Why Do Governments Delegate Authority to Quasi-Autonomous Agencies? The Case of Independent Administrative Authorities in France


Título: Why Do Governments Delegate Authority to Quasi-Autonomous Agencies? The Case of Independent Administrative Authorities in France

Autor: ROBERT ELGIE

Año: 2004

Fuente: European Policy Journal

Descarga

Why Do Governments Delegate Authority to Quasi-Autonomous Agencies?

The Case of Independent Administrative Authorities in France

ROBERT ELGIE, Dublin City University

In recent years, there has been a considerable degree of delegation from governments to quasi-autonomous agencies. Various reasons have been put forward to explain why governments decide to delegate authority in this way (Thatcher 2002, 129-139). Some reasons are based on a transactions cost approach, such as credible commitments. Other reasons are more contextual. For instance, governments may be responding to a process of cross-national policy transfer. In the literature on delegation some hypotheses have already been tested. Specifically, Gilardi (2002) has found evidence to suggest that governments create agencies to credibly commit to particular policy choices. However, other hypotheses, particularly ones based on contextual explanations, have proved much more difficult to operationalise. This article aims to help fill this gap. It does so by focusing on the creation of Independent Administrative Authorities (Autorités administratives indépendantes – AAIs) in France. We examine the reasons for their creation. Why have successive governments created so many AAIs in the last couple of decades? Does the qualitative evidence in this particular case corroborate the quantitative studies that have been undertaken elsewhere? What does the French example tell us about the more general literature on delegation?

Explaining the Delegation of Authority to Quasi-Autonomous Agencies

The origins of the political science work on delegation lie in the study of the relationship between Congress and agencies in the United States (see, for example, McCubbins, Noll, and Weingast 1989). Increasingly, though, the study of delegation has incorporated a comparative dimension. For example, Mark Thatcher and Alec Stone Sweet (2002) have focused on delegation by governments to non-majoritarian institutions. They define delegation as “an authoritative decision, formalised as a matter of public law, that (a) transfers policy making authority away from established, representative organs (those that are directly elected, or are managed directly by elected politicians), to (b) a non-majoritarian institution, whether public or private” (ibid., 3). They define non-majoritarian institutions as “governmental entities that (a) possess and exercise some grant of specialised public authority, separate from that of other institutions, but (b) are neither directly elected by the people, nor directly managed by elected officials” (ibid., 2). Examples of such institutions include courts, central banks and independent regulatory authorities (IRAs).

Much of the comparative literature on delegation has focused on the reasons why governments delegate authority to non-majoritarian institutions in the first place (e.g., Huber and Shipan 2002). What are the reasons why political actors grant discretion to quasi-autonomous agencies?

Here, the literature to date has been dominated by principal-agent theory, which is based on a transaction costs approach. The logic of this approach is entirely functional. As Mark Thatcher (2002, 131) states: “IRAs performed functions that were useful for elected officials….” More specifically, Mark Pollack (2003, 21) states: “institutions serve to lower transaction costs and facilitate mutually advantageous cooperation among rational egoists….”

Pollack (ibid.) identifies two main ways in which delegation may reduce transaction costs. It can help to reduce the costs associated with incomplete information. In other words, in areas of high policy complexity governments can benefit from delegating authority to technical experts. It can also help governments to credibly commit to various policy decisions. Governments have an inherent interest to renege on their promises in order to maximise their short-term self-interest, usually re-election. However, if governments delegate decision-making authority to a non-majoritarian institution, then they show that they can be trusted not to intervene in the decision-making process and policy-making may be more optimal.

Thatcher and Stone Sweet (2002, 4) identify two further ways in which delegation to non-majoritarian institutions may reduce transaction costs. Similar to the policy complexity argument, delegation can help to increase the efficiency of rule making. It can also help governments to shift the blame for unpopular policies from themselves to the representatives of the institution in question (Weaver 1986). Finally, Terry Moe (1995, 124) has argued that governments may choose to delegate power to an independent agency so as to prevent their political opponents from controlling the policy-making process when they take power. This argument assumes, rightly, that uncertainty is inherent in democratic political life. As a result, governments know that at some stage they will lose power and so they try to institutionalise their own preferences in the design of independent agencies. Moe states (1995, 136): “The group’s task in the current period … is to build agencies that are difficult for its opponents to gain control over later … this often means building agencies that are insulated from public authority in general­­—and thus from formal control by the group itself”.

In the comparative literature on delegation, some of these hypotheses have already been tested. For example, Gilardi (2002) has tested the credible-commitment argument and has confirmed that governments do indeed seem to delegate powers to independent regulatory agencies in order to enhance the credibility of their policies. In other work, we have argued that the decision to delegate is a function of the desire to make a credible commitment and factors relating to policy complexity (Elgie and McMenamin 2005). Thus, evidence is starting to accumulate suggesting that a major motivation behind governments’ decisions to create quasi-independent agencies is to establish a credible commitment to policy outcomes.

Even so, much work still needs to be done. To date, hypotheses have usually been tested on the basis of quantitative proxies and multiple regression analyses. This work is valuable. However, some variables are more tractable than others in this regard. For instance, it is notoriously difficult to test the blame-shifting hypothesis in these terms. Moreover, in his work Thatcher (2002) has been keen to move beyond such accounts. He argues “there is no automatic link between functional advantages of delegation and the creation of IRAs. Instead, the choice of delegation as a response must be explained rather than assumed. It must be analysed in its context as well as a response to pressures and functional advantages for governments and legislators” (ibid., 136).

In this regard, Thatcher identifies four other explanations (ibid., 136-139). Firstly, the decision to delegate may be the result of a more general process of institutional isomorphism, or cross-national policy transfer. This process may include pressure from the European Union to harmonise regulatory structures (Majone 1996). Secondly, the decision to delegate may be mediated through long-standing state traditions. In some countries delegation may be a normal part of the political process. In others, it may mark a radical break with political practice. Thirdly, the decision to delegate may be the result of political leadership, or political entrepreneurship. A particular leader may wish to pursue delegation as part a personal or party political agenda. Finally, the decision to delegate may be part of a wider process of state reform. In some countries, there has been a general introduction of new public management reforms, of which the creation of agencies may be a constituent part. In other countries, though, the degree of reform has been much less pronounced.

