Third-party punishment is more effective on women: experimental evidence

Abstract

Existing experimental studies mainly focus on motivations and choices of thirdparty punishers, but only few of them detect sanction efficacy contradictory results. Our paper wants to shed light on this point. In particular, we want to detect whether the threat of being punished for unfair actions is credible and affects subjects’ choices thus, making it rational to behave fairly. To disentangle the effect of expected punishment on behaviour, we implement in the lab two experimental games - the standard Dictator Game, that is used as baseline, and the Third-Party Punishment Game that incorporates a third player who observes and may punish the Dictator. The idea is that, if the Dictator in treatment TPP believes punishment is a credible threat, s/he may decide to change her/his behaviour, that is, to behave generously in order to avoid sanctions. We find a clear gender bias: women reacted to the punishment threat by increasing their transfer to the Recipient, while men did exactly the opposite.