Occupational Segregation in France, 1983–2011

More women entering paid employment does not necessarily mean that women and men begin doing the same jobs. Julie Argouarc’h and Oana Calavrezo’s Dares Analyses 079, La répartition des hommes et des femmes par métiers, examines occupational distributions in metropolitan France between 1983 and 2011. Progress becomes a finding to explain rather than a general verdict.

Concentration in particular occupations

The report uses the Insee Labour Force Survey, adjusted historical series and three-year moving averages that reduce small-sample fluctuations. It groups jobs into 86 occupational families for comparison across the period. These are classifications of occupations and skills, not interchangeable with the industries of employers.

In 2011, ten occupational groups accounted for 47% of women’s employment; the corresponding ten groups for men accounted for 31%. Care, cleaning and teaching occupy important places in women’s employment. This compares concentration within each group. It does not mean that women work only in these jobs or that every worker has no choice.

An index of 52 is not a discrimination rate

Page 4 and the methodological box explain the Duncan dissimilarity index. A value around 52 in 2011 can be read as a hypothetical redistribution: at least about 52% of women, or alternatively the same proportion of men, would have to change occupations for the two distributions to coincide.

This is a descriptive counterfactual, not a policy asking half the workforce to resign. It does not identify 52% of hiring decisions as discriminatory. Equal distributions also do not require every occupation to contain exactly 50% women: the occupation must account for the same share of women’s employment as of men’s employment. These distinctions prevent an indicator of structure from becoming an accusation against individual workers.

A decline can include disappearing jobs

The index fell from roughly 56 to 52 between 1983 and 2011. The report distinguishes changes in the relative sizes of occupations from changes in their internal gender composition. The contraction of highly feminized jobs can reduce overall segregation; women entering some skilled occupations can do so through a different route. Similar movements in an index may have very different consequences for livelihoods.

Growth in care and health occupations can increase measured segregation when their workforces remain largely female. Expanding necessary services should not therefore be treated as a statistical failure, nor should eliminating those jobs become a route to a better-looking indicator. Feminist evaluation also asks about pay, employment protection, service access and men’s entry into feminized occupations.

Description leaves causal questions open

The study reveals differences in occupational allocation. Its index alone does not identify the separate causal contributions of recruitment discrimination, educational pathways or care responsibilities. Classification affects the measured distance, and these historical findings do not describe France today.

The French study of domestic and parental time extends the question to relations between paid and unpaid work. Japan’s 2021 time-use report offers another institutional setting: when available time is already unequal, occupational choices cannot be explained through individual interests alone. The sources connect questions; they do not form a directly comparable national ranking.