From regulation to governance and representation: agenda-setting and the EU’s involvement in sport

This paper presents the origins and development of the EU’s involvement in sport through the examination of the landmark decisions that have shaped its approach over time. The initiation and development of that policy can be considered an example of ‘task expansion’, in which the EU has extended the scope of its activities as a consequence of actors instrumentalising institutional venues to their own benefit. It draws on concepts from agenda-setting to argue that the initial insertion of sport in the EU’s systemic agenda can be explained by the commercialisation of sport in the 1980s and 1990s. However, actor centered agenda-setting models may be more suitable to explore the consideration of sport in the institutional agenda after the 1995 Bosman case. Two routes of agenda-setting are identified: the high politics route and the low politics route. At first, sports policy was just regulatory in nature, being introduced through the low politics route. The response to the Bosman case from actors both outside and within the system (mainly sports federations and European Council) moved the issue to the high politics route, focusing more on the socio-cultural and educational particularities of sport.


INTRODUCTION
The Treaty of the European Communities, Art. 5 (1) provides that the European Union (EU) 'shall act within the limits of the powers conferred upon it by this Treaty and of the objectives assigned to it therein'. Article 3 TEC does not cite sport as a competence of the EU, which means that no authority has been conferred to the EU to develop any kind of policy on sport. However, the EU has got involved in sport-related matters for more than 30 years now, and especially in the last decade after the well known ruling of the Court of Justice of the European Communities (ECJ) in Bosman v. Union Royale Belge Sociétés de Football Association, (case C-415/93 [1995] ECR I-4921, hereinafter Bosman). Through the years, two different groups of decisions have shaped sport policy. On the one hand, there has been a general and overarching debate on how sport should be treated by the European institutions, as if the EU was asking, ' what is sport?' This debate is particularly reflected on the decade-long negotiations leading to the inclusion of an article on sport (Article III-282) in the draft European Constitution (OJ C 310/2004, 16 December). Upon the failure of the Constitutional treaty, recent initiatives such as the Independent European Sport Review (Arnaut, 2006) or the European Parliament's report on the future of professional football (European Parliament, 2007) have pursued this avenue. On the other hand, Commission decisions and ECJ rulings in particular cases have also shaped the way in which EU sports law and policy has developed. culminated in Bosman; second, the intervention of high politics in the debates about the inclusion of sport in the European Treaties. Finally, the article looks at the more recent interest in the governance of sport in initiatives such as the Independent European Sport Review.

WHAT IS AGENDA-SETTING?
Agenda-setting has received very little attention as a conceptual framework to analyse EU policy-making. It has been used as part of principal-agent models of European politics (Pollack, 1997;2003), whereas other accounts present a rather general description of some basic characteristics of agenda-setting in the EU (Peters, 1994;2001) but do not analyse the particular dynamics of agenda-setting. This is striking given the potential that agenda-setting might have if applied to the EU. European integration is a history of incremental development of competences, in which new areas of action are taken to the supranational level to form new policies. Moreover, policy-making in the EU is a multi-venue process that offers many entrance points to those wishing to affect the range of issues present in the agenda. In principle, one would think that agenda-setting has suitable tools to inform our study of these processes. This article draws on the vast literature on agenda-setting at national level, where it has received more attention than in EU studies (although in the United States agenda-setting has developed as an area of study in its own right). The article links some of the agenda-setting concepts used in that literature with the EU's very own institutional and legal framework, taking account in this respect of the very recent contribution of Princen and Rhinard (2006).

3
Before going into the details and disaggregating the dynamics of agenda-setting, it is necessary to clarify from the outset that agenda-setting is not just the simple decision of placing an issue into the agenda. If we want to make use of the explanatory powers of agenda-setting then the concept has to be broadened. Indeed, agendasetting needs to be understood as a cumulative process (Peters, 2001, p. 78); therefore, it is more appropriate to use the term agenda-setting process or, in short, the 'agenda process' (Robinson, 2000, p. 13). Agendasetting relates to the decisions by which new issues are recognised by policy-makers, hence jumping onto the agenda; but it also relates to the way in which those issues are defined or framed and the impact of that definition in the policy adopted subsequently. Moreover, agenda-setting does not only concern new issues entering the agenda, but also how old issues do move up and down in the agenda and get redefined along.

4
For analytical purposes, one can distinguish four main elements in the agenda process. First, issue recognition or identification is the process by which the attention of policy-makers and those around them is captured by particular problems over other ones, hence entering the agenda (Kingdon, 1995, p. 87). Second, the construction of problems (issue framing or definition): during the agenda process, issues are not only identified as important, but they are also defined and redefined whilst rising through the agenda. Getting a particular definition may affect the rest of the whole policy-making process (Peters, 2001, p. 78). Third, the specification of different policy alternatives: this is the process where policy advocates and specialists propose their preferred solution(s) to the issue that is occupying the decision-maker's agenda (Kingdon, 1995, p. 87). To put forward their preferred policy, they may need to change the ongoing issue-framing. Finally, the emergence/switch of policy arenas: when new issues rise to the agenda, there is a chance that institutional arrangements are set up in order to deal with it. Institutional actors and other groups in society will organise the necessary structures to deal with the problem in question. These structures can either be new (if the issue is deemed to be completely new) or just part of the available institutional framework (Baumgartner and Jones, 1993, p. 32).

