Self-portrayed honesty and behavioral dishonesty

A common conclusion of lab dishonesty studies is that subjects cheat to a very modest extent even when they cannot get caught. The modest level of cheating is attributed to subjects’ moral feelings which restrict their cheating to a level that enables them to retain their self-image as honest individuals. The present paper questions this claim, reporting the results of two experiments which uncover a discrepancy between self-portrayed honesty and actual dishonest behavior. The experiments reveal, first, that self-portrayed honesty is stricter than practical honesty, suggesting that when facing an opportunity to benefit from dishonest behavior, subjects trade off some of their self-image as honest persons for dishonestly-generated payoffs. Secondly, dishonesty out of lab, in subjects’ self-chosen environment, is greater than in the lab, with a considerable fraction of subjects cheating to the maximum extent possible. We suggest that it is not a concern about their self-image that holds subjects back from cheating maximally, but rather a concern about their social image. The modest level of dishonesty observed in the lab presumably reflects caution subjects exercise in an unsafe setup. Given an environment where they feel safe to reap the fruits of dishonesty without compromising their image as honest persons in the eyes of others, the image they have of themselves hardly plays a role in curbing their unethical behavior.