摘要
本文根据中国健康与营养调查(CHNS)1997年和2006年的数据,在考虑选择偏差和个体异质性的基础上,运用倾向性得分匹配法(Propensity Score Matching Method),分别估计了中国前后十年高中和大学的教育收益率。局部线性匹配法能有效减轻最小二乘估计因个体自选择和异质性所带来的不一致问题,而且可以对教育的处理效应进行分解。估计结果显示:个体之间存在显著的异质性;十年间各阶段教育的平均收益率都得到显著提高;高中的年收益率均大于大学的,而大学的三项政策效应差距有缩小的趋势。
Based on the CHNS data collected in 1997 and 2006, this paper uses the Propensity Score Matching Method to estimate the return to senior middle school and college education respectively while considering the problems of Selection Bias and Heterogeneity. The Local Linear Method can effectively reduce the former problems which may he suffered by the OLS and decompose the Treatment Effect of education. The estimated results show that there is heterogeneity among individuals; the average returns to these two kinds of education have increased in ten years; the yearly return to senior middle school is higher than that of college and the difference among the Treatment Effects of college is reducing. The political implication of this paper is to make the education fair and to improve income distribution among individuals government should first emphasize on popularizing the education of senior middle school well, but not rush to expand the college education.
出处
《经济研究》
CSSCI
北大核心
2013年第2期83-95,共13页
Economic Research Journal
关键词
教育收益率
异质性
选择偏差
倾向性得分匹配法
Return Rate to Education
Heterogeneity
Selection Bias
Propensity Score Matching Method