5
The term 'agenda', as applied to policy-making, can be defined as 'the list of subjects or problems to which government officials, and people outside the government closely associated with those officials, are paying some serious attention at any given point in time. (Kingdon, 1995, pp. 3-4). Cobb and Elder (1972) distinguish the systemic agenda and the institutional agenda. The former consists of 'all issues that are commonly perceived by members of the political community as meriting public attention and as involving matters within the legitimate jurisdiction of existing governmental authority' (Cobb and Elder, 1972, p. 85). The institutional agenda, in turn, is composed by ' that set of items explicitly up for the active and serious consideration of authoritative decision-makers' (86). The greater openness of EU's policy-making (as compared, for instance, with national political systems) seems to be apparent at least for getting issues on the systemic agenda, but it makes it more difficult to move those issues to an active institutional agenda (Peters, 2001).
6 Richard Parrish points out the similarity between the notion of the systemic agenda and the construction of the European institution's agendas and between the institutional agenda and the stage in which issues are finally defined and European policies are shaped (Parrish, 2003b, p. 40). The construction of the EU's systemic agenda can, thus, be understood as the process by which policy-makers select certain issues as worthy of their attention, although this may not involve active policy formulation. From the four components of the agenda process outlined above, the systemic agenda will be most related to issue recognition and to problem framing. The promotion from the systemic agenda to the formal institutional agenda is influenced mostly by the specification of policy alternatives and the nature of policy venues.

AGENDA-SETTING IN THE EUROPEAN UNION
The characteristics of the European Union as a political system include a fragmented decision-making system with multiple avenues for influence. The absence of a clear policy co-ordination and a multi-level system of governance, make agenda-setting 'very much akin to the model of garbage can decision-making' (Peters, 1994, p. 20). The agenda process in the EU is characterised by a large number of actors involved and a great number of policy alternatives, 'with the consequence that a policy entrepreneur may be able to expand the range of issues under consideration and with it expand the scope of Community action' (Peters, 1994, p. 20). Whilst Peters provides a good description of the consequences of the EU's cumbersome policy-making process on agenda-setting, he does not go as far as to analyse how this may work. For that, the literature on national agenda-setting can be of much help. John Kingdon's work on agenda-setting applies the 'garbage can model' to the agenda process. Kingdon considers that in a political system there are three different 'streams' -problems, policies and politics -each one governed by a different dynamic (Kingdon, 1995, pp. 87-89).

8
First, the problems stream is that of situations within a polity. Problems develop independently of other factors within the political system and we should focus our attention on 'how problems capture the attention of decision makers' (Kingdon, 1995, p. 90). This is true for the European Union, which might not be a tight polity in comparison to the Member States, but it certainly has its own sphere of problems. Indeed, some theories of European integration, such as neo-functionalism (see for example Haas, 1968;Lindberg, 1963), adopt a problem centred approach to explain the logics of integration.

9
Second, the policy stream comprises the alternatives and proposals generated by both insiders and outsiders in a political system for the different problems in society (Kingdon, 1995, p. 116). The policy stream focuses, then, on the interactions of different actors while proposing their preferred policy options to a problem. These dynamics can also be recognised in the European Union, where institutional and non-institutional actors need to negotiate to find consensus around issues in policies that normally take the shape of global package deals.

10
Finally, the politics stream can be regarded as the framework in which the agenda process takes place. 'It is composed of such things as public mood, pressure groups campaigns, election results (…) or changes of administration' (Kingdon, 1995, p. 145). The politics stream, as understood by Kingdon, is composed of two elements: public sphere and institutional settings. The influence of the public sphere in EU politics is less likely to be relevant (Princen andRhinard, 2006, p. 1121). There is little evidence to support the existence of a European common space or a European public agenda, which would be necessary to create an EU-wide public sphere. The importance of the institutional setting, though, is far more important. There are two possible routes for agenda-setting in the EU: the 'high politics route' and the 'low politics route' (Princen and Rhinard, 2006).

11
Building on Kingdon's work, one can identify three major forces behind the EU's agenda process: problems, actors and institutions. This is certainly not a revolutionary finding as other conceptual approaches to European policy making have used these elements to date. The difference is that agenda-setting makes an alternative use of them. In the field of European sports policy, Richard Parrish uses the Advocacy Coalition Framework (Sabatier, 1998) to formulate his 'actor centred institutionalism' (Parrish, 2003a;2003b). In short, Parrish explains the evolution of EU sports policy as the struggle between two rival advocacy coalitions. On the one hand, there is a coalition advocating a regulatory approach to sport as an economic activity. On the other hand, there is a rival coalition advocating a more conciliatory line, focusing on the socio-cultural elements of sport. The evolution of the policy is explained in terms of the different weight of each coalition's belief systems and institutional resources.

