Isolating the Roles of Individual Covariates in Reweighting Estimation
研究了如何用重加权方法分离单个协变量对结果分布差异的影响，提出了保持其他协变量边际分布不变的新方法，并用黑人与白人男性工资差异的实证分析展示了方法差异。
A host of recent research has used reweighting methods to analyze the extent to which observable characteristics predict between-group differences in the distribution of an outcome. Less attention has been paid to using reweighting methods to isolate the roles of individual covariates. We analyze two approaches that have been used in previous studies, and we propose a new approach that examines the role of one covariate while holding the marginal distribution of the other covariates constant. We illustrate the differences between the methods with a numerical example and an empirical analysis of black–white wage differentials among males. Copyright © 2015 John Wiley & Sons, Ltd.