Social Expected Utility: Indifference to Others Can Influence Risk Preferences

Many of the choices we make affect other people. While the social component of decision making is well-studied, a social dimension has not yet been incorporated into existing quantitative models of choice under uncertainty. In this paper we review previous proposals for how best to include a social component into existing models and then introduce a new approach: the Social Expected Utility model. Our approach combines social discounting and expected utility models, and accounts not only for previous findings, but also leads to novel, testable predictions, including a surprising preference reversal effect. We present the results from four empirical studies that compare our model with alternative theories and we conclude that social discounting should be included in our efforts towards more comprehensive choice models.