Social Identity and Cooperation.

Abstract

This dissertation examines the effects of social identity in an economic context.

The first chapter studies when a common group identity improves efficiency in coordination games. To address this, we propose a group-contingent social preference model and derive conditions under which social identity changes equilibrium selection. We test our predictions in the minimum-effort game in the laboratory under parameter configurations which lead to an inefficient low-effort equilibrium for subjects with no group identity. For those with a salient group identity, consistent with our theory, we find that learning leads to ingroup coordination to the efficient high-effort equilibrium. Additionally, our theoretical framework reconciles findings from a number of coordination game experiments.

In the second chapter, we investigate the effects of lending team competition on pro-social lending on Kiva.org, the first peer-to-peer microlending site which matches citizen lenders with entrepreneurs in developing countries. We find that lenders are more similar to their team members than to those outside the team. When choosing a team, lenders pay more attention to team rank than to member characteristics. Furthermore, joining a lending team increases lending, but mostly for those who are not already engaged in the site. Lenders who join teams make 1.8 more loans per month than those who do not, which translates to at least $45 per month. Using a random sample of team forum data, we find that teams who post many links to specific loans increase their lending.

The third chapter studies the endogenous choice of groups, which is an important but underexplored aspect of social identity. While group membership is frequently voluntary in real-world situations, most studies of group effects limit their analyses to either pre-determined or induced groups. In this study, I separately examine how endogenous group choice and intergroup comparison affect pro-social behavior in a laboratory setting. Allowing subjects some choice over their group membership increases pro-social behavior, while intergroup comparison does so only when subjects care about their groups. Finally, I calibrate a learning model incorporating social preferences, yielding estimates of the subjects' group-contingent other-regarding preferences.