Score contribution per author:
α: calibrated so average coauthorship-adjusted count equals average raw count
I consider two games, a stag hunt and a prisoners’ dilemma. Each game either features non-binding, costless and free-form pre-play communication or not. I study experimentally the differential effect of communication across games and whether the frequency of verbal initiative-taking suggesting cooperation varies across games. I find that communication has a larger effect on group cooperation in the stag hunt than in the prisoners’ dilemma. I also find that in the stag hunt initiative-taking is ubiquitous and initiators cooperate more often than non-initiators. In the prisoners’ dilemma, initiative-taking is less frequent relative to the stag hunt and initiators cooperate remarkably more often than non-initiators. In this case, initiators who cooperate are also more altruistic, averse to lying, and believe others are likely to cooperate compared to initiators who defect. I also find that participants often respond to initiative with agreement. Initiators who observe the other person agreeing to their proposal cooperate more often than those who do not observe agreement, in both games.