本文利用四阶段DEA模型测算长江经济带11省市2015-2019年的生态补偿效率，通过Tobit模型剔除外生环境因素对生态补偿效率的影响，使各省份在面临相同的外部环境下测得的生态补偿效率值更加贴合实际,将调整前后的效率值进行比较，判断不同因素对不同年份生态补偿效率的影响程度。再利用均值聚类分析对11省市进行分类，其中属于相对高效型的省市有3个，属于规模高效型的有6个，属于完全低效型的有2个，表明各省市的生态补偿效率存在差异，生态补偿工作需要中央和地方政府合作加强管理。最后用malmquist指数研究不同时间区间的效率动态变化，结果说明通过提升治理技术，管理水平，优化产业结构配置进行外部条件改善，可以使效率值发生变化。

In this paper, the four stage DEA model is used to calculate the ecological compensation efficiency of 11 provinces and cities in the Yangtze River economic belt from 2015 to 2019. The Tobit model is used to eliminate the impact of exogenous environmental factors on the ecological compensation efficiency, so that the measured ecological compensation efficiency value of each province in the same external environment is more realistic, and the efficiency values before and after adjustment are compared, Judge the influence degree of different factors on ecological compensation efficiency in different years. Then 11 provinces and cities are classified by means of mean cluster analysis, of which 3 are relatively efficient, 6 are scale efficient and 2 are completely inefficient, indicating that there are differences in the efficiency of ecological compensation among provinces and cities, and the ecological compensation work needs the cooperation of the central and local governments to strengthen management. Finally, Malmquist index is used to study the dynamic changes of efficiency in different time intervals. The results show that the efficiency value can be changed by improving governance technology, management level, optimizing industrial structure allocation and improving external conditions.

国家社科基金重大项目“绿色发展下水资源—能源—粮食协同发展与安全战略研究”(19ZDA084);长江保护与绿色发展研究院重点支持专项“多主体协同共治下长江经济带绿色发展模式研究”(B200204007);中央高校基本科研业务费项目“长江流域农业面源污染形成机制与协同治理路径研究”(B210207028)