Why transfer fee systems improve market competition, and why the Bosman ruling was flawed

Over the years competitive balance has decreased in many football competitions in Europe. This means that the number of clubs with a realistic chance of winning important prizes has declined. Consequently, there is less choice for the many consumers who want to watch, or become fan of, a club from their own country that can win important prices. Put differently, in many national consumer markets competition has decreased. Unfortunately, the legal profession has paid insufficient attention to the effect of decreasing competitive balance on competition in consumer markets. This is also relevant for the Bosman ruling. The ruling prohibited transfer fees for out-of-contract players. This has led to a decrease in competitive balance. Therefore, the ruling has decreased the competition in national consumer markets. However, in making the ruling the European Court of Justice fully neglected the possibility of such a decrease. While it did apply the hard law concerning the free movement of workers, and it did take account of the specific features of the sport and the need for competitive balance more specifically (even though this was not decisive in the end), it neglected hard competition law as regards the competition in consumer markets. If it had not done so the ruling might have been different, and better for the football fans.


Introduction
In the Bosman ruling of 1995, the European Court of Justice (ECJ) argued that the then existing transfer system in football was incompatible with EU law and should be changed in two ways.First, clubs should no longer be allowed to collect a transfer fee for a player whose contract had expired.Second, national associations should no longer impose quotas on players from other EU countries. 1The second point is neglected below.This paper discusses the first point, while also paying attention to the question whether transfer fees for players still under contract are compatible with EU law.
The reason the ECJ prohibited transfer fees for out-ofcontract players was that they were not in line with free movement law, as formulated in Article 48 of the EC Treaty (now article 45 TFEU).The ECJ left open the possibility that these transfer fees were also incompatible with competition law, and article 85 of the EC Treaty (now article 101 TFEU) more specifically.The reason for thinking of this possibility was that, in a labour market with a transfer fee system, clubs did not have the same chances for engaging players as they would have had under normal competitive conditions.However, in the end the ECJ did not investigate this possibility seriously because article 48 offered sufficient reason for prohibiting the transfer fees anyway. 2 The possibility is also neglected in this paper.Below, the only question is whether transfer fees are compatible with free movement law.
Of course, the ECJ and its Advocate General (AG) Lenz considered the counterarguments.Arguably the main one was that the transfer system improved the competitive balance among football teams, and the need for competitive balance was one of the specific features of the sport which should be protected for reasons of public interest.This could be a reason for granting an exception to article 48.The ECJ discussed this argument after it had established that the transfer rules were incompatible with Article 48 in principle.At that point, it stated: "It could only be otherwise if those rules pursued a legitimate aim compatible with the Treaty and were justified by pressing reasons of public interest.But even if that were so, applications of those rules would still have to be such as to ensure achievement of the aim in question and not go beyond what is necessary for that purpose". 3Put differently, the transfer rules should meet the conditions of legitimacy and proportionality.The ECJ continued to argue that sporting activities were of considerable social importance and maintaining competitive balance was a legitimate aim compatible with the Treaty. 4So, the condition of legitimacy was fulfilled.
Regarding the condition of proportionality AG Lenz was more critical, and the ECJ followed him.First, AG Lenz thought it was doubtful whether the transfer rules could attain the objective of maintaining competitive balance, or could do so to a serious extent.Second, even if they could, there were other means of attaining that objective, which had less effect on freedom of movement. 5This implied the transfer rules did not meet the condition of proportionality.Consequently, no exception to article 48 could be granted.This line of reasoning resulted in the Bosman ruling as described above. 6he Bosman case has wider importance.The ECJ made clear that the specific features of sport should be considered in principle when applying free movement and competition law to other cases too, but that this could only affect judicial decisions if the conditions of legitimacy and proportionality were met. 7In relation to Bosman and other rulings, legal scholars have paid much attention to the question how exactly the legal authorities should deal with the specific features of the sport. 8is paper focuses on the transfer system and on one of the specific features of the sport, the need for competitive balance.It presents the following argument: When competitive balance decreases, this does not only affect the competition on the pitch but also the competition in economic markets.More specifically, a (long-term) decrease in competitive balance reduces the competition in consumer markets.Therefore, if transfer rules improve competitive balance, they also improve market competition.This can be a legal reason for maintaining such rules even if they do not fulfil the conditions of legitimacy and proportionality as discussed above.Unfortunately, this line of reasoning was neglected during the Bosman case.Therefore, the Bosman ruling is flawed.
The structure of the paper is as follows.Sections 2, 3 and 4 provide the economic foundation for the legal argument in section 5. Section 2 investigates the effect of competitive balance on the competition in consumer markets.Section 3 discusses whether competitive balance improves the pleasure of the fans, and so consumer welfare.Section 4 analyses whether and to which extent transfer rules affect competitive balance.Section 5 discusses whether, given the economic analysis, the pre-Bosman and the post-Bosman transfer system are compatible with EU law.Section 6 concludes.

