生命周期偏误与中国教育收益率元分析

Life-Cycle Bias and a Meta-Analysis of the Estimates about Returns to Education in China

由于终身收入难以获得,在实证分析中以当期收入替代终身收入是研究领域的普遍做法,有研究表明这种替代会造成生命周期偏误,在估计教育收益率的过程中,生命周期偏误可能是比内生性偏误更严重的问题。本研究将对以当期收入估计的中国教育收益率进行元分析,对其中可能存在的生命周期偏误进行评估,对教育收益率的长期变动趋势进行再认识。结果表明,现有研究得到的我国教育收益率增长趋势,可能有很大一部分要归因于人口老龄化导致的样本平均年龄增长。

As the data of lifetime earnings is difficult to obtain, a common practice in the literature is applying current income instead of lifetime earnings, and life-cycle bias emerges because current income is not parallel to lifetime earnings. When we estimate the returns to education, the life-cycle bias may be a more serious problem than the endogenous bias on which researchers have paid much attention. This paper conducts a recta-analysis to investigate how life-cycle bias could influence the estimates of returns to education and their time tendency in China. The result suggests that life-cycle bias is serious and a large part of the growth of returns to education in China is due to sample average age growth caused by an aging population.