CHAPTER TEN
THE HYDRA OF PREFERENCES
The Evasion of Prop 209
at the University of California
IN THE SPRING OF 1999, on assignment from the
New York Times Magazine, James Traub visited some campuses of the University of California to assess Prop 209’s implementation. What he saw, he generally liked. The campuses seemed to be flourishing and at peace. As Traub reported, “When I asked Kenya Coleman, a black student [at UC Riverside] who was majoring in business, whether she felt that students denied admission to Berkeley would be losing out on something at Riverside, she bristled slightly and said, ‘If they end up here it would be a blessing in disguise.’” Another black Riverside senior told him, “I think I am more prepared in terms of graduate school than I would have been if I had gone to UCLA.” Traub also noted that
Ending affirmative action has had one unpublicized and profoundly desirable consequence: it has forced the university to try to expand the pool of eligible minority students. Outreach programs have . . . proliferated; the State Legislature authorized $38.5 million for such efforts last year and has required the public schools to spend an additional $31 million on similar initiatives. UC campuses are now reaching down into the high schools, the junior highs, and even the elementary schools to help minority students achieve the kind of academic record that will make them eligible for admission.
Traub pointed out that a 1996 poll of Berkeley students showed that most of them opposed racial preferences, but they were happy to vote for a $3 per student increase in student fees to support improved minority outreach. He also observed that UC administrators, like Chancellor Robert Berdahl of Berkeley, did not think that students necessarily knew what was good for them. Those running the UC system remained overwhelmingly convinced of the goodness of the racial preferences they had been forced to give up.
Our research found the same patterns. On the ground, students were adapting well. The warming effect strongly implied that black and Hispanic high school seniors embraced the opportunity to attend a university that had accepted them without regard to their race. Once they arrived at UC, as we saw in Chapter Nine, they performed exceedingly well. Unprecedented outreach efforts were expanding the base: By 2001 black applications to the university would be 10 percent above the previous, pre-209 record, and Hispanic applications would be more than 30 percent above pre-209 levels.
Yet top UC administrators were virtually unanimous in viewing the post-209 landscape with disgust. They deeply resented the intrusion of the regents and the voters into their realm, and they felt continually on the defensive in their dealings with the rest of the academic world. They condemned what had happened, and their condemnation took on a self-fulfilling character: Their expectation of disaster made them incapable of seeing anything else.
Consider, for example, a report prepared several years after Prop 209’s implementation by Susan Wilbur, the university-wide director of admissions. Wilbur was asked to examine the impact of Prop 209 on minority enrollment patterns. As we discussed in Chapter Eight, the formal adoption of race-neutrality was accompanied by rising applications and record rates of acceptance among blacks and Hispanics given offers of admission. But Wilbur described the situation differently:
I will demonstrate that the loss of talented African American and Latino students who are admitted to the university [i.e., by turning down offers of admission] has contributed to the decline in diversity experienced at the UCs. Underrepresented minority students (and especially African American students) are more likely than others to decline an offer of admission from the UC in favor of an offer from a top-tier institution . . . in a post-209 world, underrepresented minority students are taking advantage of alternative offers of admission. . . . The loss of diversity incurred by 209 may have a more negative effect on the educational opportunities of those who attend the UCs than it does on those who do not.
How did Wilbur reach this powerful conclusion? How could she miss the startling surge in yield rates that followed Prop 209? It was easy: Wilbur completely ignored data from the pre-209 era. She only examined yield patterns among UC students admitted long after Prop 209 had been implemented. Her findings were based on the simple fact that academically strong blacks (and, to a lesser extent, Hispanics) were more likely than whites and Asians to turn down a UC offer in favor of a competing offer from another elite school. But of course, because blacks (and, to a lesser extent, Hispanics) are much more likely, other things being equal, than whites and Asians to have a competing offer from an elite school, Wilbur was measuring nothing more than the continued existence of racial preferences at other schools. Had Wilbur done the exact same analysis of, say, students offered admission by Harvard or Stanford, she would have found the exact same patterns.
