A normative framework for delegation:

Transcription

1 A normative framework for delegation: A case study of the Dutch Authority for Financial Markets (AFM) Student: Jessica Endert, S Supervisor: Dhr. Dr. P. Overeem Second reader: Mevr. Dr. A. Wille 06/06/2016 MASTER THESIS, PUBLIC ADMINISTRATION, track GOVERNING MARKETS REGULATION AND COMPETITION

2 Abstract The research on NMIs primarily focuses on the consequences of, or the reasons for, delegating authority to NMIs, thus presuming the presence of NMIs to be a fixed component of the political system. The decision to delegate authority to NMIs is rarely topic of discussion. This thesis is a response to the identified deficit in the academic literature. Therefore, a normative framework is proposed to guide or evaluate the decision to delegate authority to NMIs. Based on Dahl s theory on democracy, the framework suggests that a set of principles (related to a distinction between empirical and ethical judgments, notions of political equality, and citizen participation) can perform this guiding or evaluative function. By applying these normative principles as minimal requirements, values underpinning traditional democratic institutions can be ensured, while also allowing for governing institutions to be adaptive. The framework provides a useful theoretical benchmark against which both politicians and the general public can evaluate and weigh the risks and benefits of delegating authority to NMIs from a normative perspective.

3 Acknowledgments Writing this master thesis has been a challenging and time consuming endeavor. I am proud of the final result. There are a couple of people who have made a significant contribution to the final product. It is here in the acknowledgement section that I would like to pay a special tribute to these people. First and foremost, I would like to thank my thesis supervisor, Patrick Overeem. Without his extensive knowledge on ethics in public administration, critical and elaborate feedback, and excellent guidance this thesis would not have reached this level, nor would it have been finished. Further acknowledgements go out to those who supported me during the writing of my thesis. Some of you provided me with valuable input and helped me further sharpen my arguments and my focus. Here too it goes that the end result would not have been the same without you. Thank you, to my parents and my sister Simone Endert.

4 Index Chapter 1. Introduction The regulatory state Advocating closer scrutiny of the decision to delegate The institutional design of NMIs A roadmap for this thesis 5 Chapter 2. The omnipresent NMI: concepts and theory Introduction Constructing a normative framework Defining the non-majoritarian institution Commitment problems, information asymmetries, efficiency, and blame avoidance Setting the democratic context Conclusion 15 Chapter 3. Against or in favor of NMIs Introduction A critical perspective on NMIs The rise of the unelected Conclusion 26 Chapter 4. Guiding principles for the decision to delegate authority to NMIs Introduction Separating ethical from empirical judgments Knowledge does not equal acting ethically The costs of delegating authority to NMIs Conclusion 44 Chapter 5. The case of the Dutch Authority for Financial Markets Introduction Applying the normative framework Testing the first principle separating judgments Testing the second principle acting ethically Testing the third principle Institutional reforms and functionalist arguments Conclusion 56 Chapter 6. Conclusion A normative framework for delegation Practical implications of applying the framework Recommendations and future research 61 References 64

5 Chapter 1. Introduction 1.1 The regulatory state Since the beginning of the 1980s, there has been a marked shift towards a regulatory mode of governance, both at the national and supra-national level (Thatcher and Stone Sweet, 2002; Vibert, 2007). Governments of established democracies have transferred far-reaching executive, legislative, and even judicial authority to nonmajoritarian institutions (NMIs) (Daintith, 1989; Ogus, 1995; Osborne and Gaebler, 1993). This distinctive style of governance is characterized by an attempt to steer the behavior of public and private actors, who operate at some distance of the state, in a direction that is valued by society. This is done through methods of promulgation and enforcing systems of rules, which are typically established by institutionally distinct regulators such as non-majoritarian institutions. Keeping central banks and other administrative agencies at arm s length may not seem like a new development, considering that many democracies have a longstanding tradition of insulating the judiciary from the political arena (Brisben, 2005; Greve, 1999). Yet, the thing that has changed is the scope of the authority that governments have delegated to NMIs. Some authors call this phenomenon of shifting governing institutions the rise of the regulatory state. In this regulatory state the role of traditional democratic institutions is changing if not diminishing (Majone, 1994; Majone, 1996; McGowan and Wallace, 1996). The heavy reliance of liberal democratic states on institutions that are exempted from electoral pressures has sparked concern among observers. It leads to the devaluation of the electoral process (Mair, 2005, 2013). But what exactly constitutes a heavy reliance? In the UK there were 550 NMIs operating at the national level in 2009 (Gash et al. 2010: 25), in Ireland the count of administrative bodies operating outside of ministerial control is roughly 300 (Hardiman and Scott, 2010: ), and in the Netherlands an estimated 630 NMIs were counted in 2005 (Jongeneel, 2005: 48-50). These are indications of how ubiquitous NMIs are in the modern political system. Page 1 of 76

