Moreover, there is little to suggest that abolition of the rules on foreign players might lead to players possessing the nationality of the relevant State becoming a small minority in a league. (Bosman, Lenz A-G, para 146)

England v The Rest of the World. Statistics collated from yesterday’s Premier League games…. Players: England 61; The World 115. (Johnson, 2007)

The world of sports law had been awaiting the result of SA Sporting du Pays de Charleroi v Federation Internationale de Football Association (FIFA) (Case C-243/06, (‘Charleroi’)) on whether clubs could be forced to release players for international duty without compensation. The application of competition law to football was avoided by the full court in Union Royales Belge des Societes de Football ASBL v Jean-Marc Bosman (Case C-415/93 (1996) CMLR 645 (‘ Bosman’), but advised on by the Advocate General. The question of how far sporting interest justifies restricting ordinary commercial freedoms would be squarely before the Court of Justice of the European Communities. And then, as part of a settlement including the disbanding of the G-14, the Charleroi case disappeared (BBC Online 2008a). Much will be written, not least in the sports pages, as to what this means for the future of football. From a legal perspective it means that we are left where we were when the last major European Commission activity drew to a close in 2002. We must guess as to whether recent compromises regarding the collective selling of television rights, transfer fees and player contractual freedom will withstand the scrutiny of the ECJ when the time comes. We can speculate from the settlement of Charleroi as to what the parties thought of their chances of success before the ECJ, but such speculation is of little or no value. We remain in doubt and there are still ‘rich pickings for lawyers’ (Wetherill, 2003, p. 92-93).

The settlement of Charleroi also means that when we look for guidance as to how European law applies the proportionality test to whether the interests of sport justify restrictions on Community Law freedoms, the principal authority remains the analysis of Advocate General Lenz in Bosman. There are many examples of this case’s dominance in academic literature. One prolific writer on European sports law has at great length endorsed the Bosman view on ‘mutual interdependence’ (Weatherill, 2003, pp. 52-54). Van Bogaert thought it remarkable that the German Handball Federation should use arguments as to sporting interest (for example, the link between club and country, player development) that had been refuted by the Advocate General in Bosman (Van Bogaert, 2004, pp 273-274). Similarly, Miettinen and Parrish note that FIFA’s proposals to re-introduce a form of national quota are ‘ill conceived given…the unequivocal refusal of the Court to accept this type of agreement in Bosman’ (Miettinen and Parrish, 2007, para 26), a view that has been borne out by the European Commission’s rejection of the proposals (BBC Online 2008b). For Boyes, exploring the legality of player restrictions by the English Cricket Board, the question is always whether the specific peculiarities of cricket allow Bosman to be distinguished (Boyes, 2005).

It will be argued that the Bosman analysis of the needs and interests of football has been proved wrong in most material respects. The continued application of Bosman’s general principles and particular assertions must give way to a more realistic analysis of professional sport, particularly football, and a better appreciation of the limits of adjudication when applying the proportionality test to sport. Applying Fuller’s celebrated work on the role of polycentricity in creating natural limits to what can be determined by adjudication (Fuller, 1978, 393-405), it will be argued that the ECJ can beat a general retreat without weakening the application of those principles at the heart of EU law freedoms.

This article provides very much an English perspective on the Bosman case, although supported by consideration of the European elite’s relationship with the Champions League. The basic effect of Bosman has been to increase the free movement of players between Member States and to either reduce or remove the fees payable on transfers. Obviously, this leaves significant room for differences in the effect of the decision and the responses of national associations: some nations have become major exporters of player talent, whereas, as we shall see, English football is strongly and almost exclusively in the import business. Nevertheless, it is suggested that it must be a serious criticism of the Bosman analysis that it runs almost entirely counter to the experience of one of Europe’s most prestigious leagues. It also must be a particularly serious criticism that Europe’s richest league fails to conform to the Bosman analysis as to what might be respected in terms of the redistribution of resources.

The sporting exception

This article does not review in detail the so-called ‘sporting exception’ to Community Law, but instead addresses the application of the proportionality test where sport falls within the EU’s jurisdiction. This concept has its origin in Walrave v Association Union Cycliste Internationale (1974) ECR 1405 (‘Walrave’), which held that ‘the practice of sport is subject to Community law only in so far as it constitutes an economic activity’ (paragraph 4), and that Community law has nothing to say on matters of ‘purely sporting interest’, (paragraph 8).

The distinction made in Walrave is a difficult one. IbañezColomo (2006, para 3-7) highlighted the starkly different fates in Walrave and Bosman of the nationality rules in World Cycling Championships and in European club football competitions. Some argue that the ECJ’s assertion of jurisdiction to undertake a proportionality review in respect of anti-doping rules in Meca-Medina v Commission (Case C-519/04) (‘Meca-Medina’) means that there will be few rules of sporting interest that fall outside the European jurisdiction (Miettinen and Parrish, 2007, at para 6). But this conclusion overlooks that the ECJ (para 45) expressly noted that banning performance enhancing drugs was inherent to sporting competition. At issue in Meca-Medina was whether it was proportionate to conclude that an athlete had taken performance enhancing drugs on the basis that he has 2ng/ml of Nandrolone in his urine. This is a scientific matter one step removed from the inherently sporting purpose behind drugs testing procedures – and given that the athlete’s right to work throughout the EU was in issue, it should not be a surprise the ECJ held that such procedures must be proportionate (para 55).

It is perhaps wrong to read into Meca-Medina a fully theorised approach to sports law, a subject of only occasional interest to the ECJ. The sports law jurisprudence suffers perhaps from having a limited number of cases to analyse. Although space prevents considering the point in detail, it is worth noting that there are other areas outside the scope of Community, where national competence must nevertheless be exercised ‘consistently with Community law’. Taxation is an obvious example, with fine distinctions being made as to where a rule is held to be inoffensive not withstanding that it causes disadvantage to intra-Community money-flows, and where it is held to be a breach of EU law, subject to proportionality (see, for example, ACT Class IV (C-374/04) and Franked Investment Income (Case C-446/04)). The ‘sporting exception’ may not be exceptional after all, but an application of general principles inherent in the limited nature of the EU’s jurisdiction. That, however, would demand an article to itself. This article assumes a breach of Community law freedoms, and asks whether this breach is ‘ proportionate’. As such, sporting arguments are pleaded by way of justification not as an assertion of immunity to jurisdiction, even if the practical difference may be narrow at times.

Proportionality and the hidden judicial regulator

The application of the proportionality test determines which infringements of EU law freedoms are or are not justified. Stated simply, a measure is proportionate if it is the ‘ least drastic means’ for achieving a legitimate end. This can be deduced from the very idea of justification: an action is not justified if: a) it does not further a legitimate aim; b) it does not achieve that aim; or, c) the aim could be achieved with a less drastic effect on rights. Added to this there is also ‘the need to balance the interests of society with those of individuals and groups’ as a further essential (if more nebulous) part to proportionality, see Huang v Secretary of State for the Home Department (2007) UKHL 11 at para 19 and R v Oakes (1986) 1 SCR 103, 139.

In theory, the test should give rise to a single answer: the ‘least drastic means’ for achieving the ends (Nicol, 2006, p. 734). In reality, whether it be the US or Canadian Supreme Courts, the European Court of Human Rights (ECtHR), the ECJ or our own judiciary, judges pull back from driving the test to that conclusion. They create instead what for the ECtHR is the ‘margin of appreciation’, for British judges is sometimes called the ‘ margin of discretion’ and is well described by some jurists as the ‘ indifference curve’ on which may be found a number of resolutions to the problem which are equally acceptable to the court (Greer, 2004, p. 416). The limits of the ‘indifference curve’ represent the limit of the adjudication process.

In practice, it is the ‘least drastic means’ limb which is most likely to be the determinative part of the proportionality test (Edwards, p. 872). Indeed, in the first 15 years of the Canadian charter on fundamental freedoms, every law that satisfied the ‘least drastic means’ limb passed the proportionality test (Trakman et al, 1998, 100-101). So, the more willing the court is to enquire into and speculate into possible ‘ less drastic means’ alternatives, and the more sceptical the court is as to justifications given from rights-infringements, the narrower the indifference curve. Equally, the less willing the court is to engage with the process, the wider the indifference curve. The same applies, mutatis mutandis, to the extent to which courts are willing to assert value judgements as to ‘ legitimate aims’ or ‘rational connections’, and, although these factors are generally less likely to be determinative than those relating to ‘less drastic means’, they will be highly relevant when considering the ‘quotas’ section of Bosman.