Wilbur’s analysis might have been wildly misleading, but it was prepared for an audience of UC colleagues who very much wanted to be misled. UC officials regularly made public statements about the harmful effects of Prop 209 upon the university; to our knowledge, they made no comparable statements about the many ways in which black and Hispanic performance surged during those years, the more even distribution of minority students across campuses, or the steady rise of black and Hispanic enrollment numbers UC-WIDE. The dominant mindset was that Prop 209 was harmful. The question that nagged at administrators and many faculty, in a way that increased over time, was whether Prop 209 was so illegitimate that there was an affirmative duty to evade it.
One of us (Sander) was active in the university community throughout the period of Prop 209’s implementation. Over the years he discussed with many other UC faculty and administrators how race-neutrality operated within their schools and programs. It was exceedingly rare for any of his colleagues to wonder whether the new era might actually benefit blacks and Hispanics. Sander’s colleagues contended either that Prop 209 was an unavoidable evil that must be accommodated or that it was an evil that must be resisted. In terms of admission policy, these two schools of thought argued, respectively, for admissions methods that would produce disproportionate racial dividends or for the covert violation of the law.
In December 1996, just weeks after Prop 209 won at the polls, Sander was invited to a UC system-wide meeting in Oakland to discuss ways that graduate programs would adapt to race-neutrality. The meeting was dominated by a discussion of the feasibility and merits of evasion versus the “indirect racial dividend” approach. For many small programs (e.g., graduate programs in areas like psychology, history, or German literature) faculty spoke of the feasibility of evasion: The number of students was so small, and the criteria for selection so subjective, that outside investigators could not easily detect racial discrimination. For larger programs, such as law schools or business schools, that would obviously be more difficult. Sander advanced the notion that socioeconomic (SES) preferences were a viable and desirable alternative to racial preferences and would yield a particularly strong racial dividend if they took into account not only family circumstances (e.g., parental income and education) but also neighborhood conditions (e.g., the level of poverty in the applicant’s neighborhood). The others listened with interest, but there was a good deal of skepticism that SES preferences would really produce the desired effect because, it was repeatedly pointed out, there were a lot of low-income whites and Asians.
Soon after, a colleague from a UC medical school told Sander that the medical schools had the ideal antidote to Prop 209: interviews. Because medical schools have very large faculties relative to the number of students admitted each year, it is feasible to interview finalists for admission and to give as high a mark as necessary to black and Hispanic interviewees to produce the desired level of racial diversity.
Boalt Hall Law School, on the Berkeley campus and the most elite of the university’s four law schools, appeared to play things completely straight in the first year of race-neutrality. Unlike UCLA’s Law School, which adopted a version of Sander’s socioeconomic strategy and ended up with ten black first-year students (down from nineteen in 1996), Boalt seemed to focus narrowly on its own version of the academic index and enrolled only a single black student in 1997 (down from twenty in 1996). This became an emblematic moment for the university; the “single black” received national news coverage and Doonesbury did a long riff. Even Prop 209 proponents criticized Boalt for the precipitous drop in black enrollment, suggesting that Boalt had neglected its outreach responsibilities.
The decline at Boalt was, however drastic, a foreseeable consequence of race neutrality. Unlike the undergraduate colleges, UC’s top law schools operated in a national market. Most of Boalt’s students came from outside California, and nearly all of them believed that the best path to career success was to attend the most elite law school that would have them, wherever it might be located. (As we saw in Chapters Four and Six, the evidence for this view is shaky at best.) Any black student admitted to Boalt without a racial preference would probably have a strong enough academic index to be admitted to any other law school in the country with a racial preference, and often with a race-based scholarship as well. Boalt had actually admitted eighteen black students in 1997 (down from seventy-seven in 1996), but these eighteen were generally the very strongest black applicants, with other highly attractive offers, and few if any of them would have accepted Boalt even in the pre-209 era.
Nonetheless, Boalt’s students were unhappy, and some of its faculty felt humiliated by negative attention focused on its “one black” class. The next year a committee was established to review many admissions files and exercise more “subjective” judgments. Black admissions rose by over 40 percent, and eight blacks enrolled at Boalt in 1998. By 2002 Boalt was admitting forty-four blacks and enrolling twelve. There was little doubt that Boalt decision makers had reinstituted racial preferences. In some ranges of the academic index, blacks were ten times as likely as whites to be admitted. According to a colleague from Boalt, when asked how the admissions committee managed to admit so many blacks, a faculty member of the committee smiled and said, “No one can know what’s in my head.”
