VoxEU Column Frontiers of economic research Labour Markets

Upping their game? The impact of new contracts on football referees’ performance

After losing the football world cup final in South Africa, the Dutch press blamed the “chump” of a referee from England for losing control of the game. Yet this column presents evidence that, as one of the few countries where referees are paid a salary, English referees have the incentives to be among the best.

The only Englishman on the pitch for the world cup final in South Africa was the referee, Howard Webb. Like many referees officiating in the top tier of English football today he is paid a salary. A decade ago he and his colleagues were paid by the match – effectively a “piece rate”. The change was introduced by football authorities to improve referee performance.

Despite generating billions of euros in revenue each year, the performance and incentive structure facing referees has been little studied (Frick et al. 2008). In this column we discuss our recent research on the impact of the new forms of contract (see Bryson et al. 2010).

Until the introduction of salary contracts in the early 2000s, all referees in English professional football were paid a match fee for each game they officiated. The change was introduced to “professionalise” the top tier of refereeing.

Motivating the men in black

The new contracts offered greater income security to referees. Under the match fee system, referees could be dropped from the list of officials immediately following a poor performance, whereas the salary contracts were initially renewable after two years (and subsequently renewable annually).

The salaries also offered considerably more money than match fees. In return, salaried referees were required to attend fortnightly off-the-job training sessions to improve their fitness and ability to make the right decisions on the pitch.

The motivation for introducing the salary contracts was the huge growth in revenues generated by football as a result of lucrative television deals and the realisation that referees’ decisions were coming under increasing scrutiny in the media and among fans. The employer – the Professional Game Match Officials Board – was intent on driving up standards through salary contracts.

Economic theory suggests that the employer was doing the right thing. A classic review of research on incentives and worker performance notes that when a worker is paid a salary, “despite the fact that there is no immediate relation between pay and performance, he is likely to have incentives to exert effort because good performance will improve future contracts. Such reputational concerns imply that effort exertion can occur without explicit pay-for-performance contracts” (Prendergast 1999).

The introduction of contracts could also have a second impact, attracting better quality workers. Theory suggests that the result will be some “sorting” of workers, in which better ones seek out the salary contracts and less able workers remain on match fees or leave the occupation entirely.

A “field” experiment

Our research takes advantage of the fact that when salary contracts were introduced, referees officiating in the Championship – the second tier of English professional football – continued to be paid match fees. This gives us a “natural experiment”, in which some workers move to salary contracts and others continue to be paid under the old system.

By comparing the relative performance of referees in the Premier League and the Championship in the pre-salary period with their relative performance in the post-salary period, we estimate the impact of the salary contracts.

This approach would be problematic if Premier League referees remained in the top tier throughout and thus obtain salaries, while Championship officials remain on match fees in the second tier throughout. But this is not what happened. All referees in both divisions became eligible for the new contacts, and some referees officiate in both divisions at various points.

There are two other concerns that might affect our ability to draw inferences about the impact of salaries on referees’ performance. The first is the possibility that other changes occurred at the same time as the switch to salaries, which may have had different effects on the performance of referees in the Premier League compared with the Championship. But our analysis suggests that this is not the case.

The second concern is that the employer offered the salary contracts to the most able referees so that the contract is simply an indicator of a more able referee. We are able to deal with this problem by controlling for referee “fixed effects”.

Our data include nearly all the matches played in the top two divisions of English football in the 12 seasons between 1997/98 and 2008/09 (we have data for 11,169 of the 11,184 matches). They include all 168 referees officiating at these games over the period. Our measure of referees’ performance is the number of yellow and red cards that they issue in a game.

Yellow cards are issued as a warning to a player when he has broken the rules, either by fouling an opponent, handling the ball, or for showing dissent. A red card is shown if the player commits a second offence worthy of a yellow card. Red cards can also be shown for a first offence if the offence merits it, such as violent conduct or a foul that directly prevents a goal-scoring opportunity. Red cards lead to the player being sent off and suspended from subsequent games.

Issuing many cards is often a sign that the referee has lost control of the game. Good referees are able to deal with most incidents without brandishing cards by communicating firmly with players from the outset. This may not be true in every game, and perhaps the World Cup final is a good example of where this measure is more controversial. But studies have shown a high negative correlation between number of cards awarded and subjective assessments of referees’ performance by expert panels (Frick et al. 2008).

We find that referees’ performance improves among those who moved onto salary contracts relative to those who do not. The salaried referees issued an average of half a card less per game – a decline of one-sixth.

Our finding is robust to controls for referee fixed effects, which indicates that the result is not driven by salary contracts being awarded to better referees, although this does account for part of the effect. Nor is it sensitive to workers sorting into or out of the profession. Thus it appears that one can improve officiating at football games by using good personnel economics. Perhaps officiating standards in international competitions will begin to rise when other countries join the likes of England and Spain in paying salaries to their top referees.


Bryson, Alex, Babatunde Buraimo, and Rob Simmons (2010), “Do Salaried Workers Perform Better than Piece Rate Workers?”, CEP Discussion Paper, forthcoming.

Bernd Frick, Oliver Gürtler, and Joachim Prinz (2008), “Men in Black: Monitoring and Performance of German Soccer Referees”, in Helmut Dietl et al. (eds), Soccer: Economics of a Passion, Hofmann.

Canice Prendergast (1999), “The Provision of Incentives in Firms”, Journal of Economic Literature, 37:7-63.

525 Reads