Competitive Balance in Football: Trends and Effects: September 2004
Competitive Balance in Football: Trends and Effects: September 2004
Competitive Balance in Football: Trends and Effects: September 2004
discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/283569348
CITATIONS READS
5 1,257
2 authors:
Some of the authors of this publication are also working on these related projects:
All content following this page was uploaded by Christine Oughton on 08 November 2015.
Jonathan Michie
Christine Oughton
Jonathan Michie1
Christine Oughton2
Abstract
It is well known that sports leagues need a certain degree of competitive balance in
order to survive and flourish. Without uncertainty over the outcome of individual
matches and the league championship, sports leagues can become predictable and
boring. Lack of competitive balance means that the number of spectators is not
maximised and that the league runs the risk of losing spectators in the long term.
Unbalanced leagues also result in other risks, such as: the bankruptcy, or threat of
bankruptcy, of lagging clubs; the threat of rival leagues; and increased risk as a result
of large income gaps within and between leagues. This paper focuses on the following
research questions: What has happened to competitive balance over the past 50 years
or so in the big five European leagues and has there been a marked change in recent
years? What determines competitive balance? What is the impact of competitive
balance on the long-term performance of the football industry? In the light of our
analysis we make a number of policy recommendations.
Acknowledgments
Thanks are due to Matthew Holt, Lee Shailer and Geoff Walters for valuable
assistance with data collection. We are grateful to Gerry Boon and Austin Houlihan
at Deloitte for providing financial data and to the Premier League and UEFA
Professional Football Division for providing data on the distribution of broadcasting
revenues. We would also like to thank Philip French at the Premier League and Alex
Phillips at UEFA for useful comments and suggestions. We are also grateful to
SportsNexus for sponsoring this research. The views expressed and any errors are
those of the authors alone.
1
Professor of Management and Director of Birmingham Business School, University of Birmingham
Birmingham, B15 2TT, j.michie@bham.ac.uk
2
Professor of Management and Director of the Football Governance Research Centre School of
Management & Organizational Psychology, Birkbeck, University of London, London WC1E 7HX
c.oughton@bbk.ac.uk
Competitive Balance in Football: Trends and Effects
1. Introduction
English football has gone through something of a transformation over the past 15
years or so. Investment in stadia has increased - albeit with financial help from the
Football Trust3 - attendances are up, revenue from a variety of sources has been rising
and there are encouraging signs that a number of clubs are adopting improved
methods of corporate governance.4 At the same time, it is evident that there are
important issues that the industry has to address to ensure its long-term success.
Prime amongst these is the question of competitive balance.
It is well known that sports leagues need a certain degree of competitive balance in
order to survive and flourish. Without uncertainty over the outcome of individual
matches and the league championship, sports leagues can become predictable and
boring. Lack of competitive balance means that the number of spectators, both match
going fans and those watching televised matches, is not maximised and that the league
runs the risk of losing spectators in the long term. Unbalanced leagues also result in
other risks, such as:
1. the bankruptcy, or threat of bankruptcy, of lagging clubs;
2. the threat of rival leagues being formed by top clubs that seek competitive
balance in a new or rival league – suggestions of a European Super League
and the Scottish case provide examples of the possible threat to league
stability from new and rival leagues; and
3. the creation of large income gaps within and between leagues that result in
huge financial gains from making it into the European Champions League
or the Premier League. These gaps increase business risk by producing a
financial incentive structure that encourages clubs to gamble on success.
The recent experience of Leeds United provides an example of what can
happen should the gamble fail.
3
Between 1990 and 1998 the Football Trust (now the Football Foundation) provided £190m of grant
aid for the modernisation of stadia (Deloitte and Touche, 1998, p. 80).
4
See FGRC (2003) for a review of corporate governance in football from 2001 to date.
1
Even a cursory look at competitive outcomes in the Premier League suggests that
there may be a lack of competitive balance. A simple indicator that most football
supporters of Premiership clubs will be familiar with is ‘domination’, reflected in the
fact that since the formation of the Premier League 12 years ago, the title has been
won by one club (Manchester United) in 8 out of 12 seasons and by just two clubs
(Arsenal and Manchester United) in 11 of the 12 seasons. However, long-term
domination is not a new phenomenon, unique to the Premier League. There have
been other periods in the past where the top flight has been dominated by one club,
such as Liverpool’s run in the 1970s and 1980s when the club finished 1st or 2nd in 14
out of 15 seasons between 1973 and 1987, winning the league in 10 of those seasons.
The above discussion illustrates that while there has been talk of a decline in
competitive balance, much of the evidence is anecdotal and there has been little
statistical analysis of exactly what has happened to competitive balance over time,
particularly since the formation of the Premier League. The few studies that have
5
Cairns (1987) identifies three types of uncertainty of outcome: match uncertainty; seasonal
uncertainty and inter-seasonal uncertainty or lack of dominance across seasons.
2
been carried out conclude that there is no clear trend.6 However, these studies are
based on limited and inadequate measures of competitive balance, as we explain
below. In short, we have lacked the necessary data to understand exactly what has
been happening to competitive balance over the past 15 years or so and how this
compares with previous trends. The current study addresses this gap by focussing on
the following three questions:
1. What has happened to competitive balance over the past 50 years or so and has
there been a marked change in recent years?
2. What determines competitive balance and in particular what is the relationship
between the distribution of revenue and competitive balance?
3. What is the impact of competitive balance on the long-term performance of
the football industry?
The paper is organised as follows. Section two provides a statistical analysis of trends
in competitive balance using a variety of indicators. The indicators include a number
of new measures that capture both seasonal uncertainty of outcome and dominance.
We focus on seasonal measures of competitive balance for two reasons. Firstly,
empirical studies suggest that seasonal uncertainty of outcome has a more robust
effect on attendance than match uncertainty (see Borland and MacDonald, 2003).
Secondly, by analysing seasonal competitive balance over successive years we gain
an understanding of what has happened to competitive balance over time. This
enables us to answer questions such as, ‘Are the top 5 clubs becoming more or less
dominant?’ Thus, we focus on trends in seasonal competitive balance. Using this
approach our results show that competitive balance has declined significantly over the
past 10-15 years.
In section three, we look at the factors associated with this decline, including the
relationship between the distribution of revenue, expenditure on wages and
competitive balance. This section also considers the impact of declining competitive
6
For example, Szymanski (2001, p. F76) looks at the standard deviation of win percentages and
concludes that there is no trend rise or fall in competitive balance.
3
balance on the medium to long-term performance of the football industry. Our
conclusions and recommendations are set out in section four.
Competitive balance refers to balance between the sporting capabilities of teams. The
more evenly balanced the competitive strengths of the teams that make up a league,
the more uncertain the outcome of each match. Similarly, the more evenly balanced
the teams, the more uncertain the outcome of the championship race. In a perfectly
balanced league each team would have an equal chance of winning each match and
each team would therefore have an equal chance of winning the league title.
Moreover, in a perfectly balanced league it would be impossible to predict with any
certainty which teams would be more likely than others to win the league title next
year or the year after. It follows that there are three dimensions to competitive
balance: match uncertainty, seasonal uncertainty, and dominance or long-term
uncertainty.7 However, each of these three dimensions has the competitive balance of
the teams and uncertainty of outcome of the individual matches at its core. If there is
uncertainty of outcome over individual matches, then there is uncertainty of outcome
over the league title both this year and in future years.
7
See Cairns (1987).
