State bans on the use of affirmative-action preferences by public colleges have resulted in significant declines in black, Hispanic, and Native American enrollments in graduate programs, with some fields hit much harder than others, concludes the first study to broadly examine the impact of such prohibitions on graduate education.
Moreover, the only race-conscious admissions policies allowed elsewhere—"holistic” admissions policies that purport to consider race and ethnicity as part of broad, subjective evaluation of applicants—cannot necessarily be trusted to promote diversity in graduate admissions, suggests a second study, of an unnamed medical school.
Taken together the studies, both presented here Friday at the annual conference of the Association for the Study of Higher Education, offer grim news to those seeking to enlarge the share of advanced degrees earned by black, Hispanic, and Native American people, especially in fields such as science, technology, engineering, and medicine.
Hispanic people account for about 16 percent of the nation’s total population but only about 6 percent of all graduate students. Black people account for about 12 percent of the population and 11 percent of total graduate enrollment. Both subsets of the population account for only about 4 percent of all graduate students in science and engineering, and they complete graduate programs at disproportionately low rates, prompting the federal government, college associations, philanthropies, and other key players in higher education to pump money into efforts to diversify graduate education in certain key fields and try to get more minority students through the process of earning advanced degrees.
The U.S. Supreme Court, in its 2003 Grutter v. Bollinger and Gratz v. Bollinger rulings involving the University of Michigan at Ann Arbor, has made clear that the only race-conscious admissions policies it regards as constitutional are those that consider race or ethnicity as part of a subjective, holistic evaluation of individual students. And in six states that account for a large share of the nation’s black and Hispanic populations—Arizona, California, Florida, Michigan, Nebraska, and Washington—public colleges have been barred from considering applicants’ race or ethnicity at all.
The study of the impact of bans on affirmative-action preferences found that, in the states examined, they produced a one-percentage-point decline in overall black, Hispanic, and Native American enrollment in the affected public graduate schools, and a two percentage-point drop in some key fields such as the natural sciences and engineering. Considering that black and Hispanic students each account for just 4 percent of enrollment in graduate science and engineering programs around the nation, the study suggests that the widespread adoption of such bans would have a devastating impact on minority enrollment in certain fields.
Even the loss of a few hundred prospective minority students in certain graduate fields “can make a meaningful difference,” the study’s author, Liliana M. Garces, an instructor and doctoral candidate in education at Harvard University, said Friday in presenting her findings.
Although the study of holistic admissions to medical school examined just one institution, it clearly establishes that the rhetoric surrounding such policies’ attention to diversity does not always translate into actual diversity in the classes admitted.
“The holistic review process is not really understood by everyone who is involved,” said Jeffrey F. Milem, a professor of higher education at the University of Arizona and one of the study’s authors. Moreover, Mr. Milem, who is involved in an effort by the Association of American Medical Colleges to develop and promote effective holistic admissions policies, said the problems the medical school studied had in applying holistic admissions “are problems that we hear happening at other medical schools around the country.”
Surveying Fields
The study of race-conscious admissions bans presented by Ms. Garces is the first to broadly examine their impact outside law and medical schools, where they have been found to lead to similar minority-enrollment declines. Ms. Garces attributes the lack of research on the bans’ impact on other fields to the specialized nature of graduate education, in which admissions criteria differ sharply by discipline, and to a lack of research consistently tracking graduate enrollment in certain fields.
Ms. Garces based her analysis on data from an annual national survey of graduate-school enrollment sponsored by the Council of Graduate Schools and the Graduate Record Examinations Board. She examined graduate enrollments at public universities in four states where such institutions have been barred from considering race in admissions for at least some period of time: California and Washington, where the change came through ballot initiatives that remain in effect; Florida, where public colleges have race-blind undergraduate and graduate admissions as a result of actions taken by state officials; and Texas, where public colleges were barred from using such admissions preferences under a 1996 federal court decision that the U.S. Supreme Court effectively overturned with its 2003 Grutter ruling upholding such policies as constitutional if narrowly tailored to promote diversity that yields educational benefits.