6 1.2 The decision to delegate NMIs are often taken for granted and are considered to constitute a fixed component of the political system. This is reflected in the existing research on NMIs, which primarily focuses on reasons for and consequences of delegating authority. Scientists studying political institutions have more to say about the institutional effects of NMIs than about their origins and change. This has led to a functionalist treatment and research agenda for this type of institutions. This is, for instance, reflected in theories that emphasize the rational choices of individual actors that underlie political activity and the efficient nature with which they respond to collective social needs (Pierson, 2000: ). Functionalist arguments take the following form: outcome X exists because it serves the function of Y (Idem: 476). Functionalist thinking is, therefore, a good place to start if one is trying to explain institutional formation. After all, if actors act purposively then the effects of an institution can serve as an explanation for emergence, persistence, and potentially change (Keohane, 1984). However, as good as a starting point functionalist thinking may be, it should not be an end-point for analysis. By relying on functionalist arguments for analyzing the presence of NMIs in the political system the controversial normative aspects of institutional emergence and change are avoided. Rather than dealing with these issues head-on, one starts by examining existing institutions. The assumption is that these institutions exist because they are functional for the actors that instated them, which is generally to solve a type of collective action problem. The functionalist approach to studying NMIs touches on a wide variety of subjects. There is a body of literature, strongly related to principal-agent theory, that tries to answer why governments delegate authority to NMIs by relying on the function these institutions perform. Four common rationales are distinguished: efficiency gains (Dunlop, 2010; Franchino, 2002; Majone, 2001), reducing information asymmetries (Bawn, 1987; 1995; Miller, 2005; Weimer, 2006), resolving credible commitment problems (Bendor, et al., 2001: ; Gilardi, 2002; Helm, et al. 2003), and blame avoidance (Daugbjerg and Swinbank, 2007; Weaver, 1986; Weimer, 2006). Page 2 of 76

7 Another functional account of studying NMIs can be found in research focusing on matters of legitimacy and accountability. Again, here too, existing NMIs are analyzed in an attempt to reconcile the function of these institutions with some of its potential problems. Two common types of concerns can be distinguished in this literature: 1) there is a lack of popular input in the governance of NMIs which creates potential legitimacy problems (Boedeltje and Cornips, 2004; Majone, 1999; Scharpf, 1997; Vibert, 2007), and 2) there is a decrease in the possibility of holding institutions accountable resulting from delegating authority to NMIs (Bianculli, et al., 2015; Majone, 1996; Persson, et al., 1997; Vibert, 2007). Other research on NMIs focuses on the separation of powers and problems stemming from principal-agent relationships (Bawn, 1995; Carolan, 2009; Gilardi, 2002; Weimer, 2006). What should become clear is that the decision to delegate authority to NMIs is rarely analyzed. The normative aspect of institutional emergence, formation and change is often left untouched. This thesis is a response to this deficit in the academic literature. The fact that the decision to delegate government authority to NMIs is now a common one, does not mean that it should not be subject to scrutiny. Therefore, a normative framework is proposed to guide or evaluate the decision to delegate authority to NMIs. The research question central to this thesis is: What normative principles can be used to evaluate and guide the decision of governments to delegate authority to NMIs? The need for this kind of normative framework can be linked to a 2005 rapport, written by the Dutch Foundation for Economic Supervision (SEO). In this rapport the SEO warns about the consequences and tensions stemming from dividing responsibilities over multiple actors, weakening traditional democratic accountability mechanisms, and problems having to do with legitimacy in relation to the functionalist description of NMIs (Baarsma, 2005: 11). The proposed framework will go beyond the usual functionalist accounts of research on NMIs. It will have a strong normative component and rely on concepts such as democracy, political equality, and citizen participation. Hereby, adding to the existing academic literature on NMIs. This normative framework may be relevant to Page 3 of 76

8 legislators, politicians, citizens, and all who are interested in analyzing those institutions that fall outside conventional electoral control. The primary focus is on developing the philosophical arguments that are needed to build the normative framework. In an attempt to show that this is more than a futile philosophical endeavor, the practical relevance of such a framework is demonstrated by performing a case study of the Dutch Authority for Financial Markets (AFM). The case study is a first attempt to demonstrate the practical application of the normative framework. Application of the framework results in answering the question whether or not the government s decision to delegate authority to the AFM is justifiable. 1.3 The institutional design of NMIs Designing the right kind of governing institutions is an important and difficult task. For too long research has tried to keep what is and what ought to be separate, both intellectually and in policy. Empirical research on norms has shown that ought to often becomes is (Finnemore: ). This finding can also be used to analyze the decision of governments to delegate authority to NMIs. It is on the basis of concepts such as political equality, citizen participation, and the distinction between empirical and ethical judgments that one can instigate the long-overdue normative discussion on whether or not a government should decide to delegate authority to NMIs. The governing institutions of today should not solely be analyzed and evaluated on account of what is but also on account of what ought to be. The framework will take the form of a set of principles that can be applied in two instances. First, the main purpose of the principles is to function as a set of normative minimal requirements that need to be met prior to making the decision to delegate authority. Second, the principles can be used to analyze the decision to delegate after it has been made, in which case they take on an evaluative character. Ultimately, the normative framework facilitates the debate on the role NMIs should play in society as governing institutions. This debate should, at least to some degree, be about the ability to unify the presence of NMIs in the political system with traditional democratic values. This may mean coming to terms with the conclusion Page 4 of 76

