The Escalation of Lies: An Experimental Study of the Repeated Deception Game

Managers' window-dressing behavior is a major concern in accounting. To prevent window dressing, it is important to clarify what situations lead people to tell lies and whether they escalate lying behaviors. We develop a lying aversion hypothesis that people tend to avoid lying, particularly when others are likely to experience serious damage from the deception, and a lying escalation hypothesis that people tell a small lie initially and escalate lying behavior subsequently, even if such escalation would eventually cause serious damage to others. Both hypotheses are supported by laboratory experiments. Our study supports the view that building internal control accounting systems is essential to prevent managers from telling lies and to identify small lies, if any, in the early stage of lying.