Working Papers
"Return to Over-education for Graduate Workers in China: The Role of Skills Heterogeneity" (job market paper) - with Melanie Jones and Ezgi Kaya. link
The wage effect of over-education and its correlation with individuals’ skills remains problematic in Chinese literature. This paper focuses on the wage payoff to over-education among Chinese graduates by taking into account skills heterogeneity in explaining the over-education wage differences. We include a newly formed survey of China Family Panel Studies (CFPS) in the analysis, which has three clear advantages to studying over-education in China. First, it is nationally representative that covers 30 provinces in China. Second, it provides detailed information on occupations and individuals’ self-assessments on required education levels in their jobs, making it feasible to use three measurements to define over-education. Third, it also provides information on individuals’ cognitive and non-cognitive skills, which enables us to test alternative theories to explain over-education by disentangling skills from education. Our results show that in the Chinese labour market, there are up to 40% of graduate workers are over-educated. In addition, the wage penalty for over-education ranges from 17% - 24% depending on the specific measure of over-education. However, skills heterogeneity, regardless of utilisation or levels, shows a minor effect on explaining the over-education wage penalty. We also implement a Propensity Score Matching (PSM) method to check the robustness of results obtained in linear regression but could not find large differences between estimators.
"Return to Education for Workers in the Waged Sector in China: A Comparison between Urban and Rural Areas"
This paper focuses on the wage gap between individuals with different education levels/years in the Chinese labour market. A regional analysis is also conducted where the returns to education in urban and rural areas are compared. Based on the Human Capital Theory and Mincer wage equation, a significant return to education is found in China, and urban workers enjoy a higher payoff than rural workers. Also, we examine the labour market inequalities according to subgroups such as gender and sector and find out a higher return to education for women in urban areas. In addition, we cover the analysis of individuals' employment choices and find out that the responsibility for taking care of family members would motivate both men and women into self-employed and unemployed (temporary and long-term), in different areas. Different statistical methods are used to solve the endogeneity and self-selection problem in the Ordinary Least Square (OLS) regression, including Instrument Variable (IV) and Heckman Two-step methods. Results show that the return gap between urban and rural areas is largely eliminated after controlling for self-selection bias.
"Return to Education Qualities and Subjects for Chinese Graduates"
This paper studies the wage disparities among Chinese graduates in different education qualities and subjects. Education qualities refer to the qualities of institutions, which are roughly divided into three categories: College, Ordinary University and Key University. Subjects are also classified into three groups: STEP, LEM and SSAH. Empirical results show that graduates from better institutions would enjoy significantly higher wages, especially for Key Universities. However, no significant wage differences in subject groups are found. In addition, individuals’ skill levels, including cognitive and non-cognitive skills, could not largely explain the wage gaps between different education qualities.
Work-In-Progress
"Return to over-education in recent China: A penal analysis"
Though over-education condition for graduates in the Chinese labour market is found in the previous literature, China is still implementing a higher education expansion policy. This paper focuses on the time variation in the over-education incidence and the over-education wage penalty from 2014 to 2020, using data from China Family Panel Studies (CFPS). We compare the development of over-education incidence between years. In addition, the over-education wage penalty is estimated empirically using two measurements of over-education, which are objective and statistical methods. The effect of unobserved heterogeneity remains problematic in most studies on over-education in the literature. For example, innate abilities and family backgrounds would affect both individuals' selections into over-educated jobs and wages. In this analysis, penal analysis is conducted by using the fixed effects method to eliminate the possible bias on coefficients resulting from the unobserved heterogeneity.
Status: Draft in preparation