels of said protection reduced if the EU decides to adopt their high standards.
The main role on the supply side is played by the Commission, which is assumed to have “a utility function, which it attempts to maximize, subject to constraints” (Majone 1996: 64). It is assumed that the Commission wishes to maximise its influence, which Majone measures by the scope of Commission competences. Majone argues that the Commission is still a bureaucratic organisation under development and, therefore, “the definition of competences” is the key issue for the Commission (Majone 1996: 65).
The characterisation of the EU as a regulatory state has significant normative implications for our conceptualisation of legitimacy and accountability.
Proponents of the regulatory state are likely to argue that the regulatory state achieves legitimacy through outputs rather than inputs. The difference between input-oriented and output-oriented ideas of legitimacy can be illustrated by the statements ‘government by the people’ and ‘government for the people’. A political system that relies on input legitimacy (government by the people) is a political system where the focus is on participation, and where individuals do not fear majority rule due to the existence of “a pre-existing collective identity” (Scharpf 1999: 10). Output-oriented legitimacy, by contrast, can be achieved from a “capacity to solve problems requiring collective solutions because they could not be solved through individual action, through market exchanges, or through voluntary cooperation in civil society” (Scharpf 1999: 11). For this to succeed, Scharpf (1999: 11) argues, it is sufficient that individuals in the political system perceive to have “a range of common interests that is sufficiently broad and stable to justify institutional arrangements for collective action”. If we accept output legitimacy as sufficient for the EU we have also made a normative stand on the issue of the democratic deficit of the EU.
As Majone (1997: 159) argues, the regulatory state, where “technocratic experts” employed by various agencies perform a broad variety of tasks, will undoubtedly suffer from a democratic deficit if democratic legitimacy is defined as “direct responsibility to the voters or to the government expressing the current parliamentary majority”. However, the Madisonian model of democracy, the goal of which is “to protect minorities form the ‘tyranny of the majority’, and the judicial, executive and administrative functions from representative assemblies and from fickle mass opinion” provides an alternative model of democracy where delegation is less problematic from a legitimacy point of view (Majone 1997: 160). Here delegation is used as a strategy to restrain rule by the majority by giving authority to non-elected officials, who “have limited or no direct accountability to either political majorities or minorities” (Majone 1997: 160).
Moreover, agencies can enjoy “[p]rocedural legitimacy”, which means that they have been created by and following correct procedures (Majone 1997: 160). For example it can imply that agencies have been created in accordance with existing rules, that their creation, objectives and legal authority are decided upon by elected officials and that their activities are open to review (c.f. Thatcher 2002b: 958). Arguably, the establishment of independent regulatory agencies in many parts of Europe has lead to “increased openness in the practice of decision-making” as decision-making concerning utilities now “remain subject to political and administrative public scrutiny” rather than being “largely closed and private processes” (Thatcher 2002a:966).
However, delegation at the EU-level may upset the norm of “institutional balance” and the “reciprocal duty of loyal cooperation” shared by EU institutions and national authorities (Majone 2002: 327). Delegation to yet other bodies than the ones created by the Treaties could be perceived as a violation of “fundamental, and presumably immutable, principles of the communitarian system” (Majone 2002: 321).
Definitions of “Agencies” and the “Agency Programme”
The term “agency” is not easily defined because, as Pollitt et al. (2001: 273) point out, public law varies between countries, and the institutional design of agencies is “characterized by extreme empirical heterogeneity” (Gilardi 2002: 874). Therefore, an important task when developing a framework for the study of Community Agencies is to identify critical dimensions to the “agency concept”. A review of relevant agency literature provides the following definitions.
Thatcher (2002b:956) defines an independent regulatory agency as “a body with its own powers and responsibilities given under public law, which is organizationally separated from ministries and is neither directly elected nor managed by elected officials”. Whilst acknowledging that the formal institutional status of an agency is important, Thatcher (2002b: 955) also clearly points out that they are incomplete and do not determine everything about an agency’s behaviour; control mechanisms and powers outlined in the formal documents can be used in a variety of ways.
Talbot (2004: 4) shows that “agency” can be defined “by exclusion”. For instance, “state-owned enterprises whose primary existence is within the market (albeit with subsidies) rather than in public services, and which are usually joint stock companies founded in private law” can be excluded as can “social, charitable and voluntary organisations even where their primary funding comes from the state” (Talbot 2004: 4-5).
After having specified these exclusions, Talbot (2004: 5) sets up five criteria that must be satisfied for an organisation to be considered an “agency”. He specifies that the organisation ought to be:
- at arm’s length (or further) from the main hierarchical ‘spine’ of central ministries/departments of state
- carrying out public tasks (service provision, regulation, adjudication, certification) at a national level
- staffed by public servants (not necessarily ‘civil servants’)
- financed (in principle) by the state budget (in practice some are financed up to 100 per cent from their own revenues, but even here the state remains liable for their financial condition)
- subject to at least some public/administrative law procedures (i.e. they are not predominantly or entirely private law bodies).