In the 2000–2001 academic year, faculty at UCLA’s law school followed Boalt’s lead down the race-conscious road, voting to create a special admissions path for students interested in “Critical Race Studies” (CRS). CRS, in legal studies, is a school of thought that contends that racial categories are deeply embedded in the reasoning and underlying policies of the law. The CRS path was seen by some faculty as a more subtle way to sidestep Prop 209 than Boalt’s approach because the school would not overtly be choosing students based on race but based on their substantive interest in a program that would have the effect of producing more students of color. There was also a plausible argument that by admitting students into a substantive program, the CRS admissions path would contribute to intellectual diversity at the school. Subtlety, however, went out the window when, during the program’s first full year of admissions, nearly as many whites as blacks applied to the program. The program’s special admissions committee admitted eight of the black applicants and none of the white applicants, even though many of the rejected whites had higher academic indices than any of the accepted blacks. The CRS students received preferences very similar to the racial preferences that UCLA used before Prop 209.
Meanwhile, the political climate in the state and on the UC Board of Regents was changing. Pete Wilson stepped down as governor at the end of 1998; his successor, Democratic governor Gray Davis, had been a vocal opponent of Prop 209. His appointments to the regents gradually changed its politics. In May 2001 the regents voted unanimously to rescind its 1995 bans on preferences. These precursors to Prop 209 (as discussed in Chapter Seven) had been made largely moot by it, or so university officials argued. By repealing the measures and reaffirming the university’s commitment to diversity, UC administrators suggested, the regents could help dispel the impression that the university was somehow hostile to minorities—in other words, to dispel a “chilling effect” that seemed to exist mostly in administrators’ imaginations. The repeal did have one important substantive effect, however: It ended a requirement that between 50 percent and 75 percent of the students on each campus be selected primarily on “academic” grounds. This opened the path toward less formulaic and more “holistic” admissions, under which each student could be rather mysteriously assessed based on “all” her characteristics—much as Boalt Law School was now doing.
Accompanying the repeal of SP-1 came a new path to “UC eligibility” called “Eligibility in the Local Context” (ELC). As the reader will recall, students had long become UC eligible by taking a particular regimen of courses and achieving a combination of test scores and high school grades that placed them in the top eighth of California seniors. In the years leading up to Prop 209 administrators had used a back door, “special admissions,” to increase on a case-by-case basis the number of blacks and Hispanics eligible for UC admission. The end of race-conscious special admissions had been the main reason for the drop in minority enrollment after Prop 209, a drop that had since vanished for Hispanics and was now (by 2000–2001) only about 10 percent below pre-209 levels for blacks. ELC would expand UC eligibility by making students UC-eligible if they ranked among the top 4 percent of their own high school class. The main beneficiaries of ELC would be students who had attended weak high schools, had done badly on standardized tests, but had studied conscientiously and received strong grades.
ELC was appealing as a social justice measure; it reached out to many of the state’s poorest communities and reduced the emphasis on test scores. Because high schools participating in ELC essentially had to “partner” with the university in determining eligibility, it fostered UC’s growing connection to the K-12 pipeline. But it also had an obvious potential to increase mismatch problems. It meant, in effect, that one’s high school became a major predictor of admission, but in a way negatively—not positively—related to high school quality. This was exacerbated when some of the more elite UC campuses began to use a measure of California high school quality as an explicit factor in admissions, with students attending weak high schools having much higher chances of admission to Berkeley and UCLA than otherwise similar students who had attended strong high schools. In other words, ELC became a pipeline not just to the UCs but often to the most elite UCs, who were most worried about their low black and Hispanic numbers.
The following year Berkeley decided something more was needed. In 2003 it announced that the admissions process would be reorganized into a “holistic” system. In the past, admissions officers had used a variety of academic screens to narrow the applicant pool. Under the holistic system, two specially trained readers would examine an application in its entirety, considering every aspect of a student’s background and qualifications, and would assign an overall score to that applicant. If the two readers came up with a similar holistic score, then the average of their scores would be the final evaluation of the applicant and would determine his or her place in line for admissions. If the reader scores diverged, a more senior reader would help to determine which “read” would be determinative.
