Gender pay-gap: exploring the school-to-work transition of graduates in Italy

Purpose - In this article, the authors study the gender pay-gap (GPG) among graduates in Italy (2011 cohort) who were employed four years after graduation. The authors focus on individuals who are new entering in the labour market or who match a low level of experience with a high level of education. Design/methodology/approach - Aimed at estimating the amount of the differential between male and female average wages, the authors have applied the Oaxaca–Blinder (O–B) decomposition. The results identify the presence of a GPG at the very beginning of graduates’ careers given that, shortly after graduation, women receive lower salaries than men, even after controlling for several characteristics (individual, academic, job and local labour market). The authors completed the analysis with the reweighted O–B decomposition using the recentered influence function (RIF) and the Juhn, Murphy and Pierce and Machado and Mata decomposition approaches. Findings - The results show that the GPG is already present at the very beginning of graduates’ careers, and it increases when correcting for women’s lower level of participation in the labour market. The authors also identified sticky floor and the glass ceiling effects due to the existence of a relevant high GPG both at the bottom and the top of the graduates’ wage distribution. Originality/value - By focussing attention particularly on graduates, this paper adds to the existing literature a deeper understanding not only of individuals who have recently entered the labour market, but also those who are highly skilled but have little on-the-job experience. In fact, the authors are looking at a particular sample (graduates who are all transitioning from university to work during the same period) with small heterogeneity which allows the authors to compare very similar young men and women graduates and gain a deeper understanding of GPGs in early careers while controlling for confounding and hidden sources of variability.