The 35-year-old debate about affirmative action in university admissions has often focused on whether the supposed benefits to black and Hispanic students justify the costs to whites and Asians who lose out, and the resulting racial divisions and resentments.

Lately, some critics of racial preferences (including me) have also speculated that many of the supposed beneficiaries might be better off without preferences. They are so much less qualified academically than their white and Asian classmates, this argument goes, that they end up near the bottom of their classes and drop out in disproportionate numbers.

Now comes Richard Sander, a UCLA Law School professor, with the first in-depth statistical study of how preferences actually affect black law students' academic performance, passage rates for the bar, and job prospects. His stunning and copiously documented conclusion: "Blacks are the victims of law school programs of affirmative action, not the beneficiaries."

One reason is that racial admissions preferences "significantly worsen blacks' individual chances of passing the bar by moving them up to schools at which they will frequently perform badly." Indeed, says Sander, so debilitating is the effect on the academic performance of black law students that ending preferences would increase the number who pass the bar and become lawyers.

To be sure, Sander says, the number of blacks admitted to law school would decline (by about 14 percent) if admissions were colorblind. And almost all of those admitted would end up at less elite law schools, with the percentage of blacks at the very top schools plunging from about 8 percent to 1 or 2 percent. But their grades, graduation rates, and chances of passing the bar would improve so dramatically that the number of African-Americans making it into the legal profession would rise by more than 8 percent. And most of those who make it would end up with better jobs than they get now.

Sander's 113-page study, which the Stanford Law Review plans to publish in a week or two, has already provoked furious debate in legal academia. Critics have been on the attack since Sander circulated a draft for comment in May. The controversy has found its way into several newspapers since a critic leaked the Sander draft a few weeks ago.

Four of the most vocal critics argue in a 30-page draft "empirical critique" that Sander is unrealistic to the point of being "irresponsible" in his optimism about the effects of abolishing law school admissions preferences. In fact, say David Chambers and Richard Lempert of the University of Michigan Law School, Timothy Clydesdale of the College of New Jersey, and William Kidder of the Equal Justice Society, a move to colorblind admissions would lead to declines of 35 to 45 (not 14) percent in the number of African-Americans admitted to law schools and to a decline of 25 to 30 percent in the number who graduate and pass the bar.

Remarkably, these four critics and other defenders of racial preferences have not challenged some of Sander's key factual premises, even though these premises belie the rosy portrayal of affirmative action in the mainstream media and in the June 2003 Supreme Court opinion upholding the University of Michigan Law School's racial preference program (Grutter v. Bollinger):

* Very large racial preferences are prevalent at almost all law schools, not just the most elite ones.

* The preferences used by the top-ranked schools have the "cascade effect" of pulling in many underqualified black students who would do much better at the second-tier or third-tier schools that they now shun. These less competitive schools must in turn go far down in their own applicant pools to fill their racial targets. The result is a "game of musical chairs where blacks are consistently bumped up several seats in the law school hierarchy, producing a large black-white gap at nearly all law schools in the academic credentials of students"—that is, in their undergraduate grades and LSAT scores.

* Once in law school, "around 50 percent of black students [end up] in the bottom tenth of the class; and around two-thirds of black students are in the bottom fifth." Only 8 percent of black students place in the top half of their classes.

* Contrary to the widespread view that preferences give black students opportunities to "catch up" with more-privileged whites, black law students' grades "go down a little from the first to the third year."

* About 19 percent of black law students never graduate, a dropout rate more than double that of whites.

* More than 40 percent of entering black law students never pass the bar and never become lawyers. Blacks are four times as likely as whites to fail their first bar exams and six times as likely to fail on each of five successive attempts.

The 48-year-old Sander, who co-founded a civil-rights group to enforce the fair-housing rights of minorities in Southern California and has a half-black son, notes in his article that "the material harms to whites from affirmative action in law schools are comparatively slight" and that racial admissions preferences may be justified "if the benefits to minorities substantially exceed the costs to minorities."

But the benefits to blacks (on whom Sander focuses) do not exceed the costs to blacks, he has concluded after years of empirical study, with the help of his doctorate in economics and his position as director of UCLA's Empirical Research Group.

In an interview, Sander acknowledges that he is making an educated guess in predicting that ending preferences would increase the number of blacks passing the bar exam, "because we don't know what would happen to the number of applications." But he expresses confidence that at worst, the number of new black lawyers would drop by no more than 10 percent.

The solid pillars of Sander's statistical analysis are that 1) black students do just as well in law school as whites with comparable undergraduate grades and LSAT scores; 2) this fact proves that blacks' academic difficulties stem from racial preferences, not from any racial problems in law schools; 3) black students who pull C's at the most elite law schools could be getting B's at second-tier schools, and so on down the pecking order; 4) the strongest black students would perform much better—and thus would learn much more—at the less prestigious law schools they would attend but for preferences; and 5) most employers would rather hire young black law graduates with decent grades from run-of-the-mill schools than those who barely got by at elite schools.

The Chambers-Lempert-Clydesdale-Kidder group challenges each of these five points, but not very persuasively, in my tentative view. Perhaps their most surprising point is that—contrary to the usual pro-preference argument that grades and test scores undervalue blacks' academic potential—blacks do worse in law school than whites do with similar grades and LSAT scores. The problem, the four suggest vaguely, is "something about the atmosphere of law school."

Sander, who is preparing a detailed response, sees this critique as an effort "to rally the troops with arguments that are technically complex but devoid of any basis in the real data." He concedes that racial preferences have benefits—especially for white students, who are exposed to a more diverse group of classmates. And he stops short of urging complete abolition.

But law schools should at least end what Sander calls the "pervasive public dissembling" by which they have concealed the costs imposed by preferences on their own black students, especially the many who will invest years of their lives and thousands of dollars without ever becoming lawyers. The law schools should start by adding something like the following warning message (my words, not Sander's) to their congratulatory letters to admitted black applicants:

"In choosing among law schools, you should be aware that our racial preferences in admissions have costs as well as benefits. About half of our black students end up in the bottom tenth of their classes. While we have done our utmost to help them succeed, their dropout rates are far higher, and their bar-exam passage rates far lower, than those of our other students.

"If you choose a less selective school, you are likely to be more successful academically. Indeed, according to one study, you may even have a better chance of passing the bar—a point vigorously disputed by other studies. We hope that you will take your chances here and will be glad you did."

Such a warning would no doubt be discordant. But would it be worse than selling would-be black lawyers a bill of goods?

We want to hear what you think about this article. Submit a letter to the editor or write to letters@theatlantic.com.