8
See Borland and MacDonald (2003) for a review of empirical studies that explore the relationship
between seasonal uncertainty of outcome and attendance. All the studies of seasonal competitive
balance in soccer reviewed in the article find a positive and significant effect of competitive balance on
attendance.
4
product. This is part of the business logic behind sports leagues adopting regulatory
rules to redistribute income and promote competitive balance. Competitive balance is
also important to ensure league stability. Unbalanced leagues face increased risks of
bankruptcy of lagging clubs and threats of league break-up from new or rival leagues.
Price
D2
D1
Number of Spectators
5
whole.9 In order to guard against this instability, as well as the threat of bankruptcy of
lagging clubs, or of breakaway leagues being formed by leading clubs, virtually all
sports leagues redistribute income from the most successful to the least successful
clubs. This redistribution is not a result of the benevolence of the top clubs but rather
a result of the fact that redistribution is necessary to prevent the league from
becoming unbalanced and to maintain and increase demand: redistribution is
functional in that it enhances competitive balance, attracts more spectators and
viewers and preserves the integrity of the league.
The higher this measure (SDW), the lower the degree of competitive balance in the
league. Estimates of competitive balance based on this indicator show that there has
9
These essential characteristics of sports leagues make the economics of professional sports leagues
‘peculiar’. The peculiarity stems from interdependence between clubs. There is a vast literature on the
peculiar economics of professional sports leagues, much of it dating from Neale’s (1964) seminal
article. For surveys of the literature see: Cairns, Jennett and Sloane (1986) and Szymanski (2003).
6
been no clear trend in competitive balance in over time (Szymanski, 2001, Buzzacchi,
Szymanski and Valletti, 2003).10 However, there are three major limitations of this
approach. The first is that the above measure was derived with US major sports
leagues in mind where drawn matches are rare or non-existent. This is not the case in
football where drawn matches are commonplace. Given that the index has not been
designed with drawn matches in mind, it is not the most appropriate index to measure
competitive balance in football leagues where the number of wins per season varies.
A second problem is that SDW is based on the standard deviation of win percentages
from the mean, which is one measure of dispersion or variation, however, this
measure does not take into account other aspects of the distribution of win
percentages, such as, whether or not it is always the same group of clubs that are
winning (i.e. is there dominance)?11 A further problem is that the measure has little
intuitive appeal since the standard deviation of win percentages is not bound to lie
within a range that is easy to interpret. In particular, measures of competitive balance
based on the standard deviation of win percentages do not neatly capture the balance
of competitive power between leading clubs (or championship contenders) and the
rest, and yet this is one of the features of competitive balance that matters to fans in
terms of seasonal uncertainty of outcome. A review of studies based on the SDW and
other measures is provided in Appendix Table A.1 which shows that the few studies
that focus on changes in seasonal competitive balance in football over time have
typically used the SDW measure or variants of it.
10
For example, Szymanski (2001, p. F76) looks at the standard deviation of win percentages between
1977-98 and concludes that there is no trend rise or fall in competitive balance, while Buzzacchi et al
present 10 year averages of the SDW for 5 decades and find this average rose in the 1960s, 70s and 80s
but fell in the 1990s to a level below the average for the 1970s.
11
See Humphreys (2002) who shows that the SDW indicator is insensitive to dominance.
12
Also known as the Herfindahl-Hirschman index of concentration
7
While all of these measures are, to the best of our knowledge, new measures of
competitive balance, they have been adapted from indexes that have been widely used
in industrial studies to measure the competitiveness of industries (see Davies 1979)
and market power. The Lorenz Curve has also been used in studies that measure
income inequality; it has one major advantage over the other two measures in that it is
insensitive to changes in the number of clubs that make up the league.
2.3 The Five-Club Concentration Ratio (C5) and Index of Competitive Balance
(C5CIB)
In a standard industry, the five firm concentration ratio measures the extent to which
an industry is dominated by the five largest companies. When applied to football, the
five club concentration ratio measures inequality between the top five clubs and the
rest of the league and may be calculated using the following formula:
5
= ∑s
i =1
i
where si is the share of points of the ith club. The C5 ratio is a function of the number
of firms or clubs in an industry or league and the degree of inequality between the top
5 clubs and the rest where increases in the index reflect a reduction in competitive
balance and increased domination by the top five clubs (see Davies, 1979 and Hart,
1975 for a discussion of the impact of inequality and the number of firms (clubs)). In
a standard industry, with no restrictions on the number of companies or the market
share of each company, the index would lie between 0 (reflecting pure competition
with an infinite number of companies) and 1 (pure monopoly). However, in football,
the fact that the number of clubs is fixed and that it is impossible (given the points
scoring system) for the top 5 clubs to win all the points, the bounds of the index lie
between 5/N (where N is the number of clubs) and M/(M + T) where M is the
maximum number of points attainable by the top 5 clubs and T is the absolute
minimum number of points that the remainder teams could end up with: since they
8
have to play each other, they must gain some; if it were possible for them to gain none
then the upper bound would of course be 1, with the number of points won by all
clubs being the same as the number of points won by just the top five clubs. For a 20
team league the C5 ratio lies between 0.25 and 0.55.13 Since the number of clubs is
normally constant (except for years where the size of the premiership and old first
division changed) changes in the C5 ratio reflect changes in inequality and dominance
of the top 5 rather than changes in the number of clubs.
Data for the C5 ratio are presented in Figures 2.2 and 2.3 for the period 1947-2004
and 1993-2004 respectively. It can be seen that over the period 1947-2004 the five-
club concentration ratio increased significantly and that most of this increase occurred
during the 1990s. Taking the period as a whole there was a 21 per cent increase in C5
between 1947 and 2004, however this change includes a change in the index caused
by a fall in the number of clubs from 22 to 20. Changes in the number of teams
clearly affect the competitiveness of the league. However, it is also instructive to
abstract from these changes. The most straightforward way to abstract from changes
in league size so as to focus on changes in equality is to compare years with equal
numbers of teams. Using this comparison reveals that between 1947 and 1987 the C5
ratio was unchanged: however, between 1989 and 2004 the index rose by 6.4 per
cent.14 These changes are significant given the constraints imposed on the value of
the index by the points scoring system. Thus, inequality between the top 5 clubs and
the rest was roughly constant over the period up to 1987 but has experienced a
significant increase since then.
13
M = 510 (top teams winning all matches against lower ones), and T = 420 (all draws).
14
Between 1990 and 2004 the rise was 10 per cent.
9
Figure 2.2 Dominance of the Top 5 Clubs
Percentage share of points held by top 5, C5 (%)
40
38
36
34
32
30
28
26
24
48
52
56
60
64
68
72
76
80
84
88
92
96
00
04
19
19
19
19
19
19
19
19
19
19
19
19
19
20
20
Figure 2.3 The Dominance of the Top 5 Clubs in the Premier League:
Percentage Share of Points held by Top 5 Clubs, C5 Ratio (%)
39
37
35
33
31
C5 Ratio
29
27
25
93
94
95
96
97
98
99
00
01
02
03
04
19
19
19
19
19
19
19
20
20
20
20
20
10
which is normally deflated by the standard deviation that would be attained in a
perfectly balanced league. This adjustment gives us the C5 Index of Competitive
Balance (C5ICB) as shown below.15
where N is the number of teams in the league. Thus, the C5 Index of Competitive
Balance is adjusted to correct for changes in the size of the league. For a perfectly
balanced league of any size the index takes the value of 100. Reductions in
competitive balance are reflected by an increase in the ratio. For example, if there
were a 25 per cent decline in competitive balance the index would take the value of
125. Data for the C5 Index of Competitive Balance are presented in Figure 2.4.
150
145
140
135
130
125
120
115
110 C5 Index of Competitive Balance
105
100
47
51
55
59
63
67
71
75
79
83
87
91
95
99
03
19
19
19
19
19
19
19
19
19
19
19
19
19
19
20
The five-club concentration ratio looks at inequalities or imbalance between the top 5
clubs and the rest. However, it does not capture changes in imbalance within the top
15
Note that this index can be used to capture dominance by calculating the share of points of the top 5
over a number of years, for example 5 years.
