The Persistence of Lying

We study experimentally if lying is persistent. In a strategic game, we manipulate the incentives to lie to others. Some participants are first exposed to high incentives and then lower incentives; for others the reverse. We conjectured that participants will lie more often when the incentives are high, and will continue to lie when the incentives are then reduced. Contrary to our expectation, we do not find any such persistence in behavior, and there even seems to be an overcompensation effect: participants formerly exposed to high incentives are less likely to lie when confronted with lower incentives than participants who were never exposed to high incentives. The overcompensation is still present (but not statistically significant) after correcting for a general time trend.