For gay-marriage litigants, July has been the cruelest month. Prior to today's 5-4 Washington Supreme Court decision in Andersen v. King County, there were two substantive state marriage decisions against them (New York and Connecticut), one quasi-substantive federal decision against them (the 8th Circuit, whose broad dicta went beyond the state constitutional ban at issue), and three procedural decisions against them upholding the propriety of ballot initiatives (Massachusetts, Tennessee, and Georgia).

But this may turn out to be the hardest day of all. Andersen is the most careful, closely reasoned, and comprehensive judicial opinion to date rejecting constitutional claims to gay marriage. It is much better, as a matter of conventional legal analysis and craftsmanship, than the New York Court of Appeals decision in Hernandez v. Robles rejecting gay-marriage claims a couple of weeks ago. Since the principles and arguments on this issue from state-to-state, and even in the federal courts, are not that different, the Washington decision will deserve close attention from other courts. Among the courts next to consider claims for gay marriage, the New Jersey Supreme Court in particular should grapple with Andersen.

There is a bright spot in Andersen for gay couples. The court practically invites future litigation and legislation resulting in a Vermont-style civil unions resolution, granting the benefits and protections of marriage to gay couples without the status of "marriage." This seems the likely direction for future litigation and legislative action in Washington and elsewhere. I'll say more about this below.

Since I addressed many of the constitutional issues in earlier posts reacting to Hernandez v. Robles I won't go over all that territory again. Instead, I'll address here some of the distinctive features of the Andersen opinion.

(1) While there was some speculation about possible political motivations for the 17-month period to issue the opinion, it seems in retrospect that the justices were simply taking the time to be very careful on an opinion they knew history would judge them for: writing and re-writing sensitive passages, negotiating with each other, trading drafts among the chambers, waiting to see how other courts might come out. The close result, and the opening paragraphs in particular, chiding not just the dissents for the usual abandonment of legal principles but also the concurrence for needlessly extending the opinions, suggests a court that was sharply divided over how to proceed and struggling to reach broader consensus. Yes, judges can be political and Machiavellian; but sometimes they're just doing their job as conscientiously as they can. In general, I'd be wary of the conspiracy theories about courts that spring up from the populist resentment of judicial activism.

(2) Unlike the New York Court of Appeals, the Washington court dealt substantively with the question of suspect-class status for gays under equal protection principles. It also avoided the bizarre contextualized analysis of the issue the New York court gave us, and instead dealt with the question whether gays generally constitute a suspect class.

The Andersen court is more confident about the elements of the suspect-class analysis (history of discrimination, immutable and usually irrelevant trait, and political powerlessness) than I think is justified, but its rejection of heightened scrutiny is certainly defensible under the precedents. It acknowledges, as every court addressing the issue has, a history of discrimination against gays. But it says the trait defining the class is not immutable, and seems to hold that immutability is essential to heightened scrutiny. Both halves of this holding are questionable. The Andersen court isn't clear what it thinks the "defining trait" of this class is -- homosexual acts or homosexual orientation -- and this ambiguity muddies things a bit. If homosexuals acts are the "defining trait," it's not immutable; if homosexual orientation, the best evidence is that it is immutable or very close to immutable. It's also not clear that immutability is even required, as opposed to one factor among many to be weighed, in the analysis of which groups get heightened scrutiny.

The Andersen opinion also shows how legislative advances for gays are a double-edged sword in litigation over marriage. Some courts upholding gay-marriage or civil unions claims have cited legislative progress -- eliminating sodomy laws, making adoptions more widely available, passing employment non-discrimination laws -- as evidence that times and attitudes are changing and as support for the idea that the legislature has no very good reason to withhold this last bit of progress from them.

In Andersen, by contrast, the court cites legislative progress as a reason to deny gays suspect-class status, on the dubious theory that a truly powerless group wouldn't have made such democratic progress. This part of the analysis is questionable, since historical work by William Eskridge and others has shown that groups tend to get heightened scrutiny only after they've made legislative progress. And once they've gotten suspect class status, they don't lose it simply because the legislature begins to take their concerns seriously.

I have always thought there was room for a good suspect-classification argument, but there's no doubt the weight of precedent is on the side of the Andersen court.

(3) On whether the plaintiffs have a fundamental right to marry, the Andersen opinion is also better than Hernandez. It avoids directly resolving the level-of-generality problem, i.e., is the issue a "fundamental right to marry" or a "fundamental right to marry a person of the same sex"? Instead, it resolves the issue in good lawyerly fashion by looking closely at the Supreme Court's cases dealing with the fundamental right to marry, including cases the New York Court of Appeals ignored, like Turner v. Safley (involving marriage for prison inmates). Andersen reads these cases as linking marriage to procreation. That's not the only way to read the precedents and it doesn't make much sense of the modern conception of marriage, but as an interpretation of doctrine it's defensible and perhaps the best view.

