Judicial Sunsets and Affirmative Action

As I discussed last month, both the majority opinion and (to an even greater extent) the separate opinion by Justice Thomas in Northwest Austin Municipal Util. Dist. No. 1 v. Holder raise interesting jurisprudential questions for originalists about how to explain why it's permissible for the application of a principle to change as social attitudes change but it's impermissible for the principle itself to change with social attitudes. Here I want to put aside the more abstract jurisprudential questions to focus on the doctrinal nitty-gritty itself. My contention will be that "sunsetting jurisprudence"--judge-made legal principles that expire by their own terms--are actually quite common. I'll then use that point to debunk a silly but surprisingly common misconception.

Recall that in NAMUDN1 the constitutional issue was whether the pre-clearance requirement of the Voting Rights Act was a valid exercise of Congressional power to enforce the 15th Amendment. The Court ducked the question through statutory interpretation but strongly hinted that the answer would be either "not anymore" or "not for much longer." Justice Thomas would have reached the constitutional question and would have said "not anymore." The core point is largely an empirical one: Circumstances in 1965 warranted a presumption that changes in voting procedures in covered jurisdictions were efforts to disenfranchise African Americans (and perhaps other racial minorities), but absent further evidence, circumstances in 2009 no longer warrant that presumption.

Here are a few other examples of laws that could be valid at time T1 but invalid at time T2:

a) Under the common law in force at the Founding, death was the penalty for rape and other felonies (although in fact, it was rarely imposed against white men). By the time the Supreme Court decided Coker v. Geogia in 1977, only Georgia authorized capital punishment for that crime. Thus, under both the "evolving standards" approach the Court has taken to the Eighth Amendment and under the latter's literal prohibition of "unusual" punishments, we can say that a practice that was once valid became invalid.

b) Under the Supreme Court's Miller test, whether material is obscene depends in part upon whether the average person, applying contemporary community standards, deems the material at issue to appeal to the prurient interest. Thus, some material that would have been deemed obscene in California in 1973, when Miller was decided, would likely be deemed non-obscene today. The application of the same obscenity law under the same constitutional test to the same picture would have been valid in 1973 but invalid in 2009. (Miller upheld an obscenity prosecution for the distribution of brochures that contained "pictures and drawings very explicitly depicting men and women in groups of two or more engaging in a variety of sexual activities, with genitals often prominently displayed." That sounds like much contemporary advertising!)

c) In Grutter v. Bollinger, Justice O'Connor, writing for the Court, expressed her expectation that affirmative action programs of the sort upheld in that case would "no longer be necessary." Of course, if such a program were unnecessary, it would not be "narrowly tailored," as required by the legal standard applied in Grutter, and thus would be unconstitutional. Accordingly, the same admissions program that was upheld in 2003 would be invalid in 2028.

Grutter is not unusual in allowing for the possibility that the application of a constitutional principle can change with changed circumstances. But it is quite unusual for its invocation of a specfic sunset date. Writing in Monday's NY Times, Ross Douthat had this to say:
It was a characteristic O’Connor move: unmoored from any high constitutional principle but not without a certain political shrewdness. In a nation that aspires to colorblindness, her opinion acknowledged, affirmative action can only be justified if it comes with a statute of limitations. Allowing reverse discrimination in the wake of segregation is one thing. Discriminating in the name of diversity indefinitely is quite another.
Douthat is right that the 25-year figure was made up but wrong in two further respects. First, Justice O'Connor does not appear to have been making any sort of political calculation. Rather, she noted that Grutter got to the Supreme Court 25 years after Regents of Univ. of Calif. v. Bakke, and so she projected forward another 25 years. There really doesn't appear to have been anything more calculated than that.

Second, and more importantly, Douthat's larger argument is surely wrong. He says that given current demographic trends, by 2028 there will be no single racial majority. So far so good. He then goes on to say that at that point, programs of race-based affirmative action will no longer be morally justified or politically acceptable as the majority disadvantaging itself to compensate a wronged minority, but simple racial spoils. That hardly follows as a matter of logic.

Indeed, we might think that societies in which a racial minority disproportionately holds wealth and political power are precisely the ones that need to take the clearest affirmative steps to disestablish the plantation arrangements. I have in mind here South Africa especially where, even after the end of apartheid, the black majority was dramatically underrepresented in many elite institutions in the country. Minority status may explain why some racial groups need special judicial solicitude: They are disadvantaged in the political process. But even a majority can be disadvantaged in other respects.

None of that is to say that race-based affirmative action will necessarily survive past (or until) 2028, in large part because Douthat's views--are widely shared if not entirely sensible. It is significant that affirmative action bans have been adopted in three generally blue states: California, Washington, and Michigan. But that's a point about the political shelf-life of affirmative action, not its moral standing.

Posted by Mike Dorf