By Farai Jena & Barry Reilly
In a recently published book (Money and Football: A Soccernomics Guide) Stefan Szymanski examines, inter alia, the concept of club dominance in European football leagues over the last 50 years. Szymanski reports, using data from 20 European leagues, that an average of only 10 different clubs have won their domestic league titles over this half century, ‘….way below what you would expect to see if there were balance in the league’. The League of Ireland, providing 15 different champions over this period, is identified by Szymanski as one where the number of different winners is considerably above the average. This could be taken to suggest that it is one of the better balanced leagues in Europe.
There are a number of alternative methods available to determine a league’s long-run competitive balance. One such measure is known as the Herfindahl Index of Competitive Balance (HICB). In an extreme hypothetical case where all teams complete the season with the same number of points (i.e., a perfectly balanced league), the HICB value is one. In contrast, the greater the inequity in the distribution of points across teams at the end of the season, the higher above unity is the HICB and the poorer is the league’s competitive balance.
In order to explore the issue of competitive balance for the League of Ireland we use the HICB to compare its degree of balance with a set of neighbouring leagues in England, Scotland, Wales and Northern Ireland over 16 recent seasons. The leagues are selected on the basis of arguably possessing comparable playing standards to the League of Ireland.
Figure 1 plots the HICB for the League of Ireland Premier Division, the average of tiers three to five in England, and the average of tiers two and three in Scotland. The averages are used here because little material difference in index values is detected across these leagues over the relevant seasons. In contrast to the leagues in England and Scotland, competitive balance is found to be markedly inferior in the League of Ireland. In addition, the league’s index also exhibits a greater degree of volatility over time particularly with respect to its English counterparts.
In a recently published book (Money and Football: A Soccernomics Guide) Stefan Szymanski examines, inter alia, the concept of club dominance in European football leagues over the last 50 years. Szymanski reports, using data from 20 European leagues, that an average of only 10 different clubs have won their domestic league titles over this half century, ‘….way below what you would expect to see if there were balance in the league’. The League of Ireland, providing 15 different champions over this period, is identified by Szymanski as one where the number of different winners is considerably above the average. This could be taken to suggest that it is one of the better balanced leagues in Europe.
There are a number of alternative methods available to determine a league’s long-run competitive balance. One such measure is known as the Herfindahl Index of Competitive Balance (HICB). In an extreme hypothetical case where all teams complete the season with the same number of points (i.e., a perfectly balanced league), the HICB value is one. In contrast, the greater the inequity in the distribution of points across teams at the end of the season, the higher above unity is the HICB and the poorer is the league’s competitive balance.
In order to explore the issue of competitive balance for the League of Ireland we use the HICB to compare its degree of balance with a set of neighbouring leagues in England, Scotland, Wales and Northern Ireland over 16 recent seasons. The leagues are selected on the basis of arguably possessing comparable playing standards to the League of Ireland.
Figure 1 plots the HICB for the League of Ireland Premier Division, the average of tiers three to five in England, and the average of tiers two and three in Scotland. The averages are used here because little material difference in index values is detected across these leagues over the relevant seasons. In contrast to the leagues in England and Scotland, competitive balance is found to be markedly inferior in the League of Ireland. In addition, the league’s index also exhibits a greater degree of volatility over time particularly with respect to its English counterparts.
Notes to Figure 1: The HICB values for England are an average of the third, fourth and fifth tiers of English Football. The HICB values for Scotland are an average of the second and third tiers of Scottish Football. Given the introduction of summer soccer in Ireland in 2003, the years reported on the horizontal axis relate either to the second year of a conventional playing season played across two calendar years or a summer soccer season played within one calendar year.
Figure 2 plots the League of Ireland’s HICB with those for the top tiers in Wales and Northern Ireland. These two close neighbours register a fairly high degree of competitive imbalance in common with the League of Ireland, though the index for the latter is characterised by a greater degree of volatility compared to the former two.
Notes to Figure 2: Given the introduction of summer soccer in Ireland in 2003, the years reported on the horizontal axis relate either to the second year of a conventional playing season played across two calendar years or a summer soccer season played within one calendar year.
Further, and not reported here in detail, the top tier leagues described by Syzmanski as the most plutocratic (viz., England, Germany, Spain and Italy) all register lower average HICB values than the League of Ireland over these 16 seasons, though the most dominated league in Europe (i.e., Scotland’s top tier) averages slightly higher.
A reason why a league’s long-run competitive balance might actually matter is because of its potential relationship with attendance. A poorly balanced league is likely to prove unattractive to spectators. Figure 3 plots average attendance and the HICB values (both standardized) for the League of Ireland over these 16 seasons in order to discern any informative patterns. The plots reveal an inverse relationship between competitive imbalance and average attendance. The correlation coefficient is computed at –0.60 and is statistically significant at the 5% level using a t-test with 14 degrees of feedom. It should be emphasized that this finding is best interpreted as suggestive since nothing informing the causal relationship between these two variables can be inferred from this statistic. However, competitive imbalance and attendance appear to move inversely in this league.
A reason why a league’s long-run competitive balance might actually matter is because of its potential relationship with attendance. A poorly balanced league is likely to prove unattractive to spectators. Figure 3 plots average attendance and the HICB values (both standardized) for the League of Ireland over these 16 seasons in order to discern any informative patterns. The plots reveal an inverse relationship between competitive imbalance and average attendance. The correlation coefficient is computed at –0.60 and is statistically significant at the 5% level using a t-test with 14 degrees of feedom. It should be emphasized that this finding is best interpreted as suggestive since nothing informing the causal relationship between these two variables can be inferred from this statistic. However, competitive imbalance and attendance appear to move inversely in this league.
Notes to Figure 3: The HICB and attendance variables are standardized by subtracting the relevant mean and dividing by the relevant standard deviation to ensure both variables are expressed in comparable units of measurement for plotting purposes.
The domination of a league over time by a small number of teams may suggest long-run competitive imbalance but it does not imply that less dominated leagues by this defintion are competitively balanced. The descriptive evidence reported here covering 16 recent seasons, and over which time seven different teams won the League of Ireland title (suggesting even weaker dominance than that reported by Syzmanski’s analysis), reveals the domestic league is competitively imbalanced when compared to either neighbouring lower tier leagues in England and Scotland or even Syzmanski’s set of plutocratic European leagues. The very casual empirical evidence provided here also suggests that the lack of competitive balance within the League of Ireland might actually matter for attendance.
Farai Jena is a Teaching Fellow in the Department of Economics at the University of Sussex. Her research interests are in the area of applied microeconomics and include the economics of migration, migrant remittance behaviour, and the demand for football.
Barry Reilly is Professor of Econometrics in the Department of Economics at the University of Sussex. His research interest include labour economics and the economics of sports. He has published research on developing country labour markets and on racial discrimination in football.
Farai Jena is a Teaching Fellow in the Department of Economics at the University of Sussex. Her research interests are in the area of applied microeconomics and include the economics of migration, migrant remittance behaviour, and the demand for football.
Barry Reilly is Professor of Econometrics in the Department of Economics at the University of Sussex. His research interest include labour economics and the economics of sports. He has published research on developing country labour markets and on racial discrimination in football.
RSS Feed