What does this mean for football? The greater the willingness to engage in a detailed proportionality analysis, the more the law acts as a hidden regulator, albeit as an umpire to be appealed to rather than a referee intervening proactively. This is because the extent to which administrators administer, and governing bodies govern, is naturally circumscribed by the view the courts take on what is or is not proportionate, and the various parties’ expectations of how an appeal to the judicial umpire will come out in fact. Football associations, clubs and players all calculate their rights and powers by reference to their estimation of what the law will or will not allow. And that estimation of how the ECJ will apply the proportionality test in any given situation will come partly from applying the high level norms of EU law, but largely from applying the specific approaches, principles and assertions found in the relevant judicial decisions. In short, it will come largely from the Bosman analysis of football’s needs and interests, and from the willingness of the court to put sporting justifications to strict scrutiny.

Given that the rights that EU law positively upholds in the footballing context are commercial and economic freedoms, and that sporting interest arguments will typically be in derogation to those freedoms, one cannot understate the importance of how far the ECJ is willing to involve itself in a proportionality analysis, particularly as regards the availability of ‘less drastic means’. Outside the ‘indifference curve’, economic rights are trumps.

The limits of adjudication

Put at the highest level of generality, the tasks ‘inherently unsuited to adjudication’ are those where it will be impossible ‘to preserve the meaning of the affected party’s participation through proofs and arguments’ (Fuller, 1978, 393). Fuller argued that one reason why adjudication may be inappropriate is that the problem is polycentricity - there is no clear centre to the problem.

Fuller further illustrated the heart of the polycentricity problem with a vivid metaphor (Fuller, 1978, p. 395):

We may visualize this kind of situation by thinking of a spider web. A pull on one strand will distribute tensions after a complicated pattern throughout the web as a whole. Doubling the original pull will, in all likelihood, not simply double each of the resulting tensions but will rather create a different complicated pattern of tensions. This would certainly occur, for example, if the doubled pull caused one or more of the weaker strands to snap. This is a ‘polycentric’ situation because it is ‘many centred’ – each crossing of strands is a distinct centre for distributing tensions.

However, this spider’s web is not re-created in the court environment. The court has a limited time to hear limited evidence from a limited number of parties. It must then produce a reasoned decision in which, ideally, the court’s chosen outcome follows inexorably from applying objective legal rules to facts – as Julius Stone describes (Stone, 1964, p. 200):

In hypothetical legal norms the antecedent term (IF A) refers to certain conditions, which become conditions for the application of the legal norm, and can therefore be briefly called ‘legal consequences’. The consequent terms (THEN B) specify what ought to be done or not done, or what may be done or may not be done, if those legal conditions arise, they designate the ‘legal consequences’

Fuller argued that complex problems frequently do not fit readily into the forms and conventions of judicial adjudication, and that ‘ instead of accommodating his procedures to the nature of the problem he confronts, (the judge) may reformulate the problem so as to make it amenable to solution through adjudicative procedures.’ Thus adjudication ‘solves a polycentric problem only by grossly simplifying and distorting it’ (Fuller, pp. 404 and 399). By this, we take to mean that the court seeks to encapsulate the case with a clear ‘IF A… THEN B’ when complexity does not allow such a rule. In the context of football, a Court might commit this error by creating a guiding principle that the need for competition creates ‘ mutual economic dependence’ and thus ‘each club has an interest in the health of the other clubs’ (Bosman, A-G, para 227), when, as will be seen, the reality is different. Thus, Fuller’s point about simplification and distortion within litigation applies when such nuances are overlooked as the Court seeks a reasoned decision in not too many pages. The problem is not one of law; it applies to all systematisation of human life. Unfortunately, the dynamic undermines the economic studies that provide the best academic analysis of the footballing economy. For example, when Hoehn and Syzmanski refer to clubs as ‘producers of substitutable brand identities’ it is hard to know if they are describing football as any non-economist would understand it (Hoehn and Syzmanski, 1999, p. 213).

Polycentric problems can be highly contained. For example, if two galleries had to divide a collection of paintings we have a problem where adjudication must flounder, even if mediation might prove fairly straightforward (Fuller, 1978, 394):

(T)he disposition of any single painting has implications for the proper disposition of every other painting. If it gets the Renoir, the Gallery may be less eager for the Cezanne but all the more eager for the Bellows, etc. If the proper apportionment were set for argument, there would be no clear issue to which either side could direct its proofs and contentions.

Hence, the problem of polycentricity can apply to the permanent allocation of finite objects amongst two parties. But more often polycentricity is found in the unpredictable repercussion of an ostensibly private dispute (Fuller, 1978, 394-395):

The point that comes first to mind is that courts move too slowly to keep up with a rapidly changing economic scene. The more fundamental point is that the forms of adjudication cannot encompass and take into account the complex repercussions that may result from any change in prices or wages. A rise in the price of aluminium may affect in varying degrees the demand for, and therefore the proper price of, thirty kinds of steel, twenty kinds of plastics, an infinitude of woods, other metals, etc. Each of these separate effects may have its own complex repercussions in the economy.

A court may thus adjudicate on a snapshot picture, when the world is changing rapidly; or it may analyse a problem as a matter of cause and effect within a closed system, whilst harming the interests of those beyond its notice. Closely related to this is that ‘a decision may act as a precedent, often an awkward one, in some situation not foreseen by the arbiter’ (Fuller, 1978, 397). Even if all affected parties have a fair hearing, it is one thing to do justice in the instant case, another thing to lay down the law as to how their future interactions should be governed.

An interesting part of Fuller’s polycentricity analysis is that of incommensurability: ‘you have to judge pumpkins against pumpkins, not pumpkins against cucumbers, especially when there are some relevant cucumbers not entered in the show’ (Fuller, 1978, 403). To take Walrave, how can a court analyse whether it is proportionate to organise a world championship on national lines? It must weigh the consequences of reorganisation (‘ pumpkins’) against the effect on the labour market in the sport (‘ cucumbers’). It is perhaps no surprise that, when faced with realistic sporting arguments wholly outside EU law’s normal frames of reference, that it avails itself of the ‘sporting exception’ to avoid comparing incomparable considerations. It is necessary to stress ‘realistic’, for reasons that will come clear when we come to the first half of Bosman: a vague or speculative assertion of sorting interest will not suffice.

Fuller applied to proportionality

Fuller’s analysis concentrated on what governments and legislatures should or should not commit to an adjudicative process. It is not obvious why the analysis should apply where a law or treaty has expressly or implicitly entrusted a court with a task. A court cannot refuse to adjudicate a claim on the basis that the facts are too complicated. Similarly, if the law requires the court to decide whether a rights infringement is justified, ‘modest estimates of competence’ cannot mean rights are repealed in certain contexts (see A v Secretary of State for the Home Department (2004) UKHL 56 at para 39). As one American jurist noted, abdication of judicial responsibility is the greatest activism of all (Frantz, 1963, pp 741-2).

Nevertheless, Fuller’s analysis of polycentricity is commonly stressed in literature on proportionality to explain why the courts should leave a margin of discretion to those who carry out the ordinary legislative, executive or administrative functions (Allan, 2006, p. 693; Rivers, 2006, pp. 175-176). Of course, the fact that a social or economic problem is polycentric does not mean that the resulting adjudication is: a decision to raise or lower a tax is polycentric given the unpredictable economic ramifications, but these are of no importance to the court when applying the taxing statute. However, the polycentricity problem begins if the court is supposed to take a broader view. If a court is required to consider if a measure is proportionate or justified, there is no end to the facts that need to be considered, nor can it confine itself to the immediate or short term interest. Ideally it will consider everything apart from the wholly fanciful. As will be seen, such enquiries may strain the limits of adjudication to breaking point.

The limits of proportionality are where the ‘ indifference curve’ begins, and these should coincide with the limits of adjudicating the problem set by the law concerned. The court is indifferent because it cannot with integrity adjudicate any further. The relevance of Fuller’s theory is obvious, although it lacked a theory as to how a court should adjudicate on its own limits. We shall come to these theories later, but first we must see how far the ECJ overstepped those limits in Bosman, and how, using the English experience as the principal example, this led it to the wrong conclusion about football.

