Bad Arguments Defending Racial Preference

Carl Cohen

Greetings from Michigan and from the University of Michigan. My university has been a leader in defending ethnic preferences in admission and, as you know, the Supreme Court case Grutter v. Bollinger (Lee Bollinger was our president) resulted in a 5–4 decision in which the use of racial categories was held to be, in restricted circumstances, consistent with the Equal Protection Clause of the Fourteenth Amendment.1

The disheartening majority opinion of Justice O’Connor in this case regurgitated the arguments of the university. It also triggered the movement to put the Michigan Civil Rights Initiative (MCRI) on our statewide ballot. But getting a proposition on the ballot is never easy; it takes money and enormous energies. The money came from contributors, in Michigan and around the country; the energy came from some seventeen hundred Michigan volunteers. With the impetus and passion provided by Ward Connerly and others, and the overall management of Jennifer Gratz, we did succeed. The essence of the amendment now incorporated in the Michigan constitution is well known to you. The one critical sentence of its second paragraph reads:

The State shall not discriminate against, or grant preferential treatment to, any individual or group on the basis of race, sex, color, ethnicity or national origin in the operation of public employment, public education, or public contracting.2

Our opponents believed, correctly, that if Proposition 2 were ever to reach the ballot it would very probably pass, and they therefore struggled bitterly, using every conceivable device, to keep it from the ballot. The campaign, during most of 2006, was intense and agitating, but also very satisfying, because it gave us the opportunity to put the case against preferences, widely and forcefully. We did that. An overwhelming flood of signatures put Proposition 2 on the ballot; at the election it passed statewide by a ratio of 58–42.

Some of the opposition was dishonorable. One disreputable argument was particularly nasty, because it addressed not the merits of the issue, but our characters. The claim was that the advocates of Proposition 2 had systematically defrauded citizens when soliciting their signatures by deceiving them, causing them to think that our proposition had the reverse of the effect it really did have. The attack was insidious. Our opponents claimed that in Detroit so many signatures were collected fraudulently that, in spite of the ocean of support, Proposition 2 should not be approved for the ballot. Although their argument was flatly rejected by the courts, the negative publicity and unrelenting accusation of fraud was surely damaging to us. Many among my friends and colleagues were persuaded by it.

That argument—I am obliged to speak plainly—was disgusting. The premises were false, of course—but they were also known to be false. It was a deliberate and dishonest moral smear. I offer three brief responses here:

  1. Every petition signed by a Michigan citizen carried at its top, in block capital letters, the brief explanation that the proposition would “PROHIBIT…THE STATE…AND ALL…STATE ENTITIES…FROM DISCRIMINATING OR GRANTING PRFEFERENTIAL TREATMENT BASED ON RACE, SEX, COLOR, ETHNICITY OR NATIONAL ORIGIN.” Michigan citizens can read. Michiganders, black and white, were insulted and patronized by the supposition that they were systematically fooled by the words of Proposition 2.
  2. It is not at all surprising that many black citizens in Detroit signed our petitions. A majority of black citizens do not support giving preference by race. We know this from careful and repeated polling in Michigan and elsewhere.
  3. The numbers were overwhelming. We needed some 317,000 signatures. The petitions we carried in to the board of elections (I was proud to haul some of those boxes in myself!) carried 508,000 signatures. To vet these signatures the state uses a statistically reliable subset of five hundred names chosen at random. Those valid were just about 90 percent—very high. So, approximately 445,000 valid signatures were indubitably in hand. Some who signed may indeed have been confused about the object of the proposition, but for the argument to have any merit there would had to have been some 128,000 citizens who did not understand the words on the petitions and were deliberately deceived. Absurd. Moreover, our standards in collecting signatures were as high as they could be. The argument against the MCRI on this ground was completely without merit.

Other claims were equally false and were also repeated when known to be false. Perhaps most effective were those contending that Proposition 2, in precluding discrimination by sex, would force the closure of centers for women’s health, for the diagnosis of breast cancer, for the treatment of rape victims. Restrooms would have to be shared by the sexes, it was claimed. All utter nonsense, of course (Proposition 2 was specifically restricted to the spheres of public employment, public contracting, and public education), but one had to be prepared to meet those claims and exhibit their absurdity.