12
Whilst this article certainly uses the idea of two different policy alternatives along similar lines, it is framed differently. First, it is submitted that actor centred institutionalism focuses more on the development of sports policy, rather than in the origins. It does not help to explain the reasons why sport became an issue of attention, which has an impact on the final policy results. Agenda-setting can provide a more robust link between policy initiation and policy output. Moreover, agenda-setting, through the concept of visibility provides a more suitable explanation to the different attitudes of the Commission before and after Bosman. Second, agenda-setting allows for more flexibility in the analysis of actors' behaviour because it considers each one individually. It is debatable whether the 'socio cultural coalition' within the sports policy subsystem recognised by Parrish is actually a coalition, even if it is considered a coalition of convenience (Parrish, 2003b, p. 68-75).

The evolution of problems and the political agenda
Problems are the first variable identified in our analysis of the EU agenda. In this vision the agenda is set by the evolution of a problem, which demands some form of response. Kingdon (1995, p. 109) draws a distinction between a 'condition' and a 'problem'. Conditions are present daily in a community and in our life. A condition is very broad, it is almost everything happening: 'bad weather, illnesses, poverty…' Conditions are only defined as problems 'when it is believed that something should be done about them' (109). Problems, once identified, have two essential components. First, the social conditions (the problem as it is, its nature) and second the interpretation or perception that actors make of these social conditions (the problem as it is seen, framed or defined) (110). The possibility that a problem will rise high in the political agenda are closely linked to its nature. However, the interpretation of problems made by actors is as important, if not more so: 'for those who wish to control the dynamics of an issue, the manipulation of the public's perception of it is vital' (Robinson, 2000, p. 18). Moreover, it is highly important to note that problem definition 'is by no means an a priori given' as actors within policy communities will not necessarily agree on a definition (Cobb and Elder, 1972, p. 101).
A key element to the definition of a problem is visibility, which can be defined as the number of persons or groups that are aware of the existence of the problem and its possible consequences (Cobb and Elder, 1972, p. 43). Visibility is 'a key requirement if an item is to engage the interest of an actor or institution who will subsequently act on it' (Hogwood and Gunn, quoted in Robinson, 2000, p. 17). Thus, problem definition and visibility for the policy community become paramount in the rise of problems to the agenda. Therefore, actors in policy-making will try to manipulate in their favour the perception of these two dimensions. According to problem centred models of agenda-setting, problems are constructed during the conflict between actors over visibility and definition (Robinson, 2000, p. 16). This is especially important because the definitional bias which a problem reaches the agenda with will condition the development of a policy. The interest of focusing in the definition of a problem lies in the way in which such a definition can condition policies. The nature of the problems is likely to have more influence in the early stages of the agenda process, the formation of the systemic agenda.

The role of actors in the agenda process
Actor centred models consider the agenda process from a conflict perspective, where actors try to retain control over the scope, intensity and visibility of an issue (Robinson, 2000, p. 23). The activities of actors are important both in the recognition and framing of problems and in the specification of policy alternatives. Therefore, actors have impact in both the systemic and the institutional agenda, although it is in the latter where this variable tends to have a greater explanatory power. The EU being an open political system with a large number of actors, it is only natural to find conflicts among actors about the suitable policy solution to the problems present in the agenda.

16
In the early stages of the agenda process, insiders to the policy community will try to restrict the scope of any rising issue to maintain it within the core of the policy community, hoping to retain problem definition of the issue (and therefore a policy action) suitable to their interests. If there is agreement over the definition of the issue and the suitable policy to be adopted, the agenda will be controlled by insiders. However, if an insider group does not find its interests well served by the initial policy proposals, it will try to expand the scope and visibility of the issue to include actors outside the policy-making core, generating first an internal [to the policy community] legitimacy crisis in the system Jones, 1991, p. 1056). Early action could prevent a low salience issue from getting to the high levels of the political agenda; policy images and problem definitions are better controlled at the outset, when rival conceptions tend to be scarce. However, if insiders succeed in challenging the prevailing status of a problem, attracting the interest of others outside the policymaking core, an 'external legitimacy crisis' may develop, provoking a reassessment of the policy definition (Baumgartner and Jones, 1993, p. 68).

17
Outsiders, on the other hand, find it more difficult to influence the agenda process. If they are sufficiently informed, they will try to propose their preferred definitions and policy proposals when issues are growing on the agenda. If they are not closely related to the policy subsystem, they will be at a disadvantage because they will struggle first to enter into the subsystem and afterwards to get their options considered as policy alternatives. For outsiders, a key method to affect policy is to use indirect methods, by moving the discussion out of the normal core arena and making it open to the public (Cobbet al., 1976). One way for outsiders in the EU to affect the agenda is the use institutional resources at their disposal, such as legal challenges, and direct contacts with political representatives (MEPs, Commission officials, political groups, lobby groups, etc). If successful, the legal challenge is obviously a strong avenue due to its direct enforcement, but on the other hand it may be time consuming.