Bosman and Quotas

Bosman was, of course, a case of two halves. The first half related to rules restricting the number of foreign players eligible to play in European competitions. Unsurprisingly, this was a prima facie breach of free movement of workers under Article 39. The issue then turned to proportionality. The Advocate General made the assertion with which this article starts: that there was no prospect of home players becoming a small minority in national leagues. A further contention by Advocate General Lenz provides an interesting example to illustrate Fuller’s spider web analogy (Bosman, Lenz A-G, para 146):

The argument that the rules on foreign players are needed to ensure that enough players develop for the national team is also unconvincing…Winning the World Cup, for instance, generally brings about increased interest of spectators in national league matches as well. It is therefore in a country’s clubs’ very own interests to contribute to the success of the national team by developing suitable players and making them available…Moreover, the national teams of the Member States of the Community nowadays very often include players who carry on their profession abroad, without that causing particular disadvantages…In the German national team which became world champions in 1990 there were several players who played in foreign leagues. It is therefore not evident that the rules on foreigners are necessary in order to ensure the strength of the national team.

Remove the strand of national quotas, and the strand representing the strength of national teams should be unaffected. The Advocate General assumed that ending national quotas would simply redistribute the talent around the national leagues. Yet, remove the strand of national quotas and it is far from clear what will happen to the strands representing the development of talent in leagues that have the most money to import ready made talent. Not only does the English Premier League now have a clear minority of English players, but no English players currently play in the top flight of Europe’s other leading leagues. Thirteen years after Bosman, the picture is not as the Advocate General anticipated.

Many of the stated reasons for the finding of disproportionality are vulnerable to criticism. There are good reasons to suppose that English football at least is trapped in something of a ‘perfect storm’ where player development is concerned, both in terms of it being more economically efficient to buy in talent than develop it at home, and comparatively moderate English players find it more lucrative to be substitutes in Bolton or Sunderland than to be regular starters in top continental leagues.

In fact, it is difficult to see why the Advocate General and the Court (at paras 131-135) engaged in such speculation. The better reason for rejecting the arguments is more direct: when one looks to the basic normative structure of the Treaty of Rome, it is hard to see how the arguments qualified as a ‘ legitimate aim.’ In acceding to the Treaties, the Member States subscribe to the idea that, as between themselves at least, their interests are best served by free trade, as represented by the four fundamental freedoms of the Treaty of Rome; free-movement of goods, capital, services and labour. This means that, even if some sectors of a national economy wilt under competition, the net result will be positive. The impact on the relevant national workforce may be a highly polycentric issue defying rational analysis in an adjudicative process, but Article 39 itself contains the judgment that such factors shall be considered irrelevant. A failure of English footballers to find high level employment means simply that in this area British workers are being outperformed. To structure the point in terms of proportionality, it is never a ‘legitimate aim’ for a Member State to protect its own workers from the consequences of having to compete with workers from other Member States. Talking in terms of ‘developing talent’ or ‘strength of national team’ merely disguises ways of complaining about the same failure to compete and should not have been entertained.

Similarly, the arguments as to the need for club teams in European competition to be representative of their home nation were a denial of ‘legitimate aim’ – or rather the assertions underlying that aim. UEFA was seeking to uphold a connection between club and nation which was hard to make good. Nations compete in World Cups, but it is clubs that contest the Champions League. The Advocate General was very robust on this point (Bosman, Advocate General, para 144). In any case, if we switch to the fourth part of the proportionality test (‘the need to balance the interests of society with those of individuals and groups’) arguments that internationalising teams would undermine European club competitions (for example, by eroding spectator interest) could not justify direct discrimination until such time as there was real evidence of such a dynamic. A discriminatory rule requires strict scrutiny, something we shall deal with more specifically later, and that means a higher level of justification than speculation as to what may happen in the future.

It can be seen that by going beyond the limits of adjudication and entering the realm of speculation, the Court based this part of Bosman on a flawed proportionality analysis. As such, Miettinen and Parrish (2007) and Van Bogaert (2004) should be less surprised that FIFA has revisited this ground. By straying from solid propositions that can be rooted in principle to factual speculation beyond the limits of adjudication, the ECJ itself created grounds for counterarguments. It is also worth noting that in the very recent and brief judgment in Kahveci v Real Federación Española de Fútbol (Case C-152/08), the ECJ did not consider proportionality when ruling that an EU-Turkish agreement on free-movement of labour meant that quotas for non-EU players could not apply to Turkish citizens. Quotas offended the basic norm of the EU-Turkish agreement, so justifications were irrelevant. One might wonder about the effect of this rule should it apply to the EU as against Africa and South America, but the ECJ is taking a more legally coherent approach by not speculating on such matters.

It is thus not my intention to revisit this half of Bosman in detail. The issue involved, along with recent attempts by UEFA and FIFA to introduce a watered down national quota, have been considered at length in Miettinen and Parrish (2007). My focus is instead on the decision to condemn the transfer fee system in respect of out of contract players – a decision whose effect has been greatly extended following the European Commission’s competition law investigation (Weatherill, 2003, pp. 66-69).

Bosman and Transfer fees

Bosman’s complaint was that a move from RC Liege of Belgium to US Dunkerque of France was prevented because of a dispute as to the transfer fee. As US Dunkerque would not have paid a fee for a French player (Bosman, Lenz A-G, para 32), the case could have been dealt with on the grounds of non-discrimination. The Advocate General did indeed touch briefly on how variations in transfer fee rules might lead to distortions in the single European market. However, for better or for worse, the case was resolved by considering the legality of transfer rules where there was no hint of nationality based discrimination. With discrimination removed from the picture, it is far from clear why the ECJ held that Article 39 on free movement of labour applied where nationality and crossing the border was irrelevant to the operation of the restriction. The principal definition of the free movement of labour right, in Article 39.2, talks expressly in terms of the ‘abolition of any discrimination based on nationality’. Even if Article 39 were not expressly rooted in the concept of abolishing discrimination, the closest authority (Keck) suggested that EU law would be irrelevant in a case such as Bosman. And the Advocate General largely distinguished Keck and Mithouard (Cases C-267/91 and C-268/91, (1993) ECR 1-6097) on the basis of Alpine Investments BV v Minister van Financien (Case C-384/93, (1995) ECR-0000), which greatly differed from Bosman in that the crossing of a border was key to the disruption of EU economic freedoms. However, a prima facie violation of Article 39 was held, and the case moved onto a consideration of whether that restriction could be justified. In this it should be noted that the Advocate General took the same approach to whether transfer fees were justified under Article 39 (free movement of labour) or 81 (competition), a point that Weatherill makes when assuring Bosman aficionados that the case remains relevant despite competition law having replaced free movement as the focus of EU law’s involvement in sport (Weatherill, 2003, pp. 84-85).

The justification argument was essentially this. First, it was universally accepted that ‘a professional league can flourish only if there is no too glaring imbalance between the clubs taking part’ (Bosman, Lenz A-G, para 219.) Secondly, at para 223, the Advocate General readily agreed that ‘it is of fundamental importance to share income out between the clubs in a reasonable manner’. Thirdly, recounting figures from Touche Ross, the Advocate General at para 222 noted that, in the 1992/93 season, transfer fees saw a net £13.3m transferred from the Premiership to the Football League. At this point, it is difficult to see how UEFA could lose its justification argument, but we must remember that the most important part of the proportionality test is the question of ‘less drastic means’ (Edwards, 2002, p. 872), which is to say: was there an alternative system, at least as good, for upholding competitive balance which was more compatible with EU law freedoms?

A court can scarcely assert the existence of ‘ less drastic means’ without offering suggestions. The Advocate General made two suggestions at para 226 of his Opinion:

1. A salary cap; and

2. Solidarity payments, particularly by way of redistribution of television revenues.

The salary cap need not concern us: the Advocate General noted that it was more problematic than redistributing television revenues. In this the Advocate General was doubtless correct. Not least of those problems would be the acceptability of the cap to players – the Major League Baseball cap was abandoned after a player strike). It is also notable that the success of a cap is speculative; post-Bosman studies suggesting the cap in US basketball had had no effect on league competitiveness (Dobson and Goddard, 2001, p. 129). Added to this Miettinen and Parrish have noted that a salary cap may itself be contrary to Article 81 of the Treaty as restricting competition (Miettinen and Parrish, 2007, at para 29).

It is thus Lenz’s second suggestion, the redistribution of revenues, on which the case turned. Could the redistribution of revenues (particularly television) represent an adequate alternative system for ensuring redistributing money from the richest clubs to the poorer ones? The Advocate-General could see ‘no insurmountable obstacles’, and the ECJ agreed; Hausman and Leonard noted that an agreement which split revenues and talent efficiently ‘may be impossible as a practical or political matter’ (Hausman and Leonard, 1997, p. 622). History appears to have vindicated Hausman and Leonard.