Talbot (2004: 6) then continues to discuss what he perceives as central elements to “the ‘agency’ programme”. The first idea is the idea of “[s]tructural disaggregation and/or the creation of ‘task specific’ organisations” (Talbot 2004: 6). By this, Talbot (2004: 7) refers to “the splitting up of larger bodies into a ‘parent’ body and various subordinate ‘agencies’. The following elements can be identified within this structural change (Talbot 2004: 8.
- Creating an identifiable, separate, organisational structure with its own name.
- Usually the body is given a single, or small set, of functions.
- Usually these functions are primarily about ‘delivery’, ‘execution’ or ‘provision’ and not about policy-making.
- Giving the above body a clear ‘constitution’ – in the form of some sort of legislation, or at least a formal (if not statutory) ‘framework document’ – which sets out its purpose, powers and governance arrangements.
- Making a single, named individual – usually called a Chief Executive (CE) – responsible for managing and reporting on the new body. This person I usually appointed – actually or potentially – through an open process, separate form the normal civil service recruitment.
- Making the staff of the new body different in some way from mainstream civil servants, usually by changing their formal employment status.
- Putting in place formal reporting arrangements, including separate accounts, for the new body.
The second idea is the idea of “[p]erformance ‘contracting’ – some form of performance target setting, monitoring and reporting” (Talbot 2004: 6), which refers to “any system of setting targets for, and reporting on (non necessarily publicly) the activities of an agency” (Talbot 2004: 14). The targets may be set by the agencies themselves (as is often the case in the US), by the government (typical of the UK) or by the two of them in combination, which is common in Sweden where ministries draw up the broad guidelines and leave the details to the agencies (Talbot 2004: 15).
The third idea that Talbot (2004: 6) sees as central to agencies is the idea of “[d]eregulation (or more properly reregulation) of controls over personnel, finance and other management matters”. Briefly speaking, this refers to the rules that are set up for how public bodies must function (Talbot 2004: 13). The autonomy of agencies in this respect may vary over time. As Talbot (2004: 13) writes, one strategy has been to give agencies a certain level of autonomy once and for all. Another strategy is to gradually increase an agency’s autonomy over these matters if the agency proves that it can be trustworthy.
James (2001: 235) presents an agency model made up of two “core features” derived from the UK “Next Steps” reform. The first feature is “the use of organizational units called agencies to handle distinct central government activities on behalf of ministries” (James 2001: 235). The second feature “is the creation of a ‘regulatory’ framework for each agency”. This framework, James (2001: 235) continues, consists of “a designated chief executive, who is accountable for performance, and a framework of performance requirements”. The idea behind having a chief executive officer is to ensure that there is one individual who assumes responsibility for the overall performance of the agency (James 2001: 236). The idea behind the setting of performance targets is to ensure that the interests of those people who are affected by the performance of the agency are protected (James 2001: 238).
Variation in agency design
In this section I will discuss literature related agency design, with a particular focus on how different academics have mapped differences in agency design.
Horn (1995: 25-26) lists the following five institutional choices.
i. the extent to which decisions are delegated to the administrative level rather than taken by the legislature; especially the degree of legislative vagueness and the extent of ex post legislative direction to administrators;
ii. the governance structure of the administrative agent, especially the way senior personnel are selected, the degree of statutory independence from the legislature, and the jurisdiction of the administrative agent;
iii. the rules that specify the procedures that must be followed in administrative decision making, which typically define the rights that constituents have to participate directly in the administrative decision-making process;
iv. the nature and degree of legislative monitoring of administrative decision making and the ability to use ex post rewards and sanctions;
v. the “rules” governing the allocation and use of capital and labor, in particular, the extent to which agencies are financed by sales revenues rather than taxes and the administrators’ employment conditions.
Other scholars to map agency design are Bouckaert and Peters (2004: 38-43) who present seven different types of activities that are often delegated to agencies. Implementation is the first agency activity that Bouckaert and Peters (2004: 38) lists. Agencies within this category can be involved in “direct service delivery” or they could be in charge of “the transfer of funds”, which means that they “make decisions about grants and contracts as a means of insulating these decisions from direct political control” (Bouckaert and Peters 2004: 39).
Regulation of the economy and of society is another agency activity (Bouckaert and Peters 2004: 39). A perceived advantage of this type of agency is the possibility to isolate or distance regulatory functions from political pressures. An added advantage is that a regulatory agency may involve the interests of those affected by the regulation in its regulatory activities (Bouckaert and Peters 2004: 40). Obviously also governments can do this, but the interests affected might liaise more closely with an agency actively working in their field than a government that has their field as one area to regulate among many.
A third task often delegated to agencies is to provide advice and assist in the development of policies related to the agency’s area of expertise. The idea here is that these agencies are capable of giving unbiased suggestions on policy issues of major concern, and they involve experts in their policy-making (Bouckaert and Peters 2004: 40). The fourth task mentioned by Bouckaert and Peters (2004: 41) is information gathering and dissemination, which they argue is delegated for two reasons. Firstly, delegation would mean that the gathering of data is at least partially insulated from direct political pressures. Secondly, and this concerns particularly collection and dissemination of important socio-economic data such as criminality, poverty and unemployment, there are situations where it is of uttermost importance that objectivity is maintained.