Overall, combining the various explanations, this leaves us with eight competing hypotheses. (See Table 1).

[ TABLE 1 ABOUT HERE ]

Case Selection and Methodology

In order to test the hypotheses identified in the previous section, we have chosen to examine the reasons for the creation of Independent Administrative Authorities (AAIs) in France.

In 2001, France’s highest court of administrative law, the Conseil d’État, defined AAIs as institutions that ‘act on behalf of the State without being subordinate to the Government and that, in order to carry out their tasks properly, benefit from guarantees which allow them to act with complete autonomy, such that their actions may not be influenced or sanctioned except by the courts. In order to fulfil this mission, they have varied sets of powers which, in some cases, give them the power of regulation, individual authorisation, control, injunction, sanction and, indeed, even appointment, but which, in other cases, is merely one of influence, even if this power is couched in fairly formal terms so that it gives them a genuine moral authority’ (Conseil d’État 2001, 257).

The Conseil d’État identified 34 AAIs or AAI-like institutions. They include standard regulatory authorities, such as the Commission des Opérations de Bourse (Stock Exchange Commission) and the Autorité de régulation des télécommunications (Telecommunications Regulatory Authority). They also include more diverse regulatory-like institutions, such as the Autorité de contrôle des nuisances sonores aéroportuaires (Authority for the Control of Airport Noise) and the Commission nationale de contrôle de la campagne électorale relative à l’élection du Président de la République (National Commission for the Control of the Presidential Election Campaign). Even though recent changes mean that the Conseil d’État’s list of AAIs is no longer definitive, the 2001 study is still the text of reference in this area. Therefore, this paper confines itself to the creation of the 34 institutions identified in Conseil d’État’s report.

There are various reasons as to why AAIs constitute a particularly appropriate set of institutions with which to test the hypotheses identified in the previous section. Firstly, they are consistent with Thatcher and Stone Sweet’s (2002, 3) definition of non-majoritarian institutions. Secondly, they comprise a relatively large and heterogeneous set of non-majoritarian institutions. In previous studies, various authors have sometimes chosen only to study a fairly small number of classic IRAs. In so doing, though, they may have biased their findings in favour of certain hypotheses. Specifically, an IRA-based study alone is likely to encourage support for a credible-commitment argument as, by definition, IRAs are often associated with the need for independent regulation of market-sensitive areas. So, by choosing to study a relatively large and heterogeneous set of non-majoritarian institutions, we reduce the likelihood that particular hypotheses will be favoured. Thirdly, by relying on a relatively large and heterogeneous set of non-majoritarian institutions defined by the Conseil d’État, we avoid any accusation of self-inflicted selection bias. In short, we have to work with a set of institutions over which we have no choice, rather than selecting our own.

Having chosen a set of institutions and justified the choice, we propose to test the eight hypotheses in two ways. Firstly, we wish to study the reasons for the creation of AAIs generally. Given we believe that there are advantages to studying a relatively large and heterogeneous set of institutions, it is important to explore all 34 of them. Therefore, in the next section we will take an overview of the reasons for the creation of AAIs in France and test the hypotheses on the basis of general information about the decision to delegate authority to them. We ask four questions. When were AAIs created? Which policy areas do they cover? How are they structured? What general reasons have been given as to their establishment? Secondly, while there are advantages to studying the whole set of AAIs, in so doing we may run the risk of glossing over important and more detailed evidence. Therefore, we will take two institutions and examine the reasons behind their creation in some detail. (The choice of institutions will be justified when the time comes.) We firmly believe that there are distinct advantages to an approach that combines the equivalent of a relatively large-n study and a small-n study in one paper. In short, we believe that our findings are likely to be more robust if they are based on information from different types of investigation rather than only one. In other words, we would argue that, by itself, no single piece of information proves or disproves any of the eight hypotheses identified earlier. Instead, it is the accumulation of evidence across the different types of study that matters. Let us now turn to the reasons for the creation of AAIs generally.

The Creation of Independent Administrative Authorities in France

When were AAIs created?

The 34 AAIs, or AAI-like institutions, that the Conseil d’État identified were created between January 1941 and June 2000. There is some ambiguity in specifying exactly when certain institutions were established. For example, the Commission des opérations de bourse (COB) was created in September 1967, but it was only legally classified as an AAI as part of the July 1996 Law on the Modernisation of Financial Activities (loi no. 96-597). Once again, in order to avoid any personal judgement as to when particular institutions were created as AAIs, we have decided to rely on the dates provided by the Conseil d’État itself. For example, the Conseil d’État’s chronology suggests that the most appropriate date for the COB in this regard is 1996, rather than 1967. On this basis, we find that one AAI was created in the 1940s, two in the 1950s, one in the 1960s, five in the 1970s, 12 in the 1980s and 13 in the period 1990-2000 inclusive. Further details about the 31 AAIs in the Conseil d’État’s inventory that were created during the Fifth Republic (1958-) are set out in Table 2.

[TABLE 2 ABOUT HERE ]

In terms of when AAIs were created, there is prima facie evidence to support a small number of hypotheses. For example, the vast majority of AAIs were created after the 1980s. This timeframe roughly corresponds with the increasing Europeanisation of the policy process and, hence, evidence for institutional isomorphism, as well as the spread of new public management norms in OECD countries generally. In addition, nearly 40 per cent of all AAIs were created under just two governments, the governments of Michel Rocard and Lionel Jospin respectively. These figures might provide some evidence for the importance of political leadership, particularly as both leaders, and especially Rocard, were keen to promote state reform (Elgie 2003). By contrast, ideology does not appear to have been important. Delegation occurred under both left and right-wing governments. Moreover, the most neo-liberal-minded administration during the Fifth Republic, Jacques Chirac’s 1986-88 government, was not associated with a high level of delegation. If we assume that right-wing governments are more likely to be influenced by the principles of new public management, then the evidence contradicts the new public management hypothesis. By the same token, the fact that only six of the 31 Fifth Republic AAIs were created in the 12-month period prior to a scheduled legislative election does not provide evidence of the policy uncertainty hypothesis. The legislative term is five years, so this figure corresponds to the annual percentage that would normally be expected. So, while there is some confirmatory prima facie evidence in some cases, there is also some disconfirming evidence. What is more, this sort of evidence cannot easily be used to test a number of hypotheses, such as blame shifting. (The findings are summarised in Table 3.)

