CVM研究中支付意愿问卷“内容依赖性”的实证研究——以上海城市内河生态恢复CVM评估为例

Empirical Study on "Content-dependency" of WTP for Ecological Service in CVM Research:A Case Study of CVM in Water Environmental Restoration of Urban Rivers in Shanghai

为检验意愿价值评估法(Contingent Valuation Method,CVM)中争论的焦点问题——支付意愿的问卷"内容依赖性",本文以上海城市内河生态修复为评估对象,构建CVM的假想市场,设计评估顺序、评估对象尺度、嵌套物品等四重方案,通过720份问卷平行调查对同一评估对象在不同问卷中获得的支付意愿,并进行估计值比较和t检验。研究结果显示:支付意愿的数值随评估尺度的增加并不显著增加;单独评估比作为嵌套物品评估具有更高的支付意愿值;在问卷中先被评估的物品居民给出的支付意愿较高;整体物品的支付意愿小于各部分的加总。研究结果验证了国外实证研究文献报道的"范围不敏感"、"嵌入效应"、"顺序效应"、"部分—整体效应"等现象的存在。对此现象的解释主要是收入效应和替代效应。因此,在将CVM研究结果应用于公共政策中应充分考虑上述因素引起的偏差。

The validity and reliability of CVM have been questioned by many economists for quite a long time for its survey nature in spite of its wide application. One of the main issues raised is the ＇ content-dependency＇ of WTP estimates in that they are not invariant to the inclusion of other goods in the survey questionnaire. This paper investigates the context-dependent effects of WTP for ecosystem restoration of the urban fiver in Shanghai, China. Four schemes and parallel questionnaire investigations are designed and conducted aiming at how the same policy change is valued in different contexts, such as the diffterent scale, the different order, and the amount of goods. Based on 720 sample, employing the t-test and comparing the WTP estimates, the results show that： （a） the WTP is not sensible to the dimension of surface water environment valuated; （b） the sum of WTP for different parts of the surface water environment is larger than the valuation placed on it as a whole ; （e） WTP is higher when the change in water services is valued earlier in the survey instrument than later; and （d） WTP for a river is smaller when it is valued as apart of regional fiver than that valued on its own. The exsistence of scope insensitivity, embedding effect, sequencing effects and part-whole effect of WTP was proved in this study which is consistent with the report in literatures. In this case income effect and substitute effect are the main reasonable explanation to the phenomenon above. These findings are consistent with underlying economic theory, so deviation caused by the phenomenon should be considered when results from a CVM study are adopted in policy analysis.