We did, however, also repeatedly encounter honorable arguments in opposition to the proposition. I had occasion to debate many serious opponents in forums around the state. I was among those who represented Proposition 2 in gatherings in Grand Rapids, Battle Creek, Detroit, Kalamazoo, and towns whose names I have forgotten—often at the universities: Michigan State and Wayne State and Michigan, my own university. The colleagues with whom I debated, many of them my friends, were not being dishonest. They sincerely believed that the passage of the MCRI would be a bad thing. Why?

Well, the real reason for opposition, the underlying justification of race preferences for many decent folks, is the belief that preference is only fair. Our country has been so saturated with racial oppression over the generations, and is still so plainly riven by race, that we must not act, they think, in a way that will hinder efforts to give redress for those deep and long continued injuries. That is what really drives the case for preference—the belief that it is an appropriate compensatory device that we are morally obliged to retain.

This drive to give redress helps to explain the continuing tangle over terms in the debate. “Affirmative action” in truth includes a host of programs, agencies, and policies that are not preferential—e.g., eliminating bias in housing and lending; eliminating inadvertent bias in examinations and employment qualifications, and so on. An honorable term, honorably intended from its birth in the 1960s, was captured, kidnapped by the advocates of preference. Affirmative action in its original sense we do not oppose and yet we are cast as its opponents. Why?

Well, we oppose preference. Very commonly now “affirmative action” is taken simply to mean preference. We oppose affirmative action in that sense, of course, and jumbled thinking concludes that we oppose it in every sense. What then was really being voted upon? The term “affirmative action” did not appear in our proposition—but it did appear in the ballot language: one hundred words selected by a partisan Board of Elections to summarize the force of the proposition on the ballot itself. The “banning” of “affirmative action” was put before each voter’s eyes. I was outraged.

Ward Connerly and I don’t fully agree about the impact of this semantic issue; he holds that the term “affirmative action” has itself acquired so many negative vibrations that its introduction was not necessarily damaging to our campaign. But I believe that if the ballot language had said only what Proposition 2 plainly does say—that it would ban ethnic discrimination and preferences by the state—our winning percentage would have been greater than 58, much greater.

In fact, that underlying compensatory justification of preference, although appealing to many, is entirely unsound. But in the very short times allotted by electoral debates, this is not easy to explain. University admissions preferences by color or nationality do not give redress for damage done in the past. Those who were really damaged get nothing; those who receive the preferences generally have not been damaged, else they would be in no position even to apply to the University of Michigan or its law school. And those not preferred, who pay for the alleged redress by being displaced, bear no responsibility whatever for the original damage. As an instrument of moral response, race preference is terribly unfair, crude, and a betrayal of our national ideals.

Community relief for damage done may be in order, but if it is, it is in order without regard to the color of skin or the country of origin. This was, and is, one of our central themes. Proposition 2, now part of our Michigan constitution, does not preclude conscientious social action, it precludes racial favoritism in social action. For reasons of precisely this kind the courts have repeatedly rejected this compensatory defense. It is a rotten argument—but it often comes from well-meaning hearts and is put forward honestly.

When the Supreme Court in Grutter found race preference possibly justifiable, they did not do so on compensatory grounds. Indeed, the defense of preference as remedy had been explicitly renounced by the University of Michigan—and yet the remedial defense often simmered below the surface in our debates over the MCRI; not forthrightly presented, it was therefore difficult to combat.

On the surface, the arguments we met consisted mainly of claims about how good for us preferences are. Honest opponents persuaded themselves that preferences yield real benefits. This moves the argument into an empirical realm, replete with frail predictions of consequences, where citations of studies (scholarly at least in form) become weapons, and the opinions of prominent persons (even army generals) become influential.

Here I address briefly only two aspects of the battle over the effects of preferences: their consequences for minorities and their consequences for the universities.

