Econometrica: https://doi.org/10.3982/ECTA14673

Do people always lie when it is in their interest to do so and they won’t be caught? Traditionally, economic models have assumed that the answer to this question is yes.

figure a

Izel Photography - RFC1 / Alamy Stock Photo

In contrast, a new study by Johannes Abeler, of University of Oxford, and colleagues finds that people lie much less than expected. The researchers first performed a meta-analysis of 90 studies that experimentally evaluated lying at the group level by providing participants with incentives to misreport private information and no possibility of being caught. Despite these circumstances, lying was far from universal. On average, participants gave up three-quarters of the gains they could have earned if they lied. To facilitate further exploration and analysis of their meta-study data, the authors also created an interactive website, http://www.preferencesfortruthtelling.com/. The authors next evaluated the fit of a range of formal models capturing different explanations for honesty. They found that a wide range of models could potentially explain the meta-study data. Finally, the authors conducted four new experiments to test these plausible mechanisms. They found that truth-telling is motivated both by a preference for being honest and a preference for being perceived as honest.

The study synthesizes a large volume of evidence and provides insight into the reasons behind an important and socially beneficial behaviour.