If anything is certain about The Shape of the River—William G. Bowen's and Derek Bok's massive defense of race preferences in university admissions—it is this: the book will become a primary source in every debate and lawsuit involving affirmative action for the next decade, and maybe longer. It's a book that will launch a thousand footnotes.
And that was clearly part of the authors' intention. They remind us that in 1978, when Justice Lewis Powell wrote the Supreme Court's opinion in the seminal Bakke v. Regents of the University of California case, he was willing to take on faith the word of university officials that diversity, including ethnic diversity, was an important asset in the education of university students. But this is a different age, with a different Court. Race preferences have already been prohibited in three parts of the country: by nearly identical voter initiatives in California and Washington State, and by a federal appellate court in Texas, Lou isiana, and Ark ansas. Preferences are under political and legal attack elsewhere as well—a pair of major cases challenging admissions practices at the University of Michigan will come to trial this June—and the issue is almost certain to return to a Sup reme Court that has begun to take an increasingly dubious view of affirmative action programs. Bakke itself, which allows universities to use race as one "plus factor" in choosing students, and which has been ruled a dead letter in Hopwood v. Texas, could be in jeopardy. The defenders of affirmative action urgently need good data.
And good data are precisely what the authors mean to provide. Both Bowen and Bok are former presidents of elite universities (Princeton and Harvard, respectively) that have long considered race as a plus in their admission practices. In The Shape of the River—the title refers to a quotation from Mark Twain's Life on the Mississippi about the need to know the whole shape of the river in order to navigate it—they have put together a vast amount of statistical data in an effort to demonstrate not only the success of university race preferences in the subsequent careers of thousands of individuals, but, by extrapolation, the central importance of affirmative action in creating a black upper middle class in Amer ica.
The statistics come from a large database called College and Beyond (C&B) that was created by the Mellon Foundation, which Bowen now heads. The database tracks some 45,000 students who entered one of a sample of 28 selective colleges and universities in two years, 1976 and 1989. The 28 institutions are broken into three groups. The most selective schools—with average SAT scores of at least 1300—were Bryn Mawr, Duke, Princeton, Rice, Stanford, Swarth more, Williams, and Yale; the next most selective (with average SATs between 1151 and 1300) were Barnard, Columbia, Emory, Hamilton, Kenyon, Northwestern, Oberlin, Smith, Tufts, the Uni versity of Pennsylvania, Vand erbilt, Washington University, Wellesley, and Wesleyan; in the third rank of selectivity (average SATs of 1150 or below) were Denison, Miami University of Ohio, Penn State, Tulane, Mich igan, and the University of North Carolina. Four of the institutions are public; the rest are private. Only two are in the West or South west. The database includes information on everything from each entering student's ethnicity, gender, SAT score, and college grades, to any graduate school he or she attended and his or her income and community activities long after graduation. These data are supplemented with extensive interviews with those graduates about their undergraduate experiences and particularly their relationships with students of other races. Unfortunately, though perhaps understandably, given its time frame and its focus on eastern colleges, the database concerns itself almost entirely with the results of affirmative action for blacks, not with the growing proportion of Hispanic students now going to college. And since Bowen and Bok have detailed admissions data from only five of those institutions—all private—the picture is even more limited. Still, it's probably the most ambitious such study ever undertaken.
But if The Shape of the River is an explicit attempt to demonstrate the success of race preferences in the admissions policies of those 28 institutions, it has an equally important, though perhaps unintended, subtext: the way elite institutions exploit their powerful role in selecting and certifying people for preferred positions in American society. At its core, therefore, it is almost as much about the confounding relationship between class and merit as it is about race, as much a defense of the way people are chosen and vetted by our elite colleges as an argument for race preferences in that process. It is, at bottom, a bold—and, in these days, a rather rare—assertion of establishment prerogatives in a very select category of institutions. Unfortunately that also makes their mass of data, for all its heft, of limited use in any broader examination of the consequences of race preferences, even in higher education. Their conclusions don't even seem to be very applicable to some of the institutions in their own survey which do not fall within that most select category.
