You can ban affirmative action, but can you enforce that ban?

The civil rights laws of the 1960s banned racial discrimination in many areas of public life. But they are difficult to enforce. Absent a smoking-gun confession of intent, it is hard to prove that race, specifically, was the reason a person wasn’t offered a job or an apartment. This, in turn, has led to intense battles — in the law and in social science — over what, exactly, should count as proof of discrimination. 

If, as expected, the Supreme Court soon curtails or eliminates the use of race in college admissions, conservatives will learn these lessons anew. Today, many colleges outright admit using race as a “plus factor” in admissions decisions, and that practice will certainly stop. But there are many quieter ways to go about engineering a school’s racial balance. 

As an illustration of where things are headed, take an area where colleges do not outright admit discriminating by race: the allegations that Harvard and other elite schools work to keep the Asian share of their student bodies in check, above and beyond giving preferences to underrepresented groups such as African Americans and Hispanics. There are good reasons to believe this is happening, but it’s hard to prove in court. 

In a new report for the Manhattan Institute, I look at one piece of evidence. As many observers began noting about a decade ago, the Asian share of the overall population has grown markedly in recent decades, and Asian students, on average, perform highly on numerous academic metrics. One would therefore expect the Asian presence at elite schools to grow. And yet at Ivy League schools — where Asian students long have been overrepresented, thanks to their strong academic credentials — the Asian share of the student body plateaued for an extended period of time starting in the mid-1990s. A similar trend played out at non-Ivy but highly selective schools. 

That sure looks like a quota, but there are limits to my analysis. For one thing, the federal data I use is self-reported by the schools and spotty or erratic in some cases, and some students’ racial information is missing. For another, it’s impossible to completely rule out innocent explanations for the patterns, such as that schools changed their admissions policies in legitimate ways that only incidentally happened to disfavor Asian students. Further, while the long plateau is certainly suspicious, these schools’ Asian shares have been rising again in recent years, and there’s no good way to say how high they should go before our suspicions are alleviated. 

More fundamentally, even if we accept that something is happening here, broad admissions trends cannot prove discrimination at a specific school, which is what one would need to do to force a change in policy. 

The lawsuit against Harvard demonstrates the extreme nature of the difficulty. The school disclosed detailed admissions records to expert economists for both sides, and the two analysts produced elaborate statistical analyses purporting to show that Harvard either does or does not discriminate against Asians relative to similarly qualified whites. (Both agreed that other racial groups receive preferences.) When a case relies on such information, judges and jury members, many with limited statistical knowledge, are expected to figure out which side’s models are better. That’s not a recipe for good decision-making, but it’s how the legal system works. 

There are any number of obvious ways that schools might quietly continue practicing affirmative action even if it’s banned. The simplest is merely to shift which factors get the most weight in the admissions process, favoring measures that indirectly create the desired racial balance. 

Officials redesigning the admissions process for Thomas Jefferson High School in Fairfax, for example, made the mistake of explicitly advertising their goal of racial balancing and paid the price in court recently. But the policy changes they made would have been easy enough to defend in themselves. For example, they eliminated a standardized-test requirement and relied more heavily on a “holistic” review. 

As long as a school makes innocent-seeming changes like that, provides facially legitimate reasons for them, and genuinely implements them without regard to race, it may be impossible to prove racial intent. Neither the Constitution nor federal law requires schools to, say, base admissions entirely on test scores, or to completely eliminate subjective criteria such as essays or the “personal” rating that Harvard is accused of using as a cover for anti-Asian discrimination.  

Many colleges are scaling back or eliminating their use of standardized tests. Similarly, since California banned affirmative action in the 1990s, critics have alleged that the practice, although curtailed, didn’t end. The university system retooled its admissions processes with race clearly in mind, for example. In 2014, economist Tim Groseclose, then a UCLA professor, published a book with statistical analyses suggesting the school still gave preferences to African Americans and discriminated against Asians. 

It’s one thing to suspect discrimination, or even to be pretty sure it’s happening. It’s another to prove it. That has been an obstacle to civil rights enforcement for decades, and it will be true for enforcing limits on affirmative action, too — even if the politics proves very different in that context. 

Robert VerBruggen is a fellow at the Manhattan Institute and author of a new report, “Racial Preferences on Campus: Trends in Asian Enrollment at U.S. Colleges.” Follow him on Twitter @RAVerBruggen.