Social Framing and Cooperation : The Roles and Interaction of Preferences and Beliefs

Evidence suggests that there are substantial and systematic differences in cooperation rates under varying framing conditions in social dilemmas. Several explanations of these differences have been presented. Some (e.g. McCusker and Carnevale, 1995; van Dijk and Wilke, 2000) argue that social frames mainly cause subjects' preferences to change, while others (e.g. Dufwenberg et al., 2011; Ellingsen et al., 2012) argue that frame-specific terminology, such as the name of the game, mainly affects subjects' beliefs about others' behavior, which in turn affects their own behavior. This paper advances the discussion concerning the role of frames in social dilemmas by simultaneously identifying the effects framing has on both preferences and beliefs with the same players and within the same experiment. The current experiment employes a design in which the same interaction was labeled differently, such that the interaction was referred to as a Community Game, a Wall Street Game, an Environment Game, or simply as a Game in the control condition. In all four experimental conditions, we measured the subjects' (i) social preferences, (ii) cooperative behavior and beliefs in a one-shot public goods game with the strategy method, (iii) cooperative behavior and beliefs in a ten-round iterated public goods game with random group rematching, and (iv) donation decisions to a naturally occurring public good. Overall, our results show that preferences, as well as beliefs, are both significant predictors of cooperation decisions, and that framing has significant effects on these two predictors' relative weights and also on aggregate cooperation rates. However, the impact of framing on cooperative behavior is complicated, and our results indicate that the magnitude and direction of framing effects may depend on diverse and subtle context-dependent mechanisms that are not yet fully understood