后扶贫时代中国农村相对贫困的一种测定方法与应用研究

Study on the Measurement of Rural Relative Poverty Standard in China in the“Post-Poverty”Era

研究目标:从需求层次递进结构出发,设计一种后扶贫时代中国农村相对贫困的测量方法。研究方法:根据相对贫困的内涵,改进卡方自动交互检测(CHAID)模型方法,将整体人群利用"支出-收入"回归算法进行分类,从全部"支出-收入"分类组中挑选相对低的组作为农村相对贫困样本组,以样本组最低收入值推断农村相对贫困线测度相对贫困人口。研究发现:CHAID模型的方法功能与相对贫困测量内容及指标间的理论关系吻合,能够跟踪绝对贫困消除后人的需求变化进而反映出相对贫困状态;与绝对贫困标准测量方法相比,CHAID方法保留了相同的收入测量指标,在相同需求范围内同样利用了线性回归方法,使贫困标准转换时基本原理不变,有助于国家扶贫经验的持续推进。研究创新:借鉴国际相对贫困标准比例法与指数法的收入指标与多维特征,通过非线性多元统计功能及递归原理实现了追踪人的需求变化以及归类相同需求人群的功能,实现了对相对贫困人口的测度。避免了比例法无法考察相对贫困内容及受数据分布形态影响的不足,同时也规避了指数法多维指标选取的主观性。研究价值:为绝对贫困消除后如何度量相对贫困提供了一种方案,丰富了"建立解决相对贫困长效机制"的理论体系。

Research Objectives: Starting from the progressive structure of the hierarchy of needs, designs a method to measure the relative poverty in rural China in the post-poverty era. Research Methods: According to the meaning of relative poverty, the CHAID model method is improved, the whole population is classified by using the "expenditure-income" regression algorithm, and the relatively low group is selected from the whole "expenditure-income" classification group as the rural relative poverty sample group, and the relative poverty line is inferred from the minimum income value of the sample group to measure the relative poverty population in rural areas. Research Findings: The function of the method is consistent with the theoretical relationship between the measurement content and indicators of the relative poverty, which can track the changes in human needs based on the absolute poverty measurement content and reflects the relative poverty status. Compared with the measurement method of the absolute poverty standard, the measurement method of the relative poverty retains the same income measurement indicators, and also uses the linear regression method in the same demand range. This treatment keeps the basic principles unchanged when the poverty standard is changed, which is helpful to the continuous advancement of the county’s poverty alleviation experience. Research Innovations: Based on the income index and multi-dimensional characteristics in proportion method and index method of the international relative poverty standard, this paper uses the non-linear multivariate statistical function and recursive principle to implement the function of tracking change in human needs and categorizing people with the same needs, which avoids the deficiency that proportion method cannot investigate the content of relative poverty and is affected by the data distribution, and avoids the subjectivity of the multi-dimensional index selection. Research Value: It provides a plan for how to measure relative poverty after absolute poverty is eliminated, and also enriches the theoretical system of "establishing a long-term mechanism to solve relative poverty".