Horizontal Stratification of Higher Education and Gender Earnings Gap in the Early Labor Market

This study investigates whether higher education—often regarded as an equalizing force that reduces group inequalities—can effectively address gender wage gaps. Despite the narrowing of gender disparities in access to higher education in Korea, this study examines whether obtaining a university degree reduces gender wage inequality. It focuses on the effects of college selectivity and field of study, analyzing how horizontal stratification in higher education influences the gender wage gap in the early labor market. Furthermore, it compares results from traditional linear regression with debiased machine learning, which accounts for selection bias and high-dimensional data.Using data from the Graduate Occupational Mobility Survey (GOMS), the analysis reveals that linear regression estimates a consistent gender wage gap of roughly 10% across all levels of higher education. In contrast, double machine learning identifies a 15% wage gap for graduates of two-year colleges and around 10% for those with four-year degrees. By field of study, linear regression reports consistent gaps, while double machine learning identifies a 6% gap in the humanities, 10-12% in education, natural sciences, and social sciences, and 14% in engineering. This study underscores that accounting for selection bias and high-dimensional data reveals nuanced insights, uncovering significant gender wage gaps among two-year college graduates and notably larger gaps in engineering compared to other fields.