Abstract

In most countries, women systematically outperform men in academic achievement across fields of study. Yet within a year of graduation, they earn less, face lower employment rates, and are more likely to work part-time. If human capital were the sole determinant of pay, this pattern would be difficult to reconcile. We address this puzzle by extending the statistical discrimination framework ‘a la Phelps (1972) to include not only human capital but also additional components of productivity, such as IT skills and mobility intentions -the willingness to travel or relocate for work -which might capture candidates’ technological proficiency and adaptability. Using rich microdata from the AlmaLaurea survey of master’s graduates from the University of Bologna (2015–2022), we show that while human capital alone predicts no gender wage gap in favor of men, combining it with mobility intentions reproduces the early wage disadvantage observed for women in Economics and Engineering. We further show that IT skills -an observable CV trait constructed from multiple IT-skill items- reduce the residual gender wage gap, especially in Engineering. Our findings highlight the importance of complementing human capital with field-specific preference and skill traits to explain—and potentially address—early gender wage gaps.