LAUSR

This paper examines reputation as motive for lying aversion. In a control treatment, participants roll a six‐sided die and report the outcome, which the experimenter cannot observe. In a digital die treatment, the outcome of the die roll is determined randomly on the computer. Contrary to prior literature, we reduce ambiguity in the digital die treatment by making observability common knowledge. We find that partial lying and full lying disappear when the experimenter can track participants' behavior. This result can be explained by reputational costs: Participants care about how they are viewed by the experimenter and abstain from lying.