Estimating the gender wage gap using stochastic frontiers: some modelling issues

Abstract

Purpose The purpose of this study is to analyse the gender wage gap in Italy by adopting a Stochastic Frontier Approach, with the aim of assessing whether gender differences persist after controlling for individual and job-related characteristics and how modelling choices affect the estimated gap. Design/methodology/approach The analysis is based on microdata from the 2021 wave of the European Union Statistics on Income and Living Conditions (EU-SILC) for Italy. A wage frontier model is estimated to decompose observed wage differentials into productivity-related components and wage inefficiency, with particular attention to the role played by gender in the specification of the model. Findings The results provide clear evidence of a gender wage gap in Italy, with women earning significantly less than men even after accounting for observable characteristics. Two main findings emerge. First, women display lower levels of wage efficiency, suggesting structural disadvantages in converting human capital into equivalent earnings. Second, the magnitude of the estimated gender wage gap is sensitive to modelling choices, particularly with respect to whether the gender dummy variable is included in the wage frontier or in the inefficiency term. Originality/value This study contributes to the literature on gender wage disparities by applying the Stochastic Frontier Approach to recent Italian data and by highlighting the importance of model specification in shaping conclusions about the size and sources of the gender wage gap.