Social learning and communication with threshold uncertainty

Using a threshold public bad game, we perform an experiment to test the effects of communication on coordination failure with various levels of threshold uncertainty. We apply two communication treatments to the coordination game: cheap talk between group members (unrestricted) and anonymous written communication from one generation of subjects to another (restricted). We find that the probability of groups coordinating and reaching the socially preferred equilibrium increases significantly with cheap talk and increases moderately with written communication. Repeated communication through cheap talk leads to a higher probability of achieving a set of payoff-dominant equilibria even in the face of threshold uncertainty.