Wage Inflation – Reality Bites Back

The immediate effect of Bosman was that transfer fees ceased to be payable on out-of-contract players if the player was exercising his Article 39 right of free movement of labour. It had a profound effect on the scenarios that it did not expressly rule upon. The transfer value players in-contract who might move outside their home Member State amortised to zero, depressing the value of in-contract transfers. Similarly depressed were transfer fees on domestic transactions, given that fees could be avoided by hiring a player under Bosman. Bosman was predictably followed by attempts to extend the principles to in-contract players. The attack came from the European Commission in the name of competition law, with the players’ union following up with a direct legal action. The result was a settlement negotiated under the auspices of the Commission which further reduced the circumstances when transfer fees were payable (Weatherill, 2003, pp. 66-69). The ins-and-outs of that agreement are beyond this article, but it is worth noting that it included what is now Article 6 of FIFA’s ‘Regulations on the Status and Transfer of Players’ that restricts transfers to the closed-season and a four week transfer-window during the season (FIFA, 2004/2007). In many ways, these provide greater hurdles to free movement than transfer fees, and the latter favours the largest clubs with the biggest squads (Ley, 2007). Transfer fees continue, but are doubtless lower than had Bosman been decided differently. Also, as Weatherill points out (2003, pp 69-73), the remnants of the transfer fee system remain vulnerable to legal challenge. Miettinen and Parrish (2007, para 33) have noted generally that the ECJ is supportive of sectoral agreements of this sort, but they also note that this probably would not extend to a free movement challenge.

It follows that, in Bosman, the ECJ reduced the significance of transfer fees, and asserted that any resulting loss of cross-subsidy could be made good with something more systematic which would not prejudice free movement of labour such as the redistribution of television revenues. But, as we have seen, the polycentricity critique doubts the ability of Courts to predict the consequences such a decision: a surgical removal of one strand of a spider’s web is a hazardous operation.

Consider the following example from para 224 of the Advocate General’s advice, where he speculates on how the end of transfer fees might in itself bring benefit to competitive balance:

Since the players transferred to the bigger clubs are as a rule the best players of the smaller professional clubs, those clubs are thereby weakened from a sporting point of view. It is admittedly true that as a result of the income from transfers those clubs are placed in a position themselves to engage new players, in so far as their general financial situation permits. As has been seen, however, the transfer fees are generally calculated on the basis of the players' earnings. Since the bigger clubs usually pay higher wages, the smaller clubs will probably hardly ever be in a position themselves to acquire good players from those clubs. In that respect the rules on transfers thus strengthen even further the imbalance which exists in any case between wealthy and less wealthy clubs. The Commission and Mr Bosman correctly drew attention to that consequence.

There is nothing wrong in every chain of the Advocate General’s reasoning – at least on the basis of the evidence before it. Had he known of the English system for post-contractual transfer fees, he would have seen that fees can be set by reference to the pay offer made by the new club, a system highly favourable to small clubs who have found big talent (Speight and Thomas, 1997, p. 205). Accepting that such evidential inadequacies can be attributed to the litigants and not the Court, the flaw in the Advocate General’s reasoning is that a complex economic issue has been reduced to a closed set of binary calculations of cause and effect. In the virtual world of adjudication, no transfer fees meant that small clubs could enter the market for better players. In the real world, if Manchester United can make savings on transfer fees, then it has more money to spend on players, and if players are totally free-agents at the end of their contract, their bargaining power in wage negotiations is increased. In other words, far from entering the market for the top players, smaller clubs were squeezed by wage inflation resulting, in part, from Bosman itself. Dobson and Goddard in their detailed study The Economics of Football note that wage inflation was a ‘predictable consequence’ of the decision (Dobson and Goddard, 2001, pp. 96, 421 and 430). And yet this wage inflation simply did not exist in the Advocate General’s analysis; in Fuller’s language, it was part of ‘the complex repercussions that may result from any change in prices’ (Fuller, 1978, pp. 394-395); or just another unexpected rearrangement of the strands in the web of European football.

To make matters worse, wage inflation has been highest for the best players – a further set-back for the Advocate General’s expectations. Dobson and Goddard (2001, pp. 421-422, and 430) saw this as an obvious consequence:

(I)n the post-1995, post Bosman world of free agency, an out of contract player is completely free to sell his services to the highest bidder. Small wonder the players’ earnings have rocketed as a result. And…small wonder that the earnings of the most talented players have rocketed the most, since these are the players with the greatest monopoly power as sellers of services for which few or no direct substitutes exist.

Supporters of Bosman have tended to dismiss its effect on the transfer fee system as being minor in comparison to the effect of television revenues (Weatherill, 2003, pp 67, 74; Dobson and Goddard, 2001, p. 96). Certainly, we cannot deny the importance of this factor, and other factors such as Champions League revenues and Far Eastern merchandising, have strengthened the position of the elite.

However, the existence of more immediate culprits does not acquit Bosman of making matters significantly worse – both in its effect on transfer fees and wage inflation. The picture is more polycentric. Television revenues were bound to produce both transfer inflation (which would increase the flow of money down from the Premiership and so help competitive balance) as well as wage inflation (which harms competitive balance). Bosman thus created a double blow to competitive balance – it held down transfer fees, freeing up money in the hands of the richest clubs to fuel wage inflation. When Jack Walker started buying Blackburn Rovers to the top, it could be said with considerable truth that he was a benefactor to the whole English game: his transfer fees circulated downwards throughout football. Because of Bosman, Abramovitch’s millions flow rapidly into £100,000+ per week wage packets.

The Bounty of the Big Clubs

Everything would have worked out according to the Advocate General’s expectations had the big clubs ensured that revenues were shared in a way that restored the damage to competitive balance created by the Bosman decision. This would be calculated on the loss of post-contractual transfer fees, the general effect that transfer value would amortise to zero, and the effect on wage inflation. Quantifying such effects in a world of growing television revenues and the arrival of the bankrolling owners of the Walker/Abramovitch-type would itself be a polycentric problem of no small complexity. The Advocate General (para 232) was in no doubt that a system could be found, and we have no reason to doubt that a formula could be produced by way of a best estimate. For argument’s sake, the article will proceed on the basis of the Advocate General’s ‘no insurmountable obstacles’ and not adopt Hausman and Leonard’s ‘may be impossible’ (1997, p. 622) as to the whether an appropriate method of redistribution could be designed, and concentrate on whether a redistributive solution would be acceptable in practice.

The Advocate General stressed that clubs understood that (para 227):

(F)ootball is characterised by the mutual economic dependence of the clubs. Football is played by two teams meeting each other and testing their strength against each other. Each club thus needs the other one in order to be successful. For that reason each club has an interest in the health of the other clubs…The economic success of a league depends not least on the existence of a certain balance between its clubs. If the league is dominated by one overmighty club, experience shows that lack of interest will grow.

Whilst he admits later in para 227 that the elite might not volunteer the money, he had no doubt that ‘specific measures’ could be put in place. Indeed, at para 231 of his advice, he puts forward the Champions League (1992/93) as proof:

The participating clubs received SFR 38 million (54%). A further SFR 12 million (18%) was distributed to all the clubs which had been eliminated in the first two rounds of the three UEFA competitions for club teams. SFR 5.8 million (8%) was distributed between the 42 member associations of UEFA. The remaining SFR 14 million (20%) went to UEFA, to be invested for the benefit of football, in particular for the promotion of youth and women's football.

Using the exchange rate of SFR2.3 to the pound (http://fx.sauder.ubc.ca/etc/ GBPpages.pdf), we see the following:

1. The following eight teams each received £2.1m: Marseilles; Rangers; Bruges; CSKA Moscow; AC Milan; IFK Gothenburg; Porto and PSV Eindhoven.

2. £5.2m was distributed between the teams eliminated in the first two rounds of the three European competitions. There were twenty-four such teams in each of the European Cup and Cup Winners cup, and a further forty-eight in the UEFA Cup. So each club received £56,000.

3. £2.5m went to the national associations and theoretically may have been divided up so as to give an insignificant sum to the rest of the football.

4. £6.1m went to UEFA to be invested for the benefit of football.

Far from being a model of sporting redistribution, we see that eight clubs will each receive almost as much as all the national associations put together. The only clubs to gain significantly from redistribution of television revenues were those who were actually in the Champions League and teams like Glenavon of Northern Ireland (early losers in the Cup Winners Cup) who were so small that £56,000 would actually make a difference.

