Abstract

We provide a model in which students must choose whether or not to cheat on a course exam. By assuming that the moral cost of acting dishonestly decreases as the number of other people who behave in the same way increases, our model explains one important channel by which unethical behavior of other individuals can influence observers’ behavior. Through the use of the Global Games approach of equilibrium selection, we build a framework that provides the micro-foundations of peer effects on academic dishonesty and shows how better student technology of cheating and higher disutility of effort make the cheating equilibrium more likely to be selected. By extending the model, we find that the peer effect strength is not affected by the level of homogeneity of the cohort, but decreases in the size of the class. Our approach may be seen as an important step towards reduced-form models of peer effects on dishonest behavior.