• Note: This paper is part of the PIEP (Pay Inequality and Economic Performance) Project financed by the EU under the V framework (contract N.HPSE-CT-1999-00040). We are grateful to Simona Comi, Carlo Dell’Aringa, Paolo Ghinetti, Robert Elliott, PIEP members and to the participants at seminars in Università Cattolica-Milan, Université Paris 2, Paris and XV EALE Conference held in Seville, for useful comments. We are grateful to IRESCO-CNRS (France), Bank of Italy (Italy), and U.K. Data Archive (Great Britain) for supplying national data set. The usual disclaimer applies. We are also grateful to two anonymous referees for their comments.

*Claudio Lucifora, Università Cattolica, IZA, ERMES, 1 Largo Gemelli 20123 Milano, Italy (


We investigate public–private pay determination using French, British and Italian microdata. While traditional methods focus on parametric methods to estimate the public sector pay gap, in this paper, we use both non-parametric (kernel) and quantile regression methods to analyze the distribution of wages across sectors. We show that the public–private (hourly) wage differential is sensitive to the choice of quantile and that the pattern of premia varies with both gender and skill. In all countries the public sector is found to pay more to low skilled workers with respect to the private sector, whilst the reverse is true for high skilled workers. When comparing results across countries, we find that where pay formation is more regulated (i.e. as in France and Italy) the public sector pay gap is smaller; whilst where market factors play a larger role in pay determination (i.e. as in Great Britain) the public sector pay gap is larger—particularly in the lower part of the wage distribution—and females are much better off in the public sector as compared to the private sector.