Gender differences in cooperation with group membership

We study experimentally how males and females differ in the way same-gender peers observing their action affects their social behavior. In our experiment, people play a Prisoner’s Dilemma Game with a partisan audience watching the choice. Two groups of the same size (6-10 persons) participated in each session; these groups could be both all-male, both all-female, or one allmale and one all-female. Groups were separated into two rooms. Each person in the group played the game once with an audience of the same group (“at home”) and once with audience of the other group (“away”). Participants additionally received a 1/3 share of the active group member’s payoffs in games in which they made no choices. Behavior is significantly affected by the interaction of gender and place: males cooperate substantially more often when away, while females cooperate substantially more often when at home. At home, females cooperate significantly more than males, and when away the cooperation rates are similar. We confirm the established result in the literature that there is no gender difference in aggregate cooperation rates in Prisoner’s Dilemma, but in our environment this is an artifact produced by the balance of two opposing forces. We discuss a possible explanation for this pattern: Males and females wish to signal their ingroup peers, but males wish to signal their formidability and females wish to signal their cooperativeness.