A selection analysis of returns to education in China

This article estimates the economic returns to education in China from 1989 to 2009 using the China Health and Nutrition Survey (CHNS) dataset. We find that education returns for one additional year generally increase from 2.6% in 1989 to 7.9% in 2009. Education returns, however, may reflect signals of innate ability or the accumulation of human capital. Moreover, traditional ordinary least square estimates may be biased by selection problems and mix-ups of age group heterogeneity. We therefore estimate the marginal effects of schooling with increasing labour market experience using the Heckman selection model. We find that the education returns for one additional year decline with labour market experience, which supports the human capital hypothesis for all age groups except the group educated during the Cultural Revolution. Different dynamics of education returns in the four age groups are identified with the large influence of institutional reforms in the labour market, supporting the transition explanation of the evolution of education returns in China.