中国工会存在工资溢价吗* ——基于控制样本选择性偏差的BlinderOaxaca回归分解
DO CHINESE TRADE UNIONS HAVE WAGE PREMIUM ——Based on BlinderOaxaca Regression Decomposition of the Sample Selection Bias Controlled

摘要 本文采用2008年中国社会综合调查数据，运用工具变量的方法，研究了中国工会的工资溢价问题。研究结果表明，工会会员与非工会会员之间的小时工资差异主要是由禀赋效应造成的，系数效应和交叉效应并不明显。换句话说，工会工人的资源禀赋和工作特征禀赋等是造成工资差异的主要原因，工会对工资没有明显的直接影响。工会会员在入会倾向方面具有明显的群体特征。最后本文提出了加强工会有效性的两点建议。
Abstract： This paper attempts to analyze the wage premium of Chinese trade union by adopting the method of extraction of instrumental variable, based on China social comprehensive survey data in 2008. The results show that hourly wage difference between union members and nonunion members is mainly caused by the endowment effect. The coefficient effect and interaction effect is not significant. In terms of membership tendency there have been obvious group characteristics. Final conclusion is that unions reduce the wage gap between the union and nonunion members by affecting resources endowment and working characteristics of union workers. This paper put forward two suggestions to strengthen the unions effect.