Furthermore, the Champions League revenues have not been the cure for footballing inequality, but a sizeable part of the problem and evidence of the elite’s desire to cement their advantage. Hoehn and Szymanski have demonstrated that ‘the existence of an elite international competition automatically creates domestic competition imbalance’. By 1996/97, qualification to the Champions League was worth £4m per club in broadcast revenues alone, a significant increase of the average Premier League income of £17m (Hoehn and Szymanski, 1999, pp. 221-224). The elite clubs used threats of forming of a breakaway European league to successfully press UEFA reconfigured its competitions ‘in a manner more financially favourable to the larger clubs’ (Miettinen and Parrish, 2007, at para 28). As is well known, in 1999 the Champions League was expanded so that four teams from the biggest league qualify, and two of those (soon to be three) were given a direct passage to the ultra-lucrative group stage. The number of games in the Champions League subsequently contracted, but not out of scruples as to becoming too ‘overmighty’, but because of the fixture congestion caused by combining European and domestic football, a dynamic predicted by Hoehn and Szymanski (1999, p. 225).

A. Lack of evidence

Let us consider what Football League history could have told the ECJ. The story is spelled out well by David Conn in The Football Business. One need only read to page 17 to know that the Advocate General was overly optimistic as to the availability of ‘specific measures’ to redistribute resources – put shortly, such measures are theoretically available, but depend on the support of the big clubs who generate the revenue being redistributed. The more money is at stake, the less the readiness to redistribute. And, of course, the greater the harm of such money to competitive balance.

In 1988, ITV put a proposal to the then ‘big five’ of English football. He suggested that they could all receive much more television money if they did not have to share it with the lower league clubs. The Premier League was born. Irving Scholar, then Chairman of ‘big five’ Tottenham Hotspurs, described it thus (Conn, 1997, p. 17):

Even if we only got the same money as 1988, we would each get more because, after breaking away, we would not be sharing the money out with the other smaller clubs.

It is worth starting the story of redistribution in English football at its high point. Conn explains (1997, p. 134):

After the First World War (there was) sharing of gate receipts equally between home and away teams. A small town club playing away at a big club would get half the gate, and therefore be able to compete financially. There was also a 4 per cent levy on the total gate receipts of all League clubs for the season, which were then distributed equally to every club in the League.

As for the redistribution of television revenues, originally it was distributed equally to the ninety-two football league clubs (Conn, 1997, p.139). However, from 1981 the story has been away from this redistributive paradise. The real world worked quite differently to that created in the adjudicative proceedings before the ECJ: as money came through the door, dialogue and solidarity between social partners flew out the window. Conn continues (1997, p. 140):

The first hectic whispers around the Big Five – Spurs, Arsenal, Manchester United, Everton, Liverpool – about breaking away to form a ‘Superleague’, came in 1981. They were bought off by an agreement that in future home clubs could keep all gate receipts, not have to share with the away club. This was the first step towards the bigger clubs’ total financial domination over the smaller clubs.

In 1985 came the ‘Heathrow Agreement’: the gate levy was reduced from 3% to 4% and the First Division kept 50% of the television revenues. 25% went to the Second Division and the remaining 25% went to the Third and Fourth Divisions. The creation of the Premiership saw a further decline in redistributions. The levy was replaced with a payment of £3m per year to the remaining Football League clubs, only £1m of which came from the Premier League, £2m coming from the Football Association (Conn, 1997, p. 145; Dobson and Goddard, 2001, p. 81).

B. Lack of Foresight

But does it matter that, pre-Bosman, England’s top clubs were less generous in sharing television revenues? The question was how they would respond to the effect of the Bosman decision on the cross-subsidy by way of transfer fees. The Advocate General believed they would see the sense of making good any reduction – the English and European experience suggests otherwise.

Outside the virtual reality of adjudication, the Premier League in 1992/93 paid a net amount of £13.3m to clubs in other divisions by way of transfer fees and £1m by way of direct resource redistribution. By contrast, until recently, other than a small subvention in respect of youth development, the Premiership distributed nothing of its television revenues to the Football League clubs so as to keep football competitive. And even within the Premiership, television revenues are split so as to favour the most successful clubs, and those most in demand by satellite television (Conn, 2007a). Finally, in 2007, the Premiership agreed to redistribute some of its television riches to the Football League clubs for their general use. We should not be too impressed. It amounts to £11.2m of the £830m that the Premier League will receive annually from home and overseas television and advertising rights. The Premier League believed this to be ‘generous’. For completeness, the Premiership pays £11.2m per annum to be split between Premiership clubs relegated to the Football League by way of ‘parachute payments’, to give its fallen members a competitive edge over the crowd (Conn, 2007b).It follows that in 2007/08 the Premier League redistributed by way of television revenues about as much as it had done by way of transfer fee payments in 1992/93. The average Premiership club, by contrast, will receive £28m from the sale of domestic television rights alone (Conn, 2006).

The position as regards the Champions League is similar. In 2005/06, UEFA budgeted for Champions League Revenue of €598m (£407m), of which €430m (£293m) would go the Champions League teams (72% as opposed to 54% in 1992/93), €145m (£99m) to UEFA (24% as opposed to 20% in 1992/93), and €22m (£15m) to national leagues (4% as opposed to 8%). The 18% that used to go less successful European competitors simply disappeared. Even within the Champions League competitors, distributions are greatly skewed in favour of the most successful. In distributing a €206m (£140m) fixed bonus pool, even a team that loses all games would earn an extra €3.5m (£2.4m), but a team that reached the last sixteen after a modest three wins and a draw would earn €6.7m (£4.6m). Should that team win the tournament, they would earn €11.2m, an amount equal to half the solidarity payments to national leagues, associations and clubs. A further €206m would be distributed ‘proportionate to the income from the TV contracts of each country’ – which further skewed distribution towards the elites of the Europe’s richest counries (UEFA, 2005). (The exchange used is €1.47 to £1 for the last trading day before the announcement was published on 25 September 2005, see http://www.taxfreegold.co.uk/2005 forexrates.html.)

The attitude to domestic cup competitions further demonstrates the lack of willingness to support the financial position of smaller clubs. As Dobson and Goddard point out (2001, pp. 89-90):

As with the explicit arrangements for sharing league gate revenues, however, the willingness of the leading clubs to provide these implicit cross-subsidies by participating wholeheartedly in the two main cup tournaments, the FA Cup and the League Cup, has come under strain in recent seasons.

This can be seen most clearly if we compare the League Cup in the 1980s to that now. Whereas a bottom division team had every hope of going home-and-away against a full-strength elite team in the League Cup second round and beyond, they now have a less attractive proposition of playing one game against what is in effect an elite club’s reserve side if they reach the third round. The downgrading of cup competitions is a reflection of the fact that staying in the Premiership is of paramount financial importance to smaller top-flight clubs, and Europe (qualification or participation) is only of slightly less importance to the remainder. Brian Barwick, FA Chief Executive, has argued (News of the World, 2008) that ‘(t)o stay in the Premier in League is an important thing but in 40 years’ time, will the fans remember if they finished 11^th, 12^th or 13^thin the Premier League? No.’/ In 1997 Middlesborough fans could make light of reaching the FA Cup Final and being relegated in the same season, as it was their first cup final and they had been relegated often. Notwithstanding the fate of the Big Four in the 2007/08 FA Cup, the brutal logic of Reading’s Kitson is inevitable (BBC Online, 2008c):

We are not going to win the FA Cup and I do not care less about it, to be honest. I care about staying in the Premier League, as does everybody at this club. Our league status is not protected by winning the FA Cup - simple as that.

However, the point goes beyond a simple matter of prioritisation. As Conn explains, when television revenues briefly declined at the start of the decade, the available prize fund for the FA Cup was decreased. By raising a credible threat of withdrawing from the competition, the Premiership ensured that the prize money for those reaching the later rounds remained unchanged –? the loss was borne by those who departed earlier (Conn, 2004, pp. 366-367). This clearly favoured the largest clubs, again defying the Advocate General’s expectations.

The Vice of Simplification

Let us consider another part of Fuller’s critique (Fuller, 1978, p. 401):

(W)hen one considers the nature of the problem of allocation, one wonders whether the ‘ adjudication’ here proposed could be be that in anything but name. In allocating $100m for scientific research it is never a case of Project A v Project B, but rather Project A v Project B v Project C v Project D… bearing in mind that Project Q may be an alternative B, while Project M supplements it, and that Project R may seek the same objective as Project C by a cheaper method, though one less certain to succeed etc.