11
5, or within the bottom 15 clubs (or 17, in a 22-team league). The Herfindahl index
(named after the economist who developed it) looks at inequalities between all the
firms in an industry. When applied to football it captures inequalities between all the
clubs that make up a league. In an industry context the index is based on a calculation
of the market share of every firm. These shares are then summed into a weighted
average index for the industry using each firm’s market share as its weight. We can
translate this into an indicator of competitive balance for the football industry by
looking at each club’s share of points in a season and aggregating these into an index
using each club’s share of points as weights, to give:
N
H = ∑ si2
i =1
Where si is club i’s share of points in a season, and i = 1, 2, ……N, where N is the
number of clubs in the league.16 Like the five-club concentration ratio, the H-index is
a function of the number of clubs that make up the league and the inequalities
between those clubs in terms of winning power.17 The H index reflects the degree of
competitive balance between teams. A rise in the index signifies an increase in
inequality and therefore a decline in competitive balance. In a standard industry
context, the Herfindahl index lies between 0 (with an infinite number of firms) and 1
(pure monopoly): however, in football, restrictions on the number of teams in a league
and constraints imposed by the points scoring system mean the index lies well within
this range. In a 20-team league, the lower bound of the H-index would be 0.05 (the
value attained in a perfectly balanced league) and the upper bound would be 0.07 (the
value attained in a perfectly unbalanced league with the most unequal distribution of
points attainable). The H-index is illustrated in Figure 2.5 for the period 1947-2004.
16
See Depken (1999) who uses a measure of competitive balance for major league baseball based on
the Herfindahl-Hirshmann index, however, our index is based on the share of points whereas Depken’s
is based on the deviation of the sum of squares of win percentages from 1/N where N is the number of
teams.
17
See Hart, 1975 and Davies, 1979 for a discussion of the sensitivity of the H-index to inequality and
the number of firms (clubs).
12
Figure 2.4 The H-Index
Competitive Balance inTop Flight Football 1947-2004
0.057
Change in number of
0.055 clubs from 20 to 22
0.053 H-Index
Change in number of
clubs from 22 to 20
0.051 Change in number of
clubs from 22 to 20
0.049
0.047
0.045
48
52
56
60
64
68
72
76
80
84
88
92
96
00
04
19
19
19
19
19
19
19
19
19
19
19
19
19
20
20
Figure 2.6 shows what has happened to the H-index during the life of the Premier
League.
0.056
0.054
0.052
0.05
H-Index
0.048
0.046
0.044
0.042
92
93
94
95
96
97
98
99
00
01
02
03
04
19
19
19
19
19
19
19
19
20
20
20
20
20
It can be seen from Figure 2.5 that the H-index was roughly constant between 1947
and the 1980s, and that after that it showed a rise. Over the period as a whole the H-
index increased by around 13 per cent. However, part of this increase is due to the
13
reduction in the number of teams (from 22 to 20). While this has an important effect
on competitiveness, it is useful to abstract from these changes to focus on inequality.
If we compare years for which the team numbers are the same we find that between
1947 and 1987 the H-index fell very marginally by –0.4 per cent. Between 1989 and
2004 the H-index then rose by 3 per cent - a significant increase.
Like the C5 ratio, the Herfindahl index is sensitive to changes in the number of teams.
This can be corrected for by dividing the index by the value of H that would be
attained in a perfectly balanced league to give the H Index of Competitive Balance
(HICB) as shown below:18
⎛ H ⎞
H Index of Competitive Balance = ⎜⎜ ⎟⎟ * 100
⎝ (1 / N ) ⎠
where H = ∑ si2 and 1/N = ∑ pi2 where pi equals the share of points attained by club
i in a perfectly balanced league. In a perfectly balanced league of any size the
Herfindahl index of competitive balance would take the value of 100. A decline in
competitive balance is reflected by an increase in the index. Data for the Herfindahl
Index of Competitive Balance are presented in Figure 2.7.
18
Note that this index can be used to capture dominance by calculating each team's share of points over
a number of years, for example 5 years.
14
Figure 2.7 The H-Index of Competitive Balance:
FA Premier League, 1947-2004
112
110
108
106
104
102
H Index of Comp Balance
100
47
51
55
59
63
67
71
75
79
83
87
91
95
99
03
19
19
19
19
19
19
19
19
19
19
19
19
19
19
20
The analysis of both the C5 ratio and the H-index shows that there has been a
significant decline in competitive balance between 1947 and 2004 with the H-index
and C5 ratio showing increases of 13 per cent and 21 per cent respectively. Further
analysis of the evolution of this trend reveals that there was no real change in
competitive balance between 1947 and 1987 and that virtually all of the increase has
occurred since 1989. Part of this increase is accounted for by changes in league
competitiveness associated with changes in the number of teams that make up the
league. However, even when abstracting from these effects, we observe a significant
increase in inequality as captured by both the C5 Index of Competitive Balance and
the H-Index of Competitive Balance. The increase is higher for the C5ICB which
suggests that growing inequality and imbalance between the top 5 clubs and the rest
has been a significant source of the decline in competitive balance.
15
2.6 Properties of the C5 and Herfindahl Indexes of Competitive Balance
∑ (N − X )
i −1
i
rest.
∑ (N − X
i =1
i )
where N is the number of clubs and
Xi is the league position of club i
with the top club ranked 1 and the
bottom club ranked N.
Sensitive to the number of
H-Index 2 teams in a league N and the
⎛ ⎞
⎜ ⎟ relative competitive strengths
1 ⎜ (N − X i ) ⎟
N
to ∑ ⎜ N ⎟
of all teams (as measured by
each team’s share of points).
⎜ ∑ (N − X i ) ⎟
N i =1
Reflects the relative
⎝ i =1 ⎠ competitive strengths of all
teams in the league.
C5ICB M N Takes a value of 100 for a
100 to . perfectly balanced league of
(M + T ) 5
any size. Sensitive to
changes in the relative
competitive strength of the
top 5 teams (measured by
their share of points) vis-à-
vis the rest. Controls for
variation in league size.
HICB 2 Takes a value of 100 for a
⎛ ⎞
⎜ ⎟ perfectly balanced league of
⎜ (N − X i ) ⎟ . N
N
100 to ∑ ⎜ N ⎟ 1
any size. Reflects the
relative competitive strengths
⎜ ∑ (N − X i ) ⎟
i =1
of all teams. Controls for
⎝ i =1 ⎠ variation in league size.
19
This reflects the fact that C5 measures inequality between the top 5 and the rest and ignores
inequality within the top 5 and inequality within the group of clubs outside the top 5. In contrast the
Herfindahl index measures inequality between all teams, hence the maximum value of HICB coincides
with the value for a perfectly unbalanced league.