In principle, holistic admissions was far from a radical idea. This was pretty much how admissions had always been done at small liberal arts colleges. In the context of Prop 209, however, holistic admissions was seen—by both supporters and critics—as another way of getting around the ban on racial preferences and increasing minority enrollment. The reader could guess the race of the applicant and let that influence the holistic score, while the admissions office could further hone the correlates of race and instruct readers to give those correlates special weight. As we will see in Chapter Thirteen, the Supreme Court would, in its June 2003 decision in Grutter v. Bollinger, contemporaneously endorse holistic admissions as a constitutional method of weighing the overall diversity contributions of individuals (including their race) to academic programs.
The effect of holistic admissions at Berkeley was very modest, both in its racial impact and its academic effects. There was only a slight uptick in black enrollment, and the general level of academic disparity between blacks and Hispanics, compared with everyone else, increased only slightly.
Holistic admissions would soon take on an entirely different cast at UCLA. In the spring of 2006 the Los Angeles Times published a front-page story headlined, “A Startling Statistic at UCLA.” According to the story, UCLA would be enrolling only 96 blacks among the incoming freshman class, down from 118 the year before. (In the end, 95 blacks would matriculate.) The article quoted Professor Darnell Hunt, a UCLA sociologist and head of the Bunche Center for African American Studies, who contended that Berkeley’s “holistic” system was fairer than those used at UCLA and UC San Diego (the next-most elite campus) and produced larger minority—and, in particular, black—enrollments. Hunt criticized the school for using factors of “questionable validity” in rejecting blacks, and the article seemed to endorse his suggestion that something sinister and harmful to minorities was afoot at UCLA.
Over the next several months the issue of declining black enrollment at UCLA became a low-boiling crisis. A widely expressed feeling among faculty, students, and black alumni was that the school had gone from race-neutrality to a regime in which blacks were gradually fading away. Administrators reacted defensively, agreeing that the numbers were alarming and promising to do better.
In reality, there was no crisis. Far from neglecting black applicants, UCLA admissions officers performed elaborate gymnastics to keep black numbers as high as they were. Despite the ban on racial preferences, black freshmen in 2006 entered with a mean SAT score of 1091, compared to 1311 for whites. True, black freshman enrollment at UCLA had fallen from 118 in 2005 to 95 in 2006; but black transfer enrollees had risen by essentially the same amount (from 85 in 2005 to 104 in 2006). True, Berkeley enrolled 148 black freshmen in 2006, but its average black freshman enrollment since Prop 209 had been virtually identical to UCLA’s, and UCLA consistently beat Berkeley in the number of black transfers it attracted. Throughout the post-209 period UCLA had the highest aggregate black enrollment of any of the three elite UC campuses.
Much more importantly, black outcomes at UCLA had improved dramatically since the implementation of Prop 209. From 1992 to 1997 entering black freshmen had an average four-year graduation rate of about 20 percent and a six-year graduation rate of about 55 percent; those rates had risen to nearly 40 percent and 70 percent since Prop 209. Indeed, despite the very large impact Prop 209 had on black freshman enrollment at UCLA, the number of blacks actually receiving bachelor degrees from UCLA had declined rather modestly—around 20 percent—from pre-209 levels, and the number of UCLA bachelor degrees going to blacks and Hispanics combined had remained essentially unaffected. And it was true at UCLA, as at the UC campuses generally, that rising graduation rates among blacks and Hispanics were accompanied by rising GPAs and higher concentrations in STEM fields. This was a wonderful story: It meant that UCLA was doing a dramatically better job than before of turning minority students into college graduates with promising futures.
In short, the Los Angeles Times story, misleading as it was, provided a perfect learning opportunity for the UC community—a chance to reveal and explain how the prohibition on explicit racial preferences had become on the UC campuses a way of preserving and enhancing opportunity; how, of all groups affected by Prop 209, the greatest beneficiaries had been blacks and Hispanics in substantively improved outcomes; how rising yield rates among those groups provided implicit confirmation of this truth.
FIGURE 10.1. Minority Enrollment Falls While Graduation Numbers Hold Steady
Source: UC Office of the President. Note that the number of graduates is adjusted to parallel freshman cohorts; thus, graduates for the 1992 cohort is an average of those receiving B.A.s in 1996 through 1998. For all years, B.A.s include those received by students transferring to UCLA.