High politics and low politics in agenda-setting in the EU
The interaction between problems and actors is at the core of agenda-setting in the European Union. There is, however, another key variable that needs to be analysed, for actors and problems do not interact in a vacuum, but rather in a structured (and quite complex) institutional framework. The importance of policy arenas in agenda-setting cannot be underestimated because there is a strong link between the framing of an issue and the venue in which it is being dealt with (Baumgartner and Jones, 1993). The EU offers multiple opportunities for venue shopping, as issues move from one institution to the other during the long and complex policy-making process. Thus, there are many opportunities to set or modify the agenda. Prince and Rhinard (2006) have identified two routes in which agenda-setting operates through the EU's institutional framework. They build on the old distinction between high politics and low politics.
The high politics route (Princen and Rhinard, 2006, pp. 1121-1123 is primarily a political one. It commences in the European Council, where problem recognition and some issue framing/definition takes place. The nature of the problem, especially if coupled with a focusing event, is a major factor in the initiation of the agenda in the high politics route. The European Council, due to its own nature, will normally limit itself to providing general principles about the issue in question, so normally the bulk of problem framing will go down one level to either the Commission or the Council of Ministers, who have to interpret the decisions of the political leaders and act accordingly (1121). This is how the systemic agenda is formed through the high politics route. It is a top-down approach. Once the problem in question is recognised and framed, it moves outside the initial circle of policy initiators to the formal institutional agenda, where different policy alternatives will be considered by the Commission, Council of Ministers and European Parliament. These changes of venues may be suitable for outsiders to try to influence policy output. The high politics route has the advantage of providing momentum to EU policy making, as the Commission and other institutions may feel compelled to follow the European Council's recommendations. On the other hand, once the issue has left the political leaders, it may be watered down during the formulation of the policy, especially if the debate reverts to be a low key and technical issue (1122). If the political momentum is maintained and the issue remains framed in political terms, it may be more difficult to change with technical arguments.

20
The low-politics route (Princen and Rhinard, 2006, pp. 1121-1123 is mainly technocratic and technical. Issues do not originate here due to a political decision, but rather through the professional and technical concerns of people working in the same issue area. This is rather typical of the Commission and Council working groups. There are also opportunities for individuals (such as Bosman's lawyers) to use the institutional framework to promote issues from below. Issue framing in the low political route will take part in the Commission and the Council, but the process is likely to be framed in much more precise and technical terms than the high politics route due to the work of expert groups (1121). This is the formation of the systemic agenda. Once the issue is recognised and framed as a problem, it has to be brought to the formal agenda and go through the decisionmaking process. The dynamics of the low politics route in agenda-setting are bottom-up (1122). Moving issues up to the institutional agenda in this route is difficult, because the receiving institutions can possibly veto or modify the problem as presented to them. For issues to be promoted to the institutional agenda in the low politics route, advocates need to gradually build support around them, so there is a general consensus that the new problem deserves higher agenda status. The risk is that problems might be blocked, re-defined or even 'hijacked' by other actors when moving up in the agenda.

21
The high politics route is more likely to affect problem recognition and problem framing, whereas the low politics route can have a major impact in the re-definition of a given problem or the elaboration of policies once problems are in the agenda. It is necessary to understand the two routes as the two ends of a spectrum, as it is unlikely for an issue to follow purely just one of them (1122). Indeed, issues may change from one route to the other. A particular problem may start as low politics, but be promoted to high politics by a focusing event and the reverse is also possible. Moreover, both processes can unfold simultaneously and have reciprocal effect on each other (1123).

THE EUROPEAN UNION AND SPORT: A REGULATORY START
This section looks at the early stages of the EU's involvement in sport related matters. This period is mostly characterised by a regulatory approach in which European institutions were required to analyse whether the rules adopted by sports governing bodies were in line with EU law, normally at the request of third parties, the ECJ ruling in Bosman being the peak of this regulatory activity.

23
Before Bosman, European institutions suffered what could be considered as relatively low-key encounters with sport. It is worth noting that both Walrave and Koch v. Association Union Cycliste Internationale, (case C-36/74 [1974] ECR 01405, hereinafter Walrave) and Donà v. Mantero, (Case C-13/76 [1976] ECR 01333, hereinafter Donà) were cases brought to the courts to challenge rules adopted by sports governing bodies (cycling in Walrave and football in Donà) that particular individuals deemed to be contrary to their rights as established in the TEC. It is also noticeable that, despite the ECJ rulings, no other EU institution took especially robust measures in the field of sport. There was some activity, but nothing comparable with the post-Bosman period. The European Commission took a contradictory approach (Parrish, 2003a) because it did not pursue action against football governing bodies despite claiming that the restrictions highlighted by the ECJ in Donà should be lifted. Instead, the European Commission preferred to reach a negotiated solution (European Commission, 1996). In the meantime, national governments meeting in the Council did not seem to be very interested in sport, except with problems regarding doping (see for example Council of Ministers, 1990) and the European Parliament was more interested in finding a European dimension to sport, although it did remind the Commission to 'take energetic steps against rules that limit the freedom of movement and establishment of citizens engaged in certain sports' (European Parliament, 1984, paragraph 8).
The ruling in Bosman seemed to give new impetus to the European Commission, especially DG Competition. Karel van Miert, the Commissioner in charge of that area, was rather quick to stress that 'UEFA has to evolve, whether they like or not' as he threatened football governing bodies with sanctions if they did not apply the ruling immediately (Kempson, 1996;Hopquin, 1995). In its commitment to make sport organisations accountable to European law, DG Competition started proceedings in four high profile cases: (i) UEFA's rules on football broadcasting hours (European Commission, 2001a); (ii) the possible dominant position of motor-sports' governing body (FIA) in relation to the organisation of Formula One (European Commission, 1999b); (iii) FIFA's transfer system (European Commission, 2002a) and (iv) the central marketing of Champions League's television rights (European Commission, 2001b). All were settled after negotiations with the affected sports governing bodies, without the Commission having to resort to sanctions or formal decisions.

