中文摘要：

利用2013年S高校大学生就业统计数据对大学生起薪的性别效应进行实证分析,分位数回归和分解的结果表明：均值分布和条件分布上,户籍、政治面貌、获奖层次、GPA、学科专业和就业地点对大学生起薪的影响具有显著的性别差异;就业企业、就业行业对大学生起薪的影响的性别效应仅在均值和部分分位数上显著;随着分位数的提高,大学生性别工资差异扩大,在收入分布的低端,特征效应是性别工资差异的主要原因,在收入分布的高端,性别工资差异主要由歧视引起,女生的＂天花板效应＂严重。

英文摘要：

Using college students employment statistics of S university in 2013,this paper empirically studies gender wage gap.Quantile regression and decomposition results show that ： The impact of household registration,political affiliation,awards level,GPA,professional disciplines,and the location of the employment on college students＇ starting salaries has significant gen-der differences in the average distribution and conditional distribution;Gender effect of enterprise type and industry type to college students＇ starting salaries is significant in the average distribution and some quantile condition. With the improvement of quantile,college students＇ gender wage gap widenes;In low wage distribution,characteristics differential is the main reason for the gender wage gap;In high wage distribution,the gender wage gap is mainly caused by discrimination, the ceiling effect of girls is more apparent.