Research is the fifth type of agency activity discussed by Bouckaert and Peters (2004: 41-42). The rationale behind delegating research tasks to independent agencies is the wish for objectivity as regards results, but also a desire to give research organisations the opportunity to decide themselves what to research. As Bouckaert and Peters (2004: 41) writes, governments want to “prevent the formation of some form of official science”. However, this is not without problems as ultimately the governments are most often in charge of the money, and hence are able to influence the research agenda.
Furthermore, Bouckaert and Peters (2004: 42) mention agencies functioning as tribunals and public enquiries. These agencies perform tasks of a judicial nature, such as assessing the legality of activities and actions within a specific policy area. It has been suggested that these agencies can be very powerful. Indeed, Braithwaite (2000: 229) argues that business regulatory agencies have grown to be “more significant law enforcers than the police”. A reason behind the creation of these agencies could be that the regular court system in welfare states simply cannot cope with the adjudication required for all the many government programmes devised (Bouckaert and Peters 2004: 42).
Finally, Bouckaert and Peters (2004: 43) argue that there are agencies created to provide “representational and participatory opportunities for segments of the civil society”. As an extreme example of the latter they mention the Norwegian Sami Parliament (Sámediggi), which, Bouckaert and Peters (2004: 43) argue has been “described as an affiliated organization of the Department of Local Government and Regional Development in Norway”. However, this description is inaccurate as the overarching responsibility for Sami politics in Norway comes under the Ministry of Labour and Social Inclusion with other departments responsible for the implementation of Sami politics within their field of work (c.f. Arbeids- og Inkluderingsdepartmentet 2007, http://odin.dep.no/aid/norsk/tema/same/p30003434/nn.html).
Bouckaert and Peters (2004) also mention some other dimensions along which agencies may vary in practice. Firstly, there are differences between agencies as regards “their relationships with ministerial authority” (Bouckaert and Peters 2004: 29). There can also be differences as regards legal status, “the degree of managerial involvement of ministries and other political entities or officials” and the financial relationship between the agency and the government (Bouckaert and Peters 2004: 37). The concept of autonomy is central to all these dimensions, and agency autonomy over issues such as financial matters, appointments of personnel and policy decisions may vary substantially (Bouckaert and Peters 2004: 29). As stated by Bouckaert and Peters (2004: 29), a simple conclusion that an agency is autonomous “may disguise substantial variation in their ability to act on their own and their capacity to influence public policy”.
Difference in legal status can be illustrated by Gains (2004) and McCubbins and Page (1987). Gains (2004: 56-58) outlines how political administration in the UK traditionally has worked following the so-called “Whitehall model”, which can be described as a model of non-statutory political administration organisations in which involved actors respect formal and informal rules, in particular the sovereignty of the Parliament and the importance of ministerial responsibility. In contrast, the American Congress draws up legal documents where scope, regulatory instruments delegated and procedures for the use of these instruments are seriously considered (McCubbins and Page 1987). Bans, quotas and taxes are a few of the legal instruments that may be assigned to agencies (McCubbins and Page 1987: 415). As suggested by the name, “scope” refers to the specification of what the agency may address, and it includes “the statement of the purpose of the act” of delegation (McCubbins and Page 1987: 412).
In their explanations of variation in agency design Levy and Spiller (1996: 1) stress the interaction between political and social institutions, economic conditions and regulatory processes. Levy and Spiller’s (1996: 7-9) analysis centres on how to match regulatory systems to the institutions of a political setting. To facilitate the mapping of the choices involved they construct a “decision tree”. Firstly, Levy and Spiller (1996: 7) argue that there is a distinction between “countries that have domestic institutions capable of credibly refraining from arbitrary administrative action and those that do not”. Here they suggest that countries with independent judiciaries whose decisions are enforced and regarded as impartial have better chances of achieving credible commitment. Countries that can achieve credible commitment can then be divided into two groups, those that use legislation to achieve credible commitment successfully, and those that “best achieve such commitment by embedding their regulatory systems in the operating licenses of private companies” (Levy and Spiller 1996: 7). Thirdly, Levy and Spiller (1996: 9) argue that there is a distinction between countries in which credibility can be achieved by specific, substantive rules and countries in which more flexible regulatory processes can be used. The fourth distinction mentioned by Levy and Spiller (1996: 9) is the one between countries with “strong administrative capabilities and those without”.
Agency design can also be explained by the perceived risk of “capture”, which refers to situations in which regulatory agencies get captured by the regulated interests (c.f. Thatcher 2002b: 958). A plausible idea related to this has come from “Chicago school” scholars who argue that there is a greater risk of regulatory agencies being captured by producers than by consumers. The reason for this is that producers, who are few and easily organised, often benefit significantly from regulation. Collective action is harder to achieve for consumers who, in contrast to producers, are numerous and dispersed (c.f. Thatcher 2002b: 956-958, Moran 2002: 394).