9 that NMIs are a necessity for the functioning of modern democratic states, and that traditional democratic values, at least partially, will be eroded from within. 1.4 A roadmap for this thesis Chapter 2 outlines the relevant concepts and theories. It provides an account of what constitutes a normative framework, gives a definition of delegation and NMIs, elaborates on the four rationales for delegation, and also sets the democratic context in which the rest of the arguments are to be regarded. Chapter 3 begins with a short oversight of the existing academic literature that is critical of NMIs as governing institutions. This is supplemented with a detailed account of the arguments in favor of delegating authority to NMIs as put forth by Vibert, in his book The Rise of the Unelected (2007). This enables the reader to assess how the normative framework differs from, and contributes to, the existing academic literature. Chapter 4 presents the arguments in support of the proposed set of normative guiding principles. The following three conditions need to be met for a government s decision to delegate authority to an NMI to be justified: 1) it is possible to distinguish between the empirical and ethical judgments related to the task that is being delegated, 2) for an NMI s task performance passing ethical judgments is minimized as much as possible, and 3) the benefits of delegation outweigh the democratic costs of delegation. Chapter 5 demonstrates the applicability of the normative framework by performing a case study on the decision to delegate authority to the Dutch Authority for Financial Markets (AFM). A conclusion is reached on whether or not the government s decision to delegate public authority to the AFM was justified. Chapter 6 contains the final conclusion of this thesis, what the consequences are for governing institutions, the limitations of this study, and suggestions for further research. Page 5 of 76

10 Chapter 2. The omnipresent NMI: concepts and theory 2.1 Introduction Those who engage in the debate on NMIs and the challenges brought about by this type of governing institutions often use different terms in their analysis. This is due to the fact that NMIs are a subject of study for many academic disciplines, such as, Public Administration (Coen and Thatcher, 2005; Lindgren and Persson, 2010; Weaver, 1986), Law (Lawson, 1994; Strauss, 1984), Political Science (Hall and Taylor, 1996; Majone, 1999; Thatcher and Sweet; 2002), and Economics (Hemphill, 2004; Kovacic and Shapiro, 2000; North, 1993). To avoid misunderstandings some definitional matters, concepts, and theoretical issues need to be settled prior to developing the normative framework. 2.2 Constructing a normative framework The publicness of public administration This thesis aims to preserve the publicness of public administration by constructing a normative framework that can be used to guide a government s decision to delegate. It does not fit into the tradition of public administration as a science. While scientific knowledge definitely has a place in the field of public administration the political should not be occluded. Normative theory points public officials, legislators and anyone interested in NMIs as governing institutions in the direction of usable knowledge. This type of knowledge is neither pure science nor conventional wisdom. Public administration as a field should do more than describe and explain the institutions it studies, it should represent the constitutively contestable dimensions of these institutions. Normative theory serves this purpose and is therefore a crucial source for the publicness in the field of public administration (Stivers, 2000:10). Following Arendt, the scientific approach to public administration cannot help us understand the meaning-making character of intersubjective realities. These realities embody public life, where knowledge is gained and decision are made without relying on scientific benchmarks (1977). What makes public administration public, is that public life itself consists of irreducible differences on which basis Page 6 of 76

11 people need to find a way to live together. A scientific approach to public administration has little value if one is trying to answer the bigger questions that are fundamentally open and irresolvable. If there were scientific answers to these questions public life would no longer consist of these irreducible differences and it would cease to exist. Public administration should, at least for the sake of discussion, restore the tension between right answers and normative questions. While most scholars acknowledge the political or public dimension of their field of study they often do so in a descriptive, functionalist or instrumental manner (Stivers, 2000: 14-15). This is, for instance, reflected in the functionalist approach to studying NMIs. The contestable question whether or not a government should delegate authority to an NMI in the first place is rarely acknowledged. Instead, the proper role of NMIs in the political system is addressed in terms of the utility these institutions have for the political actors that instate them. The connection between empirical science and political values is left unmentioned rather than questioned. By constructing a normative framework this thesis does connect empirical science to political values A normative framework consisting of principles The ambition to construct a normative framework to help guide or evaluate an individual s, organization s, or government s behavior is not unique to this thesis. It is a practice that is used in other scientific disciplines and by other scientists. A common feature in this work is that a set of principles or conditions is formulated based on normative arguments that are supported by existing empirical research. If the requirements of the principles are lived up to one can arrive at a conclusion that something is right. Examples of normative frameworks following this structure include: Wilson s work on the role justice should play in public health ethics and policy (2009: ), the circumstances under which mixed bundling is most likely to be effective in pricing services, as presented by Guiltinan (1987: 75), the analysis presented by Lopez, that can be used to reason why certain norms must be adopted, and why an adopted norm must be complied with (2006: 247), Black s work on the conditions that are needed for a regulator to give in to legitimacy claims (2008: 156), and Underwood s analysis of how the principles of communicative competence can be Page 7 of 76

