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

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。

Recent Developments in Econometric Methods of Income Inequality Study

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.