무응답 자료 처치모형을 이용한 서울시 수돗물 수질개선편익 추정

willingness-to-pay (WTP) survey data for implementation of the contingent valuation method often have a significant number of missing responses. In this case, an analysis of the observed data after deleting the missing data leads to incorrect welfare analysis results. This study attempts to develop a new model to correct for the missing response observations in a simpler univariate setting rather than in a complicated bivariate setting adopted in previous studies. The new model is applied to the empirical analysis of the survey data for measurement of tap water quality improvement benefits in Seoul. Non-response bias was not detected. However, no correction for the missing observations overestimated household monthly WTP by 146 to 181 won, which amount to 6.51 to 8.07 billion won when expanded to Seoul residents. The household monthly WTP estimated from the new model ranges from 3,808 to 4,434 won.