12 used as a benchmark to evaluate the level of distortion in product labeling and packaging (1998: 211). Application of the principles presented in these works varies in different respects, such as, strictness, specificity, additional conditional requirements, universality etc. In following the example of these authors the normative framework of this thesis will take the form of a proposed set of normative principles that can function as guiding principles, or minimal requirements, when analyzing a government s decision to delegate authority to NMIs. 2.3 Defining the non-majoritarian institution The tasks and duties performed by NMIs are incredibly diverse and there is more than one way to refer to this type of governing institution. Some researchers speak of quasi-nongovernmental organizations (quango s), regulatory agencies, nongovernmental public bodies, unelected bodies, or authoritative agencies (Pollitt and Talbot, 2004). In this thesis the term non-majoritarian institutions is used because this reflects an important aspect of this type of institutions. NMIs, as opposed to traditional democratic institutions, do not rely on the majority of the electorate for their legitimacy. They are not democratically accountable, nor majoritarian, and, although less relevant for the discussion at hand, also not minoritarian. NMIs derive their prominent place in the political system and their legitimacy from other sources, such as their independence, objectivity, ability to formulate empirical judgments, and use of scientific methods. This is discussed more extensively in upcoming paragraphs. Following the definition of Thatcher and Stone Sweet, an NMI meets the following two criteria: 1) NMIs possess and exercise a degree of specialized public authority, which is distinct from other institutions, and 2) they are not directly elected by the people, nor directly managed by elected officials. This definition of NMIs excludes those powers that are organized within the bureaucracy, for they do fall under the direct control of ministers and elected officials. This definition does include institutions that are linked to a ministry in formal ways, as long as the institution is not a department or administrative office of a larger bureaucratic entity (Thatcher and Stone Sweet, 2002: 2-3). Page 8 of 76

13 Delegation is an authoritative decision, which is formalized through public law, where (policy making) authority is transferred away from existing representative institutions to an NMI (Idem). Delegation to NMIs is often a gradual process in which institutions are constructed, revised, and adapted, over extended periods of time. However, in democratic states there is always a moment in time when the decision to delegate is formalized though public law. This moment is considered the deciding moment at which point the principles of the normative framework can offer guidance as to whether or not the decision to delegate is justifiable. It is possible to further distinguish between five broad categories of nonmajoritarian institutions. There are: 1) service providers, 2) risk assessors, 3) boundary watchers, 4) inquisitors, and 5) whistle-blowers (Vibert, 2007: 20). Each of these categories can be subdivided into more narrowly defined categories. For instance in the category of boundary watchers there are NMIs that monitor competition, monitor the spending of public funds, are in charge of administrative and policy control, monitor the quality and stability of an industry, and monitor developments (Baarsma, 2005: 1). An NMI can fulfill more than one function, both with regards to the broad and more narrowly defined categories. Vibert has defined four additional features that the wide variety of NMIs has in common and that set them apart from other kinds of (political) institutions. Firstly, NMIs reside in a technical world that is highly dependent on expert knowledge to which they also make an active contribution. Secondly, the sources of information and knowledge they use in their task performance almost always lie outside of government influence. Thirdly, NMIs form separate communities based on this specialized knowledge and information. Fourthly, NMIs have their own disciplines (2007: 30-33). NMIs are defined following the two criteria as formulated by Thatcher and Stone Sweet, and the four features as described by Vibert. This leads to a more narrow description of what constitutes an NMI, while also clarifying some of the contextual properties of the functioning of these institutions. An NMI is an institution that is part of a separate community based on specialized knowledge and information, it holds and exercises a degree of specialized public authority which is distinct from other institutions, it is not directly elected by the people, nor directly managed by elected Page 9 of 76

