Using choice experiments to value non-market goods and services: Evidence from field experiments

Critics of stated preference methods argue that hypothetical bias precludes survey techniques from providing reliable economic values for non-market goods and services, rendering estimation of the total economic benefits of public programs fruitless. This paper explores a relatively new methodology to obtain the total value of non-market goods and services-choice experiments-which conveniently provide information on the purchase decision as well as the characteristic value vector. The empirical work revolves around examining behavior in two very different field settings. In the first field study, we explore hypothetical bias in the purchase decision by eliciting contributions for a threshold public good in an actual capital campaign. To extend the analysis a level deeper, in a second field experiment we examine both the purchase decision and the marginal value vector via inspection of consumption decisions in an actual marketplace. In support of the new valuation design, both field experiments provide some evidence that hypothetical choice experiments combined with ""cheap talk"" can yield credible estimates of the purchase decision. Furthermore, we find no evidence of hypothetical bias when estimating marginal attribute values. Yet, we do find that the ""cheap talk"" component might induce internal inconsistency of subjects' preferences in the choice experiment.