Job Quality, Labor Market Segmentation, and Earning Inequality : Effects of Economic Restructuring in the 1980s by Race and Gender

The authors examine the effects of employment restructuring in the 1980s on white, black, and Hispanic men and women within a labor market segmentation framework. Cluster analysis is used to determine whether jobs can be grouped into a small number of relatively homogeneous clusters on the basis of differences in job quality. With data centered on 1979, 621 occupation/industry cells covering 94% of the workforce are analyzed with 17 measures of job quality, ranging from earnings and benefits to skill requirements and working conditions. The paper finds strong support for dual and tripartite schemes that closely resemble those described, but never satisfactorily verified, by the segmented labor market (SLM) literature of the 1970s: the "primary" (independent and subordinate) and "secondary" segments. But the findings also show that each of these three large segments consists of two distinct and easily interpretable job clusters that are significantly different from one another in race and gender composition. The job structure has become more bifurcated in the 1980s, as "middle-class" jobs (the subordinate primary segment) declined sharply and the workforce was increasingly employed in either the best (independent primary) or the worst (secondary) jobs. White women became much more concentrated at the top, while white men and black and Hispanic women were redistributed to both ends of the job structure. Black and Hispanic men, however, increased their presence only in the two secondary job clusters. Meanwhile, the quality of secondary jobs declined considerably, at least as measured by earnings, benefits, union coverage, and involuntary part-time employment. As these results would suggest, the paper research found that earnings differentials by cluster, controlling for education and experience, increased in the 1980s. The male and female wage gap also increased, as did the portion of these increasing differentials that were accounted for by changes in the distribution of racial groups among clusters