The Chronicle of Higher Education
December 14, 2001
Colleges Should Take 'Confirmative Action' in Admissions
By LANI GUINIER
"A liberal education," Lee C. Bollinger declares, "is largely about crossing sensibilities, about seeing the world through different eyes and realizing that your perceptions may not be accurate." Bollinger, the newly selected president of Columbia University, emphasizes that "people learn more and learn better in an environment where they are part of a mix of people where there are substantial differences, with people not like themselves." And he continues: "Understanding race in America is a powerful metaphor for crossing sensibilities of all kinds."
Bollinger has become a national figure with his aggressive defense of diversity as crucial to the mission of higher education at the University of Michigan at Ann Arbor, where he has been president since 1997. But despite support from a large contingent of college and university presidents, the diversity defense, like affirmative action itself, is under sustained legal assault. The most recent attack came from the U.S. Court of Appeals for the 11th Circuit, which ruled that the University of Georgia's decision to award applicants of color an extra half point in the admissions process was unconstitutional.
Georgia officials chose not to seek review by the Supreme Court, contending that two lawsuits involving admissions decisions at the University of Michigan, due to be reviewed this month by the U.S. Court of Appeals for the Sixth Circuit, would present a stronger defense of racial diversity. Georgia vowed instead to increase recruitment efforts to enlarge the pool of black applicants. That alone will not do much to overcome its own history; the university barred black students for the first 160 years of its existence, until it was forced to integrate in 1961. It now has a student body that is only 5.7 percent black, in a state with a population that is more than 25 percent black.
Beyond that, and more important, Georgia's response that its hands are now otherwise tied in its efforts to diversify sends the troubling message that courts of law alone will decide how we can function in a multiracial society. In fact, the legal fights over affirmative action have deflected public attention from Bollinger's astute juxtaposition of diversity and merit and distracted many colleges from a much needed exploration of what constitutes fairness for all people. What's more, prolonged court battles have obscured the role that diversity plays in helping institutions -- especially public universities -- fulfill their educational and democratic missions.
Indeed, despite the appellate court's focus on Georgia's use of race to benefit students of color, considerations of racial diversity were relevant to the admission of only a tiny fraction of Georgia's applicants. In addition, it is hardly accurate to say that the institution's use of race as a "plus factor" for students of color gave a slight edge to nonwhite applicants. The vast majority of its successful applicants were white students who benefited from criteria, such as aptitude-test scores and legacy status, that correlate closely with race but weakly, if at all, with merit in terms of future educational and professional success. Those criteria overwhelmingly favor middle-class and upper-middle-class white students.
Georgia's difficulty attracting black students -- a result of such pervasive yet almost invisible preferences -- illustrates an increasingly disturbing reality. The 11th circuit and many other courts have failed to understand what Bollinger so forcefully argues: Merit in a public institution of higher education cannot exist independent of considerations of diversity.
What the appellate court, for example, failed to evaluate was the role that diversity should play as an integral part of Georgia's institutional mission governing the selection and education of all students. Diversity is a key element of Georgia's educational mission. According to its mission statement, the University of Georgia "endeavors to prepare the university community and the state for full participation in the global society of the 21st century. ... It seeks to foster the understanding of and respect for cultural differences necessary for an enlightened and educated citizenry."
In a statement released last month, Georgia's president, Michael F. Adams, affirmed that mission. A "mix of students supports the energy and academic environment of a place charged with pushing back the boundaries of knowledge," he said, and helps create a place that "is a powerful force for economic development and discovery for the state."
To produce the "enlightened" citizenry anticipated by its mission statement and the "powerful force" for development and discovery that a mix of students helps create, Georgia needs to move beyond the marginal use of affirmative action and make what I call "confirmative action" central to its entire admissions process. By framing diversity as simply about preferences for persons of color, the current controversy over affirmative action fails to examine how the conventional "merit-based" criteria that we assume to be fair systematically exclude poor and working-class people of every racial group, including whites. Such criteria also fail to predict the most important elements of merit in a multiracial democracy: who will contribute to the society as a whole after graduation.
By contrast, the process of confirmative action ties diversity to the admissions criteria for all students, whatever their race, gender, or ethnic background -- including people of color, working-class whites, and even children of privilege. Linking diversity to merit, it confirms the public character and democratic missions of higher-education institutions. Diversity becomes relevant not only to the college's admissions process but also to its students' educational experiences and to what its graduates contribute as leaders in our democratic polity.
