Focal Points, Gender Norms and Reciprocation in Public Good Games

We examine the impact of information regarding other people's choices on individual choice in a public good experiment with two separate treatments. In the implicit treatment, subjects do not see the average contribution of others in their group, but they can calculate it from the information available. In the explicit treatment, subjects see the average contribution of others in their group. If subjects are rational calculating agents as suggested in mainstream economic theory there should be no difference in observed behavior across treatments: agents should use all available information to make decisions. What we see instead is quite different and consistent with the presence of social norms: first, players change their behavior in response to the change in displayed information; second, changes in individual behavior produce identical group outcomes, in terms of total payoffs or efficiency across the two treatments. How does this happen? The display of the average contribution of others results in behavior consistent with a focal point (Schelling, 1960), i.e., more subjects behave as reciprocators (conditioning their contributions on the contributions of others), and fewer behave as cooperators or free-riders (unconditionally contributing a lot or a little, respectively). This change in behavior differs by gender: women behave similarly to men when they see the average contribution by others; when they cannot, they behave differently, favoring unconditional strategies of free-riding or cooperation. Men's behavior, in contrast to women's adaption, does not adjust to social cues, as suggested by Croson and Gneezy (2009).