AN EXPERIMENTAL APPROACH TO INVESTIGATING HONESTY AND ASSET LEGITIMACY.

Abstract

Many people forgo benefits by being honest, even when there is no chance of being caught lying. However, factors affecting honesty have not been fully explored. In this paper, I conduct an experiment on Amazon Mechanical Turk to study whether asset legitimacy and social preferences affect dishonesty. Participants could lie without getting caught, and experimental groups differed along two dimensions: whether lying benefits oneself or others, and whether the benefit was earned by the beneficiary. Contrary to the extant literature, I find evidence of dishonesty across all groups, with cheating rates ranging from 36% to 82%. I also find that when asset legitimacy is experimentally induced, participants are 13.2 percentage points more likely to report the reward-yielding outcome if the reward accrues to oneself, but are not more likely to report the reward-yielding outcome if the reward accrues to others.