农村家庭多维贫困测度与分析

Measurement and Analysis of Multidimensional Poverty in Peasant Households

多维贫困理论与方法更有助于对贫困的精准识别和量化。本文构建了新型多维贫困指标体系,利用中国家庭追踪调查(CFPS)数据,引入人工神经网络方法,测度并分解了农村家庭多维贫困的广度、深度和强度水平。研究结论表明:随着贫困维度的增加,多维贫困的广度、深度和强度指数下降,表明农村家庭不易发生多维极端贫困;农村家庭多维贫困指数(MPI)呈西高东低态势,表明农村家庭多维贫困具有典型的区域分布特征。此外,多维贫困指数分解结果显示,收入、金融和教育等因素是我国农村家庭致贫的主因。其中,东部地区金融因素影响最大,而中西部地区则为收入因素。研究结论为贫困的识别和精准扶贫提供了政策依据。

Multidimensional poverty theory and technique are more helpful to precisely identify and quantify poverty. This paper introduces a new muhidimensional poverty index （MPI） system and the artificial neural network method. By using the China family Panel Studies （CFPS） data, it measures and decomposes the breadth, depth and intensity level of Chinese household multidimensional poverty. The research conclusion show that multidimensional poverty breadth, depth and intensity index decrease with the increase of the dimension of poverty, that is, the peasant households less prone to multidimensional extreme poverty households; and the MPI of peasant households in the west is high while that is low in the east, which means the typical regional distribution. In addition, the MPI decomposition results show that factors of income, finance and education are the main reason of rural households poverty in our country. In the east region, finance is the main factor, while in the central and western regions, income is the main factor. It provides the policy basis for the identification of poverty and the poverty alleviation.