We assess the predictive power of a model of other-regarding preferences - inequality aversion - using a within-subjects design. We run four different experiments (ultimatum game, dictator game, sequential-move prisoners' dilemma and public-good game) with the same sample of subjects. We elicit...