行业属性对劳动报酬的边际效应及其细部特征——基于分位数回归模型的实证研究

目前学术界普遍使用的基于最小二乘法的传统线性回归方法不但不能分析行业属性对劳动报酬边际效应的细部特征,而且行业劳动报酬分布具有的非正态分布特征还会严重影响模型的估计结果,误导分析结论,而现代计量经济学中的分位数回归模型可以有效地解决上述问题.文章使用分位数回归模型方法对影响劳动报酬的行业属性变量进行了选择,测算了有关行业属性变量在不同部门、不同分位点上对劳动报酬的边际效应,分析了边际效应的细部特征与变化规律.

The Effects of Industry Attributes on Labor Remuneration and Their Detail Features: An Empirical Research Based on Quantile Regression Model

The detail features of marginal effects of industry attributes on labor remuneration can't be analyzed with the traditional linear regression method based on OLS used in current academic literatures,and the non-normal characteristics of labor remuneration distribution can seriously affect the estimated results of model and therefore mislead the conclusions,but such problems can be solved effectively by the using of quantile regression method of modern econometrics.With the quantile regression model,the industry attribute variables impacting the labor remuneration were selected,the marginal effects of relevant variables on labor remuneration were measured in different sectors and different quantiles,and their detail features and changing patterns were analyzed in this article.