The issue of pay differentials between male and female workers has recently become the focus of considerable theoretical interest and empirical research. For the most part, empirical studies in this area have attempted to estimate the extent of wage discrimination by sex in the labor market. Wage discrimination against women may be said to exist when the ratio of male to female wages is greater than the ratio that would result in the absence of discrimination (Becker, 1957, p. 9).
In empirical work, the general practice has been to estimate the proportion of the male-female pay differential that is attributable to differences in those characteristics of male and female workers that are related to productivity. The residual, or unexplained portion of the differential, is then ascribed to discrimination. A recent review of seven studies of male-female pay differences found that in six cases the unexplained earnings gap was estimated at between 29 percent and 43 percent of male earnings (Sawhill 1973, pp. 387, 391). The exception, a study by Sanborn, found a differential of 12 percent after standardizing for detailed occupational classification (1964, p. 535).
The existence of large pay differentials between male and female workers that cannot be attributed to individual differences in productivity-related characteristics, and the large reduction in the size of the unexplained differential when detailed occupational classifications are taken into account have led a number of scholars to focus on the prevalence of sex segregation in the labor market. The existence of sex segregation has generally been deduced from aggregate census data which indicate that a large number of detailed occupational categories tend to be either predominantly male or predominantly female, and that such segregated occupations involve relatively high proportions of the male and female labor forces respectively. For example, in 1970, 73 percent of female workers (as contrasted with 10 percent of male workers) were in detailed census classifications in which women constituted 50 percent or more of the incumbents; on the other hand, 80 percent of male workers (as contrasted with 16 percent of female workers) were in occupations whose composition was less than 30 percent female. A pay differential between men and women with similar productivity-related characteristics is commonly seen to result from this labormarket segmentation, and the consequent "overcrowding" of the female sector.
The purpose of this book is to shed further light on the causes and consequences of sex segregation in the labor market by going to the level of establishment to examine the relationship of differences in the employment . . .