From shocks to solidarity and superstition: Exploring the foundations of faith

Abstract

Additive shocks can substantially increase cooperation in otherwise standard public goods game experiments. We study shocks that randomly adjust players’ earnings by a fixed positive or negative amount reported at the end of each round. These adjustments change neither the return to players’ contributions nor the information about other group members. We compare results across four treatments that employ the same group-level adjustment algorithm but frame it differently, with pre-play descriptions that range from omitting all useful information to accurately revealing its 50/50 random nature. In each treatment, overall contributions run about 50% higher than those obtained in the standard no-adjustment game. Contributions run higher still, nearly 100% over baseline, in a treatment that individualizes the adjustments, truthfully describing them as 50/50 random and separately calculated for each player. Our results contrast with those of previous studies, which add risk to public goods games in ways that directly interact with players’ contributions and typically reduce cooperation. Players’ contributions and post-play feedback strongly suggest that our results trace back to a pair of deep-rooted impulses that boost solidarity in response to external risk and rationalize the response with superstitious thinking.