14 officials, it operates on the basis of expert knowledge to which it actively contributes, and the sources of its knowledge lie outside government control. 2.4 Commitment problems, information asymmetries, efficiency, and blame avoidance Principal-Agent theory and delegation Giving a definition of NMIs, identifying their common properties, and establishing their pervasiveness, does not in itself provide a reason for why they have become such important governing institutions. In order to understand these institutions one needs to know something about why they are crafted and sustained. This also means having an, albeit simple, understanding of the way they impact society in different settings (Ostrom, 2005: 3). Principal-agent theory, in combination with theories on delegation, can provide a useful tool for gaining such understanding. Principals are the elected officials that delegate authority to agents. Agents are the NMIs who govern through exercising the delegated authority. Delegation is functional for principals because the expected benefits outweigh the costs. This has to be the case, for principals are initially in control, in the sense that they decide whether or not to constitute a new agent. If the expected benefits do not outweigh the costs a principal would not delegate. Furthermore, they also decide on the institutional design and an agent s remit (Thatcher and Stone Sweet, 3-4). The upcoming paragraph combines the arguments of principal-agent theory with the arguments of theories on delegation. Together they provide an explanation for the emergence and perseverance of NMIs as well as for their growing importance as governing institutions The four most common rationales for delegation In the academic literature on delegation of public authority to NMIs four common rationales for delegation can be distinguished: Firstly, delegation is used to solve commitment problems. Agents are expected to enhance the credibility of the promises made by principals. These promises are either made to the electorate or to other principals. Commitment problems often arise given underlying collective action problems (Bendor, et al., 2001: ; Gilardi, Page 10 of 76

15 2002; Helm, et al. 2003). NMIs are established to ensure that the achievements and ideas of the enacting coalition survive, seeing how there is no guarantee that those who are in power today will not be subject to authoritative direction from others tomorrow. Commitment problems can be a consequence of regular electoral competition or of a shift in the preferences of the principal due to majority-rule instability (Elgie, 2005: 534). An NMI needs a degree of independence to credibly solve a commitment problem. This independence is required because it makes it difficult for opponents, or other principals, to take control over these institutions. This allows the institution and the policies it enacts to persist. Ultimately, this results in institutions that are insulated from public authority in general. This means that they are also independent from the enacting coalition responsible for its design (Moe, 1995: ). The creation of an NMI can thus be seen as an exercise in institutional design that tries to resolve current problems, and predetermine future possible actions in ways that benefit the creator and its constituents (Horn, 1995: 43). Secondly, through delegation it is possible to overcome information asymmetries in technical areas of governance. Agents are expected to develop and use their knowledge and expertise to either construct the appropriate policy option or help principals in this process. The knowledge and expertise made accessible by NMIs would not be available to the principals if they did not decide to delegate. Though it is possible for government bureaucracies to gather this knowledge themselves, this is less likely due to time and money restrictions (Bawn, 1995; Demski and Sappington, 1987; Miller, 2005; Weimer, 2006). Globalization has made the need for reducing information asymmetries in policy making more important. Due to globalization there is a marked shift in developed nations from manufacturing industries to service-oriented economies. People consider the government to be part of this service economy and judged the government s performance on the same standards as they did the private sector (Vibert, 2007: 34-36). In order to live up to these standards and make appropriate decisions, especially in technically complex areas of policy making, the government needs to supply itself with expert knowledge. Public officials lack the resources or incentives to develop this knowledge themselves. Therefore, these areas of policy Page 11 of 76

16 making require the attention of independent expert institutions rather than political ones (Elgie, 2005: ). Thirdly, delegation enhances the efficiency and responsiveness of rule-making and the policy process. According to classic economic theory, efficiency gains can be realized through the division of labor among workers and the process of specialization (Smith, 1965: 11). The productivity of specialists depends for a great deal on how much knowledge they have. The dependence of specialization on the available knowledge, ties the division of labor to economic progress. For this type of progress depends on growth in both human capital and technology (Becker and Murphy, 1994: ). Agents are expected to concern themselves with relatively specific problems and issues that arise, while principals can then determine and update public policy in more general terms (Dunlop, 2010; Franchino, 2002; Majone, 2001). Delegation to NMIs results in efficiency gains based on the principle of the division of labor. They are institutions that are specialized in a specific policy area and they make use of, and contribute to, the available knowledge in the policy process. Finally, delegation is a tool that allows principals to avoid blame for unpopular policies. NMIs do not directly rely on the electorate for their support or legitimacy. They are expected to be able to maximize policy goals when these go against the preferences of important societal groups. This allows principals to avoid, or rather shift, the blame (Daugbjerg and Swinbank, 2007; Weaver, 1986; Weimer, 2006). NMIs rely on their superior performance, relative to the results elected officials and bureaucracies are able to bring about, to justify their existence. This is what they derive their legitimacy from (Thatcher and Stone Sweet, 2001: 18-19). The same reasoning extends to the politics of blame avoidance. In the case of policies not being popular among the electorate, NMIs can bring about the additional benefits of efficiency gains through the better use of knowledge (Idem). The legitimacy of NMIs becomes a problem when agents are corrupt or fail to realize the expected efficiency gains (Majone, 2000). If principals wish to reap the benefits of delegation this requires granting the agent some discretion. A principal can do this by sharing part of their authority to govern. This discretion might result in agency loss, which occurs when the agent does not generate the outcomes preferred by the principals who have delegated authority (Thatcher and Stone Sweet, 2002: 4-5). Page 12 of 76

