Increased cooperation in stochastic social dilemmas: Can it be explained by risk sharing?

A potential mechanism to explain changes in cooperativeness in the presence of risk may be opportunities for informal risk sharing. Using a novel experimental design, we show that the presence of both independent and correlated risk prevents the typical decay of cooperation in a laboratory social dilemma game. Notably, this result seems to rule out risk sharing as a possible mechanism behind the cooperation increase. Exploratory analyses tentatively suggest that behavior consistent with a risk sharing account may emerge late in the game, congruent with previous theorizing of slow learning in stochastic environments.