Returns to Schooling for Urban Residents and Migrants in China: New IV Estimates and a Comprehensive Investigation

This paper uses a new dataset, the 2009 Rural Urban Migration in China (RUMiC) to estimate returns to schooling in China using an instrumental variable (IV) methodology. After identifying a set of instruments, we conduct comprehensive validity and relevance testing of different combinations of instruments as well as robustness analysis of our estimates for rural to urban migrants and urban residents in China. We find that our estimates are in a fairly tight band for all four sub-samples examined (urban men, urban women, migrant men and migrant women). Estimates for men range from about 9.5% for urban workers to about 10-10.5% for migrant workers and are slightly higher than the corresponding estimates for women, which range from 7.5% for female urban workers to 8-9.5% for female migrant workers. Thus, private returns to education in urban China in 2009 were substantial and of similar magnitude to those for other transition countries, as well as to worldwide and developing country averages. We also find that the attenuation bias due to measurement error is generally large and more important in the migrant sample compared to the urban sample.