中文摘要：

环境服务的支付意愿（Willingness to pay,WTP）受被调查者收入、教育水平、对环境保护认知等多种因素的影响,居民对流域水质改善政策的实施和支付意愿也具有不同的公共偏好。分位数回归（Quantile Regression,QR）方法可以明确WTP高支付和低支付群体的主要影响因素,揭示不同支付层次、支付能力和收入水平的群体对环境政策的需求和偏好,从而为流域保护政策制定提供更为精准的定量参考。本研究以密云水库为例,利用断点回归、Tobit模型和分位数回归（QR）模型3种方法,分析居民流域水质改善的支付意愿及其公共偏好问题。结果发现：3个模型的模拟结果都显示WTP与家庭收入具有显著的正相关说明高收入群体对水质改善有更大的偏好。Tobit模型和分位数回归模型对WTP具有较好的拟合效果,分位数回归分析的结果显示,在Tobit和断点回归分析中许多不显著的指标,在WTP分布的某些分位上也变得显著了,在不同分位数上,WTP影响因素存在显著性差异,仅有家庭收入在所有分位上具有显著性,在低分位上显著性指标较少,在中位数以后,显著性指标增加,尤其在99%高分位数上,所选择的9个指标都具有很强的显著性,表明高收入人群、受更高教育者以及对水质要求更高的群体是密云水库水质改善政策制定的主要目标人群。

英文摘要：

Willingness to Pay （WTP） for watershed water quality improvement is influenced by mcome level, education, environmental awareness and other factors, which appears a certain public preference. Through analyzing the preference of WTP for the water quality improvement of groups of different levels of paying, affordability and income may help to formulate policy more objectively. By approaches of comparatively examining interval regression （IR） model, Tobit model and quantile regression （QR） model, we stimulated public preferences of the maximum WTP for the water quality improvement of watershed, analyzed the demands and policy objects of various payment groups, and accordingly provided more precise quantitative reference for policy makers. This re- search took Miyun Reservoir as a case study, stimulated public preferences and influencing factors of the maximum WTP. The results of all three models indicate WTP and household income have positive correlation at the 1% significance level. The QR simulation shows that the effectiveness of the WTP factors was difference at different distribution quantile, especially in the high quantile. The chosen 9 indicators have strong significance. Except the indicator of age which has a negative correlation; the rest of the indicators, such as household income, environmental attitudes, awareness, environmental improvement demand, years of education and gender, have significantly positive correlation, which indicate who with higher incomes, higher education and higher demand of water quality is the main target groups.