Second-Best Considerations in Correcting Cognitive Biases

Studies in psychology and behavioral economics have found that decision-making is replete with cognitive biases. Using stylized examples of time inconsistency, regret, and overconfidence, this paper illustrates possible interactions among them. As is generally true in second-best environments, the existence of biases does not imply the possibility of welfare improvements from correcting them. If only some biases are known, even correction of all known biases has ambiguous effects. With costly correction, the presence of some biases may be optimal. Further, if the correct decision is unknown, then the presence of biases does not imply that mistakes are made.