The economics of sport

The claim that professional team sports are no different from conventional industries can be challenged; and whether club owners are profit maximisers or utility maximisers remains an area of debate

By Peter J Sloane

The economics of sport is some 50 years old. Simon Rottenberg’s paper on the baseball players’ labour market in the Journal of Political Economy (1956) is generally recognised as the first academic paper to conduct a formal economic analysis of the industry.

There are, in fact, three distinct areas of study: first, professional team sports such as football, cricket, rugby, baseball, hockey and basketball; second, non-team sports such as athletics, tennis, boxing, golf, horse-racing and motor sports; and third, amateur or participation sports. Economists have tended to focus most attention on professional team sports where the allocation of players among different clubs competing in leagues has led to various forms of anti-competitive behaviour designed to achieve some degree of competitive balance. The peculiar economics of such arrangements derive from the fact that a single team can produce no output on its own.

However, economists have also observed cartel behaviour in action with sports leagues, which is rarely possible in more conventional industries because of their illegality under competition laws. Further, one can also measure productivity of players much more easily than is the case for workers elsewhere, because of the availability of detailed statistics on athletic performance. Non-team sports have been analysed much less, presumably because their allocation problems are less severe.

However, undue dominance may be a problem too, as was the case in Formula One motor racing when Michael Schumacher was winning most races. Amateur sports involve different issues, being regarded as merit goods because of positive externalities in relation to health and, therefore, subject to government subsidies. Yet, professional sports, more debatably, have also benefited from government assistance. All these activities, however, amount to less than two per cent of GDP in most countries – clearly, the sector is not particularly large in relation to gross national product.

This paper follows Rottenberg in focusing on professional team sports, but also notes the distinction in the ensuing literature between what is referred to as the North American and European models of professional team sports. This literature is extensive as reflected in the fact that there is now an International Association of Sports Economists, the Journal of Sports Economics and several textbooks on the economics of sport.

The North American profit-maximising approach

What, then, is Rottenberg’s legacy? The defining feature of his 1956 paper was a belief that the economics of professional sports leagues could be analysed using the same economic framework as for any other industry. But he did note that there were two unusual features. First, monopsony power (which refers to buyer monopoly or control over the player through contracts restricting the ability to move to other clubs) takes an extreme form. Second, competitors must be of approximately equal size if any of them are to be successful. That is, in order to produce a successful product, differences in the quality of rivals should not be ‘too great’. Fans value uncertainty of outcome.

Rottenberg believed that baseball team owners were rational profit maximisers. If one team became too strong it would be in its own interests to allow players to move to other clubs where their output would be more valuable. Hence, he claimed that the free market would be just as efficient as the reserve clause, which limited player mobility, in terms of resource allocation, through what has been referred to as the ‘invariance principle’. Whether clubs do, indeed, attempt to maximise profits, the extent to which uncertainty of outcome matters and whether the invariance principle holds have been much debated in the ensuing literature.

The invariance principle is a version of the Coase theorem, which states that in competitive markets and given that resources are freely exchangeable, the distribution of their ownership is irrelevant in ensuring that they are used efficiently. However, the assumptions of this model are quite restrictive – perfect competition, perfect information, the absence of transactions costs and the absence of wealth and income effects. It is doubtful whether these assumptions hold in the context of a professional sports league in which clubs are local monopolies and particularly so where clubs are of different sizes.

The European utility-maximising approach

I challenged the Rottenberg thesis in two papers written in 1969 and 1971. Based on the experience of English professional football, I noted that loss-making was endemic and clubs seemed to pursue sporting success to the detriment of profits. I suggested a utility maximising model which assumed that sportsmen-owners were more driven by the urge for power or the desire for prestige than in making money, thus treating a sports club as a consumption activity by putting particular emphasis on playing success. By and large, European sports economists have followed such a model by assuming that clubs attempt to maximise playing success subject to a break-even constraint, whereas in North America, where there is also some recognition that sportsmen-owners do exist, the profit maximisation assumption has prevailed.