Bowen and Bok acknowledge that the blacks in the 28-institution survey entered with far lower average SAT scores than whites—blacks with combined verbal and math scores between 1200 and 1249 had a 60 percent chance of being admitted; whites had a 19 percent chance. And not surprisingly, blacks earned significantly lower grades in college. The average black student's grade point average (GPA) is at the twenty-third percentile of the survey group; the average white's is at the fifty-third percentile. But Bowen and Bok also point out that, contrary to critics like Stephan and Abigail Thernstrom, who argue that affirmative action fosters black failure by putting students into colleges that are too demanding for them to compete successfully, blacks who attend the most-selective institutions—even those admitted with relatively moderate SAT scores by the standards of the 28 institutions—graduate (within six years) at a higher rate and enjoy more subsequent success than those with similar SAT scores who attend less-selective institutions.
This pattern persists through life. After graduation, while black men earn somewhat less than their white classmates, black graduates of select institutions, male and female, get advanced degrees at the same rate as their white classmates. They earn considerably more than average white college graduates from run-of-the-mill universities, are more active in community organizations, and seem more satisfied with their college experience. At the same time, support for affirmative action seems reasonably high both among white and black alumni at most of these more-selective schools. More important, black graduates of these colleges now make up a substantial proportion of the black leadership class in America—as partners in law firms, public officials, corporate officers, and in other leadership positions.
As Bowen and Bok relate in their book:
To take but a single illustration: since starting to admit large numbers of black students in the late 1960s, the Harvard Law School has numbered among its black graduates more than one hundred partners in law firms, more than ninety black alumni/ae with the title of Chief Executive Officer, Vice President or General Counsel of a Corp oration, more than seventy professors, at least thirty judges, two members of Congress, the mayor of a major American city, the head of the Office of Man agement and Budget, and an Assistant U.S. Attorney General.
If you get admitted to Yale or Princeton or Williams or Stan ford, in other words, your chances of graduating are higher, regardless of race, despite your modest test scores and relatively weak academic performance in college, than if you went to a less-selective place. And what follows from this, of course, is that you also have better chances of getting into a good professional or graduate school (where your chances of graduation are also high), of making useful career connections, and of landing in a prestigious law firm or corporate office and thereby making lots of money.
The Old College Try
Bowen and Bok say—a little disingenuously—that they're not certain exactly why this is true. Is it because the selective institutions pick better students, or is it because they do a better job educating them once they get there? They tilt toward the latter explanation, suggesting that the rich resources of the elite colleges provide sufficient help, and provide it soon enough, that most of their students make it. What they don't say is that institutional wealth and prestige make up for a lot of deficiencies. More to the point, this is hardly a new phenomenon. Long before there was affirmative action, it was common belief that one had to do something really dreadful to get tossed out of Harvard or Yale. And once you had the degree—the college tie that prestigious institutions provided—other doors opened up as well. Grad uation from an elite institution makes you a member of a selective club. "The more selective the school," as Bowen and Bok say, "the more the student achieved subsequently." (The same thing appears to be true for many medical schools.) The big hurdle has always been getting in, not getting a degree.
Call it the Ivy bonus. There was a time a couple of generations ago when there was a lot of talk in higher education—and in the elite colleges particularly—about the well-rounded man and the gentleman's C. Charitably, these terms applied to students who weren't academic stars but had other desirable attributes—as athletes, musicians, poets, sons of alumni, farm boys, and the deserving poor—that the college considered beneficial, even if only because their admission would guarantee the continued engagement or financial support of alumni. Uncharitably, this group also included a high concentration of wealthy and well-connected WASPs from Deerfield or Andover, who skated through in their white shoes and went straight to cushy jobs on Wall Street and Madison Avenue.
