Abstract
Without question, the topic of race- and gender-based workplace inequality has been a mainstay of sociological research for the past several decades. Studies in this broad area have attempted to bring us closer to understanding both the extent and the consequences of employment discrimination for women and racial ethnic minorities. However, as Kaufman notes in Race, Gender, and the Labor Market: Inequalities at Work, analyses often fail to take into account intersections between race and gender that produce distinct differences in labor market opportunities and outcomes. To address this deficiency, Kaufman develops a comprehensive strategy for modeling segregation and earnings determination processes for four race–gender groups: black men, black women, white women, and white men. It is through meticulous estimation, solid theoretical grounding, and thoughtful interpretation that we gain much insight into how race–gender stereotyping and devaluation processes lead to worker-level differences.
Kaufman begins the book by illustrating how race–gender groups come to be defined using examples of two pairs of common jobs. This provides the spring board for outlining his unique approach and for highlighting employment and earnings trends over time by race and gender. At the outset, Kaufman also makes a strong case for why such research is essential and the importance of his two-step procedure to effectively account for worker- and position-level factors.
After a thorough review of supply-side and demand-side theoretical approaches to explaining workplace inequality, Kaufman outlines his integrated perspective founded on six basic principles explicitly linked to the literature on queuing preferences. Of particular interest are the connections made between job sorting processes, product and labor market locations, and differential earnings among race–gender groups that extend “beyond individual differences in credentials, productivity, and trainability” (p. 35). It is through the proposed two-level modeling, based on comprehensive data for workers and positions, that Kaufman charts new territory.
For the technically minded reader, chapter three along with Appendix A offer a wealth of detailed information pertaining to data sources, definitions and measurements, as well as methods of analysis. Although some may question the use of 1990 data sources, Kaufman provides clear justification for the choice pointing out that most, if not all, recent data lack the kind of detailed worker- and position-level information to adequately address the questions posed in this set of studies.
At the heart of the book are analyses examining segregation and then earnings differentials among the race–gender groups examined. For each set of analyses, Kaufman uses a base model that tests whether the effect of each predictor varies with values of the other factors, and then adds a more complex model examining potential moderating effects of employment growth, market power, and profitability. The final set of analyses, presented in chapter six, explore the role of regional context as defined by differences between the southern and northern United States.
Although there is much to digest from Kaufman’s research findings, perhaps some of the most insightful results are those that highlight the moderating effects of external pressures and internal resources on stereotyping and devaluation processes. Also of particular interest are the findings supporting a race–gender labor queuing perspective. As Kaufman notes, positions at all levels have an ordering of race–gender groups that create distinct opportunities and outcomes. Lastly, the earnings analyses makes evident the way in which skills and worker characteristics stereotypically associated with women or minorities are devaluated to the point of having negative or no return to earnings, while those commonly identified with whites or men add value to earnings.
As with any research, the scope of projects, types of analyses, and even the kinds of questions to be addressed are often limited by available data sources. Perhaps, pushing the envelope that has been opened by Kaufman will require large-scale data that allow for fine-grained measures of skill assessments and preferences as well as employers’ perceptions and use of race and gender stereotypes to fill positions. In addition, data that include sufficient representation of other racial ethnic–gender groups is likely to assist in assessing the extent to which labor market patterns outlined here exist more broadly. In all, there is much to learn from Kaufman’s complex modeling strategy and almost endless results that will certainly strengthen theory, refine research, and inform policy directed at improving employment inequality.