基于混合正态模型的居民收入分布的演化研究  

Research on Evolution of Resident Income Distribution Based on Mixed Normal Model

作　　者：周雪娇 李群 张宝学 Zhou Xuejiao;Li Qun;Zhang Baoxue(School of Statistics,Capital University of Economics and Business,Beijing 100070,China)

机构地区：[1]首都经济贸易大学统计学院,北京100070

出　　处：《统计与决策》2020年第3期21-26,共6页Statistics & Decision

基　　金：国家自然科学基金面上项目(40431957340001)。

摘　　要：收入分布是反映居民收入分配结构和变化趋势的重要概念,是经济发展过程中收入分配结果的变现,是定量研究居民收入问题的基础。文章选用1989—2015年CHNS中的居民家庭人均收入数据,从微观视角探究居民收入分布的演化过程和规律。结果显示:居民收入分布在明显的向右迁移且越来越扁平化,同时右拖尾的情况也呈现愈加严重的趋势,表明收入水平在不断地提高,但收入差距也在不断地拉大。通过Yeo-Johnson转换改进了收入分布的离散程度,结合EM算法和数值法来估计收入分布中的参数值,其分布出现显著的双峰特征,在此基础上拟合居民收入服从混合正态分布,估算出基尼系数。Income distribution,an important concept reflecting the structure and change trend of residents’income distribution,is the realization of income distribution results in the process of economic development,and the basis of quantitative research on residents’income.This paper selects the data of per capita household income in CHNS from 1989 to 2015 to explore the evolution process and law of household income distribution from a micro perspective.The results show that the distribution of residents’income is moving obviously to the right and becoming flatter,and at the same time the situation of right trailling also presents a more serious development trend,indicating that the income level is constantly improving,but the income gap is also widening.The dispersion degree of income distribution is improved through Yeo-Johnson transformation,and the parameter value of income distribution is estimated by combining EM algorithm and numerical method.The distribution shows a significant bimodal feature.On this basis,residents’income obeys mixed normal distribution,with Gini coefficient estimated.

关 键 词：收入分布 Yeo-Johnson转换 收入差距 混合正态分布 

分 类 号：F222[经济管理—国民经济]