Online Fisher symposium: Race-neutral nonsense

The following submission to our Fisher symposium comes from Michael Rosman. Mr. Rosman serves as General Counsel at The Center for Individual Rights and authored an amicus brief in support of the petitioner in Fisher v. University of Texas.

The Court states that, as a part of the narrow-tailoring inquiry that is the second step in strict scrutiny analysis applicable under the Equal Protection Clause of the Fourteenth Amendment to racially disparate treatment by state entities, courts should look at whether the defendant gave serious consideration to race-neutral alternatives before adopting the race-conscious program at issue.  It is obviously a crucial part of Fisher v. University of Texas because the petitioner argues that the University of Texas had already achieved sufficient “racial” diversity (if not other kinds) through the use of its Top Ten Percent Plan.  That plan, adopted by the Texas legislature in the wake of the Fifth Circuit’s ruling in Hopwood v. Texas that diversity was not a compelling governmental interest, and that the University of Texas School of Law had no compelling interest at all to justify its use of race in its admissions process, provided that the top ten percent of each high school class in Texas was guaranteed admission to the University.  Because many of Texas’s high schools were themselves racially identifiable (that is, had student populations with one predominant racial group), this actually led to a significant number of minorities being admitted to the University each year and a not unreasonable amount of racial diversity.

Unfortunately, the “race-neutral alternative” jurisprudence is hopelessly bollixed.  It would be nice if the Court could use Fisher to straighten it out, or, better yet, jettison it altogether, but I personally do not have high hopes.

To understand the reason why this jurisprudence is so confused, one need only consider the fact that its primary exponent has been now-retired Justice Sandra Day O’Connor.  To understand the confusion itself requires a bit more study.

Justice O’Connor first articulated the idea of a “race-neutral alternative” in 1989 in City of Richmond v. J.A. Croson Co..  That case involved a set-aside program, The Richmond Plan, for city construction contracts, a program that the Court found violated the Fourteenth Amendment’s guarantee of equal protection.  Here is the discussion of “race-neutral alternatives”:

As noted by the court below, it is almost impossible to assess whether the Richmond Plan is narrowly tailored to remedy prior discrimination since it is not linked to identified discrimination in any way.  We limit ourselves to two observations.

First, there does not appear to have been any consideration of the use of race-neutral means to increase minority business participation in city contracting. . . .  Many of the barriers to minority participation in the construction industry relied upon by the city to justify a racial classification appear to be race neutral.  If [minority-owned businesses] disproportionately lack capital or cannot meet bonding requirements, a race-neutral program of city financing for small firms would, a fortiori, lead to greater minority participation.

Justice O’Connor – speaking for the Court in this particular part of her opinion – may be saying one of two different things here.  One is that the barriers to minority participation (like capital or bonding requirements) are race neutral; that is, that the problem the City should have been examining is race neutral.  On this view, the paucity of minority-owned firms contracting with Richmond was not caused by past racial discrimination, but rather by race-neutral requirements that disproportionately affected such firms.  This interpretation might fit in with the first sentence – that it is difficult to determine narrow tailoring in the absence of identified discrimination.

If that is the correct view, then Justice O’Connor is presumably saying nothing more than that state actors should examine the prerequisites to contracting, hiring, or anything else to determine whether they (1) have a disproportionate impact and (2) are truly justified requirements.  This would create an incentive structure not unlike the one that employers subject to possible disparate impact lawsuits under Title VII face.  Indeed, one would think that employers (or state contractors) should eliminate gratuitous and unnecessary prerequisites regardless of whether they have a disparate impact, even if there’s no legal obligation to do so.  By definition, unnecessary prerequisites eliminate qualified applicants.

