不可观测能力、群分效应和行业工资差距:行业分割视角  

Unobservable Ability, Sorting and Inter-Industry Wage Differentials: A Perspective of Industrial Segmentation

作　　者：田柳 周云波 沈扬扬 Tian Liu;Zhou Yunbo;Shen Yangyang

机构地区：[1]清华大学创新发展研究院 [2]中国特色社会主义经济建设协同创新中心 [3]北京师范大学经济与资源管理研究院

出　　处：《世界经济》2018年第12期98-120,共23页The Journal of World Economy

基　　金：国家自然科学基金面上项目(71672086);教育部人文社会科学重点研究基地重大项目(12JJD790050)的资助

摘　　要：可观测特征、工作条件相似的劳动者仅仅因为进入不同行业就会获得显著不同的工资报酬,本文试图从行业分割的视角分析这一问题。不同行业的劳动者存在系统的差别,因为不可观测能力异质性的存在,个体理性决策使得不可观测技能接近的劳动者聚集在相同行业,不同行业间的劳动者处于隔离状态。这种行业分割和行业工资水平的分化本质上是由群分效应引起的,而且群分效应与个体的自选择效应形成联动机制,进一步扩大行业工资差距。本文利用2013年中国家庭收入调查的城镇数据,并采用平均处理效应估计方法进行计量分析,发现控制其他可观测变量后不可观测能力基本可以全部解释行业工资溢价。Why are workers with similar observable characteristics and working conditions receiving significantly different wages merely because they enter into different types of industries?This paper attempts to answer this question from the perspective of industrial segmentation.The workers indifferent industries will have systematic differences,especially in the heterogeneity of unobservable abilities,so that the equilibrium result of the individual s rational decisions leads to the workers with similar unobservable abilities concentrated within the same industry,and the workers in different industries will thus be segregated.This type of industrial segmentation and the inter-industry wage differentials are essentially caused by the sorting effect.Sorting and worker s self-selection effects will lead to the formation of a linkage mechanism,further strengthening the expansion of inter-industry wage differentials.The paper uses urban data from the China Household Income Survey(CHIP)of 2013 and the average treatment effect estimation method to implement econometric analysis.We find that after controlling for other observed variables,the unobservable ability can almost fully explain the inter-industry wage premium.

关 键 词：不可观测能力 群分效应 行业工资差距 平均处理效应 

分 类 号：F249.24[经济管理—劳动经济]