Unlike the medieval institutions of Europe and the American colonies, which trained future clerics and oligarchs, most public colleges and universities in the United States were, in fact, established to develop leaders and active, critical citizens in a participatory democracy. And according to Mindy Kornhaber, an assistant professor of education at Pennsylvania State University at University Park, who has studied the mission statements of several public universities, the training of leaders remains pre-eminent today. It certainly has been key to serving the needs of the state of Georgia, President Adams explained last month. "I don't need to recount the number of current legislative, legal, business, industry, and civic leaders who have been educated here." The training of leaders for public service, Kornhaber concludes, may be the key contribution of American public higher education to all universities -- public and private, here and abroad.
The current definition of merit, however, relies on uniform paper-and-pencil tests that cannot measure the capacity of students to solve complex problems or to lead others in resolving the complicated issues of our day. Rather, such tests measure a student's learned intelligence to make quick, strategic guesses with less than perfect information. Such learned intelligence is an inappropriate admissions criterion in part because it is often based on information or forms of analysis that students have not learned in school. As Richard C. Atkinson, president of the University of California, argues, sorting and ranking by aptitude-test scores, such as the SAT I, fail to assess what is in fact most important to colleges and universities. Instead, the practice sends "a confusing message to students, teachers, and schools" that "students will be tested on material that is unrelated to what they study in their classes. It says that the grades they achieve can be devalued by a test that is not part of their school curriculum."
Research by Howard Gardner, a professor of cognition and education at Harvard University, shows that the ability to work collaboratively and to learn from diverse perspectives constitutes a kind of emotional intelligence that can also be learned and that is highly correlated with career advancement and leadership, but is not measured by numerical scores on aptitude tests. The current timed paper-and-pencil tests do not tell us how someone will function in situations where a range of intelligences -- including personality factors such as drive, motivation, creativity, and problem-solving skills -- are relevant.
Indeed, research by Derek Bok and William G. Bowen in The Shape of the River (Princeton University Press, 1998), and a Harvard study of three classes of its own graduates over a 30-year period, discovered just this point: High test scores that count so much in admissions decisions do not measure attributes of leadership. To the contrary, the Harvard study reported that low SAT scores and a blue-collar background were the two factors that correlated with financial and career satisfaction as well as community leadership.
Timed, standardized aptitude tests also fail to predict academic competence. The SAT's are frequently hailed as the essence of objective merit, yet SAT's correlate less with college students' first-year grades than with either their parents' or their grandparents' socioeconomic status. Similarly, the LSAT nationwide, some say, predicts the variance in first-year law-school grades only 9 percentage points better than what might be produced by a random assortment of scores and grades.
Honing leadership skills and training future problem-solvers for active participation in a multiracial democracy mean that students must be encouraged to ask questions, to seek knowledge from those with whom they disagree, and to participate in open and honest debate. Sadly, standardized tests appear to have the opposite effect. SAT's and other standardized aptitude tests socialize students who do well to believe they "deserve" opportunities to succeed; the tests don't encourage a commitment among those students to give back in any way to the taxpayers who subsidize such opportunities.
In fact, at Bollinger's University of Michigan Law School, a study found that high LSAT's actually correlate with career dissatisfaction, have no bearing on income as a lawyer, and don't necessarily lead to either community service or leadership. Students who scored the highest on the law school's admissions index, which combines undergraduate grade-point averages and LSAT scores, were less likely to be mentors for younger colleagues or serve on community boards once they became lawyers. Underscoring Bollinger's defense of diversity, it was disproportionately the black and Latino students who became leaders as lawyers.
High-stakes testing may have begun as a noble effort to open up opportunity, but excessive dependence on aptitude tests has instead reproduced the very hierarchy that they were originally intended to topple. Reliance on test scores tends to favor the children of the rich over those of working-class families. At the University of California at Berkeley, for instance, nearly 42 percent of the white students entering in 1997 came from families earning more than $100,000 a year.
Similarly, at the University of Texas at Austin, when SAT scores were a dominant admissions criterion, 75 percent of the freshman slots were occupied by students from 10 percent of the state's high schools. Those feeder schools, which represented 150 of the 1,500 high schools in Texas, were the resource-rich suburban public and private institutions serving the upper-middle class of the state's white population. Some counties in rural West Texas, with large poor and working-class white populations, had not sent any graduates recently to the state's flagship public universities.
