Optimistic behavior when a decision bias is costly: An experimental test

Abstract

The existence of optimism when biased decisions are costly is investigated experimentally. Subjects make an informed guess about an exogenously determined event: the outside temperature a couple of days later. Payments in the control group depend only on how close the guess was to the realized temperature, while payments in the treatment group also depend positively on realized temperature. The data show the treatment group subjects to guess higher temperatures, that is, higher payments, compared with the control group. Results thus indicate that there is optimism also when a decision bias is costly. © 2010 Western Economic Association International.