Which Policy Areas Do AAIs Cover?

In the French literature on AAIs, various attempts have been made to classify the different areas in which AAIs operate. For example, the Conseil d’État (2001, 307) states that AAIs are usually said to operate in two basic areas: regulation and the protection of civil liberties. Indeed, this is exactly the distinction made by Gentot (1994, 48). Moreover, consistent with this dichotomous classification, Jorion (1998, 41) states that AAIs operate in two sectors: the economy, as well as information and elections. That said, some writers have indicated that a broader classification schedule is more appropriate. For example, Guédon (1991, 17) states that they operate in three general areas: the battle against red tape, the regulation of the market economy, and the information and communications sector. What is more, she then goes on to suggest (ibid., 18) that in this latter sector there are three sub-sectors: the regulation of broadcasting, political life, and scientific and technological activity. Finally, the Conseil d’État (2001, 307) itself has declared that a broader perspective is more appropriate. In its report, the Conseil d’État identified five general areas in which AAIs now operate: ombudsmen and mediation, regulation, the protection of civil liberties, as institutions that safeguard the public against arbitrary decisions by the state, and in a residual area that brings together a heterogeneous set of institutions, such as the anti-doping authority, the Conseil de prévention et de lutte contre le dopage.

The presence of AAIs in the above areas provides evidence to support various hypotheses. Most importantly, it helps to support the credible commitment hypothesis. The literature on this topic invariably focuses on the economy as well as the more general need for independent agencies to regulate activity in areas where otherwise the government would be tempted to intervene. Whatever the classifications that individual writers have proposed, all of them state that AAIs operate in the area of the economy and regulation. Moreover, many state that they also operate in areas where individuals or groups need to be protected from the possibility of government intervention. In addition, there is also some support for the institutional isomorphism and new public management hypotheses. In terms of institutional isomorphism, while it is beyond the scope of this article to demonstrate the point, we can assert with some justification that several of the AAIs in France are similar to institutions found elsewhere, notably classic IRAs such as the Autorité de régulation des telecommunications. As for new public management, the various ombudsmen, namely the Médiateur de la République, the Défenseur des enfants and the Médiateur du cinema, are entirely consistent with the basic principles of new public management reform. In addition, if we accept Guédon’s classification, then there is also some support for the policy complexity hypothesis. The creation of AAIs in areas where there is scientific and technological activity suggests that the government has accepted the need to delegate decision-making authority to policy experts. Finally, the list of AAIs seems to contradict some of the established wisdom about the tradition of state intervention in France and, thus, weakens this hypothesis. In some respects, the fact that AAIs exist at all runs contrary to the French tradition of administrative law and the organisation of the state. Indeed, Jorion (1998, 42) says that they mark a “completely remarkable rupture with the Napoleonic conception of a hierarchical system where the administration is subject to the orders of the executive” (all translations from the French are by the author). However, the fact that there are AAIs operating in areas such as the regulation of phone tapping, the Commission nationale de contrôle des interceptions de sécurité, and the classification of defence secrets, the Commission consultative du secret de la défense nationale, is an even more tangible sign that attitudes towards the appropriate role of the state have changed. Traditionally, the French state has maintained close control over ‘high’ politics, including issues relating to national security. The fact that the French government has decided to delegate decision-making authority to quasi-autonomous agencies in some aspects of this domain is an indication that the state tradition argument is not appropriate in this regard at least.

How Are AAIs Structured?

AAIs comprise a very heterogeneous set of institutions. As a result, and given space constraints, it is impossible to go into any great detail about how they are structured. That said, two observations can be made.

The first observation concerns the structure of their boards of governors. Here, it is noticeable that political appointees can be found on many of these boards. In some cases, such as the Autorité de régulation des télécommunications (ART), all the board members are appointed by political representatives. In other cases, such as the Commission de la sécurité des consommateurs (CSC), the percentage of political appointees is relatively small. The fact that AAIs often include political appointees on their boards provides some support for the political uncertainty hypothesis. It might be argued that governments are aware of the fact that they will lose office at some time and so construct AAIs in such a way that they can still maintain a majority on their boards. That said, another noticeable feature about AAIs is that political appointees are often required to have professional expertise. So, for example, only ‘qualified’ people can be appointed to ART’s board and collectively its members are required to have expertise in law, technical issues and the economy. The prevalence of this ‘expertise’ requirement provides support for the policy complexity argument. That is to say, in many cases the government has expressly delegated authority to policy experts. What is more, in typical French fashion, members of the grands corps are often appointed to the boards of governors. These people are civil servants, but they are members of the highest administrative institutions in the land, notably the Conseil d’État, the Cour des Comptes, and the Cour de Cassation, and are experts in matters relating to law, including administrative law, and public finances. Thus, there is further evidence for the complexity argument. Finally, there is also evidence for the credible commitment hypothesis. In many cases, the boards include external representatives. For example, the CSC includes representatives of consumer groups and professional associations. In this way, it might be argument that government have delegated authority to institutions that are populated by people with no strategic interest in either the government or the state. This is entirely consistent with the credible commitment hypothesis.

The second observation is that many AAIs are able to exercise quite considerable powers. True, there are some that are relatively toothless. Most notably, the Commission d’accès aux documents administratifs has only an advisory role. Equally, the role of the various ombudsman institutions, see above, is relatively limited. Most AAIs, though, have the power to make a range of decisions, including the imposition of sanctions. This point is particularly important in regard to the credible commitment hypothesis. The creation of an AAI is not sufficient in itself to show that governments wish to avoid any future intervention in the policy process. In order to make such a commitment a credible one, governments need to transfer some significant degree of powers to the authorities in question. It is beyond the scope of this article to demonstrate that governments have done, even though this issue will be addressed in the case studies that follow. In general terms, though, it can be asserted that most AAIs have considerable decision-making responsibilities and are not just examples of political window-dressing.