Whether clubs follow one or the other objective has implications for the impact of revenue sharing. Since clubs attempting to maximise playing success subject to a break-even constraint will always spend additional income on improving playing performance, more equal revenue sharing will always improve competitive balance. If, however, clubs follow profit maximising behaviour, this outcome is less certain, though more equal revenue sharing is likely to increase the profitability of poorer clubs. The pursuit of playing success may well result in reduced competitive balance compared to the pursuit of profit maximisation, since if some clubs are successful in winning more games the zero-sum nature of competition means that other clubs will lose more games and likely incur losses. The pursuit of playing success at the expense of profits may then tend to destabilise the league from an economic perspective and increase the importance of revenue sharing or other forms of cross-subsidisation as a mechanism for attaining competitive balance.

The importance of uncertainty of outcomes

The question of whether uncertainty of outcome is important in maintaining fan interest has been the subject of much investigation. Rottenberg suggested that attendances would be a function of the general level of income in a team’s locality, the price of admission, the goodness of leisure substitutes, population size, the location of the stadium, average league standing of the team (i.e. team quality) and the dispersion of the percentage of games won in the league by different teams (i.e. uncertainty of outcome). By and large, subsequent studies have included such variables in regression models designed to estimate and forecast attendances, in a range of team sports, though with variable success, particularly in relation to the uncertainty of outcome variable.

However, uncertainty of outcome can be measured in different ways and may refer to short-run uncertainty of match outcome, seasonal uncertainty in terms of how many teams remain in contention for the title or long-run domination in terms of whether the same team (or teams) wins the title in successive seasons. Few studies include all these measures or interact them. Uncertainty of match outcome also needs to allow for home advantage, so the most even match will occur when the away team is somewhat stronger than the home team. Also, season ticket holders make their decision prior to the start of a season and this must be based on the performance in the previous season and expectations prior to the start of the current season, whereas those who purchase on the day will be influenced by recent performances. A further problem relates to stadium capacity. If there is a full house then one is not able to observe the true demand curve. For such reasons it is perhaps not surprising that the uncertainty of outcome variables used in attendance models have not always been significant and where they are significant the effects are often small. Improving the measurement of such variables is an important area for future research.

Revenue sharing arrangements

The actual arrangements for revenue-sharing differ across the various sports. In North America, the National Football League (NFL) is the most equal with gate revenue divided 60 per cent to the home team and 40 per cent to the away team. In baseball, the division is 80:20, and in the National Basketball Association (NBA) it is 100:0. Accordingly, it has been found that competitive balance is higher in the NFL than in the NBA. In Europe, 100:0 is the norm and it is interesting to speculate why gate revenue sharing should be less pronounced in Europe than in North America.

The major source of revenue in most sports today is television. In North America, collective selling of television rights was sanctioned through the passing of the 1961 Sports Broadcasting Act, which enables the major leagues to divide national television income equally among the clubs. The current NFL contract is the richest in the world, amounting to US$3 billion per annum.

In Europe, the collective selling of television contracts is a more contentious issue. In Spain and Italy, such arrangements are outlawed, whereas political intervention in France and Germany has permitted them. In England, the English Premier League was successful in defending collective selling before the Restrictive Trade Practices Court, but subsequently the European Union has insisted that the product must be offered for sale in a number of bundles to allow more than one broadcaster to obtain the rights. The English Premier League shares 50 per cent of its television income equally among member clubs – 25 per cent on the basis of league position (positively) and 25 per cent on the basis of a facilities fee related to the number of times a given club’s games are televised. While this is claimed to be relatively equalising, the growth in the relative importance of this source of income means that it has become less equal over time. Thus, according to Deloitte (Annual Review of Football Finance, 2004, London), sources of revenue of English Premier League clubs in 1992 were £82 million from gate receipts, £15 million from broadcasting and £73 million from commercial activities. By 2003, these figures had risen to £363 million, £543 million and £340 million respectively, representing growth ratio of 342 per cent for gate income, 3520 per cent for broadcasting income and 365 per cent for commercial income. Television income now dominates other sources of income.

