Much is being made of the "federalism" challenge to the Defense of Marriage Act. Leading conservative scholars like Michael McConnell and Jonathan Adler, and conservative commentators like George Will, have argued in briefs filed with the Supreme Court or in print that Section 3 of DOMA, which defines "marriage" for purposes of federal law, violates core principles of federalism. These arguments have gained even more traction following oral argument on March 27, as they seemed to have found a sympathetic ear in Justice Anthony Kennedy, even while the Solicitor General of the United States and counsel for the plaintiff Edith Windsor conceded that DOMA was not a violation of federalism principles, and despite the fact that it was not even included in the questions presented to the Court.
The argument runs like this. Domestic relations law is a core function of state governments, one of those powers not delegated to the federal government and thereby, as the Tenth Amendment notes, reserved to the States or to the People. When the federal government adopted a definition of marriage for federal law purposes that was different than might be adopted in the individual states, it impermissibly intruded on that core state function. Hence, DOMA is unconstitutional.
The arguments reflect a serious misunderstanding of federalism, as a simple analogy will demonstrate. Property law, too, is a core function of state governments. But every year, millions of Americans deduct from their income taxes mortgage interest paid on their primary residence. Suppose Massachusetts were to redefine "residence" to include "automobile." Under the faux federalism argument being urged to the Court, any Massachusetts citizen would thereby be able to deduct from their federal income taxes the interest on their automobile loan, and the federal government would have no power to define "residence" for federal purposes as it had always been understood, lest it intrude on that core state power. Such a conclusion is absurd.
DOMA does not force states to accept the federal definition of marriage. That would indeed run afoul of federalism principles. But it does prevent any single state from forcing its definition on the federal government (Section 3), or on any other state (Section 2). That allows each state to chart its own path, as federalism requires, but it does not allow any state to chart a path for the rest of the country. That is federalism as it should be.
The exponents of the faux federalism argument seek to bolster their position by the factual claim that the federal government has never before sought to define marriage but, as George Will puts it, always "took state law as it found it." That factual claim is both false and irrelevant. It is false because, as Paul Clement rightly noted during oral argument, the federal government has on occasion adopted for its own purposes a definition of marriage different than that applied in the states. When Utah was a territory, for example, the federal government refused to recognize (even criminalized) polygamous marriages. That was permissible because the federal government actually exercises something akin to state government power when it regulates for territories that are not yet states, but the federal government continued to adhere to its anti-polygamy definition of marriage even after Utah became a state, making it an irrevocable condition of Utah statehood. A couple of additional examples bolster the point. The federal government refuses to recognize December divorces and January remarriages designed to skirt federal tax laws, even if the state would recognize those actions. And it refuses to recognize marriages of convenience between a citizen and a non-citizen, designed to end run federal immigration policy, even where a particular state would recognize the marriage.
Moreover, even if such examples did not exist, the fact that the federal government had previously been willing to take "state law as it found it" says nothing about whether it is required to continue to do so if some states begin to experiment with radical redefinitions of marriage. The Windsor case arose because the federal court in New York anticipated that New York would recognize a lesbian marriage lawfully performed in Canada. But suppose New York also decided to recognize a polygamous marriage lawfully performed in Saudi Arabia, after the multiple spouses took up residence in New York. Would the federal government be obligated to pay out social security survivor benefits to multiple spouses upon the death of any one of them, because "federalism" required it to take New York's law "as it found it"? And would it then be able to cut off those benefits if those surviving spouses all moved to North Carolina, which did not recognize the polygamous marriage? Of course not. DOMA does nothing more than retain the understanding of "marriage" that existed when that word was used in over 1100 federal statutes adopted prior to the same-sex assault on the traditional understanding. Some of those statutes might exceed Congress's enumerated powers and thereby violated federalism principles in their own right, if we were to return to the original understanding of the Commerce and Spending powers of Congress, but that would be true federalism, not the faux federalism being advanced in an admitted attempt to give the Court a plausible way to avoid having to decide whether DOMA violates the constitutional guarantee of equal protection.