The legal definition of “discrimination”; the declining significance of occupational segregation?

I’d like to focus on two issues that are not addressed in-depth in this otherwise wonderful book.  First, the book overlooks the importance of the substantive legal doctrine that emerged between 1966 and the 1980s.  Title VII says that employers may not “discriminate against any individual with respect to . . . employment, because of such individual’s race, color, religion, sex, or national origin.”  But what exactly does it mean to “discriminate . . . because of”?  Over time, the courts converged on the view that the employer must have been consciously motivated by the relevant characteristic at the time of making the adverse employment decision.  Under this definition, only a very narrow range of behavior gives rise to legal liability.

This kind of consciously motivated discrimination certainly continues to exist, but it is less common than it once was.  Today, if employers allocate opportunities or rewards unfairly, it is more likely to be the result of unconscious or “implicit” bias—either perceptions based on group stereotypes or unduly favorable impressions of members of the decision-maker’s “in-group.”  The law does not prohibit employment decisions that are influenced by unconscious biases.

Also, as blatant discrimination has receded, racial and sex segregation have become increasingly influenced by what we might call “pre-labor market” factors that influence workers’ educational and occupational choices.  That doesn’t mean that white women, black men, and black women freely and rationally sort themselves into less desirable occupations.  Their choices are constrained by culture, family, schools, and social networks.  But the point is that these constraints operate before individuals knock on an employer’s door, and thus are outside the purview of Title VII.

Although this case law was developed by conservative courts, it reflects American culture’s deeply entrenched belief that individuals can and should be responsible for their own actions and fates.  As a result, attempts to push employment discrimination law toward a more expansive understanding of employer motivation, or toward requiring employers to remedy pre-labor market disadvantages, have met with strong cultural resistance.  So far, neither legal scholars nor movement leaders have been able to frame an expansion of prohibited employer behavior than can be widely accepted as legitimate.

Second, I’d like to raise the possibility that while the concept of segregation at work was highly meaningful in the 1960s, it is becoming less so as the nature of work and the way it is organized are changing.  In any context—employment, education, or residential—we care about the segregation of groups into different social locations when those locations (a) have clear boundaries around them; (b) permit little or no mobility among them; (c) are isolated from each other; and (d) are unequal in status and rewards.   If these conditions hold, segregation implies both lack of intergroup contact and durable group inequality.  We tend to think both conditions are undesirable.

How well are these conditions met in the context of occupations within workplaces?  In the early 1960s, pretty well.  At that time, the typical large establishment was a manufacturing plant with a bureaucratic structure.  Jobs consisted of fixed bundles of tasks grouped together on occupational lines, assigned to individuals in a stable way, and linked by hierarchical chains of command, which in turn were distinguished on the basis of function.  Both vertical and horizontal occupational boundaries were usually quite clear.  Occupational groups were physically isolated from each other, even within a single workplace.  Locations in the hierarchy were clearly unequal in status, and mobility to higher-status occupations was sharply restricted.  In that setting, racial segregation between, say, operatives and managers was highly meaningful and entailed both minimal intergroup contact and inequality of rewards.

Now consider contemporary firms that provide expert or professional services—one of the fastest growing employment sectors today.  In such firms, occupational boundaries are much blurrier than in the 1960s factory.  Although employees have areas of responsibility, they are not assigned to fixed bundles of tasks; rather, their tasks change over time as the firm grows and changes the mix of services it offers.   Most employees shift among tasks that could be considered professional, sales, and managerial work.  For example, the same employee might develop a service product, meet with prospective clients, and engage in strategic planning or supervisory tasks.  There is no physical separation: people performing different tasks interact frequently in a team production process.  Finally, insofar as professional, managerial, and sales occupations can be identified, they have equal status as occupations (although there is individual inequality within occupations).

Setting aside the likelihood of measurement error in classifying individuals into occupations, suppose that most of the “professionals” in a firm are women and most of the “sales workers” are men (or vice versa).   If there is neither isolation nor inequality, does it matter?  Maybe so, but why exactly?  Furthermore, managers typically start their careers in non-managerial professional occupations.  The meaning of gender or racial segregation between managers and non-managerial occupations is quite different in this context than in the 1960s manufacturing plant.

To be clear, I am not suggesting that racial and gender inequality at work is about to disappear.  We will still see unequal chances of hiring, retention, and promotion.  First of all, note that these unequal chances do not necessarily imply occupational segregation — the separation of race-gender groups into different occupations.  There could be (and likely will still be) within-occupation inequality.  Second — and this is the main point I am trying to make — even if race-gender groups do tend to move into different occupations, that separation may be less meaningful or important than it was in the past if those different occupations (a) have blurrier boundaries between them; (b) permit greater mobility between occupations; (c) are less physically and socially isolated from one another; and (d) are more equal in status.  As work evolves toward a greater emphasis on knowledge-based services, employment relationships grow more tenuous and contingent, and organizations shift toward flatter hierarchies and network structures, such changes in the occupational structure may or may not be taking place; the outcome is not yet clear.  But we should not simply assume that because occupational segregation was significant in the past, it will always be significant in the future.  Instead, we need to think carefully about what occupational segregation means, what its consequences are, and why those consequences are important.