Ms. Garces excluded from her study historically black colleges and institutions from which data were missing. To help ensure that the enrollment changes she documents were attributable solely to bans on affirmative-action preferences, she took into account enrollment numbers in a comparison group of 17 states with demographic characteristics, labor markets, levels of educational attainment, and graduate-program offerings comparable to those of the states that were her primary focus. She excluded from her comparison group Southern states that were under desegregation agreements that would have influenced enrollments, and Michigan, where, during the period studied, race-conscious admissions were under so much attack that minority students might have shied away from the public university’s graduate programs.
To account for the possibility that the departure of minority students from states with bans might have artificially inflated minority enrollments in graduate programs in states without them, Ms. Garces also conducted analyses comparing only states that were a significant distance from one another.
Along with finding that bans on race-conscious admissions produced a one percentage-point drop in black, Hispanic, and Native American students’ share of enrollments at all public graduate schools in the states studied, Ms. Garces found that such bans produced declines of two percentage points in graduate programs in the natural sciences, engineering, the social sciences, and education. The bans led to a one percentage-point drop in their share of enrollment in the humanities, and no impact at all on their share of enrollment in graduate business schools.
In translating her findings into actual enrollment numbers, Ms. Garces estimated that a 2-percentage-point drop in black, Hispanic, and Native American enrollments amounted to nearly 210 fewer entering the natural sciences, nearly 100 fewer entering engineering, about 70 fewer entering the social sciences, and about 130 fewer entering education at public graduate schools in the four states studied in a single year.
In presenting her findings Friday, Ms. Garces argued that the graduate-enrollment declines she charted might have been much steeper had public universities not taken steps to try to blunt the impact of the bans, such as stepping up recruitment efforts or adopting as part of their admissions criteria rough proxies for race, such as being part of the first generation of one’s family to attend college.
Her paper says minority students may be hesitant to apply to graduate programs in states that ban race-conscious admissions policies because they are less certain of their admissions prospects or fear that they will be less welcome at such institutions. In discussing her findings, she said she was open to the possibility that some minority students are attracted to universities without race-conscious admissions policies because they believe that attending such institutions will not expose them to the risk of being stigmatized as less-qualified than their classmates.
The Hole in Holistic
The study of holistic admissions at a medical school was conducted by Mr. Milem along with Manuel S. Gonzalez Canche and Farah Sutton, both graduate students at the University of Arizona, and Brendan Cantwell, a postdoctoral associate at the University of Georgia. They agreed to identify the school they examined only as “Western medical school” as a condition for obtaining the information they needed.
The medical school they examined had adopted holistic review, applying it for the first time to students entering in 2009, to promote diversity after a period without any policy explicitly considering the race and ethnicity of applicants.
The four researchers examined all students who applied to the medical school from 2005 through 2009, drawing their data from the American Medical College Application System, a centralized system, operated by the Association of American Medical Colleges, through which all applications to allopathic medical schools in the United States are processed. The data included Medical College Admission Test scores, previous educational transcripts, and self-reported data on students’ sex, race and ethnicity, and parental education. In studying the medical school’s policies and practices in depth, the researchers reviewed its holistic review directives and observed how admissions officials applied the policy.
Upon examining the students admitted under holistic review, the researchers found that the policy had failed to bring about nearly all of the changes in admissions practices that the rhetoric surrounding it suggested it would produce. Neither MCAT scores nor undergraduate science grades had any less weight in predicting whether students would be admitted, and students who were female, black, Hispanic, Native American, from less-privileged family backgrounds, or graduates of that state’s high schools appeared no more likely to gain admission than they had been before.
In fact, both the socioeconomic diversity of the entering class and the share of students who had expressed a commitment to serving the state’s rural communities—a trait the medical school hoped to see associated with attending an in-state high school—had actually declined. About the only factor that appeared to have more weight under holistic review were the ratings assigned to students based on their personal interviews, but the interviewers did not appear to be working from the same page, and the judgments they made did not produce a more diverse class.
In discussing the study’s findings, Mr. Milem said the culture of the medical school was not one that seemed especially open to giving weight to applicants’ race, and the study’s results might have been different if, for example, the medical school examined was one in which holistic review had replaced some other explicitly race-conscious admission policy.
The researchers’ paper summarizing their findings cautions that their study only examined one medical school over a short period of time, and it might have produced different results if they had examined the school over a longer period, giving admissions officers more time to learn how to apply the holistic review policy, or if more medical schools had been included in their analysis.
“We need to do more studies to see if holistic review is effective in different contexts,” Mr. Cantwell said Friday.
