一种新的收入差距研究的计量方法——基于分布函数的半参数化估计  

Recent Developments in Econometric Methods of Income Inequality Study

作　　者：黎波[1] 迟巍[1] 余秋梅 

机构地区：[1]清华大学经济管理学院 [2]国家统计局城市司

出　　处：《数量经济技术经济研究》2007年第8期119-129,共11页Journal of Quantitative & Technological Economics

基　　金：国家社会科学基金项目"当前部分社会成员收入分配差距扩大的现状;成因和对策的研究"(编号:06CJY009)资助;本文为该项目的阶段性成果。

摘　　要：收入差距扩大的成因分解方法的最新研究强调描述整个收入分布,并且对收入分布的变化进行分解,进而得到各个统计量的分解,这类方法的优点是可以显示收入差距的变化主要集中在哪些收入群体中。本文着重介绍了两类对收入分布函数进行成因分解的半参数化方法:一类是Lemieux及其合作者们发展起来的权重重置法;另一类是Machado和Mata首创的基于分位数回归的分解方法。To decompose a rise in income inequality, recent developments recommend estimating the wage distribution and decomposing the change in the wage distribution to different factors. Once this is done, the decomposition of a specific index follows immediately. Moreover the decomposition of entire distribution reveals which groups of individuals, upper or lower earners, experience a greater increase in income inequality. For this reason, these methods are particularly useful in empirical research. We introduce two semi-parametric methods of income distribution decomposition. One is the reweighting method developed by Lemieux and his collaborators. The other is the approach based on conditional quantile regressions, initially proposed by Machado and Mata.

关 键 词：收入差距扩大 分布函数 参数化 统计量 分解方法 估计 收入分布 分位数回归 

分 类 号：F126.2[经济管理—世界经济]