Back to the law stuff.
As we hinted at previously, there are two different analyses in play simultaneously that intersect, because of the complicated message of the interracial marriage case, the delightfully named Virginia v. Loving. The Court did not see the need to comment on how the two issues interacted there, because they both lead to the same result. The Equal Protection Clause is involved, and because the act of classification Virginia engaged in was racial, this triggered strict scrutiny. Virginia would have to show that classifying people by race for purposes of their marriage statute was necessary for achieving a compelling state interest. The state failed to show this, so the statute violates Equal Protection. But it also said that marriage is a fundamental civil right, thus triggering the Due Process Clause.
Brief digression on this. Opponents of gay marriage sometimes say that people forget that marriage is a creature of government, and they are right. That is why, instead of writing “fundamental right,” which might induce people to think of natural rights, I wrote “fundamental civil right.” But that does not mean that a right is any less important for all that. The Bill of Rights is largely concerned with what lawyers call procedural rights (e. g., freedom from double jeopardy). There are no procedural rights where there are no procedures, and so these too are creatures of government, and no less important for all that. And if one argues that these procedural rights are rooted in natural interests (consider, in this light the Eight Amendment right to not be subjected to cruel and unusual punishments), well, obviously the intimate relationships marriage as an institution are founded on express all sorts of natural interests too. Anyway.
Now a fundamental civil right also triggers strict scrutiny, so we end up getting the same result. Virginia can burden that right only if doing so is necessary to achieve a compelling state interest. Virginia loses.
Same-sex marriage has the same constitutional duality. First, it classifies people. Another digression: how exactly does it classify people? There are two ways to look at this. If Virginia v. Loving is the model, it classifies people by sex, by gender, not by sexual orientation. We didn’t talk about this before, but this is an Equal Protection issue as well. It triggers an intermediate form of scrutiny (because basically we’re OK with some sex/gender classifications, so we don’t want to make them all but impossible). Intermediate scrutiny says that the state must show that its action is an important means to a substantial state interest. You might think it odd that a government action which burdens gays and lesbians should be thought of as a form of sex discrimination, but there is precedent afield for this, in a non-constitutional context. There was a case, Oncale v. Sundowner Offshore Services, in which an employee sued his private employer for sexual harassment; the facts were men harassing a man because [correction: they thought that] he was gay. The Court held that because they would never have treated a female employee this way, this was treating someone differently because they* were a man (not orientation discrimination, which under this law, is perfectly fine as such). Delightfully, the author of the opinion was my favorite rigorist, Justice Scalia.
If, instead of gender classification, we see the California marriage statute as classifying (implicitly) by sexual orientation, then we’re in rational basis review. And as we saw, if you think of this one way, the state wins, because we usually let states win with rational basis review: the state can discriminate against people who drive fast, or people who practice medicine without a license, or any number of other things, and we don’t inquire too deeply into its motives. If, however, the discrimination seems to be driven by nothing but animus (against, for example, the mentally challenged) state action can flunk even rational basis review.
So now our ducks are lined up. The two sides in the constitutional question characterize the issue in the following ways (I’m going to help both sides out a bit because they haven’t put their best foot forward, at least not in court).
Statutes that implicitly restrict some generic legal privilege to heterosexuals are subject to rational basis review, and ordinarily in such cases we would not look too closely at the goals and means the state is after in such cases, unless there is animus. But discrimination against homosexuals is driven by nothing but animus, and so we must use rational basis review “with teeth.” (In any case, the Supreme Court has already told us this, in Romer). So what is the state interest? Procreation and child-rearing. But the state has not produced one scintilla of evidence that restricting marriage in this way has any tendency to further anything of interest in this area. But it’s not clear that this is even the right level of scrutiny, because on analogy with the non-constitutional, employment discrimination model, this statute classifies by gender (after all, in Loving v. Virginia, the Court didn’t say that the marriage statute classified by out-race orientation, but by race simpliciter). This heightens the scrutiny, making the state’s case even worse. Though procreation and child-rearing are probably important enough to serve as important state interests, intermediate scrutiny is flunked for the same reason rational basis review is flunked: there is no evidence that the restriction does the slightest thing to further the goal, let alone that it is substantially related to that goal. And this would be the case if we were not talking about, of all things, marriage. But marriage is a fundamental civil right, and this triggers strict scrutiny. Stipulate that procreation and child-rearing are compelling state interests, the restriction still has no evidence supporting it, and so surely can’t be necessary to achieve it. No matter what scrutiny you use, no evidence is no evidence.
