Women with Discontinuous Employment Histories
An innovation in the 1991 Census was to obtain information on the last job held in the previous ten years (if any) for everyone not currently in work. The 1971 and 1981 Censuses had moved in this direction by collecting information on the last job of the unemployed and retired, but the 1991 Census went a lot further in collecting systematic information for a very much larger group of people who had recent work experience but no current job, either as employee or self-employed. In effect, the 1991 Census collected employment data for a ten years' reference period as well as the usual 'last week' reference period ( Hakim, 1995a). The change had the greatest impact on data for women, who move in and out of jobs, and in and out of the labour market over the life cycle far more frequently than men, in the USA as well as in most European countries ( Hakim, 1996c). The conventional 'last week' reference period for employment data has become less and less useful as women's labour force participation has become more discontinuous and dispersed across the whole life cycle instead of being concentrated in early adult life, before marriage and childbirth ( Hakim, 1979: 10-12, 1996a: 132-9). The pure marriage career has almost disappeared in Britain although it remains very common in Greece and some other European countries ( Hakim, 1997: 43). It has now been replaced by the modern marriage career, in which women work as secondary earners, often part-time, after marriage or childbearing ( Hakim, 1996a: 135-8, 1996b). This chapter analyses the SARs data on jobs held in the last ten years and provides an important complement to data on current jobs analysed in Chapters 2 and 3. The new data also provides an opportunity to test theories on the differences between women who work continuously and women with discontinuous work histories.
Hakim's thesis on the polarization of the female workforce and of women of working age ( Hakim, 1996a) suggests that discontinuous women workers will be more similar to women currently working part-time than to women currently working full-time. The thesis also leads us to expect sharp differences between women in continuous employment and women with discontinuous work histories. Women with discontinuous work histories should display a lower investment in human capital and in employment careers than career-oriented women. More specifically, this chapter tests the hypothesis