We study the effects of different communication and monitoring structures on the emergence and evolution of cooperation in the indefinitely repeated prisoner’s dilemma. For this purpose, we employ an experimental 3×3-design. Subjects are either allowed to communicate (chat) before a supergame starts, repeatedly before every round of it, or not at all. They play under different monitoring structures: perfect, imperfect public or imperfect private monitoring. Under all three structures, communication boosts efficiency. However, under imperfect (noisy) monitoring, cooperation is only stable with repeated communication. Subjects fail to coordinate on complete contingent plans beforehand and thus need to renegotiate when bad signals occur. Beyond improving coordination, communication increases efficiency by making subjects’ play more lenient and forgiving.
(joint work with Fabian Dvorak (University of Konstanz, Germany)