How Do Occupational Characteristics Contribute to the Explanation of Occupational Sex Segregation? Results from a Dynamic Fixed-Effects Panel Analysis for the German Labour Market

This paper aims to investigate how occupational characteristics can contribute to the explanation of occupational sex segregation by using dynamic fixed-effects panel analysis. Occupational segregation is a distinct feature of the German labour market and has a wide impact on labour market outcomes. Although prior research has shown that occupational segregation leads to social inequality, there is still a research gap concerning the causes of this labour market feature.

This paper focuses on the influence of structural conditions represented by occupational characteristics, such as working conditions or qualification requirements, on the share of women in occupations. So far, only few studies exist which analyse the interplay between occupational characteristics and the share of women in occupations on the occupational level. These studies are, however, based on cross-sectional data and are, hence, limited in interpreting their effects as causal. Building on this, we aim to improve causal statements with the help of panel data analysis.

First, we build a dataset on the occupational level by aggregating individual data of the 1996 to 2010 waves of the German micro census. For this reason, we merge the information of all respondents that work in the same occupation. Relying on causal interpretations of parameters obtained by dynamic fixed effects panel models we then test whether occupational characteristics influence the share of women in occupations. Our key findings show that rising working-hours and fixed-term contracts lead to a decrease in the share of women in occupations. We trace this back to the influence of structural conditions on the probability of female occupational choices. Furthermore, higher mean incomes and the representation of academics in occupations reduce the respective percentage of women. These effects are in line with the assumption of men being preferred employees due to attributions of higher productivity and less labour turnover.