Old Saybrook Case Study

Based on the descriptive statistics for private risk measures presented in Table 2.3, residential private properties in the town of Old Saybrook are, on average, more vulnerable to flood-related risks than those in Waterford. Accordingly, it is plausible to hypothesize that the impacts on WTP of private risk are different across these communities. That is, the influence of perceived risks (or actual vulnerability) on the WTPs for adaptation methods and outcomes for one group (e.g. residents in Old Saybrook) is not the same as for the other (residents in Waterford). Such an investigation could be achieved by pooling the two samples, and estimating a single utility function that allows for the potential differences in utilities across communities and over measures of private risk variable(s). Prior to this investigation, I disregard the variations in perceived (or actual) risk across the two communities, and estimate a pooled mixed logit model, in which I allow for potential differences in marginal utilities across the two groups via interactions of all attributes with the Old Saybrook binary indicator. The

The results are reported in Table 2.14. Despite the different specification of the mixed logit model, estimated mean implicit prices are generally similar to those based on the results of the restricted models for Old Saybrook and Waterford (Table 2.7 and 2.8). WTP differences suggest that an average household in Old Saybrook (as compared to a resident in Waterford) is willing to pay $7.33 higher to protect one percentage point of the threatened homes in the community. Also, in comparison to Waterford, an average household in Old Sayrbook is willing to pay significantly less for an adaptation policy that emphasizes hard defenses relative to the base