Replication data for: Inequality, Aspirations, and Social Comparisons

We develop a model of adaptive learning with social comparisons. Actors are more likely to choose actions that recently yielded satisfactory payoffs; satisfaction is evaluated relative to an aspiration level that reflects previous payoffs and possibly other players' payoffs. This captures the phenomenon of social comparison via reference groups. We show that if agents compare themselves to those who are doing better than they are then in stable outcomes all payoffs must be equal. If, however, agents' aspirations are driven by less ambitious social comparisons then very unequal distributions can be stable. We apply our general results to collective action problems in socio-political hierarchies and derive conditions for stable exploitation. Finally, we develop a computer simulation to illustrate the roles of inequality and reference groups.