The impact of monitoring in infinitely repeated games: Perfect, public, and private

This paper uses a laboratory experiment to study the effect of a monitoring structure on the play of the infinitely repeated prisoner's dilemma. Keeping the stage game fixed, we examine the behavior of subjects when information about past actions is perfect (perfect monitoring), noisy but public (public monitoring), and noisy and private (private monitoring). We find that the subjects sustain cooperation in every treatment, but that their strategies differ substantially in the three treatments. Specifically, we observe that the strategies are more complex under public and private monitoring than under perfect monitoring. We also find that the strategies under private monitoring are more lenient than under perfect monitoring, and less forgiving than under public monitoring.