As the Supreme Court has made clear, under highly deferential rational-basis review, a statutory classification must be upheld “if there is any reasonably conceivable state of facts that could provide a rational basis for the classification.” (FCC v. Beach Communications (1993); see generally the House of Representatives’ Ninth Circuit brief at 30-33.) The Defense of Marriage Act’s definition of marriage in section 3—for purposes of provisions of federal law only—easily passes this review.
The permissible rationale for section 3 that is perhaps simplest to grasp is the federal government’s interest in uniform eligibility for federal benefits. Just as, say, it is a reasonable option for the federal government to apply the same definition of “disability” for purposes of Social Security benefits, so it is a reasonable option to apply a uniform definition of marriage, for purposes of eligibility for federal benefits (and of liability for federal burdens), as “only a legal union between one man and one woman as husband and wife.”
1. The federal government could accept whatever any state defines as a marriage.
2. The federal government could define what it regards as the essential components of marriage and could accept state variants on incidental features.
As to option 1: If “marriage” were truly an empty vessel that the states could define however they wish for purposes of provisions of federal law, that would mean, say, that a state that recognized polyamorous marriages would trigger federal benefits for members of those marriages. And, even if another state chose to limit marriage to a two-person union, it could allow adult siblings to marry. Yet another state could deem any two business partners, or any two neighbors, who are otherwise unmarried to be married. (After all, if the term can mean anything, it can mean virtually nothing.) In all these instances, under option 1, the federal government would have to extend the federal benefits of marriage in accordance with the states’ wildly varying definitions.
If some or all of these examples strike you as absurd, that simply proves the point that it would be absurd to posit some meta-principle (whether resting on confusion over what federalism entails—see point 3.b here—or on any other ground) that would compel the federal government to distribute federal benefits (and burdens) based on whatever the various states choose to call a “marriage.”
Now let’s consider option 2: Option 2 is the approach that section 3 of DOMA takes (and, as it happens, it is the approach that the federal government has always taken—see House brief at 7-9). Section 3 sets forth three essential components of a marriage for purposes of provisions of federal law: (a) the marriage must be a “legal union”; (b) it must be male-female; and (c) it must be between two persons. Beyond that, it treats existing state variations on, say, age of consent or degrees of consanguinity as incidental.
Most opponents of DOMA object only to the second of the three essential components of marriage that section 3 sets forth. They thus implicitly acknowledge that option 2 generally—that is, the federal government’s definition of the essential components of marriage for purposes of the benefits and burdens of provisions of federal law—is a permissible approach. They recognize, in other words, that it is entirely reasonable for the federal government to use its own uniform definition of the essential components of marriage as the basis for distributing federal marital benefits.
The only coherent position for these opponents of DOMA is to maintain that it is nonetheless somehow wholly irrational, for purposes of federal benefit eligibility, to use a uniform definition of marriage that has an essential male-female component. But how could that be?
If any of the federal judges who have invalidated DOMA have any anything cogent to say on this, I’ve missed it. (If you think I have overlooked something, please let me know.) The First Circuit, for example, simply failed to address the federal interest in uniform eligibility for federal benefits. (See near the end of my point 3.c here.) A federal district judge in New York invoked a confused and misplaced notion of federalism.
Federal district judge Vanessa Bryant, in her sloppy ruling on Tuesday, contends that DOMA “in fact infuses complexity and inconsistency into the conferral of federal marital benefits.” (Slip op. at 102.) But apart from the fact that the federal interest in uniformity cannot be reduced to ease of administration, there should be nothing at all complicated about having a male applicant for federal benefits certify that his marriage is to a woman, and vice versa. (That’s a lot less complicated than trying to figure out the problem that the House brief poses (p. 40) and that Bryant simply ignores: the confusion regarding the “status of a same-sex couple that obtains a marriage license in a state where same-sex marriage licenses are available but resides in a state where same-sex marriage is not permitted.”) And the “inconsistency” that Bryant alleges is (as I’ve explained before) simply a pejorative way of recasting the interest in uniformity, not a way of refuting it: the very purpose of a uniform federal definition of the essential elements of marriage is to ensure that marriages recognized under state law that don’t have those elements won’t be treated the same (under federal benefits programs) as those that do.
In sum, it ought to be clear that the only way to reach the conclusion that the male-female component of DOMA’s definition of marriage is irrational is to maintain that it is also irrational for the states to define marriage as male-female. In other words, much as opponents of DOMA try to pretend otherwise, the case against DOMA would also compel broader invention of a federal constitutional right to same-sex marriage that would trump contrary state laws.