17 This exploration of principal-agent theory has provided the four most common rationales for delegation. These rationales for delegating authority to NMIs are all in some way related to relying on an NMI s scientific knowledge, objectivity, and independence. This does not mean that these four rationales are the only rationales for engaging in acts of delegation. When authority is delegated to NMIs it is possible that other rationales play a part. These four rationales are, however, identified as the most common ones in the literature on delegation. 2.5 Setting the democratic context In order to construct the normative framework it is important to set the democratic context. NMIs as governing institutions operate within this context, and in liberal democratic states they have to share the political arena with traditional democratic institutions. What democracy is, what institutions it requires, and what purpose these institutions serve is contested. There are a number of democratic theories that each emphasize different aspects of democracy. According to liberal democratic theory the most important aspect of democracy is that the state respects the individual liberties of its citizens through the rule of law (Diamond, 1999:3). A procedural account of democracy is put forth by Schumpeter, he states that democracy is an institutional arrangement for political decision making, where political leaders compete for the votes of the electorate (2005: ). Adherents of deliberative and participatory democratic theory stipulate how democracy is the only way to organize political power and political decision making in such a way that it leads to democratically legitimate decisions (Estlund, 2007; Saward, 2000). They argue that democracy is the only political system in which policies emanate from the population to which they apply. The use of political power does need to be conducive to the general interest of the population (Saward, 2000: 68). There is no model of democracy that is universally acceptable. However, following Dahl, the importance of political equality as a defining feature of democracy is reflected in almost all democratic theories and democratic societies in the world (Dahl, 2000: 62-80). Therefore, this thesis uses an account of democracy that emphasizes democracy s normative and practical dimensions by relying on the Page 13 of 76

18 concept of political equality. Such an account of democracy reflects the ambition of this thesis to construct a normative framework that is of practical use. Dahl s democratic theory best fits this description. Democracy is both an ideal and an actual set of practices and institutions. Ideal democracy is a political system designed in such a way where its members are willing to treat each other, for political purposes, as political equals. Members of such a system may still, and probably will, regard one another as unequal in other respects. Despite this being the case political equality is the essence of democracy and should be strived for (Dahl, 2000: 38). Though ideal democracy is perhaps unachievable, it is nonetheless highly relevant when it comes to judging political systems, creating strategies of democratization, and designing the appropriate institutions for democratic government (Dahl, 1999: 10-13). In striving for ideal democracy one is often left with, what Dahl calls, polyarchy. In polyarchical democracy power is vested in multiple people rather than just one person or the whole population, it is therefore neither a dictatorship nor a true democracy. The fundamental principle underlying polyarchical democracy is that the government is responsive to the preferences of its citizens, and that they are considered political equals with unimpaired opportunities (Dahl, 1971). Political equality can, however, mean different things. According to Beitz, it refers to the notion that all citizens preferences are weighed equally in the decision making process (1989: 3-4). It can also be equated with citizens having equal procedural opportunities to influence political decision making (Knight and Johnson, 1994: 280; Phillips, 1995: 154). In this thesis political equality is considered to comprise both of these aspects, which also corresponds with the concept of political equality that features prominently in Dahl s democratic theory. There are five principles on which democratic government is based. Each of these principles is related to the ideal account of democracy and contributes to realizing political equality. These principles are: 1) effective participation, 2) equality of votes, 3) informed insights, 4) popular control over the political agenda and 5) inclusion (Dahl, 2000: 37-40). In turn, six (political) institutions are required to give substance to these principles and political equality. In order for democracy to function properly one needs: 1) elected representatives, 2) free, fair, and recurring Page 14 of 76

