Dynamic Incentives and Markov perfection: Putting the 'Conditional' in Conditional Cooperation

Abstract Many economic applications, across an array of fields, use dynamic games to study strategic interactions that are dynamic in nature. While these games will generically have large sets of possible equilibria, Markov perfection (MPE) is the main criterion for selection in applied work. Our paper experimentally examines this assumed selection across a number of simple dynamic games. Starting from a two-state modification of the most studied static environment–the infinitely repeated PD game—we work outward, characterizing the response to broad qualitative changes to the game’s features. Subjects in our experiments show an affinity for conditional cooperation, readily conditioning their behavior not only on the state but also the recent history of play. More-efficient history-dependent play is the norm in many treatments, though the frequency of MPE-like play can be predicted with a modification to an index developed for infinitely repeated games. A dynamic extension of the basin of attraction is shown to have predictive power for the selection of MPE outcomes.