Gossip in the Dictator and Ultimatum Games: Its Immediate and Downstream Consequences for Cooperation

= 240) across 39 sessions, participants acted in different roles (i.e., Person A, B, and C) in two different games-Person A was first assigned to allocate an amount of resource to Person B in a dictator game or an ultimatum game. Afterward, Person C interacted with Person A (i.e., trustee) as a trustor in a trust game. Prior to their decisions, participants (a) learned that Person B could gossip by sending evaluations about Person A's behavior to Person C prior to the trust game or (b) did not receive this information. Findings replicate previous research showing that potential gossip by one's partner greatly increases cooperation that is revealed in the resources allocated to the partner. Yet, compared to the dictator game, the presence of strategic considerations in the ultimatum game does not significantly enhance cooperation, and even makes people less likely to reciprocate others' behavior in the subsequent interaction. Interestingly, when there is no gossip, those who have played the ultimatum game, compared to the dictator game, are more trusted by others but do not vary in reciprocity in the subsequent interaction. However, when there is gossip, those who have played the dictator game, compared to the ultimatum game, are more trusted and also more likely to reciprocate others' behavior in the subsequent interaction. These findings imply that gossip invariably promotes cooperation across strategic and non-strategic situations, but the potential rejection by one's partner weakly promotes cooperation, and even undermines future cooperation especially when paired with reputation sharing opportunities. We discuss the implications of these findings for implementing reputation systems that can promote and maintain cooperation cost-effectively.