The effect of league design on spectator attendance: a regression disontinuity design approach

This paper exploits a sharp regression discontinuity design to identify the causal impact of the Scottish Premiership League (SPL) “split” on spectator match attendance. We use data drawn from all 19 completed seasons for which this institutional arrangement has been in place. The causal effect of the “split” is to induce, for the last five rounds of games played in the season, a differential in average attendance of about 24% between the clubs that just qualify for the “Championship Play-off” section and those that do not. However, the annualized effect for the season is found to be modest.