« John Marshall Harlan (I) and Giles Hotchkiss on Same Sex Marriage Issues
Andrew Hyman
| Main | Randy Barnett on Michael Dorf on Originalism
Michael Ramsey »


Obergefell, Craig, Nguyen, and Somin (Updated)
Chris Green

Just below, Andrew makes an interesting argument about the relationship of the Obergefell sex-discrimination argument to Giles Hotchkiss's comments at the very end of the February 1866 Bingham Amendment discussion. Hotchkiss hoped that Bingham's proposal would be replaced with a provision "to provide that no State shall discriminate between its citizens and give one class of citizens greater rights than it confers upon another," or a provision that "no State shall discriminate against any class of its citizens," or a ban on "discrimination to the injury or exclusion of any class of citizens from the privileges which other classes enjoy." I think Hotchkiss's hope, expressed in terms of privileges and citizenship, was clearly fulfilled in the Privileges or Immunities Clause, not the Equal Protection Clause. Andrew's basic point about the way class legislation was understood in 1866 is still relevant, however, mutatis mutandis.

For my part, I wanted to make a few points about the relationship of the sex-discrimination argument to current doctrine. 

Chief Justice Roberts received (see here at 61-64) two replies from Michigan's lawyer John Bursch to the sex-discrimination argument for same-sex marriage long made by academics like Akhil Amar and Andy Koppelman. 

First, unlike cases applying intermediate scrutiny since Craig v. Boren in 1976, the gender distinction in traditional marriage definitions is symmetric.

Second, in 2001's Nguyen v. INS, Justice Kennedy held for the Court that the preference for children of married American parents or unmarried American mothers over children of unmarried American fathers--particularly the requirement that the unmarried American fathers commit to child support--satisfied intermediate scrutiny because encouraging paternal bonding and support was deemed "important," even "exceedingly persuasive."

Ilya Somin, who made the argument as an amicus alongside Koppelman and others, provides some follow-up here. Some follow-up of my own after the jump.

1. Intermediate Scrutiny for Symmetric Gender Classifications?

The classic reply to the symmetric-discrimination point, which Somin duly makes (like Koppelman here), is that the segregation and racial-endogamy laws struck down in Brown and Loving were likewise symmetric. It is not at all obvious, though, that we can draw a strict analogy of the form (symmetric racial distinctions) : (racial exclusion) :: (symmetric gender distinctions) : (gender exclusion).

Observers from Chief Justice Taney in 1857 to Charles Sumner and his many Republican allies in 1872 to Justice Harlan in 1896 to Brown in 1954 have recognized that racial segregation typically connotes racial insult. Few, however, think that gender segregation in bathrooms, athletic teams, or the like typically connotes gender inferiority. This is so even though such segregation is obviously based on stereotypes about athletic ability and heteronormative sensitivity to same-gender intrusions of privacy. Likewise, one might see traditional marriage definitions as the sort of integration requirement that was struck down in Parents Involved, but which poses very different issues when gender, not race, is involved.

Socially-constructed artificial kinds like race are different from chromosomally-constructed natural kinds like gender. This matters first to their likelihood of being relevant to legitimate governmental purposes, which is why we see the difference between strict and intermediate scrutiny. It also matters to the typical social meaning of such distinctions' symmetric use. We thus can get a difference, perhaps, between gender-based segregation/integration requirements and gender-based exclusions. Merely appealing to the repudiation of Pace in Brown, McLaughlin, and Loving does not explain heightened scrutiny in this context.

Moreover, the invention of intermediate scrutiny in Craig was based in part on the very rarity of the sorts of gender discrimination involved there. The sort of discrimination at issue was a "now almost universally rejected tradition," as Justice Stevens put it, while the majority noted regarding alcohol sales to Native Americans, "The repeal of most of these laws signals society's perception of the unfairness and questionable constitutionality of singling out groups to bear the brunt of alcohol regulation." Gender segregation in certain contexts, however, was ubiquitous in 1976 and remains ubiquitous today. If intermediate-scrutiny doctrine were to apply mechanically to symmetric segregation or integration requirements, these statements from Craig make no sense. If the rarity and abandonment of gender-based exclusions suggest more scrutiny, the commonness and persistence of symmetric gender classifications suggest less.