When the ECJ announces that ‘ interdependence’ and the need for competitive balance should lead to a willingness to share television revenues, it is as if there were only ‘Project A’ (to share) or ‘Project B’ (not to share). In truth, there are an infinite number of variables. The ECJ’s analysis begs certain questions: what level of competitive balance, and between whom?; and to share what, and with whom?

Let us consider the following table which represents varying degrees of competitive level, and different possibilities for redistributing television revenues:

In considering their attitude to redistribution, each club has three priorities, a) to maintain position; b) to improve their position; and c) to ensure the general health of football insofar as it affects themselves. The Premiership elite undoubtedly have an interest in having a few direct competitors, but they have no interest in the likes of Aston Villa (the Premiership residents) challenging them for Champions League places; Model 1 is undoubtedly their preference. The Premiership residents may resent Model 1 harming their ability to challenge for the title or even the coveted fourth place, but their true fear is Models 3 and 4 which would expose them to greater competition from clubs below as well as a significant reduction in revenues for their shareholders. Their ideal is Model 2, for obvious reasons. As regards occasionals and residents, it is foolish to believe they have long-term influence. As should be clear from the recent Premiership proposals for an ‘overseas round’, clubs outside the Premiership were not consulted. It is also worth noting the ‘ overseas round’ controversy to show the relative bargaining strength within the Premiership. Richard Scudamore, Chief Executive of Premiership, explained that every club agreed with the idea because, if they did not, there were five clubs with predictable identities who would go it alone (Dunn, 2008).

This analysis contradicts an important strand of sports law thinking. Weatherill used a revolt against the Old Firm to demonstrate the Bosman analysis of interdependence to show that elite can be made to spread wealth that they generate throughout football (Weatherill, 2003, pp. 52-54). Recent events have shown that Weatherill’s example cannot be generalised. Indeed, the possibility of the game’s patricians seceding from national leagues to create a European superleague is clearly a far greater factor in the balance of power in the European game than the possibility that a ‘secession of the plebians’ might lead to increased redistributions.

As Pearson has noted in a review generally positive about the effect of European law on football, the ‘ commercial aspirations of the larger clubs’ makes it unfeasible that there will be any agreement for larger redistribution (Pearson, 2003, 115). It simply is not in the interest of elite clubs to foster challenges to their own dominance, whatever might have appeared logical to the Bosman judges. As Hoehn and Syzmanski state (1999, p. 229): ‘a system (of redistribution) might be difficult to implement within the existing structure where the beneficiaries are potential future competitors of the top clubs, but might be more acceptable within a closed Superleague.’ Even if it now has the most predictable top-four in European football, the Premiership did not create a closed circle of winners - twelve of the Premiership’s twenty-two members in 1992/93 were no longer in the top-flight fifteen years later (Robinson, 2007, p. 80). However, this has not translated into a desire to make lower league status less frightening – merely the anti-competitive parachute payment system aimed at assisting ex-Premiership members to bounce back (Conn, 2006). Beyond this, Hoehn and Syzmanski’s observation remains true: there is no desire to assist competitors breaking into the charmed circle.

It is worth noting that revenue figures also explain why clubs simply cannot afford to be generous (Eason, 2008):

It is difficult to see a consensus for redistributing broadcasting incomes. Those with Champions League revenues (the top four) will not redistribute away their advantage. That being the case, Tottenham and Newcastle have no interest in any agreement as regards domestic revenues. Added to which, any significant increase in redistribution of television revenues would only increase the power of those whose main spending power comes from a wealthy patron, whether it be Chelsea’s Roman Abramovitch or Manchester City’s Abu Dhabi syndicate. Further, we must also note that the investors and shareholders in what are increasingly businesses will want all revenue possible so as to get a return. Curiously, the Bosmananalysis on economic interdependence overlooked this point, despite ostensibly looking at sport firstly as a form of economic activity.

One might add, that with such a lack of common interest at the club level, it is difficult to talk rationally in terms of ‘social partners’ for a process of social dialogue. Similarly, one wonders what ‘democracy’ in football can mean if professional clubs are so far from forming a ‘demos’: the expansion of the G-14 to include other top clubs is more akin to a widening of the aristocracy than real democracy (BBC Online, 2008a).

Systematic error

There is one leading part of the Bosmananalysis that needs to be considered, that transfer fees represent a ‘rather arbitrary and inefficient’ means for a cross-subsidy (Dobson and Goddard, 2001, 99). The Advocate General noted at para 233 of his Opinion:

If a club can reckon with a certain basic amount which it will receive in any case, then solidarity between clubs is better served than by the possibility of receiving a large sum of money for one of the club’s players. As Mr Bosman has rightly submitted, the discovery of a gifted player who can be transferred to a big club for good money is very often largely a matter of chance. Yet the prosperity of football depends not only on the welfare of such a club, but also on all the other clubs being able to survive. That, however, is not guaranteed by the present rules on transfers.

Whilst the ECJ preferred redistribution to transfer fees because it was a more certain method, Hoehn and Syzmanski note that this certainty brings with it dangers (1999, p. 229):

Redistribution among the clubs is essentially a way of softening the effects of competition. Rather than promoting competitive balance, certain mechanisms may reduce the incentive to compete and worsen competitive balance.

The Advocate General believed that skilful design could avoid the problem, and certainly there is no evidence that Premiership clubs who benefit from the redistributions of television wealth are competing less hard in that league – although it is clear that the certainty of vast wealth simply by coming seventeenth in the League has led to teams competing less hard in cup competitions. As we have seen, this tends to harm the smaller clubs for whom cup competitions provide a valuable ‘cross-subsidy’, as Dobson and Goddard have noted (p. 89-90) – although, if one follows the Advocate General’s logic, it would not harm the small clubs if the cup competitions were abolished because a lucrative cup run is as much a matter of luck than anything else, and it is thus an ‘arbitrary and inefficient’ means for a cross-subsidy.

The problem with the Advocate General’s reasoning is that luck is at the heart of sporting success. It is as Gary Player explained: ‘The more I practice, the luckier I get.’ Any English football fans know that Dario Gradi’s Crewe Alexandra practiced the art of player development, and became exceptionally lucky (Conn, 1997, 240), creating a significant upturn for a club more accustomed to apply for re-election to the Football League than fight for promotion. Wimbledon’s rise from non-league to FA Cup winners and Premier League residents was likewise propelled by player development and transfer fees.

As often happens, adjudication struggles with complex reality, and the advantage of the court’s proposed system is not that it is better but that it can be understood as a system. The law could only look with incredulity at an outwardly random route for success, rather than understand that it brought opportunity to the industrious. In the post-Bosmanreality, clubs look to benefactors more than long term management to bridge the gaps in competitive balance – which is ironic, if we are supposed to view football as a commercial activity.

Cures for Polycentricity

Having considered the various errors in the Bosman approach to football, it is time to consider how to prevent perpetuating such mistakes without reinventing the ‘ sporting exception’ in the name of judicial restraint. Fuller argued that (1978, pp. 398-399):

There are polycentric elements in almost all problems submitted to adjudication. A decision may act as a precedent, often an awkward one, in some situation not foreseen by the arbiter…It is a question of knowing when the polycentric elements have become so significant and predominant that the proper limits of adjudication have been reached.

In the midst of the above quote, he gave an example:

(S)uppose a court in a suit between one litigant and a railway holds that it is an act of negligence for the railway not to construct an underpass at a particular crossing. There may be nothing to distinguish this crossing from other crossings on the line. As a matter of statistical probability it may be clear that constructing underpasses along the whole line would cost more lives (through accidents in blasting, for example) than would be lost if the only safety measure were the familiar ‘Stop, Look & Listen’ sign. If so, then what seems to be a decision simply declaring the rights and duties of two parties is in fact an inept solution for a polycentric problem, some elements of which cannot be brought before the court in a simple suit by one injured party against a defendant railway.

It should first be noted that such an action would doubtless be a claim for damages against the railway. As such, it should always be remembered that the burden of proof is on the claimant. The more obscure and polycentric the issues the harder it will be for the claimant to convince the judge that he is right on the balance of probabilities; and the easier it should be for the defendant railway to raise significant doubts. The deeper vice of resolving such a polycentric issue in favour of the Claimant is in precedent. It is one thing to make the railway pay damages, it is another thing to order it to continue building underpasses regardless of changes in circumstances and future discoveries. Similarly, it is one thing to decide that UEFA had not justified the restraint of Mr Bosman’s free movement of labour and order damages to be paid, another to set in stone that transfer fee revenues can be replaced by redistribution of television revenues. It is the distinction between resolving an individual dispute and setting down a law of general application.

