In three sets of experiments involving 5,432 subjects, we show that agents make more errors when doing so allows them to justify selfish behavior. We show that errors relating to addition arise when they can help to justify selfishness but are eliminated when selfish motives are removed. In...

We study cheating as a collective-risk social dilemma in a group setting in which individuals are asked to report their actual outcomes. Misreporting their outcomes increases the individual's earnings but when the sum of claims in the group reaches a certain threshold, a risk of collective...