What General Reasons Are Given as to Why AAIs Have Been Created?

In the literature on AAIs, the most common reason that is put forward to explain why they have been created is the idea that governments are responding to new policy problems and, specifically, to a more complex decision-making environment. For example, Teitgen-Colly (1988, 24) argues AAIs “are, in effect, a new way in which the public powers are responding to the problems that they face in certain policy sectors”. More specifically, she refers to the “complexification” of social life as a factor that explains their creation (ibid., 27). Similarly, Guédon (1991, 16) writes that AAIs “were a response either to the emergence of new problems or to problems that had become more acute; they were a sign that traditional state structures were unable to respond to the evolving and more complex set of societal problems with which the state was faced”. Gentot (1994, 38) discusses their creation in the context of “technological revolutions” over which citizens had no control. Jean-Jacques Daigre (2002, 8) writes “everything starts from the observation that the State was no longer really able to regulate a certain number of activities because of their technical nature …”. All told, there is no doubt that those who write about AAIs provide evidence that supports the policy complexity hypothesis.

Another very common reason that is used to explain the creation of AAIs is that they are part of the response to make political action more credible. In this regard, it is worth quoting Guédon at length. She writes (1991, 24): “The essential objective is to engage in certain sectors in activity that is separate from the influence of political power, meaning executive power. The ‘crisis of the State’ and the discrediting of political life have combined to weaken the credibility of public power”. In its official report, the Conseil d’État (2001, 275) also emphasises this reason. The report states that AAIs are created so as to “offer public opinion a reinforced guarantee of the impartiality of the State’s actions …”. The report then goes on to link the complexity and credible commitment hypotheses (ibid., 276). AAIs, it is stated, are created to “associate professionals with the process of determining the rules that are applicable in technical areas, a process which, in order to be credible, requires the support of economic actors …”. Again, there is no doubt that the literature on AAIs supports the credible commitment hypothesis.

The situation is somewhat less clear in regard to the new public management hypothesis. In some respects, much of the language that has been used to describe the creation of AAIs is perfectly compatible with the general thrust of the new public management literature. For example, one of the first writers on the subject emphasised that AAIs were part of “a new vision of State/society relations” (Chevallier, 1989, 178) and were a sign that “a pluralist vision of the State” was beginning to emerge (ibid., 178). For another writer, AAIs were a manifestation of the desire to establish a “more modest state” (Demarigny 1996, 158). Guédon (1991, 26) states that AAIs “seem to be the sign of a new type of relationship between the administrative and the political”. Even though the language is compatible with the new public management hypothesis, writers do not explicitly make reference to this factor as an explanatory variable. In part, this may simply be because in comparative perspective the overall level of public sector reform in France is not very great (Pollitt and Bouckaert 2000, 231). Whatever the reason, while the evidence suggests that new public management norms may be consistent with the decision to create AAIs, there is little evidence to suggest that such norms were an explicit part of the decision to delegate authority.

In addition to these reasons, other reasons are mentioned less frequently, but nonetheless seem to have some relevance. In this regard, the impact of institutional isomorphism has been implied by a number of authors. Perhaps most explicitly, the Conseil d’État (2001, 270) suggested that foreign agencies, such as US Securities and Exchange Commission, acted as a model that France adopted. The Conseil d’État (ibid., 271-274) also made explicit reference to the impact of the European Union on the decision to create AAIs. In addition, one author hints at blame shifting as a possible motivation. Georges Dupuis (1988, 17) talks, cynically, about the “cowardice of power”. He states that Pontius Pilate may act as a role model for certain politicians and that it is a natural temptation for them to “wash their hands” of difficult problems. While the language is a little undiplomatic and no specific evidence is provided to back up the claims, the implication is clear.

Table 3 summarises the findings in this section. In general terms, there was strong evidence to support the credible commitment, policy complexity and institutional isomorphism hypotheses. There was also a limited degree of evidence to support the political entrepreneurship, political uncertainty, blame shifting and spread of new public management reforms hypotheses. By contrast, there was evidence that contradicted the state tradition hypothesis.

[ TABLE 3 ABOUT HERE ]

Explaining the Creation of AAIs: The Conseil Supérieur de l’Audiovisuel and the Commission de Régulation de l’Électricité

In this section, we deepen the level of analysis. We have chosen to examine the circumstances surrounding the creation of two particular AAIs, the Conseil supérieur de l’audiovisuel (CSA), the broadcasting regulator, and the Commission de régulation de l’électricité (CRE), the electricity regulator. These institutions have their counterparts in other countries. As a result, we cannot be accused of stacking the deck against certain explanations by focusing on relatively idiosyncratic agencies that perhaps only exist in the French case. What is more, in the French literature on AAIs the CRE is usually classified as an economic regulator, while the CSA is often classed as a social regulator. Thus, they operate in the two general areas that are often identified in the literature on this topic. Finally, the CSA was established in 1989, whereas the CRE was created in 2000. Therefore, if there has been a change in the motivations for the creation of AAIs over time, we may be able to capture this change by looking at these two institutions. Overall, we do not claim that the CRE and the CSA are typical of AAIs as a whole, but we would claim that they do not bias the findings in favour of particular hypotheses and that there is sufficient variation between them for the salience of the different hypotheses to be tested.

Having chosen the two case studies, we focus on the parliamentary debate that took place at the time of their creation and try to identify the motivations of the decision-makers at the time. More specifically, we follow the debates that took place in the National Assembly and examine the arguments put forward by both the government and opposition. The law that created the CSA was debated in the lower house on the 5, 6 and 15 December 1988. The law that created the CRE was debated there on the 28 January, 16, 17, 18 and 19 February and 2 March 1999. The 1988 law was almost solely confined to the establishment of the CSA, while the creation of the CRE was only part of the 1999 law. As a result, in the latter case we will concentrate on those parts of the debate that dealt directly with the CRE. In both cases, we try to summarise the arguments that were put forward to determine the level of support for each of the eight hypotheses identified in section one.

