Abstract:

This chapter discusses the sensitivity of willingness to pay (WTP) estimates based on a double-bounded dichotomous choice question format. Different estimation versions are structured by the following four categories: (1) exclusion/inclusion of particular groups of individuals;

(2) treatment of zero WTP, “do not know” answers, and protest responses;

(3) welfare statistics; and (4) distribution assumptions. Based on this classification the range of potential WTP values and their sensitivity depending on explanatory variables is analyzed when combinations vary.

Monthly mean WTP ranges between AC 3.76 and AC 8.05 (mean values) and between 0 and AC 4.04 (median estimates), respectively. Although several measures (e. g. log likelihood values, standard errors, McFadden R2) provide information about the fit of the models, the question which approach is appropriate to calculate WTP values cannot be answered without taking into account the intended purpose of the study.

KEYWORDS: avalanche accidents, double-bounded dichotomous choice, sensitivity of willingness to pay, value of statistical life