Association for Public Policy Analysis & Management

Policy impacts in conventional cost-benefit analysis (CBA) are monetized based on either the observation of revealed preferences or the elicitation of stated preferences. In practice, this process often involves “benefit transfer,” or the application of shadow prices derived from available empirical research to the predicted impacts of a proposed policy. A large body of evidence from behavioral economics, however, suggests that in many situations, individual choices do not necessarily reveal true, or at least stable, preferences. Therefore, use of existing empirical evidence in benefit transfer requires not only assessments of it internal and external validities, but also an assessment of its benefit validity if individuals make choices inconsistent with the rational choice axioms underlying neoclassical welfare economics. This paper considers the implications of four behavioral challenges to neoclassical economics for benefit validity: violations of expected utility, large differences between willingness to pay and willingness to accept, non-exponential discounting, and addiction.