19 elections, 3) freedom of speech, 4) the right to alternative information, 5) freedom of association, and 6) inclusive citizenship (Idem). 2.6 Conclusion In this chapter a potential conflict has been uncovered. This conflict exists between the values associated with governance as performed by NMIs and those associated with governance as performed by traditional democratic institutions. The potential conflict resides in the following contradictory account of what constitute the proper values for organizing the political system. NMIs predominantly rely on academic knowledge, scientific objectivity, and their independence for improving the quality of public policy and political decision making. The notion is that NMIs are capable of separating fact from value in the performance of their tasks and that they are legitimate and accountable institutions particularly because of their independence and objectivity. In sum, knowledge takes primacy which often implies political inequality elsewhere in the political system. Traditional democratic institutions rely for the quality of decision making on the process of informed communication and deliberation. Democracy has political equality as its highest ideal and democratic government embodies the commitment to this ideal. Democratic decision making is the only way to reach political decisions that carry any weight on account of legitimacy and accountability. In sum, political equality takes primacy. NMIs do not necessarily pose a threat to the values associated with democracy. This is partly dependent on whether one adheres to an ideal account of democracy, or whether one accepts that democratic government is, at best, the closest approximation of certain ideals. Arguments have been provided, however, why NMIs might threaten traditional democratic institutions and values. This warrants closer scrutiny of the decision to delegate authority to NMIs. Page 15 of 76

20 Chapter 3. Against or in favor of NMIs 3.1 Introduction This chapter presents another aspect of the academic debate on delegating authority to NMIs. It first discusses four criticisms on NMIs as governing institutions and it ends with one defense of the practice as formulated by Vibert. In combination with the four functionalist rationales for delegation, as discussed in chapter 2, it provides an overview of what has been written on the subject. This chapter also clarifies how the proposed normative framework constitutes an addition to the existing literature on NMIs. For the arguments of this chapter are substantially different from the ones in support of the normative framework. The difference is that the former are based on functionalist lines of reasoning, whereas the latter relies on making normative claims. 3.2 A critical perspective on NMIs Legitimacy problems The first criticism focuses on NMIs lacking democratic legitimacy. Scharpf explains how democratic legitimacy is a two-dimensional concept consisting of both the inputs and the outputs of a political system. The input side refers to the mechanisms and procedures required to be able to link the preferences of citizens to political decisions (1997: 18-21). In liberal democratic states these mechanisms and procedures take the shape of representative institutions, in which the decision maker can be held accountable through elections. The output aspect of legitimacy is related to the ability of these democratic procedures to generate responsive outcomes. A democracy would fail if it is not capable of achieving the goals that the population collectively cares about (Scharpf, 1997: 18-21). NMIs tend to score low on input legitimacy and high on output legitimacy. Several authors signal that there is a trade-off between input and output legitimacy (Boedeltje and Cornips, 2004; Lindgren and Persson, 2010; Majone, 1999). When input legitimacy of NMIs would be increased through a process of making them more Page 16 of 76

21 democratic this creates the danger of decreasing outcome effectiveness and undermining the overall performance of the system (Elsig, 2007: 88; Lindgren and Persson, 2010: 254). Boedeltje and Cornips explain this trade-off by pointing to the different requirements that need to be met in order to achieve input and output legitimacy. They state that input legitimacy relies on the concept of fairness. Decision making and deliberative procedures should provide equal opportunities for all relevant stakeholders to participate and influence the outcome of such a process. Output legitimacy relies on the concept of competence. They use Habermas concept of the ideal speech situation, which has been highly influential in deliberative democratic theory, to define competence as the ability to use language to create understanding and agreement. Competence is needed to translate political procedures into successful outcomes (2004: 7-9). The two criteria of fairness and competence are at odds with each other. Higher levels of fairness can only be reached at the expense of competence and vice versa. This is why, when NMIs score high on output legitimacy they tend to score low on input legitimacy. The inability of elected officials to transfer both public authority and their own input legitimacy to NMIs is another reason why NMIs score low on input legitimacy. Transferring public authority from a principal to an agent will always be accompanied by costs, irrespective of the potential benefits that it offsets. The loss of input legitimacy can therefore become one of the most important agency costs of deciding to delegate authority to NMIs (Majone, 1999: 7). This becomes even more problematic when delegation diminishes the importance of traditional democratic institutions, which in turn might lead to a loss in the associated value of political equality. When losing input legitimacy coincides with diminishing political equality, this loss of equality becomes an important aspect of the agency costs of delegation Accountability problems The second line of criticism is that NMIs are not sufficiently accountable. This line of criticism is often discussed in relation to the first. Accountability and legitimacy, though strongly related concepts, are different. Legitimacy is about how powers are justified, and accountability about how they are exercised and accounted for. There Page 17 of 76

