Inequality Aversion in Ultimatum Gameswith Asymmetric Conflict Payoffs

Assuming inequality averse subjects as modeled by Fehr and Schmidt (1999) or in the ERC model by Bolton and Ockenfels (2000) in ultimatum games with asymmetric conflict payoffs allows to make predictions especially concerning responder acceptance thresholds. These predictions are tested in a laboratory experimenteliciting proposer offers and respondent's acceptance thresholds using the strategyvector method. By and large both models make good predictions. However, they are unable to convincingly explain the observed selfishness on behalf of responders in ultimatum games favoring them in conflict. Overall, observed behavior gives rise to a context dependent interpretation of inequality aversion and to Knez andCamerer's 1995 observation that subjects form `egocentric assessments of fairness'.