The "Defense of Children Act": Making More Sense of the Freestanding Federalism Argument in Windsor

By Mike Dorf

Thanks to readers for a lively debate on my post on the federalism argument in the DOMA case.  As I noted in the original post and in reply to various comments, I'm not persuaded by the federalism argument standing alone (as opposed to the argument that federalism issues knock out potentially otherwise-legitimate grounds for states barring SSM.  That seems to me a stronger argument, but I have a hard time evaluating it because I think the potentially otherwise-legitimate grounds do not in fact justify laws barring SSM).  However, some commenters suggested that the freestanding federalism argument is crazy or incomprehensible.  I don't go that far so I've been trying to think of a hypothetical example that might make the case for the freestanding federalism argument more forcefully.  I've come up with the following:

Suppose that, following United States v. Lopez, Congress enacted the "Defense of Children Act" (or DOCA).  DOCA defines marriage for purposes of federal law as excluding all marriages in which any children, grandchildren or great-grandchildren of either spouse attend a public or private school that is not protected by a state Gun Free School Zones Act.  Valid?

I think not.  Sure, other things being equal, Congress has the power to define the terms of federal laws, including marriage (or, following Justice Alito's suggestion, any term used as a substitute for marriage).  However, under McCulloch, Congress may not use that power pretextually.  And here that's pretty obviously what Congress is doing.  Under Lopez, Congress may not directly enact the Gun Free School Zones Act.  And under New York v. United States, Congress may not "commandeer" the states by requiring them to enact laws as a means of circumventing the limits on the affirmative powers of Congress.  So . . . it follows that Congress may not, under the guise of defining the terms of a federal statute, pressure the states adopt a law it could not directly require them to adopt.

What exactly do I mean by "pressure"?  If I were trying to come up with a test here, I think I'd simply borrow the test under the Spending Clause articulated in South Dakota v. Dole and applied to invalidate the Medicaid extension rules in the ACA case (even though I think the latter was a misapplication of the doctrine).  The idea would be that Congress may not use its power to define terms of federal statutes in order to "coerce" states into changing their laws.  My hypothetical DOCA appears to do just that.

Does the actual DOMA?  I tend to think not--and we have pretty good evidence of that fact: States have adopted same-sex marriage largely undeterred by the fact that in doing so, they are not penalized by DOMA; their same-sex couples simply don't get the benefits of being treated as married under federal law.  For similar reasons, I think a federal law extending marital benefits to same-sex couples regardless of state law--what I called the Defense of Same-Sex Marriage Act (or DOSSMA) in my last post on the subject--would be permissible.  So too, to use an example offered by a reader in a comment, would a federal law in the early Republic defining "property" under federal law to exclude slaves.

All of the foregoing is simply another way of affirming my view that the freestanding federalism argument against DOMA should fail.  But I don't think it's a slam-dunk, mostly because my understanding of DOMA as a failure to extend rather than a penalty depends on a fairly arbitrary baseline assumption.  At the margin, we can well imagine that DOMA would discourage states from adopting same-sex marriage, just as DOSSMA would encourage them to adopt it.

Moreover, it's not obvious that I'm right that the test for pretext should be equivalent to the test for coercion under the Spending Clause.  Maybe the federal government is simply forbidden from pursuing policies that are tied up in the exercise of the states' reserved powers.  In that case, DOMA, DOSSMA, and even the slaves-aren't-property law would all be invalid as beyond the enumerated powers.  In a time when the Constitution was widely regarded as protecting property in slaves, it would not have been so odd for Congress to be seen as lacking the power to de-property-ize slaves under federal law.  Immoral, sure.  But not necessarily odd.

Or perhaps the test is one of motive: When, as in DOMA, the language and context of federal law defining terms clearly pursues a policy with respect to a reserved power, it's beyond the scope of federal power.  That also wouldn't be crazy.  The very idea of "pretext" suggests some inquiry into motive.

Again, I want to be clear that I'm not persuaded by any of these arguments.  I think DOMA is invalid because I think it denies equal protection but I think that as a matter of federalism, the federal government otherwise has broad power to define the terms of its statutes.  But I continue to think the contrary argument is merely wrong, not crazy.