Competitive balance and market competition
Professional football clubs compete with other clubs for consumers.They do so on their domestic market to a large extent.For instance, Ajax Amsterdam seriously competes for dedicated fans and other consumers with other Dutch clubs, while it competes less seriously with German clubs for consumers.This means the national consumer markets are relevant markets.In many countries, competitive balance has decreased over the years.This means that the number of top clubs, i.e., clubs which have a serious chance of winning national or even European titles one day, has decreased.Therefore, consumers who want to watch, or even become fan of, a top club from their own country have less choice.Put differently, market competition has decreased.As a result, consumer prices have increased, to the detriment of welfare. 9This is not to deny that clubs are also competing for consumers with foreign clubs to some extent.For instance, Barcelona competes with Liverpool for Dutch TV viewers.A Dutch consumer may choose to become a fan of Barcelona, or some other foreign club (perhaps as a second favorite after a Dutch club), watching some or many of its games.In other countries consumers make similar choices among foreign clubs.So, the European consumer market is a relevant market too.Consumers often prefer a foreign club that can win the highest European title one day, so they get less choice when the number of clubs that can win this title gets lower.Thus, in the European consumer market too, less competitive balance means less competition.However, this problem may still be relatively small, as the number of clubs that have a serious chance of winning the European title one day is still well above ten.
The conclusion most relevant for this paper is this: a reduction in competitive balance within a country means that the number of domestic clubs that can win important prizes (in the domestic competitions and possibly in European competitions too) decreases, and because of this the competition in the national consumer market deceases.

Competitive balance and consumer welfare
This section discusses the effects of competitive balance on consumer welfare.The consumer welfare generated by the professional football sector is equal to the sum of the utility, or pleasure, which all consumers (football fans) taken together get from watching the games.Obviously, a competition in which one team wins all its matches is not optimal for welfare; the related lack of uncertainty of outcome would strongly reduce the utility of the football fans watching it, including the fans of the winning team.At the same time, the following holds: as long as the number of wins of a club does not surpass a certain (high) level, the utility of the fans of the club increases the more matches (and prizes) are won.Therefore, welfare is higher when clubs with a large fan-base win more often than clubs with fewer fans, as in that case more fans get pleasure from seeing their own team win.These two points, plus some other ones, lead to the following conclusion: to maximize welfare, clubs with a large fan-base should win more often than clubs with fewer fans, but at the same time uncertainty of outcome (and so competitive balance) should not be too low. 10A more difficult question is which level of competitive balance exactly maximizes welfare.As yet no clear answer has been given here.Another question is whether, in the real world, the competitive balance is below the welfare-maximizing level, which would imply that an increase in the balance improves welfare.This question has been investigated in many empirical studies.The overview studies by Szymanski (2003), Borland and   MacDonald (2003), Downward et al. (2009) and Dobson  and Goddard (2011) discuss these empirical studies, and they all conclude that the results are mixed.
The overview study by Van der Burg (2023) investigates the reasons why the results of studies can be mixed, and it takes account of more recent empirical studies too.In relation to this, its conclusions are less ambiguous than those of the older overviews.More precisely, the first part of the study focuses on empirical research published before 2014.After having observed, like the other overview studies, that some studies conclude that reductions in competitive balance have a negative effect on welfare while other studies conclude that there is no such effect, Van der Burg argues that some of the latter studies are based on an imperfect research approach.For this and other reasons, he is positive about the hypothesis that in the years before 2014, and possibly before 2000 too, competitive balance was below the welfare-maximizing level in many competitions at least, but he acknowledges that the empirical studies offer no certainty here.Subsequently, Van der Burg discusses the empirical studies published since 2014, which concern situations in relatively recent years.As competitive balance has been decreasing over the years, these studies concern cases where competitive balance is relatively low, making it more likely that lack of competitive balance is a serious problem.Indeed, after investigating the empirical studies published since 2014, Van der Burg concludes that it is highly plausible that in recent years the competitive balance has been below the welfare-maximising level in most competitions at least.