But an equally plausible interpretation of the opinion – and probably the better one given the full quotation in context – is that Richmond should have considered whether the elimination of capital or bonding requirements would increase minority participation in construction contracts regardless of whether it had a good reason for those requirements.  That is, the elimination of such requirements would not be for the purpose of expanding the pool of truly qualified applicants of all races, but for the express purpose of having more minority applicants regardless of whether they were “qualified” under the criteria that Richmond believed best.  Under this view, the problem is insufficient numbers of minority-owned firms getting city construction contracts, which may have been caused by past discrimination, but the solution is to eliminate some of the requirements needed to qualify for those contracts.

The first – and maybe the last – question to ask about this interpretation is why eliminating valid prerequisites for the purpose of changing the racial mix of the applicants would be better than outright preferences based upon race?  A good case, it strikes me, can be made for the proposition that it is not, and may be worse.  Making direct race preferences, after all, at least provides you with those minority applicants who best meet the criteria we assume (for purposes of this discussion) to be valid.  Eliminating a valid criterion expands the pool by including those who fail to meet that criterion; thus the minority applicants may not even be the best minority applicants.  In the case of something like bonding requirements, it may also increase the pool by including less-qualified non-minorities.  One has to choose from this now-expanded pool of applicants, and, by hypothesis, one cannot use the criterion (bonding requirements or whatever) that one deems best.

To put the point more clearly, suppose Richmond had decided to include singing ability as a criteria for determining which companies would get construction contracts because it believed that minority-owned businesses would have better singers and it wanted to increase the number of minority-owned businesses getting contracts.  Would that really be better than a direct racial preference?  Would it really have been race-neutral?

So, too, with the Top Ten Percent Plan.  It is widely believed that the Texas legislature did not adopt that plan because it decided that class rank alone, without benefit of any score from a standardized test, was the best means of assessing who would be the best students at the University; rather, it did so to increase the eligible minorities.  Indeed, Texas’s brief to the Supreme Court takes this position.  It asserts that “increas[ing] minority admissions” was “[a]n acknowledged purpose of the law,” but that the increase came “at significant cost to educational objectives.”

Notably, the Court – again through Justice O’Connor – at least hinted at this analysis in Grutter v. Bollinger in response to the argument by the United States (as an amicus in support of plaintiff) that “percentage plans” like those used in Texas were a race-neutral alternative to the race-conscious admissions plan employed by the University of Michigan School of Law:  in a passing reference to this argument, the Court observed that “even assuming such plans are race-neutral” – at least intimating that it thought they might not be.  And Justice O’Connor’s opinion went out of its way to reject other so-called “race-neutral” plans, like a lottery or lower admissions standards, on the ground that they would force the University to sacrifice its stated criteria for academic excellence.  (On the other hand, the Court also said that, as part of its narrow-tailoring obligation, colleges and universities “can and should draw on the most promising aspects of . . . race-neutral alternatives [in states like California, Florida, and Washington where the consideration of race was prohibited by state law] as they develop.”)

So, in Grutter, the Court hints that perhaps the first interpretation of Croson is the correct one: the only “race-neutral” alternatives that really need to be “considered” are the elimination of unnecessary criteria that are blocking the truly qualified members of all races from competing, without any consideration of the racial consequences.  But if that’s the case, what does this have to do with “narrow tailoring” under strict scrutiny?  One does not get to the narrow-tailoring prong, after all, unless the state has a compelling interest to use race in the first place.  Let’s say the compelling interest is the present effects of identified past discrimination.  If there are gratuitous criteria eliminating qualified applicants for a government program, how is eliminating those criteria a remedy for that past discrimination?  After all, the government agency running the program should eliminate gratuitous criteria anyway, since they interfere with the most efficient function of the program.

How will all of this play out in Fisher?  I have no idea.  It is a most bizarre case for that purpose.  Fisher’s best argument may be that a facially neutral plan adopted for racial reasons, questionably constitutional itself, precludes the explicit use of race.  In essence, she contends that the (not so) surreptitious use of race precludes the overt use of race.  And the university’s best argument may be that Fisher’s “race-neutral alternative” (the Top Ten Percent Plan) is not really a race-neutral alternative at all, and thus (as suggested in Grutter) cannot be considered such under narrow tailoring.