2. Nguyen v. INS.

What about the Bursch's claim that traditional marriage definitions can satisfy intermediate scrutiny? His use of Nguyen illustrated a stark difference between the petitioner- and respondent-side briefs in Obergefell. The 4343 pages of petitioner-side briefs contain only two very short references to Nguyen, both in the reply briefs (here at 12 and here at 20). The Koppelman-Somin sex-discrimination amicus brief neglects it, as do all other petitioner-side amici. The 3244 pages of respondent-side briefs, however, harp on Nguyen's approval of gender distinctions in biologically-based contexts repeatedly (see here at 12, here at 28 n.27, here at 24, here at 5 n.2, here at 32, here at 9 & 12, here at 50, here at 38 & 56, and here at 3, 20 & 30). 

Somin reads Justices Ginsburg, Sotomayor, Kagan, and Nguyen author Kennedy as objecting to Bursch's deployment of Nguyen. The transcript does not read that way to me. Justice Ginsburg certainly engages with the Nguyen argument at pages 62-63 and again at 63-64, but Sotomayor, Kagan, and Kennedy were instead objecting to Bursch's mention of out-of-wedlock birth statistics in the middle of page 64, which Kennedy thought cuts in favor of allowing those in same-sex relationships to adopt such children while being honored as marriages. Kennedy was clearly responding to Bursch's use of these statistics, not his invocation of Nguyen.

Justice Ginsburg's objection to the use of Nguyen stressed that the burden on children of unmarried American fathers was not particularly severe. Unwed fathers could still establish paternity by committing to support their children, for instance. It is not clear to me, however, that the severity of a burden to a plaintiff is relevant in the application of tiers-of-scrutiny doctrine once we have decided what level of scrutiny to apply.

Truth be told, like Justices Marshall, Stevens, Burger, and Scalia, I am skeptical of the entire framework of tiers. I present an originalist argument that race, gender, and age should be subject to the same basic sort of equality analysis under the Privileges or Immunities Clause in chapter 6 of my book. That said, the idea is supposed to be that we first (a) decide what level of scrutiny applies based on what sort of distinction is at work (intermediate for gender), and then (b) decide whether we have enough fit ("substantial relation" for intermediate scrutiny) to a sufficiently-big governmental interest ("important" for intermediate scrutiny). The size of the burden--i.e., just how hard the law makes it for unmarried fathers or their children--is not at play when we apply intermediate scrutiny. Only the size of the governmental interest and the fit are then at issue.

It would be different if we used Marshall's general sliding-scale arbitrariness requirement, which would weigh both the costs and benefits of a classification at the same (single) stage. But then the sex-discrimination argument would have to show arbitrariness in this particular case, rather than relying on an intermediate-scrutiny thumb on the scale. We can, then, abandon tiers-of-scrutiny doctrine and with it the basis for special skepticism of all gender classifications, or we can apply it, in which case Ginsburg's distinction of Nguyen seems to me irrelevant. 

Update: Somin replies here. I'm not sure he appreciates the simplicity of Bursch's proposal that intermediate scrutiny does not apply if a statute does not "involve[] treating classes of men and women differently." That's not an inquiry into motive, as Somin suggests, but merely a proposal that symmetric gender distinctions deserve categorically less scrutiny than gender exclusions. My rationale for lessened scrutiny was the typical lack of a connotation of inferiority in usual cases of segregation, such as in bathrooms or athletics, but I offered this as a reason to adopt Bursch's idea, not a suggestion that we make ill motive an additional threshold requirement before applying intermediate scrutiny. The proposal to exempt symmetric distinctions from intermediate scrutiny would still allow for the Craig result, of course.

Somin adds nothing about the relevance of the frequency of sex distinctions in Craig, which struck down a single state's rule, about the implausibility of applying heightened scrutiny to commonplace instances of gender segregation like bathrooms and athletics, or about the oddness of distinguishing the intermediate-scrutiny analysis of Nguyen by pointing to the relative burden on different plaintiffs.