16
2.7 The Lorenz Seasonal Balance Curve
The Lorenz curve is an indicator of inequality that is frequently used in studies that
look at inequalities in income or wealth. It is constructed by calculating the share of
income or wealth held by each percentage of the population and comparing this to an
ideal or equal distribution. Under an equal distribution, 10 per cent of the population
would have 10 per cent of income, 20 per cent of the population would have 20 per
cent of income and so on. Calculation of the actual figures using Lorenz’s method
might show for example that the poorest 10 per cent of the population had only 5 per
cent of income. The extent of inequality is measured by the gap between the equal
distribution shown by the diagonal line and the Lorenz curve which shows the actual
distribution of income. The wider this gap the greater the degree of inequality (see
Figure 2.8).
17
Figure 2.8 The Lorenz Curve: Inequality in Top Flight
Football 1950-2004
100
Cumulative percentage of points
80
60
1950
1993
2004
40
20
0
0 20 40 60 80 100
Cumulative percentage of Clubs
The results are presented in Figure 2.8 for the years 1950, 1993 and 2004. It can be
seen that between 1950 and 1993 there was little change in competitive balance, in
fact the lines coincide at some points. However, comparing 1993 and 2004 we see
that there has been a significant decline in competitive balance. The results from this
analysis are therefore consistent with the results from our other two indicators.
The decline in competitive balance experienced in the English Premier League raises
the question of whether other European Leagues have experienced a similar pattern.
Data on the C5 and H indexes of competitive balance are presented in Figures 2.9 to
2.12 for the top-flight leagues in Italy, Germany, France and Spain.
It can be seen that in Italy there has been a marked deterioration in competitive
balance since 1992, so that at the end of the period (2004) Italy had the highest degree
of imbalance of the top 5 leagues. In Germany, the increase is less marked but it is
possible to discern a rising trend (decline in competitive balance) over the past 10
18
C5 Index of
Competitive
Figure 2.9 Competitive Balance in Serie A, Italy: H Index of
The C5 and H Indexes of Competitive Balance Competitive
Balance
Balance
150 114
145
112
140
135 110
130
108
125
106
120
115 104
110
C5 Index of Competitive Balance H Index of Competitive Balance 102
105
100 100
56 58 60 62 64 66 68 70 72 74 76 78 80 82 84 86 88 90 92 94 96 98 00 02 04
19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 20 20 20
C5 Index of
Competitive Figure 2.10 Competitive Balance in the German Bundesliga: H Index of
Balance The C5 and H Indexes of Competitive Balance Competitive
Balance
145 114
140
112
135
110
130
125 108
120 106
115
104
110
C5 Index of Competitive Balance H Index of Competitive Balance 102
105
100 100
64
66
68
70
72
74
76
78
80
82
84
86
88
90
92
94
96
98
00
02
04
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
20
20
20
19
C5 Index of H Index of
Figure 2.11 Competitive Balance in the French Ligue 1:
Competitive Competitive
Balance The C5 and H Indexes of Competitive Balance Balance
145 109
140 108
135 107
130 106
125 105
120 104
115 103
110 102
105 101
C5 Index of Competitive Balance H Index of Competitive Balance
100 100
54
56
58
60
62
64
66
68
70
72
74
76
78
80
82
84
86
88
90
92
94
96
98
00
02
04
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
20
20
20
C5 Index of Figure 2.12 Competitive Balance in the Primera Liga, Spain: H Index of
Competitive Competitive
Balance The C5 and H Indexes of Competitive Balance
Balance
145 112
140
110
135
130 108
125
106
120
115 104
110
102
105
C5 Index of Competitive Balance H-Index of Competitive Balance
100 100
56
58
60
62
64
66
68
70
72
74
76
78
80
82
84
86
88
90
92
94
96
98
00
02
04
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
20
20
20
20
years or so. In France, there appears to be no clear trend in competitive balance
though in the past 6 years both indexes have hit all time highs. In Spain, competitive
balance has followed a cyclical pattern declining between 1956 and 1976 and rising
thereafter.
2.9 Summary
Using a number of indicators for the period 1947-2004 we find that competitive
balance in the Premier League remained roughly constant for 40 years or so between
1947 and 1987. However, since then all three indicators show a significant decline in
competitive balance. Comparing the HICB and the C5ICB it is apparent that the
decline in competitive balance is most marked using the C5 index of competitive
balance which suggests that growing inequality between the top 5 clubs and the rest is
a major factor contributing to the decline in overall competitive balance. One possible
explanation for this is that it is the top 5 clubs that have enjoyed increased revenue
streams from broadcasting and access to European competitions and associated
funding streams. This and other factors are discussed in Section 3. Similar trends in
competitive balance have been experienced by the top-flight leagues in Italy and
Germany. In France there is no clear trend over the period, though there are signs of
increases in the indexes from 1992 to date. Spain has experienced a more cyclical
pattern with a moderate improvement in competitive balance between 1956 and 1976
and a moderate decline (captured by an increase in the indexes) from 1976 onwards.
21
3. Revenues, Wage Expenditure and Competitive Balance
One of the factors that makes football clubs particularly difficult to manage
financially is the dual objective of football clubs to attain sporting success and to run a
commercial enterprise. These twin objectives are set out in most clubs’ Memorandum
and Articles of Association constituted at the time of incorporation of the club as a
company.20 Unlike most companies that have to concern themselves only with the
commercial objectives of the business (making a profit), football clubs have to
balance sporting and commercial objectives. This balancing act is complicated by two
well-established relationships that characterise the football industry and sports leagues
more generally:
20
Note that this is typically the case in the UK though some clubs are incorporated as trusts or
associations.
22
Conversely, clubs at the bottom of the league find themselves trapped in a vicious
circle of poor performance on the pitch, leading to a loss of support and revenue
streams from match day income, broadcasting rights, sponsorship and cup runs,
which, in turn, reduces the resources available for investment in players and may thus
lead to a further deterioration in sporting results. The existence of such virtuous and
vicious circles and the inherent instability of the industry is the reason why sports
leagues redistribute income from the richest to the poorest clubs. Without such
redistribution the performance gap between the top and bottom clubs widens and the
league becomes unbalanced and unstable. Such instability results in various risks,
such as loss of spectators, bankruptcy of clubs and the threat of breakaway leagues.
Over the past decade or so a variety of factors have led to a decrease in competitive
balance and a rise in risks facing the industry.
In the discussion that follows we explore what has happened to revenue growth and
the relationship between wage expenditure and performance and consider how these
relationships have been affected by recent trends in the industry.
Over the past decade Premiership clubs have experienced unprecedented growth in
revenue. Between 1992 and 2003 turnover of the Premiership increased by around
650 per cent, that is a 6.5 fold increase in revenue. By far the largest source of growth
in turnover came from increased income from broadcasting rights, which rose by over
3500 per cent (or more than 35 times) from just £15m in 1992 to £543m in 2003.21
However, income from match-day gate receipts and other commercial activity also
showed significant increases, though not on the same scale as broadcasting revenue.
Income from gate receipts increased by just under 350 per cent (or 3.5 times) to
£363m, while revenue from other commercial activity (sponsorship, conference
facilities, catering, merchandise sales and other activity) increased by just over 350
per cent to £340m. As a result, broadcasting income, which accounted for less than
10 per cent of Premiership income in 1992, now accounts for the biggest share of
21
Source: Deloitte (2004a) p. 28 and Deloitte and Touche Annual Review of Football Finance various
issues.
