Could There Be Diversity
If Affirmative Action Ends?

Would there be diversity after affirmative action?

That's the pressing question for college admissions observers as the U.S. Supreme Court prepares to hear arguments Tuesday on challenges to race preferences at the University of Michigan.

The court's conservative majority has touched off expectations of a decision curbing preferences -- and prompted new analysis of what the outcomes of such a decision might be.

A study expected to be released Monday concludes that black and Hispanic enrollment would drop to 4% from 12% at the nation's 146 most selective colleges -- constituting the top two tiers in a Barron's guide to colleges -- if affirmative action is eliminated and admission is based on grades and test scores. The study is from the Century Foundation, a liberal public-policy think tank in New York.

Anticipating an adverse ruling on rigid racial preferences -- Michigan awards undergraduate applicants 20 extra points in a 150-point admissions formula simply for being in a minority group -- the study's authors propose an alternative based on socioeconomic disadvantages. That is but one in a variety of alternatives, including everything from lotteries to geographic quotas, that are being examined as ways to avert an expected steep decline in minority enrollment.

The Century Foundation study, by Anthony Carnevale, vice president of the Educational Testing Service, which designs the SAT college-admission test, and Stephen Rose, of survey-researcher Macro International, also evaluates the racial impacts of a variety of possible admissions policies. It comes out in favor of "much more vigorous" preferences for economically disadvantaged students.

The report says black and Hispanic enrollment would fall only slightly, to 10%, if affirmative action were replaced by preference for students of low socioeconomic status, as measured by such factors as parental income and education and% of students qualifying for free or reduced-price lunches at the applicant's high school.

LOSING DIVERSITY? Blacks and Hispanics make up 12% of enrollment at elite colleges. A study by testing experts looked at how that percentage would change if affirmative action ended and 'race-neutral' policies went into effect: • Admission based on grades, test scores only: 4%   • Automatic admission for the top 10% in class rank: 12%   • Preference to low-income students: 10%   • Selection by lottery of all students with minimum 900 SAT score: 9%   Source: Socioeconomic Status, Race/Ethnicity, and Selective College Admissions, by Anthony P. Carnevale and Stephen J. Rose March 2003

After analyzing federal education data, Mr. Carnevale and Mr. Rose concluded that low-income students are scarcer than minorities at the 146 elite colleges. Although many selective colleges claim to give an edge to low-income applicants, the researchers found wealthy students with the same grades and test scores are slightly more likely to be admitted. Students in the bottom quarter of U.S. society constitute only 3% of enrollment at these colleges -- compared with 74% for the top quarter. Low-income students "are hugely under-represented at selective colleges, and many more of them could be admitted and succeed," the authors write.

Low-socioeconomic preferences have already been adopted in some places. In California, where voters struck down affirmative action in 1996, the University of California system factors in economic status and "life challenges," such as family illness, immigration hardships and living in a high-crime neighborhood. The criteria have helped boost Hispanic and black enrollment. But they have rankled some Asian-Americans, who feel their disadvantages aren't given as much weight.

The criteria also have been criticized for encouraging applicants to exaggerate any disadvantage. Meanwhile, a number of public school districts, including San Francisco, have moved to compensate for affirmative-action bans by considering socioeconomic factors in admissions to selective schools known as exam schools. The economic preferences would probably hold up against legal challenges, unless they are tailored to favor one race over another.

Besides California, state universities in Texas and Florida also have been banned from practicing affirmative action. The three states have sought to compensate by admitting the top tier of students based on class rank alone. Texas, for instance, automatically admits the top 10% of students from all of its high schools to the state univerity of their choice.

This class-rank approach -- endorsed by the Bush administration -- makes the best students at heavily minority high schools eligible for admission regardless of their SAT scores. In some cases, such plans have minimized declines in minority enrollment. At the same time, they have been criticized by liberals as inadequate, and by conservatives as affirmative action in disguise. Also, they are unlikely to be equally effective for graduate schools, private colleges and schools in less segregated states.

The question of whether minority status or poverty is the more fundamental disadvantage in U.S. society has long divided the liberal community, Mr. Carnevale notes in an interview. Some advocates of race preference, he says, have accused supporters of economic preference of pandering to public opinion. "Economic preference is a lot easier to be for at the polls," says Mr. Carnevale, who favors a combination of economic and race preference if the court upholds the latter. The dispute "has stopped us from moving forward."

The economic-disadvantage preference appeals to some conservative thinkers who have no use for affirmative action. Edward Blum, a senior fellow at the Center for Equal Opportunity, which has submitted a brief to the Supreme Court calling for ending racial preference, says he favors setting aside from one-tenth to one-fourth of seats at elite colleges for economically disadvantaged applicants whose SAT scores are no more than 15% below average for incoming freshmen.

Write to Daniel Golden at dan.golden@wsj.com²

Updated March 31, 2003