Communication in Repeated Games Under Private Monitoring : An Experiment

We experimentally examine whether communicating private signals after each stage game affects cooperation in an infinitely repeated two-player prisoner’s dilemma where players can only observe imperfect and private signals about the partner’s actions. Based on evidence from two specifications of stage payoffs, we find that when the return to cooperation is low, communication increases the probability of cooperation; however, this effect is not significant when the return to cooperation is high. We also show that subjects use various strategies to reveal their private observations, with the majority of their messages being truthful. By comparing games with communication to games with imperfect public signals, we further demonstrate that subjects respond to the message received from the partner in the same way that they respond to the public signal about their own past action. This finding suggests that the mechanism by which communication works is similar to that of public monitoring