The effects of a transfer fee system on competitive balance
4.1 Basic theory and some first evidence Késenne (2007) provides a theory of the effects of a transfer fee system on competitive balance in football in Europe.His model assumes that the objective of football clubs is to win as many games as possible.It takes the initial distribution of young talents over the clubs to be given, with part of the talents playing for small clubs.The model shows that if this part is sufficiently large, the later transfers of the talents will benefit the small clubs on balance; the amount of transfer fees flowing from large clubs to small clubs will exceed the amount flowing the opposite way.As a result, the small clubs will increase their players' wage budgets, and this improves competitive balance.Késenne acknowledges that clubs could also use the benefits from transfers to improve profits instead of the team, in principle.However, the objective of (most) European clubs is to win games rather than make profits.Therefore, it is realistic to assume that transfer surpluses are used to improve the team.
To improve their first team, clubs can pay salaries to firstteam players and transfer fees to other clubs, as in Késenne's model, but they can also do other things such as hiring good coaches, improve players' diets, invest in a youth development programme, etc.Including such factors makes the analysis more realistic.However, it will not change the conclusion that benefits from the transfer system will improve the first team (sooner or later).Therefore, it is reasonable to neglect these complicating factors.In relation to this, section 4 will assume that every club will use any net gain from transfers to increase its wage bill for players, while a net loss from transfers will lead to an equal decrease in this bill. 11he idea that the initial distribution of talents over the clubs may be such that the small clubs benefit financially from the transfer system, is supported by empirical evidence.Dobson and Goddard (2011) investigate the transfers within the four highest English divisions in four periods of three seasons : 1976-78 and 1986-88 (both before Bosman) and  1996-98 and 2006-2008 (after Bosman).In each period, the clubs in the highest division taken together had a transfer deficit in their dealings with clubs from the three lower divisions (which means they paid more transfer fees than they received), while each of the three lower divisions had a transfer surplus.The next section will provide additional evidence.Given all the evidence, it can be concluded that the transfer system improves competitive balance in football in Europe, both before Bosman and after it.
In addition to Késenne's theory some remarks can be made about the specific result of the Bosman ruling, the abandonment of transfer fees for out-of-contract players.Before Bosman transfer fees could be asked for all players, which meant that a club which had a contract with a player could, in principle, force the player to remain with it forever by asking an excessive transfer fee which no club would pay.Thus, there was no time pressure to sell talents before their contracts expired.If, for instance, the transfer fee of an 18-year-old talent was expected to be highest when he reached the age of 25, his club could keep him until that age and sell him then.The talent would then improve the first team for quite many years, and the club would also get a high transfer fee.This makes it plausible that the abandonment of transfer fees for out-of-contract players has decreased competitive balance.
There is no reliable statistical evidence in favor or against this argument. 12However, there is anecdotal evidence.In the pre-Bosman area, Ajax Amsterdam managed to retain its home-grown talent Johan Cruijff until 1973, when he was aged 26 and had just won the European Cup I with Ajax three times in a row.In 1973 Ajax sold Cruijff for a record transfer fee to Barcelona.If the Bosman ruling had been issued in 1960, Cruijff could have left Ajax earlier and for a lower transfer fee.In that case, Ajax might not have beaten the financially bigger clubs Inter Milan and Juventus in the European finals of 1972 and 1973.All in all, it is plausible that the abandonment of transfer fees for out-of-contract players has had a negative effect on competitive balance.

