I find this essay troublesome. I have almost no difficulty with the argument when taken one line at a time. But its basic thrust seems misguided. Sturm and Guinier lay out a new basis for an affirmative action public policy in employment and education by setting up “merit” as an independent policy objective. This seems to point toward a set of government policies that would likely be even more cumbersome and invasive than the policies already in place (although they never discuss specific policy). Moreover, recasting equal opportunity in their terms would obscure the original justification for affirmative action and make the existing policy almost incomprehensible to a generation of Americans who were born after the policies were in place and are increasingly the sons and daughters of immigrants whose families were not even in this country at that time.

Historically, equal employment opportunity policy arose as part of an attempt to dismantle a social system that systematically sought to stigmatize certain social groups and confine them to subordinate and demeaning social roles. It began as an effort to dismantle the system of white supremacy and discrimination against blacks, and was then extended to the system of male supremacy and discrimination against women. Ultimately, it has been used–or discussed–in relation to other ethnic and racial minorities, as well as the physically handicapped and gay men and lesbians. Whether or not white supremacy or male dominance remains a sufficient threat to justify the continuation of the policy, and whether or not the other systems of dominance and subordination are sufficiently invidious to warrant its extension, are debatable. But that, it seems to me, is the debate we need to have. All of these systems of dominance and subordination are of course non-meritocratic, but in exactly the same sense that genocide is murder.

I also have trouble with the way the authors lay out the issue of meritocracy. It is not clear what they mean by merit. It appears to mean “economic performance,” or relative economic performance. They are careful to distinguish between test scores and economic performance–indeed that is a major point of their article. But they seem to imply that test scores are the primary criteria for judging economic performance on the job, or for predicting it. I know of no precise figures on how employment is allocated, but test scores are certainly not the dominant allocative mechanism; in high-level jobs they are rarely used at all. The authors also distinguish between actual economic performance and the performance that would be possible if the job structure and technology were to be reoriented to take greater advantage of “diversity,” but here too it is unclear operationally what this might mean. I believe, in fact, that the economy is being reorganized and reoriented in just this way–but not toward the kind of diversity that is central to the social problems with which the authors are concerned.

I do not see reward on the basis of relative economic performance as an independent moral principle. Much, if not most, of the difference in economic performance among individuals is due to traits that have nothing to do with the basic “worth” of people, such as intelligence, physical prowess, beauty, or social characteristics. Parents pass these traits on directly, or through the education that they enable their children to obtain.

Reward on the basis of relative economic performance is, however, arguably instrumental in the pursuit of principles that are moral. The argument that it is instrumental in this way is that it creates a more efficient economic system, one that maximizes national income and, hence, material well being. That income, once it has been generated, can then be redistributed in accordance with whatever set of principles do have moral validity.

The optimal institutional structure for achieving such a system is a competitive market economy. In such an economy, the agents have a strong incentive to allocate jobs on the basis of relative economic performance because they are rewarded for doing so by higher returns and, in the extreme, punished for failing to do so by bankruptcy. Those agents, moreover, operate at a level where they have the maximum information required to distinguish among individual workers. If they nonetheless fail to allocate jobs on the basis of merit, it may be because the differences in merit among people are small relative to the costs of making those distinctions among them, or because the distinctions simply cannot be made at all. In any case, if employers cannot make this distinction when they are close to the job and under strong market pressure to do so, I doubt very much that an external agent would be able to do so. The kind of surveillance involved in trying to do so, is, moreover, very costly and invasive, as we have learned in the pursuit of race and sex discrimination. Exactly what the relationship is (or was) between these social systems of domination and subordination and the competitive marketplace has never been precisely clear. What is clear is that the systems, and the mechanisms and belief structures through which they were promulgated and maintained, completely overwhelmed whatever competitive pressure might have countervailed them. There is no reason to believe that they would have disappeared without the kinds of equal employment policies we have pursued in the last three decades. It is not clear that these polices, however costly and invasive, have been sufficient to eliminate the problem. But the compromises the market made on race and sex–compromises that virtually every institution in our society also made–do not invalidate presumptions in favor of the market in general.

Finally, I think there are special dangers in linking the case for equal employment opportunity to the efficiency of diversity. The success of the American economy in the 1990s is arguably due to the increasing diversity of its workforce. Certainly, taken company by company, the labor force in the high-tech industries, which have fueled the expansion and restored the country to a preeminent position in global markets, is probably the most ethnically diverse in the world. The diversity has probably contributed to the success of these companies in several ways: it has made it possible for them to draw upon a high-quality, extremely flexible labor force; it has contributed to the creative tensions that foster the development of new and innovative products; and it has made the companies more attuned to the foreign markets in which these products are sold. Many of these factors can be measured, and eventually it will become possible to test the contribution of diversity to economic performance of companies in a relatively precise way. There is already a movement in this direction within the accounting profession as part of the effort to better reflect the role of high-level human resources in the value of particular companies in the new economy.

The diversity of these high-tech companies is the direct product of the openness of our system of higher education to foreign students and faculty and to the changes in our immigration laws in the 1960s. The openness of education and immigration were undoubtedly facilitated by the ideology of tolerance and diversity, which grew out of the country’s response to the civil rights movement. But the diversity in these companies is concentrated among certain groups (such as Asian Americans and recent immigrants) at the expense of other racial minorities (such as native-born blacks), and probably of native-born women as well. The latter groups are not well represented in the diverse labor force of the high-tech industries, and without the new immigration there would be a great deal more economic pressure to facilitate their upward mobility. Thus, while the ideology of diversity may ultimately contribute to economic success, and that ideology is bound up with the forces that have promoted equal employment opportunity, the diversity that has contributed to success at the level of the firm does not have much to do with the groups that have been the target of equal opportunity policies.