Inequity aversion models have dominated the behavioral economics landscape in the last decade. This study uses variants of dictator and trust games to provide empirical content to these models. We manipulate market features-such as competition over resources-to demonstrate that extant models cannot explain realistic manipulations of either game. For example, we show that if socially acceptable actions provide one player with a greater portion of the rents, she will put forth extra effort to secure them, to the detriment of the other person. When she can earn more than the other player through customary actions, we find a preference for selfishness.