4.2.1
The study of Szymanski and Kuypers (1999) Szymanski and Kuypers (1999) are among the first to estimate the effects of the transfer system on competitive balance.They analyse the four highest football divisions in England in the 1996/97 season.That season, the Premier League clubs (taken together) had a transfer deficit of £50 million.It consisted of a deficit with the three lower English divisions of 3 million and a deficit with Scotland and overseas countries of 47 million.According to the authors, the small deficit with the lower English divisions means that 12 Goddard et al. ( 2012) present an overview of empirical studies that investigate how the Bosman ruling affected competitive balance in England.All studies investigate the effect of the ruling as a whole, so the combined effect of the abandonment of transfer fees for outof-contract players and the abandonment of quota for foreign players (from other EU countries).The studies make clear that competitive balance decreased in the first fifteen years after Bosman.However, Goddard et al. argue it is unclear to which extent this was the result of Bosman, as there were contemporaneous developments that also affected competitive balance, such as the strong increase in gate and broadcasting revenues.Isolating the effect of Bosman from the effects of the other developments is difficult.Norbäck et al. (2021) address this problem in their analysis of the competitive balance in the Champions League and its predecessor, the European Cup I tournament.In the period 1996-2014, so after Bosman, the balance was lower than in 1980-95.The authors discuss three changes other than Bosman that may have been a reason for this decease: a change in the tournament structure, a change in the distribution of the tournament's revenues, and a change in the prize money.They show that the three changes cannot fully explain the decrease.Therefore, they conclude that Bosman had a negative effect on competitive balance.
However, the authors neglect still other changes that were not directly related to the Champions League but that could affect, or did affect, the competitive balance there.For instance, the rise of pay TV has increased the revenues of larger clubs especially, thus harming competitive balance in all competitions including the Champions League (Groot 2008).In addition, even if it could be shown that Bosman decreased competitive balance, it would still be possible that this effect was caused by the abolishment of the quota on players from other EU countries.All in all, there is no reliable statistical evidence about the effects of the abandonment of transfer fees for out-of-contract players. 11This does not mean that a club cannot use transfer revenues to pay transfer fees to other clubs.It means that if it does so, its net transfer surplus will be lower so that less money can be spend on wages.
the amount of money trickling down the English leagues is "negligible". 13This could suggest that the impact on competitive balance is also negligible, and the authors seem to share this idea.
However, such an idea would be debatable.In the same season, the wage bill of the Premier League clubs was £207 million. 14Without their 50 million transfer deficit, the clubs could have spent 257 million on wages.The wage bill of the clubs in the three lower divisions was £117 million. 15Without their 3 million transfer surplus in their dealings with the Premier League, these clubs could have spent 114 million on wages.(Their dealings with Scotland and overseas countries are unknown and neglected here.) The wage bill for all players in the 1996/97 Premier League was approximately £160 million. 16Keep in mind that this is 77% of the total wage bill (of 207 million; see above).Without their 50 million transfer deficit, the players' wage bill of the Premier League clubs would have been 210 million (given the assumption of section 4 that a transfer surplus or deficit of a club always leads to an equal increase or decrease in its players' wage bill).Put differently, the transfer system reduced the players' wage bill of the Premier League from 210 to 160 million, a decrease of 24%.It is not unreasonable to take if that the players' wage bill of the clubs in the three lower divisions was also (about) 77% of their total wage bill (of 117 million), so 90 million.Thus, it would have been 87 million without the transfer surplus of 3 million with the Premier League.Put differently, the transfer system increased the players' wage bill of the clubs in the three lower divisions from 87 to 90 million, an increase of 3.4%.It follows that the transfer system increased the ratio of the players' wage bill of the clubs in the three lower divisions to the players' wage bill of the clubs in the Premier League by 36%. 17This will have enabled the former to compete better for players with the latter, and so to get better results in their games against larger clubs.All in all, the transfer system had a significant, and possibly substantial, positive effect on competitive balance.
To take a different perspective: in 1996/97 the players' wage bill of the average Premier League club was, with the transfer system, 6.4 times as large as that of the average club in the three lower divisions. 18Without the transfer system, it would have been 8.7 times as large. 19So, the transfer system reduced the ratio from 8.7 to 6.4.This might suggest that the system solved only a small part of the problem of competitive balance, since with the transfer system the inequalities were still quite large.However, such an interpretation would be wrong.As explained in section 3, a situation with perfect competitive balance is undesirable, so the problem of competitive balance is not that all clubs do not have equal playing strengths.The problem is that the level of competitive balance is below the welfare-maximizing level.Let us assume, just for the moment, that this optimal level was at the point where the ratio of the players' wage bill of the average Premier League club to that of the average club in the three lower divisions equalled 6.4.This would imply that, in 1996/97, the transfer system solved the problem of too low competitive balance completely.Of course, this is merely a hypothetical conclusion since, as noted in section 3, the optimal level of competitive balance is unknown.Anyway, this theoretical exercise has hopefully made clearer why it is impossible to make precise statements about the extent to which the transfer system helps solve the problem of competitive balance.
The main conclusion remains this: the figures of Kuypers and Szymanski show that the transfer system increased competitive balance in English football in the 1996/97 season to a significant and possibly substantial extent.

The figures of AG Lenz (1995)
During the Bosman case, AG Lenz also investigated the four highest English divisions.He focused on the 1992/93 season, when the Premier League clubs had a transfer deficit of £18.5 million.By contrast, the clubs in the First, Second and Third Division had a transfer surplus of 9.3 million, 2.4 million, and 1.6 million, respectively, so 13.3 million in total. 20So, the smaller clubs benefitted from the system.Still, AG Lenz was not optimistic about the effects of the system on competitive balance. 21His full economic argument will be discussed in section 4.4.Here, I give my own interpretation of Lenz's figures. 13 Szymanski and Kuypers (1999, p.113). 14Ibidem, p.81. 15 Ibidem, p.81. 16  Szymanski and Kuypers (1999, p.80) say it is a reasonable estimate that each club had a first-team squad of 40 players, and the average player earned £200.000.With 20 clubs, the players' wage bill is 160 million in total. 17(1.034)/(1-0.24)= 1.36. 18For the 20 Premier League clubs this bill was 160 million, so 8 million on average.For the 72 clubs in the three lower divisions the bill was 90 million, so 1.25 million on average.8/1.25 = 6.4. 19For the 20 Premier League clubs this bill would have been 210 million without the transfer system, so 10.5 million on average.For the 72 clubs in the three lower divisions, it would have been 87 million, so 1.21 million on average.10.5/1.21 = 8.7. 20Opinion of AG Lenz, n 2 supra, para 222. 21Ibidem, para 225.
In 1992/93, the total wage bill of the Premier League clubs was roughly £110 million. 22The total wage bill of the three divisions below was roughly 69.6 million. 23Roughly 77% of these wage bills was for players' wages. 24Given these estimates, the transfer system improved the ratio of the players' wage bill of the three lower divisions to that of the Premier League by 62%. 25 Thus, it enabled the smaller clubs to spend roughly 62% more on players' wages compared to the Premier League clubs.This means the effect on competitive balance was significant and possibly (very) substantial.

