Group Outcomes As A Public Signal, Focal Point Effects And Reciprocity

Download

Description/Abstract

We study experimentally how group membership affects an agent's individual behavior in an almost minimal group setting, consisting of (a) a trivial group-assignment, (b) no social interaction, and (c) anonymity. Our experimental design introduces two novel features. The �first is a careful choice of payments in the Dictator and Trust game to ensure that pecuniary incentives stay the same across the two games. The second feature is the display of group outcomes which bears no significance on agents' monetary payoffs, albeit serves the role of a public signal. In data analysis, we develop a structural model and test two hypotheses. One is that group membership in influences behavior through social identity. The other is that group membership operates via the display of group outcomes which facilitates coordination on a focal point for the group, or focal coordination in short. Our findings support the second hypothesis. The display of group outcomes induces a group cooperation effect in both the Dictator and the Trust game. In addition, we find evidence in support of conclusions found in recent experimental studies, which demonstrate that allocation choices are sensitive not only to the choice set available to the agent contemplating an action, but also to the behavior of the agent that generated the choice set.