... I was at the World Bank presenting this (preliminary, i.e., not ready for prime time) paper:
Joint Estimation of Revealed and Stated Preference Averting Behavior: An Application to Improved Municipal Water Supply in Kathmandu, Nepal
by Subhrendu K. Pattanayak and John C. Whitehead
Abstract. In this paper we demonstrate an approach to combining and jointly estimating revealed and stated preference data for drinking water quality and improved water supply. The application is to piped-water services in Kathmandu, Nepal. The revealed preference model uses averting behavior method data on the decision of whether to treat water. The stated preference model uses contingent valuation method data on the decision of whether to hook up to an improved water system. The bivariate probit model is used to jointly estimate the determinants of the decisions. We test for common coefficients across the revealed preference and stated preference data and find evidence that parameters related to the willingness to pay for health risk reductions are statistically equal while other coefficients vary. Households are willing to pay $3/month (US$) to reduce health risks. Averting behavior data suggests that water treatment generates an additional $3/month in joint production benefits (e.g., improved taste). The willingness to pay for the reduction in health risks, joint production and improved water supply is $18/month.
by Subhrendu K. Pattanayak and John C. Whitehead
Abstract. In this paper we demonstrate an approach to combining and jointly estimating revealed and stated preference data for drinking water quality and improved water supply. The application is to piped-water services in Kathmandu, Nepal. The revealed preference model uses averting behavior method data on the decision of whether to treat water. The stated preference model uses contingent valuation method data on the decision of whether to hook up to an improved water system. The bivariate probit model is used to jointly estimate the determinants of the decisions. We test for common coefficients across the revealed preference and stated preference data and find evidence that parameters related to the willingness to pay for health risk reductions are statistically equal while other coefficients vary. Households are willing to pay $3/month (US$) to reduce health risks. Averting behavior data suggests that water treatment generates an additional $3/month in joint production benefits (e.g., improved taste). The willingness to pay for the reduction in health risks, joint production and improved water supply is $18/month.
Feel free to download the paper (PDF) and/or presentation (PDF) while supplies last.
