August 5, 2002
"Diversity Dodge"--A Commentary by Professor Peter H. Schuck
(This essay originally appeared in the July 2002 issue of The American Lawyer.)
By Peter H. Schuck, Simeon E. Baldwin Professor of Law
After a quarter-century's hiatus, the issue of the constitutionality of affirmative action in higher education will finally return to the Supreme Court. In a 5-to-4 decision handed down in May, an en banc federal appeals court upheld the University of Michigan Law School's affirmative action program. With sharp division on this issue among and within the U.S. Courts of Appeal for the Fifth, Eleventh, Ninth, and (now) Sixth Circuits, the high court will likely review the new decision, Grutter v. Bollinger.
One hopes that the Court will be clearer this time than in its 1978 decision in Regents of the University of California v. Bakke, which involved race preferences for admission to a California state medical school. Bakke splintered into six opinions; the Court could not agree on a standard of review, criteria for assessing specific programs, or a rationale for affirmative action (antidiscrimination, more black doctors, or educational diversity). Lower courts still disagree over whether Justice Lewis Powell's swing opinion, which justified diversity-based preferences under certain conditions, was dictum, spoke for a majority of the Court, or represented only himself. Moreover, they disagree as to precisely what his Delphic words meant. In later noneducation rulings, the Court has resolved a few of the issues that Bakke left open. Racial preferences are now subjected to strict scrutiny, and they cannot be used to remedy past social discrimination or continuing disadvantage. This brings us to the Michigan case. The opinions in Grutter are remarkably churlish even by affirmative action case standards. In a "procedural appendix," the dissenters charge the majority judges with manipulation; the majority bitterly denounce this charge as "shameful" and harmful to the nation. More importantly, the opinions on both sides are analytically weak at crucial points.
Chief Judge Boyce Martin wrote the majority opinion. (The concurrences challenged the appendix and emphasized the expert testimony about the educational value of diversity.) Judge Martin did not dispute the extremely large preferences enjoyed by ethno-racial minorities at Michigan, seeking to use that fact to buttress his case. Without such preferences, he wrote, minority enrollments (not the same as admissions, a point discussed below) would decline from 14.5 percent of the class to only 4 percent-a "token" level, in Michigan's view. The school's admissions decisions, Martin noted, took into account not only grades and test scores but a variety of "soft" variables, including race/ethnicity targeted by past discrimination, that would bring a different perspective to the school. Further, the school did not use a specific quota to pursue this diversity but merely aimed at enrolling a "critical mass" that enabled minority students to feel comfortable there. Insisting that the court was bound by Powell's Bakke opinion and also by Powell's approval of Harvard's admissions policy, Judge Martin read that policy to allow race/ethnicity as a "plus" along with other traits but not to insulate the minority applicants so favored from having to compete with other applicants on an individual basis. Because Michigan's policy was virtually identical to Harvard's, he concluded, it passed muster.
The most central of these points--the majority's total reliance on the authority of Powell's view in Bakke--is highly debatable, as leading scholars such as Chicago's Cass Sunstein argue that the eight other justices in Bakke explicitly rejected Powell's view. But what really takes the majority off the rails is its brief analysis of the "narrow tailoring" that strict scrutiny demands of race-conscious plans such as Michigan's. Even setting aside the dissent's claim that the "critical mass" of racial/ethnic minorities sought by Michigan functions as a quota because it varies within a fairly narrow range around 14 percent year after year, the majority's reasoning fails.
Reduced to its essentials, the majority's position is that diversity means what Michigan says it means and that any sincere effort by Michigan to achieve the critical mass satisfies the narrow tailoring test. To every suggestion of race-neutral alternatives, the majority's invariable answer is that the alternative would not yield the numbers (critical mass) that Michigan wants. This answer is almost certainly true as a matter of fact. But it wholly begs several crucial questions that the narrow tailoring test--if the test means anything--must put at the center of analysis whenever the supposed rationale for ethno-racial preferences is diversity.