But the leaders of UCLA did not do this. Though one of us (Sander) knows many of these leaders personally and truly admires the wisdom and moral purpose of each of those he knows, the leadership exercised in this case fell seriously short. UCLA leaders sought to ally themselves with those accusing the campus of racism and promised to make reforms.
We have an unusual window into what happened next because of an aide-memoire produced in 2008 by Tim Groseclose, a political scientist at UCLA and a member of its admissions committee from 2004 until 2007. Groseclose is right of center and therefore sometimes at odds with his more liberal colleagues, but he is an eminent scholar and his account remains, so far as we know, completely undisputed.
According to Groseclose, UCLA’s chancellor Norman Abrams asked to meet in the early fall of 2006 with UCLA’s Committee on Undergraduate Admissions and Relations with Schools (CUARS). Groseclose reported Abrams’s opening remarks as follows:
First, I want to say how much I favor and respect faculty governance. I don’t want to pressure you. But at the same time, we worry about many of the same things. I want to report to you what we are hearing from the outside world. Several constituencies of UCLA are distressed and upset about the very low numbers of African American freshmen. The political angst and concern is enormous. I don’t feel the pressure. I sublimate very well. But there is pressure exerted upon me. The numbers of underrepresented minorities on campus are too small. . . . I ask that you make the whole admissions process holistic. Not only that, I have a further request: This is that you do it quickly and adopt the exact same process that Berkeley currently uses.
Soon afterward CUARS voted to move UCLA toward the Berkeley holistic system. Groseclose (along with an African American colleague) voted against the proposal. Groseclose argued that holistic admissions, especially in this environment, would open up opportunities to engage in forms of discrimination prohibited by Prop 209. In relatively short order, the admissions staff had started setting up the mechanisms for holistic admissions, including the hiring of many part-time readers to do the intensive “holistic reads” of individual files. Groseclose soon learned that of the 160 readers hired to handle the job, 40 were African American (in a labor market where black college graduates made up about 7 percent of the total pool). Asian readers were underrepresented.
Six months later, in the spring of 2007, UCLA announced a dramatic increase in African American freshman enrollment for the coming year—an exact doubling, from 95 to 190. The mood of campus administrators was celebratory; it was not generally pointed out that Hispanic freshman enrollment at UCLA, which had been at or close to pre-209 levels for several years, slightly dipped. Clearly other forces besides the change to holistic admissions had played a role. Abrams and other campus leaders had mobilized black alumni to help improve recruitment among talented black high school students and to develop private funding sources that could channel recruitment scholarships to blacks accepted by the campus.
Still, the question remained what role admissions had played. In another New York Times magazine article, this one in September 2007, David Leonhardt generally praised UCLA’s efforts to foster opportunities for blacks, but noted, “The big question that hangs over UCLA’s success, of course, is whether the university broke the law. Looking at the numbers, it’s hard not to conclude that race was a factor in this year’s admissions decisions.” The key numbers Leonhardt pointed out were the SAT scores of admitted students, which fell sharply for blacks while holding steady for other racial groups. This, along with the dip in Hispanic enrollment and the 50 percent jump in the black admission rate, was damning indeed. If the purpose of the holistic process was to do a better job of considering one’s success in overcoming personal disadvantage, how could it produce huge jumps in black admissions while leaving Hispanics—who, in California, are fully as disadvantaged as blacks—untouched?
Groseclose wanted to look closely at the numbers, and he asked the staff director of admissions to provide him a sample from the university’s admissions data covering several hundred students from before and after the “holistic” era. He received a variety of deflecting responses. The director expressed his view that “we should not study or do any analysis of holistic admissions at UCLA until we have at least four or five years of the outcome to avoid normal annual fluctuations.” The faculty chair of CUARS suggested that although a study was a good idea, it would be best if done by CUARS as a whole rather than by individual faculty. This implied that no individual faculty member would actually get the raw data. Groseclose brought matters to a head by introducing a formal motion that all faculty members of a workgroup on the consequences of holistic admission be given an anonymous database of a sample of applications. His motion failed on a 3–3 vote.
