The Development and Estimation of a Latent Choice Multinomial Logit Model with Application to Contingent Valuation

We offer a new approach to investigate hypothetical bias in contingent valuation using a latent choice multinomial logit model. To develop this model, we extend Dempster, Laird, and Rubin's 1977 work on the expectations maximization algorithm to the estimation of a multinomial logit model with missing information on category membership. Our model can be used to determine within-choice heterogeneity. Using data on the preservation of Saginaw wetlands in Michigan, we find evidence for two types of Yes responders in the data. We suggest that one set of Yes responders consists of yea-sayers who answer Yes to the hypothetical question but are less likely to pay the bid amount if it were real. We suggest that the second group of respondents does not suffer from hypothetical bias and are more likely to pay the bid amount if it were real. Even if the connection to hypothetical bias cannot be made, our method can be used in sensitivity analyses of willingness-to-pay estimates.