Examining the wage gap between public and non-public sector in China: Insights from a robust quantile selection model

This study examines the wage gap between public and non-public sectors, which constitute a key component of overall wage inequality in China. While quantile models are traditionally employed to analyze such wage gaps, their conventional implementations face methodological challenges, particularly in addressing endogeneity issues stemming from restrictive assumptions. Our empirical analysis of China's sectoral wage gap reveals that these conventional models fail to satisfy model specification tests, primarily due to bias introduced by residual terms in regression coefficient predictions. To address these limitations, we employ a robust quantile selection model that mitigates the restrictive assumptions. The results demonstrate a significant public sector wage premium—approximately 25 % on average—with the disparity being most pronounced among low quantiles. The analysis also reveals that previous models' prediction systematically underestimated public sector wage spillovers, and in some cases, mistakenly concluded that public sector wages were lower than those in the non-public sector. A detailed decomposition indicates that observable characteristics can explain over 80 % of the total wage gap in 2018. In comparison, the more extensive implementation of market-oriented reforms in state-owned enterprises corresponds to a reduced wage premium and unexplained portion. Finally, factor analysis identifies returns to human capital, demographic composition changes, and risk preferences emerge as key drivers, while the impact of industrial restructuring appears negligible.