贫困问题与反贫困是我国面临的重大理论与实践课题，城乡居民医疗保险制度的反贫困功效备受关注。本项目在“准自然实验”和追踪研究设计框架下，采用中国健康与养老追踪调查（CHARLS）、中国健康和营养调查（CHNS）等大型微观数据，以及江西省2018-2020年追踪调查数据，综合运用双重差分法（DID）、倾向得分匹配法（PSM）等研究方法，建立面板数据固定效应模型，在全国和江西省两个层面，从城乡居民医疗保险制度影响居民医疗负担（直接的反贫困功效）和健康水平（间接的反贫困功效），并影响贫困发生率和贫困脆弱性（综合的反贫困功效）等三个方面，对其减贫效应进行评估。对于截面数据，运用倾向得分匹配法，排除人群间异质性带来的选择性偏倚；对于两期面板数据，运用双重差分法，剔除不可观察变量带来的混杂偏倚。项目研究对于优化城乡医疗保险制度设计，提升其反贫困功能，具有重要价值。

Poverty and anti-poverty are major theoretical and practical issues facing our country. The anti-poverty effects of Medical Insurance System （MIS）for Urban and Rural Residents have attracted much attention of the public. Based on the design of quasi-natural experiment and longitudinal study，using datas collected in China Health and Retirement Longitudinal Study（CHARLS）,datas collected in China Health and Nutrition Survey（CHNS）,data series collected from longitudinal survey of 2018-2020 in Jiangxi province, and building panel data models of fixed effect , we will evaluate the effects of MIS on the disease burden and health condition for urban and rural residents, and then evaluate its effects on poverty incidence and poverty vulnerability, with the methods of PSM（Propensity Score Matching Method） and DID（Difference-in-Differences Method）,on both level of the State and Jiangxi province. In dealing with cross-section datas, we will use PSM method to remove the selection bias thanks to the heterogeneity of persons between treatment and control groups. In dealing with panel datas, use DID method to elliminate the confounding bias aroused by unobservable viariables.The research project will be of great value to the optimization of the Medical Insurance System and the improvement of its anti-poverty functions.