I’ve long been a proponent of affirmative action based on economic disadvantage and have argued that a U.S. Supreme Court decision curtailing racial preferences in Fisher v. University of Texas – due any day now – could lead colleges to switch from race-based to class-based preferences in admissions. My reading of the evidence, outlined in A Better Affirmative Action, is that most universities become interested in class-based affirmative action only when they are banned from using race and begin using economic status in admissions to indirectly promote racial and ethnic diversity.

Two recent pieces – a front-page story by Richard Pérez-Peña in the New York Times, and a major article by Columbia University President Lee Bollinger in Slate – take up this precise argument. Bollinger’s article is subtitled: “The Supreme Court should not force universities to trade affirmative action for socioeconomic diversity. Schools can have both.” Meanwhile, Pérez-Peña raises the opposite possibility: that a Supreme Court decision could lead colleges to pursue neither, because class-based affirmative action is relatively expensive.

Bollinger argues that class- and race-based diversity are not “mutually exclusive goals…the right course is to pursue both.” He cites Columbia as an example and does not cite any other. But careful research by William Bowen of Princeton University finds that while selective universities give a substantial boost to under-represented minorities in admissions, they give no preference whatsoever on average to low-income students. Places like Amherst College have made real progress on socioeconomic diversity under former president Anthony Marx, but most other institutions didn’t follow suit. At the most selective 193 colleges and universities, according to Georgetown University’s Anthony Carnevale and Jeff Strohl’s 2010 report, Rewarding Strivers, 70 percent of students come from the richest quarter of the population, and just 5 percent from the lowest quarter.

If Bollinger argues that universities already care about socioeconomic diversity, Pérez-Peña’s piece raises the concern voiced by educators that even with a negative Supreme Court decision on race, universities might not aggressively pursue affirmative action based on socioeconomic status. The article opens: “Opponents of race-based affirmative action in college admissions urge that colleges use a different tool to encourage diversity: giving a leg up to poor students. But many educators see real limits to how eager colleges are to enroll more poor students, no matter how qualified – and the reason is money.”

California, which Pérez-Peña profiled in an earlier front-page New York Times article, has taken substantial steps to promote economic diversity in the face of a bar on racial preferences, making U.C. Berkeley and UCLA among the most economically diverse selective colleges in the country. But the University of Michigan, which is also banned by voter initiative from using race, is less socioeconomically diverse and has not, like the UC system, banned legacy preferences for the children of alumni, Pérez-Peña notes. At U.C. Berkeley, 34 percent of students are eligible for Pell Grants, while at the University of Michigan, the figure is 16 percent.

The article correctly points out that the U.C. system, which has had since 1996 to develop alternative programs, has done more to promote socioeconomic diversity, than Michigan, whose ban on affirmative action came 10 years later. Just as universities vary in their commitment to racial diversity, so they are not uniform in their commitment to socioeconomic diversity.

But the article clearly shows that even at Michigan, the ban on racial preferences had a clear effect on the institution’s socioeconomic diversity policies. As Pérez-Peña reports, “many” of Michigan’s programs for economically disadvantaged communities “have existed only since Michigan voters banned affirmative action.” Moreover, while Michigan has a ways to go on socioeconomic diversity, “the proportion of students at Michigan from low-income families has grown significantly in the last decade” and is far greater than at competitors, like the University of Virginia, where race continues to be used in admissions.

Pérez-Peña is right on the mark when he notes that class-based affirmative action programs are more expensive than race-based programs, which often are aimed at fairly advantaged students of color. That’s precisely why Bollinger’s call for universities to consider race alongside socioeconomic status rings hollow: universities, which have to constantly decide whether to build new dormitories, or raise faculty salaries, or increase financial aid, almost always give short shrift to low-income students, who lack political power.

When racial preferences are removed from a university’s tool kit, however, the political equation changes. Universities, which are rightly embarrassed when they lack racial diversity, begin seeking socioeconomic diversity as an indirect means to boosting minority enrollment. And interesting political coalitions arise, as in Texas, where rural white Republican legislators have formed an alliance with urban black and Latino Democrats, to promote admissions policies that benefit low-income and working-class students from both groups.

Of out the ashes of race-based preferences has arisen a smarter, stronger, more politically and legally sustainable affirmative action for disadvantaged students of all races.