The study of Szymanski (2015)
This study analyses the top divisions of 24 European football countries from 2006 to 2011.Over this five-season period, there were 9 countries where the clubs in the top division had a transfer deficit: Turkey, England, Ukraine, Spain, Russia, Italy, Greece, Germany, and Rumania.The deficit as a percentage of total wage spending was 18% in England, 17% in Spain, 10% in Italy, and 8% in Germany, for instance.In the other 15 countries, the top division had a transfer surplus.In Portugal.Sweden, the Netherlands and Belgium it was, respectively, 32, 21, 14 and 11% of total wage spending.Szymanski concludes that the larger football nations (England, Germany, Italy, Spain) have a deficit.At the same time, his interpretation of the figures is that "overall the pattern does not suggest a system in which redistribution through transfer fees plays a significant role". 26his last conclusion is debatable.Given the assumption of section 4 that a transfer surplus or deficit of a club always leads to an equal increase or decrease in its players' wage bill, the figures imply that the transfer system increased, for instance, the ratio of the players' wage bill of the Dutch top division to that of the English top division with roughly 54%. 27This means the transfer system improved competitive balance to a significant and possibly (very) substantial extent.

The study of Hoey et al. (2021)
This study provides information about the transfer surpluses and deficits of the 202 clubs that played, during the five seasons from 2013 to 2018, for one season at least in the top division of one of the following eight countries: England, France, Germany, Italy, the Netherlands, Portugal, Scotland, and Spain.The clubs are ranked according to market size, with a distinction between small, mid-sized and large clubs.Almost all small and mid-sized clubs had a transfer surplus over the five-season period.Especially some of the midsized clubs, such as Benfica, Monaco, Lyon, Sevilla, and Ajax, had big surpluses.The biggest deficits were experienced by some of the large clubs.Overall, the transfer system reduced the financial inequalities among the 202 clubs.
Comparing the eight top divisions, the Premier League clubs (taken together) had by far the largest transfer deficit.The German and Italian top division also had a deficit, but it was considerably smaller.Spain and Scotland had still smaller deficits.France and the Netherlands had serious transfer surpluses, and Portugal had a very high surplus.This shows that, overall, the transfer system reduced the inequalities among the countries.The study also analyses the inequalities within each national top division and within the Champions League.The transfer system had the 23 In 1992/93, the average wage bill of a Premier League club was roughly £5 million (n 22 supra).According to Szymanski and  Kuypers (1999, p. 96), the average club wage bill in the Premier League (which had 22 clubs), the First Division (24 clubs), and the Second Division (24 clubs) was, respectively, 12.26, 4.24, and 1.88 times higher than the average club wage bill in the Third Division (23 clubs), This implies that the average club wage bill in the Third Division was (5/12.26=) 0.41 million, and the total wage bill (23•0.41=) 9.4 million.In the Second Division the total wage bill was (24•1.88•0.41=) 18.5 million, and in the First Division it was (24•4.24•0.41=) 41.7 million, making the total for the three lower divisions 69.6 million. 24In section 4.2.1, this percentage was estimated to be 77 in the 1996/97 season.I have no similar estimate for the 1992/92 season, so the best I can do is to take the percentage in 1992/93 to be 77 too. 25The players' wage bill of the Premier League clubs was (0,77•110 =) 84.7 million, that of the lower division clubs (0,77•69.6=) 53.6 million.The 18.5 million transfer deficit of the Premier League clubs implies that their players' wage bill would have been 103.2 million if there had been no transfer system.The 13.3 million transfer surplus of the lower divisions clubs implies that their players' wage bill would have been 40.3 million without the transfer system.Consequently, the ratio of the players' wage bill of the three lower divisions to the players' wage bill of the Premier League was 0.63 with the transfer system, while it would have been 0.39 without it.This means the transfer system improved the ratio by 62%. 26 Szymanski (2015, p.10). 27This can be explained as follows.In the 2016/17 season, the clubs in the top divisions in England, Germany, Spain, France, Italy, Portugal, and the Netherlands, spent on average 74 percent of their wage bill on players' wages (according to UEFA 2017).Given this figure, it is not unreasonable to assume that, in the 2006-2011 period, the top division clubs in both The Netherlands and England also spent on average 74% of their wage bill on players' wages.Given the figures in the main text, it follows that, with the transfer system, the Dutch transfer surplus was (14/0.74=) 18.9% of the Duch players' wage bill, and the English transfer deficit was (18/0.74=) 24.3% of the English players' wage bill.The Dutch players' wage bill with transfer system is called a, the English one b, so the ratio of the former to the latter is a/b.Without the transfer system, the ratio of the Dutch players' wage bill to the English one would have been 0,811a/1.243b,so 0.65(a/b).A change from 0.65(a/b) to a/b is an increase of 54%. 22Figure 3.2 of Szymanski and Kuypers (1999, p. 96) gives the average wage bill of a Premier League club from 1978 to 1998.In 1992/93 it was roughly £5 million.This probably involves an error of a few percents as I inferred this number from a line in the figure.With 22 Premier League clubs, the total wage bill was roughly 110 million.
strongest redistributive effect in the Champions League.It also reduced the inequalities in the French, Italian, and Spanish top division.In the other five countries, however, the effects on the distribution of revenues were not really important (from the perspective of the top division as a whole).All in all, the transfer system reduced the inequalities at the European level and within some but not all individual countries.
However, Hoey et al. emphasize that these redistributive effects are small.They demonstrate this with two Lorenzcurves.The first one concerns the pre-transfer revenues of the 202 clubs, i.e., the revenues excluding the net transfer revenues (which can be positive or negative).The second one concerns the post-transfer revenues, which include the net transfer revenues.The Lorenz curve of the post-transfer revenues is not far above that of the pre-transfer revenues. 28he authors conclude that, by and large, "the transfer system hardly puts a dent in the large pre-existing inequality in pretransfers revenues." 29owever, a different interpretation is also possible.Consider the smallest 50% of clubs.The post-transfer revenues of these clubs were 13.8% higher than their pre-transfer revenues. 30In the figure this appears as a small upward move of the Lorenz curve.However, it also implies that the transfer system increased the ratio of the players' wage budgets of the smallest 50% of clubs to the players' wage budgets of the largest 50% of clubs with about 35%. 31 Thus, the transfer system improved competitive balance to a significant and possibly substantial extent.