To get back on track, we need to view "merit" as a functional rather than a generic concept, one that is intimately connected with each institution's specific mission and the democratic purposes of higher education. Reframing how we view merit, a key aspect of confirmative action, would indeed confirm the benefits of our democratic ideals -- but not simply to people of color. Merit would become a forward-looking function of what our democratic society needs and values rather than a fixed, quantifiable, and backward-looking entity that, like one's family tree or inherited assets, can be chronicled with proper instruments. Merit, in other words, would reflect the benefits to the institution's educational mission and learning environment of the diverse experiences of people from all backgrounds, especially those who are committed to contributing back to the larger society and to the taxpayers who subsidized their education in the first place.
The challenge of embedding an institution's mission into admissions should also prompt us to re-emphasize issues of curriculum, learning theory, and ways to motivate all members of a diverse population to reach their full potential while making larger contributions to society. That means that institutions must become more accountable for nurturing diversity in the curriculum or in pedagogy, not just in admissions. If college presidents like Lee Bollinger and Michael Adams, as well as social scientists like Howard Gardner, are right that the ability to learn from and use diverse perspectives is instrumental in constructive problem-solving, then we need to move beyond the complacent "chalk and talk" format so ubiquitous on college campuses. We must reconsider the ways that teachers teach and students learn.
The misleading suggestion that diversity is in tension with merit limits their dual potential to inform and inspire more group-based learning and collaborative teaching -- approaches that better prepare students for what they will do in both the workplace and the public sphere after they graduate. For example, Uri Treisman, a calculus professor at the University of Texas at Austin, developed group-based peer workshops when he was a professor at Berkeley that significantly improved the performance of his black students. After researchers discovered that the Chinese-American students who excelled in his first-year calculus classes were studying in groups, Treisman carefully observed their study habits. He realized that it was the process of intellectual engagement with peers that enabled the students to master the conceptual challenges of calculus, and that all his students would benefit if he altered how he taught calculus to include more group problem-solving.
In the admissions process, a confirmative-action approach suggests, for example, that the University of Georgia could require all students, including those with high SAT scores, to explain why they should be admitted to a public university that values diversity and hopes to train its students to become leaders and critical thinkers. The 11th circuit wrote: "If the goal in creating a diverse student body is to develop a university community where students are exposed to persons of different cultures, outlooks, and experiences, a white applicant in some circumstances may make a greater contribution than a non-white applicant."
Assuming that the court is correct that white applicants contribute to the university's diversity and that Georgia should therefore "assess each applicant as an individual rather than as a member of a particular racial group," then Georgia and other institutions should consider requiring every applicant to complete a diversity essay. That essay would describe how the applicant would contribute to the diversity of an intellectual community that is dedicated to preparing its graduates to function consistently with the institution's democratic mission. In that way, admissions decisions could be tied to evidence of a long-term commitment to public service, leadership, and group problem-solving skills -- especially those involved in overcoming obstacles or working with diverse communities -- and an ability to contribute to a successful multiracial educational experience for other students.
The advantage of a confirmative-action approach is that all students would compete against one another based on their ability to contribute to the university and society rather than on largely arbitrary, biased factors like SAT scores and other proxies for privilege -- such as legacy status, which was also the basis of a preference in the Georgia admissions process. Admissions decisions would not be entrusted to those who design nationwide tests or those who inflexibly administer selection criteria determined by others.
Instead, a structured decision-making process could draw from the wisdom and experience of relevant stakeholders -- including members of local communities, current students, faculty members, administrators, and leaders from the world of business and the public sphere. Those individuals would be encouraged to collaborate to design admissions policies reflecting an institution's mission and consistent with changing local needs.
Such a democratically inspired approach was used in Texas to eliminate the SAT as a requirement. In lieu of SAT scores, the University of Texas now automatically admits any applicant who graduated in the top 10 percent of a Texas high-school class. This is not the only way to gain admission, of course. But it has increased the number of Latino students and maintained the number of black students, compared with those admitted under affirmative action, and created a more economically diverse student body.
Because the adverse impact of the SAT was most obvious in its effect on the admissions prospects of black and Latino citizens, those groups led the fight against the use of aptitude tests. They also discovered that class rank in high school more accurately predicted first-year college grades. Working-class and poor whites joined those efforts once they realized that the tests also had a less visible but equally significant class bias.
The Texas Legislature adopted the plan with the support of several conservative white lawmakers. They realized that changes were necessary to give their own constituents access to a valuable public resource -- taxpayer-subsidized higher education. Gov. George W. Bush signed the plan, which has now been in effect for almost three years. The result? The grade-point average of freshman students admitted under the 10-percent plan is higher, for every racial group, than it was when SAT scores dominated admissions criteria.
