Fairness in Risky Environments: Theory and Evidence

Abstract

Theories of fairness have typically used the assumption of ex-ante known pie size. Pie size, however, is rarely known ex ante. Using three simple allocation problems generally known as dictator, ultimatum and trust games, we explore the influence of ex-ante unknown pie size of varying degrees of risk on individual behavior. We derive theoretical predictions for two of these games using utility functions that capture additively separable constant relative risk aversion and inequity aversion. We test the theoretical predictions experimentally on two different subject pools: students of Czech Technical University and employees of Prague City Hall. We control for the risk attitude of our subjects through a variant of the Holt-Laury assessment instrument. We find statistically significant differences in giving behavior as a function of the degree of risk, and the degree of risk aversion, across individuals. We also find differences across the two subject pools but show that, once we control for various socio-demographic and cognitive characteristics, these differences evaporate. We discuss the policy and methodological implications of the results of our artefactual field experiment, as well as the implications for theories of fairness of reciprocity and their experimental test.