Thatcher (2002b: 963) examines the risk of “capture” of agencies by firms with help of three indicators. Firstly, he considers “the extent of the ‘revolving door’”, which refers to the possibilities of individuals to move between employment by a regulated firm to a regulatory agency, and back again. This, Thatcher (2002b: 963) maintains, gives “an indication of the ‘relation distance’” between the regulatory agency and the regulated firms. Secondly, he considers how many mergers have been blocked or have been subjected to conditions. This indicator aims to highlight relations between regulated firms and regulatory agencies in the competition field (Thatcher 2002b: 964). As a third indicator, Thatcher uses the number of times decisions made by regulatory agencies have been legally challenged. This, Thatcher (2002b: 963) argues, shows us whether or not there is a high level of conflict between the regulatory agencies and the regulated firms. A high number of legal cases against a particular regulatory agency would suggest that that agency has not been captured as “legal action represents a public and hostile challenge” (Thatcher 2002b: 963).
Thatcher (2002b) also examines the relationship between independent regulatory agencies (IRAs) and elected politicians. He lists the following five indicators of how elected politicians can control agencies (Thatcher 2002b: 959):
· party politicization of appointments: the greater the politicization of regulators, the lower the likely independence of regulators and the greater control by elected politicians;
· departures (dismissals and resignations) of IRA members before the end of their term (. . .): the lower the level of early departures, the more likely IRA independence;
· the tenure of IRA members: the longer their tenure, the greater their likely independence from elected politicians;
· the financial and staffing resources of IRAs;
· the use of powers to overturn the decisions of IRAs by elected politicians.
To conclude, it can never be sufficient to focus on formal definitions when studying agencies. This is particularly true as regards Community Agencies as the EU have chosen to give them all the same broad definition. Legal status, functions, governance structures, finance and personnel matters must be considered, and the concept of autonomy is central to all these aspects.
Drivers behind agency creation
An early contribution to the delegation literature is made by McCubbins and Page (1987), who in their analysis of delegation from the American Congress argue that delegation is perhaps most often explained by complexity arguments. By this they mean that the complexity of many problems faced by legislators requires considerable resources, and as resources are scarce, legislators decide to delegate legislative powers to agencies (McCubbins and Page 1987: 409). Degrees of complexity vary, and a logical assumption would be that we could expect more delegation in policy areas where there is a high degree of complexity. However, as we witness delegation in fields not normally deemed as enormously complex, it seems that complexity arguments cannot explain all different forms of regulation. McCubbins and Page (1987: 409) acknowledge this, and argue that delegation can also be explained by the “shift-the-responsibility” model, which they accredit to Fiorina. This model predicts that “the legislative choice to delegate authority to an administrative agency in preference to judicial enforcement of a legislative enactment follows when the act of delegating disguises the costs of the regulation to a larger extent than it disguises the benefits” (McCubbins and Page 1987: 410).
McCubbins and Page (1987: 414) also argue that “the degree of conflict and amount of uncertainty are the principal determinants of what form of delegation Congress chooses”. They suggest that:
with greater uncertainty and greater conflict there is a stronger incentive for Congress to pass the hot potato to the agency by broadening the scope and instruments of delegated authority. But while passing on the hot potato, there is no need to send over a loose cannon. With one hand, Congress broadens the scope and instruments; with the other, Congress tightens the procedural requirements. (McCubbins and Page 1987: 419).
Bawn (1995) puts forward a similar argument. Her central hypothesis is that “the degree of agency independence on any particular policy reflects the legislature’s willingness to trade uncertainty about policy consequences for uncertainty about agency behavior” (Bawn 1995: 63). Applied to Community Agencies Bawn’s (1995) and McCubbins’ and Page’s hypotheses would imply that a wide scope of activity and a large number of legal instruments would be given to agencies that operate in areas where there is considerable conflict and/or high degree of uncertainty. At the same time we could expect said agencies to be restricted by tight procedural requirements.
Horn (1995: 24) argues that explanations for variation in agency design can be explained by examining transaction costs, and his overarching argument is that “legislators choose those administrative arrangements that best address the transaction problems they encounter”. Transaction costs can be determined by “constitutional differences among countries” (Horn 1995: 9) and the level of uncertainty and conflict in a given policy area (Horn 1995: 15). High levels of uncertainty and conflict increases the legislators costs of adopting detailed, refined legislation wherefore high levels of conflict and uncertainty tend to lead to vague legislation, Horn (1995: 15) argues. Furthermore, Horn (1995: 11) maintains that “the institutional arrangements faced by administrators are likely to have a systematic influence on the type of person who seeks public sector employment, the type of employee who ends up being promoted to a position of responsibility, and the incentives he or she faces once appointed”.
Huber and Shipan (2002) takes a different view and argue that legislators tend to devise precise legislation when there is a high level of policy conflict, when legislative capacity is high, when there is no conflict between the two chambers in a bicameral system, and when “the political system does not allow the legislative majority to rely on nonstatutory factors to influence policy implementation” (Huber and Shipan 2002: 215). Hence, it follows that great discretion in delegation ought to be commonplace when the opposite is true. As regards operationalisation of “conflict”, Huber and Shipan (2002: 217) point out that it is necessary to consider the workings of the political system under study.
