Why do third party punish?: Second-order cooperation of in-group members

Abstract

Recent experimental studies have demonstrated that third-party participants are willing to pay a cost in order to punish cheaters in social exchanges, even if no discernable advantage can be gained by the individual him or herself. In this paper, we argue that third-party punishment is actually a form of second-order cooperation used to maintain first-order cooperation during social exchange. Given that first-order cooperation is often extended preferentially toward ingroup rather than outgroup members, we predicted that third-party punishment should be similarly biased towards ingroup members. Shinada, Yamagishi & Ohmura (2004) showed that cooperators punish non-cooperative in-group members more severely than they do non-cooperative out-group members, but noncooperators punish non-cooperative out-group members more severely than in-group members. We argue that the competitive social motivation to enhance one's relative standing against others is the dominant motivational basis of punishing out-group members. To test the hypotheses that third-party punishment is directed to in-group members, and out-group punishment is reduced punishing the other does not put one at an advantage, participants were given the opportunity to punish ingroup members (undergraduates from the same department) and outgroup members (from another department) during a gift-giving game. The results supported our hypothesis; both cooperators and non-cooperators paid more money to punish ingroup cheaters than outgroup cheaters. (PsycINFO Database Record (c) 2012 APA, all rights reserved) (journal abstract)