Dynamics of returns to vocational education in China: 2010–2017

The global vocational education and training (VET) sector has been central to recent economic reforms. As economies develop, high value-added industries replace low value-added ones, increasing demand for skilled workers. VET plays a crucial role in adapting education systems to meet this demand. China, a major economic player, has also upgraded its VET system, enacting policies to improve skilled workers' social status. Despite this, a long-standing tradition of undervaluing vocational graduates persists, leading to lower secondary graduates only pursuing vocational credentials if their academic performance is insufficient for academic degrees. Figures presented show a shift from lower to higher education qualifications between 2010-2013 and 2015-2017, reflecting improvements in labor supply quality. However, a mismatch between the supply and demand of skilled labor remains due to theory-oriented university curricula and graduates' preferences for "white-collar" jobs. The Chinese government's efforts to promote vocational education since the 1990s have not fully addressed this skilled labor shortage, potentially due to persistent negative perceptions of vocational education and insufficiently high labor market returns. This paper addresses the latter by examining returns to different education qualifications from 2011 to 2017 in China, focusing on the dynamics of returns to vocational education and their policy implications.

The paper cites several studies that have examined the dynamics of returns to education, including Arias and McMahon (2001), Söderbom et al. (2006), and Crespo Cuaresma and Raggl (2016). It also acknowledges the endogeneity issue in estimating returns to education, referencing Card (1999), Woodridge (2010), and Castellano et al. (2014). The authors highlight the contributions of their research, specifically addressing the dynamics of returns to vocational education in China, which is a less-researched area. They also note their use of rigorous econometric techniques such as instrumental variables (IV) and the Lewbel method to address endogeneity concerns, comparing their findings to those of Dai and Martins (2020) and Chen and Pastore (2021).

The study utilizes data from the Chinese General Social Survey (CGSS) from 2010, 2011, 2012, 2013, 2015, and 2017, focusing on waged workers. The authors employ several econometric techniques to address the endogeneity of education choices: Ordinary Least Squares (OLS), Instrumental Variables (IV) with multiple instruments (proportion of university graduates relative to the total population, upper secondary enrollment ratio, parental education, and a post-1985 education enrollment dummy), and the Lewbel heteroskedasticity-based instrumental variables method. They also employ propensity score matching (PSM) to address potential selection bias when comparing vocational and academic upper secondary education. The IV approach utilizes various instruments to predict the likelihood of individuals choosing specific educational paths, aiming to isolate the causal effect of education on wages. The Lewbel method exploits heteroskedasticity in the error term to create instruments. PSM aims to create a balanced comparison group to determine the impact of choosing vocational vs. academic education at the upper secondary level. The study includes numerous control variables such as age, gender, experience, marital status, hukou status, and occupation, as well as provincial fixed effects to control for unobserved regional variations. The data is analyzed to determine the wage premiums associated with different education levels and to compare returns to vocational and academic education at both upper secondary and tertiary levels.

The study's key findings reveal that returns to all education levels increased between 2010 and 2017. Compared to those with only compulsory education, upper secondary graduates earned roughly 20% more, vocational college graduates 50%, and university graduates 75.3%. At the tertiary level, academic education consistently yielded higher returns than vocational education (45.3% higher), although this gap narrowed over time. At the upper secondary level, OLS estimates suggested a 6% wage premium for vocational graduates, but this result was not robust to IV, Lewbel, or PSM estimations, leaving the return difference indeterminate. The study also notes that returns to upper secondary education increased more substantially than those of tertiary education relative to their baselines, potentially reflecting the aftermath of higher education massification. Additionally, a subsample analysis suggests a widening marriage wage premium for men over women during this period.

The findings address the research question by showing that while returns to vocational education are increasing, they remain lower than those of academic education at the tertiary level. The inconsistency of the upper secondary level results underscores the complexities of comparing vocational and academic pathways, highlighting the importance of using multiple econometric methods and carefully considering potential selection biases. The narrowing gap in returns between tertiary vocational and academic education suggests the effectiveness of policy efforts to improve the status and attractiveness of vocational training. The observed increasing marriage premium for men raises concerns about growing gender inequality in the Chinese labor market, requiring further investigation. The study's comprehensive methodological approach improves understanding of the dynamics of returns to education in China, offering valuable insights for education policy.

This paper contributes to the literature by providing a comprehensive analysis of the dynamics of returns to education in China, with a specific focus on vocational education. The key findings highlight increasing returns to all education levels, a narrowing gap in returns between tertiary vocational and academic education, and an indeterminate return difference at the upper secondary level. Future research could investigate the factors contributing to the widening gender wage gap and explore the long-term impacts of policy interventions aimed at enhancing vocational education.

The study's reliance on cross-sectional data from the CGSS limits the ability to make definitive causal inferences. While multiple econometric methods were used to mitigate endogeneity, residual confounding factors might still exist. The study's focus on waged workers excludes the self-employed, potentially biasing results. Further limitations relate to the specific instruments used in IV estimations and the reliance on a particular definition of vocational education.