Économie et Statistique n° 464-465-466 - 2013Inequalities and Discriminations: measurement issues - Indirect approaches: discrimination as an unexplained component of inequalities - Experimental approach: the contribution of testing - Subjective approaches: measuring felt discriminations
Origin and gender discrimination in hiring: is mistrust undifferentiated, or targeted to certain groups?
The existence of racial discrimination in recruitment on the labour market has been widely documented in most western economies. While this phenomenon has been confirmed for the French labour market by various studies, the causes and sources remain largely unknown. This article presents the findings of a study designed to empirically assess the main sources of discrimination in hiring in the Île-de-France region. The evaluation is based on the controlled sending of applications in response to job offers posted on public information sites. The study shows that discrimination in hiring against candidates of immigrant origin is in the order of 40% on average. This inequality of treatment between applicants affects all applications of foreign origin, independently of the origin of the candidate, a fact that contradicts the hypothesis of targeted mistrust of particular waves of immigration. Based on these findings, it seems that discrimination is much more a question of opposition between the majority “native” group and the individuals of immigrant origin as a whole than of specific difficulties with regard to certain groups. In this respect the results confirm the existence of an ethnic “homeophilia” underlying discrimination - in other words, an undifferentiated mistrust among employers of any applicant who does not belong to the majority ethnic group. Next, our findings indicate that female applications are favoured above their male equivalents and are relatively less affected by origin discrimination. Lastly, the inclusion of an explicit signal of linguistic ease on half of the applications eliminates all origin-related discrimination for the female candidatures. However, the effect of this signal is much weaker on discrimination between male candidatures.