Standards, presumably, are tougher these days. It's probably correct that, as Bowen and Bok argue, everybody who attends the most-selective institutions on their list these days is well prepared. While they estimate that more than half of the black students in their survey would not have been admitted without consideration of race, Bowen and Bok try to remind grumpy alumni from the good old days that today's entering blacks have higher average scores and far richer academic backgrounds than did the white boys in the class of 1940. In any case, they argue, merit is not defined solely as high grades and test scores; it depends greatly on "what one is trying to achieve." The job of admissions officers—"an eclectic and interpretive art," they call it—is not to determine "who has a 'right' to a place in the class, since we do not think that admission is a right possessed by anyone." What they do is choose those applicants "considered individually and collectively, [who] will take fullest advantage of what the college has to offer, contribute most to the educational pro cess . . . and be most successful in using what they have learned for the benefit of the larger society." But only a strong patrician institution that is immune from the political demands imposed on public universities could make such an assertion.
Still, the old tradition has been a perfect fit for the new policies. To all the other characteristics, we've simply added race—which, as Powell said in Bakke, could pass constitutional must er as another "plus factor" in a college's effort to achieve a rich and diverse student body. The gentleman's C presumably is gone, despite allegations like those of Lino Graglia, the University of Texas law professor who argues that affirmative action students—unable to compete with students admitted according to the regular process—seek out academic safe havens like black studies and multicultural studies where they find easier going and more support. Those charges seem to be false, at least in the C&B schools. "Blacks and whites," Bowen and Bok write, "were equally likely to have majored in philosophy, economics, the natural sciences and engineering," while only a few majored in African-American or black studies.
But if the gentleman's C is gone, the minority C seems to be alive and well. And it's at this point that Bowen and Bok open themselves to attacks from conservatives like Shelby Steele, who accuses these "two white guys" of "massaging data" that utterly fails, in his view, to demonstrate in any way that the affirmative action students are in fact "competitive," despite their subsequent high salaries and fancy-sounding jobs. "Those blacks," he said in an interview with me last fall, "are being artificially propped up through life"—career-long affirmative action babies getting breaks from guilty whites trying "to reclaim their moral authority."
The hollowness of their argument, Steele charges, is underlined by the authors' use of community involvement as one of the criteria of success. Since it's largely the marginal people in big organizations—those who won't make it to the very top—who comp ensate by immersing themselves in such activities, Bowen and Bok inadvertently prove the emptiness of their own case. In his new book, A Dream Deferred: The Second Betrayal of Black Freedom in America, Steele writes, "the top quartile of black American students—often from two-parent families with six-figure incomes and private school educations—is frequently not competitive with whites and Asians from lower quartiles. But it is precisely this top quartile of black students that has been most aggressively pursued for the last thirty years with affirmative action preferences." Affirmative action, in his view, is "the institutionalization of low expectations."
But some people are always going to be in the bottom quarter of the class, and their chances of moving into prestigious jobs by virtue of their college's status and connections are just as great as their classmates'. While Steele may be right to worry that affirmative action casts a shadow on the achievements of the thousands of blacks who would have gotten into these colleges even without any race preference, the C&B data for at least the most selective (private) institutions appear to show no great concern on the part of blacks, nor any widespread resentment among whites who were rejected. Bowen and Bok also provide a long string of quotations from blacks and whites about the benefits, as they see them, of being exposed to students from other racial backgrounds. But those arguments, coming from a very selective segment of the population, seem a bit contrived. One of the ways you demonstrate your high status is by not showing your racial resentment.
Still, guilt may not be the most important impulse here. At a time when society itself is becoming increasingly diverse, race-based affirmative action—whether in admissions offices, government jobs, or private business—is not so much an effort to reclaim moral authority as to reinforce institutional legitimacy. To paraphrase Bowen and Bok, it's not just that the affirmative action minority students need the elite colleges; it's also that the elite colleges need the minority students. More important, perhaps, if one harkens back to an older view of the mission of the elite institutions, it's rather hard to determine to what extent those institutions created the establishment, and to what extent they merely served to train and polish those who were already destined for it. In an age that badly needs minority leadership, the kind of affirmative action practiced at Yale or Princeton is not all that different.