25
The initial stance of the Commission's Competition authorities was received by the affected sports organisation rather by surprise, especially in the case of FIFA and UEFA (Interview, Gerhard Aigner, former UEFA General Secretary and CEO, Nyon, 6 February 2007). However, the long negotiations proved to be, at the end, constructive for football representatives, which now tend to see that period as a tough but positive one in which to get acquainted to a new reality (Interview, UEFA top senior official, Nyon 15 February 2007).

From the systemic to the institutional agenda: Visibility, venues and definition
The most direct consequence of Bosman, of course, was the abolition of the old football transfer system and of the nationality quotas in club competitions. In terms of our analysis for EU sports policy, the main consequence of Bosman was that sport was propelled high up in the EU agenda because the ECJ's ruling forced the Commission and other institutions to take a position on the regulation of sport, which they had failed to do previously. There was a general concern about sport in the EU as reflected, for instance in the 1994 European Parliament Report by Jessica Larive (European Parliament, 1994), but it was not very intense; indeed, it was rather patchy. Before the Bosman explosion sport was decidedly low on the agenda. It could be said to have reached the EU's systemic agenda, as it was an area of limited interest but it had no real institutional agenda status because there were no real efforts to formulate a coherent approach towards sport.

27
Agenda-setting concepts can shed some light on the very different reactions of the EU institutions after Bosman, as compared with the virtual indifference that followed Donà. In a problem centred agenda-setting approach, one can consider sport pre-Bosman as having a low level of visibility, as only few actors involved in policymaking were interested. With such a low visibility level there was no real conflict about the definition of sport within EU institutional venues and it was not yet considered a problematic issue. Thus, most sport decisions were taken on a case-by-case basis at the level of low politics. At this point, the definition or framing of the issue 'sport' was done rather in economic, technocratic and regulatory terms. In short, sport was not a hot and politically salient issue. This changed with Bosman. First, the visibility of the issue was enhanced. The popularity of the affected sport (football) multiplied the media coverage. Moreover, football in 1995 was a completely different sport to the one in the years of Donà. Professional sport in general, and football in particular, benefited massively from the large influx of money from digital television channels (García, 2006a). The commercialisation of professional sport contributed to the Commission's early framing of the issue in economic terms. Thus, with Bosman sport gained in visibility as it was placed in the institutional agenda because the ECJ's ruling forced the Commission and other institutions to take a position on the regulation of sport by the EU.

28
Once on the institutional agenda, the change in the definition of sport and sports policy took a while. The actions of the Commission in the aftermath of Bosman were logically focused on the regulation of sport as a market place because the ECJ had stated in Walrave, Donà and Bosman that European law was only concerned with the economic side of sport. Thus, the actions of the Commission were led by the DG Competition Policy and DG Internal Market. Indeed, Bosman enhanced the definition of sport within the EU as just an economic activity in the eyes of these Directorates General within the Commission, but this was about to change. Shortly after Bosman the issue of 'sport in the EU' became politicised with the interventions of the European Council. Moreover, external actors such as sports governing bodies tried to change the ongoing definition of sport in the EU agenda as an economic activity. It was time for high politics and new policy arenas if the policy was going to be redefined.

THE 'HIGH POLITICS' OF SPORT IN THE EU: SPECIFICITY AND AUTONOMY
Despite earlier warnings in Walrave and Donà, Bosman came as a real shock for sports governing bodies, not least for the football authorities who abruptly realised the far reaching consequences that European law could have for their activities (Interview, Jonathan Hill, Head UEFA Brussels Office, 17 March 2006). It is difficult to understand how this was possible given the clear case-law of the court. Be that as it may, sports governing bodies felt the need to engage with the EU to redress the situation created after Bosman and the high-profile Commission investigations. Sports federations' main goal at that moment was to reduce the regulatory activity of the Commission (Interview, Christophe De Kepper, IOC's Chief of Staff, Lausanne 16 February 2007); they wanted to see introduced into the Treaty the necessary provisions that would allow for a softer application of 30 European law to sport, or even for a complete exemption from it (for details on the demands see for example UEFA, 2001).
The arguments of the sporting movement were structured around two concepts: specificity and autonomy (Interview, Christophe De Kepper, Lausanne, 16 February 2007). Specificity is a topical, but rather elusive, concept that is referred to constantly by sports governing bodies. However, sports organisations are still to produce a clear definition of what the specificity of sport is. Very broadly speaking, the so-called specificity of sport can be understood as the inherent characteristics of sport, both as a social and economic activity, which can justify a tailored application of EU law and policies. The most common example is that of the necessity of balanced competitions, as recognised by the ECJ in Bosman (paragraph 103). So that, rules aimed at maintaining that balance should be deemed to be compatible with EU law. On the other hand, the autonomy of sport has been better defined by the sporting movement. It is the idea that sport, as a civil society movement that emerged on the margins of public authorities' regulation, should remain self-governed by the structures and bodies that have done so over the years (organisations such as FIFA in football, FIBA in basketball, the International Olympic Committee and so on), with minimal intrusion by public authorities.

