New study assesses potential impact of eliminating affirmative action
A new study published in the Journal of Labor Economics which attempts to estimate the potential impact of eliminating affirmative action has found there could be striking implications for minority enrollment, particularly at the nation’s most selective schools.
The study, published by California State University, Sacramento professor of economics Jessica Howell, found that if race-neutral college admissions were to be implemented nationwide in college admissions (as has already happened in California, Texas, Florida, Michigan and Washington state), black and Hispanic enrollment would decline two to four percent nationwide, and 10 percent at schools ranked as most selective in admissions.
Howell’s research finds that 30 percent of 2006 applicants to four year colleges were minorities, adding that other evidence has shown that more selective colleges, schools which fall within the top 20 percent of SAT distribution, tend to factor race more heavily in admissions. Howell cites a study published by Ohio State University professor Thomas J. Kane, which finds “the preference given to black applicants to be equivalent to 400 points on the SAT or two-thirds of a GPA point.”
Howell draws on data from the National Education Longitudinal Study, a survey first conducted in 1988 where eighth-graders were asked their preferences on school and work and followed up with in 1990 and 1992. The study, conducted by the National Center for Education Statistics, part of the U.S. Department of Education Institute of Education Sciences, found that 56 percent of high school seniors apply to at least one four-year college, and that 32 percent apply to multiple four-year schools. The study found that 51 percent of college applicants were female, 69.5 percent where white, 9.5 percent were black, 11.9 percent were Hispanic, 8.2 percent were Asian, and one percent was Native American. In the study, 70 percent of schools to which respondents applied were public, and 73 percent of the colleges respondents ultimately chose were public. Additionally, 64.2 percent of students applied to schools ranked less or somewhat selective, while 18.1 percent applied to schools ranked very selective, and 16.8 percent applied to the most selective schools, according to Howell.
The NELS study also found that males are 4.9 percent less likely than females to apply to college at all, and that black students were 15.1 percent less likely to submit no four-year college applications than white respondents.
The NELS data finds that the average white applicant would likely apply to a four-year college if he or she had an SAT score of about 800 points, before the SAT had 2400 total points, and that the average black student would apply with a median SAT of 787. The study also finds that students likeliest to apply to four-year schools came from high schools with the highest proportion of students in college preparatory classes.
Howell’s research does find some institutional preference for minority students at schools ranked most selective.
“Black and Hispanic applicants are 23 and 9.2 percentage points, respectively, more likely than observationally equivalent white applicants to be offered admission,” Howell wrote. Further, her research states that being a minority is statistically insignificant in admissions at all schools except ones ranked most selective.
Howell’s study concludes that her simulations find that affirmative action has a positive and statistically significant effect on minority admission to four-year colleges, most pronouncedly at the most selective universities. Therefore, she finds, “an affirmative action ban is likely to have the largest effects on minority applicants at a relatively small subset of selective four-year colleges.”
The study, however, finds that most minority students would not change their application decisions.
Approximately 95 percent of minorities would submit the same number of applications even with an affirmative action ban, and 92 percent would apply to colleges of the same level of selectivity, according to Howell’s regressions.
Ultimately, Howell’s study finds that an affirmative action ban would not bode well for minority students applying to college, but also would not dramatically alter the demographics of colleges and universities.
2.5 percent less minority students would receive multiple admissions, and 1.8 percent less would receive no admissions, according to the study. Admissions of minorities would drop by 0.6 percent at all colleges, although they would decline by 10.2 percent at the most selective schools.
Unsurprisingly, experts on both sides of the affirmative action debate at the national level chose to interpret the study according to their take on the issue.
Richard Kahlenberg, an education fellow at The Century Foundation, a non-profit Washington think tank which specializes in monitoring public policy, said he believes college and universities should switch to giving preference to students of lesser socioeconomic status, rather than giving a boost to minority students.
“I think the study confirms that today universities are providing a preference based on race and that the most selective universities have larger preferences in place, and therefore, if you eliminated them, you’d see the most dramatic impact at the most selective colleges and universities,” he said.
Kahlenberg added that he feels that even if affirmative action were to be eliminated, colleges and universities would seek an alternate route to bolster diversity.
“I don’t think it’s realistic to assume that universities will simply drop race-based affirmative action without trying something different as an alternative, so in the places where the use of race has been banned such as Calif., Fla., Mich., and Wash., in each case, universities have responded by trying to find alternative paths to creating racial diversity,” said Kahlenberg.
Kahlenberg suggested giving benefits to students who have faced socioeconomic adversity.