Conclusions of section 4.2
It is reasonable to assume that every club uses its net gain or net loss from transfers to increase or decrease its players' wage budget.Given this assumption, the transfer system increases the ratio of the players' wage bills of small (and medium-sized) clubs to the players' wage bills of large clubs with about 36, 62, 54 and 35% in the example cases above.This means that a transfer fee system improves competitive balance to a significant and possibly substantial extent.

Transfer fees as a barrier to entry
Some clubs have been bought by rich investors willing to spend much money on the team.Such new owners often start with a relatively weak team, and improving it is hampered by the transfer fee system.Therefore, this system is a barrier to entry for clubs that wish to compete with the top European clubs (Szymanski 2015).This is in line with the view of the ECJ that it is possible that transfer fees are incompatible with competition law because, in the labour market, clubs do not get the same chances for engaging players as they would have under normal competitive conditions.However, the ECJ did not seriously investigate this possibility (see section 1), and this paper follows this approach.
This leaves us with the economic question how the entry barrier created by the transfer system affects the impact of new owners on competitive balance.I can only discuss some examples here, starting with Manchester City.This club was bought by Sheikh Mansour in 2008.In the previous 30 years, it had never finished in the top three of the English top division.In the first four years after the 2008 takeover, the results improved gradually but significantly.In 2011 City ended third in the Premier League, and in 2012 first.This suggests the new owner improved competitive balance in England during his first years.However, in the five seasons 31 This estimate is based on the following points: 1.Over the five seasons from 2013 to 2018, the 202 clubs in the eight countries spent on average 62% of their (pre-transfer) revenues on wages.This is the average of two estimates (rounded off).The first estimate concerns England, Spain, Italy, Germany, and France, where, over the five seasons, the top division clubs spent on average 60.4% of their (pre-transfer) revenues on wages (Deloitte 2015; 2016;  2017; 2018; 2019).The second estimate concerns Portugal, the Netherlands and Scotland.For these countries, Deloitte gives no relevant information and I turned to the UEFA Club Licencing Benchmarking Reports.I could only find the reports that concerned the 2014/15 season, the 2016/17 season, and the 2017/18 season (UEFA 2015; 2017;  2018).During these three seasons the top divisions of the three countries spent on average 64% of their (pre-transfer) revenues on wages.
2. Over the five seasons the 202 clubs in the eight countries spent, on average, 74% of their total wage spending on players' wages.This is based on the estimate given in note 27 supra, according to which the relevant percentage was, in 2016/17, 74 on average in the top divisions in England, Spain, Italy, Germany, France, Portugal, and the Netherlands.I have no information about Scotland and the other four seasons.
3. The transfer system increased the players' wage bill of the smallest 50% of clubs with 30.1%.This follows from the figure of 13.8% in the main text, points 1 and 2 above, and the assumption that a trans- 28 The two curves correspond to pre-and post-transfer Gini coefficients of 0.597 and 0.571, respectively. 29Hoey et al. (2021, p.9). 30 This information was provided to me by co-author Thomas Peeters in his mail of 27 June 2023.fer surplus or deficit of a club always leads to an equal increase or decrease in its players' wage bill.Note that (13.8/0.62)/0.74= 30.1.
4. The transfer system decreased the players' wage bill of the largest 50% of clubs with 3.7%.This follows from the mail I received from Thomas Peeters (n 30 supra), according to which the pre-transfer revenues of the largest 50% of clubs are 88.9997% of the pre-transfer revenues of all clubs, while this percentage is 87.4818 for the post transfer revenues.Consequently, the post-transfer revenues of the largest 50% of clubs are 1.7% below their pre-transfer revenues, and (1.7/0.62)/0.74= 3.7.5.It follows from points 3 and 4 above that the transfer system increased the ratio of the players' wage bill of the smallest 50% of clubs to the players' wage bill of the largest 50% of clubs with 35%.Note that 1.301/(1-0.037)= 1.35.
Footnote 31 (continued) from 2018 to 2023 City won four English titles.This raises the question whether the owner has decreased competitive balance in recent years.
Paris SG was bought by Qatar Sports investments in 2011.In the fifty seasons before, the club had won the French title twice, in 1986 and 1994.After the takeover, results quickly improved.From 2013 to 2016 Paris SG won four French titles in a row.In the next seven seasons it won five.This suggests the new owner decreased competitive balance soon after the start.There might have been more competitive balance if the entry barrier created by the transfer system had been larger, making it harder for the new owner to improve results on the pitch.
More generally, the question whether the barrier to entry created by the transfer system reduces or increases competitive balance, is not an easy one.More research is needed here.For now, there is no reason to assume that the existence of this barrier can change our earlier conclusions about the effects of the transfer system on competitive balance (significantly).

