基于空间相关性的人力资本模型估计

Estimation of Human Capital Model Based on Spacial Correlation

未考虑样本个体间相关性的人力资本模型高估了教育收益率。基于CLDS 2012的微观调查数据,引入经济地理学中的空间相关性,采用空间人力资本模型考察了引入空间相关性对非线性人力资本模型的影响作用。研究结果表明:(1)引入空间相关性的教育收益率虽仍然较高,但受个体所处环境影响较强,较未考虑空间相关性时的教育收益率有所降低;(2)个人工作经验、性别、家庭背景以及个人工作能力等个体特性受所处环境的影响较弱;(3)人力资本市场仍然存在相当程度的性别歧视。因此,需要改善教育培养方式,减小就业环境对毕业生收益率的冲击;需要有效提升代际收入流动性,破除社会阶层的进一步固化;需要继续完善性别平等制度体系,进一步释放市场潜能。

The human capital model without considering individual correlation among the samples over-estimates education's rate of return. Based on the microscopic survey data of CLDS 2012 and by introducing the concept of spacial correlation in economic geography,this paper examines the impact of spacial correlation on non-linear human capital model by applying the spacial human capital model.The result shows that( 1) The education's rate of return is comparatively lower when introducing spacial correlation because it is influenced by the environment to greater extent.( 2) The individual character such as personal working experience,gender,family background and person working ability is influenced by the environment to a lesser extent.( 3) Gender discrimination is still quite obvious in human capital market. Therefore,the reform of higher education is necessary in order to alleviate the shock of employment environment on graduates' rate of return; the inter-generational income mobility should be fastened to break the class consolidation; the gender equality system should be built to exploit the biggest market potential.