Language and lies

Does an individual's aversion to a lie depend upon the language used to communicate the lie? We adapt the Lopez-Perez and Spiegelman (2013) dot experiment to measure how a “weak” vs. “strong” message affects individuals' propensities for truthfulness when there is a monetary incentive to lie and no other person is affected by the communication. Weak messages state a fact, whereas strong statements “solemnly swear” to the fact. In our first (between-subject) experiment, strong (vs. weak) statements increase the percentage of subjects choosing to tell the truth by approximately 30 percentage points in each of three different payoff scenarios that favor lying to a different extent. Because lies increase payoffs in the experiment, the weaker aversion to weaker lies is socially advantageous. In a second (within-subject) experiment participants choose between messages of different strength and we find (1) a preference for lying with weak (vs. strong) language, and (2) a significant fraction of subjects who are willing to pay a positive amount to avoid a strong vs. weak lie. From both experiments, we conclude that our subjects tend to be intrinsically less averse to dishonesty when a lie is conveyed with weak vs. strong language.