(4) Once it rejected any reason to apply heightened scrutiny to the exclusion of gay couples from marriage, the Andersen court applies a standard rational-basis review. The court rightly describes this form of review as "extremely deferential" and granting the state "nearly limitless" power to make policy as it sees fit. Here, the court does a much better job than did Hernandez explaining why it's constitutionally rational for the state to create a special status for heterosexual couples. The underinclusiveness and overinclusiveness of the classification in relation to the claimed interests -- promoting procreation and child welfare -- are very poor policy reasons to exclude gay couples from marriage, and the court implicitly recognizes this at several points in the opinion. But they're sufficient for traditional rational basis review, absent some good indication of animus. The court explains -- unlike the New York court — that the issue is not whether excluding gay couples from marriage advances these interests in any way (the exclusion of gay couples does not plausibly advance them) but whether including straight couples in marriage advances these interests (it clearly does).

(5) Andersen quickly, and I think correctly, dismisses the claim that due process "privacy" interests command the inclusion of gay couples in marriage. Marriage is not a purely private matter, though important aspects of the marital relationship are private and it is an intensely personal experience for couples and families. Marriage is full of public benefits and privileges that make it a public investment and a matter of public concern.

(6) The Washington opinion also rejects the sex-discrimination argument much more convincingly than did the New York court. Andersen does as good a job as I've seen in a judicial opinion of refuting "the Loving analogy," which suggests that excluding same-sex couples from marriage is constitutionally analogous to excluding interracial couples from marriage. Moreover, the opinion deals not just with the formal aspects of the sex-discrimination argument (marriage laws classify on the basis of sex) but also with the sociological claim (that marriage laws have the impermissible purpose or effect of perpetuating gender stereotypes). The sociological claim about gender stereotypes seems a crabbed view of marriage, its public purposes and effects.

(7) Finally, there's a potentially significant passage in the middle of the opinion that might get overlooked on a first read:

We do not dispute that same-sex couples raise children or that the demographics of "family" have changed significantly over the past decades. We recognize that same-sex couples enter significant, committed relationships that include children, whether adopted, conceived through assisted reproduction, or brought within the family of the same-sex couple after the end of a heterosexual relationship. We do not doubt that times have changed and are changing, and that courts and legislatures are increasingly faced with the need to answer significant legal questions regarding the families and property of same-sex couples. (Citations omitted).

We are also acutely aware, from the records in these cases and the briefing by the plaintiffs and the amici supporting them, that many day-to- day decisions that are routine for married couples are more complex, more agonizing, and more costly for same-sex couples. A married person may be entitled to health care and other benefits through a spouse. A married person's property may pass to the other upon death through intestacy laws or under community property laws or agreements. Married couples may execute community property agreements and durable powers of attorney for medical emergencies without fear they will not be honored on the basis the couple is of the same sex and unmarried. Unlike heterosexual couples who automatically have the advantages of such laws upon marriage, whether they have children or not, same-sex couples do not have the same rights with regard to their life partners that facilitate practical day-to-day living, involving such things as medical conditions and emergencies (which may become of more concern with aging), basic property transactions, and devolution of property upon death.

Lest you think these are just crocodile tears from a gutless court delivering gay families to the tender mercies of the heartless legislature, the court continued:

But plaintiffs have affirmatively asked that we not consider any claim regarding statutory benefits and obligations separate from the status of marriage. We thus have no cause for considering whether denial of statutory rights and obligations to same-sex couples, apart from the status of marriage, violates the state or federal constitution. (emphasis added)

This qualification is repeated at the end of the opinion, along with another recitation of the ways the denial of marriage harms gay families and with a strong suggestion that "the legislature may want to reexamine the impact of the marriage laws on all citizens of this state."

To the state legislature, the message seems to be this: "Get moving on addressing the hardships faced by gay couples and their children, some of which we've listed for you. You don't have to give them marriage and maybe not even all of the rights of marriage, but something needs to be done. If you don't act, we might."

To gay-marriage litigants, the message seems to be this: "Go to the legislature and see what can be done about the sorts of problems you've identified and that we agree exist. If the legislature is unresponsive, come back to us not with a claim for the status of marriage, but with a remedial claim for the benefits and protections of marriage for your families."

My guess is that this dual message was necessary to get the five votes needed to uphold the state's marriage laws.

I've said before that courts confronting gay-marriage claims may now see three choices: (1) ordering full marriage (Massachusetts); (2) denying the claims (New York); and (3) compromising on civil unions, with instructions to the legislature to decide on implementation (Vermont). Choice #3 involves many complications and permutations.

Though superficially opting for #2, the Washington court would like very much to give #3 a try. It's a sensible direction for litigants, legislatures, and courts.