The Conseil Supérieur de l’Audiovisuel

The debate in the National Assembly concerning the creation of the CSA was instructive in a number of ways. Specifically, four observations can be made.

The first observation is that in the case of the CSA there was little or no supporting evidence for two hypotheses: the spread of new public management reforms and blame shifting. Evidence for the spread of new public management reform hypothesis would have been found if deputies had discussed the creation of the CSA in the context of wider administrative reforms. However, at no point did they do so. So, we can discount this hypothesis. The same is true for the blame-shifting hypothesis. Even if we assume that a government is unlikely to reveal publicly that blame shifting is one of its motivations, what is noticeable is that at no point in the debate did the opposition try to suggest that the government had this motivation in mind. So, there was no evidence for this hypothesis.

The second observation is that there was limited evidence to support two hypotheses: policy complexity and institutional isomorphism. On two occasions, the complexity of the policy area was implied. At the very start of the debate, the government’s main rapporteur, Jean-Jacques Queyranne, justified the bill by saying that the organisation and structures of the broadcasting sector had been “overtaken by the very rapid changes” that had affected the area (Assemblée nationale, compte rendu intégral (henceforth, ACRI), 5 December, 1988, 3097). The rapporteur of one of the parliamentary committees, Jean-Pierre Michel, provided the same justification, stating that the reasons for yet another bill in this policy sector was because of the “technological evolution” in this sector (ACRI, 5 December, 1988, 3102).

There is also limited evidence to support the institutional isomorphism hypothesis. In this regard, it should be noted that Europeanisation was not a motivation. Indeed, one of the government’s main spokespeople, Bernard Schreiner, stated that the government would wait until 1992, and the creation of the Single European Market, to take another look at the CSA’s role in the area of telecommunications regulation. So, European integration was only in the background at this time. That said, the government did note the importance of policy transfer in other respects. For example, the government’s decision not to give the CSA powers to regulate the telecommunications sector was justified on the basis that this was the situation in other countries such as Japan, Britain, Germany and Spain (ACRI, 5 December, 1988, 3104), although it might be noted that the opposition, which wanted the CSA to be given powers in this area, cited the experience of countries where the broadcasting regulator did have powers in the telecommunications sector (ACRI, 6 December, 1988, 3200). Equally, on a number of occasions Schreiner justified the way in which members of the CSA’s board would be appointed by saying that an equivalent process was used in countries like the US, Britain and Canada (ACRI, 6 December, 1988, 3121-22, 3170 and 3189). All in all, while there is not strong evidence to support the institutional isomorphism hypothesis, there is at least some evidence that the government was motivated by reasons of cross-national policy transfer.

The third observation is that there is good evidence to support the political leadership hypothesis. During the 1988 presidential election campaign, President Mitterrand promised to abolish the existing regulator, the CNCL, and replace it with a new authority. Indeed, in his de facto election manifesto, Lettre à tous les Français, Mitterrand even gave the CSA its name (Mitterrand 1988, 14). It must be remembered that the President is constitutionally prohibited from taking part in parliamentary debates. Even so, it is noticeable that throughout the National Assembly debate, Mitterrand’s name was closely associated with the reform. For example, Jean-Jacques Queyranne quoted Mitterrand as part of his justification for abolishing the CNCL (ACRI, 5 December, 1988, 3098). The junior minister with responsibility for the bill, Catherine Tasca, quoted Mitterrand at length when she tried to justify the controversial way in which the government proposed to appoint the board of the CSA (ACRI, 6 December, 1988, 3163). In fact, so frequent were the references to the President that at one point the opposition noted ironically that the Lettre à tous les Français had already acquired almost “biblical” status (ACRI, 15 December, 1988, 3695). In these ways, there is no doubt that Mitterrand was directly responsible for the creation of the CSA. This provides support for the political leadership hypothesis.

The final observation concerns three hypotheses – the policy uncertainty hypothesis, the state tradition hypothesis and the credible commitment hypothesis – where the evidence is mixed. The policy uncertainty hypothesis is a good case in point. The usual evidence to support this thesis is found in the idea that governments will create non-majoritarian institutions in the period immediately prior to an election that they know they will lose in order to ensure that the new government cannot control the policy sector in question when it comes to power. Given that the CSA was created immediately after a new government was elected, the usual evidence to support his thesis is absent. That said, there is still some evidence consistent with the policy uncertainty hypothesis. As we have already noted, the composition of CSA was a particularly controversial issue. The government proposed that there would be six members of the board and that the President of the Republic, the President of the National Assembly and the President of the Senate would each make two appointments. To the extent that the socialists controlled two of these three offices at the time, the opposition repeatedly accused the government of trying to ensure that the socialists would a majority on the new authority in the future. (For example, ACRI, 8 December, 1988, 3177). Needless to say, the government defended itself very strongly against this charge, but we can say that there is at least limited evidence to support the policy uncertainty hypothesis.

The situation with regard to the state tradition hypothesis is similar. The CSA was one of the first AAIs to be created. Perhaps as a result, the CSA was not portrayed as being part of a French tradition of delegation to non-majoritarian institutions. Indeed, Catherine Tasca stated that the principle of creating an independent and effective regulator in the broadcasting area was “new” (ACRI, 5 December, 1988, 3128). All the same, throughout the debate the government made reference to the first broadcasting regulator, the Haute Autorité, which had been established by the socialist government in 1982. Repeatedly, the government claimed that the Haute Autorité had broken with the French tradition of state-controlled media and argued that the CSA was following in its footsteps. (For example, ACRI, 5 December, 1988, 3102, and the intervention by Jean-Pierre Michel). It must be noted that equally repeatedly the opposition, which had abolished the Haute Autorité in 1986 and replaced it with the CNCL, took issue with this argument, stating that there was no need to abolish the CNCL and arguing that whereas the CNCL was independent the CSA would not be. (For example, Gérard Longuet, ACRI, 5 December, 1988, 3115). All in all, as we might expect, the parliamentary debate provides evidence of claim and counter-claim. Even so, if we assume that the 1982 reform had started a new tradition and that the 1988 bill was another manifestation of this new tradition, then the fact that the government did at least try to justify the creation of the CSA by reference to the Haute Autorité does provide limited evidence for the state tradition hypothesis.

