Conditional cooperation under uncertainty: The social description-experience gap

Conditional cooperation is typically studied in experimental settings where the behavior of others is known to subjects. In this study, we examine conditional cooperation under uncertainty. Using a novel experimental design, we exogenously manipulate the likelihood that a subject's partner in a Prisoner's Dilemma will cooperate. Information about the partner's cooperation is either presented descriptively or learned through experiential sampling. We observe a description-experience gap: subjects are more likely to cooperate under experience when their partner's probability of cooperation is low, while the opposite holds when it is 50% or higher. This result contrasts with expectations deriving from the individual choice literature, where rare events are typically underweighted in experience-based decisions. We find that the gap we observe is driven by conditional cooperators being less responsive to social information acquired experientially than descriptively. Furthermore, we show that stronger priors held by subjects under social uncertainty compared to individual uncertainty can account for the disparity with the individual choice literature.