Somin also misapprehends, I think, the precise relevance of the length of the tradition in favor of opposite-sex definitions of marriage. There was a tradition in favor of coverture rules and the like in favor of the rights of husbands, of course, a tradition repudiated in America beginning with Mississippi's 1839 Married Women's Property Act. This Lawrence-style emergent tradition became unanimous in 1979, when Louisiana repealed its "head and master" rule, and in Kirchberg in 1981, Louisiana conceded that its legislation had been constitutionally compelled. There is, of course, no such emergent consensus, let alone a unanimous one, in favor of same-sex marriage. It is one thing for the Constitution to protect a tradition that spread gradually across the entire country over the course of 140 years, quite another to suggest that the Constitution requires the repudiation of most states' laws in favor of a mini-tradition still only 12 years old and 17 states wide.

Update 2 (5/1): Somin adds more (still here), but, tellingly, I think, saying nothing about the social meaning of gender segregation in general, i.e., in cases like athletics or restrooms. This is important. As I see it, we must distinguish three aspects of heightened-scrutiny doctrine: (a) the rationale for such scrutiny, which includes such matters, among other considerations, how typical a sort of distinction has become, whether there is a national consensus against it, and the like; (b) the judgement of general suspiciousness involved in giving a sort of classification heightened scrutiny; and (c) the arbitrariness of particular applications of the doctrine, e.g., gender-segregated bathrooms or athletics, or the traditional scope of marriage.

As Somin points out, Craig obviously does not make rarity or consensus a factor in applying heightened scrutiny to gender classifications. At the same time, the relative rarity of such classifications is one of the factors motivating the adoption of intermediate scrutiny. It is hard to tell exactly what the precise rationale of Craig is; the Court passes the buck to "earlier cases" for adopting its "important governmental objectives" and "substantially related" standards, though earlier cases did no such thing. It certainly seems to me that the obsolescence of sex distinctions in practice was important to Craig. It called them "old," "archaic," and "outdated." Brennan's Frontiero plurality, advocating strict scrutiny, appealed to congressional "increasing sensitivity to sex-based classifications." Because these descriptions do not apply to typical cases of symmetric gender classifications, like bathrooms or athletics, it seems best to interpret the doctrine not to apply to them. That is not to say that we should interpret the doctrine itself to apply only to rare distinctions; rather, because symmetric distinctions were and are commonplace and uncontroversial, we should interpret doctrine not to apply to them.

I am not, then, claiming simply that in the particular case of opposite-sex-only marriage, the gender classification is not degrading or insulting. Heightened scrutiny in this context only makes sense if gender segregation in general is suspicious, and the lack of suspiciousness to most reasonable observers of of gender-segregated bathrooms and athletics shows that they are not. This is so even though, as I note above, these examples themselves are based on gender role-typing stereotypes like heteronormative increased sensitivity to opposite-sex privacy intrusions. Somin apparently thinks that separate restrooms are in fact suspicious. But I think that most reasonable observers in 1976 or today, even those on the Court, would disagree. 

We could, of course, junk heightened scrutiny based on the general suspiciousness of all gender classifications and merely talk only about the justifiability of the particular distinction at issue in traditional marriage definitions. If I were a justice, that's the door I'd choose. But that would abandon the argument from Craig and its progeny.

Update 3: Somin agrees that under his approach, same-sex bathroom and athletic requirements are indeed suspicious, but might pass intermediate scrutiny. If we regard gender role-typing as pernicious, it's very hard for me to see why that would be.  Separate restrooms tell women to conform to the gender stereotype of being less concerned about privacy intrusions from other women than about privacy intrusions from men, and likewise for men. 

Whether separate bathrooms or athletic teams would be struck down under Somin's interpretation of Craig, however, I think that they're not even suspicious.  If we ask if symmetric gender distinctions are generally suspicious, moreover, the numbers of suspicious or non-suspicious cases within that denominator matters, and restroom counterexamples are everywhere. Examples of innocuous gender-integration requirements abound as well, I think. Given the issues at stake in the relative numbers of men and women attending law school, it would not be suspicious to require that our admissions committee contain both men and women. Gender-integrated decisionmaking in all sorts of areas of life--business, education, government--strikes me as unworthy of a Craig-style raised eyebrow.

Update 4 (5/5): Somin replies again, but my reply is brief enough for Twitter.