The starting point for Epstein and O’Halloran’s (1999: 9) analysis is an assumption that decisions about delegation are taken in order to maximise the political goals of legislators. Thus, delegation will take place when the benefits of delegation are perceived to outweigh the costs of delegation. After having reviewed previous literature on delegation, Epstein and O’Halloran (1999) concludes that there are three broad sets of explanations for delegation. The first one is “reasons of constituency relations”, which Epstein and O’Halloran (1999: 30) argue can come in three varieties. The first variety, and, according to Epstein and O’Halloran (1999: 30), the most benign, is that “legislators may wish to free up time to spend on constituency service, simultaneously taking advantage of agency expertise”. This argument can be found in a number of other texts on delegation (c.f. Pollack 2003; Pollitt et al. 2001). The second variety supposes that “legislators may gain by exposing their constituents to the vagaries of administrative regulation, content to play the role of ombudsmen who intervene if this process goes awry” (Epstein and O’Halloran 1999: 30). Finally, the third variety of constituency relations explanations “cast[s] legislators in the role of leading an organized ‘protection racket,’ threatening uncooperative constituents with harmful regulation if they do not contribute to legislators’ coffers” (Epstein and O’Halloran 1999: 30-31). The second set of explanations can be related to a notion that Epstein and O’Halloran (1999: 31) terms “regulatory lottery”, which supposes that interest groups that cannot agree in the legislative process may prefer delegation to agencies and a following regulatory uncertainty to a certain outcome that contradict their interests. Expressed in terms of a concrete hypothesis this would mean that the higher the level of uncertainty, the more delegation. Thirdly, Epstein and O’Halloran (1999: 32) argue that delegation to bureaucrats is often a preferred choice when “the policy in question has concentrated benefits and dispersed costs”. If this is true we could expect the extent of delegation to co-vary with the extent of benefit concentration and cost dispersion.
The wish to organise the public sector along market lines to force public sector organisations to be more efficient is another driver behind agency creation (Bouckaert and Peters 2004: 29). It has been argued that unbundling of the public sector into “distinct, single-purpose organizations will make it easier for key stakeholders to identify, participate in, and be consulted about the work of the organization” (Pollitt et al. (2001: 277). It could also improve “rank-and-file motivation” of agency staff, allow for more professional management, and improve various aspects of personnel management (Pollitt et al. 2001: 277).
A well-spread hypothesis concerning agency creation is the so-called credibility hypothesis, which stipulates that governments delegate powers to independent agencies in order to increase the credibility of their commitment to particular policies (c.f. Gilardi (2002). Credibility is a problem for elected politicians as their commitment to a particular policy tends to change over time due to the pressure of public opinion and the democratic process. With other words, policies are time inconsistent (c.f. Epstein and O’Halloran 1999; Gilardi 2002; Pollack 2003). How significant a problem this is depends on the policy area in question and on the level of impact the delegated activity has on society. However, on a general note Gilardi (2002: 876) concludes that “a lack of credibility is problematic because rational actors can anticipate future policy changes and thereby prevent the government from attaining its objectives”. This problem is exacerbated when the government cannot implement its policies through coercion (Gilardi 2002: 876). Given the EU’s lack of coercive powers, and the at times slow implementation of EC directives into national law, one could hypothesise that credibility is of significant importance for the EU, and, subsequently, that this is a common driver behind the establishment of Community Agencies. The more important credibility is perceived to be in a policy area, the more autonomous we could expect agencies in that policy area to be.
As a first step of an empirical assessment of the credibility hypothesis Gilardi (2002: 876-878) presents three related hypotheses. Firstly, Gilardi (2002: 877) hypothesises that “the more an economy is subject to ‘international interdependence’, the more delegation there should be”. International interdependence is measured “by the sum of trade, foreign direct investment (FDI) and international portfolio investment as a share of gross domestic product (GDP)” (Gilardi 2002: 880). The second hypothesis is that “delegation is more likely in sectors that have been recently subject to market opening” (Gilardi 2002: 877), and, thirdly, Gilardi (2002: 878) suggests the hypothesis that “there is a significant link between veto players and delegation”. Veto players is measured by a proxy variable “based on the number of parties in the executive and in the legislature, taking into account divided government and coalitional heterogeneity” (Gilardi 2002: 880). The dependent variable of Gilardi’s (2002) study is agency independence, which is measured by an index focused on formal independence. The index takes the following aspects into account: “the agency head status, the management board members’ status, the general frame of the relationships with the government and the parliament, financial and organizational autonomy, and the extent of delegated regulatory competencies” (Gilardi 2002: 880). Gilardi (2002: 884-888) concludes that there is a positive link between market opening and agency creation, but economic interdependence is not proven to have any significant impact. An intriguing result of the study is that veto players are relevant for the institutional design of agencies, but, contrary to what one might assume, there is a negative association between veto players and agency independence.
Blame-shifting is another potential driver behind delegation. The core idea behind blame-shifting as a driver behind agency creation “is that politicians seeking to claim credit and avoid blame from voters face a choice of direction or delegation within any policy domain, while voters or citizens choose between praising or blaming those who have direct responsibility in public policy” (Hood 2002: 17). One way for politicians to manage blame is blame avoidance through “agency strategies”, i.e. “the selection of institutional arrangements to minimize or avoid blame, for example in choosing between direct control and delegation” (Hood 2002: 16-17).
