Unequal Pay for Women and Men: Evidence from the British Birth Cohort Studies

Citation

Neuman, S. (2001), "Unequal Pay for Women and Men: Evidence from the British Birth Cohort Studies", International Journal of Manpower, Vol. 22 No. 5, pp. 475-486. https://doi.org/10.1108/ijm.2001.22.5.475.3

This book presents the results of an extensive cohort study of wages of British men and women who are members of two birth cohorts of 1946 and 1958. Babies who were born in 1946 and 1958 were interviewed when they reached their thirties.

The focus of the empirical analysis is on trends and determinants of gender wage differentials between British men and women, and among the latter between full‐time and part‐time workers.

While gender wage differentials have been intensively studied and documented for Britain and for many other countries, this study makes a number of significant contributions and has high value added:

• •

  It employs a unique rich data set. The existence of two highly comparable cohort studies provides an opportunity to follow individuals from birth to adulthood. The major advantage of longitudinal data in an analysis of wage differentials is that they provide, alongside adult earnings, information on innate ability, education, work histories, family background, and occupation. Using two comparable surveys (with 7,547 employee respondents in the first and 2,087 in the second), allows for cross‐cohort, as well as intra‐cohort, analysis.

• •

  The authors explore and highlight the relationship between the legislative and the economic systems. Two major acts were passed in the 1970s: The Equal Pay Act and the Employment Protection Act that introduced statutory maternity leave. It appears that being a “maternity leaver” cancelled out the pay benefits of motherhood, and the enforcement of the Equal Pay Act led to a significant drop in the full‐time wage gap, mainly at the bottom of the wage distribution. Cross‐country comparisons show that women’s pay relative to men’s has risen most sharply in countries with stronger equal pay legislation and enforcement (e.g. Scandinavian countries, France, Belgium, Australia, New Zealand). These facts bring us to the conclusion that the legislative body affects the economy, namely, pay trends in the labour market. The relationship is symmetric in the sense that statistical economic analysis might serve the legislative body: econometric regression analyses have been used in US courtrooms as evidence for pay differentials. In Britain too, tribunals have used statistical data.

• •

  Joshi and Paci show that political policy matters too. Interestingly, Margaret Thatcher, a female prime minister, embarked upon a labor market agenda of deregulation which resulted in dismantling of collective bargaining, subcontracting work to agencies and contractors, and, as a result, women were less protected against discriminatory hiring and promotion practice.

• •

  Special emphasis is given to the analysis of trends and determinants of the full‐time/part‐time pay gap among women. Most studies of pay differentials pay little attention to this type of wage differentials, although a significant percentage of women are employed in part‐time jobs. The full‐time/part‐time pay gap among women, in combination with the gender full‐time gap, reflects the overall gender gap between men in full‐time employment and all women. While there was progress in closing the full‐time wage gap, the disadvantages of being a British woman are compounded if she is constrained to take a part‐time job. Less than half of the full‐time/part‐time gap can be accounted for by differences in measured human capital endowments. Firm, job and occupation characteristics also explain part of the differential. However, about half of the difference stems from the constant term reflecting the “part‐time job effect”. Investigating the relationship between family responsibilities and women’s low pay, reveals that mothers tend to be poorly paid because they tend to take part‐time jobs (and not the opposite, i.e. the high concentration of mothers in part‐time jobs causes low average pay of part‐timers). The careful investigation of the performance of female part‐timers is an important component of this study.

The authors present a rich statistical analysis of wage differentials. However, the study suffers from some shortcomings:

1. 1.

   (1) No reference is made to the modifications and refinements of the standard Oaxaca/Blinder wage decompositions. To unravel the components of the wage gap, the authors are using the standard Oaxaca (1983) and Blinder (1973) decomposition, correcting for selectivity bias. Joshi and Paci ignore the many refinements and extensions of this basic decomposition. An empirical study employing such a rich database calls for the use of more rigorous decomposition methods, i.e.:

• •

  decompositions that relate to the role of occupational segregation, which is a major component in the explanation of male‐female pay differences (Brown et al., 1980; Miller, 1987; Neuman and Silber, 1996);

• •

  breakdown of the wage difference which gives different weights to the various characteristics (Neumark, 1988; Oaxaca and Ransom, 1994); and

• •

  statistical analyses that look at the whole wage distributions of the two genders, rather than at the difference between averages.

1. 1.

   (2) There is no decomposition of the selection term (λ). The authors correct for selectivity bias (using a probit or a multinomial participation equation for the estimation of λ). Such a correction results in an addition of another explanatory variable (λ) to the wage equation. It is not clear from the presentation of methodology and results how the difference in the respective lambdas is decomposed. There seem to be several alternative decompositions that include λ (e.g. Neuman and Oaxaca, 1998), each resulting in a different breakdown implications.

Overall, this work adds another important building block to the numerous studies of gender differentials in pay. It helps understand the functioning of the British labour market and leads to important policy implications.