The same conclusion can be drawn in relation to the credible commitment hypothesis. The opposition criticised the CSA, arguing that the board would be composed of political appointees and that the regulator had not been given sufficient powers. Indeed, at one point an opposition spokesperson explicitly stated that the CSA’s credibility could be reinforced if it were to be given greater powers (ACRI, 5 December, 1988, 3147). As a rejoinder, Bernard Schreiner accused the CNCL of lacking “authority and credibility” (ACRI, 5 December, 1988, 3115). Moreover, he defended the credibility of the CSA, particularly in relation to the composition of its board (ACRI, 7 December, 1988, 3311). In fact, the parliamentary debate as a whole was characterised by the government arguing that the CSA would be independent – and, hence, credible – and the opposition arguing that it would not. What are we to conclude from this war of words? We should conclude that the credible commitment hypothesis is not strongly supported. The government did not justify the creation of the CSA in the usual terms of establishing a credible commitment in the context of market opening. At the same time, though, the government did try to argue that it was creating an independent regulatory and did make explicit reference to the concept of credibility on a couple of occasions. So, there is limited support for the credible commitment hypothesis.

The Commission de Régulation de l’Électricité

As with the CSA, the debate in the National Assembly was instructive as to the various motivations for the creation of the CRE. This time, three observations can be made.

The first observation is that in the case of the CRE there was little or no supporting evidence for either the political leadership hypothesis, the spread of new public management reform hypothesis, the blame-shifting hypothesis, or the political uncertainty hypothesis. There was absolutely no evidence to support either the political leadership hypothesis, or the spread of new public management reform hypothesis. At no point in the debate was the creation of the CRE associated with any given political leader, nor was it considered in the context of a wider set of administrative reforms. So, we can discount these hypotheses entirely. We can do the same for the blame-shifting hypothesis. As with the CSA, there was no attempt by the opposition to claim that the government was trying to shift the blame for unpopular policies to the CRE. The situation with regard to the political uncertainty hypothesis is more interesting, even if the conclusion is the same. Like the CSA, the CRE was not created in the immediate run up to an election. So, the sort of evidence that is usually called upon to support this hypothesis is once again absent. That said, most of the members of the CRE’s board are political appointees. Therefore, as with the CSA, we might have expected the opposition to accuse the government of trying to create an institution that it could control even when it lost power in the future. However, this did not happen. An amendment was tabled to reform the composition of the CRE’s board (ANCRI, 18 February, 1999, 1776), but the whole exercise was a rather half-hearted affair. In contrast to the extremely acrimonious debate surrounding the creation of the CSA more than a decade earlier, at no point on this occasion did the opposition accuse the government of trying to find ways of keeping the CRE under its long-term political control. So, we can put aside the political uncertainty hypothesis.

The second observation is that there is only a limited degree of support for the policy complexity and state tradition hypotheses. As regards the former, one of the parliamentary rapporteurs for the bill argued that any regulatory authority must have the “capacity for expertise” (ANCRI, 18 February, 1999, 1773). To the extent that he was defending the creation of the CRE, this can be taken as at least some support for the idea that the policy complexity was relevant to the creation of the CRE. By the same token, on a couple of occasions the opposition argued that the government had not given the CRE sufficient resources in regard to expertise, thus implying that policy complexity was important in this area (ANCRI, 16 February, 1999, 1499; and 17 February, 1999, 1571). All the same, at no point did the government cite the complexity of the policy area as one of its motivations for the creation of the CRE. So, the level of support for this hypothesis is limited. The situation is very similar with regard to the state tradition hypothesis. At one point in the debate, a parliamentary rapporteur noted that the CRE was just one of a number of regulatory authorities and said that deputies had been “quick” to resort to the creation of “so-called ‘independent’ authorities” in recent years (ANCRI, 16 February, 1999, 1485). Unlike the situation with the CSA in 1988, this statement supports the idea that there is now a tradition of establishing independent authorities in the French system. In this way, there is some support for the state tradition hypothesis. All the same, the level of support is low.

The third observation relates to the credible commitment and institutional isomorphism hypotheses. There was explicit support for the credible commitment hypothesis. Indeed, on two occasions the issue of credibility was explicitly mentioned. A parliamentary rapporteur noted that “the mode of regulation is a cardinal element in the success of the reform” and that “what France does will be judged by the credibility of it…” (ANCRI, 16 February, 1999, 1482). Similarly, the Secretary of State responsible for the bill stated that the role of the CRE was “indispensable” to the credibility of the market-opening process (ANCRI, 16 February, 1999, 1496). This sort of language provides good support for the credible commitment hypothesis. That said, on many occasions the Minister made it very clear that he was only willing to give the CRE regulatory powers in certain areas and argued strongly that the government needed to maintain responsibility for the regulation of various aspects of the policy sector. (See, for example, ANCRI, 18 February, 1999, 1775). So, when the opposition tried to extend the powers of the authority their amendments were rejected, which on a couple of occasions led to some fairly sarcastic remarks about the supposed independence of the CRE. (See ANCRI, 18 February, 1999, 1685; and 18 February, 1999, 1703). Given the opposition’s doubts as to the independence of the CRE, this casts some doubt on the evidence for this hypothesis. All in all, while the support for the credible commitment hypothesis is not unequivocal, there is good evidence that it was an explicit motivation of the government.

