Statistical dribble that never hits goal

What happens when the dismal science tackles the beautiful game? Stefan Szymanski sees no result.

In the days when it was unfashionable to admit to being a football fan or a Labour Party member, the conversation in the bar at economics conferences typically revolved around socialism and the selection policy of the England team. The sociology of economists is an interesting phenomenon. While generally perceived as market-worshipping crypto-fascists, a very high proportion are either active or fellow-travelling socialists. They also tend to love football - several public figures in the economics world are fans, among them Lord Burns, former permanent secretary to the Treasury and now chairman of the National Lottery Commission, (QPR); Gavyn Davies, chairman of the BBC (Southampton); Sir Howard Davies, chairman of the Financial Services Authority and future director of the London School of Economics (Manchester City).

But few economists have written about football. Perhaps the economics of a sport is not a weighty enough matter. The Economics of Football claims to be the first detailed economic analysis of the professional game and demonstrates that there is no end of statistical work that can be done.

Stephen Dobson and John Goddard have been working in this field for the past decade, and this book brings most of that work under one cover, held together by a detailed historical introduction to the development of soccer and a conclusion on its present condition.

Statistical analysis (econometrics or economic tricks, as you wish) dominates the book. It is aimed squarely at economists - chapter one explains that non-technical summaries have been provided for the non-econometrician, but such a reader must take a lot on trust. I can see three reasons for conducting this kind of analysis: the pleasure of football statistics itself (anorak tendency), the desire to test hypotheses drawn from economic theory (ivory-tower tendency) and the wish to derive policy prescriptions (wonkers).

The book is heavily weighted to the first of these activities, contains something of the second and is rather disappointing on the third. For example, chapter three deals with one of the most important issues in football - competitive balance. We all know that the income distribution has become increasingly skewed toward the big clubs - is this good or bad? Can anything be done about it? After reviewing the theory, the authors go to great lengths to show that the home team's probability of winning has fallen over the past 30 years. This is an interesting fact - but what follows from it? Should there be more income sharing, redistribution mechanisms or salary caps? We are not told.

Chapter four reviews the statistical variation of transfer fees in relation to the characteristics of players and of buying and selling clubs. This an atheoretical exercise - every possible explanatory variable is thrown into the pot with little reasoning as to what should and should not matter (for example, what determines that a player is traded in the first place?). This model is being used by some clubs to determine what constitutes a reasonable fee. Regrettably, we are not told how this experiment is progressing. I can imagine economists on the terraces hurling abuse at the manager for having paid two standard deviations over the fitted value.

Chapters five and six are an extended attempt to rank the contributions of football managers. This is anorak territory. What can we draw from the conclusion that John Gregory was, statistically, the most efficient manager in the Premier League between 1993 and 1998? The 50 per cent of the population who are not Manchester United fans will be pleased to discover that Sir Alex Ferguson did not make it into the top 25 despite winning the championship in four out of the six seasons; he will just have to console himself with the silverware.

Chapter seven returns to economic theory and a statistical account of match attendance - the demand for "live" football. Elsewhere, attendance data has been used to address important policy issues such as the impact of competitive balance and the broadcasting of live matches on TV. Here, however, the authors examine only the effect of prices (economists will be reassured that higher ticket prices lead to lower attendance) and team performance (answers on a postcard). Dobson and Goddard also try to explain why some teams have a larger fan base - statistically, the main factors are population, length of time in the league and unemployment (close the factories and everyone goes to watch the footy).

Chapter eight examines the efficiency of the stock market and football betting odds. Efficiency here means that all information about a team's performance is incorporated instantly into a share price or the bookie's odds. The authors find that share prices react quickly to information about team performance. They are less convinced that the betting market is efficient, but, as they ruefully remark: "The inefficiencies do not appear to be sufficiently large to create profitable betting strategies."

So what policy conclusions can be drawn from all this data? In chapter nine, we are told that the main cause of competitive imbalance in English football is not owners or players but TV. (Not much we can do about that.) We are told that a breakaway Superleague is now unlikely but that the Atlantic League is a definite possibility. Player wage inflation is a consequence of an arms race driving all the clubs to penury, encouraged by the European Union ( qv the Bosman ruling). Media interest in football club ownership is set to continue, with the internet becoming increasingly important. But "whether the kinds of changes unleashed by market forces in the highly charged world of modern-day professional football are good or bad is a normative question, of a kind that economists are usually ill-equipped or unwilling to address".

Call me a wonker, but I think we have more to offer than this.

Stefan Szymanski is professor of economics, The Business School, Imperial College London (S****horpe United).