Hood (2002: 28) outlines three different blame-shifting strategies through delegation. Politicians may avoid blame for financial failure through privatisation. Operational and organisational problems can be avoided through managerialisation, and, blame for “judgemental failures” such as “false assumptions, faulty forecasts, technical errors or ignorance” can be avoided by “expertization”. The intensity of the wish to shift blame may vary between policy areas and subsequently this will be reflected in agency autonomy.
Agency creation can also be explained by historical institutionalism and the idea of path dependency. In her study of ‘Next Steps’ agencies in the UK, Gains (2004) argues that the new rules and norms that came about through agencification in the UK had to be developed alongside already existing rules. Gains (2004) maintains that the pre-‘Next Steps’ organisation of certain British departments has influenced what agencies were set up in connection to the unbundling of these departments and how the agencies came to operate; the UK’s choice to establish agencies on an administrative rather than a legal basis reflects a historical preference for non-statutory organisations. The Community Agencies, by contrast, have a clear legal foundation. Thus, a framework for the study of Community Agencies ought to consider whether a legal preference has led to the EU following a “legal path” in its agency creation. Previous EU work in the areas where Community Agencies have been established must also be considered.
Perceived benefits of a particular institutional form can also explain delegation to agencies (c.f. James 2001; McNamara 1998; 2002). The notion of ideas as a potent force is particularly relevant when considering the spread of an agency model from one country to another, and we can hypothesise that the degree of preference for an institutional form determines the likelihood of its spreading. “Policy-transfer”, which refers to “voluntary” as well as “coerced” adoption of policies, and “lesson-drawing”, which “includes most observation of an reflection on other countries’ experiences in order to draw positive and negative lessons” are two related terms used to describe processes through which policy ideas or tools developed in one political setting are exported to other political settings (James 2001: 241; cf. Dolowitz and March 1996; Rose 1993). Thus, interdependencies between countries and the ability and/or wish of international organisations to promote the spread of agencies are potential drivers behind agency creation (Gilardi 2005: 85).
DiMaggio and Powell (1991: 64, 66) argue that the fact that some organisational forms are “normatively sanctioned” makes it more likely that these forms are going to be adopted and that this homogenization of “organizational forms and practices” is best captured by the concept of institutional isomorphism. DiMaggio and Powell (1991: 67) discuss three types of isomorphism: coercive isomorphism, mimetic isomorphism and normative isomorphism. Coercive isomorphism “results from both formal and informal pressures exerted on organizations by other organizations upon which they are dependent and by cultural expectations in the society within which organizations function” (DiMaggio and Powell 1991: 67). Mimetic isomorphism occurs when organisations in times of uncertainty model themselves on other organisations (DiMaggio and Powell 1991: 69). Interestingly, “[t]he modelled organization may be unaware of the modelling or may have no desire to be copied; it merely serves as a convenient source of practices that the borrowing organization may use” (DiMaggio and Powell 1991: 69). Normative isomorphism refers to the diffusion of normative rules about the behaviour of professionals and organisations through similar education and the development of professional networks (DiMaggio and Powell 1991: 71). DiMaggio and Powell (1991: 74-77) suggest the following hypotheses concerning institutional change:
· The greater the dependence of an organization on another organization, the more similar it will become to that organization in structure, climate, and behavioural focus.
· The greater the centralization of organization A’s resource supply, the greater the extent to which organization A will change isomorphically to resemble the organizations on which it depends for resources.
· The more uncertain the relationship between means and ends, the greater the extent to which an organization will model itself after organizations it perceives as successful.
· The more ambiguous the goals of an organization, the greater the extent to which the organization will model itself after organizations that it perceives as successful.
· The greater the reliance on academic credentials in choosing managerial and staff personnel, the greater the extent to which an organization will become like other organizations in its field.
· The greater the participation of organizational managers in trade and professional associations, the more likely the organization will be, or will become, like other organizations in its field.
· The greater the extent to which an organizational field is dependent upon a single (or several similar) source(s) of support for vital resources, the higher the level of isomorphism.
· The greater the extent to which the organizations in a field transact with agencies of the state, the greater the extent of isomorphism in the field as a whole.
· The fewer the number of visible alternative organizational models in the field, the faster the rate of isomorphism in that field.
· The greater the extent to which technologies are uncertain or goals are ambiguous within a field, the greater the rate of isomorphic change.
· The greater the extent of professionalization in a field, the greater the amount of institutional isomorphic change.
· The greater the extent of structuration of a field, the greater the degree of isomorphism.
As delegation to agencies is a key aspect of the regulatory state, the literature on the regulatory state is a logical starting point when devising a framework for the study of Community Agencies. In contrast to the Keynesian welfare state, which places a significant focus on redistribution, the focus of the regulatory state is on regulatory policy-making. As the EU appears more geared towards regulatory policy-making than redistribution, a case could be made for considering the EU a regulatory state. However, this has significant implications for conceptualisation of the legitimacy of the EU political system. This is so because legitimacy in the regulatory state tends to be focused more on output-oriented legitimacy (government for the people) than input-oriented legitimacy (government by the people). One could also argue that delegation to Community Agencies upsets the delicate inter-institutional balance in the EU.