Cream of the Crop
The real difficulties—both in the use of this study and for race preferences generally—lie elsewhere. Preferences notwithstanding, more than 75 percent of the black students entering the C&B schools in 1989 had entering SAT scores above 1100, and nearly half scored over 1200, which puts them at the very top of the pool of black test takers. They are already winners. Most black high school students never get near those numbers, assuming they even take the SAT. And while the black-white gap in reading has shrunk by half in the last generation, and the math gap by a third (as measured by the National Assessment of Educational Progress), the differences are still substantial. [See Christopher Jencks and Meredith Phillips, "America's Next Ach ieve ment Test," TAP, Sept ember-October 1998.] The C&B data, in short, have little relevance for the vast majority of institutions and black students.
Even within the nation's selective institutions, estimated to be about 20 to 30 percent of all colleges and universities, a great deal of what is labeled as affirmative action is very different from the practices Bowen and Bok have in mind. At places like Berkeley, where the mean score of entering students in 1989 was 1176, and the University of Texas Law School, where academic race preferences were eliminated after coming under the most intense attacks, the admissions decisions were not made, as in the Ivy League, on the basis of a candidate's whole record and background. Instead, they were based largely on a numerical formula that included little more than grades, test scores, and race. Under that formula, blacks and Hispanics were admitted with substantially lower scores than other applicants. In Texas for a time, the law school maintained an entirely separate admissions process for minority applicants. Those systems have little in common with Bowen's and Bok's broad definition of merit as something that goes far beyond grades, test scores, or even the hard work that many students do to get them. And thus they were always hard to defend.
What made the task even harder was the lack of candor from public institutions that pretended that they weren't really giving preferences, or that they were using them only in the most marginal ways. When it was finally disclosed how large the black-white gaps sometimes were, advocates defended preferences as ways of creating opportunities for individual minority students—most of whom were well enough qualified to go elsewhere. What they failed to do was proclaim the minority students' much greater importance to the institution and the larger society—the fact that "merit," as Bowen and Bok would have it, should itself be conditioned by broader institutional objectives, "permitting students to benefit from diversity on campus, and addressing long-term societal needs."
Unfortunately, those broader objectives are far easier to maintain in private institutions than they are in the small circle of public institutions—Berkeley, UCLA, the University of Texas Law School, and a couple of others—where admission is highly competitive and where race preferences have therefore been a major issue. Even at the University of Michigan, which is part of the C&B survey and is now being sued for reverse discrimination in its admissions practices, admission seems to rest not on an evaluation of each candidate's complete record, but on a formula that automatically gives a large edge to Hispanics, blacks, and American Indians. The university—to quote the complaint in one of the suits—does not "merely use race as a 'plus' factor or as one of many factors to attain a diverse student body. Rather, race was one of the predominant factors . . . used for determining admission." And contrary to the authors' conclusions that affirmative action does not breed widespread resentment but rather fosters understanding and integration, Michigan's own survey, as reported recently in the Chronicle of Higher Education, showed that two-thirds of its white students agreed that "students of color are given advantages that discriminate against other students."
None of this is surprising. In part the public universities just don't have the resources to read every word in every applicant's folder, in every essay and letter of recommendation. In part they face intense political pressures for "fairness" which make it hard to maintain any selection process that seems to depend very much on subjective judgments. The public universities operate in a very different social and political environment, no matter how respected their academic programs are. In Texas, after the Fifth Circuit Court of Appeals not only prohibited race preferences but declared the Bakke decision a dead letter, the legislature adopted a system requiring the university to admit the top 10 percent of the graduates of every Texas high school. In California, when the regents abolished race preferences in the summer of 1995, they decreed that the University of California (in the name of "merit") had to admit at least 60 percent of applicants strictly on the basis of grades and test scores. In both cases, that was tantamount to replacing one arbitrary formula with another.