This has impacted strongly on the amounts shared amongst the clubs. While the ratio of television income of the top club to the bottom club rose from 2.2 times in 1992/3 to only 2.4 times in 2003/4, the absolute difference increased from £1.3 million at the start of the period to £19.4 million at the end of it. The gap will be even greater under the most recent three-year contract in operation from 2007-2010, which totals £2.7 billion. This guarantees a Premier League team at least £40 million per annum and increases the gap between clubs in the Premier League and those in the Football League from which promotion and relegation operates for the three bottom and top teams respectively. Further, the top three or four Premier League Teams gain entry into the Union of European Football Association (UEFA) Champions League (and the teams just below to the UEFA Competition), substantially increasing the income of the top clubs relative to the rest. Thus in 2004 the top three clubs – Manchester United, Arsenal and Chelsea – obtained a total of £56 million from European Champions League Broadcasting revenue.

Salary caps

This has caused UEFA to consider the introduction of salary caps, even though these would be difficult to administer in a European context, given the differences in the sizes of clubs across participating countries. These are commonplace in North America and amount to a form of revenue-sharing between teams and players. They are not to be confused with maximum wages for individual players as existed in the English Football League up to 1961 and which have been applied in the NBA for certain players. In the case of salary caps, the parties agree upon defined gross revenues for sharing purposes and what share will go to the players.

In some cases there are ‘hard’ caps, which means equal wage bills without exceptions, and in other cases ‘soft’ caps, which may allow exceptions (e.g. free agents who re-sign for the same club do not count against the cap). A minimum of spending close to the cap may also be required to prevent owners from simply pocketing the wealth transfer, though this would seem to be less necessary if clubs are win maximisers subject to a break-even constraint. Regardless of the objective function of the club, there is a strong probability of cheating, since clubs are prevented from being at their preferred position in terms of winning percentages. There are examples of this in all sports that have implemented such arrangements.

Summarising the differences

We have already pointed to certain similarities and differences between North American and European leagues, often referred to as the North American and European Models. These may be summarised as follows.

In relation to objectives in the North American Model, profit maximisation is regarded as the norm, while in the European Model the norm is utility maximisation subject to a profit constraint.
In North America there are closed leagues with each club given a franchise, whilst in Europe there are open leagues with promotion and relegation. Each has its advantages and disadvantages. Open leagues are likely to raise team quality as the fear of relegation with its consequent financial costs will force lower teams to strengthen their squads, whilst the possibility of promotion will encourage the better teams in the lower league to do likewise. Promotion and relegation will also raise fan interest as a higher proportion of games will matter in terms of teams being in contention for either league champion?ship or relegation. However, open leagues make it more difficult for clubs to choose their preferred win-loss ratio and force a departure from an optimal distribution of teams, as teams of higher quality than those promoted may be relegated. Comparison of the relative merits of the two types of league is, however, difficult as there is a relative over-provision of clubs in Europe relative to North America.
This leads to a third difference, namely size, as there are a restricted number of clubs per head of population in North America compared to Europe.
A fourth difference relates to the geographical pattern of clubs. In North America, clubs have exclusive territories, but franchise mobility, whilst in Europe there is restricted geographical movement of clubs. This enables clubs in North America to move to other locations where potential attendances are perceived to be greater, though they may have to compensate other franchises in order to do this.
A fifth difference is the role of international competition, which is important at both club and national level in Europe, but absent in North America. This has led to conflicts in Europe between club and country, as players may get injured playing for their national team without compensation to the club. Participation in international club competitions also means that it may be necessary for clubs to be ‘too strong’ for their domestic competitions in order to challenge for international competitions, a possibility ignored by Rottenberg in his analysis of baseball.
In North America, player drafts are commonplace whereby clubs are entitled to pick the most promising young players in reverse order of team finish in the previous season in order to improve competitive balance. Such arrangements are absent in Europe.
The sale of players for cash is commonplace in Europe, whereas restrictions on sales of players for cash apply in North America.
Roster limits on the number of players contracted to an individual club are extensively used in North America, but are absent in Europe.
As mentioned earlier, a ninth difference is the more extensive use of revenue sharing in North America than in Europe.
Salary capping is also more extensive in North America than in Europe, though here the gap appears to be narrowing.
Finally, in Europe there has recently been a trend for clubs to float on the stock market, while in North America this possibility is restricted. Interestingly, there is no evidence that this has changed objectives more towards profit maximising behaviour and the payment of dividends to shareholders has been infrequent. Further, a number of clubs have been taken over and subsequently delisted. This has not prevented some economists from suggesting that the two models are converging, a process that may be accelerated by the takeover of several English Premier League clubs by American entrepreneurs.