31
Sport authorities launched an intensive lobbying campaign at the highest level during the 1997 IGC, ultimately convincing the German Chancellor and the Belgian Prime Minister of the necessity of amending the Treaty to protect sport (Husting, 2004, p. 517;Van Miert, 2000). The European Parliament also suggested the inclusion of an Article on sport in the Treaty of Amsterdam, but with a much more justified case for it and excluding categorically the possibility of a legal exemption for sport (European Parliament, 1997). The European Council's intervention took the form of the rather slim Declaration Number 29 on Sport, attached to the final act of the IGC reforming the Maastricht Treaty (hereinafter the Amsterdam Declaration on Sport).

ALLIES WITHIN THE COMMISSION: THE SPORTS UNIT
Shortly after the Amsterdam Declaration on Sport, the Sports Unit within the Commission's Education and Culture DG emerged as a key actor to find the equilibrium between the commercial side of sport and a better attention to the specificity of sport (Parrish, 2003b, pp. 178-179). The Sports Unit initiated a process of dialogue and consultation with the sports world. As a result, the European Council decided to invite the Commission 'to submit a report to the Helsinki European Council with a view to safeguarding current sports structures and maintaining the social function of sport within the Community framework' (European Council, 1998).

33
In response to the European Council's request, the Commission adopted the so-called Helsinki Report on Sport in December 1999. The report acknowledges that the commercial development of sport in Europe 'risks weakening its educational and social function' because 'new phenomena are calling into question the ethics of sport to the detriment of a more balanced development of sport' (European Commission, 1999c, p. 3). The Helsinki Report on Sport presents a general view of the suitable European Sports Policy. The key concept in the report is 'partnership'. The Commission calls for a new approach to sport both at European and national level, with greater consultation between stakeholders, both institutional and non institutional. Sports federations are recognised for their vital role in the governance of sport, but they are reminded of their responsibility to ensure the solidarity and democracy within the sports community (European Commission, 1999c).

34
Straight after the adoption of the Helsinki Report on Sport the European Council intervened again, backing the Commission's vision as stated in the Helsinki Report. Under French presidency, the European Council that adopted the Treaty of Nice in 2000 included the Nice Declaration on Sport in the presidency conclusions of the summit. The aims of the declaration are perfectly summarised by its title: 'Declaration on the specific characteristics of sport and its social function in Europe, of which account should be taken in implementing common policies' (European Council, 2000). The Nice Declaration is longer than the one adopted in Amsterdam and it calls for EU institutions to take into account sports organisations' opinions when formulating their policy on sport. Yet, it is again a non-binding political declaration. It is interesting to note that the Helsinki Report on Sport was drafted by DG Education and Culture within the Commission. The vision of sport in this document differs from the earlier regulatory impetus of DG Competition. The nature of the Commission as a single monolithic institution with a common vision of policies is certainly challenged in the case of sport. It is outside the scope of this article to detail the internal dynamics of policy-making within the Commission, but it is worthy of note as an issue for further research. It is even more interesting to note how the European Council intervened to change the balance in favour of DG Education and Culture.

35
After the Nice Declaration on Sport, the dialogue between sport and the European institutions intensified, trying to find a way to build upon the momentum. The sporting movement regarded the so-called European constitution as possibly the last chance to get political recognition for sport in the Treaty (Parrish, 2003c). Sport was initially introduced in the Convention's first draft as part of a generic article on Youth, Education and Vocational Training. During the IGC, under Italian presidency, the Convention's proposal was amended in the form of Article III-282 <http://ec.europa.eu/sport/action_sports/article/docs/articlesport-final.pdf>. It was an article for sport on its own, giving it political recognition, but it did not award the degree of independence the governing bodies were demanding.

A new definition for EU sports policy: The impact of high politics in the agenda
The Treaty Establishing a Constitution for Europe has never come into force due to the negative referenda in France and the Netherlands that stopped the ratification process. However, the whole debate on the constitutionalisation of sport in the EU with the interventions of the European Council and the Commission's Sports Unit had an effect on the framing of the issue 'sport', thus redefining sports policy. Indeed, the arguments put forward on the specificity and the autonomy of sport (as recognised in the Nice Declaration on Sport) informed the ECJ and the Commission's decisions in particular cases. The Court recognised in Christelle Deliège v. Ligue Francophone de Judo et Disciplines Associées ASBL (Joined cases C-51/96 and C-191/97 [2000] ECR I-2549, hereinafter Deliège) that the rules for the selection of sportspersons to take part in international competitions are inherent to sport, hence they are compatible with EU law as long as they remain not discriminatory and proportionate despite being prima facie an obstacle to the freedom to provide services. In the ruling the Court made reference to the Amsterdam Declaration on Sport as a reinforcement of the recognition of sport'? s specific features (Deliège, paragraph 42). Similarly, in Jyri Lehtonen and Castors Canada Dry Namur-Braine ASBL v. Fédération Royale Belge des Sociétés de Basket-Ball ASBL (case C-176/96 [2000] ECR I-2681, hereinafter Lehtonen) the Court considered that the limitation of players' transfers to short periods during team competitions (the so-called transfer windows) can be accepted as a means to preserve the integrity of championships, despite being an obstacle to the freedom of movement. The Commission also took account of the new arguments when settling investigations such as the one on the central marketing of the UEFA Champions League TV rights (European Commission, 2002b), where it accepted that central marketing is a valid tool to redistribute TV income among the participating clubs, thus ensuring a balanced competition. Similarly, the Commission cleared UEFA's rule on the ownership of clubs participating in European clubs' competitions (European Commission, 1999a).