23
revenue (44 per cent), with the other 56 per cent accounted for by match-day income
(29 per cent) and commercial revenue (27 per cent).22
These dramatic changes in revenues and wage expenditures have had a significant
impact on competitive balance. Figures 3.1 and 3.2 show the relationship between
wage expenditure (£000s) and league performance (measured by each team’s share of
points) for the years 1994 and 2003. It can be seen that there is a strong positive
association between these two variables, that is, the more clubs spend on wages the
better their league performance. Comparing the Figures for 1994 and 2003 it can also
be seen that the relationship between wage expenditure and performance has become
stronger: in 1993 the correlation coefficient between the two variables was 0.68, while
by 2003 it had increased to 0.73 – with both coefficients being statistically significant
at the 1 per cent level (0.99 confidence interval); that is, the line has pivoted to the left
and become steeper. The shift reflects the fact that whereas in 1994 the wage
expenditure of the top club was four times that of the wage expenditure of the bottom
club, by 2003 the ratio of the highest to lowest wage expenditure had risen to eight
times.
Moreover, it can be seen that a gap has opened up between the top 5 clubs and the
rest. In 1994, with the exception of the highest wage spender (Manchester United),
wage expenditure of the other top 5 clubs was matched, or almost matched by a
number of lower position clubs. By 2004 the top 5 clubs were all spending over £40m
22
Source: Deloitte (2004a) p. 28.
23
Source: Deloitte (2004a) Appendix 1 and 7 pp. 4-5 and 8-9 and Deloitte and Touche Annual Review
of Football Finance various issues.
24
Figure 3.1 Wage Expenditure and League
Wage Performance: FA Premier League Clubs 1994
Expenditure
£000s
12,000
Manchester United
10,000 Liverpool FC
8,000
6,000
4,000
2,000
Swindon Town
0
2 3 4 5 6 7 8
Percentage Share of Points
Wage
Figure 3.2 Wage Expenditure and League
Expentiture Performance: FA Premier League 2003
£000s
100,000
40,000
20,000 Sunderland
0
1 2 3 4 5 6 7 8 9
Percentage Share of Points
25
on wages, which was significantly above the level of all of the remaining clubs bar
one. The exception was Leeds United which was saddled with high wage costs
following its (failed) strategy of attempting to qualify for a Champions League place.
Figures 3.1 and 3.2 illustrate that the dominance of the top 5 clubs is increasingly a
reflection of their ability to spend more on wages as compared with the lower
positioned clubs.
This ability is, in turn, a reflection of the greater revenue earning power of the top
clubs. Figure 3.3 shows the share of revenue earned by the top 5 clubs over the 11-
year period from 1993 to 2003. It can be seen that in 1993 the share of revenue held
by the top 5 was 26.8 per cent, slightly above the 22.7 per cent that would reflect an
equal distribution of income. By 1994, the 5 firm concentration ratio (C5) had risen to
over 40 per cent and by 1997 to over 47 per cent, well above the 25 per cent level that
would reflect an equal distribution in a league of 20 clubs. Since 1997, the index has
remained more or less flat with the top 5 clubs accounting for just under half the
income of the league. The concentration of revenue within the hands of the top 5
clubs has enabled them to increase their share of total wage expenditure in the
Premiership from 32 to 39 per cent. Given the positive association between
expenditure on wages and league position, this increase has enabled these clubs to
dominate the top few positions in the league table and is a significant factor
underlying the decline in competitive balance. The above analysis raises the question
of what factors have led to: (a) the increase in revenue; and (b) the greater
concentration of revenue in the hands of the top 5 clubs?
50
45
40
35 C5 Index
30
25
20
93
94
95
96
97
98
99
00
01
02
03
19
19
19
19
19
19
19
20
20
20
20
Year
26
3.2 Sources of Revenue Growth in Premier League Football
We have already touched on the main sources of the increase in revenue enjoyed by
the Premier League: by far the biggest increase has occurred in monies from the sale
of TV broadcasting rights. Table 3.1 reports the main sources of revenue and their
growth over the period 1992-2003. Against this background of rising revenues,
inequality has increased and the top clubs have been able to secure a bigger share of
the revenue coming into the league. There are three main sources of this increase in
inequality in revenue:
The introduction of the current revenue sharing procedures within the Premiership
allocate domestic TV broadcasting income according to the Premier League’s
50:25:25 rule – 50 per cent of broadcasting revenue is divided equally among the 20
Premier League clubs; 25 per cent is a merit payment awarded according to league
position at the end of the season with the top club receiving the biggest payment and
the bottom club the least; and 25 per cent is distributed according to the number of
times each team takes part in a match broadcast live, with each team guaranteed a
27
minimum number of appearances (see Findlay et al, 1999 for a more detailed
discussion). The rapid growth in broadcasting revenues and the application of these
rules has resulted in a widening gap between the revenue of the leading clubs vis-à-vis
lagging clubs. Figures showing the allocation of revenue between 1992/93 and
2003/04 are presented in Table 3.2.
It can be seen that while the distribution rules have kept the ratio of revenue earned by
the top to bottom club roughly constant at around 2.2, the gap between the top earner
and bottom earner has increased dramatically from £1.3m in 1993 to £19.4m in 2004.
This revenue gap has had two effects. First, it has enabled the top clubs to spend a
much larger amount on wages as compared with clubs at the bottom end of the league.
Given the relationship between wage expenditure and share of points this has tended
to reduce competitive balance. Second, it has set up a financial incentive for clubs to
gamble on success by spending more on players’ wages in the hope of recouping the
28
money by moving up the league table. Thus, compared to the situation where
broadcasting revenues were shared equally, risk in the industry has increased.
In addition, television exposure is one of the factors that has attracted increased
support for football clubs and it is the top clubs whose matches are screened more
regularly, who are therefore the main beneficiaries. These clubs can then benefit
disproportionately in terms of increased sponsorship and other revenues, including
gate receipts (in the case of clubs that are not already attaining capacity attendances).
Since the formation of the UEFA Champions League in 1992, clubs finishing in the
top 1 to 4 places in the Premiership have had access to TV broadcasting and match-
day revenue streams and sponsorship monies from playing in the UEFA Champions
League.24 Data on the allocation of revenue from the UEFA Champions League to
Premiership clubs in 2003 are presented in Table 3.3.
The figures reported in Table 3.3 exclude match day income from playing UEFA
Champions League matches and sponsorship and other revenue that arises as a result
of participation in the Champions League. To put these figures in context it can be
seen that even at Manchester United, the UEFA broadcasting revenue income alone
covers around 17 per cent (almost one fifth) of the total wage bill. When match-day
income and revenue from sponsorship rights (which are marketed by UEFA) are taken
24
Provided those not qualifying directly do get through the necessary qualifying round or rounds. Prior
to 1992 revenue was distributed under the European Cup competition.
29
into account, Manchester United’s income from the UEFA Champions League25 was
£26.8m in 2003, equivalent to just over 33 per cent of its total wage bill. We estimate
that when these revenue streams are taken into account for all four clubs in Table 3.3,
income from the UEFA Champions League accounts for around 15 per cent of the
increase in the share of revenue generated by the top five clubs over the period 1993-
2003.
