The New York marriage decision raises many interesting questions. I plan, over a few posts, to deal with some of them: (1) the due process claim; (2) the equal protection claim; (3) the legal rationality of excluding same-sex couples from marriage; and (4) some implications of the decision for the future of gay marriage. First up, due process.

The plaintiffs argued that there is a fundamental "right to marry" guaranteed by the state constitution's due process clause (which New York courts interpret more broadly than the analogous federal clause), and that this right applies to same-sex as well as opposite-sex couples. Thus, their exclusion should be subjected to strict scrutiny, which hardly anyone seems to think a state could satisfy. The Court of Appeals held that while there is a fundamental right to marry, there is no fundamental right "to marry someone of the same sex," (p. 11) and thus that only rational basis scrutiny applies. It differed from the Massachusetts court in 2003 in Goodridge by holding that the exclusion of gay couples from marriage is rational.

The doctrinal key in these fundamental rights/substantive due process cases has long been the proper way to characterize the right at stake. Much academic commentary has been offered on this issue. Where courts characterize the right at stake broadly (e.g., "this case is about 'the right to privacy'") plaintiffs win (see e.g., Griswold v. Connecticut, Eisenstadt v. Baird, Roe v. Wade, Casey v. Planned Parenthood, and Lawrence v. Texas) because their activity can always be said to partake of the broad right. Where courts characterize the right at stake narrowly (e.g., "this case is about the 'right of nonmarital biological fathers to visitation with their children adulterously conceived'" or "this case is about the 'right to engage in homosexual sodomy'") plaintiffs lose (see e.g., Bowers v. Hardwick, Michael H. v. Gerald D., and Washington v. Glucksberg), because there is rarely the kind of historical and traditional support for a narrow right that would allow it to be ranked as fundamental. In every due-process case, it is possible to articulate the claimed right very broadly, very narrowly, or somewhere in between.

So the issue is largely decided, doctrinally, at this initial stage of characterizing the right. The rest of the analysis is usually straightforward, though there have been exceptions (e.g., where a law meets strict scrutiny or fails rational basis review). To its credit, the New York court recognized the tensions in the precedents on this issue, with courts sometimes choosing broad characterizations and sometimes narrow ones. The U.S. Supreme Court has never given us a methodology for choosing the level of generality, which has led to the strong suspicion that the cases in this area are result-driven.

Which path to follow in a gay-marriage case, that of Glucksberg and its narrow conception of the claimed right, or that of Lawrence and its broad conception of the claimed right? The question is not a hypothetical one; it will confront every court at the state and federal level that deals with a gay-marriage claim grounded in due process. The New York court offers this methodology for making the choice: "The difference between Lawrence and Glucksberg," reasoned the court, "is that in Glucksberg the relatively narrow definition of the right at issue was based on rational line-drawing. In Lawrence, by contrast, the court found the distinction between homosexual sodomy and intimate relations generally to be essentially arbitrary." (p. 12) Thus, the New York court chooses the Glucksberg model and rejects the Lawrence model. (Yet another indication, if we needed more, that Lawrence is not leading courts quickly to gay marriage.)

The New York court's approach is a novel, but I think unsatisfactory, way to resolve the level-of-generality dilemma in due process cases. It resolves the problem only by deciding the result. Thus, in Glucksberg, under the New York court's approach we would choose the more specific level of generality ("this case is about a narrow 'right to physician-assisted suicide' rather than a broad 'right to die'") only after deciding that the law meets the very test likely to be applicable to a right characterized at that specific level: it is rational for the legislature to ban physician-assisted suicide so we characterize the right as involving only the very narrow practice that it is rational for the legislature to ban, not some broader principle of human liberty, like a "right to die with dignity," which the legislature could not deny.