37
The period that extends from the aftermath of Bosman in 1996 to the adoption of the Constitution in 2004 saw a rise in the importance of sport in the EU agenda due to the politicisation of the issue caused by a focusing event (Bosman). During the rise to the top of the institutional agenda, there was also a change in the direction of sports policy as a result of the redefinition of 'sport' as a problem in the EU's political agenda. In agendasetting terms, this is a period focused on policy framing/definition in the formal institutional agenda, whereas the years that culminated in Bosman can be understood as a period of issue recognition and entrance in the systemic agenda. Sport, which was seen before as an economic industry, was re-defined as a socio-cultural activity with important benefits for European society as a whole. Some European institutions now see sport as a tool for policies in areas such as health (fight against obesity) or social inclusion (fight against racism) (Interview, DG Education and Culture Official, Brussels 12 May 2006). Two groups of actors were instrumental in this redefinition. On the one hand the European Council and DG Education and Culture, as insiders to the political system and the policy community. On the other hand, sports governing bodies, which started as outsiders to the policy community but gained importance over time.

38
Agenda-setting explains this redefinition of issues and policies in terms of actors exploiting new policy venues. The shift of policy arenas is one of the most powerful instruments at hand to change policy status-quo (Baumgartner and Jones, 1993). Once sport emerged on the institutional agenda after Bosman, DG Competition in the Commission was the first to act. Sport was still being dealt in the low politics route. However, sports federations (and also the European Parliament) hoped for an alternative to be considered. Unhappy with the regulatory efforts of the Court and the Commission, the sporting movement proposed a new definition of sport to the EU. They found allies within the EU's institutional framework in the form of the European Council. The interventions of the European Council changed the low politics route in which sport was located at that time, proposing a different approach to sports policy to take more into account of the specificity and autonomy of sport. This is an example of the difficulties faced in moving an issue up the agenda. The intervention of the European Council modified the course of sports policy that had arisen through the low politics route. It was then left to the Commission's Sports Unit (with input from the European Parliament and the sports organisations) to give particular content to the redefinition started by the political leaders of the Council. The negotiations culminating in Article III-282 could be considered as the point in which both routes (high and low politics) and both policy definitions (regulation v. politicisation) met.

A LOOK TO THE FUTURE: THE QUEST FOR GOOD GOVERNANCE
Despite the failure to ratify the European Constitution, sport has remained an active item in the European agenda. The majority of the actors involved in EU sports policy agree that the key issue in the last two years has been the role of European institutions in the governance of sport and the autonomy of sport governing bodies (interviews, passim). Whilst the specificity of sport could be considered as addressed by the decisions reviewed in the section above (even if sport governing bodies are unhappy with the final result of this settlement), the exact degree of self-regulation that should/could be granted to sport was (and probably is) still to be decided. This section deals with the latest developments in sports policy, namely the Independent European Sport Review (Arnaut, 2006) The Independent European Sport Review (IESR) is an initiative of British sports minister, Richard Caborn (for a detailed analysis of the IESR see García, 2006b). The IESR was initially as a report on European football, but it has been later transformed in a blue print for the governance of professional sport in the 21 st century. The recent corruption scandals surrounding European football in the last year, such as match fixing in Italy, prompted European ministers to find a solution to the consequences of massive commercialisation that professional sport in Europe has suffered in the last decade (García, 2006a). One of the particularities of the IESR is that despite its claims for independence it has been 'commissioned by UEFA' (Independent European Sports Review, 2006). The final report of the IESR concludes that professional sport in Europe 'is not in good health' (Arnaut, 2006, p. 13) and it goes on to propose a series of recommendations to improve the governance of sport.

41
The IESR defends the so-called European Model of Sport (European Commission, 1998). The Review asks for EU institutions to empower sport governing bodies as their natural interlocutors in the governance of sport. Indeed, the report proposes a series of legal measures by which the EU could reinforce the role of federations in the governance of sport, giving them what is called ' legal certainty' to perform their role as governing bodies. We can see the IESR as an attempt to reinforce the on-going definition of sports policy, but giving an even more important role to the autonomy of sport federations in relation to public authorities and also in relation to other stakeholders such as clubs and leagues: UEFA has played an important role in the initiation and development of the IESR by framing the corruption scandals around football as problems of governance related to the mismanagement of money at club level and has also benefited from its experience in dealing with EU-matters and its status now as an insider closer to the sports policy-making core in Brussels. Indeed, UEFA senior officials dedicate now a good portion of their time to traveling regularly to Brussels (Interview, UEFA top senior official, 8 February 2007). Focusing events, such as the match-fixing allegations in Italian football, helped European football's governing body to put forward its particular agenda. Again, the choice of venue was also important, as the political pressure of national governments (high politics) helped to raise the profile of the IESR and it is hoped that it can influence the next steps of the Commission regarding sports policy. For instance, the Review could feature heavily in the upcoming European Commission White Paper on Sport (Financial Times, 2006).

