我国工资性别差异的分布效应研究.docVIP

我国工资性别差异的分布效应研究 ――基于反事实分析法的分位数模型论证 山西财经大学 林剑、程鑫、李雪静 目 录 Abstract - 2 - 摘要 - 4 - 一、问题的提出 - 5 - 二、建模思想与模型设定 - 6 - （一）理论分析与假设 - 6 - （二）工资分布的非参数估计对性别工资差异的刻画 - 8 - （三）男性与女性工资方程的分位数回归模型 - 8 - （四）分解方法的选择 - 9 - 1. 性别工资均值差异的Oaxaca-Blinder分解方法 - 10 - 2. 性别工资差异的分位数分解模型 - 10 - 3. 反事实工资分布的构造 - 11 - 三、数据说明 - 12 - 四、初步的描述统计分析 - 13 - （一）男性与女性平均工资的差异 - 13 - （二）“天花板效应”和“地板效应”的初判 - 14 - 五、男性和女性工资方程的分位数回归系数差异分析 - 15 - 六、性别工资差异的分位数分解结果分析 - 18 - 七、结论与建议 - 20 - 附录 - 22 - 参考文献 - 23 - Abstract Using CHNS(2009) data, we firstly apply nonparametric estimation methods to observe the overall gender differential situation in China and to get an initial understanding whether there exists the glass ceiling effect or the sticky floor effects in China by plotting the gender wage gaps in the different income groups. Then we use quantile regression method to analyze coefficients differential of the influential factors in the male and female wage equations. By the establishment of quantile decomposition model based on counterfactual distribution we empirical the glass ceiling effects or the sticky floor effects from the point of distinction between urban and rural in our country. The results show that there is a stick floor effect in the urban of our country, while a strong glass ceiling effect exists in the rural of our country. The results also reveal that a lion share of the wage differential between female and male is attributable to discrimination rather than to the human capital difference between the two sexes. Comparing the productivity features return level of male with female on different quantile, we conclude that on the low position of wage distributions, education returns proportion of female is lower than that of male, and on the same time ,increasing of experience only make an significant effect on female wage which is on the low position of wage distributions. Returns proportion is lower than that of male. we also get the gender wage gaps are more obvious in blue-collar occupations，and