The economic arguments of AG Lenz in 1995
During the Bosman case, AG Lenz argued that the transfer system cannot be justified by the objective of maintaining competitive balance.Among other things, he wrote the following regarding the suitability of transfer rules for maintaining competitive balance: "it must first be observed that the rules currently in force probably very often force the smaller professional clubs to sell players in order to ensure their survival by means of the transfer income thereby obtained.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." 32ections 4.1 and 4.2 have argued differently: the transfer system improves competitive balance to a significant and possibly substantial extent.In addition, the abandonment of transfer fees for out-of-contract players, which was the specific result of Bosman, reduced competitive balance, especially because it made it difficult for small clubs to retain their best talents for a longer period.
Unfortunately, AG Lenz fully neglected the possibility that the pre-Bosman transfer system enabled small clubs to retain their best talents for a longer period, and to sell them for a higher transfer fee after that.Instead, he emphasized that small clubs will probably hardly ever acquire good players from big clubs.Of course, this is true.However, the transfer system does not aggravate this problem, as AG Lenz suggested.Once a top player is playing for a large club he gets a high salary, plays big matches, and can win prestigious prizes.Because of this, he will normally not move to a small club, until he is no longer good enough for a large club because of age.At that time, no large club will wish to spend money on him.If he still has a contract with a large club, this club will try to let him move to a small club, for no transfer fee if necessary.The player may then make this move or end his career.Thus, it is generally impossible for small clubs to acquire good players (in their best days) from large clubs, both with and without transfer fees.So, it would have been better if AG Lenz had focused on the more important question of whether the transfer system lets young talents stay longer with small clubs.
In the quote above, Lenz also said the current transfer rules probably very often force the smaller professional clubs to sell players in order to ensure their survival, to the detriment of competitive balance.This is not realistic.First, as discussed in section 4, the transfer system is financially beneficial for the small clubs on average, implying their financial problems can be smaller.Second, once a small club has serious financial problems, it can no longer pay the salaries it used to pay both with and without transfer fees.Consequently, it cannot keep all its good players.This inability will be most serious without a transfer system, because a transfer system still gives the club the opportunity to sell some good players for a significant fee, using part of the fees to reduce its debts while using another part to keep some other good players by offering them good salaries.All in all, the economic arguments of Lenz above were weak.

Conclusion of section 4
A transfer fee system improves competitive balance to a significant and possibly substantial extent.

Some modest criticism
Presenting a legal argument that is fully satisfactory to a court can be much more difficult than a legal layman, such as me, may think.In relation to this, none of the arguments below pretends to be such a fully satisfactory argument.Rather, the analysis represents the view of an economist who thinks he may bring some useful points to the table but who also thinks his contribution has little value without a further discussion by legal scholars.
In Section 5, 'Bosman ruling' means Bosman ruling insofar it concerns transfer fees.As discussed in section 1, this ruling was based on the idea that the transfer system was incompatible with free movement law, as formulated in article 48 of the EC Treaty, in principle.However, the ECJ also made clear that if the system improved competitive balance, this could be a reason for an exception to article 48, but only if the system met the conditions of legitimacy and proportionality.According to the ECJ it met the condition of legitimacy, but it did not meet the condition of proportionality for two reasons.First, the transfer system had at best a limited effect on competitive balance.This first reason is not a good one in my view; as argued in section 4, the effect is significant and possibly substantial.
The second reason was that there were other measures for maintaining competitive balance that did not restrict free movement of workers.Now, such measures exist indeed. 33However, they are not implemented (to a sufficient extent) in modern times, because the large clubs, which have much power and prefer to win many prizes themselves, do not want them (if taken too far). 34Therefore, the measures are not available to the small clubs, the national associations and UEFA.Unfortunately, the ECJ did not discuss this problem at all.This is a serious omission in my view; courts should deal with the real world instead of a hypothetical world where power plays no role.
I do not know whether this type of criticism can help justify an exception to article 48.For clarity, section 5.2 will assume that it cannot, and that the pre-Bosman transfer system was incompatible with article 48 without any room for exceptions based on the specific features of the sport.

