Experiments in Identity Economics

Abstract

This dissertation furthers our understanding of the influence of identity on behavior using experimental methods. My first project investigates the influence of social norms on behavior by testing for the presence of a gender gap in competitive preferences across two populations that vary in culture. Using a laboratory experiment in which subjects select among a competitive and non-competitive payment scheme, we find a stark contrast across the two populations with one exhibiting the previously found gender gap in competitive preferences while the other exhibits no such gap. Our results suggest that nurture in the form of social norms may explain the persistence of the gender disparity in labor market outcomes rather than nature. My second project looks at group behavior more generally by studying the mechanism underlying in-group bias in trust. Using a laboratory experiment that separately identifies the ways in which the decision to trust differs from a similarly risky decision, we study how individuals’ willingness to trust depends on the identity of their counterpart. We find that individuals exhibit little to no preference over relative earnings whether the subject’s counterpart is a member of their group or another group. On the other hand, individuals are significantly less likely to take a risk when its outcome is determined by a member of another group as opposed to a member of their own group. Our results imply that the in-group bias in trust is the result of differences in the perceived emotional cost of betrayal rather than altruistic preferences towards the in-group. My final project returns to studying gender norms but does so in the context of social interaction. We study the strength of gender norms in low promotability task allocation and completion through the introduction of heterogeneous costs into a laboratory setting. We find that explicit cost differences produce an overall gender gap that is similar in magnitude to previous studies, but that the gap among low cost individuals is smaller. In an additional treatment, we find that when managers who request another participant to complete the task are introduced that the inclusion of explicit cost differences crowds out the previously found gender bias. Taken together, our results suggest that gender beliefs play a large role in the gender disparity in low promotability task allocation but that making managers aware of opportunity cost differences among their employees may alleviate this trend. My three projects illustrate that identity is influential in the development of individual preferences. Individuals recognize and conform to the behavioral prescriptions of the groups to which they belong leading to large differences in behavior between and across groups. My results suggest that a deeper understanding of the influence of norms and identity-based behavioral prescriptions is necessary to develop a deeper understanding of heterogeneity in behavior.