The situation with regard to the institutional isomorphism hypothesis is similar. There is good evidence to support this hypothesis. The bill as whole was motivated by the requirement for the French government to transpose Directive 96/92/EC of the European Parliament and of the Council of 19 December 1996 concerning common rules for the internal market in electricity. On countless occasions, the government stated that this was the context in which the bill was being proposed. Moreover, explicit mention of the CRE was made in this regard (ANCRI, 18 February, 1999, 1773). Further evidence to support this hypothesis relates to the concept of policy transfer more generally. On more than one occasion, the government implied that the institutional architecture of the CRE was inspired by equivalent authorities both in Europe (ANCRI, 18 February, 1999, 1702) and the world more generally (ANCRI, 16 February, 1999, 1495). Equally, at one point a parliamentary rapporteur stated that specific aspects of the CRE’s powers had been inspired directly by the Authority for the Regulation of Telecommunications, which had been established a short time previously (ANCRI, 18 February, 1999, 1787). In all these ways, there is good evidence that the creation of the CRE was motivated by a general trend towards cross-national, or cross-sector, policy transfer. That said, evidence in support of this hypothesis does need to be qualified somewhat. It should be noted that Directive 96/92 required the creation of a regulatory authority that was separate from electricity suppliers, but not one that was independent from the government. So, the government’s decision to create an AAI in this area was not necessarily a European requirement. Indeed, on occasions the government made a virtue of the fact that the government had a degree of flexibility as to how the directive was transposed. (For example, ANCRI, 16 February, 1999, 1495). By the same token, even though the government admitted that it had studied the organisation of electricity regulators in other countries, it was also stated that they constituted a “disparate” set of institutions and so the government was justified in choosing its own form of institutional architecture (ANCRI, 18 February, 1999, 1773). Finally, even if some of the CRE’s powers resembled those of the telecommunications regulator, more often than not the government stressed the electricity sector was different from the telecommunications sector (ANCRI, 18 February, 1999, 1774) and indeed that the electricity sector was “specific” (ANCRI, 18 February, 1999, 1702). Overall, we can conclude by saying that there is strong prima facie evidence for the institutional isomorphism hypothesis, but that some of this evidence should be qualified somewhat.

[ TABLE 4 ABOUT HERE ]

Conclusion

There are various explanations as to why governments should decide to delegate authority to quasi-autonomous agencies. Some of these explanations are based on a transactions cost approach. Others emphasise more contextual factors. To date, only a small number of these explanations have been systematically tested and, even then, attention has tended to focus on a relatively small number of more quantitatively tractable hypotheses. This paper has helped to fill this gap.

The existing work has provided considerable support for both the credible commitment and policy complexity hypotheses. The qualitative work in this article has suggested that these hypotheses remain strong. At both the general level and in the two case studies, evidence consistent with both explanations was found. This finding is important. Given there is now quantitative and qualitative evidence to support both hypotheses, their salience is strengthened. The need to delegate decision making to acknowledged experts and, in particular, the desire to make a credible commitment really do seem to explain in large part why non-majoritarian institutions are created.

In addition, this paper has also demonstrated support for other explanations. Perhaps most notably, there was good all-round evidence to back up the institutional isomorphism hypothesis. This evidence was apparent in both the general study of all 34 AAIs as well as in the case studies, particularly as regards the CRE. This finding is important because the institutional isomorphism hypothesis is not easily tractable in statistical terms. In addition, there was evidence in both the general study and the case studies to support the political uncertainty and political leadership hypotheses. Again, evidence for the latter is particularly important because this hypothesis has not proved amenable to statistical tests. That said, we perhaps need to qualify the strength of the evidence for the political leadership hypothesis. The main evidence for it lay in the fact that a large proportion of AAIs were created under two governments and because the CSA was closely associated with François Mitterrand. The former evidence is not compelling and the latter is perhaps institution-specific. Even so, the study has provided at least some evidence for an explanation, support for which, so far, has not been found elsewhere.

Finally, there was only very limited evidence for the blame-shifting hypothesis, the state tradition hypothesis and the spread of new public management reforms hypothesis. There was very limited evidence to support the blame-shifting hypothesis in both elements of the study. However, in each case, the evidence was either limited or equivocal. By the same token, an argument could be made that there may now be a growing tradition of delegation to AAIs in France. In this way, the state tradition argument may be more applicable to France in the future than it has been up to now. In general, though, the state tradition hypothesis was only weakly supported in this study. For its part, there was some evidence consistent with the spread of new public management reforms hypothesis in the general study, but none in the case studies. Again, on the whole this hypothesis can be largely discounted. Overall, these findings are also important. These hypotheses correspond to arguments that have been put forward in the literature but which have not been amenable to statistical tests. Therefore, even though evidence had not yet been found to support them, they were still plausible. Now, though, we have tried to test them and little evidence was found to support them. So, while we have to acknowledge that this study was based on one set of institutions in one country, these hypotheses appear to be the least applicable as things stand.

All in all, this article has helped to explain why governments have established so many AAIs in France. More than that, it has also helped to strengthen or weaken the validity of the various reasons that are put forward to explain the delegation of authority to quasi-autonomous agencies in general. In this latter regard, it has raised certain questions that need to be explored more fully, but it has helped to advance the received wisdom in this area.

Bibliography

Barro, Robert J., and David B. Gordon. 1983. “Rules, Discretion and Reputation in a Model of Monetary Policy.” Journal of Monetary Economics 12: 101-22.

Chevallier, Jacques. 1989. “Les Autorités Administratives Indépendantes.” In 1989 Universalia. La Politique, Les Connaissances, La Culture en 1988. Paris: Encyclopaedia Universalis, 177-178.

Conseil d’État. 2001. “Rapport Public 2001. Jurisprudence et Avis de 2000. Les Autorités Administratives Indépendantes.” Études et Documents no. 52. Paris: La Documentation Française.

Daigre, Jean-Jacques. 2002. “Ombres et Lumières. Examen Critique du Fonctionnement des Autorités Administratives Indépendantes.” In Le contrôle démocratique des autorités administratives indépendantes à caractère économique, ed. Hervé de Charette. Paris: Economica, 5-13.

Demarigny, Fabrice. 1996. “Independent Administrative Authorities in France and the Case of the French Council for Competition.” In Regulating Europe, ed. Giandomenico Majone. London: Routledge, 157-179.

