Necessary Conditions for Valid Benefit Transfers

Benefit transfer is an approach to estimating costs and benefits of policies in the absence of original data collection. Many different methods have been proposed (Navrud and Ready 2007). They all use estimates from one or more previous studies to predict the benefits from a new policy at a different point in space, time, or both. Under what conditions will these predictions be valid? Previous work on benefit transfer methodology has addressed this question by suggesting informal guidelines for the selection of previous studies (Water Resources Research 28(3), 1992 and Ecological Economics 60(2), 2006). There is consensus that the study sites should be “similar” to the policy site in terms of their consumer populations and in their provision of the good being valued. The need for similarity is also emphasized in Environmental Protection Agency (EPA) guidelines for benefit-cost analysis. However, there is no formal test for judging whether two sites are sufficiently “similar” to conduct a benefit transfer.