“Using socioeconomic status would buffer that decline,” he said of minority enrollment, were racial preference to be eliminated. “I wouldn’t say it would offset it 100 percent, but it will stem the decrease.”
Roger Clegg, president and general counsel at the anti-affirmative action, pro-colorblind public policy, Center for Equal Opportunity in Falls Church, Va., said he believes the study proves that ending affirmative action would have a limited impact.
“What struck me was that the effect of eliminating racial and ethnic preferences would actually be rather minor,” he said. “These are relatively minor impacts, and when you consider all the costs of engaging in racial and ethnic discrimination in college admissions, it’s just not worth it, you’re sacrificing the principle of equal opportunity and equal treatment without regard to race.”
Donna Stern, the national coordinator at By Any Means Necessary (BAMN), a pro-affirmative action coalition formed in response to the 1996 Proposition 209 ruling in California ending affirmative action at the University of California system, said she believes the study’s findings are an underestimate, citing statistics from the University of Michigan system’s enrollment after Michigan’s Gratz v. Bollinger decision.
“I think that’s an underestimation,” she said of Howell’s findings. “The drop was not 10 percent, the drop was about 50 percent for literature, science and arts undergrads” at the University of Michigan.
In 2005, minority students accounted for 15.1 percent of literature, science and arts students at the University of Michigan, while in September 2009, they accounted for 9.7 percent. In 2005, minority students represented 8.6 percent of engineering students. Last year, they were 5.8 percent, according to the Office of the Registrar at the University of Michigan.
Enrollment numbers at the University of California system, however, do not indicate much of a drop-off in minority enrollment, and in fact reflect gains across all ethnic groups.
According to a September, 2008 study by professor Charles L. Geshekter at California State University, Chico, enrollment increased across all racial categories between 1998, after Proposition 209, and 2006.
The study, titled “The Effects of Proposition 209 on California: Higher Education, Employment and Contracting,” finds that black students enrolled jumped from 1,368 in 1998 to 2,326 in 2006, a 55 percent increase. In that interim, Latino enrollment leapt from 1,348 across the system to 2,743, an 80 percent increase, while Chicano representation soared from 4,155 to 8,453, an 83 percent increase. Asian enrollment, often cited as a factor in California’s Proposition 209 ruling, also increased heavily, as Asians went from 10,427 of the UC’s students to 17,035, a 60 percent increase. In this interim, white students also made gains, as white student enrollment increased from 15,201 in 1998 to 22,471 in 2006, a 33 percent increase.
Ward Connerly, a former member of the University of California Board of Regents who participated in the Proposition 209 debate and opposes affirmative action, said he believes simulation numbers are not reflective of what happens in reality, and thus that he did not accept the study’s findings.
“I’m not very enamored with the study,” Connerly, the chairman of the American Civil Rights Institute, a Sacramento-based organization which works to end racial and gender preference in public education, said. “These simulation models are not very effective when you apply them to matters of student behavior and how universities will react.”
Connerly said that when the UC Board of Regents debated ending affirmative action, it ran multiple models, few of which ultimately yielded accurate results.
“You just can’t predict, number one, how people are going to react once a certain policy is changed, and more impotently you can’t predict how the global university community will react once they’re all on the same footing.”
Connerly said he sees the fact that private schools can still offer racial preference in states that have elected to ban the practice as a negative, as students make their application decisions based on where they feel they will be admitted.
“Right now, you can make the case that when Calif., Fla., Wash., Neb., and Mich. eliminate preferences, those students who would have gotten into selective institutions, they might apply to historically black colleges or private colleges, but if everyone’s on the same footing and as an applicant you know you can’t go and get a better deal at Stanford than UC Berkeley, your decisions are different.”
Connerly did, however, concede that an achievement gap exists between white and Asian students and black and Hispanic students.
“If we understand the fact that there is a profound academic achievement gap between black and Hispanic students on the one hand, and Asian and white students, and if we know that that gap is likely to continue as trends suggest, and if institutions of higher education place emphasis on those academic standards, then that drop in the number of underrepresented students of two percent to four percent, that might be true,” he said.
What might such a hypothetical change mean for the University of Massachusetts?
22 percent of incoming freshmen for the fall of 2009 were African, Latino, Asian, Pacific Islander and Native American (ALANA) students, accounting for roughly 907 students out of this year’s incoming class of 4,124, according to the Office of Institutional Research’s undergraduate admissions and enrollment statistics survey.
Were enrollment figures to drop at UMass along the same lines as Howell’s study indicates, this would mean that minority enrollment would drop in the incoming class by between 25 and 30 students.
Sam Butterfield can be reached at email@example.com.