The first English clubs to enter the Champions League were Manchester United in
1994/5 and Blackburn in 1995/96. Initially broadcasting revenues were low: in the
first year of the competition they amounted to 38m Swiss Francs. However, by
1998/89 the figure had risen to 614m Swiss Francs and by 2002/03 to 814m. The
growth in broadcasting revenue received by English participating clubs over the last
seven years can be seen from Table 3.4. The data illustrate that revenue streams from
European competitions have grown rapidly. Given that these revenues are only
attainable by the top clubs in the league, the growth in UEFA Champions League
revenue streams is a significant factor contributing to growing inequality in the
distribution of income between the top 4 to 5 clubs and the rest.
Table 3.4 Broadcasting Revenue from the UEFA Champions League 1998-2004,
£m
1998 2002 2003 2004
Manchester United 5.9 20.0 13.1 18.3
Newcastle United 3.1 10.0
Liverpool 14.0 8.4
Arsenal 13.4 18.6
Chelsea 19.0
Total 9.0 34.0 44.9 55.9
Source: Deloitte and Touche and FourFourTwo (1999) and Deloitte (2004a) and Deloitte and Touche
(2003). Figures converted from Euros to Sterling using the following exchange rates: 0.62 in 2002;
0.64 in 2003 and 0.67 in 2004 (European Central Bank web site www.ecb.int).
30
view to strengthening non-participating teams and promoting competitive balance.
Similarly, 700,000 Swiss Francs was distributed to the Football Association and its
clubs; equivalent to around £300,000 at current exchange rates.26 Clearly, the
principle of solidarity payments is extremely important and the fact that UEFA
specifies that this money is to be spent on youth training means that these payments
help strengthen the sporting capabilities of teams and promote competitive balance.
However, the payments would appear to be too small to make a significant
contribution to closing the revenue gap between Premiership clubs qualifying for the
Champions League and the rest. To put the figures in context, the average payment to
non-participating clubs amounted to less than 2 per cent of the average Champions
League broadcasting revenue distributed to Manchester United, Arsenal and Chelsea
in 2004.
A further effect of the large sums of money now attached to gaining a Champions
League place is that it has provided an additional incentive for clubs to gamble on
success. With up to almost £30m to be gained from a successful Champions League
campaign, clubs may be tempted to overspend on players’ wages in the hope of
recouping this expenditure via qualification for the Champions League. Of course,
such a strategy is risky because there is only a limited number of Champions League
places. One of the last clubs to take this risk was Leeds United who failed to qualify
and subsequently experienced a financial crisis followed by administration and
relegation.
While the growth in broadcasting revenue has been the most significant single factor
underpinning the spectacular growth in revenue experienced by the Premier League,
revenue growth from match day income and other commercial income has also been
exceptional. Together these two sources of income still account for over 55 per cent
of total income.27
26
This sum comprises: the set payment for each participating national association of 300,000 Swiss
Francs (CHF); 240,000 CHF for Manchester United for 3 rounds (including one qualifying round) of
the UEFA Cup; 80,000 CHF for Southampton for 1 round of the UEFA Cup; and 80,000 CHF for
Blackburn for 1 round of the UEFA Cup – see UEFA (2004b).
27
The latest available figures are for 2003.
31
As discussed in section 3, clubs at the top of the league can experience a virtuous
circle of success, with success on the field and greater exposure on television leading
to increases in support, match day income, merchandising, sponsorship and other
commercial income. The data in Table 3.4 provide support for this and show that
although the top 5 clubs attain almost 40 per cent of broadcasting revenue they also
dominate in terms of match day income (accounting for 48 per cent of total match day
income generated in the Premiership) and other commercial revenue (accounting for
42 per cent of the total). While part of the growth in match day and commercial
income is the result of success on the pitch fuelling commercial success, it would be
wrong to conclude that all of the growth in revenue was attributable to such a virtuous
circle that is unrelated to clubs’ business strategies, marketing, financial management
and governance. This is illustrated by the fact that there are differences in the
rankings of the top 5 clubs in terms of broadcasting revenue, match-day revenue and
commercial revenue. In addition, it can be seen that the financial performance of
Manchester United stands out in all three categories, and particularly so in terms of
match-day income where this is more than double that of any other club.
The exceptionally high match day income earned by Manchester United is partly a
reflection of strategic investment decisions made over the previous decade to expand
capacity by enlarging the stadium and improving facilities. Similar investments were
made by Newcastle who have the second highest match day revenue.
Table 3.4 Premier League Revenue Streams for the Top 5 Clubs, 2002/03
Match Day Gate Broadcasting Commercial
Receipts Revenue Revenue
Manchester United £64.9m £53.5m £42.5m
Arsenal £25.7m £47.9m £22.1m
Liverpool £26.4m £42.0m £27.2m
Newcastle £30.1m £38.5m £20.3m
Chelsea £28.0m £27.6m £30.0m
Total of Top 5 £175.0m £209.7m £142.1m
Total Premiership £363.0m £543.0m £340.0m
Share of Top 5 (%) 48% 39% 42%
Share of Manchester United (%) 18% 10% 13%
Source: Deloitte (2004) The Rich List and Deloitte (2004) Annual Review of Football Finance. Data
from The Rich List converted from Euros to Sterling using an exchange rate of 0.64 Pounds to the Euro.
32
The business performance of clubs is normally measured by profitability. Four of the
top five clubs in 2003 made an operating profit in each of the last 5 years: however,
only one club, Manchester United made a pre-tax profit after transfers in all five years
between 1999 to 2003. Within the top 5 clubs, four clubs made a pre-tax profit in
2003. The exception was Chelsea, which shows that sporting success does not
automatically produce economic success. Outside the top 5 clubs, only one club,
Birmingham City, made a pre-tax profit in 2003.
To summarise, although the sporting success of the leading clubs increases their
revenue earning potential, the extent to which this potential is fulfilled depends on the
business strategy, marketing, financial management and corporate governance
practices of the clubs. Clubs can and do enhance their revenue earning potential by
the adoption of successful business strategies, sound financial management and good
governance. The fact that some clubs have dealt with the peculiar economics and
governance issues surrounding the football industry better than others is one of the
factors that has widened the revenue gap between leading and lagging clubs and
contributed to a decline in competitive balance. Within the Premier League and
elsewhere there are examples of best practice. Rolling out best practice across the
league would help to boost the income of lagging clubs and close the revenue gap.
28
The other two clubs were Manchester United and West Bromwich Albion.
33
3.6 Revenues and Competitive Balance in Europe
The growth in revenue over the past decade has not been confined to the English
Premier League. All of the 5 big European leagues – England, Italy, Germany,
France and Spain have enjoyed significant increases. However, both the rate of
revenue growth and the absolute level of revenue differ markedly across leagues. The
Premier League has experienced the highest rate of growth with revenue reaching
€1,790m in 2003, significantly above the total revenue of the next two largest leagues:
Italy, with a revenue of €1,162m and Germany with a revenue of €1,108m. France
and Spain were somewhat behind with revenues of €847m and €689m, respectively.29
Given that broadcasting revenue is not shared equally, it is not surprising that the
rapid increases in revenue to such high levels in Italy and Germany have led to
deteriorations in competitive balance. The decline has been most marked in Italy
where redistribution rules are the least egalitarian. In France, where revenue has
grown less rapidly and the redistribution rules are more egalitarian, there is no clear
trend decline in competitive balance. Similarly, in Spain where revenue growth has
been modest, competitive balance has not shown any marked decrease over the past
decade.
