In recent years I have become uneasy with the position that we should eliminate all racial and ethnic preferences in college, university, and professional school admissions. This was a position that I held until the last few years. In 1975, 1 published Affirmative Discrimination, one of the first books, I believe, to deal with the shift in affirmative action from "soft" to "hard." In its early phases, affirmative action called for reaching out, advertising, recruiting, and preparing members of minorities for jobs, promotions, admission to selective colleges, and the like. A new phase, what I would call "hard" affirmative action, dates from the end of the 1960s and the beginning of the 1970s, and in this phase numerical targets were set by federal agencies to monitor progress in reaching affirmative action goals. I argued against this, as violating the spirit of the Civil Rights Act of 1964 -the specific assurances of what that law meant given by its Congressional sponsors, and its very specific language prohibiting quotas. Further, I argued, such race-conscious measures were against the basic spirit of what America was intended to be, as defined by the Declaration of Independence, the Constitution, and the key post- Civil War amendments to the Constitution.

In that book, I discussed "affirmative action" in its new numerical phase as it affected employment, efforts to root out segregation in public education, and efforts to overcome through numerical targets segregation in housing. I did not discuss specifically two other important areas in which affirmative action operates: (1) granting of public contracts, and (2) college admissions, the subject of our panel today. But from the positions taken in that book it was clear that I also opposed racial preference in those areas.

In this paper, I discuss the considerations that have led me to change my mind concerning racial preferences in college admissions.

First, a few words about language. In the early days of numerically targeted affirmative action, the defenders of racial preference called its goals "targets," the opponents called them "quotas." A "target," if there is power to enforce it or to punish the failure to achieve it, rapidly becomes a "quota." It is clear why the defenders of affirmative action resisted the use of that term. It was specifically prohibited in the section of the Civil Rights law proscribing discrimination in employment and promotion, and it reminded people of the infamous practice, which was still widespread only a few years before the passage of the Civil Rights Act, of limiting Jewish admissions to undergraduate colleges and medical schools by establishing a ceiling, a "quota." Of course the "quota" for blacks in selective colleges was a very different thing: it set a lower limit to be reached, not a higher limit not to be exceeded, which was what Jews faced. But both resulted in differential treatment of individual applicants on the basis of their group affiliation, religious, ethnic, or racial.

A new war of words exists today, as exemplified in the exchange between President Bill Clinton and Abigail Thernstroin on 3 December 1997 in the televised "conversation" on race. In discussing the policies that have flowered under affirmative action, are we to talk about "affirmative action" or "racial preference"? The defenders of affirmative action prefer the former, the opponents the latter, for understandable reasons. Affirmative action includes both such measures as reaching out, advertising, selectively recruiting and training, as well as the harder form of affirmative action involved in setting a numerical target or goal. Who can be against the kind of affirmative action called for in the original executive orders of Presidents Kennedy and Johnson, which did not specify "targets" and "goals"? So the defenders of affirmative action prefer the broader term, which includes practices few object to, as well as those that many object to, hoping the good repute of the former will lead to general public acceptance of the whole package of affirmative action. Naturally the opponents want to focus on the feature that is most objectionable, direct racial preference, preferring one to another simply because of race. And as we know, the majority of people will vote against "racial preference" (as in the vote on California's Proposition 209), but not against "affirmative action" (as we saw in the vote on affirmative action in Houston).

But how we name the policies we oppose does not help us determine just what in these policies we oppose, or what we plan to do, if anything, to replace the policies we oppose. At present, the debate, and the legal cases moving through the courts, center on situations, quite common, in which students with academic records and test scores well below the minimum acceptable to an academic institution are admitted specifically because of their race, a procedure challenged in legal suits by applicants with better records who were not admitted. But, as we know, there are other practices driven by the policy of increasing the number of black and some other minority students enrolled in colleges and universities and professional schools. For example, some colleges send minority students on recruiting trips to their old high schools or to other high schools with high percentages of minority students, and will pay for such trips and arrange for students to be excused from classes. There once existed many programs-some still exist-in which minority students were brought to college campuses for a summer of intensive work to overcome academic weaknesses. An example of such an effort is the program at West Point designed to increase the number of minority students who can qualify for West Point and ultimately become officers in the United States Army. Hundreds of institutions acquire lists of minority students who have achieved a certain level in their SAT tests from the Educational Testing Service in order to target them for recruiting. All these and other programs are specifically motivated by the desire to increase the number of minority students enrolled in the institutions. Are we to oppose such programs too? They definitely demonstrate a "racial preference."

So we can see that even the apparent clarity of the term "racial preference" is not clear enough to guide us. What kind of racial preference do we wish to ban? Do we really seek to outlaw even these racially targeted practices I have described?

In this paper, I concentrate on only one issue in the entire world of affirmative action, the issue of admission to selective colleges, universities, and professional schools, and I limit the discussion to only one group, blacks or African Americans. It is the potential impact of the elimination of racial preference for blacks in selective schools that raises the most difficult and sobering questions, because universal access to these institutions is, in one sense, emblematic of our quest for an inclusive society. Some act of bureaucratic mindlessness thirty years ago included Asians and Hispanics (of all kinds) in the groups that were to be benefited by affirmative action programs. But clearly the primary intention of affirmative action was to help blacks. Hard affirmative action started in the building trades because they excluded blacks. There was no important dispute in those professions over the denial of entry to Asian Americans, or any lengthy history, I believe, of discrimination against Hispanics. Blacks were the group specifically excluded from the building trades unions, though there was undoubtedly discrimination against others, too. It has created great confusion in discussing this issue that Asians-for a while-and Hispanics-until today-were included as beneficiaries of affirmative action. (American Indians or Native Americans were also included, from the beginning, but are really not at issue. They number less than I percent of the population. Because of their distinctive history they have long been beneficiaries, on the basis of their race, of certain benefits-e.g., the right to fish for salmon in certain rivers, to open gambling casinos on reservation land. Even if we had a strict quota requiring I percent of all admits to be Native Americans, it wouldn't agitate us much.)

