Abstract: Based on the CGSS data, this paper uses the quantile regression method to estimate and decompose the wage differentials of formal employment and informal employment. The results show that the wage of formal employment is signally higher than that of informal employment, and on the three representative quantiles, the rate of return to education is higher in formal employment than in informal employment. At the same time, in both formal and informal employment, the rate of return to education has a tendency to rise first and then fall with wage quantile changing from low to high. Compared with formal employment, the gender discrimination in informal employment is more serious and becomes more and more serious with wage quantile rising from low to high. Wage decomposition shows that market discrimination is significantly higher than feature differences.

Key words: Wage Differentials, Formal Employment, Informal Employment, Quantile Regression

摘要： 基于CGSS数据,本文利用分位数回归方法对中国正规就业与非正规就业的工资差异进行估计并对工资差异分解分析,发现正规就业工资水平显著高于非正规就业,在三个代表性分位数上正规就业的教育收益率均高于非正规就业,同时两种就业方式的教育收益率随工资分位数水平由低到高而呈现先升后降的趋势。相比于正规就业,非正规就业的性别歧视更为严重并随着工资分位数水平由低到高而愈强。工资差异分解表明市场歧视造成的差异占总差异的比例显著高于特征差异。

关键词: 工资差异, 正规就业, 非正规就业, 分位数回归