Do People Accurately Anticipate Sanctions?

We provide lab data from four different games that allow us to study whether people have accurate expectations regarding monetary sanctions (punishment/reward) and nonmonetary sanctions (disapproval/approval). Although the strength of the sanction is always predicted with some error (particularly in the case of monetary sanctions), we observe that (i) most subjects anticipate correctly the sign of the average sanction, (ii) expectations covary with sanctions, (iii) the average expectation is very often not significantly different than the average actual sanction, and (iv) the errors exhibit no systematic bias, except in those situations where rewards are frequent. In this line, we find some evidence that punishment is better anticipated than rewards.