Eventually, the committee determined (in consultation with campus administrators) that UCLA should retain an empirical specialist to study the effect of the holistic process. The administration offered to fund the study at $100,000. Groseclose, who was a distinguished empiricist and thought the key analytic questions were quite simple, offered to do the study for free on condition that the entire CUARS committee could have access to the data. His offer was declined. Groseclose’s inference was that the university was unwilling to have a skeptical faculty member of its own admissions committee scrutinize a process that had produced suspiciously discriminatory results. Groseclose resigned in protest over this lack of transparency in September 2008.
Groseclose’s resignation made local news, and Sander was keenly interested in the story. Soon afterward they met for the first time. Sander had just had some success (in collaboration with others) shaking loose pre- and post-209 data on student outcomes from the UC Office of the President; perhaps the two of them could secure data on holistic admissions through a public records request? They did and, in due course, secured a dataset in many ways superior to the one Groseclose had initially sought (e.g., it included all applicants, not just a sample). UCLA provided Groseclose and Sander with a dataset containing roughly twenty variables on every applicant to UCLA during the three years before holistic admissions began (2004–2006) and the first three years of its implementation (2007–2009).
The admissions data files did indeed show that UCLA had restored explicit racial preferences with the introduction of holistic admissions in 2007. Blacks were admitted under standards different not only from those applied to whites and Asians but even from those applied to Hispanics. But the actual mechanism of discrimination was surprising. The holistic process helped black and Hispanic applicants because, other things being equal, students who went to weaker public schools or had less affluent parents received stronger evaluations. It helped them further in ways that our data could not capture—perhaps through the stories of disadvantage told in their essays or from the holistic readers’ subjective evaluations of letters of recommendation. All told, blacks and Hispanics with a particular academic index were twice as likely to get a holistic score strong enough for admission as were whites and Asians with similar academic indices.
But as it turned out, the holistic system had been oversold. The preholistic system had already been capturing the same sorts of “soft” factors—sympathetic personal histories, family hardship, and so on—that conferred racial dividends in the holistic system. The special readers UCLA hired to do the holistic scoring were not—at least to any measureable degree—“cheating” in favor of black applicants. As at Berkeley, holistic admissions by itself would barely budge UCLA’s admissions rate for black applicants. This was not an acceptable outcome for those running the admissions system. They had been urged in the strongest possible terms to produce more black freshmen.
We cannot say exactly what happened next. We can say two things with certainty. First, blacks with middling or poor holistic scores were admitted at rates far higher than those of any other racial group. Second, the particular black applicants admitted tended to be those who had attended UCLA outreach sessions.
Here is our speculation about what happened. UCLA admission officers found the results of the holistic method disappointing. The holistic process generated some racial dividend, but no more than the preholistic system and certainly not enough to produce the dramatic increase in black freshmen they wanted. So as they reviewed files in particular holistic ranges, they set aside those of blacks and, to a lesser extent, Hispanics that they remembered from outreach sessions. These students would be admitted despite their modest holistic scores.
The upshot was that over a quarter of all blacks admitted to UCLA during the holistic period had holistic scores that were ordinarily disqualifying. The same was true for only about 4 percent of East Asian admits and about 6 percent of the white admits. Of the total increase in black enrollment produced by all of the extraordinary efforts launched in 2006, it appears that perhaps one-quarter came from recruiting a stronger applicant pool, one-quarter came from more effectively wooing those who were accepted, and one-half came from explicit racial preferences. These last tended, of course, to be applicants with particularly weak academic preparation and, thus, a particularly acute vulnerability to mismatch.
UC San Diego (UCSD) is generally regarded as the new “star” in the University of California firmament. Over the past thirty years it has built extraordinarily strong science departments, and in recent years it has often eclipsed UCLA in national or worldwide rankings. It has also become, in admissions, by far the most selective UC campus after Berkeley and UCLA. In racial matters, however, UCSD was seen as the campus least concerned with diversity. Even before Prop 209, UCSD used much smaller racial preferences than did UCLA or Berkeley, and after Prop 209’s implementation it had notoriously admitted a class (the freshman matriculants of 1998) in which blacks had virtually the same objective qualifications (and subsequently the same performance) as white students. Characteristically, USCD got no credit for this achievement; it was instead criticized for matriculating only fifty black freshmen. And throughout the decade after Prop 209’s implementation, central administrators often regarded UCSD with vague suspicion on racial matters.