Elgie, Robert. 2003. “Governance Traditions and Narratives of Public Sector Reform in Contemporary France.” Public Administration 81: 141-162.

Elgie, Robert, and Iain McMenamin. 2005. “Credible commitment, political uncertainty, or policy complexity? Explaining variations in the independence of non-majoritarian institutions in France”. British Journal of Political Science 35: 531-548.

Gentot, Michel. 1994. Les Autorités Administratives Indépendantes. Paris: Montchrestien. Second edition.

Gilardi, Fabrizio. 2002. “Policy Credibility and Delegation to Independent Regulatory Agencies: A Comparative Empirical Analysis.” Journal of European Public Policy 9: 873-893.

Guédon, Marie-José. 1991. Les Autorités Administratives Indépendantes. Paris: Librairie Générale de Droit et de Jurisprudence.

Huber, John D., and Charles R. Shipan. 2002. Deliberate Discretion? The Institutional Foundations of Bureaucratic Autonomy. Cambridge: Cambridge University Press.

Jorion, Benoît. 1998. “Les Autorités Administratives Indépendantes.” In Droit administratif et administration. Paris: La Documentation Française.

Majone, Giandomenico. 1996. “A European Regulatory State.” In European Union, Power and Policy-Making, ed. Jeremy Richardson. London: Routledge, 263-277.

McCubbins, Mathew D., Roger G. Noll, and Barry R. Weingast. 1989. “Structure and Process, Politics and Policy: Administrative Arrangements and the Political Control of Agencies.: Virginia Law Review 75: 431-82.

Mitterrand, François. 1988. Lettre à tous les Français. Livry Gargan: SAGIM.

Moe, Terry M. 1995. “The Politics of Structural Choice: Toward a Theory of Public Bureaucracy.” In Organization Theory. From Chester Barnard to the Present and Beyond, ed. Oliver E. Williamson. New York: Oxford University Press, 116-153.

Pollack, Mark A. 2003. The Engines of European Integration. Delegation, Agency, and Agenda Setting in the EU. Oxford: Oxford University Press.

Pollitt, Christopher, and Geert Bouckaert. 2000. Public Management Reform. A Comparative Analysis. Oxford: Oxford University Press.

Teitgen-Colly, Catherine. 1988. “Les Autorités Administratives Indépendantes: Histoire d’une Institution.” In Les Autorités Administratives Indépendantes, ed. Claude-Albert Colliard and Gérard Timsit. Paris: Presses Universitaires de France, 21-73.

Thatcher, Mark. 2002. “Delegation to Independent Regulatory Agencies: Pressures, Functions and Contextual Mediation.” West European Politics 25: 125-147.

Thatcher, Mark, and Alec Stone Sweet. 2002. “Theory and Practice of Delegation to Non-Majoritarian Institutions.” West European Politics 25: 1-22.

Weaver, R. Kent. 1986. “The Politics of Blame Avoidance.” Journal of Public Policy 6: 371-398.


TABLE 1

Hypotheses Explaining Delegation to Quasi-Autonomous Agencies.

1. Policy complexity/the increased efficiency of rule making

2. Credible commitment

3. Blame shifting

4. Political uncertainty

5. Institutional isomorphism, including Europeanisation

6. State tradition

7. Political entrepreneurship

8. The spread of new public management norms


TABLE 2

Data relating to the creation of AAIs during the Fifth Republic.

Under which presidencies were AAIs created?

de Gaulle

(1958-69)

1

Pompidou

(1969-74)

0

Giscard d’Estaing

(1974-81)

5

Mitterrand

(1981-95)

15

Chirac

(1995-end 2000)

10

Under which types of premiership were AAIs created?

Left-wing

12

Cohabitation with

left-wing PM

6

Right-wing

10

Cohabitation with

right-wing PM

3

Under which premierships were AAIs created?

Rocard

(1988-1991)

7

Jospin

(1997-

end 2000)

6

Barre

(1976-1981)

5

Mauroy (1981-1984)

4

Juppé

(1995-1997)

4

Chirac

(1986-1988)

3

Pompidou

(1962-1968)

1

Bérégovoy

(1992-1993)

1

At what point in the electoral cycle were AAIs established?

More than 12 months before

a scheduled election

25

Fewer than 12 months before

a scheduled election

6


TABLE 3

Summary of the findings for all 34 AAIs.

When created

Areas

Structure

Reasons

Policy complexity

No evidence

Limited evidence

Good evidence

Good evidence

Credible commitment

No evidence

Good evidence

Good evidence

Good evidence

Blame shifting

Not applicable

No evidence

No evidence

Limited evidence

Political uncertainty

Limited evidence

No evidence

Good evidence

No evidence

Institutional isomorphism

Limited evidence

Limited evidence

No evidence

Limited evidence

State tradition

No evidence

Evidence against

No evidence

No evidence

Political leadership

Limited evidence

No evidence

No evidence

No evidence

Spread of NPM norms

Evidence for and against

Limited evidence

No evidence

Limited evidence

TABLE 4

Summary of the findings for the CSA and CRE.

CSA

CRE

Policy complexity

Limited evidence

Limited evidence

Credible commitment

Limited evidence

Good evidence

Blame shifting

No evidence

No evidence

Political uncertainty

Limited evidence

No evidence

Institutional isomorphism

Limited evidence

Good evidence

State tradition

Limited evidence

Limited evidence

Political leadership

Good evidence

No evidence

Spread of NPM norms

No evidence

No evidence

Responder

Por favor, inicia sesión con uno de estos métodos para publicar tu comentario:

Logo de WordPress.com

Estás comentando usando tu cuenta de WordPress.com. Cerrar sesión / Cambiar )

Imagen de Twitter

Estás comentando usando tu cuenta de Twitter. Cerrar sesión / Cambiar )

Foto de Facebook

Estás comentando usando tu cuenta de Facebook. Cerrar sesión / Cambiar )

Google+ photo

Estás comentando usando tu cuenta de Google+. Cerrar sesión / Cambiar )

Conectando a %s

A %d blogueros les gusta esto: