Using Team Discussions to Understand Behavior in Indefinitely Repeated Prisoner’s Dilemma Games

We compare behavior of two person teams with individuals in indefinitely repeated prisoner dilemma games with perfect monitoring. Team discussions are used to understand the rationale underlying these choices and how these choices come about. There are three main findings: (i) Teams learned to cooperate faster than individuals, and cooperation was more stable for teams. (ii) Strategies identified from team dialogues differ from those identified by the Strategy Frequency Estimation Method. This reflects the improvisational nature of teams’ decision making. (iii) Increasing cooperation was primarily driven by teams unilaterally cooperating in the hope of inducing their opponent to cooperate. (JEL C72, C73, C92)