National Origin Differences in Wages and Hierarchical Positions #Evidence on French Full-Time Male Workers from a matched Employer-Employee Dataset

This paper estimates the differences in wages and hierarchical positions that can be attributed to national origin in France. Our data come from a matched employer-employee wage survey performed in 2002. The business survey provides very reliable wage data which are matched to many individual-level variables collected in a household survey. The sample of male full-time workers is decomposed into three sub-samples according to the parents' birthplace (France, North Africa and Southern Europe). The large number of executives in the sample allows us to perform a switching regression model of wage determination and occupational employment. We adapt and extend existing decomposition methods to this framework: while usual methods only take care of selection issues, we develop here a methodology which also properly takes into account composition effects due to differences in hierarchical positions when comparing mean wage gaps. Moreover the method we use only requires model estimation on the reference population and therefore yields more precise results when the sample size of the potentially discriminated group is small. Our results show no wage discrimination but a certain degree of occupational segregation yielding composition effects. Moreover, differences in the returns to some of the individual characteristics including higher diplomas might reveal mechanisms of statistical discrimination on the labor market.