My literature review reveals that due to varying terminology and legal traditions between countries there is no single definition of what exactly an agency is. However, as all Community Agencies by nature is part of the same political system, that of the EC/EU, finding a definition to be used within my thesis ought not to be a significant problem. Rather the challenge is to what extent findings on agencies in other contexts, perhaps defined by other parameters, can be applied to the Community Agencies and vice versa.
A key part of my PhD project is to map the Community Agencies with regards to their functions, organisations, powers and so on. A number of relevant issues have been identified in the literature. For instance, it is important to consider the legal status of an agency and its relationship with central decision-making institutions. It is also important to consider financial, staffing, management and performance issues. For the mapping of the agencies to be useful it is essential to seek to explain this variation. Existing literature suggests that the functions an agency is expected to carry out has an impact on the design on the agency. The relationship between agencies, decision-making bodies and judicial bodies can also explain differences. Important factors to consider here are the level of outside influence on the agency as regards personnel, management and finance. The legal status of an agency, how well defined its mandate is, and what control mechanisms are imposed on the agency are other issues also of central importance that can explain agency operations. Variations in agency design could also be explained by the level of uncertainty and political conflict in the area that the agency is going to operate in as well as by the perceived risk of capture. As regards the impact of uncertainty and political conflict on agency design, somewhat contradictory hypotheses can be derived from the literature, which indicates that more research is clearly needed in this area.
Finally, this paper has sought to identify drivers behind agency creation. Different authors term their identified drivers somewhat differently, but there are certain ideas that recur. Delegation as a response to the wish to lower political transaction costs is one such idea. Political transaction costs can be lowered through increased efficiency in decision-making and the use of developed agency expertise. Efficiency can be gained if delegation of detailed policy-making to agencies leave decision-makers more time to deal with broader issues, and if the relevant policy-making expertise is present amongst agencies, these agency experts might be able to take complicated decisions faster than a non-expert politician. Another “version” of the expertise argument claims that as society has become more complex, decision-makers cannot have all expertise required wherefore delegation is the best way to ensure informed regulation in complex matters. Blame-shifting is another way to lower political costs. This means that decision-makers delegate with an intention to blame failed and/or unpopular decisions on the agency. This may be particularly attractive to politicians in policy areas where benefits of particular legislation are concentrated and costs dispersed. The need for credible commitment, particularly in areas where policy choices may not be consistent over time, is another often cited reason behind delegation. The essence of this argument is that by delegating responsibility for a particular policy to an agency, decision-makers cannot back down from their policy choice.
Other explanations for delegation are provided by sociological and constructivist perspectives. Authors from this tradition suggest that delegation to agencies take place because of a normative constructed belief of the superiority of this particular form of organisation. The concepts of policy transfer and isomorphism are central to this idea, wherefore constructivist explanations may prove particularly useful for the explanation of the rapid spread of agencies.
Arbeids- og Inkluderingsdepartmentet. 2007. “Samepolitikken: Ansvarsfordeling og Samordning”. http://odin.dep.no/aid/norsk/tema/same/p30003434/nn.html, accessed 5/02/2007.
Bawn, K. 1995. “Political Control Versus Expertise: Congressional Choices about Administrative Procedures”. The American Political Science Review, Vol. 89, No. 1, pp. 62-73.
Bouckaert, G. and Peters, G. B. 2004. “What is Available and what is Missing in the Study of Quangos?” in Pollitt, C. and Talbot, C. Unbundled Government: A Critical Analysis of the Global Trend to Agencies, Quangos and Contractualisation. London and New York: Routledge, pp. 22-49.
Braithwaite, J. 2000. “The New Regulatory State and the Transformation of Criminology”. British Journal of Criminology. Vol. 40, pp. 222-238.
Caporaso, J. A. 1996. “The European Union and Forms of State: Westphalian, Regulatory or Post-Modern?”. Journal of Common Market Studies. Vol. 34, No. 1, pp. 29-52.
Community Agencies. 2007. http://europa.eu/agencies/community_agencies/index_en.htm, accessed 2007-04-03.
DiMaggio, P. J. and Powell, W. W. 1991. “The Iron Cage Revisited: Institutional Isomorphism and Collective Rationality in Organizational Fields” in DiMaggio, P. J. and Powell, W. W. The New Institutionalism in Organizational Analysis. Chicago: The University of Chicago Ltd. Press, pp. 63-82.
Dolowitz, D. and March, D. 1996. “Who Learns What from Whom? A Review of the Policy Transfer Literature”. Political Studies, Vol. 44, pp. 343-357.
Epstein, D. and O’Halloran, S. 1999. Delegating Powers – A Transaction Cost Politics Approach to Policy Making under Separate Powers. Cambridge: Cambridge University Press.
Gains, F. “Adapting the Agency Concept: Variations within ‘Next Steps’” in Pollitt, C. and Talbot, C. Unbundled Government: A Critical Analysis of the Global Trend to Agencies, Quangos and Contractualisation. London and New York: Routledge, pp. 53-74.
Gilardi, F. 2002. “Policy Credibility and Delegation to Independent Regulatory Agencies: a Comparative Empirical Analysis”. Journal of European Public Policy, Vol. 9, No. 6, pp. 873-893.
Gilardi, F. 2005. “The Institutional Foundations of Regulatory Capitalism: The Diffusion of Independent Regulatory Agencies in Western Europe”. The ANNALS of the American Academy of Political and Social Science. Vol. 598, pp. 84-101.
