Social preferences on networks

Abstract

Social preferences are a powerful determinant of human behavior. We study their behavioral implications within the context of a network game. A key feature of our game is the existence of multiple equilibria that widely differ in terms of their payoff distributions. Determining which equilibrium is most plausible is thus a key concern. We show that introducing social preferences into the game can resolve the problem of equilibrium multiplicity. However, the selected equilibria do not necessarily yield more efficient or egalitarian payoff distributions. Rather, they just reinforce the inequality that is already inherent in a network structure. We validate these predictions in an experiment and discuss their implications for managerial practice and behavior in larger networks.