Not surprisingly, Bowen and Bok don't think much of the Texas alternative. While it may increase the number of minority admissions from de facto segregated schools, the chances of admitting students who are incapable of the work are far greater, even as the chances of admission for able minority students from academically tough schools decline. Nor do the authors align themselves with the left-wing critics who argue that the SAT is biased against minorities and should be abandoned altogether. On the contrary, they point out that the SAT, for unknown reasons, actually overpredicts how well blacks do in the C&B institutions—or to put it another way, blacks, for reasons still not clear, perform less well in college than the test predicts. Nor, again, do Bowen and Bok think much of Richard Kahlenberg's thesis that "class-based affirmative action"—choices based on low-income or other measures of socioeconomic status—can be used as surrogates for race, since there are still many more poor whites—or, in places like California, poor Asians—than poor blacks. In fact, as they point out, most of the blacks who make it into the C&B institutions are not the sons and daughters of poor people; they are, for the most part, as upper-middle-class as everyone else. With rare exceptions, the Ivy League has never been a launching pad for the poor.
Outside the Ivory Tower
But neither is the Bowen-Bok formula a paradigm for anything in the larger world. While the Bowen-Bok data, which included only four public institutions in the 28-institution sample, may become part of a defense against the judicial demolition of Bakke, they won't be very useful against state-level anti-preference ballot initiatives like Proposition 209 in California and Proposition 200 in Washington State. They may not even be very effective in fending off suits against institutions like the University of Michigan, which will have a hard time showing that it fits the C&B model. Applying Bowen-Bok data becomes particularly difficult in highly diverse states like California, where it's not simply a question of admitting a few more blacks—and thus reducing the percentage of whites ever so slightly—but rather one of accommodating claimants from a highly diverse spectrum of ethnic groups, particularly Asians and His panics, both of whom have been victims of widespread discrimination in the recent past. If blacks are admitted to Berkeley or UCLA at demonstrably higher rates than are Chinese applicants with higher scores, even the most eloquent defense of an institution's larger social obligation may sound like a hollow rationale for yet another form of Asian exclusion.
That's not to deny that the price of the race-blind equity that follows the abolition of race preferences in California and Texas may be steep, at least in the short run—on the one hand a sharp decline, already evident in California, in minority admissions; on the other hand, mounting political pressures to lower the threshold for everybody. One of the backers of Proposition 209, a Berkeley law professor named John Yoo, later told the New Yorker's Jeffrey Rosen that he didn't realize until after the vote "that affirmative action, as it was applied by the schools, allowed you to have some racial diversity and at the same time maintain intellectual standards for the majority of your institutions." Institu tions that are legally forced to ignore race in admissions decisions will, inevitably, either search for superficially color-blind fudges or adopt lower standards for everybody to avoid resegregation. Boalt has already dropped the practice of adjusting its applicants' college GPAs according to the presumed academic difficulty of the schools they attended. Hence the Boalt admissions process rates a B from Harvard or Rice as equivalent to a B from Chico State or Howard. And as a result, the number of minority students Boalt admits, which was down to virtually zero in the first year after race preferences were prohibited, has begun to climb again.
All of which leaves selective public institutions with a singular dilemma. Bowen and Bok make a powerful case for the consideration of race in the sophisticated admissions processes of the elite private institutions. But it's not a process that most tax-supported institutions can use. Thirty years ago, University of California President Clark Kerr proudly marked the moment when Berkeley's graduate programs were rated as being on a par with, and sometimes above, Harvard's. But as the affirmative action story makes clear, no tax-supported university, regardless of the prestige of its faculty and research programs, can ever behave like an elite private university. What the Ivy Leagues do may be both successful and admirable, but it rarely applies to the rest of America.