收入识别与长期多维贫困:基于中国家庭追踪调查数据的实证分析

Income Identification and Long-Term Multidimensional Poverty:An Empirical Analysis Based on Data from China Household Tracking Survey

通过利用多层Logit模型来考察以家庭平均收入作为唯一识别标准的贫困瞄准精确度,同时利用多项Logit模型比较低保户、误保户及漏保户这三类家庭的特征,用以分析瞄准偏差的影响因素。研究发现,单维识别的精确度较低,且误保户与低保户在受教育程度、工作情况方面的差异较小,以此证明低保户的实际确定过程中会依据如家庭成员的受教育程度、患病、住房情况等其他因素,从侧面验证了扶贫工作中'两不愁三保障'原则的正确性,即在义务教育、基本医疗、住房安全等方面同样需要统筹兼顾。另外,结合A-F法和跨期持续时间分析法,构建长期多维贫困指数,得到各指标对多维贫困的贡献率,将多维贫困的研究拓展到跨期的动态领域,对贫困识别路径进行更深入的探讨。

This paper firstly uses the multi-level logit model to examine the poverty targeting accuracy with the average household income as the unique identification standard.At the same time,using multiple logit models to compare the characteristics of the three types of households,namely,the low-income households,the mis-guaranteed households and the life-saving households,The factors that influence the deviation.The study found that the accuracy of single-dimensional identification is low,and the difference between the mis-guaranteed households and the low-income households in other dimensions is small,which proves that the actual determination process of the low-income households will be based on the education level of the family members.Other factors such as illness and housing have verified the correctness of the principle of'two guarantees and two guarantees'in poverty alleviation work,that is,in compulsory education,basic medical care,and housing security,it is also necessary to make overall plans.The second section combines the AF method and the Foster(2009)intertemporal duration analysis method to construct a long-term multidimensional poverty index,and obtains the contribution rate of each index to multidimensional poverty.The research on multidimensional poverty is extended to the inter-period dynamic field to identify poverty.The path is discussed in more depth.