Thoughts on the Supreme Court case about race-conscious admissions ??

I was rereading the original district court case against Harvard around race-conscious admissions, and I was reminded that the lawsuit had two components (= two different claims): Discrimination vs Affirmative Action, often conflated by both sides in the suit and by the media.

The broader legal issue raised by the plaintiffs was that affirmative action and race conscious admissions practices in general are, they argued, unconstitutional (Counts 2-6).

But there was also a narrower factual issue. According to the plaintiffs, Harvard intentionally discriminates against Asian Americans in practice (Count 1). This claim did not and does not implicate affirmative action.

Legally it mattered that Harvard's evaluation policy is facially race neutral, meaning: on its face, their policy does not discriminate against Asian Americans.

(A counterfactual example of a facially non-race-neutral practice would be: We can accept only X number of this group per year, or Y percent of the class has to be from that group.)

So for Count 1, in determining whether Harvard actually discriminates against Asian American in practice (this was not a claim about affirmative action), it came down to a statistical analysis about whether discrimination was occurring in practice in this facially neutral process. The plaintiffs had to prove not just that discrimination against Asian Americans was happening, but that it was happening intentionally.

The data did show that Asian Americans are held to a higher standard with respect to the academic qualifications compared to other groups of applicants.

But the academic metric is only one among several that Harvard uses, and there were other metrics, especially the "personal" and the "overall" categories in Harvard's holistic evaluation process, that were in play here. Harvard has never claimed — and no selective university in the United States has ever claimed — that the admissions outcome is based on academics alone.

The squishy categories in the holistic process leave a lot of wiggle room for schools to engineer their incoming class in ways that reflect their institutional priorities, which include ones that aren't just past academic performance.

In the battle of expert testimony and competing regression analyses, Harvard argued that the plaintiff's expert had failed to show that discrimination was taking place (because of flawed methodology), let alone that there was intent to discriminate.

Specifically, the experts (economists) disagreed over whether to include in their models a variable for the “personal” ratings assigned to applicants and to what extent, if any, the personal rating incorporates race at all or bias against Asian Americans specifically.

The district court rejected both of the plaintiff's claims, including the factual claim that Harvard discriminates against Asian Americans. The court decided the evidence was "inconclusive" and did not prove intentional discrimination or what accounts for the difference.

In its requested remedies, the plaintiff also asked the district court to strike down affirmative action, which they didn't even contend was the injury against Asian-Americans.

And they expressly disclaimed any interest in dismantling Harvard's legacy preferences, which also benefit whites over Asian Americans.

In the appellate review upholding district court ruling, it was decided that "while being Asian American was negatively correlated with the personal rating, it was more likely that “factors external to Harvard . . . like personal essays and recommendations" were the cause of the correlation, "rather than any racial animus." (That to me is the weakest part of the court's finding, because reading essays and writing recommendations could also involve unconscious bias, although I doubt anyone would argue it's with intentional animus.)

Here's where the technical aspects come into play: the appellate court also decided that the lower court had not made an error when it used Harvard's statistical model with the personal rating as a variable, instead of the plaintiff's expert's alternative model.

This is where judges are expected to be statisticians, which is also fraught, because that's not their area of expertise. “Without the personal rating, the model would suffer from omitted variable bias," they concluded.

None of these factual issues are supposed to be relitigated or reevaluated at the Supreme Court level; it is supposed to come down entirely to the legal issues, not the fact finding. I'll be curious to see how much they do draw on the factual findings from the district court, which were on Harvard's side and upheld by the appellate court (meaning, the appellate court found no error).

Will the Supremes ignore the factual finding that no discrimination was proven and therefore no injury? We'll have to wait until next week to find out.

In the meantime, I'd like to provide a bigger-picture admissions perspective:

As a practical matter, all admissions decisions at selective institutions involve discrimination. You let some in, and not others, and when your acceptance rate is in the low digits, that's a lot of people who aren't getting in.

The reality is that a school like Harvard turns away plenty of applicants with perfect GPAs and fantastic test scores. They break a lot of hearts every year.

It's important to point out that most institutions are NOT selective —?they accept the majority of their applicants — so we're really talking about the practices at universities that affect a small percentage of students in higher education in America.

In many ways this lawsuit comes down to our (differing) societal assumptions about who "deserves" to attend one of those highly selective institutions, who "belongs" there, and who is "entitled" to the cultural capital that accrues. That's where the fault lines are.