“Football is a simple game. Twenty-two men chase a ball for 90 minutes and at the end, the Germans always win.” While being an accurate assessment, this statement by English record world cup scorer and part-time philosopher Gary Lineker hides another truth about football: some Germans win more than others. Thus, the beginning of the new Bundesliga season on Friday will not only spur discussions about tactical details and whether or not the referee has chosen the wrong profession, but also about the justification of skyrocketing footballers’ salaries and whether the earnings are fairly distributed among teams and among players. To shed some light on these issues, I have built a data set that comprises the incomes of all 454 players that stepped on the football field during the 2014/2015 Bundesliga season (the most recent season where good data is available). Most of the income data was obtained from the site fussball-geld.de and some additional estimations and calculations were conducted (See the footnote for a summary of the data collection process). Because players and teams do not disclose their contracts, all data has to be based on expert estimates and market-value approximations and should not be understood as hard facts (In general, think of this as an economist’s version of bar-room philosophy rather than a dead serious academic article) . Nevertheless, they offer a unique way to think about issues of income inequality and its implications.
The chart above shows the annual incomes of the 454 players that played at least in one game in one of the 18 teams that constituted the Bundesliga during the 2014/2015 season (bonuses and advertising revenue are not included). One can easily observe that incomes are distributed not very equally and that talking about “the rich football pros” is a very broad-brush notion. The highest income is 12 million Euros (exhibited by two players from Bayern München), about 286 times the smallest one at approximately 42 thousand Euros (earned by a player from the Hamburger SV). The median income is 800 thousand Euros, the average income is 1.39 million Euros and total incomes amounted to 644.7 million Euros. To measure inequality systematically, there are two popular measures, which have also been discussed in the article on income inequality three weeks ago. The first one is the Gini coefficient which measures the deviation from an equal distribution where zero depicts perfect equality and one absolute inequality. The Gini coefficient of the income distribution of the footballers’ income in the season 2014/15 is 0.51, comparatively high when one considers Germany as a whole which exhibited a Gini index of 0.30 in 2011, according to the World Bank, and 0.31 in 2014, according to Eurostat. When it comes to inequality as measured by the Gini index by international organisations, the Bundesliga rather resembles Brazil or Chile than Germany. To better understand the concept of the Gini coefficient, one can look at the Lorenz curve which plots the share of total income y in dependence on the share of the poorest x*100 percent of the total population.
For example, the incomes of the “poorest” 50 percent add up to about 17 percent of all incomes and the incomes of the “poorest” 83 percent combined amount to 50 percent of the total income that is earned by all players. The Gini coefficient measures the area between the Lorenz curve and the curve representing perfect equality. Another way to measure inequality is to look at income shares of certain top quantiles. The highest earning 25 percent of all Bundesliga players in 2014/15 earn 62 percent of the total income, the top ten percent get 37.7 percent and the top one percent of the players (consisting of five players from Bayern München, one of them only partially weighted) receive 7.6 percent of all incomes. The estimated shares for Germany in 2010 are 39.5 percent of all incomes for the top ten percent and 12.8 percent for the richest one percent (see Alvaredo et al., 2016) and therefore indicate that despite substantial income disparities, the Bundesliga is slightly more equal in terms of incomes than Germany as a whole. This contradicts the findings of the Worldbank and Eurostat, which is also evident from the fact that the Worldbank estimates the top ten percent income share in Germany with 23.7 percent in 2011. This considerable discrepancy is likely caused by the fact that Alvaredo et al. (2016) make use of tax data which usually tends to be fairly accurate while the Worldbank and Eurostat rely on survey data which usually underestimates high incomes due to underreporting and other difficulties (see Moore et al., 2000). This example goes to show how cautious one has to be when dealing with data. Nevertheless, two things are worth pointing out. First, if additional player incomes like bonus payments for won matches and trophies and advertising revenue were to be included in the data, the income inequality displayed by the Bundesliga would most probably be higher than in Germany as a whole, independent of the measure or source that one uses. This is due to the fact that one would expect these additional incomes to predominantly benefit those players who already receive a high regular earning. Second, the “elite” that causes high income inequality in the Bundesliga is a (relatively) bigger subgroup that is more equal within itself (one could draw the line at the upper 15 percent that earn almost 50 percent of the total income; when considering additional income sources, this group might slightly shrink down) than the small top one percent group in Germany that is almost the same to the upper 10 percent that the upper 10 percent are to the total population.
Where do the big income disparities in the Bundesliga come from? Naturally, a lot of it is caused by disparities in performance. For example, Battré et al. (2009) find that performance indicators explain 60 percent of the variance in income and Bryson et al. (2009) determine that both-feet players enjoy a pay premium of more than 50 percent while left-footed players receive a statistically significant premium of 15 percent. These findings can also be supported by the data. After a contract has been signed, the general income of a football player without bonuses is independent of the games that he plays during a season. However, there is a strong correlation between games played and the income of a player.