The fuse was thus set for ignition in the winter of 2010, when a group of white and Asian fraternity students at UCSD had, off campus, a social event billed as the “Compton cookout.” The students had apparently held a series of outdoor picnics featuring and making low-key fun of a variety of cuisines. “Compton” referred to a working-class black neighborhood in Los Angeles that had gained some notoriety during the 1992 Los Angeles riots, and the “Compton cookout” featured both authentic and stereotypical black dishes. The event was both sophomoric and offensive, and when word of it reached campus groups, the reaction was one of predictable outrage and condemnation. Matters were further inflamed when a student discovered a noose lying on a shelf in the university library; although it turned out that the noose had been put there by an Hispanic student trying to make a political statement and seeking to convey her own feeling of vulnerability, this fact was widely ignored, and the two events were conflated into proof that racism was widespread and out of control on the UCSD campus.
These events precipitated a reaction among the university-wide leadership similar in nature though much larger in scale and consequence to the reaction at UCLA when the number of black freshmen fell below one hundred. Almost immediately there was talk of requiring UCSD to adopt the same holistic admission system used by UCLA and Berkeley. This broadened into the idea that all the UC undergraduate programs (which, since the opening of UC Merced in 2004, now numbered nine) should adopt holistic admissions—a proposal that UC authorities formalized and adopted in the winter of 2011.
Almost at the same time the university decided to expand Eligibility in the Local Context so that the top 9 percent (rather than the top 4 percent) of students at each public high school would be eligible for UC admission. This meant that the university could no longer guarantee enrollment to other students who placed in the top eighth of high school seniors statewide, and it meant that, for many students, the single-most promising path to UC was to attend a public school with academically weak classmates.
The university had the full support of the UC regents in taking these steps. The personnel of the regents had completely turned over since the mid-1990s heyday of race neutrality, and most of the regents wanted the university to do more—not less—to promote black and Hispanic enrollment. Influential state legislators regularly pressed university officials to be more aggressive in pursuing diversity, and three times the state assembly and state senate passed bills that directed the university to take race into account in UC admissions. (All three bills were vetoed by Governors Schwarzenegger and Brown as in conflict with the California Constitution, as modified by Prop 209.) UCOP leaders knew that they would face minimal resistance and perhaps receive rewards for efforts to use any devices available to increase black and Hispanic representation.
Meanwhile, back at UC San Diego, administrators had determined that an effective way to inject more racial diversity into the school was to lower entrance requirements for transfer students. By the 2011–2012 academic year it was possible to transfer to UCSD with a B average in community college courses. The racial dividend was modest—most beneficiaries of the easier transfer policy were white and Asian—and the lowering of academic standards was disturbing to many faculty. In the fall of 2011 one of us met with a distinguished professor who had been dragooned onto the admissions committee. She was not pleased with the logic of a policy that substantially increased the number of academically weak students at UCSD while having only a small impact on the school’s overall diversity. She shook her head and sighed, “One question is whether racial preferences are a good thing. Another question entirely is whether we can get rid of racial preferences without self-destructing as an institution.”
At this writing, the University of California system is still, formally, race-neutral, but in practice it has come very close to a form of racial proportionality. The university still requires all freshman admittees to be high school graduates and to have completed a standard “college-prep” high school curriculum (e.g., at least three years in high school of English and math). When one examines the pool of California public high school graduates who have completed the core curriculum, Hispanics and blacks are as likely as whites to go on to a UC campus, even though the disparity in test scores and grades remains quite large.
The quest for racial diversity without (in general) the explicit use of race means that the university scoops up many academically weak students in the hope that a disproportionate fraction of them will be black or Hispanic. As these devices move from the margins to the center of campus admissions policies, we are starting to observe an even wider academic gulf between the most-prepared and least-prepared students than existed before Prop 209. And because most of the work-around devices are formally race-neutral, they also bring into the system many whites and Asians with weak academic preparation. Students of all races still tend to be sorted across the UC campuses in a way more closely related to their level of academic preparation than was the case before Prop 209. But gaps in academic preparation within campuses are large and resurgent. Mismatch is likely to grow and to be more multiracial now than it was before Prop 209.