22 are three components to accountability: 1) one needs to be answerable for one s actions, 2) powers can only be exercised within defined limits, and 3) sanctioning needs to be possible somewhere in the system in case powers are misused (Vibert, 2007: ). In liberal democracies, public authority resides in institutions that can be held accountable to the electorate. NMIs as governing institutions do not fit this description. The individuals who are in charge of NMIs are appointed, not elected, and yet they yield significant public authority (Majone, 1996: 4-5). Those who look at NMIs with suspicion often do so because they worry about democratically controlling the exercise of this public authority (Follesdal and Hix, 2006; Person, et al., 1997; Scott, 2000). Jordana provides three explanations why these worries are valid reasons for concern. Firstly, principal-agent literature has pointed to the potential increased opportunity of NMIs to shirk, misuse, and abuse political authority relative to ministerial departments because of their independence. Despite this being the case it is also emphasized that there is a de facto incentive for NMIs to remain accountable. For principals may decide that the suffered agency losses are too significant. In these cases the mandate of the agent could be modified, its operating procedures changed, or it could even be dismantled. It is the threat of political intervention that affects an agent s behavior and makes it accountable (Jordana, et al., 2015: 13-14). Secondly, there is debate on the extent to which NMIs are accountable beyond the national state and its territorial boundaries. The foundation of the argument is that NMIs become an integral part of network governance, which relies on new structures of cooperation between national, regional, and international agencies, in which traditional democratic accountability is less important. Instead, NMIs derive their accountability from interacting with broader audiences that are part of these governance networks. This does require understanding the ways in which NMIs become part of these networks and the accountability mechanisms through which they interact with their wider audiences (Ibidem: 14-15). Finally, there is discussion on how NMIs can be held accountable through alternative means. Examples include inter-institutional accountability, accountability towards consumers, and accountability towards citizens. Responsiveness to consumers and clients might in this case supplement, or partially substitute, Page 18 of 76

23 democratic accountability. This form of accountability ties in with a more reduced version of citizenship. In this case citizenship is tied not to a political conception but to an economic consumer-like one (Ibidem: 13-14). Accountability concerns should always be considered in a wider context that also includes other institutional checks and balances Problems with the separation of powers The third line of criticism discusses the wider institutional context in which NMIs are set. It focuses on problems that NMIs pose for the separation of powers and the principle of checks and balances. The separation of powers discussion analyses how different governmental structures result in, or manage to prevent, an absolutist government (Strauss, 1984: 622). What is common in almost all of the literature is that it relies on a three-way classification of state power in a legislative, executive and judicial branch. An approximation of the basic principle to separate the branches of government, through a system of checks and balances, is still found in liberal democratic states today. Differences arise not only over the validity of this initial threefold classification of the governing functions of the state, but also over the institutional allocation of these powers (Carolan, 2009: 18-24). The legislative-executive-judicial division of state power is an important element that prevents absolutist government. The sheer size of modern democratic states, significant regulatory responsibilities, and growing complexities due to globalization and technological developments, places some serious constraints on this eighteenth-century model (Strauss, 1984: ). NMIs pose a threat to the separation of power for two reasons. Firstly, an NMI can unify more than one of the three distinct separate functions of government within the same institution, by simultaneously engaging in legislative, executive, and judicial activities (Sunstein, 1987: ). And secondly, it is not possible to unify both the variety and dispersion of NMIs within the tripartite model, where government power is neatly divided into three radically separate pillars. This has led to a situation in which NMIs have been uncomfortably and artificially forced into the legislativeexecutive-judicial division, despite them functioning as an additional fourth branch of government (Carolan, 2009: 44-47). Under normal circumstances each branch has a role in holding the others in place, through a process of checks and balances. Page 19 of 76

24 NMIs have not yet successfully been integrated in this existing system of checks and balances (Vibert, 2007: 15). Some observers disagree with qualifying NMIs as a fourth branch of government. For this would mean that NMIs are equal to and operate on the same level as the other three branches, which is not the case. The difference can be found in the fact that NMIs, as opposed to the legislative, executive and judicial branch, do not derive their legitimacy directly from the constitution. The decisions made by NMIs are in principle subject to review, where the other three branches have definitive decision making power (Overeem, 2012: ; Vile, 1998) Majone points out how the two signaled problems are reflected in the structure of the European Union, a political institution that is frequently used as an object of study in research on NMIs. One of the prominent features of this institution is the impossibility of attributing legislative, executive, or judicial authority to specific institutions. For instance, the European Union has a legislative process instead of a legislature, in which the Council, Parliament, and Commission, all play a different part. Similarly it is not possible to identify an executive power, since the powers are vested in the Council, the Commission, and Member States, depending on the situation (2002: ). This is not to say that the European Union is the only example of an institution that is confronted with these problems, it often applies to states too, for instance with regards to bicameral systems. Majone continues that the complexity of European Community policy making should be reflected in assigning a greater autonomous role to NMIs, which essentially function as a regulatory state. His notion of the regulatory state is in some ways analogous to the idea of a fourth branch of government. This fourth branch denotes the distinctiveness and unity of the underlying regulatory process, as well as its importance in terms of modern governance. This regulatory process is not simply an extension of the executive function. It represents the scope of rights normally exercisable by government as a whole, though limited by the boundaries set by the principal that has statutory authority (2002: ) Principal-agent relationship problems A final line of criticism finds its origins in principal agent-agent theory. The basis of this theory states that there is an information asymmetry between an ignorant Page 20 of 76