95

Blue collar territories are rather unconducive to women’s employment; it all depends on the presence of local industries that are intensive in female labour. For given population density and local share of blue collar women, these territories have much lower female employment rates. This link is analysed in detail: in particular, the hypothesis of a spatial mismatch unfavourable to female employment is examined. The analysis, successively in cross-section and panel, is conducted on census data for the years 2006-2012. In cross-section, once adjusted for confounders and spatial interdependencies, the relation of interest (between the share of blue collar and women’s employment) is reduced but remains very significant. The results of the panel analysis suggest that two-thirds of this is a long-term link compatible with a spatial mismatch scenario.