1. Introduction
It is uncommon to encounter truly experimental data outside of the laboratory, especially for
high-stakes outcomes. But the Covid-19 pandemic has created exceptional circumstances,
which allow us to contribute to the small literature investigating the effects of a home crowd
on the outcomes of sporting fixtures and referee impartiality (e.g., Dohmen and Sauermann,
2016; Garicano et al., 2005; Sutter and Kocher, 2004).
1
We exploit the natural experiment
arising from the Covid-19 induced shutdown of football seasons and the later resumption
mostly behind closed doors. Using data from 6,481 football matches played before and after
the mid-season shutdown in 17 countries, including 1,498 played without spectators, we find
that the absence of crowds reduced home advantage persistently, with the gap between home
and away team punishments significantly narrowing.
Past studies have shown that playing behind closed doors, in one-off matches, reduces
aspects of football home advantage (Pettersson-Lidbom and Priks, 2010; Reade et al., 2020).
They also suggested that the lack of social pressure from the crowd affects the referee, with
fewer punishments for foul play for the team playing away from home. But it is unclear from
one-off matches whether the driver of reduced home advantage was the lack of social pressure
or unfamiliarity with playing and officiating without a crowd.
Answers to these questions are of direct interest to the multi-billion-dollar sports industry,
because they inform understanding of the role that officials play. More broadly, those running
sports have a responsibility to the fans and others who pay substantial sums, either on season
tickets or TV subscriptions, to see high quality contests that are competitive and refereed
neutrally. The betting and financial markets are also interested in any margins associated with
sporting outcomes and the nature of referees’ decisions. Recent articles on football played
without crowds in the Economist and Financial Times are testament to the widespread interest
in these matters beyond the sports pages.
2
This study also contributes rare evidence from a
natural experiment on whether individuals make different and potentially biased decisions in
1
Covid-19 led to most professional football being completely suspended and then resumed without crowds. In
the English Premier League, for example, the last match was played on 9
th
March, then the sport was locked down,
and the 2019/20 season did not resume until 17
th
June. This 13-week hiatus was mirrored in many other countries
across Europe and further afield.
2
See ‘Graphic detail: Covid-19 and football’, in The Economist, 25
th
July 2020 and ‘Net benefit: Home
advantage in play but football refs are fairer’, in the Financial Times, 18
th
July 2020.