Carry a big stick, or no stick at all An experimental analysis of trust and capacity of punishment

We investigate the effect of punishment in a trust game with endowment heterogeneity in which the investor may punish the allocator at a cost. Our results indicate that the effect of the punishment crucially depends on the investor’s capacity of punishment, that is measured in our experiment by the proportion of the allocator’s payoffs that the investor can destroy. We find that punishment fosters trust when the capacity of punishment is high (i.e., when the cost of punishing is relatively low). Otherwise, punishment fails to promote trusting behavior, crowding out intrinsic motivation to trust. Trustworthiness is higher with punishment than without punishment, except if investors have a high capacity of punishment