Co-author of the book behind the article that was the subject of yesterday’s post, Stefan Szymanski, has very kindly taken the time to expand on the subject for those of us yet to receive our copies of his and Simon’s new book. He did so in the comments to the original post, but I felt they deserved a post to themselves:
In the book we don’t explain this is much detail, but I have explained the methodology elsewhere in print- in academic articles and in my now ancient book Winner and Losers: the business strategy of football. Essentially the data is taken from the accounts of football clubs and consists of the total wage spend. So this includes all employees. However, we can be fairly sure that most of the wages of a club are paid to the players. The salaries of players are largely determined contractually before the start of the season- the bonus elements are quite small, especially for the players that cost the big bucks. I even co-authored a paper once that tested whether causality ran from wages to position or from position to wages- the evidence suggested that it went from wages to position.
The point people are making about the 8% is fair enough, but remember the 8% is distributed across all clubs. For example, if every club except one got exactly waht it paid for, then the 8% would all go to one club, whose managers would clearly be geniuses. But in reality most clubs are little above or below the performance line, and so no one is doing much better or worse than expected. I have examined these relationships for English leagues between 1970 and 2007 (before that wages did not explain the variation in performance very well, because the variation in wages was quite small) and the only manager who ever really stood out for me was Brian Clough. I agree that Wenger and Ferguson are great managers, but I think their skill lies in persuading the money men to back them and their investments, rather than getting better results than the resources would justify.
The real sticking point is whether the successful managers are paid so much that their wages are explainign the variation in position. I just don’t think this is credible- managers are generally paid less than the stars, and it is their wages that make up for most of the variation in club wage spending.
Stefan ends by saying that “This might give you more ammunition to disagree, but I hope it clarifies what’s been done”. As I said in my own comments, I myself am not equipped with the kind of statistical knowledge to do more than express a comparatively unfounded opinion, but I know that those of you who do have a command of this area will find this additional information well worth having.
Many thanks again, Professor Szymanski.