Not as I Do: Hypocrisy Aversion and Optimal Punishment of Common Offenses

Law enforcement agents enforce rules that they might transgress in their private lives. Building from a theory of “hypocrisy aversion” where agents incur psychological costs from imposing a sanction on others for rules that they might break, the authors design a two-player game in which players are simultaneously placed in the roles of enforcer and potential transgressor. In this model, discretionary enforcement is endogenous to the size of the sanction. Conditional on rewards to enforcement and punishment that are both sufficiently small in the status quo, the authors demonstrate that price effects can be dominated by second-order hypocrisy effects, leading to rates of transgression thatincreasewith larger sanctions. The authors test the model within a laboratory experiment where individuals can simultaneously gamble at the expense of a third party and punish those they observe gambling. Examining the comparable testable predictions of models of (i) selfish agents, (ii) pro-social agents, and (iii) pro-social agents who are averse to hypocrisy, the authors find evidence of hypocrisy aversion in the rates of gambling, sanctioning, and the changing composition of agent strategies. Our results suggest that increasing sanctions can backfire in the deterrence of common criminal and non-criminal offenses.