Hood, C. 2002. “The Risk Game and the Blame Game”. Government and Opposition. Vol. 37, No. 1, pp. 15-37.
Horn, M. J. 1995. The Political Economy of Public Administration: Institutional Choice in the Public Sector. Cambridge: Cambridge University Press.
Huber, J. D. and Shipan, C. R. 2002. Deliberate Discretion? – The Institutional Foundations of Bureaucratic Autonomy. Cambridge: Cambridge University Press.
James, O. 2001. “Business Models and the Transfer of Businesslike Central Government Agencies”. Governance: An International Journal of Policy and Administration, Vol. 14, No. 2, pp. 233-252.
Kohler-Koch, B. and Rittberger, B. 2006. “Review Article: The ‘Governance Turn’ in EU Studies”. Journal of Common Market Studies. Vol. 44. pp. 27-49.
Levi-Faur, D. 2005. “The Global Diffusion of Regulatory Capitalism”. The ANNALS of the American Academy of Political Science. Vol. 598, No. 1, pp. 12-32.
Levy, B. and Spiller, P. T. 1996. “A Framework for Resolving the Regulatory Problem” in Levy, B. and Spiller, P.T. Regulations, Institutions, and Commitment: Comparative Studies of Telecommunications. Cambridge: Cambridge University Press, pp. 1-35.
Majone, G. 1996. Regulating Europe. London: Routledge.
Majone, G. 1997. “From the Positive to the Regulatory State: Causes and Consequences of Change in the Mode of Governance”. Journal of Public Policy. Vol.17, No 2, pp. 139-167.
Majone, G. 2002. “Delegation of Regulatory Powers in a Mixed Polity”. European Law Journal. Vol. 8, No. 3, pp. 319-339.
McCubbins, M. D. and Page, T. 1987. “A Theory of Congressional Delegation” in McCubbins, M. D. and Sullivan, T. (eds.) Congress: Structure and Policy. Cambridge: Cambridge University Press, pp. 409-425.
McNamara, K. R. 1998. The Currency of Ideas – Monetary Politics in the European Union. Ithaca and London: Cornell University Press.
McNamara, K. R. 2002. “Rational Fictions: Central Bank Independence and the Social Logic of Delegation”. West European Politics. Vol. 25, No. 1, pp. 47-76.
Minogue, M. 2002. “Governance-based Analysis of Regulation”. Annals of Public and Cooperative Economics. Vol. 73, No. 4, pp. 649-666.
Moran, M. 2001. “The Rise of the Regulatory State in Britain”. Parliamentary Affairs. Vol. 54, pp. 19-34.
Moran, M. 2002. “Review Article: Understanding the Regulatory State”. British Journal of Political Science. Vol. 32, pp. 391-413.
Műller, W. C. and Wright, V. 1994. “Reshaping the State in Western Europe: The Limits to Retreat”. West European Politics, Vol. 17, No. 3, pp. 1-11.
Pollack, M. A. 2003. The Engines of European Integration: Delegation, Agency, and Agenda Setting in the EU. Oxford: Oxford University Press.
Pollitt, C. et al. 2001. “Agency Fever? Analysis of an International Policy Fashion”. Journal of Comparative Policy Analysis. Vol. 3, pp. 271-290.
Rose, R. 1993. Lesson-Drawing in Public Policy: A Guide to Learning across Space and Time. New York: Seven Bridges Press.
Scharpf, F. 1999. Governing in Europe – Effective and Democratic?. Oxford: Oxford University Press.
Talbot, C. 2004. “The Agency Idea: Sometimes old, sometimes new, sometimes borrowed, sometimes untrue” in Pollitt, C. and Talbot, C. Unbundled Government: A Critical Analysis of the Global Trend to Agencies, Quangos and Contractualisation. London and New York: Routledge, pp. 3-21.
Thatcher, M. 2002(a). “Analysing Regulatory Reform in Europe”. Journal of European Public Policy. Vol. 9, No. 6, pp. 859-872.
Thatcher, M. 2002(b). “Regulation after Delegation: Independent Regulatory Agencies in Europe”. Journal of European Public Policy. Vol. 9, No. 6, pp. 954-972.
 The article is called “The Regulatory State and its Legitimacy Problems” and was published in West European Politics. I am yet to read this article.
 For Horn’s explanation of variation, see section on drivers behind agency creation.
 As examples of members of the Chicago school Thatcher (2002b: 958) mentions Stigler, Peltzman and Becker.
 EU definition of Community Agencies: “A Community agency is a body governed by European public law; it is distinct from the Community Institutions (Council, Parliament, Commission, etc.) and has its own legal personality. It is set up by an act of secondary legislation in order to accomplish a very specific technical, scientific or managerial task, in the framework of the European Union’s ‘first pillar’.” (http://europa.eu/agencies/community_agencies/index_en.htm, accessed 2007-04-03).
 For details of the weighing and coding of the variables in the index, see Gilardi (2002: 880).
 Blame-shifting through agency creation does not always work; sometimes politicians are blamed for events and situations over which they have no control (Hood 2002: 24).