Beyond occupation : the evolution of gender segregation over the life course

We argue that gender segregation stems from sources beyond occupation, the traditional domain of study: women and men differ not only in their occupational allocation but also in their time involvement in paid work, in their decisions to participate in the labor market at all and in their retirement age. We pool 21 Labour Force Surveys for the United Kingdom to measure and compare these various forms of segregation (occupational, temporal and economic) over the 1993-2013 period (n = 1,815,482). The analysis relies on the Strong Group Decomposability property of the Mutual Information index to add up all forms of segregation and to identify the evolution of segregation over the life course net of cohort and period effects. There are two main findings. First, over the life course, the evolution of gender segregation parallels the inverted U-shaped pattern of the employment rate. When workers are younger, measures of all concepts of segregation are small. Then, gender segregation increases due to a combination of economic and time-related components. After the prime childbearing years, gender segregation remains fairly stable for approximately 15 years, sustained by expanding occupational segregation; finally, in the later years, gender segregation decreases substantially. Second, gender segregation is consistently 20% higher than occupational segregation after the teenage years. However, as much as 44% of gender segregation at age 35 and 52% at age 64 would remain even if occupations were completely desegregated. These ages correspond to two key stages in the life course: career and family building on the one hand and retirement on the other.The authors are grateful for financial support from the Spanish government’s National Program for Research. Daniel Guinea-Martin is supported by grant CSO2011–30179–C02–02; Ricardo Mora by grant ECO2015–65204–P; Javier Ruiz-Castillo by grant ECO2014–55953–P. Additionally, Guinea-Martin acknowledges financial support from the Spanish government and uned under contract RYC–2008–03758 and Mora and Ruiz-Castillo from the Department of Economics of the Universidad Carlos III through grant S2015/HUM–3444 and from the Maria de Maeztu program through grant MDM 2014–0431