Preferences for a racial group in a social setting are very bad for that group. We hammered on that plain truth. Even the individuals within the group who are preferred may suffer—as Richard Sander shows with a penetrating analysis of the data about what has happened to minority applicants preferred by law schools.3 But, even if one grants that some individual members get “a leg up” from preference, the group as a whole is deeply hurt. Black and Hispanic students admitted preferentially perform less well. This is a statistically inescapable result. It is true not because of their color (that is a canard, of course)—but because of the corrupt process by which they are admitted. Those preferred persons are, in general, sure to prove inferior to students whose admissions credentials are not diluted by nonintellectual considerations. This disparity in performance is marked by, and associated prominently with, skin color. It cannot be hidden. The preference thus serves to forge links between color and inferior performance, and to reinforce that canard.

From the perspective of the universities and their students we often confronted the claim that preferences are not simply good but absolutely essential for quality education. “Diversity” became the ubiquitous buzzword. Now, intellectual diversity—real variety of opinion and attitude—is surely not a bad thing. But racial preferences as actually given do almost nothing to advance genuine diversity. The views and attitudes of those preferentially admitted (at Michigan, for example) are fully and frequently represented in the views and attitudes of others already among us. Having more people of different colors (as Justice Thomas likes to say) is more of an aesthetic than an intellectual concern. But it is an aesthetic concern with great political weight; university administrators who do not satisfy their constituencies with good racial numbers will not be promoted and may lose their jobs.

The numbers will be good enough only if they approach ethnic proportionality. A slow approach will require an apology and a promise to “do better.” During debates we heard the most extravagant claims: that diversity (meaning ethnic diversity) is the very heart of a liberal education, that diversity is the source of all excellence, that we simply could not do our work as educators without it. Ridiculous, of course; there are contexts in which diversity is simply irrelevant. But the claims of its centrality are difficult to put down, and honesty requires the recognition that diversity is, on the whole, a merit, even if not an overpowering one.

The administration of the University of Michigan was so determined to hang on to preferences that they urged social scientists in our university to write up, in appropriate scholarly garb, any studies that might seem to support the claims for its educative benefits. That was done, and Michigan published a book of such stuff. At the time of this book’s publication it was believed (correctly) that winning the Michigan cases (Gratz and Grutter) would call for a formal showing of the compelling need (an element of what is called “strict scrutiny”). And lo, the title of the book was The Compelling Need for Diversity in Higher Education.4 Thin social science it was, tendentious and weakly argued. At the center of the resulting controversy over empirical results was the work of my warm-hearted friend and colleague Patricia Gurin, who claimed to find (in briefest summary) that a racially diverse setting “improves learning.”5

However, explaining the failings of these claims—their assumptions, equivocations, and gaps—is not easy. In oral argument in a time-limited forum it is nearly impossible. This so-called “Gurin study,” with her claims about the great benefits of diversity, became the principle weapon of the opponents of Proposition 2; it was wielded ad nauseum.

In fact, those claims are without merit. The statisticians Robert Lerner and Althea Nagai examined Gurin’s work very closely and simply demolished it.6 Racial diversity cannot be shown to have a beneficial impact on learning. It can, however, be shown to have a markedly adverse effect on the learning environment in some contexts. Preferences are claimed to “promote cross-racial understanding,” and that is supported (in the University of Michigan materials) by reporting students’ answers to shamefully loaded questions like: “Do you feel that diversity enhances or detracts from how you and others think about problems and solutions in classes?” Social scientists of good repute were shocked by such manipulation. So a serious study, respecting scientific standards, was undertaken by Stanley Rothman, Seymour Martin Lipset, and Neil Nevitte. Without asking about diversity or its merits, they got undistorted straight reports from thousands of students and faculty at 140 institutions about the current state of cross-racial understanding on their campuses. They then correlated these reports with an independent empirical measure of enrollment diversity on those same campuses. Does diversity enhance students’ experience? The outcome was precisely the opposite of what Gurin and company had claimed. Rothman et al., found that “as the proportion of black students enrolled at the institution rose, student satisfaction with their university experience dropped, as did assessment of the quality of their education, and the work efforts of their peers.”7 And the National Association of Scholars released a report by Tom Wood and Malcolm Sherman, Why Justice Powell’s Diversity Rationale for Racial Preferences in Higher Education Must Be Rejected—a meticulous and devastating review of what is known about the consequences of preferences.8