We should thus expect in coming years to see at least a partial reversal of the academic improvements for black and Hispanic students that were so notable in the immediate post-209 period. It is difficult at this writing to tell what has happened, partly because the most dramatic steps aimed at producing racial dividends have just been implemented and partly because the university has clamped down even further on its release of information. For example, at UCLA Law School the initial classes admitted to the school’s Critical Race Studies program had quite poor bar results, comparable to or even worse than those received by blacks and Hispanics in the pre-209 era. But as of 2006 the school stopped making available detailed information on the school’s bar passage outcomes even to its own faculty. The school’s total bar passage rate plunged in 2009—clearly in significant part because of the school’s covert diversity efforts—but a report analyzing the decline was bowdlerized before being released to the faculty. Boalt Law School experienced a similar, precipitous decline in bar passage rates as it reintroduced covert racial preferences. Both schools have stonewalled public records requests aimed at making it possible to study carefully the link between large preferences and poor bar outcomes.
All of this tells us that neither voters nor state officials can end university racial preferences by a single stroke. Like the ancient Hydra of Greek myth, two heads are likely to grow in place of the original.
The University of California is by no means a special case in this regard. In Michigan, voters approved a 2006 referendum that was virtually identical to Prop 209. Mary Sue Coleman, the president of the University of Michigan, announced the next day that the university would not be deterred in its quest for diversity. In the months that followed, university officials spoke of using a new system, called “Descriptor Plus” (developed by the College Board) as a way of preserving the school’s diversity. The Descriptor Plus system used techniques similar to those common in the marketing business to divide neighborhoods and high schools into clusters with very similar characteristics. The College Board reported the detailed demography of each cluster along with the neighborhood and school clusters to which each student taking the SAT belonged. A school wanting to admit blacks, for example, could simply assign extra weight to students from black neighborhood or high school clusters.
After a typically lengthy negotiation, we obtained from the University of Michigan admissions data on freshman applicants for 2006, before the passage of Prop 2, and for 2008, after Prop 2 had been implemented. The results were surprising. The university made some use of the Descriptor Plus categories, but not much. So far as we could tell from the data—and in our opinion, the data was strong enough to sustain a suit against the university—admissions officers were simply continuing to use unadulterated racial preferences. Descriptor Plus might be a race substitute, but for the most part it was a screen for business as usual.
Proposition 209 had three major effects upon the University of California. First, it prompted a rise in applications and yield rates that was particularly remarkable and impressive among black and Hispanic students. Second, the cascading of black and Hispanic students led to campuses where their credentials more closely matched those of white and Asian students lowered their numbers at the elite campuses but led to improved academic outcomes and, most notably, higher graduation rates. But third, university officials generally ignored the good news and instead introduced a series of measures—many of them probably legal but some clearly not—that sought either to produce racial dividends in enrollment or to use racial preferences covertly.
We can now appreciate some of the complexities of the situation that higher education has gotten itself into. Racial preferences produce perverse consequences—not only mismatch but also stigma and many other problems we have discussed. However, a simple abolition of racial preferences produces perverse consequences as well: In particular, it produces evasive maneuvers by universities that can accentuate the mismatch problem. If a school uses large racial preferences to achieve a student body that is 10 percent black, a moment’s reflection makes it obvious that any indirect method of achieving the same goal will require still larger preferences. A school that is determined to offset a ban on racial preferences fully through racial substitutes will merely aggravate the mismatch problem.
Universities are generally reluctant to make available information on admissions and student outcomes, no doubt in large part because of the large size of racial preferences and the poor outcomes of many students admitted with those preferences. When we add a legal ban on preferences into the mix, the incentives for concealing information grow. Although we and our colleagues did eventually obtain enough data from the University of California to demonstrate many of the findings discussed in the past few chapters, the datasets made available to us were heavily censored. We can see enough to verify that mismatch is a serious problem at the University of California that was somewhat ameliorated by Prop 209, but the data lacks sufficient detail to answer many important questions, such as the point at which preferences are small enough to be academically benign.
Any reader capable of being persuaded by evidence should recognize by this point in our story that the mismatch problem is real. But diagnosing its exact dimensions requires better data, and it is not clear that a simple legal ban on preferences is either a necessary or workable solution. We now turn to an exploration of some of the institutional dynamics—in higher education and among those who are supposed to provide oversight and accountability for higher education—to dig more deeply into the problems.