42
However, the possibility of a sudden change remains. The Charleroi case, which is now awaiting a preliminary ruling of the ECJ, is a good reminder of this possibility. In this case, Belgian first division club Sporting Charleroi has started proceedings against FIFA's rule on the release of players for national team duty, which is considered as an abuse of dominant position by world football's governing body. Sporting Charleroi has now been joined in the proceedings by the G-14, the grouping of 18 of the most powerful professional football clubs in Europe. The argument is that clubs have to release their players for national team competitions (such as the World Cup) without being entitled to any type of compensation even if they come back to their club injured (Mateo, 2005). The clubs presented their case before Charleroi's Commercial Court (Martínez de Rituerto, 2005), who has referred it to the ECJ for a preliminary ruling under Art. 234 TEC. The ECJ has been asked whether the analysed FIFA rules 'are contrary to Arts. 81 and 82 TEC or to any other Treaty dispositions, especially Arts. 39 and 49 TEC' (OJ C 212/2006, 2 September, p. 11).

43
As the ECJ's decision is pending, it is not possible to elaborate on the consequences of this case for EU sports policy. But should the decision go in the clubs' way, it could have an important impact on the power structures within football (and sport in general) as clubs and leagues would have an important weapon against governing bodies. Indeed, it is interesting to see that for both clubs and leagues, the problem of governance is not about the management of money at club level, but rather an issue of democracy and representation in the decisionmaking structures of the governing bodies. The Charleroi case is a timely reminder that sports policy could change again through the intervention of the European Court, especially if we take this case in consideration with the recent ruling of the ECJ in Case C519/04P

44
It is outside the scope of this article to enter in the legal assessment of the consequences of Meca-Medina for the application of EU law to sport (for that, see Weatherill, 2006). It is also too early to see what the consequences in political terms can be. However, there is an interesting point in terms of agenda-setting if Meca-Medina and CharleroiBosman had in 1995, even if the judgments go against governing bodies, because the general political discourse about sport in the EU is totally different. (regardless of the outcome) are considered together. Both cases present a familiar pattern: the challenge to rules adopted by sports governing bodies, as Jean-Marc Bosman did more than a decade ago. In the current debate on the governance of sport and the independence of federations for self-regulation, the recourse to the ECJ can be seen as another change of venue in order to shift the current definition of sports policy. The current policy definition is focused on the socio-cultural values of sport, which tends to benefit governing bodies over professional clubs, leagues and athletes, more keen on the economic side of professional sport. However, it is difficult to say if the new recourses to the ECJ can have the same impact.
In practical terms, Meca-Medina casts a shadow on the IESR's legal analysis of the role of governing bodies in the governance of sport (García, 2006b). Charleroi is a reminder to sports federations that their desire of total independence from public authorities is difficult to achieve. Sports governing bodies would probably do better to accept the reality of a degree of independence similar to the concept of supervised autonomy (Foster, 2000, p. 57). If sports federations want to remain able to influence the EU agenda on sport, they need to be considered as insiders to the policy community, thus they will do better if they engage with the EU and are careful to respect European law, not make the same mistakes that led to Bosman.

CONCLUSION
European institutions did not get involved in sport of their own volition, but as a result of the powers conferred to the ECJ and the Commission to control the single market. In this respect, one could argue that the application of European law to sport can be easily understood in terms of neo-functionalism and the natural spill-over of European economic integration reaching a commercially developing area of activity such as sport (Haas, 1968;Lindberg, 1963). Indeed, this is a plausible explanation if we focus on the initial regulatory actions of both the ECJ and the European Commission. However, it is more difficult to understand the hesitancy of the Commission in 1976 after Donà and the absence in the case of sport of a political follow-up to the drive of economic integration. In the case of sport regulation, national governments saw the Court's interventions as unintended and undesirable, thus 'the connection between the functional logic and the political strategy did not happen' (Barani, 2005, p. 53).

47
Agenda-setting can provide an alternative explanation to why some sectors do complete the logic of spill-over and others do not. It can also help to understand the role of external actors to the integration process, such as the role of the sports federations. A problem centred model of agenda-setting can explain this first stage of the agenda process in EU sports policy. It can be argued that it was the commercial development of sport and the internal fights between employers (clubs and federations) and employees (players) that mainly forced sport into the systemic agenda of the EU through the cases before the ECJ. However, sport only reached the EU's institutional agenda with the ECJ ruling in Bosman. In the case of sport, visibility was initially very low and the salience of the issue was minimal in the EU's agenda. The change in sports policy after Bosman illustrates the two different routes through which agenda-setting can operate in the EU: the route of high politics or the bottom-up route of low politics (Princen andRhinard, 2006, p. 1122). It also illustrates the connections between policy redefinition and policy venue. The interventions of the European Council facilitated a change in sports policy from a regulatory approach to a more political and socio-cultural point of view that takes into account the specificity and the autonomy of sport.