The claims of great educational benefit flowing from preferences are a tissue of exaggeration and supposition and falsification. But I think we did not have great success in showing this. Our counterarguments are cumbersome to present. And no matter what we said about reliable and unreliable empirical studies, the Gurin claims were repeated endlessly and fervidly. We did not suffer unduly from them in the Michigan referendum—but in the U.S. Supreme Court our opponents actually won that argument. The Gurin claims and the brief of the University of Michigan were swallowed whole, on toast, by Justice O’Connor and the majority in Grutter. Strict scrutiny was the supposed constitutional standard of the courts where racial categories are employed. The scrutiny given by the Supreme Court to the university’s claims about the benefits of race preference was not only not strict, it was flagrantly superficial. That was, and remains, an intellectual scandal.

That is where we are. The Civil Rights Initiative is coming again to the fore in Colorado, Arizona, and Nebraska. In the electoral campaigns of 2008 we will encounter, repeatedly, false empirical claims about consequences. We must learn to show that the consequences of race preference are dreadful, that it is in fact hurtful to all, and most hurtful to the members of the minorities preferred. But that is not, in my judgment, the critical issue upon which those elections will depend. The issue will be decided at the polls on the basis of what the electorate thinks is fair. It is our job to show, emphatically, that race preference is not fair, is not a morally appropriate response to past racial oppression. We can and we will argue effectively that race preference is in fact morally wrong, and a betrayal of the highest ideals of a nation “dedicated to the proposition,” as Lincoln put it, that all men and women are equal, and that on the basis of race, and sex, and national origin the state may give preference to none.

1 Grutter v. Bollinger, 539 U.S. 306 (2003).

2 Constitution of Michigan, Article 1, Sec. 26, par (2).

3 Richard H. Sander, “A Systemic Analysis of Affirmative Action in American Law Schools,” Stanford Law Review 57, no. 2 (November 2004).

4 University of Michigan, The Compelling Need for Diversity in Higher Education (January 1999); see

5 Patricia Gurin, “The Compelling Need for Diversity in Education,” expert report prepared for Gratz and Hamacher v. Bollinger, Duderstadt, the University of Michigan, and the University of Michigan College of LS&A, U.S. District Court, Eastern District of Michigan, Civil Action No. 97–75231; and Grutter v. Bollinger, Lehman, Shields, the University of Michigan and the University of Michigan Law School, U.S. District Court, Eastern District of Michigan, Civil Action No. 97–75928 (January 1999); available at

6 Robert Lerner and Althea K. Nagai, A Critique of the Expert Report of Patricia Gurin in Gratz v. Bollinger (Washington, DC: Center for Equal Opportunity, 2001).

7 Stanley Rothman, Seymour Lipset, and Neil Nevitte, “Does Enrollment Diversity Improve University Education?” International Journal of Public Opinion Research 15, no. 1 (2003): 15.

8 Tom Wood and Malcolm Sherman, Race and Higher Education: Why Justice Powell’s Diversity Rationale for Racial Preferences in Higher Education Must Be Rejected, report, National Association of Scholars, released May 1, 2001; available at

  • Share
Most Commented

April 24, 2024


Heterodox Thinking on Evolution and Radical Enlightenment

Between the Modern Synthesis—which says that evolution is driven by accidental genetic changes—and its heterodox challenges—which argue for various forms of agency and non-......

April 24, 2024


Evolution Is Neither Random Accidents nor Divine Intervention: Biological Action Changes Genomes

Biologist James A. Shapiro believes the discovery of different biological means by which organisms can alter their genomes, along with the process of interspecific hybridization, demands a r......

Most Read

May 30, 2018


The Case for Colonialism

From the summer issue of Academic Questions, we reprint the controversial article, "The Case for Colonialism." ...

July 2, 2020


In Humans, Sex is Binary and Immutable

The idea that there are more than two sexes in human beings is a rejection of everything biological science has taught us. Unbelievably, this idea is coming directly from within the highest......

October 18, 2023


Did American Police Originate from Slave Patrols?

The claim that American policing “traces back” to, “started out” as, or “evolved directly from,” southern slave patrols, is false....