In a gay-marriage case, we have to choose between characterizing the exclusion of gay couples from marriage as implicating either (1) a "right to same-sex marriage" (narrow), or (2) a "right to marry" (broad). We make this choice, according to the New York court, by first deciding whether the exclusion itself is rational. If it is rational (as it is in the case of gay marriage, the court concludes), then we'll choose the narrow characterization of the right implicated (#1), which will mean that rational-basis scrutiny applies (since there is rarely sufficient historical support for rights narrowly conceived, and thus they aren't "fundamental"), and we've just decided that the exclusion is rational. It can't be argued with a straight face that there is a traditionally recognized "right to same-sex marriage." It follows like the night the day that gay-marriage plaintiffs lose under the New York-style due-process analysis

On the other hand, if the New York court had found — as the Massachusetts court did in Goodridge in 2003 -- that the exclusion of gay couples from marriage was arbitrary/irrational, then it presumably would have characterized the claimed right at a broader level of generality ("this case is about the 'right to marry'"), since the exclusion would not be based on "rational line-drawing"; then it would have found that the case implicated a fundamental right to marry; and hence it would have applied strict scrutiny, which the exclusion would fail, since the court already would have found it irrational.

The New York court thus gives us an approach to the level-of-generality problem in substantive due process cases that, if taken seriously, is both circular and eviscerates fundamental rights. The result in every case is practically determined by the initial analysis about the law's rationality. If the law is determined rational at the outset, it will probably be constitutional. If not, it probably won't be. But this makes a (further) mess of the fundamental rights doctrine, the underlying point of which is to say that some rights are so important they cannot be denied even if the government has rational reasons to do so.

I do not want to be too hard on the New York court on this point. The judges deserve credit for recognizing the problem and honestly attempting to grapple with it; not all courts do. It's equally possible for a court to assume that gay marriage implicates the "right to marry" without ever recognizing this issue, and thus uphold a gay-marriage claim in an equally conclusory fashion.

A less abstract way to approach the problem of choosing a level of generality would be to ask how courts have tended to characterize the right at issue in other marriage cases. Here, the most obvious precedents to look at would be the canonical marriage cases -- Loving v. Virginia, Turner v. Safley, and Zablocki v. Redhail. In all three cases, the U.S. Supreme Court chose a broad level of generality, not a narrow one. Thus, in each case, the right at stake was the "right to marry," not the "right to marry a person of another race" (Loving), the "right of prison inmates to marry" (Turner), or the "right of deadbeat parents to marry" (Zablocki). This would suggest that the right at stake in the New York case was the "right to marry," not the "right to same-sex marriage." Surprisingly, the court doesn't even mention Turner and cites Zablocki only for an unrelated point.

But this, too, would be an unsatisfactory approach. What, after all, is the content of the "right to marry"? It surely cannot mean a fundamental right to marry anyone or anything, regardless of number of partners, age, consanguinity, or even species -- a right that could only be denied if the government could satisfy strict scrutiny. "Marriage" in the fundamental rights context must have some substantive content, some underlying purpose, which would help us to distinguish the claim of an adult, unrelated gay couple from the claim of Joe to marry Fido.

I have no problem distinguishing same-sex marriages from inter-species marriages, because I have an underlying idea of marriage and the purposes it properly serves that includes the claims of the former and excludes the claims of the latter. But my view of the underlying point of marriage is sharply contested, since for many people in almost all times marriage has meant something -- uniting people whose sex procreates — that excludes gay couples.

There are good arguments against their view -- among them, that it has been for at least a century a largely anachronistic vision of marriage -- but I'm hardly so confident that I'm correct that I want courts deciding the issue at this point as if it's a settled matter of "fundamental rights." And if I were to brief this point for the anti-gay-marriage movement in a good lawyerly way, I'd have plenty of room to argue that the cases in this area seem implicitly and sometimes explicitly to link marriage to procreation, conceiving "marriage" and thus the fundamental "right to marry" as necessarily involving a man and a woman.

One could argue for a wholly different approach to protecting rights under the due process clause, one that is not tethered to long-recognized practices and traditions, or one that wholly eschews a distinction between fundamental rights and other liberties. Under such an approach, a gay-marriage claim might be upheld even though gay marriage has no grounding in our history and tradition. But, despite the powerful arguments of my co-blogger Randy Barnett and a few others, I do not believe that the Supreme Court has moved to an alternative approach. Nor have I yet seen an alternative approach that would be better than the maddening one we now have.

So it seems to me that, at this point in our history and in the development of doctrine, the New York court reached the right result on the due proces claim for rather unconvincing reasons.