Where does Australia fit in?

It is interesting to speculate whether there is a distinctive Australian model. Australian sports economists have tended to favour the utility maximisation model preferred in Europe over profit maximisation, but in other respects Australia seems closer to the North American model with closed leagues, player drafts, extensive revenue sharing and soft salary caps.

What is clear, however, is that in all three cases there has been a conflict, particularly in the labour market, concerning restrictions on the freedom of the players to move from one club to another, both in relation to the law and competition policy. In North America, baseball escaped from the rigours of the anti-trust legislation though a 1992 Supreme Court decision, widely criticised, that Major League Baseball was exempt from such policy as it was not engaged in interstate commerce. A negative court case in relation to the collective selling of broadcasting rights was overturned by the passing of the 1961 Sports Broadcasting Act and the reserve clause was only weakened substantially as a result of collective bargaining in the 1970s. In Europe, the Advocate General in the 1995 Bosman case before the European Court of Justice found that the existing retain and transfer system was unlawful as the Treaty of Rome required free mobility of labour, and following discussions between EUFA and the European Union a modified form of player contract was introduced.

The European Union stance is based on these general principles: to take account of the special characteristics of sport, to apply the competition rules in a manner that does not question the regulatory authority of sporting organisations with respect to genuine sporting rules, and to preserve the social and cultural functions of sport. It is true to say however, that European Union policy has been much more interventionist than anti-trust policy in North America. In Australia, a number of court case decisions have found restrictions on player mobility to be in restraint of trade, illustrating the common problems facing sporting organisations in attempting to restrict competition in order to assist competitive balance.

Assessing Rottenberg

In conclusion, Rottenberg’s analysis has had a major impact on the ensuing literature on the economics of sport, but it is based on the assumption that competitors are of approximately equal size, which is less true in Europe than it is in North America. His claim that professional team sports are inherently no different from conventional industries can certainly be challenged and whether club owners act as rational profit maximisers remains an area of considerable debate. The highly visible nature of professional sports leagues is likely to keep them in the spotlight as far as public policy is concerned.

Perhaps, however, there is scope for more economic analysis of non-team sports and participatory sports, which have been relatively neglected by economists in the past.

Top ^

A condensed version of a public lecture jointly organised by the Department of Economics and the Melbourne Institute of Economic and Social Research delivered at the University of Melbourne on 17 May 2007. The full paper appeared in Rodriguez, P., Kesenne, S. and Garcia, J. (eds), Sports Economics After 50 Years: Essays in Honour of Simon Rottenberg, University of Oviedo Press, 2006.

Professor Sloane FRSE, is Research Professor and Director of WELMERC, a centre for labour market research at the University of Wales, Swansea. He is Vice-President of the International Association of Sports Economics and member of the editorial board of the Journal of Sports Economics.

Volume 2 OCT 2007