Fuller offers cures for polycentricity. The most important is that the question of general law of policy should be resolved outside of adjudication. First, the legislature may take the matter out of the hands of judicial legislation to solve the problem by way of a ‘political deal’ or ‘an accommodation of interests’, meaning the normal parliamentary process (Fuller, 1978, 398-400). However, given that the ECJ was interpreting a fundamental provision of the EC Treaty, there is no realistic possibility of legislative adjustment. It is notable that for all the complaints by Member States and even the European Parliament, the Amsterdam Declaration on Sport did nothing to reverse Bosman (Weatherill, 2003, pp. 88-89).

Social dialogue

More directly, Fuller suggests that contract might be used to resolve polycentric issues in preference to adjudication. Miettinen and Parrish (2007, para 33) have suggested that football might adopt a process ‘making use of methods of social dialogue such as collective agreements’, which would amount to a contractual way for football’s stakeholders to settle their differences and remain compatible with EU law. They cite the case of Brentjens (Case C-115/97), which concerned an agreement between employers and employees within the Dutch building sector that all employers in that sector should subscribe to a particular pension arrangement, and that all employees should be entitled to the same pension terms regardless of risk.

One cannot help being sceptical. The sectoral pension agreement in Brentjens was born of a felicitous mixture of solidarity and self-interest involving employers and employees. If we return to Fuller’s example of two galleries dividing up a collection, law can adjudicate only through its knowledge of prices, the artistic value to each individual gallery of any painting or combination of paintings is outside its remit (Fuller, 1978, p. 394). These are examples where dialogue can maximise the interests of all parties – at the time of the agreement, dialogue can provide more than even a victory in litigation. This has less application where the interests are strikingly different, and divisions between rich and poor clubs are such that they are not realistically social partners for each other let alone capable of entering into such partnerships with players and governing bodies. Given the state of football, dialogue is more about settling litigation or possible litigation – a dynamic very favourable to the richest clubs seeking to exercise their economic freedoms. As we have seen, since the early 1980s, the English elite have been diluting and destroying such agreements as regards redistribution of gate and television revenues. Today, we find Chelsea’s chief executive, Peter Kenyon, expressly denouncing ‘ artificial ways to get an even platform’ (Hughes, 2008).

What does EU law offer to cajole the elite into spreading the wealth or to deter acts of selfishness? Weatherill suggests that a breakaway European league might be denied permission to collectively sell broadcasting rights and ‘thereby lose one commercially attractive opportunity’ (Weatherill, pp 79-80). The same point is made by Caiger and Gardiner (2000, p. 60), although Pearson (2003, p. 115) believes that this is unrealistic given ‘the commercial aspirations of the large clubs and governing bodies.’ The potency of such a threat obviously depends on whether the benefits of breakaway outweigh this dent in their earning power. Also the existence of a sectoral agreement will only be relevant to the application of competition law, it would not help FIFA or UEFA restrict a breakaway that invoked free-movement rights (Miettinen and Parrish, 2007, para 33). In truth, almost by definition, EU law cannot oblige parties to compromise their legal rights. The greatest contribution it can make to dialogue is to leave sufficient uncertainty as to victory so as to blunt the confidence of the elite that, should it come to litigation, their economic rights will be trumps. Of course, the law can never aspire to spread doubt as to what the law is.

Governance

Fuller’s other suggestion is that administrators can manage polycentric problems better than adjudicators, and thus a regulator may achieve the flexibility that a court lacks. Unless the Commission assumes the role of football’s regulator, this function falls to the game’s governing bodies. But the same point made earlier in respect of social dialogue applies equally in respect of governance: an intrusive Bosman- approach to proportionality assists the football elite against the governing bodies.

It is impossible to attribute the failure to redistribute resources to a failure of governance, for this presupposes that the governing bodies can enforce such solutions. Conn may well be right that the Football Association could have stopped the Premier League seceding from the Football League (Conn, 2004, pp. 49-52, 293-297), but that is old history. If we look at governance through the eyes of a Big Four chairman who is concerned only for the material consequences, he knows from Bosman that any attempts by the governing bodies to dent his commercial freedoms will be closely scrutinised by the ECJ.

Perhaps more pressing is that the governance of UEFA, FIFA and the national associations is inevitably undermined by the credible threat of the elite breaking away to form a European super league – we have already seen how the threat of a breakaway led to the expansion of the Champions League at the turn of the millennium. If, for example, football’s authorities introduced rules of governance to increase the redistribution of resources, or to restrict squad rotation, or to end the ‘ sugar-daddy’ phenomenon of billionaire owners bankrolling a club’s success, such rules of governance may well comply with EU law and even gain warm praise from the Commission and Court. However, the involvement of EU law goes deeper; should a disgruntled elite counter by threatening to create a breakaway super league, the top clubs would be relying on their fundamental Community law freedoms to trade and provide services in the EU without being largely restricted to their own national ‘market’. When the elite and the football authorities estimate their relative positions of strength, both will have in mind the countermeasures that the authorities might take to impede or deter the elite from offering their product across the EU in this new manner. For example, the authorities might prohibit the breakaway clubs from playing ‘friendlies’ against those in ‘official’ leagues or they might ban the players concerned from international competition. But one can see immediately the difficulty for the authorities in taking countermeasures. To impede the use of a Community law freedom is by definition prima facie illegal. The difficulty of justifying countermeasures as proportionate when the ECJ will exercise strict scrutiny in the manner of the Bosman decision as regards ‘sporting interest’ arguments must significantly increase the credibility of a breakaway by the elite. It follows that the fact that the elite’s most powerful weapon involves invoking their EU law rights must greatly strengthen the bargaining position of the elite and weaken that of the authorities.

Thus, the question is whether the ECJ can remedy the situation by finding a principled way to pull back from giving arguments of sporting interest the strict scrutiny given in Bosman. By doing so, EU Law would cease to provide powerful trump cards to football’s elite when the governing bodies seek to govern in the interests of the sport as a whole.

Judicial remedies to polycentricity

This leads us to Fuller’s suggestion as to how the judges might mitigate the polycentricity problems placed firmly in its court (Fuller, 1978, p. 398):

If judicial precedents are liberally interpreted and are subject to reformation and clarification as problems not originally foreseen arise, the judicial process as a whole is enabled to absorb these covert polycentric elements.

The Court should thus avoid setting general rules by limiting its precedent or by being more than usually ready to overrule itself. Unfortunately, Fuller did not elaborate on this key idea. As an idea, it exists in tension with his famous eight principles that make up the rule of law, particularly that the law should be ascertainable and should not be retrospective (Fuller, 1969, pp. 38-39, 51-62). But the basic imperative that the law should not simply repeat the oversights of the judicial lawmaker is sound. The great Oliver Wendall Holmes once wrote, a law should not persist from ‘blind limitation of the past’ (Holmes, 1897, p.469). This must apply even more so to the blind imitation of findings that have been falsified by experience.

The least that the Court could do is to overrule Bosman. However, this would be an inadequate response to the problem of polycentricity if the Court blunders from one ill-conceived assertion to the next. For example, if collective selling of television rights were taken before the ECJ, the Court might conclude that such collective selling does little to help competitive balance, and maybe transfer fees were the better way all along. Or it might recommend salary caps. Or it might recommend Hausman and Leonard’s suggestion of a flat tax on payrolls distributed evenly amongst the clubs (1997, pp. 622-633). What we need is not just a way to reverse Bosman, but a way for the court not to make the same mistake in a different way. To emphasise again, a general commitment to strict scrutiny will always render the position of the football authorities perilous: it is easy for an elite club to assert an EU law economic freedom whether one of free movement or competition; but it is very hard for a governing body to vindicate its position if put to strict proof as to necessity. This particularly if the court is willing to assert the existence of ‘less drastic means’ which are in fact speculative at the time and ill-conceived in retrospect.

