On Oct. 19, the Supreme Court granted cert in a case called Association for Education Fairness v. Montgomery County Board of Education. I am surprised it hasn’t received more press coverage because it could be the final nail in affirmative action’s coffin.
Perhaps you think the Supreme Court already hammered in that nail in Students for Fair Admissions, Inc. v. President & Fellows of Harvard College, but you’d be wrong. In that case, the court found racial preferences at Harvard and the University of North Carolina to be illegal. Nevertheless, many scholars are now urging schools to turn to what I call “race neutral affirmative action.” These are programs that seek to increase the number of some racial groups relative to others not by selecting for race itself, but by selecting for characteristics correlated with race. For example, instead of using race preferences in admissions, a university uses preferences for applicants who live in zip codes that are nearly all black, applicants who grew up on Indian reservations, or applicants for whom English is a second language. These are real-life examples of what some universities have done when they were forbidden from using race. Many affirmative action supporters think these sorts of things are perfectly legal.
But I have long argued that race proxies are just as illegal as using race itself. My arguments became moot points after the Supreme Court allowed universities to use race itself in admissions in the 2003 case of Grutter v. Bollinger. But now that SFFA has taken that off the table, the question I teed up is back in the limelight: is it constitutional to use racial proxies instead?
That’s what Montgomery County allegedly did to pick the students for its magnet middle schools. Its efforts predated SFFA because, even before the Supreme Court told universities not to use race, they had told school districts not to. In a nutshell, the alleged facts are these: Asian children make up only 15% of the school system’s population – it is majority black and Hispanic – but were taking up half the spots in the magnet middle schools. This didn’t go over well with many parents, so the school board sought to even things out. The board tried several plans but ultimately used a lottery among the applicants who scored above the 85th percentile in their neighborhood on standardized tests. Because the pool was based on how well applicants did compared only to others in their own neighborhood, the plan spread magnet school spots across the county – and, because many neighborhoods don’t have much racial diversity, it spread the spots across the races, too. The number of black and Latino students went up and the number of Asian students went down.
According to Supreme Court precedent, this is unconstitutional. We have a long history of attempts to discriminate indirectly when we couldn’t discriminate directly – and we have an equally long history of the Supreme Court telling us it’s illegal. For example, when we couldn’t deny the ability to vote on the basis of race, we tried using literacy tests and grandfather clauses and felon disenfranchisement instead. The court said those things are illegal. When we couldn’t segregate schools, we tried drawing school district boundaries around mostly black neighborhoods and mostly white neighborhoods instead. The court said we can’t do that, either.
The doctrine is this: if your “purpose” is to discriminate on the basis of race and your “effect” is to discriminate on the basis of race, then you are engaged in discrimination on the basis of race – even if you aren’t using a racial classification. It’s something like the old adage: if something walks like a duck and quacks like a duck, it’s a duck.
Most scholars supported these precedents when people were discriminating against blacks. But now not so much. Thus, I expect a bevy of amicus briefs from academics contorting their way through them. But the contortions I have seen thus far are very weak. For example, the leading theory on why these precedents don’t doom Montgomery County is the old distinction between “benign” and “invidious” discrimination: the purpose to help is different than the purpose to hurt. The precedents involved the purpose to hurt blacks, whereas Montgomery County was trying to help blacks (and Latinos); hurting Asians was just a byproduct, not the purpose.
If you thought the Supreme Court had already rejected – over and over again – the distinction between “benign” and “invidious” discrimination, you’d be right. After all, that’s how we got SFFA. As the court explained there, life is often zero-sum: if you help one race by giving them more spots, you necessarily hurt other races by giving them fewer spots. What is “benign” for one is “invidious” for the others. It’s a distinction without a difference.
The scholars say the cases rejecting the benign-invidious distinction don’t count because they involved racial classifications, not racial proxies. But the court in SFFA anticipated this move and cut it off: “[U]niversities may not simply establish through . . . other means the regime we hold unlawful today. … [W]hat cannot be done directly cannot be done indirectly. … The Constitution deals with substance, not shadows.” Yes, that’s dicta, but isn’t it very, very probative of what the very same justices will think about the very same question next year?
Moreover, the dicta didn’t come from nowhere: in fact, the Supreme Court has already struck down “benign” racial proxies – a bunch of them. When governments engage in racial gerrymandering of voting districts, they don’t use racial classifications: voters are assigned to one district or another based on where they live, not their skin color. Yet, even when these gerrymanders have been designed to help black and Latino voters, the Supreme Court has struck them down all the same, including just this past term in Louisiana v. Callais. Is not Montgomery County engaged in gerrymandering no less than Louisiana? It’s just school districts, not voting districts.
It is true that the court has required discriminators to try “race-neutral alternatives” before they can use race itself. But the court requires race-neutral alternatives only when discriminators otherwise have a compelling interest to justify their discrimination. All this means is that the court prefers racial proxies to racial classifications when forced to choose; it doesn’t mean racial proxies are insulated from constitutional scrutiny.
Still, the court may not need to get into all this because the question presented in the Montgomery County case merely asks what counts as a discriminatory “effect.” The U.S. Court of Appeals for the 4th Circuit held that the Asian students had to show a “disparate impact” – a term borrowed from Title VII law that means the challenged practice impacts one race more than another – and they couldn’t do so because, even after the gerrymandering, they still comprised a greater portion of the magnet school spots than they did of magnet school applicants.
Although it is true that disparate impact is one way to show a racial effect under the Supreme Court’s purpose-and-effect precedents, as I previously explained, it is not the only way. For example, when the Supreme Court said that school districts couldn’t evade its desegregation decrees by drawing district lines around mostly white neighborhoods and mostly black neighborhoods, it did not require the plaintiffs to show a disparate impact (e.g., schools of different quality); it was enough that one school was overwhelmingly white and the other was overwhelmingly black. When the Supreme Court has said that legislatures couldn’t gerrymander voting districts, it did not require the plaintiffs to show a disparate impact (e.g., differential vote dilution); it was enough that the racial composition of the districts had been affected.
It is true that what exactly the racial effect is in the voting cases is undertheorized. There is no obvious baseline to compare to the gerrymandered district lines in order to measure the “effect.” For this reason, some scholars think the court has all but read out the effect requirement from the purpose-and-effect test. Maybe so. The effect requirement comes from an old case called Palmer v. Thompson: there, Jackson, Mississippi, closed all their public swimming pools rather than desegregate them; the court held that this wasn’t unconstitutional racial discrimination because all the races had been affected equally. The ruling is unpopular among scholars on the left; law professor Randall Kennedy at Harvard has compared it to Korematsu. Thus, I doubt many will mourn its demise; plaintiffs would still have to show an injury to establish standing in federal court. But overruling it will only help the plaintiffs in the Montgomery County case.
Still, a lingering question that hasn’t been asked enough is whether the purpose-and-effect doctrine – or even a purpose-only doctrine if Palmer goes away – is consistent with the original understanding of the Constitution. It is hard for me to see how it couldn’t be: it is very easy to evade any rule by proxy; surely, they knew that even back in 1868. But I haven’t seen any historical analyses yet. I hope there are some in the briefing.
I appreciate that for many scholars it is inconceivable that we could end up in a world where it is illegal to do anything about racial disparities. Many of them endorse so-called “antiracism” and believe that all racial disparities must be the product of racism – and, indeed, that we, too, are racists if we don’t do something about them. This means we can sometimes have “too many” Asians. Maybe they’re right about that, maybe they’re wrong. But the 14th Amendment does not enact Mr. Ibram X. Kendi’s “How to Be an Antiracist.”