Why the Bosman ruling was fundamentally flawed
As an introduction to this section, the crucial point in the Bosman case can be repeated once again: the transfer system hampered the free movement of workers and was therefore incompatible with article 48 of the EC Treaty.In relation to this, the ECJ prohibited transfer fees for out-of-contract players.While doing so, it did not quantify the extent to which transfer fees improved free movement, but its qualitative argument was convincing.
In section 4 of this paper, qualitative arguments and empirical data have been used to show that both the pre-Bosman and the post-Bosman transfer fee system improved competitive balance to a significant and possibly substantial extent.Section 2 has shown that an increase in competitive balance leads to more competition in consumer markets.This implies that both the pre-Bosman and the post-Bosman transfer fee system improved competition in consumer markets.Section 4 has argued that the abandonment of transfer fees for out-of-contract players has decreased competitive balance, using a qualitative argument and anecdotal evidence.This implies that Bosman has reduced competition in consumer markets.
Article 85 of the EC Treaty (presently article 101 TFEU) forbids "all agreements between undertakings ….. ….. which may affect trade between Member States and which have as their object or effect the prevention, restriction or distortion of competition within the internal market."The transfer system is an agreement between professional football clubs.These clubs are undertakings (which engage in economic activities and are therefore within the scope of application of EU law). 35The transfer system has as its effect an increase in the competition in consumer markets.Rational decision-making implies that if the ECJ forbids agreements that restrict competition -as it does -it should also do its best to maintain agreements that increase competition, such as the transfer system.Indeed, maintaining agreements that promote competition should be equally important as forbidding agreements that restrict competition, in principle.Here, 'in principle' means that the size of the effects of the agreements is not yet taken into account.
There is no clause in the Treaty which says that the free movement law is more important than the competition law.This means, in my view as an economist, that the negative effects of the transfer system on the free movement of workers should be balanced against its positive effects on the competition in consumer markets.The evaluation should involve estimates of the size and the weights of the different effects, where the weights should reflect the importance of the effects in view of the goals of the Treaty.
Without any pretention to give such a complete evaluation, but just to provide an illustration, a few occasional remarks about the effects can be made here.This paper has shown that the transfer system has a positive effect on competitive balance.More competitive balance improves welfare directly as it leads to more pleasure for the fans.It also improves the competition in consumer markets.This leads to lower prices, and lower prices cause higher welfare.So, the transfer system improves welfare in two ways.Improving welfare is one of the underlying goals of competition law, and from this perspective the transfer system is in line with the spirit of competition law.
However, improving welfare is also one of the underlying goals of free movement law.The transfer system hampers the free movement of players.This will reduce the happiness of part of the players at least (because their income will be lower, and/or because they will play for a club which is not their preferred option), and players' happiness is part of welfare.From this perspective, abandoning (part of|) the transfer system is in line with the spirit of free movement law.
A next question is to which extent exactly the reduction in free movement reduces welfare.In the case of Jean-Marc Bosman, whose career was seriously harmed by the transfer system, the negative effect was large.However, when we focus on all players taken together some questions arise.For instance, if the transfer system reduces the income of players, will their happiness decrease to a relatively small or large extent?Here, it is not impossible that the decrease is relatively small for the richer players.One can even ask if a lower income could make some of the richer players happier, for instance because they would not be spoiled at a too young age.
This brings us to the core question: are the welfare benefits which the transfer system provides to the consumers larger than the welfare reductions for the players?The answer is not certain yet.If it is yes, the transfer system increases welfare on balance and is in line with one of the underlying goals of competition law and free movement law, and vice versa.There may be room for further discussion here, but this might require an effort of legal scholars.
In any case, the main conclusion of this paper remains as follows.The Bosman ruling has neglected the effects of changes in competitive balance on the competition in consumer markets.While the court paid ample attention to the fact that the transfer system reduces the free movement of workers, it neglected the fact that it improves the competition in consumer markets.This means the Bosman ruling is fundamentally flawed.
Finally, it could be useful to investigate whether the argument above has implications for more recent court cases regarding the transfer system, such as the recent case about the home-grown players rule.

Conclusion
For the ECJ and many legal scholars, the Bosman case (insofar it concerned transfer fees) was a contest between the hard free movement law of article 48, and the idea that an exception to free movement law was justified because of the specific features of the sport and the need for competitive balance more specifically.The specific features of the sport lost, at least partly because the EC Treaty had no hard law protecting them. 36his paper has provided a new perspective.The Bosman ruling should have been a contest between free movement law on the one hand, and, on the other hand, the competition law of article 85 (as applied to consumer markets) plus the argument about the specific features of the sport.With hard competition law as its ally, the need for competitive balance could then have been on the winning side, to the benefit of the football fans.