Attendance at Premier League matches rose from the mid 1980s to 2003 and although
there was a marginal fall last season, attendance at the ground is close to capacity at a
number of clubs. However, it cannot be concluded from this that attendance and TV
viewing figures are maximised or that the figures would not be higher had the decline
in competitive balance not occurred. Attendance and viewing figures are determined
by a number of factors including consumers’ income, media coverage and fashion.
29
Deloitte (2004a).
34
The number of households subscribing to pay TV is still significantly less than 50 per
cent and is well below saturation point. There is also scope to increase viewing
figures for pay-per-view matches. The value of the next TV deal will depend on
subscriptions and pay-per-view audiences and these will be influenced by competitive
balance.
It should also be noted that the two leagues where competitive balance is declining
and low are the Italian Serie A and the Scottish Premier League. Both leagues have
had problems with attendance and in both leagues attendance is considerably below
capacity. Moreover, in Italy attendance has been falling since the 1990s.
Within the English Premier League there are clear signs that the widening income
gaps associated with declining competitive balance are affecting the financial viability
of clubs and increasing risk in the industry. In the past two seasons, two clubs have
been forced into administration. Moreover, the majority of clubs are making pre-tax
losses. Part of the explanation for this is to be found in the fact that the increasing
revenue gap has provided an incentive for lagging clubs to spend more on players’
transfers and wages in order to catch up. Between 1994 and 2003 the average
expenditure on wages in the Premiership increased from 55 per cent of turnover, to 65
per cent.
Finally, when a league becomes unbalanced the leading clubs may seek competitive
balance in a new league or a rival league. The idea of a European Super League has
been mooted from time to time and although it is unlikely to provide a viable
alternative to the current structure, the concept of a ‘Super League’ remains a threat
(even if the credibility of the threat is low) to the stability of both the Premiership and
the UEFA Champions League.
4. Policy implications
35
increase their dominance of the league (as measured by, for example, share of points)
and has resulted in a decline in competitive balance. This analysis raises the question
of what has caused the gap and the inequality? The widening gap has emerged largely
as a result of rapid revenue growth, hence, it is partly a reflection of the economic
success of the Premier League. More detailed analysis reveals that there are three
main sources of revenue growth that have had an unequal impact on the fortunes of
clubs at the top and bottom of the league:
The distribution of the first two of these sources is under the regulatory control of the
Premier League and UEFA respectively. Both the Premier League and UEFA have
been extremely successful at increasing revenue streams from the sale of broadcasting
rights. However, a downside to this success is that it has widened the income gap
between leading and lagging clubs and has resulted in a decline in competitive
balance. As a result, attendance and viewing figures are not maximised and, perhaps
more ominously, the growing imbalance has resulted in increased risk, which if left
unchecked may destabilise the league. This instability is reflected in poor financial
performance, clubs going into administration and calls for the formation of a
European Super League. It is to avoid precisely such pitfalls that professional sports
leagues redistribute income.
The Premier League and UEFA can guard against these risks by re-examining their
redistribution rules. The Premier League’s rules have the effect of widening the
income gap between the top and bottom club when revenue is rising, and narrowing
the income gap should revenue fall. To prevent further widening of the gap arising
from increased revenue from broadcasting we would advocate that any increase in the
combined value of future broadcasting rights should be divided equally between the
20 clubs. All clubs would benefit from the equal share of the increased revenue, so all
36
clubs would be better off, and the lagging clubs would increase their share of total
revenue. Moreover, if the more egalitarian distribution of income enhances
competitive balance this should increase attendance and viewing figures leading to
further increases in revenue. The more egalitarian distribution of income will also
help reduce risks and stabilise the league.
In a similar fashion, UEFA can modify its redistribution rules to ensure that a larger
share of income from future television deals is earmarked for National Associations
and leagues earmarked for member clubs that are not participating in the Champions
League. Again, this will improve competitive balance in national leagues, thus
making the threat of a breakaway European Super League less likely. The same
caveat applies regarding the need to promote improvements in the governance
practices of lagging clubs. This could be done as an extension of UEFA’s initiative
on licensing.
Finally, it is apparent that part of the greater variation in revenue across clubs in the
league is due to a number of clubs adopting best practice business management and
governance strategies. As a result, these clubs have been able to take full advantage
of the potential revenue streams that may be earned in the industry. There is clearly a
need to learn from the best practice of these clubs and to roll this out across the
industry. This could be done via the establishment of appropriate executive courses
designed to provide football executives and administrators with the necessary
expertise in business strategy, financial management and governance, thus enabling
clubs to maximise their potential in the difficult and complex business environment of
the football industry. Such executive courses would dovetail neatly with the FA’s
work on developing a workable code of corporate governance for football and
UEFA’s licensing initiative.
37
Appendix 1
Table A.1 A Review of Measures Competitive Balance in US Major League Sports and Soccer
38
Eckard (2001) Dominance A Herfindahl-Hirschman index based on Finds results consistent with variance based
team shares of the top 4 places in the measure for the AL and a marginal
league over a 5-year period. improvement in competitive balance in the
NL.
b) Soccer
Author(s) and year of Type of Competitive Description of Measure Comments and Findings
publication Balance
Peel and Thomas (1988) Match Probability Odds: the inferred home win Captures match uncertainty. Used as a
probability from betting odds e.g. odds variable to explain the demand for soccer.
of 2/1 for a home team win, imply a win Authors find a U shaped relationship between
probability of one third for the home attendance and match uncertainty. Forrest and
team. Simmons (2002) suggest that betting odds
may be subject to bias.
Kuypers and Szymanski Seasonal The standard deviation of each club’s Shows that the average (over 10 seasons)
(1999) seasonal points total averaged over 10 standard deviation of points totals increased
seasons for four decades since 1946. in the first 3 decades since 1946 but fell
Sensitive to changes in the points during the last decade (1986-96).
scoring system and the number of
teams.
Kuypers and Szymanski Long-term Dominance Following Fort and Quirk (1992) Lorenz Shows that the English league had the lowest
(1999) curve measuring the cumulative degree of long-term dominance and the
percentage of championship wins over Netherlands the highest degree over the
the period 1946-1998 for 5 European period 1946-98.
leagues.
Dobson and Goddard Match Forecasting model which uses Finds that past performance can be used to
(2001) probabilities and historical data to predict future performance but that there are
predict match outcomes. still random effects (a degree of uncertainty)
39
Szymanski (2001) Seasonal and Trend SDW calculated across all Football Shows no clear trend in competitive balance.
League and Premier Leagues clubs on The measure does not capture uncertainty
an annual basis over the period 1977- over who will win each league championship
1999. This is the most commonly used as four leagues are combined.
measure in US studies, though US
studies typically deflate this measure by
the ideal standard deviation that would
occur if each club had an equal chance
of winning each match.
Forrest and Simmons Match Based on betting odds data and authors Used as a variable to explain attendance.
(2002) claim to correct for bias in betting odds Captures match uncertainty in the 1997/8
data. season for Football League Divisions 1, 2 and
3.
Buzzacchi, Szymanski Dynamic A Gini-type index that looks at the Finds that 3 closed US leagues are more
and Valletti (2003) actual number of clubs winning a competitively balanced that 3 open European
championship or entering the top 5 leagues. This result is not surprising given
places over various time periods that the (larger number of) clubs entering
compared to the theoretical distribution European top flight leagues are drawn from
assuming equal winning probabilities. lower level leagues and are therefore less
likely to win the league or enter the top 5,
than clubs remaining in the top flight.
