How does probability judgment influence contingent valuation method to estimate WTP for natural disaster reduction

Contingent valuation method (CVM) is a widely adopted approach for estimating individuals' willingness to pay (WTP) for public goods. Using national survey data from Japan, this paper examines how the estimated WTP through CVM varies with different risk probabilities and risk information presentations in the context of natural disasters. The findings show that people's WTP is insensitive to the frequency of the disaster and the baseline level of the risk; the commonly observed ratio bias (RB) effect is not significant in our investigation. However, consistent with one similar study before, “group diffusion,” an opposite effect of RB is likely to occur, especially when the disparity between different risk descriptions is large. The result of this paper suggests that WTP may be influenced by the risk information provided and the way risk is framed in communications; therefore, CVM should be cautiously treated through rigorous validity checks when it is used to estimate WTP and risk is an important factor.