Referee bias in football: Actual vs. expected additional time

Abstract

This study introduces ball-in-play time as a novel variable to assess potential referee bias in football concerning additional time decisions. The study begins by calculating the expected additional times using ball-in-play data. Hence, any systematic difference between actual and expected additional time can be considered ‘bias’ if the difference depends on the winner. By using matches played behind closed doors during the Covid-19 pandemic as a natural experiment, the analysis reveals a trend that referees exhibit a bias towards home teams, but crucially, only in the presence of fans. This suggests that social pressure exerted by fans significantly influences individuals’ decisions. Moreover, the magnitude of this bias increases with crowd size. Then, the study examines the post-pandemic period, exploring whether referees adjusted to a more impartial stance due to their experience during matches played behind closed doors. The propensity to favour home teams remains prevalent once fans return to the stadiums. Furthermore, the study shows no significant association between referee experience or quality and referee bias. The study concludes with a recommendation to mitigate the bias.