Seasonal The paper also includes estimates The SDW measure for the English top flight
(averages over five 10-year periods) of shows a decline in competitive balance until
the SDW (the SD of win percentages the end of the 1980s and an improvement
deflated by the ideal standard deviation thereafter. Cross league comparisons show no
for a perfectly balanced league). clear cut pattern of difference between the
competitive balance of US and European
leagues.
40
Morrow (2003) Dominance Looks at the identity of the top 5 clubs Produces tables of the top 5 clubs for 7
over the 5-year period 1998-2002. European leagues.
Bourg (2004) Seasonal SDW, see Scully (1989) and Fort and Looks at SDW average over 4 five-year time
Quirk (1992) above. periods between 1980-2000. Finds
competitive balance deteriorates between 86-
90, improves between 91-95 and deteriorates
between 96-2000. Also compares 2 ten-year
time periods and finds a deterioration in the
ten-year average for 1991-2000 compared to
the average over the previous decade in
England, Italy and France and an
improvement in Germany and Spain.
Gerrard (2004) Dominance Following Eckard (2001) uses the The concentration of Championship titles is
Herfindahl index to measure the used for cross section analysis to look at the
concentration of championship titles and comparative competitiveness of different
the share of the top 4 places. Also looks national leagues. The data are also broken
at the share of titles won by the top 2 down into 2 time periods 1992-97 and 1997-
and top 3 clubs, the points percentage of 02 for different indicators. The evidence is
winners and the percentage points gap somewhat mixed though 3 of the four
st nd
between 1 and 2 place. indicators for England suggest an increase in
dominance between 1997-02 compared to
1992-97 while in Italy 3 of the four indicators
show an improvement.
41
References
Borland, J and MacDonald, R (2003) Demand for Sport, Oxford Review of Economic
Policy, 19:4:478-502.
Bourg J-F (2004) Professional Team Sports in Europe: Which Economic Model? in Fort,
R and Fizel, J (eds) International Sports Economics Comparisons, Praeger,
Connecticut, London.
Deloitte and Touche (2003 and various previous issues) Annual Review of Football
Finance, Deloitte and Touche: Manchester.
Football Governance Research Centre (2003) State of the Game: the Corporate
Governance of Professional Football 2003, FGRC Research Paper Series, 2003,
No. 4, Birkbeck, University of London.
42
Gerrard, B (2004) Still Up for Grabs? Maintaining the Sporting and Financial Viability of
European Club Soccer, in Fort, R and Fizel, J (eds) International Sports
Economics Comparisons, Praeger, Connecticut, London.
Morrow, S (2003) The People’s Game? Football Finance and Society, Palgrave, London.
Peel, D A and Thomas, D (1988) Outcome Uncertainty and the Demand for Football,
Scottish Journal of Political Economy, 35:242-249.
Quirk, J and Fort, R D (1997) Pay Dirt: The Business of Professional Team Sports.
Princeton N J, Princeton University Press.
Szymanski, S and Kuypers, T (1999) Winners & Losers: The Business Strategy of
Football, Viking, England.
UEFA (2004b) UEFA Champions League: League’s Share, uefadirect, No 27 July 2004
p. 14.
43
Other Publications in the Series
Unless otherwise stated, all publications are available for £5 (including post and packaging) from the FGRC
at the address on the back cover. Please make cheques payable to Birkbeck College. Also downloadable
free of charge from: www.football-research.bbk.ac.uk
Fresh Players, New Tactics: Lessons from the Northampton Town Supporters’ Trust -
Phil Frampton, Jonathan Michie and Andy Walsh
This paper provides an in-depth analysis of the supporters’ trust at Northampton Town football club. It
investigates the challenges facing an established supporters’ trust and what can be done to facilitate the
growth of the trust, increase its influence with the club and strengthen its ties with the local community.
The State of the Game: The Corporate Governance of Football Clubs 20041 - FGRC
The 2004 State of the Game report focuses on trends in the corporate governance of clubs and for the first
time includes analysis of the Football Conference. The report also reviews developments in regulation by the
football authorities, and considers the role of stakeholders including supporters’ trusts, and local authorities,
assessing their contribution to promoting and enhancing the sustainability of football clubs.
The State of the Game: The Corporate Governance of Football Clubs 2003 - FGRC
The State of the Game: The Corporate Governance of Football Clubs 2002 - FGRC
The State of the Game: The Corporate Governance of Football Clubs 2001 - FGRC
These Model Rules were produced as a collaborative effort between Cobbetts, Birkbeck and the Co-operative
Union, along with the pioneering groups of supporters who first set up the new wave of supporters trusts. The
Industrial & Provident Society Model for a football community mutual has now been used many times by
groups of supporters advised by Supporters Direct.
A ‘Fit and Proper’ Person Test for Football? Protecting and Regulating Clubs - Matthew Holt
This research paper considers the regulation of football clubs’ owners and major shareholders. It deals with
the introduction of a ‘fit and proper’ person test as recommended by the Football Task Force and the role of
the Football Association in promoting best practice and improving corporate governance at clubs.
Building Sustainable Supporters’ Trusts in the West Midlands: Interim Report and Final Report -
FGRC and The Co-operative College
Using the principles of action research the Interim Report identifies the training and support needs of
supporters’ trusts in the West Midlands and explores the potential for trusts to develop links and joint
initiatives with wider community stakeholders at football clubs. The Final Report provides an overview of the
Building Sustainable Supporters' Trusts in the West Midlands action research project and its findings.
Building Sustainable Supporters’ Trusts in the West Midlands: A Training Manual 2 - FGRC and The
Co-operative College
The Training Manual is a set of teaching materials designed to train members of supporters’ trusts. The
materials incorporated within the Training Manual can be used by trusts as text-based resources or by
facilitators to guide a series of interactive workshops. A CD ROM accompanies the pack that enables users to
download and print out the materials, exercises and handouts for further use.
The Ownership Structure of Nationwide League Football Clubs 2002-03 - Stephen Hope
Using detail from the last annual returns and accounts from Companies House, this research paper focuses
on the ownership of football clubs in the Football League, and includes a ‘club by club’ summary of their legal
structure and shareholding concentrations.
Professional Footballers’ Association: A Case Study of Trade Union Growth - Geoff Walters
In the context of a declining trade union movement, this research paper charts the remarkable growth and
development of the Professional Footballers’ Association throughout the 1980s and 1990s.
1
The latest edition of the State of the Game 2004 report is £25, with a discounted rate of £5 for supporters’ trusts, students
and club officials.
2
The 190 page Training Manual (including CD ROM) is available in hard copy for £50, with a discounted rate of £20 for
supporters' trusts, club officials and students.
The Football Governance Research Centre (FGRC) is a recognised Research Centre
of Birkbeck, University of London. Our research and publications focus on sport
management and the governance and regulation of professional football and sports
leagues. The FGRC is part of the School of Management and Organizational
Psychology located in the Clore Management Centre. The School offers a number of
Masters programmes in Sport Management & The Business of Football, including an
MSc and an MRes. Both of these degree programmes have ERSC recognition for the
award of studentships for postgraduate study leading to a PhD degree. For further
details see: www.football-research.bbk.ac.uk
Further copies of this paper are available for £5.00 (including post and packaging)
from the FGRC at the address below. Please make cheques payable to Birkbeck
College. Also downloadable free of charge from: www.football-research.bbk.ac.uk