The Development and Estimation of a Latent Choice Multinomial Logit Model with Application to Contingent Valuation
开发了一个潜在选择多项Logit模型，用于研究条件价值评估中的假设偏差，通过密歇根州萨吉诺湿地保护数据发现两类“是”回答者，其中一类存在假设偏差，另一类则无。
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.