Women's work in UK official statistics and the 1980 reclassification of occupations

Analyses using either time series or longitudinal data can be complicated by having to disentangle real changes from the artefactual effects imposed by the reclassification of key variables. In this paper the Warwick occupational conversion program is used to create adjustment tables which correct for the artefactual effect of the 1980 reclassification, along with coder errors, on 1971-1981 patterns of occupational sex segregation in longitudinal data from the Office for National Statistics's Longitudinal Study. The crowding of large numbers of women into a few occupational groups, which is consistent through the reclassification, arises in part because classification schemes fail to recognize the diversity in their work. This reflects past practice and the undervaluation of women's skills.