Despite outliers at 15 and 20 games played, the positive correlation between number of games and income turns out to be highly significant in a simple regression of income on games played: players on average receive EUR 32,657 more income for every game they play. Note that this does of course not imply causation. Rather, it is very likely that players who perform better are deployed more regularly and at the same time are valued higher by the teams and hence receive higher incomes. Another way to test for the effect of performance on income is to analyse forwards with regard to their income and goals shot per game as a simple measure of performance. The following table shows the result of a simple regression of a player’s deviation from team average income on the player’s deviation from the team average of goals shot per games played (Both measures only consider forwards; By using deviations from team average, bias due to team fixed effects is ruled out).
The result is highly significant (at the one percent level) and suggests that a forward’s income in the Bundesliga is 218 thousand Euros higher than the team’s average income of forwards if he is 10 percent more likely to shoot a goal in a game that he plays than the team forwards on average (the average Bundesliga forward shoots 0.167 goals per game in the season 2014/15). It could be the case that goals per game is not a good gauge for performance as also other factors are important for a good football player. However, one can argue that for a forward shooting goals should be one of the top priorities and works as a good indicator of overall performance. While these findings from the data are of course somewhat suggestive and are not meant to be airtight in their derivation (think omitted variable bias and reverse causality!), they do support the theory that a big part of income inequality in the Bundesliga is caused by performance disparities and hence are consistent with the literature. Furthermore, other income drivers like age and top-level experience fit this narrative as well. However, as Lehmann & Schulze (2008) point out, also player popularity plays a role in footballers’ wage determination in the Bundesliga. Their theory can be substantiated by looking at the average income of a Bundesliga player in 2014/15 by position.
One can observe that forwards and midfielders earn on average roughly 200 thousand Euros more than defenders and goalkeepers. Intuitively, it does not seem very plausible that these income disparities are caused by a higher supply of goalkeepers and defenders or by a generally lower productivity. Rather, one possible explanation for them could be that players whose position can not be evaluated with simple statistics like goals and assists might have a weaker negotiation position. Another reason could be that fans favour offensive players and their spectacular scoring abilities and are willing to pay high prices to see them and wear their jerseys which in turn leads teams to pay these players higher salaries.
But even if one considers both performance and popularity effects, one still might wonder if they justify a difference of the factor 286. One possible explanation could be that small differences in productivity (e.g. a slightly better shot or a marginally fresher haircut) can lead to huge differences in income, e.g. due to the fact there are only a handful of players that can regularly appear on posters in football magazines and teams might be willing to pay generously for a slight edge over their opponents. To describe these mechanisms in modern markets that lead to high income inequality Rosen (1981) coined the phrase “economics of superstars”. And while Lehmann & Schulze (2008) find that the Bundesliga does not exhibit economics of superstars and disparities in performance and popularity are remunerated proportionally, they do point out that Lucifora and Simmons (2003) and Garcia‐del‐barrio and Pujol (2007) find evidence for “superstar” incomes for top clubs in other European countries, where they comparably have even more money at their disposal. Hence, the financial upsurge of German top clubs into the European Mount Olympus that could be observed in the recent past (most prominently, Bayern München) could lead to growing income inequality in the Bundesliga due to superstar effects among players as well.
This upsurge also expresses itself in high disparities in the cumulated incomes of the teams’ players. The Gini coefficient that can be derived from the income distribution among teams is 0.388 and therefore probably represents a slightly more equal distribution than among German residents (if we trust the data provided by Alvaredo et al., 2016, rather than the data from the Worldbank and the OECD).
When plotting the cumulated incomes of the teams, a picture presents itself that one already knows from the standings of recent seasons: the FC Bayern München, lonely at the top. Constituting only 5.6 percent of the Bundesliga’s teams, the club’s players earn 22.2 percent of all incomes. It appears as if there is a clear connection between a team’s players’ incomes and the team’s success. However, there are also teams like the notorious Hamburger SV that spend a lot on their players and still fail to rise to the top. Could inequality among a team’s players be detrimental to a team’s success and be one of the causes of rich teams losing?
There is economic evidence that income inequality can hinder growth for countries (see e.g. Berg & Ostry, 2011). In sports, the findings have so far been ambiguous: while Torgler et al. (2006) find evidence that inequality can be harmful for team performance in football, for Basketball, Berri and Jewell (2004) and Simmons & Berri (2011) find no or even a small positive effect of income inequality. For the 2014/15 Bundesliga season it is hard to tell from visual evidence. The most unequal teams are the FC Schalke 04 and the Hamburger SV and the most equal team is the 1. FC Köln. It appears that, generally, rich teams also tend to be more unequal. To test the effect of inequality on team success, consider a simple regression of the points a team won during the season on its Gini coefficient as a measure for inequality, controlled for the effect of its players’ total income (Do not do this at home: the sample size here is normally considered too small).
