Contingent valuation with heterogeneous reasons for uncertainty

We test the hypothesis that respondents stating divergent reasons for choice uncertainty differ in their probability to vote yes in a CV setting. We introduce the use of a follow-up question used to classify uncertain respondents according to reason for uncertainty. Results indicate that respondents whose uncertainty derived from concerns about unforeseen negative impacts associated with provision of the good were more likely to vote yes, and depending on the model, that the probability of a yes vote of these respondents was consistently different than that of respondents whose uncertainty derived from concern about cost of provision or expected benefits.