Nor will I talk about women, another beneficiary class. Discrimination against women in admissions these days hardly exists, and disparities in admissions between men and women are much more likely to be based on women's tastes than on any hostility to women on the part of admissions officers or administrators of institutions of higher education.

The key fact that gives me pause when I consider the abolition of racial preference for blacks in selective institutions of higher education is the following: If there were no preference on the basis of race, the percentage of blacks in selective undergraduate institutions and professional schools would drop from the present figure of 6 or 7 percent to I or 2 percent. This would be the case if test scores alone were used to admit students. We have known this for a long time. Robert Klitgaard first set out the facts in full detail in his Choosing Elites in 1985, and Stephan and Abigail Thernstrom also address the matter in their current book, America in Black and White, using data from the Journal of Blacks in Higher Education and other sources.
The impact of such a change for Hispanics would be considerably less.

The key group to consider then in thinking about the effect of jettisoning racial preference is blacks or African Americans. While we all know that is the case, let me spell out the specific reasons why.

First, this is the group for whom this program was initiated. Blacks were the insurgents and carriers of the Civil Rights revolution of the 1960s and 1970s. Second, they play a key role in American history, in Constitutional law, in social life, and in American consciousness, one that far transcends in its importance the role played by Hispanic Americans or Asians. The Civil War was fought to free them-whatever other explanations were given at the time, or since. The major post-Civil War constitutional amendments were passed to secure their status as free and equal to white Americans, and these amendments have become the most significant part of the foundation of current American constitutional law. We commonly test the success of American society and the achievement of our hopes for it by considering the progress of American blacks. What happens to them is far more important when we judge the quality of our society than what happens to Hispanics and Asian Americans. The latter two groups are today mostly immigrants and their children. American blacks, in contrast, while joined by many immigrants in recent years, are no newcomers to American society. They started arriving only shortly after the first colonists, and were one-fifth of our population in 1790. They are still one-eighth of our population. When we consider the role of immigrants and their children, we expect they will initially have a hard time, and will in two or three generations become indistinguishable from other Americans. Immigrants will have to deal with the burdens of learning a foreign language, and overcoming the barriers posed by a distinctive culture in assimilating to American society. The fact that immigrants do poorly at first is no judgment of American society. Its generosity properly extends to admitting them, and fairness requires that they not face discrimination on grounds of race, religion, or national origin. Otherwise, they are on their own. And in fact, any significant distinction between the descendants of immigrants and white Americans long~settled in the United States fades in two or three generations. And a third reason why we concentrate on blacks: Their academic deficiency is the greatest when we compare them with the other minorities grouped together under affirmative action, and thus the consequences of the abolition of racial preference would be greatest for them.

To return to the key fact: The elimination of racial preference means the reduction of blacks, today 12 percent of the American population, 6 or 7 percent of the students in selective colleges and professional schools, to I or 2 percent of the students there.

Well, what do we think of that? One possible reaction is to shrug our shoulders and say, that's the way it will have to be until blacks improve their academic performance. I think most people would say that such a drastic change would dash our hopes of integrating blacks into American society, the way other immigrant groups of the past have been integrated.

Is this an alarmist projection? Will the elimination of the specific use of race in making acceptance decisions lead to such a result? Four points have been made in arguing that this expectation is too pessimistic.

First, there are alternatives to racial preference in keeping up the present number of blacks in selective institutions. We can use income as a basis for preference. Or we can use hardships overcome. Or we can use living in certain low-income neighborhoods, or attending certain high schools. There are problems here, and not the least is that these approaches, however we define the basis of disadvantage, would sweep into the preference pool large groups of poor whites, Hispanics, even Asians, and so these policies of replacing direct racial preference with economic substitutes will not do much to keep up the number of blacks enrolled. This was the conclusion of a study conducted by the University of California, Berkeley, before it was required to eliminate racial preference by the Board of Regents. Furthermore, insofar as these alternatives are transparent efforts to keep up the number of blacks, they might be outlawed, too, if the courts continue to strike down racial preferences.

A second point made against this grim projection: Why do we concentrate on only the most selective institutions of higher education? The figures I have cited, from the journal Blacks in Higher Education, as reported in the Thernstroms' America in Black and White, do so; but we know that the number of such institutions is small and that almost any high school graduate can find some college that will admit him or her. Many community colleges, for example, enroll large numbers of blacks; their student bodies may be more than half black, or even overwhelmingly so. We also know that success in life is not dependant upon attending the most selective institutions. So what if the most selective institutions and professional schools drop from their present 6 or 7 percent to I or 2 percent? There are plenty of other institutions, many of them of great merit, where the numbers of black students will be maintained or even increased, as those who have failed of admission in the most selective institutions downgrade their ambitions. And, as Thomas Sowell ar