Total income is highly significantly positively correlated with points per season (Again, this is not testing for causation due to the lack of identifying assumptions). While the results indicate a negative correlation of a high Gini index with fewer points, this is not a significant effect and might be due to noise in the data. In contrast to the whole economy, in team sports, considerably high income inequality might not be detrimental to success due to various factors. The economic mobility might be higher and hence players with low incomes be more motivated to perform well in order to rise in the income distribution. Furthermore, while income inequality in the Bundesliga is probably higher than in the whole German economy, the inequality is more likely to be solely based on real performance and other factors that are important to teams like popularity and leadership qualities and hence less demotivating for players with comparably low incomes. There is less room for the players to shroud their true productivity with million football “experts” watching their every move and increasingly advanced statistics to measure their performance (This “controlled” setting – additionally with clear rules enforced by referees – is also the reason why a lot of economists are into analysing sports so much).
But while high income inequality among players within teams might not be harmful for team success, the high level of economic inequality among teams might be harmful for a majority of Bundesliga fans who have to endure boring seasons or have to constrain their hopes to one of the ranks behind Bayern München. The high inequality among teams could be curbed e.g. by the introduction of salary caps and regulated drafts of new talents based on the model of North American professional sports leagues (it’s quite interesting that sports seems to be the only department where US Americans prefer more restrictive labour markets than Europeans). On the other hand, in Germany, a high concentration of team spending might be necessary to have international success, as the Bundesliga as a whole earns less from TV rights and other licenses than for example teams in England or Spain (Do you think any German club would have been willing or even able to pay 105 million Euros for Paul Pogba?).
In the mean time, the development towards an ever-increasing income inequality in the Bundesliga most likely goes on with players like Mats Hummels going to Bayern München (where he allegedly now earns about 10 million Euros). The fuel does not seem to be running out for fiery conversations about economic inequality in the Bundesliga, coming to a sports bar near you. Friday will be a first test of how exiting this new season is going to be when Werder Bremen faces Bayern München, the main driver of income inequality among players and teams in the Bundesliga and at the same time the beacon of hope for German aspirations on the international club stage. Let’s hope that the Bundesliga cannot only teach us something about income dynamics but also bestows us richly with goals, close games and German success in the Champions league!
The data on team affiliation, games played, goals scored and position were obtained from the site kicker.de, while the data on annual income for most players can be found on the site fussball-geld.de that gathers financial data on various Bundesliga topics, summarising news reports and conducting own approximations. For 43 players where no income data could be found on fussball-geld.de, data on income came from various news reports or were approximated with player market values as provided by the site transfermarkt.de. For this, I divided a team’s players’ total income by the team’s players’ total market value and thus obtained a team specific income-to-market-value ratio that then was multiplied with a certain player’s market value. This approach is consistent with the approach used by fussball-geld.de. For players where this market-value approach resulted in a lower approximated income than the lowest team intern income as provided by fussball-geld.de, this team-intern lowest income was used instead in order to not introduce an upward bias of inequality levels. The data set for players includes 454 observations while the data set for team analysis includes 461 because a few players played in multiple teams during the 2014/15 season.
Alvaredo, Facundo, Anthony B. Atkinson, Thomas Piketty, Emmanuel Saez, and Gabriel Zucman. (2016). The World Wealth and Income Database. http://www.wid.world, 21/07/2016
Battré, M., Deutscher, C., & Frick, B. (2009). Salary determination in the German Bundesliga: a panel study. In No 0811, IASE Conference Papers, International Association of Sports Economists.
Berg, A., & Ostry, J. (2011). Equality and efficiency. Finance & Development,48(3), 12-15.
Berri, D. J., & Jewell, R. T. (2004). Wage inequality and firm performance: Professional basketball’s natural experiment. Atlantic Economic Journal,32(2), 130-139.
Bryson, A., Frick, B., & Simmons, R. (2009). The returns to scarce talent: Footedness and player remuneration in European football. Unpublished manuscript.
Lehmann, E. E., & Schulze, G. G. (2008). What does it take to be a star?-The role of performance and the media for German soccer players. Applied Economics Quarterly, 54(1), 59-70.
Moore, J. C., & Welniak, E. J. (2000). Income measurement error in surveys: A review. Journal of Official Statistics, 16(4), 331.
Rosen, S. (1981). The economics of superstars. The American economic review, 71(5), 845-858.
Simmons, R., & Berri, D. J. (2011). Mixing the princes and the paupers: Pay and performance in the National Basketball Association. Labour Economics,18(3), 381-388.
Torgler, B., Schmidt, S. L., & Frey, B. S. (2006). Relative income position and performance: an empirical panel analysis.
Written by Jonas