The power of foregone payoffs: a mousetracking study

Behavior in two-player laboratory games has been observed to depend upon choices that the other player "could have made," in violation of the principle of subgame perfection. Models of other-regarding preferences that only transform payoffs at end-nodes (e.g. inequality aversion) cannot explain this behavior, and various explanations (e.g. models of intention-based reciprocity) have been proposed. We explore the mechanisms by which foregone payoffs influence decision-making in a variety of two-player, two-stage games using mousetracking, a technology that allows us to observe which payoffs subjects attend to, and for how long, when making strategic decisions.