when applied to exclude same-sex married couples from federal employee benefits: So concluded Ninth Circuit Judge Stephen Reinhardt earlier this week, in an internal order resolving a complaint brought by a Ninth Circuit employee. Chief Judge Alex Kozinski suggested in a similar order last month that DOMA might be unconstitutional in such situations, but interpreted the federal benefits statute in a way that avoided the need to decide whether DOMA is indeed unconstitutional. [UPDATE: These orders are not binding precedent, but they are a hint of how these judges -- and presumably some of their colleagues -- may rule in future cases arising outside this context of internal Ninth Circuit employment complaints.] (Thanks to Robert Iafolla and John Roemer of the L.A. Daily Journal for breaking the story.)

Judge Reinhardt concluded that applying the Defense of Marriage Act's mandate that "marriage" under federal law shall include only opposite-sex couples violates the Due Process Clause because it is not "rationally related to a legitimate governmental purpose." The opinion concluded that:

1. The denial of benefits "cannot be justified simply by a distaste for or disapproval of same-sex marriage or a desire to deprive same-sex spouses of benefits available to other spouses in order to discourage them from exercising a legal right affored them by a state," since that would constitute "a bare desire to harm a politically unpopular group."

2. The federal government interest in "defending and nurturing the institution of traditional, heterosexual marriage," mentioned in the House report on DOMA, "is largely irrelevant to the rational basis analysis here because the same-sex couple here is already married," and the government's denial of benefits to same-sex spouses wouldn't "encourage[ gay people] to enter into marriages with members of the opposite sex."

3. As to the interest in "defending traditional notions of morality" also mentioned in that House report, "if the denial is designed to 'defend' traditional notions of morality by discouraging same-sex marriage, it does so only by punishing same-sex couples who exercise their rights under state law, and thus exhibits the 'bare desire to harm' same-sex couples that is prohibited under City of Cleburne and Romer." Moreover, "denying married same-sex spouses health coverage is far too attenuated a means of achieving the objective of 'defending traditional notions of morality," plus "Romer and Lawrence v. Texas strongly suggest that the government cannot justify discrimination against gay people or same-sex couples based on 'traditional notions of morality' alone."

4. Finally, the third interest mentioned in the House report, "the government's interest in preserving scarce government interests" "does not provide a rational basis for that policy if the policy is, as a cost-saving measure, drastically underinclusive, let alone founded upon a prohibited or arbitrary ground."

Some of these rationales might be restrictable to employee benefits alone. But the logic of this argument, I think, would ultimately invalidate the entire Defense of Marriage Act, and for that matter would read the U.S. Constitution to secure a right to same-sex marriage (at least so long as the government recognizes opposite-sex marriages).

I like the idea of giving same-sex spouses the same health benefits as opposite-sex spouses; and I support same-sex marriage as a policy matter. But it seems to me that these claims that the law lacks a "rational basis" are not sound. The principle behind the Defense of Marriage Act — and nearly all states' recognition of opposite-sex marriage but not same-sex marriages — is that (1) heterosexual relationships are better for society than homosexual ones, and that the (2) government's preferring same-sex marriages in a wide range of contexts will tend to reinforce that norm among the public. I'm skeptical of both these claims, but I don't think they can be said to be irrational, or animated by a "bare desire to harm": Many critics of same-sex marriages genuinely believe (even if in my view ultimately wrongly) that specially fostering heterosexual relationships is indeed materially better for society than treating heterosexual and homosexual relationships equally — their desire is to try to help society (even if this in the process means excluding some groups for benefits), not "bare[ly]" "to harm" people.

And this is especially so since sexual behavior is indeed alterable for quite a few people. Whatever one might think about the alterability of sexual orientation, there is strong evidence that many (perhaps about half) non-purely-heterosexual men and most non-purely-heterosexual women are actually bisexual, and may thus be not unlikely to respond to social norms in their choices of partners. As I've suggested before, I don't think we should try to pressure this choice using government benefits; but that doesn't make the contrary view irrational, it seems to me. Moreover, while I am told that there is a considerable amount of intriguing evidence about sexual orientation being innate, I'm pretty skeptical that the matter is conclusively resolved at this stage of the investigation. I thus don't think it's irrational for Congress to believe that, at least as to some people, broad social norms may affect their felt orientations and not just their choices within a felt bisexual orientation.

Of course, any particular application of the law (e.g., to employee benefits) might not have that much of a norm-reinforcing effect. But that sort of particularization isn't the proper approach under the rational basis test, I think, nor should it be: The point of this law, as of many other laws, is to have a significant effect through its aggregate incentive and norm-setting function.

Finally, I should acknowledge that the famously mysterious decision in Romer v. Evans, and the less famously mysterious but still, in my view, not well-reasoned decision in City of Cleburne v. Cleburne Living Center might be read as supporting the conclusion that DOMA is unconstitutional here. This is so precisely because they are so vague, and because the "rational basis with bite" review that they are said to create is so unclear on what that "bite" is supposed to mean, and what actual legal rule judges are supposed to use to implement this "bite." But I'm pretty sure that those decisions don't mandate the results (again, precisely because they are so vague).

In any case, let me stress again: I think wise employers, federal, state, and private, should care about keeping their employees happy and productive — including by helping the employees' provide for their families — and not about the employees' choice of spouses or life partners. I also support the recognition of same-sex marriages. But while I think the views contrary to mine are mistaken, I don't think they're irrational, either in the lay sense or in the constitutional sense.