New Salvos on Affirmative Action

With Michigan voters weeks away from a vote on whether to ban affirmative action, critics of the practice are releasing admissions statistics that they say show the extent of the gap between black and white applicants admitted to the University of Michigan.

The data reveal large differences in grades and standardized test scores, and indicate that black applicants are much more likely to be admitted, even with lower grades and test scores. These are the sort of data that have been influential in other states that have considered -- and passed -- statewide bans on affirmative action. "The people of Michigan have a right to know the extent to which discrimination is taking place," said Roger Clegg, president of the Center for Equal Opportunity, which is releasing the data today and planning a series of events in Michigan to publicize the figures.

David Waymire, a spokesman for One United Michigan, which is leading the fight against the referendum, said that the data being released were "worthless" because they did not include breakdowns by economic class. He said that he believed the gaps in scores were largely driven by class, not race and ethnicity, and that this was just "the usual half-assed job" from the Center for Equal Opportunity.

The data came from the University of Michigan, which had to release the figures in response to the center's Freedom of Information Act requests. Among the findings:

• The SAT median for black students admitted to Michigan's main undergraduate college was 1160 in 2005, compared to 1260 for Hispanics, 1350 for whites and 1400 for Asians. High school grade point averages were 3.4 for black applicants, 3.6 for Hispanics, 3.8 for Asians, and 3.9 for whites.
• Black and Hispanic applicants in 2005 with a 1240 SAT and a 3.2 GPA had a 9 in 10 chance of getting in -- while white and Asian applicants with the same scores had a 1 in 10 chance of getting in.
• For undergraduates in the most recent year for which data are available (2004), 28 percent of black students had been on academic probation at some point in their Michigan careers, compared to 23 percent of Hispanic students, 8 percent of Asian students, and 5 percent of white students.
• Similar patterns hold for law and medical school admissions. In the latter, for example, the data indicate that of applicants with an MCAT total of 41 and a GPA of 3.6 in college science courses, admit rates were 74 percent for black applicants, 43 percent for Hispanic applicants, 12 percent for white applicants and 6 percent for Asian applicants.

The debate in the weeks ahead is likely to be over what these numbers mean. To foes of affirmative action, they are the smoking gun about the use of racial preferences in admissions. To the University of Michigan, these are numbers without context or much significance at all (except perhaps politically).

Clegg of the Center for Equal Opportunity said that these data suggest that the university is paying more attention now to race and ethnicity that it was before two landmark decisions by the Supreme Court in 2003. Those decisions -- one about the system used by Michigan to admit undergraduates and one about its law school -- effectively said that colleges could continue to use affirmative action, but couldn't have separate systems in which extra points were awarded across the board specifically for race and ethnicity. Clegg's group was hoping at the time for the court to completely bar affirmative action, but he said that the data show that Michigan is violating the ruling that was handed down.

What the Supreme Court upheld was the use of race in a "limited and nuanced way," he said, which is inconsistent with the wide gaps shown in the data his group is releasing.

Julie Peterson, a spokeswoman for the University of Michigan, released a statement in which she took issue with Clegg's analysis, which she called "flawed and shallow," noting that expert witnesses in the affirmative action cases had found that such comparisons are oversimplified to the point of being misleading.

The center's analysis ignored key factors, she said, such as "the rigor of the student's high school or undergraduate curriculum, extracurricular activities, essays, teacher and counselor recommendations, and socioeconomic status." By ignoring these qualities about applicants, she said, "CEO attempts to reduce human beings to a couple of simplistic numbers. No top university admits students solely on the basis of grades and test scores. We consider many factors in order to admit a group of students who have diverse talents, who are highly motivated and who have the potential to succeed at Michigan and make a contribution to the learning environment."

Peterson noted that after the Supreme Court rulings, the university revised its undergraduate admissions process to gain more information about students. "It is just plain wrong to imply that race somehow carries a greater amount of weight than it has in the past, or than the Supreme Court allowed."

If there was one area on which Peterson and Clegg agreed, it was that the political stakes are high right now for data like the figures being released.

"It is no coincidence that CEO has released this report in the weeks leading up to a ballot proposal that would outlaw public affirmative action in the state of Michigan," Peterson said. "This is a politicized attempt by CEO to narrow the focus of the debate to college admissions at a single institution, rather than acknowledging the broader potential impact on state employment and contracting, K-12 schools and public universities and community colleges, potentially affecting financial aid, outreach, pre-college and other programs that consider race, gender and national origin."

For his part, Clegg said that he hopes the data will persuade Michigan voters to bar affirmative action. If they don't, he said that the data could be helpful to others who may want to sue the university. And if you aren't in Michigan, Clegg said that his group -- which previously did a series of studies like the Michigan one -- is planning another series.