In 1980, 44% of employed women in Britain worked part-time, at an average hourly wage of £1.60, compared with £1.90 for women in full-time jobs. Does part-time employment contribute independently to the pay gap between men and women, or do women in full-time employment tend to have larger endowments of human capital such as education and work experience which attract higher rewards? If part-time jobs are found to reward human capital at a lower rate than full-time jobs, then part-time employment does make a separate contribution to the gender pay gap.
In Discussion Paper No. 234, Robert Wright and Research Fellow John Ermisch explore this question using data on 4,075 women collected in the 1980 Women and Employment Survey (WES). Ermisch and Wright estimate an equation in which hourly earnings depend on a range of `human capital' variables, capturing aspects of a woman's educational attainment and work experience. The number of years of full- and part-time employment are reported in the Survey and are included separately in the model, in order to test whether full- and part-time work experience are rewarded at different rates.
Like most cross-section surveys, the WES only collected earnings data for women who were employed at the time of the Survey. This creates a potential difficulty in the estimation of the model: the women in the observed sample, i.e. those employed at the time of the Survey, may not have been a representative sample of all women. If a woman has significant `career' motivation or `ability' (which are not attributes that can be observed or measured) she would tend to command a higher wage and may as a result decide to postpone childbearing and to participate more in paid employment.
In order to assess the importance of this `selection bias', Ermisch and Wright model explicitly the decision to work full-time, part-time, or not at all. The choice is affected not only by the human capital variables used in the wage equation, but also by demographic, childbearing and marital status variables. The analysis reveals no selection bias for the part-time wage equation but a significant bias in the sample of women in full-time employment. Part of the observed differential between the hourly earnings of full-timers and part-timers therefore arises because some women who have taken up full-time employment have done so because they possess unobservable characteristics which allow them to command higher pay.
Ermisch and Wright also conclude, however, that there is a substantial component of the hourly wage differential between full- and part-time employment which is not explained by differences in human capital endowments. Women full-timers gain more from an additional year of full- or part-time work experience than do part-timers, while a full-timer's wage is enhanced more by full- than by part-time experience. A `typical' woman would earn about 8.5% less in a part-time job than in a full-time job. The larger proportion of women than men in part-time jobs does indeed contribute to women's lower hourly wages separately from gender differences in human capital attributes.

Women's Wages in Full- and Part-Time Jobs in Great Britain
John Ermisch and Robert E Wright

Discussion Paper No. 234, April 1988 (HR)