Abstract

Cohn et al. (2019) designed the field experiment about the lost wallets across 40 countries to examine whether people attempt to contact the owners to return the 17,000 wallets. We discussed the design flaw in their experimental settings by reanalyzing the relationship between the rates of wallet return, in the Cohn et al. (2019)’s data, and the percentage of the Internet penetration (over population) as an upper bound of proportion email users. We found that countries with limited access to email have a lower rate of wallets’ return after controlling other factors. Furthermore, we revisited the Abeler et al. (2019)’s aggregated data to study whether the dishonest behaviors in the laboratory could predict the actual honesty behavior or not. It turns out that what happens in the lab makes no sense to our reality. This comment contributes to the extant literature about an experimental designation for honesty studies.