By contrast, when Gov. Jeb Bush of Florida adopted his own version of the 10-percent plan, he failed to include the relevant stakeholders in the decision-making process. As a result, many black and Latino community leaders and legislators found fault with the One Florida plan, under which those graduating in the top 20 percent of the state's high-school classes gain admission to "a" public college but not necessarily to one of "the" flagship universities, as in Texas. Animus against the plan is credited with motivating the higher-than-usual black turnout in the 2000 election. Because Jeb Bush neglected black and Latino lawmakers, professors, and community advocates in developing the new admissions protocols, he was much less successful in fulfilling the state's higher-education mission of serving the diverse larger community.
Joining functional merit with institutional missions would go beyond supporting affirmative action for its specific beneficiaries. It would decouple merit from testable mental aptitude for all students and free us to envision many more ways to select students and allocate scarce educational resources in a manner that better comports with democratic values. Institutions throughout the country could use a more structured and participatory decision-making process to experiment and learn from best practices.
Some colleges might abandon the SAT requirement entirely, as Mount Holyoke, Bowdoin, and Bard Colleges already have done. Other institutions, such as large universities with tens of thousands of applications, might consider using the tests only as a floor -- accepting no students with scores below which no one previously succeeded in graduating. Above that floor, applicants could be chosen by several alternatives, including a lottery. Concerns about a lottery's insensitivity to an institution's particular needs or values could be dealt with by increasing the selection prospects of applicants with the skills, abilities, or backgrounds that are valued by the institution.
For example, a weighted lottery could enhance the chances that certain students will be selected by putting their names in two or three times. Depending on the state's or locality's needs, the institution could weight the names of students who come from underrepresented communities or demographic groups, or students who make five-year public-service commitments to work in disadvantaged communities upon graduation. Particularly in the education arena, where opportunity lies at the core of institutions' missions, a lottery -- although arbitrary -- might allow random selection to serve important democratic goals.
Other alternatives include reserving spaces in the sophomore or junior classes at flagship campuses for students who have maintained B averages at community colleges, where admission would be open to all. Or if, as some studies say, the SAT accurately predicts only 17 percent of first-year grades, and a college's mission emphasizes academic excellence among freshmen, that institution might consider limiting the use of the test scores to the admission of a comparable portion of its first-year class.
California, often a bellwether, is considering two proposals that are consistent with the premise of confirmative action. First, Atkinson has argued that his university system, one of the largest in the country, should no longer require applicants to take the main SAT test. If Atkinson's proposal is adopted, the system would rely -- at least temporarily -- on exams that are intended to measure student achievement in specific subjects, rather than general aptitude or reasoning in mathematics, vocabulary, reading, and other areas. Already, as a result of a second proposal, applicants will soon be rated in a dozen categories, including special talents, leadership skills, and accomplishments in the face of "personal challenges," such as economic hardship.
Practicing confirmative action, each institution would, with its specific mission in mind, regularly review and seek feedback on its admissions program. And it would ask several important questions to guide such efforts:
I am not advocating a specific magic-bullet replacement for admissions policies centered on aptitude tests. Nor do I endorse only one selection practice as the way to achieve coherence between what public universities do and what they say they want. Yet if we can move away from the contested terrain of affirmative action, on the margin of the admissions process, we can begin to rethink how colleges and universities admit and educate everyone.
To do so, the democratic challenge of distributing educational opportunity must be returned to the relevant stakeholders rather than administered by a judiciary more concerned with the appearance of formal equality than the reality of functional merit. Although Georgia decided not to appeal the 11th circuit's ruling to the Supreme Court -- a wise choice in the eyes of civil-rights practitioners -- the court's decision gives the university an unusual opportunity to refashion its admissions processes.
It is not enough for Georgia to expand the pool of black applicants. Indeed, Georgia's President Adams is willing to contemplate future changes to the entire admissions process in order to distribute "an important public resource" more fairly, to meet the needs of the individual, the institution, and the state.
That means the university must move beyond its current practices, which admit the vast majority of students based on test-centered criteria that benefit primarily white middle-class students, without obvious relation to the institutional mission. Most important, it should experiment with a range of admissions practices and then assess each practice in terms of its contribution to the mission of the institution -- and the role that the institution itself can play in helping create a genuine multiracial democracy.
Lani Guinier is a professor of law at Harvard University. She is the author of Lift Every Voice (Simon and Schuster, 1998); the co-author, with Susan Sturm, of Who's Qualified? (Beacon Press, 2001); and the co-author, with Gerald Torres, of The Miner's Canary: Enlisting Race, Resisting Power, Transforming Democracy, to be published by Harvard University Press in February.
This article from The Chronicle is available online at this address:
http://chronicle.com/weekly/v48/i16/16b01001.htm