Evidential approach

One method to mitigate polycentric problems is simply to widen the Court’s access to information or judicial investigation. This is an idea written in the famous Brandeis brief of US constitutional law whereby a Court deals with complexity by broadening its knowledge, what is described as ‘judicial education’ (Allison, 1994, 367, 374-377 and 382-383; Sedley, 1995, 398; and Karst, 1960, 99 et seq). One sees this method in the increasing practice of the English Courts to have special interest groups provide evidence and arguments that might be overlooked by the direct litigations. The general limitations of this process have been discussed by Hannett, and key amongst these are that the Court will still only receive an incomplete picture, now biased by the choice of special interest groups (Hannett, 2003). However, it is unnecessary to consider here just how far a Court can overcome evidential limitations. The point that should be clear from Bosman is that this can only be a partial solution: no amount of evidence will allow a Court to confidently make the predictions it sought to make in that case with any degree of confidence. Indeed, as was noted earlier, the polycentric problem applies not just to law but to any attempt to systematise knowledge of how human interactions work. The economic studies of football on which the ECJ would doubtless use to expand its expertise, useful though they are, are prone to the same polycentric errors when they move from descriptive to predictive. The Court’s bold assertions on football’s needs and interests are perhaps a prime case of what Fuller described as an arbitrator with a ‘desire to demonstrate virtuosity in his calling’ (Fuller, 1963).

Ultimately, all tribunals hear a limited amount of evidence over a limited time. When this time finishes they normally close their ears to further evidence. Bosman itself is a case in point, the court having refused an application to hear further evidence from UEFA (ECJ judgment, paras 52-54). We thus do not need a way for the Court to make an accurate prediction, because we shall always fall significantly short of this goal. We must seek a principled strategy for the Court to avoid predicting beyond its powers of foresight.

Judicial restraint and polycentricity

In human rights adjudication, it has long been recognised that a Court should frequently draw back from giving a positive view on questions of proportionality: a subject typically discussed under such headings as judicial restraint or deference. There are two particular reasons. First, as the Court will typically be overturning the decision of the elected authority, there is the question of democratic legitimacy. This need not concern us – the Court is not ranging its opinion against the elected legislatures of Europe. What concerns us is the second, ‘ pragmatic argument’ that the judiciary may be ‘institutionally incompetent’ to properly resolve certain complex issues, particularly socio-economic problems (Edwards, 2002, p. 859).

Variable intensity of review

The principal route by which a Court can avoid giving a definitive ruling on polycentric issues is to hold that the issues do not merit in depth scrutiny. Rivers explains it thus (2006, p. 207):

Where there is a minor limitation of a less important right, the gain to the public interest need not be large, and courts will ordinarily accept the appropriate executive body’s assessment of the degree to which the public interest is furthered and admit a range of possible policy choices. By contrast, where the limitation of rights is substantial and the right is important, the gain to the public interest must also be substantial. Furthermore, review will also be more intense: the executive will have to demonstrate that its assessments of the public interest are as reliable as they can be, and persuade the court that the cost to rights really is worth it.

To take a classic example, building restrictions that prevent building a three-storey house are less intrusive than ones which prevent even bungalows being built, and therefore require greater justification (see Justice Holmes in Hudson County Water Co v McCarter (1908) 209 US 349, 355-356). It is a commonplace that the proportionality test is to be applied more or less strictly depending on the importance of the rights infringement (Allan, 2006, pp. 685-686; and Edwards, 2002, p. 880).

The Advocate General in Bosman did consider the level of review appropriate to the case. However, he noted (para 216) that freedom of labour was of ‘fundamental importance’ and thus only a public interest of ‘paramount importance’ could justify it. He did not, however, ask himself whether some infringements of Article 39 are of less fundamental than others. The transfer rules concerned were not discriminatory and could justifiably have been held to be outside the core of the right.

Issues which are undoubtedly important within the basic norms set out in the Treaties would receive a higher level of scrutiny. For example, the effect of competition law on the sale of multi-billion pound television rights could not be portrayed as a minor matter. Importantly for Bosman, and also for the current debate on FIFA’s ‘6+5’ foreign player limitations (Miettinen and Parrish, 2007), this would justify strict scrutiny of restrictions that contradict the fundamental norms of Community Law freedoms. Sporting arguments, such as it being in essence of European club competition that there is a good presence of home nationals in each time, could be argued. However, they would have to prove such contentions to a high standard – Greer suggests that ‘clear and cogent’ evidence would be required (Greer, 2004, p. 432; and Rivers, 2006, p. 205). Discriminatory quotas could not be justified on the basis that having club teams with few home players would undermine European club competitions, because this assertion was speculative as of the time of Bosman and has not been vindicated in the subsequent years. But whilst the court may recognise clear speculation in others, how does it stop itself from committing the judicial over-reach of Bosman when strict scrutiny is unavoidable? For this we turn first to the theory of institutional competence.

Institutional Competence

It has been suggested by many that Courts should recognise the limits of their institutional competence in matters of proportionality. Such limits may be inherent within the adjudication process, or may arise from the limits of personal and professional expertise of the judiciary – and as such could be overcome if the evidence before the court and judicial training could be expanded (Jowell, 2003; Steyn 2005). As such, whilst the court may demand that the rights-infringer justifies its measures, it recognises that this may often only be done by way of intelligent speculation and plausible intuition. At this point the court may recognise (whether or not it is conscious of Fuller’s polycentricity analysis) that its own intelligent speculation is inapt as a source of a binding rule of law.

On this approach, the court may also, without abdicating its own role, legitimately give ‘due weight’ to the considered opinions of the institutional decision maker, the body closest to the situation and with a day-to-day expertise greater than the court concerned. Certainly this was the approach to proportionality adopted by a unanimous House of Lords in the recent case of Belfast City Council v Miss Behaving Ltd ([2007] UKHL 19, at paras 26, 37, 47, and 91). Whilst some legal commentators doubt the competence of football’s governing bodies, it will be difficult for the ECJ to deny that the governing bodies of world football have greater institutional competence in matters of football.

Case specific rulings

Courts can respect polycentricity problems by limiting the scope of the precedent to the facts of the case. For example, a court could stress the fourth part of proportionality (‘ balancing’) and hold that the particular harm to the specific claimant outweighed the legitimate aim, even accepting the defence case. In Bosman, the Court could legitimately have ruled that the greater good could never justify putting Mr Bosman in his dreadful predicament, but it could not have assumed that all transfer fee systems created this possibility. The court could have accepted that it was incompetent to gainsay UEFA on the benefits of the transfer system, but they could never put an innocent player in the limbo suffered by Mr Bosman. Given that the English transfer fee arbitration system did not create a risk of a player being left high-and-dry (Speight and Thomas, 1997, p. 203), the decision on Bosman went beyond what was necessary to adjudicate the case.

Conclusion

Much in football remains vulnerable to challenge before the ECJ. As Weatherill points out, almost the entire competition law settlement between the European Commission and UEFA is vulnerable with regard to extent to which transfer fees remain, the restricted transfer windows, and the rules that restrict the amount of times a player may change clubs in a season (Weatherill, 2003, 93). Collective bargaining of television rights remains vulnerable. The requirement to release players for international duty may come again into focus, despite the settlement of Charleroi. Added to which one must also consider that any measures taken by UEFA and FIFA to discourage the Premier League’s ‘ overseas round’ plan could equally lead to arguments as to abuse of a dominant position. On all these matters the ECJ will be the ultimate battleground – or at least all parties will act according to their expectations of how the Court will rule. Currently the wealthiest clubs can approach disputes with every reason to suppose that the sporting interest will be read narrowly and sceptically as a derogation from their assertions of economic freedom under the Treaty of Rome.

Many hope to see EU law as a facilitator for a social dialogue. Certainly, there is nothing like a dispute over legal rights to lead to a negotiated settlement. But such settlements are not always mutually beneficial processes, but are conducted in the shadow of how the case is likely to turn out in fact. This has proved highly favourable to the elite. Consider the settlement to the Charleroi litigation, which will see the top clubs receive a total of £128m from FIFA and UEFA in compensation for releasing players for international duty (BBC Online, 2008a). The Commission-brokered settlement on transfer fees led to the ‘one in-season transfer window system’ which was favourable to the richest clubs with the biggest squads, and leads to an artificial scramble on deadline day (Ley, 2007, and BBC Online, 2008d). The expansion of the Champions League in 1999 came from the dialogue (social or otherwise) caused by the elite’s threat to use their Community Law economic freedoms and provide services across European borders by way of a superleague.

Weatherill (2003, p. 93) might say, ‘Fair Play!’ to the EU’s contribution to football, but it is time to rethink completely the Bosman analysis of football, and the level of scrutiny that the Court gives to non-discriminatory sporting arguments. Doing so will not leave player’s rights wholly unprotected, nor completely remove the judge from the shoulder of the governing bodies. However, it is time to dent the elite’s confidence that they have the umpire on their side in disputes with the governing bodies. And this may even make ‘social dialogue’ a more attractive option than litigation for the rich and powerful.