Social preferences and group heterogeneity

Abstract

This dissertation examines how group heterogeneity shapes cooperation in public goods settings through the lens of social preferences. It argues that heterogeneity may not simply reduce cooperation, but may also transform the motivation, direction, and enforcement of cooperation. The dissertation consists of three main research chapters. The first research chapter, Chapter 2, develops a reduced-form theoretical model of preference-type heterogeneity. It shows how the coexistence of cooperators and free-riders may gradually erode normative commitment to cooperation. In the model, repeated exposure to free-riding weakens guilt from not cooperating and strengthens concerns about being exploited. The model also predicts a nonlinear relationship between free-rider prevalence and cooperation, and shows that an institutional floor on normative commitment determines long-run cooperation levels. These results should be interpreted as model-based hypotheses, not as causal evidence of preference change. The second research chapter, Chapter 3, studies identity heterogeneity through a laboratory experiment with 162 Thai university students. The design compares homogeneous and heterogeneous societies under three levels of identity strength. The results suggest that group identity may redirect cooperation rather than simply reduce it. In heterogeneous societies, strong identity shifts resources away from a universal public good toward a group-specific club good, even though total contributions remain unchanged. This pattern is consistent with parochial cooperation becoming more pronounced over repeated interaction. The third research chapter, Chapter 4, asks whether institutions can reduce enforcement bias in heterogeneous groups. In a laboratory experiment with 126 participants, it compares a default exclusion mechanism, in which the lowest contributor is automatically removed unless a majority votes to forgive, with a traditional mechanism that requires active votes to punish. The results show that default exclusion changes the enforcement process and reduces active out-group targeting. In heterogeneous societies, the mechanism operates through passive enablement rather than active punishment decisions. However, bias changes form rather than disappears. It shifts from punishing out-group members to mildly protecting in-group members. Less biased enforcement in this form also does not restore inclusive cooperation, as participants continue to allocate more to club goods than to public goods. The findings further point to a potential equity-efficiency trade-off, since higher exclusion rates may shrink the group. Taken together, cooperation in heterogeneous groups should not be assessed by total contribution levels alone. A group may cooperate less, cooperate more narrowly, or enforce cooperation in ways that depend on group boundaries. Institutions can reduce some forms of bias but may not be sufficient to fully restore inclusive cooperation. The central challenge in diverse societies is how to design conditions and institutions under which cooperation remains inclusive across group boundaries.