Marriage is a creature of government, and homosexuals are not a class which triggers heightened scrutiny, so we are in rational basis review. As a rule, rational basis review defers to the government’s actions both as to means and ends, for the simple reason that in a democracy, the default position of the courts should be to let the people rule. The kind of deference rational basis review involves means that the courts cannot demand evidence, as if the burden was on the people to demonstrate the wisdom of its legislative actions, and only then will judges allow acts to be law. That is judge-ocracy, not dem-ocracy. So the complaint about evidence is entirely misplaced. The burden is not on the people to justify its choices, because we live in a democracy. Challengers’ attempts to heighten scrutiny, furthermore, are unjustified. The heightening that takes us to rational basis review with teeth in this context requires that the state action have literally nothing whatsoever to do with sexuality; there has to be a complete disconnect between means and ends to support the inference of animus. The state action in Romer, which among other things prohibited homosexuals from benefitting from statutes prohibiting private housing discrimination, had nothing to do with homosexuality, but rather with housing discrimination. If homosexuals were able to lobby to ban housing discrimination on the basis of sexual orientation, how on earth might that interfere with the state’s interest in there being no housing discrimination? The two things do not connect up in the right sort of way, and that is why the inference that the state action is nothing but an expression of hostility is well-nigh irresistible. But marriage is all about sexuality, so this argument fails. The state can regulate housing to achieve a housing-related purpose, and it can regulate sexuality to achieve a sexuality-related purpose, and it’s not a court’s job to second-guess this. What it can’t do is purport to regulate one thing in an utterly bizarre way as a mask for trying to hurt a class of people as an end in itself.
The argument that scrutiny gets heightened to intermediate scrutiny because of gender classification fails too, because it is bizarre. Marriage in its very essence is the legal union of a man and woman. By contrast, marriage is not in its very essence the legal union of two people of the same race. This is why no state has ever “prohibited” same-sex marriage because it literally goes without saying. It makes no more sense to say that a state does not permit people of the same sex to marry than it does to say that it does not permit people to marry multiple partners simultaneously, groups, corporations, animals, vegetables, inanimate objects, regions of space, abstract objects, etc. The issue is not that we fear that these things will come about if same-sex marriage is introduced, but if we begin to use the word marriage to refer to something which is not, in its essence, marriage, and this is reflected in law, then marriage simply no longer exists and has been replaced with something else. But surely the decision to abolish marriage is a decision for the people, not for the courts? Lastly, marriage is indeed a fundamental civil right, but not only could we repeat the preceding points even under strict scrutiny, but it is clear from the various ways that we regulate marriage even within its essential nature as a civil union between two people of the opposite sex, this has never been incompatible with its structuring to limit it to parties above a certain age, or of a certain distance in kinship. Surely if restricting marriage to people who are above a certain age and not first cousins would meet strict scrutiny (well, wouldn’t it?) restricting marriage to people who are the opposite sex does as well. Yes?
Those are the issues. As you can see, the strength of the legal arguments on both sides are dependent upon how prior conceptual issues are settled: what it means to be married, and what it means to be a homosexual. After those questions have been answered, the legal questions almost answer themselves. And since both sides are satisfied that those questions have already been answered, both sides think the constitutional issue is a no-brainer. Which is why it isn’t.
*The word “they” is now a singular as well as a plural neuter pronoun. You heard it here first.