Behavioral economics--which we might define as the analysis of decisions contrary to the assumption that people make choices that best promote their self interest as they see it--has a growing role in policy evaluation.  Examples abound.  Much energy policy rests on a premise that consumers and business owners fail to investments in more efficient lighting, heating, and cooling, when the value of energy saved exceeds the cost of those investments at any plausible discount rate.  A manifestation of this in climate policy is the expectation that a considerable amount of greenhouse gas mitigation can be achieved at "negative cost," reflecting the idea that the benefits of programs exceed costs even if they have not yet been adopted.  Another prominent policy example is that whether people enroll in employer-subsidized defined contribution pension plans depends on the apparently irrelevant and trivial fact of whether they have to check a box to opt into the plan, or check a box only if they want to opt out of a plan.

Whether moving past standard economics to behavioral economics to explain behavior in laboratories, these contexts, or other practical settings, raises important methodological questions.  Our focus here, however, is policy evaluation.  For that, behavioral economics challenges the foundations of benefit-cost analysis (BCA).  A defining presumption of BCA is that one can estimate how much people are willing to pay for the benefits of a policy, or to avoid the costs of a policy, from observed supply and demand.  While making those observations presents major and familiar empirical challenges, the premise of the exercise is clear--that we can infer value of something to people from their revealed preference for it.  Behavioral economics rejects that premise, and thus compels us to find some other means of estimating benefits and costs.

The purpose of this paper will be to examine alternatives.  Any alternative will require that one has a way to ascertain or impute "real" preferences; one cannot infer them from choices.  The challenge will be to see if there is a way to come up with real preferences that avoids paternalism or substitution of the policy makers’ preferences for those of the individuals affected by the policy choices.  If behavioral economics becomes an important consideration in explaining those choices, this challenge will have to be met, especially in light of the theme of the conference--the challenges of assessing efficiency and effectiveness in an age of scarcity.