Making a promise increases the moral cost of lying: Evidence from Norway and the United States

It is commonly observed that people make promises to commit future behavior to their moral ideals. But do promises work? In particular, can making a promise effectively promote honesty by increasing the moral cost of lying? We explored this research question in three high-powered experiments using representative samples from Norway and the United States (N=7,200). Based on a one-shot “mind game”, in which dishonesty is economically incentivized but impossible to detect at the individual level, we find that about 1/4 of participants are willing to lie for the chance of winning a bonus reward of $100. Crucially, dishonesty was significantly reduced, on average by 7 percentage points (25 percent), when participants were asked to make an inconsequential promise to report accurate information. This promise effect was equally strong in both Norway and the United States, and in different participant subgroups. However, promises reduced dishonesty only when the promise required active choice, suggesting that personal engagement increases the internal inconsistency that would arise from a subsequent lie. In contrast with predictions, an experimentally manipulated expression of trust did not impact honesty by itself or boost the effect of making a promise. Given recent debates about the replicability of dishonesty research in social science, the current work provides high-quality evidence of broad relevance, documenting a moderate but robust effect of promises on subsequent dishonesty. The promise effect is formalized in a simple model integrating cognitive dissonance theory from psychology with perspectives on moral decision-making from behavioral economics.