|
The proportion of the male-female wage differential that can be attributed to gender discrimination can be estimated from regressions of men's and women's pay on variables that measure their human capital. The measure of `direct discrimination' is taken to be the difference in earnings once the effects of male-female differences in human capital have been removed. These studies conclude that gender discrimination does appear to exist, but empirical estimates of its extent vary considerably. In an influential study based on the 1975 General Household Survey, Zabalza and Arrufat found that the equal pay legislation of the 1970s had virtually eliminated gender-based pay discrimination in Great Britain. Their findings are, however, in sharp disagreement with those of other authors, especially Joshi and Newell's analysis in Discussion Paper No. 156.In Discussion Paper No. 278, Robert Wright and Programme Director John Ermisch investigate the male-female pay gap using data from the 1980 Women and Employment Survey (WES). Unlike the General Household Survey, the WES contains detailed work histories and wage data for a large sample of women aged 15-59. The availability of `real' work-experience data in the WES allows the authors to provide the first estimates of gender discrimination for a nationally representative sample of married women. Wright and Ermisch's analysis of the relationship between pay and human capital reveals that about 17% of the wage differential between married females and males can be attributed to differences in education and work experience. Between 25 and 30% is due to the `depreciation effect' <196> the reduction in earning capacity experienced by women as a result of the time they spend out of the labour force. This contrasts sharply with the 70% estimated by Zabalza and Arrufat. Wright and Ermisch find that about 45 to 57% of the gap is left unexplained. This implies that married women's wages would be 20-25% higher in the absence of discrimination, around four times the estimate of Zabalza and Arrufat, somewhat lower than that obtained by Joshi and Newell, and in line with estimates for other industrialized nations. The authors conclude that while the equal pay legislation introduced in the 1970s has helped reduce gender discrimination in the British labour market, its effect is not as important as previously thought Gender Discrimination in the British Labour Market: A Reassessment Robert E Wright and John F Ermisch Discussion Paper No. 278, October 1988 (HR) |