摘要
针对教育收益率测算中可能存在的弱工具变量问题,本文利用2006年中国健康与营养调查数据,结合工具变量估计框架下的各种模型设定检验,对我国正规就业者的教育收益率进行测算。检验和测算结果表明:受教育程度的变量存在内生性,个体配偶的受教育年限是内生变量受教育程度的强工具变量,而个体的出生季度是弱工具变量。广义矩估计结果显示我国正规就业者的教育收益率为10.1%。
For the weak instrumental variables problem in calculating the return to education, this paper used the China health and nutrition survey data in 2006 to estimate the return to education for China＇ s formal employments based on model specification test under the framework of instrumental variable estimation. Testing and estimation results show that education is a endogenous variable, and spouse＇ s education is a quarter-of-birth variable. GMM estimation results show that the is 10. 1%. strong instrumental variable for education, but not the return to education for China＇ s formal employments
出处
《统计研究》
CSSCI
北大核心
2014年第8期19-23,共5页
Statistical Research
基金
国家自然科学基金资助项目(71301173
71301160
71303264)
中国博士后科学基金资助项目(2012M520420)资助
关键词
教育收益率
正规就业者
弱工具变量
模型设定检验
Rates of Return to Education
Formal Employment
Weak Instrumental Variables
Model Specification Test
