Abstract
The Contingent Valuation Method (CVM), an environmental valuation method worthy of attention, has been widely used to estimate the value of an environmental good with a non-market property based on survey data relating to the consumer's Willingness To Pay (WTP) in past decades. Nonsampling errors from measurement instruments in a sampling survey, however, often result in biased quantitative estimates of environmental values. This paper provides a practical discussion on the problems leading to nonsampling errors in CVM, through analysis of the effects of the payment vehicle on WTP in the sampling survey data valuing the surface forest park assumed to be constructed at the reclaimed land located in Tokyo Gulf. In order to concentrate on the measurement error, the CVM survey was carried out using the dichotomous choice (DC) format, and the payment vehicle in the questionnaire includes both tax and fund scenarios. Based on results from both the single-bound DC and the double bound DC, the statistical data analysis has demonstrated that the payment vehicle can affect the estimates of WTP significantly.