Consider only the most obvious of these questions: Why does Michigan favor the particular groups it does? Michigan says that it seeks diversity of experience and perspective, not a remedy for past discrimination, but its preferences favor only a few racial groups and one language group. If viewpoint diversity is really what Michigan is after, why not look to its strongest proxy, political party? If most of its students are Democrats (as I presume), why doesn't Michigan favor Republican (or third-party) applicants? And what about religion, which also correlates strongly with perspective? More than half of law professors in the many schools surveyed during the 1990s were Jewish or professed no religion; comparable figures in the general population were only 2 percent and 8 percent, respectively. If Michigan's student body is anything like Yale's, it contains relatively few Muslims, evangelical Christians, Baha'i, or Buddhists--almost certainly in percentages lower than the population at large. A priori (which is how Michigan selects the groups it prefers), don't those groups' God-based perspectives have at least as much diversity value as those of middle-class, secular blacks or Hispanics? And if social stigma is the touchstone for favoring certain perspectives and experiences, why doesn't Michigan prefer the ugly, the obese, and the physically or mentally disabled, so long as they meet minimal academic standards?
The answer is that Michigan, like almost all other institutions that use preferences, cares less about perspective and experience, which are complex, than about skin color and surname, which are simple--and, more to the point, easily counted. It assumes the Constitution permits preferences based on perspective diversity but not on ethno-racial diversity, and it pretends that the latter is merely a proxy for the former. Under this line of reasoning, police departments could use racial profiling to fight crime--something that Michigan surely does not intend. (As I explained in an earlier column ["A Case for Profiling," January, 2002], racial profiling might be justified, even essential, under certain circumstances depending on the urgency of a quick decision; the accuracy of the proxy stereotype; the cost of acquiring better, individual-specific information; and so forth. But using skin color and surname as a proxy for perspective diversity--especially when better information on individual students' perspectives is readily available, and the decision-making process is a protracted one--could not possibly meet this test.)
The use of a crude racial proxy for diversity, then, cannot meet the narrow tailoring test, nor is it what Justice Powell had in mind in Bakke-insofar as one can divine his meaning. In invoking the diversity goal, he did not endorse what amounts to little more than color-coding and color-counting in service of color-targeting (Michigan's "critical mass"). Thus, the Grutter majority's slavish reliance on Powell is ironic as well as legally dubious. Ethno-racial diversity on campuses is highly desirable, so long as the means used to achieve it do not contradict other precious social values such as merit, candor, and formal equality. But they do. The majority ignores the unfairness to Asian and white applicants who would have been admitted absent the ethno-racial preferences. Because these nonpreferred groups are so much larger than the preferred ones, the preferences reduce only slightly the statistical chances of admission for each individual member. But in group terms, their aggregate disadvantage is immense. And group, after all, is what preferences are supposed to be about.
Were we interpreting the Fourteenth Amendment afresh, much might be said for allowing Congress (not necessarily the states) to prefer blacks--the paradigmatic stigmatized minority with special claims for constitutional solicitude--if this did not violate the amendment's heightened protection for Asians and other ethno-racial minorities. Michigan's program, however, picks and chooses among protected minorities; some receive preferences, and others (equally or more disadvantaged) do not. In short, even if the Court were to permit preferences based on historical disadvantage, Michigan's program might not qualify. The same would probably be true of other existing programs.
Lost in all the legal argument about law school admissions is the real dilemma that Michigan and other schools face: enrollments. The tragic reality is that only a small number of non-Asian minorities are academically competitive with whites and Asians admitted to elite law schools, and a zero-sum conflict rages among those schools over who will get them. Even for a top school such as Michigan, enrolling them is much harder than admitting them, for many decide to go instead to Harvard, Yale, and Stanford, often with even more generous financial aid packages than Michigan can offer. This problem is even worse, of course, for law schools further down the academic food chain. Affirmative action cannot solve this problem. Only better elementary and secondary education for minorities can.
Peter H. Schuck is Simeon E. Baldwin Professor at Yale Law School, and author of the forthcoming Diversity in America: Keeping Government at a Safe Distance (Harvard U. Press, 2003).