The 9th Circuit, in a 2-1 decision, has struck down California's Proposition 8, which had eliminated the right of same-sex couples to marry. This is, of course, a great day for equal rights and liberty in the United States. But the particular way the court went about its ruling is a little strange -- intellectually speaking, if not legally or politically.
The first thing that needs to be emphasized about this opinion is that it did not hold that all laws barring same-sex marriage are unconstitutional. Rather, it held specifically that the decision of California voters to take away the previously vested right for same-sex couples (and only same-sex couples) to marry is unconstitutional. That California briefly legalized same-sex marriage, only to later have it stripped by referendum, distinguishes it from most states (where gay marriage has never been legal at any point in time).
This leads to weirdness part one: California spent many years prohibiting gay marriage, then briefly (through judicial decision) legalized it, then outlawed it again. But how can it be that what was legal for most of the past two decades becomes illegal based on a brief, half-year interlude?
As a matter of law, this actually isn't that difficult: the holding of the California Supreme Court established that the ban on gay marriage violated the state's constitutional guarantees -- in other words, establishing a right to gay marriage. As a matter of formal law, Proposition 8 acted to strip a particular class of citizens -- and only that class -- of rights it had previously possessed under the Constitution. That this move might be considered an expression of particularly malevolent animus, in a way qualitatively different from simply refraining from establishing gay marriage in the first place, is not hard to grasp.
The problem is that this seems to misapprehend the meaning of Prop. 8 as it reflects upon constitutional interpretation. The California Supreme Court's decision in Marriage Cases was that gay marriage was protected under various provisions of the state constitution (equal protection, fundamental rights, etc.). They held, in essence, that equal protection requires recognition of gay marriage; fundamental rights requires recognition of gay marriage. But when Californians went to the polls in 2008, they did not conceive of themselves as deciding whether or not to craft exceptions to these constitutional guarantees (equal protection, except that gay marriage can be prohibited). They thought of themselves as answering whether they believed the California courts original interpretation of these clauses was correct (does equal protection require gay marriage?). It was meant to be a signal as to their belief, not that the constitution was wrong to protect gay marriage, but that the California Supreme Court was wrong to believe that it ever did in the first place.
Now, it may be that as a matter of law this has no bearing -- judicial pronouncements of "what the law is" are authoritative, and any subsequent democratic revision or modification proceeds from there. So in that sense, the 9th Circuit got it right -- the California constitution gave gays and lesbians the right to marry, and then the people of California took it away. But it is notable what this outlook says about popular constitutionalism: it gives so little credence to popular understandings of constitutional meaning that even when they immediately reverse a judicial decision via an amendment that constitutionalized a prior doctrinal understanding of the original clause, that is coded as "changing the constitution" rather than "counterinterpretation of the constitution".
Anyway, the 9th Circuit proceeds to argue that the fact that this was California taking away a right, rather than declining to extend one, makes this case indistinguishable from Romer v. Evans. Romer was perhaps the Supreme Court's first high-profile gay rights opinion, striking down a Colorado constitutional amendment which forbade the state or localities from enacting anti-discrimination laws on basis of sexual orientation (and only that category). The law had been passed in reaction to the passage of such laws in a few more liberal Colorado cities (e.g., Boulder), and the Court found that it was effectively motivated by bare animus against gays.
There are few things that make Romer notable. First, the case does not stand for the proposition that states must have anti-discrimination protections on basis of sexual orientation. Rather, the Court's holding was that the state cannot arbitrarily pluck out a particular class of citizens and make it more difficult for them to garner the protections that others enjoyed. Second, Romer applied rational basis review, which is traditionally quite easy to meet. But Romer implies that the decision to take away a previously vested right is more suspicious (even under rational basis review) than simply not extending a right. It keys the analysis to the stripping of the right, and it is often harder to articulate a rational (non-prejudiced -- Romer and other cases establish that "animus" is not a rational basis) reason for taking the affirmative step of removing a right than it is to explain why one never bothered to grant it all. Third, Romer was a Kennedy-authored opinion, and the 9th Circuit decision is aimed like a laser at Justice Kennedy in an effort to hold his vote (or even avoid a cert grant -- I might have to dissent from Orin Kerr in calling Supreme Court review inevitable given the narrowness of this particular ruling).
In keeping with the Kennedy-focus, the 9th Circuit styled its opinion as a clean extension of Romer. But it did so through oddity #2 -- the narrowness of what Proposition 8 did. All Prop. 8 did was prevent gay couples from calling their relationships "marriage", which the court accurately described as a (quite significant) dignitary harm on gays and lesbians. By contrast, the rational bases typically offered for heterosexist marriage restrictions usually are some chatter about the state's interests in procreation and child-rearing. I think those arguments tend to be irrational gibberish anyway, but for the purposes of this case, the court simply observed that Proposition 8 had precisely no bearing on these topics. California already has well-established procedures for how gays and lesbians can conceive and raise their kids, and Proposition 8 effected those not at all. So since the only thing Prop. 8 did do was foist upon gay couples this dignitary harm, and since simply denigrating the legitimacy of gay people is not a legitimate state interest (falling under the ambit of animus), the law falls under Romer. The implication is that if California voters had done more -- for example, hinging certain childcare rights on marriage in tandem with restricting marriage to heterosexual couples -- it would have been more likely to survive rational basis review (or at the very least, force the Court to make the affirmative constitutional case for gay marriage nationwide).
Again, the court's analysis isn't necessarily wrong here, so much as illuminates an oddity in how broader constitutional doctrines manifest. Judicial supremacy means that we can't even conceive of a constitutional amendment that is meant to correct a "wrong" constitutional interpretation -- there is, in this view, no such thing as a "wrong" constitutional interpretation except when the courts themselves say so. Rational basis means that legislatures need to more aggressively target the rights of unpopular minorities, because if they don't, it is easier to say they acted out of animus.
I should stress that I think that the constitutional case for gay marriage is more or less a slam dunk, and in that sense the 9th Circuit opinion definitely reaches the right outcome. But in its (tactically quite wise) efforts to fit itself within the narrowest box possible, it helps illuminate certain strange elements of our constitutional order which I felt compelled to point out.