Trends in gender and ethnic occupational segregation in England and Wales: census and longitudinal evidence

Daniel Guinea-Martin, Office for National Statistics

[Project number 20100]

The LS data will be used to draw a cohort born in 1957, 1958 and 1959 in order to compare the distribution of labour market and educational variables with the National Child Development Study (NCDS).

This project will analyse the LS and the NCDS in a complementary way. NCDS work histories will be used to examine the dynamics of employment changes during the decade (this level of detail is not available in the LS). At the same time, the LS will be used to assess the representativeness of the NCDS sample, with specific reference to the occupations held by that cohort. This is possible because the research design of the LS means that it does not suffer from attrition in the way that other longitudinal samples do.

The proposed research builds on previous research using Census and LS data. The latter showed that occupational differences between the sexes (i.e., their segregation) declined in the 1990s at a greater pace than in previous decades in Britain. Census data was used to determine this trend by ethnic group. LS data was used to contextualise the cross-sectional analysis by illustrating ethnic differences in demographic and employment continuity and change.

The LS is uniquely placed to contribute to this study because being a record linked study. As such in the LS there is no response burden and it does not suffer attrition due to non-response. Instead, in cohort studies such as the NCDS non-response is a problem which can affect the representativeness of the sample over time. For this reason the LS input will be invaluable by way of offering a sort of sampling 'benchmark' with which to judge the representativeness of the NCDS sample.