Commitment To Honesty

If dishonest behavior stems from a self-control problem, then offering the option to commit to honestywill reduce dishonesty, provided that it lowers the self-control costs of being honest. To test thistheoretical prediction, we conducted an incentivized online experiment in which participants couldcheat at a game of rock-paper-scissors. Treatment groups were randomly or invariably offered a hardHonesty-Commitment Option (HCO), which could be used to prevent cheating. Our between- andwithin-subject analyses reveal that the HCO provision significantly reduced cheating rates byapproximately 64%. Evidence suggests that the commitment device works by lowering self-controlcosts, which is more pronounced in individuals with low cognitive reflection, rather than by anobserver effect. Further analyses reveal two key dynamics. First, an individual’s frequency of not usingthe HCO reliably predicts their propensity to cheat when the option is unavailable. Second, repeatedlydeciding not to use the commitment device can become habitual, diminishing the HCO provision’seffect in reducing cheating over time. This research highlights the effectiveness of honesty-commitment devices in policy design while also noting that their disuse can become habitual, pointingto a new dynamic in the study of cheating.