Documenting Desegregation: Racial and Gender Segregation in Private Sector Employment Since the Civil Rights Act

Notes

339 Notes Introduction 1.  This is not to say that white women, black men, and black women were equivalent in their cultural position in the early 1960s. White women received racial benefits in household and educational resources. As we document in this book, these benefits, as well as the benefit of whiteness, became increasingly important for white women in private-sector employment as the society retreated from institutionalized white male advantages in employment. Chapter 1 1.  It is important to recognize that this is a study of EEO change in the United States. The U.S. state has rarely attempted to change corporate behavior through administrative fiat and typically has preferred symbolic regulation. The strength of mimetic isomorphism in U.S. EEO diffusion may simply reflect the weakness of U.S. federal coercive mechanisms. 2.  We identify whites by subtraction before 1980. See the methodological appendix for more detail. 3.  Of course, some whites and blacks counted in the EEO-1 forms are immigrants , but they are a much smaller proportion, and so we use the designations “African American” and “European American.” 4.  The EEOC reporting forms collect occupational distributions (for example, managers, professionals, technical, sales, clerical, craft, operatives, laborer, and service) and so ignore within-occupation, job-title segregation. We treat this as a form of measurement error and adjust reported segregation measures upward to take into account observed occupational heterogeneity at the workplace level. The technical discussion behind this adjustment can be found in Tomaskovic-Devey et al. (2006) and in the methodological appendix . The index of dissimilarity is computed across occupations within establishments as follows: 340    Notes D P P oex oey oe Noe = − = ∑ 1 2 1 9| | to where Poex and Poey are the proportions of group x and y, respectively, within an occupation in an establishment. 5.  The isolation index is computed across occupations within establishments as follows: x X x t i i i i N      ×               = ∑ 1 9 to where x is the number of the group in occupation i, X is the total number of group x in the workplace, and ti is the number of workers in occupation i. 6.  For example, we calculate craft representation as: Craft Representation (CR) = (((Xcit/Tcit)/∑ (Xijt/∑ Tijt)) − 1) × 100 where Xcit is the number of status group members (for example, white males) in the craft occupational category within an establishment in a given year; Tcit is the total number of individuals in the craft occupational category c within establishment i in a given year t; ∑ (Xijt) is the total number of status group X members in commuting zone j in a given year t; and ∑ (Tijt) is the sum of employment across all establishments in commuting zone j for a specific year t. 7.  Commuting zones are aggregations of counties; not confined to state boundaries , they are calculated based on decennial Census surveys that document the distance that individuals travel to work from where they live (Tolbert and Sizer 1996). Therefore, they describe local labor markets. We impose 1990 commuting zone boundaries on all years of data in these analyses for consistency purposes. The use of stable geography allows us to compare localities over time. Further discussion is in methodological appendix. 8.  We do this by statistically controlling for a fixed effect for detailed three-digit industry in statistical models to estimate time trends net of shifts in industrial structure. Technical examples of this approach can be found in TomaskovicDevey et al. (2006) and Stainback and Tomaskovic-Devey (2009). Formal models are presented in the methodological appendix. 9.  All groups dropped between 1980 and 1985, when reporting requirements by small firms were eased by the EEOC. 10.  In the sharecropping system, tenants had long-term, sometimes intergenerational ties to specific land, which was typically owned by the same large Notes   341 landholding family that had once controlled the slave-based plantation on the same land. Although formally free to leave the land, sharecropping was a system of debt peonage in which the sharecropper would get loans of money or materials from the landowner in order to produce a crop. The landowner, in return, would receive a share of the agricultural production, and it was not unusual that at the end of the season the tenant owed the landlord everything he had earned, or even more—and so indebted, he was not free to leave. Chapter 2 1.  In an earlier study, we showed...