Cross-cohort changes in gender pay differences in Britain from 1972 to 2004: accounting for selection into employment using wage imputation

Jenny Neuburger, Diana Kuh, Heather Joshi


This paper examines trends in the labour market position of British women and men from 1972 to 2004, using micro data from three British Birth Cohort Studies, of 1946, 1958 and 1970.  Women’s rates of employment and hourly pay have been lower than men’s over this period, but generally increasing.  Because employment decisions are influenced by the level of pay on offer, changes in women’s relative pay in the working population may not be representative of changes in their labour market position.  We accounted for selection into employment by imputing missing hourly wages for non-employees using observed wages of employees of the same sex and age with similar work and family histories, matched on their propensity score.  At each survey, women’s median hourly pay was lower than men’s.  Although relative pay increased across the cohorts, it decreased with age within each cohort.  Accounting for selection into employment gave a lower estimate of young women’s potential pay relative to men’s in the two earlier cohorts, flattening the within-cohort profile for the earliest cohort.  This evidence supports the view that the improvement in young women’s labour market position since the 1970s has been substantial, and is underestimated in pay trends for the working population.


Wages, Gender gaps, Employment, British Birth Cohorts, Sample selection, Imputation, Propensity score matching

Full Text:



Copyright (c) 2014 Longitudinal and Life Course Studies